foll soc entiated cardiac rehabilita tion...soc - e Lon ially valuated by g-term differ clinical ou foll entia tcomes, se PhD Ka Aa low-up ted ca econdary pre dissert thrine H Health

soc

- e

Lon

cially

evaluated by

g-term

differ

y clinical ou

m foll

rentia

utcomes, se

PhD

Ka

Aa

low-up

ated ca

econdary pre

D dissert

athrine H

Health

arhus Univ

2018

p on t

ardiac

evention an

tation

Hald

h

versity

the eff

c reha

d use of hea

fect of

abilita

alth care ser

f

ation

rvices

Long-term follow-up on the effect of

socially differentiated cardiac rehabilitation - evaluated by clinical outcomes, secondary prevention and use of health care services

PhD dissertation

Kathrine Hald

Section for Clinical Social Medicine and Rehabilitation

Department of Public Health

Health

Aarhus University

2018

Supervisors

Claus Vinther Nielsen, Professor, MD, PhD

Section for Clinical Social Medicine and Rehabilitation, Department of Public Health, Aarhus

University, Aarhus, Denmark

Mogens Lytken Larsen, Professor, MD, DMSc

Danish Centre for Inequality in Health, Department of Cardiology, Aalborg University Hospital,

Aalborg, Denmark.

Bo Christensen, Professor, MD, PhD

Section for General Medical Practice, Department of Public Health, Aarhus University, Aarhus,

Denmark.

Kirsten Melgaard Nielsen, MD, PhD

Department of Cardiology, Aarhus University Hospital, Aarhus, Denmark.

Finn Breinholt Larsen, Senior scientist, Cand.mag.

DEFACTUM, Social and Health Services and Labour Market, Central Denmark Region, Aarhus,

Denmark.

Assessment committee

Kaj Sparle Christensen, Professor, MD, PhD (Chairman)

Section for General Medical Practice, Department of Public Health, Aarhus University, Aarhus,

Denmark.

Joep Perk, Senior professor, MD, EFESC

Department of Health and Caring Sciences, Faculty of Health and Life Sciences, Linneaus

University, Kalmar, Sweden

Janus Laust Thomsen, Professor, MD, PhD

Department of Clinical Medicine, Aalborg University Hospital & Center for General Practice,

Aalborg University, Aalborg, Denmark.

Financial support

TrygFonden, the Health Foundation, the Committee of Multipractice Studies in General Practice,

the Foundation of Director Kurt Bønnelycke and Mrs. Bønnelycke, the Foundation of Karen

Elise Jensen, the Health Research Fund of Central Denmark Region, the Practice Research Fund

of Central Denmark Region & Aarhus University.

Acknowledgements

The work presented in this PhD dissertation was conducted at Section for Clinical Social

Medicine and Rehabilitation, Department of Public Health, Aarhus University & DEFACTUM,

Social and Health Services and Labour Market, Central Denmark Region in Aarhus, Denmark.

The last sentence in this PhD dissertation marks the end of a personal journey that I began in

2007 when my father died of a cardiovascular disease. In order to make sense of my

overshadowing sorrow, I decided that I would try to help patients with cardiac disease and their

families. In this way, cardiovascular disease and cardiac rehabilitation have been my main fields

of interest all the way through my education.

A lot has happened in my life in the last 12 years and I know that I would not be where I am

today if I was not surrounded by so many wonderful people in my professional and private life. I

would like to express my deepest gratitude to my supervisors Claus Vinther Nielsen, Mogens

Lytken Larsen, Bo Christensen, Kirsten Melgaard Nielsen and Finn Breinholt Larsen for their

excellent support, guidance, help and constructive feedback. I am truly honoured that such

brilliant and gifted people have been my supervisors! In particular, I would like to thank Claus

for giving me the best surroundings for my daily work, for trusting me and for giving me

confidence. Mogens, for believing in me and for his overwhelming support from the very first

meeting, for his ability to see opportunities and for all of the enriching conversations. Bo, for

giving me a chance when I came knocking on his door as a first-year MSc student with big

dreams about doing a PhD, for his huge support in the time hereafter and for his concern for me.

Kirsten, for making it possible for me to do this PhD and for her exceptional contributions when

I was writing my scientific papers. Finn, for all his help in the construction of the PhD and his

wise considerations regarding data and results in a larger perspective. I also wish to express my

sincere gratitude to Lucette Kirsten Meillier for her encouragement and priceless help all the way

through the process and for her always positive and sweet attitude towards me.

I wish to thank all of my colleagues on Marselisborgcentret in Aarhus for creating a workplace

where I really have enjoyed working for the last four years. I will always remember and

appreciate all the fun we have had together. Even though I now leave Marselisborgcentret, I will

forever have my girls (you know who you are) in my heart.

My thanks go to Thomas Maribo for interesting discussions and coaching during my teaching at

Aarhus University. Also, I wish to thank Inger Hornbech for all of her practical help specifically

regarding my PhD, and generally when computers and printers conspired against me. Thank you

to Christina Malmose Stapelfeldt for her highly skilled help with the preliminary statistical

analyses.

A special thank you goes to Cardiovascular Research Unit and Danish Centre for Inequality in

Health at Forskningens Hus, Aalborg University Hospital in Aalborg, Denmark for letting me

spend three months in their inspiring research environment. I owe a huge gratitude to Martin

Berg Johansen for his enormous help regarding statistical considerations and analyses. I wish to

thank him for his patience with me and for his endless positive approach.

I am very grateful to the patients who participated in the original project on which this PhD

dissertation is based. I would like to thank the foundations that have provided financial support.

Also, I would like to thank the general practitioners who participated in my questionnaire survey.

Without all the patients, the financial support and the general practitioners, this PhD would not

have been possible.

Last but not least, I owe my deepest gratitude to my beloved family and my dear friends for all

of their support and interest in my work. I love you.

I dedicate this PhD dissertation to my father Kurt Hald.

1 March 2019 Kathrine Hald

List of scientific papers

The PhD dissertation is based on the following papers:

I. Hald K, Nielsen KM, Nielsen CV, Meillier LK, Larsen FB, Christensen B, Larsen ML.

Expanded cardiac rehabilitation in socially vulnerable patients with myocardial

infarction: a 10-year follow-up study focusing on mortality and non-fatal events. BMJ

Open. 2018;8:1-8.

II. Hald K, Larsen FB, Nielsen KM, Meillier LK, Johansen MB, Larsen ML, Christensen B,

Nielsen CV. Medication adherence, biological and lifestyle risk factors in patients with

myocardial infarction: A ten-year follow-up on socially differentiated cardiac

rehabilitation.

Accepted for publication in Scandinavian Journal of Primary Health Care.

III. Hald K, Meillier LK, Nielsen KM, Larsen FB, Johansen MB, Larsen ML, Nielsen CV,

Christensen B. Does socially differentiated cardiac rehabilitation affect the use of health

care services after myocardial infarction? A ten-year follow-up study.

Submitted to Scandinavian Journal of Public Health.

Abbreviations

ACE inhibitors Angiotensin-converting Enzyme Inhibitors

ACME Automated Classification of Medical Entities

ACS Acute Coronary Syndrome

ATC Anatomical Therapeutic Chemical Classification

BP Blood Pressure

CABG Coronary Artery Bypass Surgery

CAD Coronary Artery Disease

CARIMAP Coronary Artery Risk Management Program

CDR The Danish Cause of Death Register

CHD Coronary Heart Disease

CI Confidence Interval

CPR Civil Personal Register

CVD Cardiovascular Disease

CR Cardiac Rehabilitation

CRS The Danish Civil Registration System

DNPR The Danish National Prescription Register

DUN Danish Education Nomenclature

EUROASPIRE European Action on Secondary and Primary Prevention by Intervention to

Reduce Events

GP General Practitioner

HbA1c Haemoglobin A1c

HDL High Density Lipoprotein

ICD-10 The International Classification of Diseases 10th Edition

IHD Ischemic Heart Disease

LABKA The Central Denmark Region Clinical Laboratory Information System

LDL Low Density Lipoprotein

MACE Major Cardiac Events

MI Myocardial Infarction

NHSR The Danish National Health Service Register

NICE British National Institute for health and Care Excellence

NOR-COR Norwegian Coronary Prevention Study

NPR The Danish National Patient Register

OR Odds Ratio

PCC Person Centered Care

PCI Percutaneous Coronary Intervention

RCT Randomized Controlled Trial

SCORE Systematic Coronary Risk Estimation

SD Standard Deviation

SES Socioeconomic Status

SIGN Scottish Intercollegiate Guidelines Network

WHO World Health Organization

Contents

1. Introduction and aim ........................................................................ 1

2. Background ........................................................................................ 4

2.1 Cardiovascular disease ..................................................................................... 5

2.1.1 Epidemiology ................................................................................................................... 5

2.1.2 Risk factors ...................................................................................................................... 5

2.1.3 Symptoms, diagnosis and treatment of myocardial infarction ........................................ 6

2.2 Cardiac rehabilitation ....................................................................................... 6

2.2.1 Definition ......................................................................................................................... 6

2.2.2 Effect ............................................................................................................................... 7

2.2.3 Core components ............................................................................................................. 7

2.2.4 Delivery and implementation .......................................................................................... 8

2.2.5 Legislation ....................................................................................................................... 9

2.2.6 Organisation ..................................................................................................................... 9

2.2.7 The role of general practice ........................................................................................... 10

2.3 Social inequality in health .............................................................................. 11

2.3.1 Definition ....................................................................................................................... 11

2.3.2 Policies ........................................................................................................................... 11

2.3.3 Social inequality in cardiovascular disease and secondary prevention ......................... 12

2.4 The original socially differentiated cardiac rehabilitation project ................. 14

3. Methods ............................................................................................ 16

3.1 Study setting ................................................................................................... 17

3.2 Study population ............................................................................................. 17

3.3 Study design ................................................................................................... 20

3.4 Intervention ..................................................................................................... 20

3.5 Data sources .................................................................................................... 21

3.5.1 Baseline data .................................................................................................................. 21

3.5.2 Follow-up data ............................................................................................................... 22

3.5.2.1 Registers .................................................................................................................. 22

3.5.2.2 Questionnaire survey in general practice ................................................................ 23

3.6 Outcomes ........................................................................................................ 24

3.6.1 Paper I ............................................................................................................................ 24

3.6.2 Paper II .......................................................................................................................... 24

3.6.3 Paper III ......................................................................................................................... 25

3.6.4 Additional analyses ........................................................................................................ 26

3.7 Statistical analysis .......................................................................................... 26

3.7.1 Baseline data .................................................................................................................. 26

3.7.2 Paper I ............................................................................................................................ 26

3.7.3 Paper II .......................................................................................................................... 27

3.7.4 Paper III ......................................................................................................................... 27

3.7.5 Additional analyses ........................................................................................................ 27

3.8 Approval and ethics ........................................................................................ 28

4. Results in summary ......................................................................... 29

4.1 Baseline characteristics .................................................................................. 30

4.2 Paper I ............................................................................................................. 32

4.3 Paper II ........................................................................................................... 33

4.4 Paper III .......................................................................................................... 36

4.5 Additional analyses ........................................................................................ 37

5. Discussion of main findings in relation to other studies ............... 43

5.1. Mortality and non-fatal events ...................................................................... 44

5.2 Medication adherence and biological and lifestyle risk factors ..................... 47

5.3 Use of health care services ............................................................................. 50

6. Discussion of methods ..................................................................... 52

6.1 Internal validity .............................................................................................. 53

6.1.1 Selection bias ................................................................................................................. 53

6.1.2 Information bias ............................................................................................................. 54

6.1.3 Confounding .................................................................................................................. 54

6.1.4 Power ............................................................................................................................. 55

6.1.5 Study population ............................................................................................................ 56

6.1.6 Study design .................................................................................................................. 58

6.1.7 Intervention .................................................................................................................... 59

6.2 External validity ............................................................................................. 62

7. Conclusions ...................................................................................... 63

8. Perspectives ...................................................................................... 65

9. English summary ............................................................................. 68

10. Dansk resumé ................................................................................. 71

11. References ...................................................................................... 74

12. Appendices ..................................................................................... 86

Page | 1

1. Introduction and aim


Page | 2

"There are many ways of going forward, but only one way of standing still."

Franklin D. Roosevelt

This year it is 70 years ago that the first study participant was enrolled in the Framingham Heart

Study. The study was initiated because of the American President Franklin D. Roosevelt's poor

heart condition and the fact that half of all deaths in the 1940s in America were caused by

cardiovascular disease (CVD). At that time, evidence concerning primary and secondary

prevention of CVD was lacking and premature death due to CVD was merely accepted as a fact

of life. The Framingham Heart Study, among others, has provided essential knowledge of the

epidemiology and risk factors of CVD.1

Great achievements in CVD prevention have been reached during the previous half century.2

However, European mortality and morbidity rates as well as years lived with disability due to

CVD are still worryingly high.2,3 There is well-established evidence of secondary CVD

prevention including cardiac rehabilitation (CR) based on international guidelines.4-8 Low

socioeconomic status (SES) is associated with risk of CVD, lower adherence to guidelines and a

poorer long-term prognosis.9-20 International and national policies have placed social inequality

in health high on the public health agenda, and the will and courage to continue the fight to

reduce social inequality in health seem to be present.21-24 In the early noughties, a project with

focus on socially differentiated CR was conducted in Denmark.25 The project showed significant

improvements at one-year follow-up in secondary CVD prevention outcomes in patients with

low SES.26 It is, however, important to follow up on this project over a longer period of time.


Page | 3

The overall aim of this PhD dissertation was to investigate the long-term effect of a socially

differentiated CR intervention compared to standard CR in socially vulnerable patients admitted

with first episode myocardial infarction (MI) at Aarhus University Hospital, Denmark between

2000 and 2004.

The main objectives were:

1. To examine the effect of the intervention on total mortality, cardiovascular mortality, non-fatal

recurrent events and major cardiac events (MACE) at ten-year follow-up (Paper I).

2. To examine the effect of the intervention on medication adherence and biological and lifestyle

risk factors at two- five- and ten-year follow-up (Paper II).

3. To examine the effect of the intervention on the use of health care services in general practice

and hospitals at two- five- and ten-year follow-up (Paper III).

Page | 4

2. Background

2. Background

Page | 5

2.1 Cardiovascular disease

2.1.1 Epidemiology European mortality rates due to coronary artery disease (CAD) have declined drastically through

the last half century. It is estimated that 2016 mortality rates are half as high as mortality rates

reported in the early 1980s.2 However, CVD and CAD remain leading causes of mortality,

morbidity and years lived with disability in Europe.2,3 The 2017 edition of the European

Cardiovascular Disease Statistics determines that CVD is the leading cause of death in Europe

and each year 3.9 million Europeans die from CVD. CVD is also the leading cause of premature

death in Europeans younger than 75 and 65 years, respectively. In 2015, the incidence of CVD in

Europe was around 11.3 million new cases while the prevalence showed that more than 85

million Europeans live with CVD. Moreover, CVD was responsible for 64.7 million disability

adjusted life years due to premature death and years lived with disability.3

In Denmark, the most common type of CVD is ischemic heart disease (IHD). IHD is caused by

arteriosclerosis in the coronary arteries. The average yearly incidence and prevalence of IHD was

21,341 and 169,099, respectively between 2010 and 2012. On average, IHD resulted in 4,681

deaths, 29,268 somatic hospitalisations, 88,311 outpatient/ambulant somatic visits and 790,497

visits in general practice per year between 2010 and 2012. Between 2010 and 2012, the average

yearly health economic burden of IHD in Denmark was more than DKK 1,400 millions for

hospitalisations, DKK 162 millions for outpatient/ambulant visits, DKK 73 millions due to visits

in general practice and DKK 203 millions were spent on medicine.27

2.1.2 Risk factors The risk factors associated with development of CVD, CAD and IHD can be divided into non-

modifiable and modifiable risk factors. The non-modifiable risk factors are age, gender and

genetic predispositions.2 The modifiable risk factors are well established in the literature. The

Framingham Heart Study and the Seven Countries Study have found that abnormal lipids,

smoking, hypertension, diabetes, obesity, psychosocial factors, unhealthy diet, excessive alcohol

consumption and low physical activity all contribute to increase the risk of CVD.1,28-30 The

Interheart Study even states that the above nine risk factors combined account for at least 90 %

of the population attributable risk of CVD.31

2. Background

Page | 6

The World Health Organization (WHO)32, the European Society of Cardiology2, the Joint British

Societies33, the British National Institute for health and Care Excellence (NICE)34 and the

Scottish Intercollegiate Guidelines Network (SIGN)35 also emphasize the importance of

acknowledging these risk factors. Different algorithms have been developed to estimate the risk

of CVD. In Europe, the Systematic Coronary Risk Estimation (SCORE) chart is most often used.

The SCORE chart combines information about age, gender, cholesterol levels, blood pressure

(BP) and smoking status and estimates the 10-year risk of fatal CVD in populations of countries

at high and low cardiovascular risk.2,36

2.1.3 Symptoms, diagnosis and treatment of myocardial infarction MI is an acute state of IHD in which the myocardial perfusion is reduced to an extent where cell

necrosis is present. This will most often be caused by a thrombus formation in the coronary

artery due to a rupture or fissuring of an atherosclerotic plaque. Consequently, the blood is

exposed to thrombogenic lipids leading to activation of platelet and clotting factors. Patients with

suspected MI will usually present with heavy chest pain, but symptoms may vary. The diagnosis

of MI includes the patient's clinical history, a physical examination, an electrocardiogram and

cardiac biomarkers. There are different kinds of highly efficient acute invasive and medical

treatment options depending on the specific kind of MI.37,38

Secondary CVD prevention should be initiated before patients are discharged after being

diagnosed with IHD. CVD prevention is a coordinated set of actions including CR aimed at

eliminating or minimizing the impact of CVD and related disabilities.2,37,38 However, many

patients diagnosed with CVD are not referred to and do not participate in secondary prevention

programmes and do thus not achieve the international secondary prevention guideline

standards.39,40

2.2 Cardiac rehabilitation

2.2.1 Definition WHO defines rehabilitation as: "A set of measures that assist individuals who experience, or are

likely to experience, disability to achieve and maintain optimal functioning in interaction with

their environments”.41 The Danish Health Authority uses a translated version of WHO's

rehabilitation definition.42 Considerations concerning help to humans with disabilities have been

2. Background

Page | 7

found in scriptures dated all the way back to ancient times. Modern rehabilitation is considered

to have arisen as a result of the First World War with the aim of reducing the soldiers' physical

disabilities.42 The Danish Society of Cardiology defines CR as a multidisciplinary approach with

the main purpose of lowering the risk of disease progression. Also, the aim of CR is to improve

the patients' physical, mental and social functional level, to eliminate or reduce disabilities and to

maximise patients' quality of life.43

2.2.2 Effect Evidence on the effect of CR is well established. The Cochrane Collaboration performed an

overview on Cochrane systematic reviews in 2014, including six reviews of high methodological

quality. The reviews together contained 148 randomized controlled trials (RCT) and more than

98,000 study participants. It was found that exercise-based CR is effective and safe in secondary

CVD prevention.4 Other recently published reviews and meta-analyses have also found that

participation in CR reduces mortality, morbidity and the number of hospital admissions caused

by CVD. Moreover, CR participation improves physical and psychological factors and increases

quality of life.5-8,44

2.2.3 Core components International and Danish guidelines agree on the importance of focusing on specific core

components of risk factor management when designing and delivering a multidisciplinary CR

programme, including2,22,33-35,38,41,43,45-47:

Adherence to prescribed medicine

Monitoring and controlling levels of lipids, BP and glucose

Physical activity

Smoking cessation

Dietary management and weight control

Psychosocial support including screening for depression and anxiety

Vocational rehabilitation

2. Background

Page | 8

2.2.4 Delivery and implementation CR is divided into three phases. Phase I is the acute phase in the hospital while the patient is still

admitted. Focus is on diagnosis and prognostic procedures as well as initiating patient

information and visitation to secondary prevention. Phase II is the early post discharge phase.

Focus is on maintenance and optimisation of symptomatic and prophylactic treatment as well as

exercise, lifestyle changes and psychosocial support. Phase III is the life-long maintenance

phase. Focus is on maintenance of symptomatic and prophylactic treatment as well as

maintenance of lifestyle changes.45,46,48

Coordinated CR must be delivered by a multidisciplinary team. The CR staff should consist of

cardiologists, nurses, and physiotherapists. The contributing staff could be dieticians,

psychologists, occupational therapists, pharmacists and social workers. If needed, consultant

staff could be cardiac surgeons, internists, diebetologists, neurologists, pneumologists, general

practitioners (GP) and primary care nurses.41,45 CR services can be provided in different settings

such as hospitals and rehabilitation centres or rehabilitation may be home-based. Centre- and

home-based CR programmes have been shown to be equally effective at the same costs.4 Long-

term CR will often be provided in primary care facilities such as community centres or in general

practice.41

Patient education interventions should be a part of CR.49 Current evidence on health education

interventions was presented in a systematic review and meta-analysis from 2017. Three types of

health education interventions in particular were identified; general health education only

including health information and psycho-educational interventions as well as secondary

prevention educational interventions, which included strategies to promote a healthy lifestyle, to

manage medications and to reduce cardiovascular complications.50 The effect of an

individualized CR programme delivered at a rehabilitation day-hospital was investigated by

Sturchio et al.51 It was found that the intervention resulted in a better risk factor control when

compared to usual care.51 Also, Saffi et al.52 found that structured and systematic nurse-led

lifestyle counselling sessions reduced the ten-year CVD risk score.52 Moreover, a RCT from

2017 found that the specific patient education 'Learning and Coping'-strategy improved

adherence to CR.53

2. Background

Page | 9

2.2.5 Legislation There is no comprehensive rehabilitation law in Denmark. Thus professionals in the field of

rehabilitation must use different legal documents and base rehabilitation efforts on the individual

patient's needs and goals. The Danish Health Act (Sundhedsloven) regulates issues concerning

health, disease and disability.42,54 It is a requirement in The Danish Health Act (Sundhedsloven)

that the five regions and 98 municipalities in Denmark must draw up health agreements.

The aim of the health agreements is to ensure that all patients are offered and receive a coherent

and high quality treatment when needed.55 One of the focus areas in the health agreements is

rehabilitation with emphasis on interdisciplinary collaboration and coordination, patient

involvement and health equality as well as documentation, quality development and research.56

As an addition to the health agreements, Central Denmark Region has developed a pathway

programme for IHD, heart failure and heart valve disease.46 The pathway programme is partly

based on the national clinical guidelines for CR47 and describes the interdisciplinary, cross-

sectional and coordinated services in hospitals, the municipal sector and general practice for

patients residing in Central Denmark Region.46

In 2018, The Danish Health Authority published their latest recommendations for

interdisciplinary and cross-sectional services for patients diagnosed with IHD, cardiac

dysrhythmia, heart failure and heart valve disease. The ambition is that patients should be

offered an individually coordinated and high-quality service.22 The quality of the treatment

offered and delivered to patients with IHD in Denmark is monitored nationally through registers

and reports. Moreover, the five regions have established a new register containing hospital-based

data and the goal is also to include data from municipalities and general practice.22

2.2.6 Organisation The Danish Health Reform from 2007 altered the cross-sectional organisation and

responsibilities related to CR. The Danish government has the overall responsibility for health

care planning. It is the regions' responsibility to draw up rehabilitation plans and to initiate CR

before patients are discharged from hospital. The regions also hold the responsibility for the

referral to non-pharmacological phase II CR either in the patient’s home municipality or in

outpatient clinics. In all cases it is recommended that a cardiologist holds the overall medical

responsibility including the responsibility for the pharmacological treatment during phase II CR.

Moreover, the regions are responsible for the services delivered by general practice. The GP

2. Background

Page | 10

holds the medical responsibility for the pharmacological treatment in the long-term phase III CR

and is required to assess the patient’s mental health. The patient’s home municipality holds the

responsibility for inviting the patients to clarifying consultations and to offer non-

pharmacological phase II and phase III CR.22,42,4654,57

2.2.7 The role of general practice The GP is a key person when it comes to initiating, coordinating and providing long-term phase

III CR, which also includes preventive care and monitoring of chronic disease.2,39,58,59 GPs hold

this unique role because of their ability to identify CVD risk factors based on the patient's risk

profile and to assess the patient's eligibility for interventions. Also, the role as gatekeepers and

the fact that the GP is the patient’s first contact in the health care system provides the GP with a

unique possibility to provide continuity of care and to manage any comorbidity with a patient-

centred approach. A successful effort by the GP based on international guidelines reduces the

number of recurrent MIs and hospital admissions.2,39,58 A systematic review and meta-analysis

found that interventions aimed at improving clinician and patient adherence to guidelines are

associated with a reduced risk of all-cause and cardiovascular mortality at four to six years

follow-up.60 In Denmark, all patients diagnosed with IHD are entitled an annual chronic care

consultation in general practice. The consultation should be based on the resources, level of self-

care and comorbidity of each individual patient.61

The GPs are considered to hold a responsibility that extends beyond clinical practice. GPs

should be involved in integrating policies at national and regional level with a strong focus on

outlining their own role in the long-term secondary CVD prevention.62 To maintain their future

role as key persons, GPs must focus even more on rehabilitation. There is a need for a

comprehensive approach not just based on the patients' diagnoses but on their overall

functioning. The successful integration of rehabilitation in the every-day life in general practice

depends on the use of63:

1) Patient education and involvement of relatives to broaden the patients' knowledge about their

own diseases, improve their self-care and provide support to succeed.63

2) Problem solving therapy when addressing physical and mental disorders.63

2. Background

Page | 11

3) Multidisciplinary teams based on the collaborative care model to collaborate about the

patients and the provided rehabilitation.63

4) New technology that will create more time for rehabilitation and improve the patients' level of

self-care.63

2.3 Social inequality in health

2.3.1 Definition The Danish Health Authority defines social inequality in health as systematic differences in

health between citizens divided by socioeconomic criteria. The main factors influencing social

inequality in health are the development in the first years of life, education, health behaviour,

work environment, local environment, unemployment, social vulnerability and inequalities in the

use of health care services.64

2.3.2 Policies Health inequality was already on the agenda back in 1981 where the WHO announced their

'Global Strategy for Health for All by the Year 2000'. The aim of the strategy was to achieve the

highest possible level of health for all by the year 2000. The strategy emphasised that the

existing inequality in health status was of concern and had to be drastically reduced. Also, it was

emphasized that it was fundamental to work towards an equitable distribution of health resources

both within and among countries, which would lead to universal accessibility to primary health

care and supporting services.65 The strategy was followed up in 1985 where the WHO announced

38 targets which were seen as fundamental requirements for health of people in the European

region of the WHO by the year 2000. Target one was 'Reducing the differences'. The goal was

that by the year 2000, differences in health status within and between countries should be

reduced by at least 25 %. It was estimated that this goal could be reached if all were provided

with the same prerequisites for health, if lifestyle-related risks were reduced, if the health aspects

of living and working were improved and if everyone had access to high-quality primary health

care.66 Another policy framework followed in 1999, where the 38 targets were reduced to 21.

One of the targets was 'Ensuring equity in health through solidarity in action' with focus on

inequality in health due to poverty, unemployment, gender, ethnicity and disability.67

2. Background

Page | 12

In 1999, the Danish government led by the Social Democratic Party presented their ten-year

public health programme. One of the two overall goals was to reduce social inequality in health

as much as possible through an intensified effort to improve the health among the most

disadvantaged groups.68 The goal was continued in 2002, where the next liberal government

presented their national public health strategy. But unlike the previous programme, this

programme focused on the eight most common Danish diseases including CVD. Moreover,

emphasis was put on rehabilitation to improve the patients’ quality of life.69 To this day, fighting

social inequality in health has been on the public health agenda and it still has political

awareness in Denmark and in the European Union.23,24

Despite the many years of focusing on reducing social inequalities in health, Denmark and its

neighbouring countries are challenged by a phenomenon known as 'the Scandinavian Welfare

Paradox of Health'. Despite the many years with an extensive social welfare system resulting in

small income inequalities and no serious financial barriers to healthcare, it has not been possible

also to reduce social inequalities in overall mortality.70,71 It is believed that inequality in

mortality is the result of both social differences in the likelihood of getting ill and in social

differences concerning survival. The average age standardised mortality rates for men and

women with a high education is 4.5, 3.1 and 3.3 per 1000 in Denmark, Norway and Sweden,

respectively. Compared with this the average age standardised mortality rates for men and

women with a low education is 8.5, 6.5 and 6.1 per 1000 in Denmark, Norway and Sweden,

respectively.72

2.3.3 Social inequality in cardiovascular disease and secondary prevention A WHO rapport from 2010 states that evidence on social determinants and inequalities indicates

an inverse relationship between SES and CVD incidence and mortality. Other factors such as

low social support also impact on the development and outcomes of CVD through a life course.

The WHO emphasizes that protection of CVD health in lower socioeconomic groups should be a

priority in population-based prevention. Primary care is considered the most appropriate health

sector entry-point for addressing social inequality in health.21 A report from 2015 by the Danish

Health Authority showed an inverse relationship between educational level and IHD. Having a

primary school education or a short education was associated with a higher IHD incidence, a

higher all-cause mortality, more life years lost until the age of 75 years, a higher number of

2. Background

Page | 13

admissions and outpatient hospital visits as well as a higher number of visits in general practice

compared to having a medium cycle or long cycle education.27

Numerous studies have examined if and how socioeconomic factors and social support is

associated with CVD. Three studies from Denmark, Norway and Sweden, respectively highlight

the previously mentioned welfare paradox. The studies found that low income, low educational

level and living in areas with lower SES were associated with mortality and a poorer prognosis

after MI.9-11 The role of SES in mortality risk prediction of MI was examined in a 13-year

follow-up study. SES was divided into individual SES and neighbourhood SES. Individual SES

consisted of income, education, and employment before MI as well as living with a partner.

Neighbourhood SES was measured through a developed index. It was found that low educational

level, lack of a steady partner, below average income, and pre-MI unemployment were

significantly and independently associated with mortality.12 Being unmarried or living alone was

also found to predict 30-day hospital readmission after invasive surgery.13 Low social support or

lacking a close confidant was moreover found to be a risk factor for and associated with adverse

outcomes post-MI in patients with CHD patients.14,15 However, a Danish study found that low

educated patients and patients living alone who were admitted with MI had a significantly lower

six-moth risk of readmission. At two-year follow-up, educational level and cohabitation status

were found not to be significantly associated with the number of contacts to GPs.73

A newly published study based on register data from the 'European Action on Secondary and

Primary Prevention by Intervention to Reduce Events' (EUROASPIRE) IV survey examined

smoking cessation rates in a European population of patients diagnosed with coronary heart

disease (CHD). The study found that successful smoking cessation was associated with higher

levels of education.16 Higher education was in the same survey also associated with a

significantly better control of CHD risk factors defined as smoking, overweight and obesity,

physical activity and lipid control in men and obesity, BP, physical activity and lipid control in

women.17 An Australian cohort study examined the socioeconomic variation in incidence of

primary and secondary CVD events and found that incidence increased with decreasing

education level.18 A recent study examined the impact of educational level on ten-year prognosis

in patients diagnosed with acute coronary syndrome (ACS). All-cause mortality and number of

recurrent ACS-events were significantly higher in the group with a low educational level

compared to those with a medium cycle or long cycle education.19

2. Background

Page | 14

A systematic literature review found that patients with lower SES are less likely to make

behavioural changes post-MI. It is thus recommended that an increase in participation in CR

among patients with lower SES should be prioritised.74 Another review investigated the

effectiveness of interventions targeting secondary prevention of CVD, including combination

pharmacotherapy and CR among lower SES groups. It was found that patients from lower SES

groups achieve comparable risk factor control as patients from higher SES groups when applying

combination pharmacotherapy. Also, CR was found to be effective to modify risk factors and to

improve dietary, exercise and smoking habits.20 A psychosocial CR intervention was also found

to have a positive effect in patients with low social support.14 A RCT from Sweden examined the

effect of an intervention given to patients post-ACS compared to usual care. It was found that

patients without post-secondary education benefitted significantly from the intervention when

evaluated on a composite score.75 Another RCT from Germany examined the effect of an

intervention compared to usual care in a study population that consisted of CAD patients who

were almost only male and with low educational level. The patients with low educational level

who received the intervention had a significantly better risk profile and a higher quality of life

score than patients with low educational level who received usual care.76 Despite the beneficial,

positive and significant effects of CR among lower SES groups, a recent Danish study found that

patients with low SES receive significantly less information about CR and are significantly less

willing to participate in CR. If the patients with low SES were informed and willing to

participate in CR, they were more often referred to community-based CR than to specialised

CR.77

2.4 The original socially differentiated cardiac rehabilitation project

The former Aarhus County - the current Central Denmark Region since the Danish Health

Reform in effect from 2007 – conducted a project from 2000 to 2004 in cooperation with

Department of Cardiology at the former Aarhus County Hospital – the current Aarhus University

Hospital. The project was supported by the Ministry of Social Affairs, the Ministry of Health,

Aarhus County and the Danish Heart Foundation. The aim of the project was to examine the

possibility of systematically offering and implementing a socially differentiated CR intervention

to reduce social inequality in health among socially vulnerable patients admitted with first

episode MI.25 The follow-up after one year showed significant improvements regarding

medication adherence, lipid values, systolic BP and body mass index when comparing socially

2. Background

Page | 15

vulnerable patients who received a socially differentiated CR intervention with socially

vulnerable patients who received standard CR.26

Years have passed since the project was conducted. However, social inequality in health has not

been eradicated neither in Denmark nor internationally. A systematic review from 2018 found a

limited number of studies describing expanded CR in addition to standard CR, and none of the

studies reported specific socially differentiated CR interventions.78 This, combined with the

significant results at one year follow-up, and the fact that the Danish national clinical guidelines

for CR47 and the Central Denmark Region pathway programme for IHD, heart failure and heart

valve disease46 both refer to the original socially differentiated CR project in their guidelines

regarding psychosocial efforts makes it relevant and important to investigate the long-term effect

of the intervention.

Page | 16

3. Methods

3. Methods

Page | 17

3.1 Study setting

The original socially differentiated CR project was performed at Department of Cardiology at

Aarhus County Hospital in Denmark. A new hospital structure has merged many smaller

hospitals, and Aarhus County Hospital is today a part of Aarhus University Hospital, which is

located in Central Denmark Region, which is one of the five regions in Denmark. Central

Denmark Region has 19 of the 98 municipalities in Denmark and the population in 2010 was

1,253,998. The area of the region is 13,124 km2 equivalent to 30 % of the total area of

Denmark.25,79

3.2 Study population

From 1 April 2000 to 31 March 2002 (24 months), 205 patients below the age of 70 years were

admitted with first episode MI. Of these patients, 171 (83 %) were offered to participate in CR.

A total of 146 patients participated in CR (71 %) and 133 (65 %) patients gave written, informed

consent. From 1 September 2002 to 31 December 2004 (28 months), 303 patients below the age

of 70 years were admitted with first episode MI. Of these patients, 279 (89 %) were offered to

participate in CR. A total of 256 patients participated in CR (85 %) and 246 (81 %) patients gave

written, informed consent (Figure 1).25,26,80-82 Criteria for being diagnosed with MI were clinical

history, significant electrocardiogram changes and / or raised biochemical markers.25 Patients

were excluded if they had serious comorbidities such as stroke, dementia, retardation, severe

alcohol abuse or mental disease. Patients with depression or anxiety were not excluded.25,26,80-82

Participation in CR was defined as attendance for at least one consultation with a cardiologist

and attendance for at least three consultations with a cardiac nurse for patients receiving standard

CR and attendance for at least four consultations with a cardiac nurse for patients receiving the

intervention.25 The study population consisted of 379 patients admitted with first episode MI who

participated in CR and who gave written informed consent.25,26,80-82

3. Methods

Page | 18

The 379 patients in the study population were categorised as socially vulnerable or non-socially

vulnerable by a project nurse before discharge. Patients were defined as socially vulnerable if

they had a low educational level and / or if they lived alone. Low educational level was defined

as having an education coded as 1- 4 in the Danish Education Nomenclature (DUN) in patients

who were 55 years old or younger and 1-3 in patients who were older than 55 years.

DUN level 1 is equal to not having finished primary school. DUN level 2 is equal to primary

school as the highest education completed. DUN level 3 is equal to primary school as the highest

education completed supplemented with some courses of a shorter duration. DUN level 4 is

equal to upper secondary education as the highest education completed for example high school

or a craftsman education.83 Of the 133 patients admitted during the study period from 1 April

2000 to 31 March 2002, 78 patients were categorised as socially vulnerable and 55 patients were

categorised as non-socially vulnerable. Of the 246 patients admitted during the study period from

1 September 2002 to 31 December 2004, 130 patients were categorised as socially vulnerable

and 116 patients were categorised as non-socially vulnerable.25,26,80-82

3. Methods

Page | 19

Figure 1. Flowchart of the study population25,26,80

All patients < 70 years admitted at Aarhus University Hospital with

first episode MI from 1 April 2000 – 31 March 2002.

N = 205 (100 %)

All patients < 70 years admitted at Aarhus University Hospital with

first episode MI from 1 September 2002 – 31 December 2004.

N = 303 (100 %)

Patients referred to CR.

N = 171 (83 %)

Patients referred to CR.

N = 270 (89 %)

Patients who participated in

CR and gave informed consent. N = 133 (65 %)

Patients who participated in CR and gave informed consent.

N = 246 (81 %)

Socially differentiated CR

N = 379

Patients categorised

as socially vulnerable.

Received standard CR.

N = 78


as non‐socially vulnerable.


N = 55


as socially vulnerable.

Received expanded CR.

N = 130


as non‐socially vulnerable.


N = 116

3. Methods

Page | 20

3.3 Study design

The study was performed as a prospective cohort study.80-82 Baseline was defined as the date of

admission with first episode MI.25,26,80-82 Ten-year follow-up was conducted in Paper I80 and

two-, five- and ten-year follow-up was conducted in Papers II and III.81,82

3.4 Intervention

During the study period from 1 April 2000 to 31 March 2002, clinical practice regarding

standard CR was observed. The 133 patients admitted in this study period received standard CR

according to the guidelines84-86 applicable at the time regardless of being categorised as socially

vulnerable or not. During the study period from 1 September 2002 to 31 December 2004, the

intervention was performed. The 116 patients categorised as non-socially vulnerable received

standard CR according to the guidelines84-86 applicable at the time. The 130 patients categorised

as socially vulnerable received an expanded CR intervention in addition to standard CR.25,26,80-82

In phase I, standard CR and expanded CR consisted of the same components. Medical and

possibly surgical treatment was initiated. Also, secondary prevention concerning medication,

smoking, diet and physical activity was initiated. Moreover, the patients and their relatives were

offered psychological and social support. In phase II, standard CR consisted of five to six weeks

of CR. This involved three consultations with a cardiologist, four consultations with a cardiac

nurse, two consultations with a dietician, an exercise programme lasting from six to 12 weeks

and a screening for depression and anxiety performed six weeks after admission. Expanded CR

had the same components as standard CR. In addition, it consisted of two extra weeks of CR and

one extra consultation with a cardiac nurse. Patients were entitled to extra consultations if

necessary. Moreover, the patients played an active role in drawing up individual rehabilitation

plans which were sent to the patients' GPs. In phase III, standard CR consisted of a referral to the

GP and information about activities in the municipal sector and in the Danish Heart Foundation.

Expanded CR had the same components as standard CR. In addition, it consisted of a referral to

a 30-minute preventive consultation at the GP, a referral to activities in the municipal sector and

in the Danish Heart Foundation as well as a telephone follow-up performed by a cardiac nurse

two months after the completion of phase II (Table 1).25,26,80-82

3. Methods

Page | 21

Table 1. Content of socially differentiated CR80

STANDARD CR

EXPANDED CR

PHASE I

Start of medical and surgical treatment

Start of secondary prevention

Psychological and social support to patients and relatives

Standard CR

PHASE II

5‐6 weeks of CR

3 consultations with a cardiologist

4 consultations with a cardiac nurse

2 consultations with a dietician

6‐12 weeks of exercise programme

Screening for depression and anxiety 6 weeks after admission

Standard CR and:

Extra 2 weeks of CR

1 extra consultation with a cardiac nurse

Individual rehabilitation plan

Sharing of individual rehabilitation plan with GP

PHASE III

Referral to GP

Information about activities in the municipal sector and in the Danish Heart Foundation

Standard CR and:

Referral to 1 preventive consultation at the GP

Referral to activities in the municipal sector and in the Danish Heart Foundation

Telephone follow‐up conducted by a cardiac nurse 2 months after completion of phase II

3.5 Data sources

3.5.1 Baseline data The baseline data consisted of demographic, socioeconomic and clinical information about the

patients. Baseline data were collected before the patients were discharged after their first episode

of MI. Data regarding age, gender, total cholesterol, high-density lipoprotein (HDL) cholesterol,

low-density lipoprotein (LDL) cholesterol, triglyceride, fasting blood glucose, body mass index,

prescribed medication and the possible presence of hypertension, hyperlipidaemia, diabetes

mellitus and other diseases were collected from patient records by a project nurse. Data

regarding educational level, cohabitation status and smoking status were collected from a

questionnaire completed by the patients. Baseline data were typed in by a project nurse and

safely stored.

3. Methods

Page | 22

3.5.2 Follow-up data

3.5.2.1 Registers

The Danish Civil Registration System (CRS)

The CRS was established in 1968 and has registered all Danish citizens since 1968. It is required

by Danish law that all permanent residents have a unique ten-digit Civil Personal Registration

number (CPR number).87,88 The first six digits refer to the date and year of birth and the last four

digits distinguish persons who are born on the same day and year.80-82,87,88 The CPR number is a

key component in the linkage between the study population and the selected registers as it

provides the possibility for highly valid and almost complete (99.7 %) long-term follow-up.87

The Danish Cause of Death Register (CDR)

Since 1871, it has been required by Danish law to fill in a death certificate any time a death

occurs in Denmark. In 1875, a systematic national registration of cause of death began, which

has been digitalised since 1970. The CDR contains information on all Danish residents who have

died in Denmark. The International Classification of Diseases 10th edition (ICD-10) codes are

used when registering cause of death.89

The Danish National Patient Register (NPR)

The NPR was established in 1977 and holds information on all somatic hospital admissions,

outpatient contacts as well as emergency and psychiatric contacts to the healthcare system. The

admission diagnosis is defined by the doctor in charge of the patient using ICD-10 codes.90

The Danish National Prescription Register (DNPR)

The DNPR was established in 1994 and contains information on all prescription medicine sold in

Denmark. The medicine is divided into groups and identified through the Anatomical

Therapeutic Chemical Classification (ATC) codes.91

The Central Denmark Region Clinical Laboratory Information System (LABKA)

The LABKA system was initially implemented in the clinical departments on Aarhus County

Hospital in 1985. Today, LABKA holds information on all blood sample test results retrieved at

public and private hospitals and by GPs in Central Denmark Region, North Denmark Region and

the Capital Region.92

3. Methods

Page | 23

The Danish National Health Service Register (NHSR)

Data from the NHSR have been available for use in research since 1990. The NHSR contains

information about all contacts to health professionals in primary care, including GPs, who have a

contract with the public health care system in Denmark.93

3.5.2.2 Questionnaire survey in general practice

A questionnaire was developed to be used in general practice. The questionnaire was reviewed

by an experienced GP. Based on comments from the GP, the questionnaire was adjusted and sent

to three of the patients' GPs for pilot testing. The pilot test gave no rise to changes and the

questionnaire was then forwarded to all of the patients' GPs including an invitation letter

(Appendix IV). The invitation letter contained information about the patient's CPR number and

the questionnaire contained information about the date the specific patient was admitted with a

first episode MI. The GP was requested to fill in the questionnaire with information from the

medical record concerning the patient's BP (mm/Hg) and smoking status (smoker / non-smoker)

at follow-up. The medical record is a workings paper for the GP and the GP does not

systematically register health behaviour information such as the patient's dietary habits or

physical activity level. However, it is common that the GP registers the patient's BP and smoking

habits. The GP received a compensation of DKK 253.30 for each completed and returned

questionnaire. A reminder was sent out twice to GPs who did not return the questionnaires to

increase the response rate. All questionnaires were sent by registered letters to the GPs as they

contained confidential personal information. Each questionnaire had a unique serial number

which could be linked to the specific patient in question. Thus, the GPs did not send any

confidential personal information when returning the questionnaires by ordinary mail. The data

from the returned questionnaires were entered into the data documentation programme

'Epidata'94 by two different assistants. All data entries were compared and in case of any

discrepancy, the questionnaires were re-evaluated by the two assistants.81

3. Methods

Page | 24

3.6 Outcomes

3.6.1 Paper I The outcomes were:

All-cause mortality, cardiovascular mortality, non-fatal recurrent events and MACE.

Information on all-cause mortality and cardiovascular mortality was obtained from the CDR.

All-cause mortality was defined as any cause of death reported in the register during follow-up.

If a patient had died during follow-up, data on the underlying cause of death was obtained from

the CDR using the automated classification of medical entities (ACME) based on the ICD-10

codes.95 All underlying causes of death were reviewed and defined as 'cardiovascular mortality'

or 'other cause of death' in cooperation with an experienced cardiologist. The ICD-10 codes

defined as 'cardiovascular mortality' will be reported in the results section. Information on non-

fatal recurrent events was obtained from the NPR. In cooperation with an experienced

cardiologist, it was defined that a non-fatal recurrent event had occurred if a patient appeared in

the NPR during follow-up with at least one of the following ICD-10 codes: DI21, DI219, DI200,

DI200B and DI200C. MACE was defined as cardiovascular mortality and non-fatal recurrent

events pooled together in a composite outcome. MACE was considered to have occurred if a

patient had at least one non-fatal recurrent event or had died from CVD during follow-up.80

3.6.2 Paper II The outcomes were:

Medication adherence to antithrombotics, beta-blockers, statins and angiotensin-converting

enzyme inhibitors (ACE inhibitors), total cholesterol, HDL cholesterol, LDL cholesterol,

triglyceride, haemoglobin A1c (HbA1c), BP and smoking status.

Information on medication adherence was obtained from the DNPR and defined as the purchase

of at least one prescription every year of follow-up using ATC-codes. The following ATC-codes

were used81:

Anti-thrombotics: B01AC04, B01AC06, B01AC56

Beta-blockers: C07AA, C07AB

Statins: C10AA, C10AB, C10AD, C10AX09, C10BA

ACE inhibitors: C09AA, C09BB, C09CA, C09D

3. Methods

Page | 25

Information on total cholesterol, HDL cholesterol, LDL cholesterol, triglyceride and HbA1c was

obtained from the LABKA system. Total cholesterol, HDL cholesterol, LDL cholesterol and

triglyceride were measured as mmol/L. HbA1c was measured as mmol/mol. Each of the blood

test outcomes were mean values computed from all values obtained during each follow-up.81

Information on BP and smoking status was obtained through a questionnaire survey in general

practice and filled in by the patients' GPs. BP was measured as mm/Hg and smoking status was

reported as 'smoker' or 'non-smoker' at each follow-up (Appendix IV).81

3.6.3 Paper III

The outcomes were:

Participation in annual chronic care consultations in general practice, contacts to general

practice, all-cause hospitalisations and cardiovascular readmissions.

Information on participation in annual chronic care consultations in general practice and contacts

to general practice was obtained from the NHSR. In cooperation with a experienced GP,

participation in annual chronic care consultations in general practice had occurred if a patient

appeared in the NHSR with one the following codes each year during follow-up: 0106, 0104,

2304 and 0120. Participation in annual chronic care consultations was measured dichotomously

(yes/no). Contacts to general practice were defined as any contact involving the general practice

clinic and were assessed each year during follow-up. This included visits to the GP, telephone

consultations, email consultations, home visits as well as contacts and visits to and from other

health care professionals in the general practice clinic. Contacts to general practice were

measured continuously (number of contacts) and were defined as the number of times a patient

appeared in the NHSR with one of the following codes during each year of follow-up: 01, 02, 03,

04 and 05.82

Information on all-cause hospitalisations and cardiovascular readmissions was obtained from the

NPR and measured as number of hospital admissions during follow-up. All-cause hospitalisation

was defined as any kind of admission reported in the register during follow-up. In cooperation

with an experienced cardiologist it was defined that a cardiovascular readmission had occurred if

a patient appeared in the NPR during follow-up with at least one of the following ICD-10 codes:

DI20, DI21, DI50, DI61 and DI63. All-cause hospitalisations and cardiovascular readmissions

3. Methods

Page | 26

were measured continuously (number of hospitalisations / number of cardiovascular

readmissions).82

3.6.4 Additional analyses The outcomes were all outcomes from Paper I80, Paper II81 and Paper III82.

3.7 Statistical analysis

The statistical analyses in Paper I, Paper II and Paper III compared socially vulnerable patients

admitted from 1 September 2002 to 31 December 2004 who received expanded CR (N=130) and

socially vulnerable patients admitted from 1 April 2000 to 31 March 2002 (N=78) who received

standard CR. To accommodate the possible differences between the two calendar periods, a

supplemental analysis of the non-socially vulnerable patients in the two corresponding time

periods was performed. All analyses were based on yearly survivors. Data management and

analyses were performed using Stata/MP version 14.1, 14.2 and 15.0 and SAS/STAT version

9.4. A significance level of 0.05 was applied; P-values below 0.05 were considered statistically

significant.80-82

3.7.1 Baseline data Categorical variables were presented as numbers with the associated percentages. Continuous

variables were presented as means with the associated standard deviations (SD). All baseline

data were presented with P-values.

3.7.2 Paper I Data on all-cause mortality, cardiovascular mortality, non-fatal recurrent events and MACE were

shown as numbers with the associated percentages. Comparisons were evaluated using statistical

regression analysis, in specific adjusted logistic regression models were used to mitigate

potential confounding bias. All associations were estimated by odds ratios (OR) with 95%

confidence intervals (CI) and P-values. Analyses of all-cause mortality and non-fatal recurrent

events were adjusted for gender, age and diabetes mellitus status. Cardiovascular mortality was

adjusted for gender. MACE was adjusted for gender, age, diabetes mellitus and smoking status.

Only patients who were not diagnosed with a non-fatal recurrent event in the first month after

3. Methods

Page | 27

admission due to first episode MI were included in the analyses concerning non-fatal recurrent

events and MACE.80

3.7.3 Paper II Evaluations of the outcomes were performed as statistical tests. Data on medication adherence

outcomes were shown as proportions of patients who had purchased at least one prescription for

each type of medicine at each time of follow-up. Comparisons were assessed by calculating

ratios of proportions. The ratios were presented with logit-transformed 95% CIs and P-values

from χ2 tests. Data on blood test outcomes were shown as means of all blood samples retrieved at

each time of follow-up. Comparisons were evaluated using t-tests and reported as differences in

means. The differences were presented with 95% CIs and P-values. Questionnaire data

evaluating BP outcomes were shown as means at each time of follow-up. Comparisons were

evaluated using t-tests. The differences were presented with 95% CIs and P-values.

Questionnaire data on smoking status were shown as proportions at each time of follow-up.

Comparisons were evaluated using Χ2 tests. The ratios were presented with logit-transformed

95% CIs and P-values.81

3.7.4 Paper III Evaluations of the outcomes were performed as statistical tests. Participation in annual chronic

care consultations in general practice was shown as proportions at each time of follow-up.

Comparisons were evaluated using Χ2 tests. The ratios of proportions were presented with logit-

transformed 95% CIs and P-values. Contacts to general practice, all-cause hospitalisations and

cardiovascular readmissions were shown as means at each time of follow-up. Comparisons were

evaluated using t-tests. The differences were presented with 95% CIs and P-values.82

3.7.5 Additional analyses The additional statistical analyses compared all socially vulnerable patients in the study

population, N=208 (130+78) to all non-socially vulnerable patients in the study population,

N=171 (116+55) regardless of the type of CR received. The comparisons were performed as

descriptive analyses including Kaplan-Meier estimates and plots of summary statistics.

3. Methods

Page | 28

3.8 Approval and ethics

An informed, written consent was obtained from all patients participating in the study. The

Danish Data Protection Agency approved the study (case number: 1-16-02-684-14). Ethical

approval was not required.

Page | 29

4. Results in summary


Page | 30

A thorough presentation of the findings and visualisation of findings at follow-up in the entire

study population divided into socially vulnerable and non-socially vulnerable patients are

accessible in Paper I80, Paper II81 and Paper III82 in appendix I, II and III. The following

summary provides an overview of the findings of expanded CR compared to standard CR in

socially vulnerable patients. Moreover, additional descriptive analyses are presented at the end of

the summary.

4.1 Baseline characteristics

The baseline characteristics of the entire study population divided by social vulnerability,

rehabilitation type and time period are shown in Table 2.

Mean age among the socially vulnerable patients, primarily men, was 55.5 years. The two groups

of socially vulnerable patients were not significantly different in relation to age (P 0.65),

educational level (P 0.66) or cohabitation status (P 0.51), which were the baseline variables

determining whether the patients were categorised as socially vulnerable or not. The two groups

of socially vulnerable patients were significantly different in relation to the baseline variables

total cholesterol (P 0.00), LDL-cholesterol (P 0.02) and triglyceride (P 0.02) (Table 2).


Page | 31

Table 2. Baseline characteristics of the study population (N=379) divided by social vulnerability, rehabilitation type

and time period. All data are given as numbers (percentage) or as means (SD) with P‐values.

Socially vulnerable patients

Non‐socially vulnerable patients

Rehabilitation typeN

Time period

Rehabilitation type N

Time period

Standard CR

N=78 2000‐2002

Expanded CR

N=130 2002‐2004

P‐ Value

Standard CR

N=55 2000‐2002

Standard

CR N=116

2002‐2004

P‐ Value

Age at admission, years

56 (8.2) 55 (8.5) 0.65 60 (7.6) 57 (73) 0.02

Gender, male 57 (73) 93 (71) 0.81 42 (76) 94 (81) 0.48

Educational level (DUN)

3.2 (1.2) 3.3 (1.4) 0.66 4.8 (1.1) 4.8 (1.2) 0.79

Living alone 27 (35) 51 (39) 0.51 0 0 ‐

Total cholesterol, mmol/L

5.7 (1.5) 5.2 (1.0) 0.00 5.2 (0.9) 5.2 (0.9) 0.88

HDL cholesterol, mmol/L

1.2 (0.3) 1.2 (0.3) 0.13 1.3 (0.3) 1.3 (0.4) 0.89

LDL cholesterol, mmol/L

3.5 (1.0) 3.2 (0.9) 0.02 3.3 (0.9) 3.2 (0.8) 0.58

Triglyceride, mmol/L 2.0 (1.1) 1.7 (1.0) 0.02 1.5 (1.0) 1.7 (0.9) 0.25

Fasting blood glucose, mmol/L

7.5 (4.6) 6.9 (2.8) 0.25 6.8 (3.3) 6.7 (2.0) 0.69

Prescribed anti‐thrombotics

72 (92) 126 (97) 0.13 47 (86) 112 (97) 0.01

Prescribed beta‐blocker

71 (91) 116 (89) 0.67 49 (89) 107 (92) 0.50

Prescribed statins 20 (26) 99 (76) 0.00 10 (18) 104 (90) 0.00

Prescribed ACE‐inhibitors

24 (31) 55 (42) 0.09 20 (36) 49 (42) 0.46

Current smoker 59 (76) 83 (64) 0.28 34 (62) 60 (52) 0.29

Body Mass Index 27.3 (4.4) 26.3 (4.1) 0.10 26.4 (4.0) 26.5 (3.1) 0.77

Hypertension 18 (23) 28 (22) 0.80 11 (20) 23 (20) 0.98

Diabetes mellitus 10 (13) 16 (12) 0.91 6 (11) 10 (9) 0.63

Other diseases 39 (50) 82 (64) 0.06 27 (49) 68 (59) 0.27


Page | 32

4.2 Paper I

A full follow-up of yearly survivors at ten-year follow-up was achieved (Table 3). The reviewed

underlying causes of deaths defined as 'cardiovascular mortality' had the following ICD-10

codes: I109, I219, I249, I251, I252, I255, I258, I259, I489, I501, I509, I519, I639, I649, I702

and I718. No significant differences between expanded CR and standard CR were seen in

adjusted analyses comparing socially vulnerable patients who received expanded CR versus

standard CR concerning all-cause mortality, cardiovascular mortality, non-fatal recurrent events

or MACE (P > 0.05) at ten-year follow-up (Table 3).80

Table 3. All‐cause mortality, cardiovascular mortality, non‐fatal recurrent events and MACE among socially

vulnerable patients who received expanded CR and standard CR at ten‐year follow‐up.80

Total

(N = 208) Standard CR (N = 78)

Expanded CR (N = 130)

OR (95 % CI)

P‐value

All‐cause Mortality*

35 (17) 12 (15) 23 (18) 1.29 (0.6;2.9) 0.53

Cardiovascular Mortality**

19 (9) 8 (10) 11 (8) 0.80 (0.3;2.1) 0.65

Total

(N = 176***) Standard CR (N = 61***)

Expanded CR (N = 115***)

OR (95 % CI)

P‐value

Non‐fatal recurrent events*

30 (17) 8 (13) 22 (19) 1.62 (0.7;3.9) 0.29

MACE**** 41 (23) 14 (23) 27 (23) 1.31 (0.5;2.4) 0.75

* Adjusted for gender, age and diabetes mellitus. ** Adjusted for gender. *** Only patients who did not have a recurrent event during the first month after admission were included in the analysis. **** Adjusted for gender, age, diabetes and smoking status.


Page | 33

4.3 Paper II

Medication adherence

A full follow-up of yearly survivors at two-, five- and ten-year was achieved (Table 4).

Medication adherence to antithrombotics was approximately 90 % throughout the ten-year

follow-up in socially vulnerable patients. No significant differences between expanded CR and

standard CR were seen in socially vulnerable patients during follow-up (P > 0.05). Medication

adherence to beta-blockers varied and decreased during the ten-year follow-up in socially

vulnerable patients. Socially vulnerable patients receiving expanded CR had a significantly

lower proportion of adherence at two- (ratio 0.7 95% CI 0.6;0.8 P 0.00) and five-year (ratio 0.8

95% CI 0.6;1.0 P 0.02) follow-up but not at ten-year follow-up (P > 0.05) compared to socially

vulnerable patients receiving standard CR. Medication adherence to statins was more than 80 %

throughout the ten-year follow-up in socially vulnerable patients. Socially vulnerable patients

receiving expanded CR had a significantly higher proportion of adherence at two-year (ratio 1.1

95% CI 1.0;1.3 P 0.04) follow-up but not at five- and ten-year follow-up (P > 0.05) compared to

socially vulnerable patients receiving standard CR. Medication adherence to ACE inhibitors

increased during the ten-year follow-up in socially vulnerable patients. No significant differences

between expanded CR and standard CR were seen in socially vulnerable patients during follow-

up (P > 0.05) (Table 4).81

Table 4. Medication adherence among socially vulnerable patients receiving expanded CR and standard CR,

respectively at two‐, five‐ and ten‐year follow‐up.81


Standard CR Expanded CR

Year of follow‐up

Proportion

N Proportion

N Ratio P‐value

Anti‐thrombotics

2 0.90 78 0.95 128 1.1 (0.9;1.1) 0.37

5 0.92 75 0.89 125 1.0 (0.9;1.1) 0.46

10 0.94 66 0.90 107 1.0 (0.9;1.0) 0.33

Beta‐blockers

2 0.82 78 0.55 128 0.7 (0.6 0.8) 0.00

5 0.68 75 0.52 125 0.8 (0.6;1.0) 0.02

10 0.58 66 0.47 107 0.8 (0.6;1.1) 0.16

Statins

2 0.82 78 0.91 128 1.1 (1.0;1.3) 0.04

5 0.88 75 0.86 125 1.0 (0.9;1.1) 0.74

10 0.91 66 0.90 107 1.0 (0.9;1.1) 0.79

ACE‐inhibitors

2 0.38 78 0.47 128 1.2 (0.9;1.7) 0.23

5 0.44 75 0.53 125 1.2 (0.9;1.6) 0.22

10 0.50 66 0.56 107 1.1 (0.8;1.5) 0.43


Page | 34

Blood tests

The number of patients contributing with blood test outcomes varied at each follow-up and for

each outcome (Table 5). The mean values of total cholesterol, LDL cholesterol and triglyceride

decreased during the ten-year follow-up in socially vulnerable patients. Socially vulnerable

patients receiving expanded CR had a significantly lower mean value of total cholesterol

(difference -0.6 95% CI -0.8;-0.4 P 0.00 and -0.4 95% CI -0.6;-0.2 P 0.00), LDL cholesterol

(difference -0.5 95% CI -0.7;-0.4 P 0.00 and -0.3 95% CI -0.4;0.1 P 0.00) and triglyceride

(difference -0.4 95% CI -0.7;-0.2 P 0.00 and -0.2 95% CI -0.4;-0.0 P 0.04) at two- and five-year

follow-up but not at ten-year (P > 0.05) follow-up compared to socially vulnerable patients

receiving standard CR. The mean values of HDL cholesterol remained nearly stable throughout

the ten-year follow-up in socially vulnerable patients. Socially vulnerable patients receiving

expanded CR had a significantly higher mean value of HDL cholesterol at two-year (difference

0.1 95% CI 0.1;0.2 P 0.01) follow-up but not at five- or ten-year follow-up (P > 0.05) compared

to socially vulnerable patients receiving standard CR. The mean values of HbA1c varied during

the ten-year follow-up in socially vulnerable patients. Socially vulnerable patients receiving

expanded CR had a significantly lower mean value of HbA1c at two (difference -8.4 95% CI

-12.6;-4.2 P 0.00) and five-year (difference -6.4 95% CI -11.3;-1.4 P 0.01) follow-up but not at

ten-year (P > 0.05) follow-up compared to socially vulnerable patients receiving standard CR

(Table 5).81

Table 5. Blood tests among socially vulnerable patients receiving expanded CR and standard CR, respectively at

two‐, five‐ and ten‐year follow‐up.81



Year of follow‐up

Mean N Mean N Difference P‐value


2 5.1 78 4.4 130 ‐0.6 (‐0.8; ‐0.4) 0.00

5 4.7 74 4.3 116 ‐0.4 (‐0.6; ‐0.2) 0.00

10 4.3 68 4.2 105 ‐0.1 (‐0.3 ‐ 0.2) 0.51


2 1.2 78 1.3 130 0.1 (0.1;0.2) 0.01

5 1.3 74 1.3 114 0.0 (‐0.1; 0.1) 0.95

10 1.2 68 1.3 103 0.1 (‐0.0 ;0.2) 0.11


2 3.0 77 2.5 130 ‐0.5 (‐0.7;‐0.4) 0.00

5 2.6 74 2.3 111 ‐0.3 (‐0.4;‐0.1) 0.00

10 2.4 68 2.3 102 ‐0.1 (‐0.3 ;0.1) 0.37

Triglyceride, mmol/L

2 2.0 78 1.5 130 ‐0.4 (‐0.7;‐0.2) 0.00

5 1.8 74 1.6 112 ‐0.2 (‐0.4 ‐0.0) 0.04

10 1.7 68 1.6 103 ‐0.03 (‐0.3;0.2) 0.80

HbA1c, mmol/mol

2 50.7 30 42.3 117 ‐8.4 (‐12.6;‐4.2) 0.00

5 52.1 40 45.7 71 ‐6.4 (‐11.3;‐1.4) 0.01

10 48.9 43 44.9 93 ‐4.0 (‐8.1;‐0.1) 0.06


Page | 35

Blood pressure and smoking status

A total response rate of 79 % was achieved in the questionnaire survey in general practice among

the 379 patients. Not all of the 301 returned questionnaires were completely filled in. The

response rate for questionnaires concerning socially vulnerable patients receiving expanded CR

was 81 % and 77 % for questionnaires concerning socially vulnerable patients receiving standard

CR (P = 0.50). Mean systolic and diastolic BP varied and increased slightly during follow-up

(min. mean BP 131.4/78.3 - max. mean BP 134.1/80.0) in socially vulnerable patients. The

proportion of smokers declined from baseline (min. proportion 0.64 – max. proportion 0.76 –

Table 2) to two-year follow-up (min. proportion 0.46 – max. proportion 0.52). The proportion of

smokers increased slightly from two- to five-year follow-up (min. proportion 0.47 - max.

proportion 0.59) and decreased again during the last five years of follow-up in socially

vulnerable patients (proportion 0.37). No significant differences in mean systolic BP, mean

diastolic BP or the proportion of smokers were seen throughout follow-up when socially

vulnerable patients receiving expanded CR were compared to socially vulnerable patients

receiving standard CR (P > 0.05) (Table 6).81

Table 6. BP and smoking status among socially vulnerable patients receiving expanded CR and standard CR,

respectively at two‐, five‐ and ten‐year follow‐up.81



Year of follow‐up

Mean*

Proportion** N Mean*

Proportion** N

Difference* Ratio**

P‐value

Systolic BP, mm/Hg

2 131.8* 53 131.4* 93 ‐0.4* (‐6.0;5.1) 0.89

5 132.7* 54 132.6* 96 ‐0.1* (‐6.8;6.5) 0.97

10 134.1* 55 132.4* 92 ‐1.7* (‐7.5; 4.2) 0.57

Diastolic BP, mm/Hg

2 79.5* 53 79.0* 93 ‐0.5* (‐3.9 ;2.9) 0.77

5 78.3* 54 79.0* 96 0.7* (‐2.7 ;4.1) 0.68

10 80.0* 55 79.4* 92 ‐0.5* (‐4.2;3.2) 0.78

Current smoker

2 0.52** 33 0.46** 70 0.9** (0.6;1.4) 0.58

5 0.59** 39 0.47** 68 0.8** (0.6;1.2) 0.23

10 0.37** 43 0.37** 75 1.0** (0.6;1.6) 0.98


Page | 36

4.4 Paper III

A full follow-up of yearly survivors at two-, five- and ten-year follow-up was achieved for all

outcomes (Table 7). The proportion of socially vulnerable patients participating in annual

chronic care consultations in general practice increased throughout the ten-year follow-up.

Socially vulnerable patients receiving expanded CR had a significantly higher proportion of

participation in annual chronic care consultations in general practice at two- (ratio 7.3 95% CI

1.0;55.2 P 0.02) and five-year (ratio 2.6 95% CI 1.4;4.6 P 0.00) follow-up but not at ten-year

follow-up (P > 0.05) compared to socially vulnerable patients receiving standard CR (Table 7).82

The mean number of contacts to general practice varied during the ten-year follow-up in socially

vulnerable patients (min. mean number 18.3 – max. mean number 25.0). No significant

differences between expanded CR and standard CR were seen in socially vulnerable patients at

two- and five-year follow-up (P > 0.05). At ten-year follow-up, socially vulnerable patients

receiving expanded CR had significantly fewer contacts to general practice compared to socially

vulnerable patients receiving standard CR (difference -5.8 95% CI -11.0;-0.6 P 0.03) (Table 7).82

No significant differences between expanded CR and standard CR were seen in socially

vulnerable patients concerning all-cause hospitalisation and cardiovascular readmission (P >

0.05) at two-, five- or ten-year follow-up (Table 7).82

Table 7. Participation in annual chronic care consultations in general practice, contacts to general practice, all‐

cause hospitalisations and cardiovascular readmission among socially vulnerable patients receiving expanded CR

and standard CR, respectively at two‐, five‐ and ten‐year follow‐up.82

Socially vulnerable patients Standard CR Expanded CR

Year of follow‐up

Proportion*Mean**

N Proportion*Mean**

N Ratio*

Difference** P‐value

Participation in annual chronic care consultations in general practice

2 0.01* 78 0.09* 128 7.3* (1.0;55.2) 0.02

5 0.15* 75 0.38* 125 2.6* (1.4;4.6) 0.00

10 0.47* 66 0.36* 107 0.8* (0.5;1.1) 0.13

Contacts to general practice

2 19.0** 78 20.3** 128 1.3** (‐3.6;6.1) 0.60

5 21.5** 75 18.3** 125 ‐3.2** (‐8.1;1.8) 0.21

10 25.0** 66 19.2** 107 ‐5.8** (‐11.0;‐0.6) 0.03

All‐cause hospitalizations

2 0.8** 78 0.6** 128 ‐0.2** (‐0.6; 0.3) 0.46

5 0.7** 75 0.5** 125 ‐0.2** (‐0.6; 0.2) 0.30

10 0.5** 66 0.4** 107 ‐0.1** (‐0.3;0.2) 0.81

Cardiac readmissions

2 0.3** 78 0.2** 128 ‐0.1** (‐0.29;0.07) 0.24

5 0.1** 75 0.2** 125 0.1** (‐0.1;0.2) 0.43

10 0.1** 66 0.1** 107 0.0** (‐0.1;0.1) 0.98


Page | 37

4.5 Additional analyses

Additional descriptive analyses were performed to present visual comparisons of all socially

vulnerable patients compared to all non-socially vulnerable patients at follow-up. All patients

regardless of social class had similar survival and event-free probability during the ten-year

follow-up (Figure 2).


Page | 38

Figure 2. Kaplan‐Meier estimates of all‐cause mortality, cardiovascular mortality, non‐fatal recurrent events and

MACE among socially vulnerable patients and non‐socially vulnerable patients during ten‐year follow‐up.

Cardiovascular mortality and major cardiac event analyses only included patients who were not diagnosed with a

non‐fatal recurrent event during the first month after admission.

All‐cause mortality

Survival probability

Socially vulnerable 2000‐2004

Non‐socially vulnerable 2000‐2004

Cardiovascular mortality

Survival probability



Non‐fatal recurrent events

Event‐free probability



MACE

Event‐free probability




Page | 39

The proportion of medication adherence to anti-thrombotics, beta-blockers, statins and ACE

inhibitors were similar among socially vulnerable and non-socially vulnerable patients,

respectively at two-, five- and ten-year follow-up (Figure 3).

Figure 3. Medication adherence among socially vulnerable patients and non‐socially vulnerable patients at two‐,

five‐ and ten‐year follow‐up.


Page | 40

The socially vulnerable patients had a higher median total cholesterol at two- and five-year

follow-up but not at ten-year follow-up. The same tendencies were seen regarding LDL-

cholesterol. The socially vulnerable patients had lower median HDL-cholesterol levels and

higher median triglyceride and HbA1c levels throughout the ten-year follow-up (Figure 4).

Figure 4. Blood tests among socially vulnerable patients and non‐socially vulnerable patients at two‐, five‐ and ten‐

year follow‐up.


Page | 41

The socially vulnerable patients had slightly higher mean systolic BP throughout the ten-year

follow-up. Non-socially vulnerable patients had slightly higher mean diastolic BP at two- and

five-year follow-up but not at ten-year follow-up. The proportion of smokers in the group of

socially vulnerable patients was markedly higher throughout the ten-year follow-up (Figure 5).

Figure 5. BP and smoking status among socially vulnerable patients and non‐socially vulnerable patients at two‐,

five‐ and ten‐year follow‐up. Values are based on data from the questionnaire survey in general practice.


Page | 42

The same patterns in use of health care services were seen regardless of SES during the ten-year

follow-up. However, there was a tendency that the proportion of socially vulnerable patients who

participated in annual chronic care consultations in general practice was a bit lower (Figure 6).

Moreover, there was a tendency that the mean number of contacts to general practice, all-cause

hospitalisation and cardiovascular readmission was a bit higher among socially vulnerable

patients (Figures 6 and 7).

Figure 6. Participation in annual chronic care consultations in general practice and contacts to general practice

among socially vulnerable patients and non‐socially vulnerable patients during ten‐year follow‐up.

Figure 7. All‐cause hospitalisation and cardiovascular readmission among socially vulnerable patients and non‐

socially vulnerable patients during ten‐year follow‐up.

Page | 43

5. Discussion of main findings in

relation to other studies

5. Discussion of main findings in relation to other studies

Page | 44

5.1. Mortality and non-fatal events

At ten-year follow-up, only non-significant differences in all-cause mortality, cardiovascular

mortality, non-fatal recurrent events and MACE were seen when comparing expanded CR with

standard CR among socially vulnerable patients. No major differences in mortality probability

and event-free probability were observed when doing a ten-year descriptive evaluation of

socially vulnerable patients compared to non-socially vulnerable patients.

A ten-year follow-up of a CR intervention is really rare. However, an older study by Hedbäck et

al. from 200196 examined the long-term effect after ten years of CR compared to usual care in 49

patients undergoing coronary artery bypass surgery (CABG). CR consisted of patient education

with focus on risk factor control, outpatient and home-based exercise training, psychosocial

support and follow-up visits once a year for five years after the intervention. It was found that

CR had a significant effect at the ten-year follow-up concerning total number of cardiac events

(P < 0.01) and mean time to first cardiac event (P < 0.05). No significant effect was seen in

mortality. The content of the CR intervention in Hedbäck et al.96 was quite similar to our

intervention but the patients in this study had annual follow-up visits for five years. We

compared standard CR to expanded CR whereas Hedbäck et al.96 compared standard CR to usual

care. The discrepancies between our non-significant long-term outcomes and the significant

outcomes in Hedbäck et al.96 may partly be explained by standard CR in Denmark at the time of

the intervention being of such a quality that significant improvements could be difficult to detect

when adding something extra. Also, Hedbäck et al.96 did not address potential socioeconomic

differences and it is thus not possible to know if the significant outcomes were mainly caused by

only the non-socially vulnerable patients benefitting from the offered CR.96

This PhD dissertation examined the effect of expanded CR compared to standard CR. The effect

of expanded CR in addition to standard CR, standard care or usual care was examined in a RCT

and evaluated in a systematic review. The Swedish RCT by Plüss et al.97 from 2011 conducted a

five-year follow-up of an intervention in patients diagnosed with MI or patients undergoing

CABG. All patients received standard CR consisting of physical training for 60 minutes twice a

week for three months. The physical training was individually designed with low, medium or

high intensity. Standard CR also consisted of one consultation involving the patient, spouse and

a cardiologist and individual counselling by a social worker.


Page | 45

Moreover, all patients were offered to participate in a so-called heart school, which consisted of

a three-hour consultation with a nurse, a physiotherapist, a dietician and a social worker and an

offer to participate in a smoking cessation programme. In addition to standard CR, expanded CR

consisted of a five-day stay at a patient hotel with focus on information and physical training, 20

group stress management sessions for one year and three cooking sessions and diet counselling

with a dietician. At five-year follow-up, patients randomized to expanded CR experienced

significantly fewer non-fatal MIs compared to patients randomized to standard CR (P = 0.047).

No significant differences were seen regarding mortality.97 The follow-up time in Plüss et al.97

was only half as long as in our study; they found a significant effect of their intervention and we

did not, which could be a sign of a declining effect of the intervention over the years. The

intervention in Plüss et al.97 was much more comprehensive and long-lasting than the

intervention in our study and this could explain the observed significant long-term effect in their

study. Similar to Hedbäck et al.96, potential socioeconomic differences between the patients were

not addressed in Plüss et al.97

A systematic review from 201878 evaluated the effectiveness of expanded CR concerning

numerous outcomes including mortality in patients with CHD. Thirty RCTs were included of

which six contained data on mortality. Only one of these studies found a significant effect of

expanded CR compared to standard CR regarding a three-year cumulative survival rate (P =

0.003). The content, duration and follow-up of the expanded CR in the different studies were not

consistent as they were complex interventions which were often individually tailored and in

some cases home-based. Moreover, information about the content of standard CR, standard care

or usual care was not available in many of the studies. Also, potential socioeconomic differences

were not specifically addressed.78 Thus, it was not possible to compare the outcomes in the

review with our results.

Murphy et al.98 examined the effect of a complex intervention with a duration of 18 months

compared to usual care offered to patients with CVD and delivered in general practice in an Irish

RCT from 2015 with six-year follow-up. The intervention consisted partly of tailored practice

plans and education given to GPs concerning medication and health behaviour and partly of

tailored patient plans. The patients were invited to the GP three times a year and played an active

role in defining the aim of their secondary prevention. At six-year follow-up no significant

differences were seen in all-cause and cardiovascular mortality when comparing the intervention

with usual care.98


Page | 46

The non-significant effect of the intervention in Murphy et al.98 is comparable to our study. Even

though Murphy et al.98 did not directly focus on socially vulnerable patients, the intervention was

tailored to fit the individual patient. However, the intervention was of a much longer duration

than ours and performed in a primary care setting in contrast to our hospital-based intervention.

A study with focus on socially vulnerable patients was the German RCT with three-year follow-

up by Mayer-Berger et al.76 from 2014. This study population consisted of 600 patients

diagnosed with CAD. The 271 patients in the intervention group and the 329 patients in the usual

care group consisted almost only of men (87.5 % vs. 90.8 %) with a low educational level

defined as ten years of education or less (88.4 % vs. 90 %). All patients received three weeks in-

patient CR consisting of psychological and educational support and 90-180 minutes of physical

training / day. The intervention consisted of one extra day of CR six months after in-patient CR.

The content of this day was group sessions and individual counselling. Moreover, the patients

received regular telephone follow-up throughout the three-year follow-up. The patients in the

intervention group who almost all had a low educational level had experienced significantly

fewer cardiac events (P < 0.05) at three-year follow-up compared to the patients, also with a low

educational level, who received usual care. No significant differences were seen regarding

mortality.76 The study population in this RCT was much larger than our study population making

it more likely to detect any possible significant differences. The intervention was very similar to

the intervention for our patients, but the contact to the patients in the study by Mayer-Berger et

al.76 was maintained for a longer time. It is worth mentioning that the intervention in Mayer-

Berger et al.76 was not tailored specifically to socially vulnerable patients; by chance, this study

population mainly had a low educational level.

A Swedish RCT from 2016 by Fors et al.75 examined the effect of a person-centred care (PCC)

intervention in patients diagnosed with ACS. The PCC intervention consisted of a 'shared

decision making' process in the creation and implementation of the patients' individual health

plans. The plans took the patients' resources and challenges including social support into

account. The patients had follow-up visits at an out-patient clinic four and eight weeks after

discharge. At six-months follow-up, patients without post-secondary education in the

intervention group showed significant improvements compared to patients without post-

secondary education who received usual care when measured with a composite score including

mortality (P = 0.041).75 The content of the intervention had features similar to our intervention.


Page | 47

However, the follow-up time was very short and the composite score made it difficult to assess

the effect of the intervention on mortality only.

5.2 Medication adherence and biological and lifestyle risk factors

No persistent long-term differences between expanded CR and standard CR in socially

vulnerable patients were found at two-, five- or ten-year follow-up. Socially vulnerable patients

who received expanded CR had a significantly higher adherence to statins at two-year follow-up

and a significantly lower adherence to beta-blockers at two- and five-year follow-up. Levels of

total cholesterol, LDL-cholesterol, triglyceride and HbA1c were significantly lower at two- and

five-year follow-up, and levels of HDL-cholesterol were significantly higher at two-year follow-

up among socially vulnerable patients who received expanded CR. No major differences in the

proportion of medication adherence were observed when doing a ten-year descriptive evaluation

of socially vulnerable patients compared to non-socially vulnerable patients. In general, the

socially vulnerable patients had slightly higher median total cholesterol, LDL-cholesterol,

triglyceride, HbA1c and BP levels and slightly lower HDL-cholesterol levels. The proportion of

socially vulnerable smokers was distinctly higher than the proportion of non-socially vulnerable

smokers.

An Italian RCT from 2012 by Sturchio et al.51 examined the effect of an outpatient coronary

artery risk management programme (CARIMAP) in 168 patients with CAD. All patients

received intensive CR. After this, the control group was referred to usual care at their GP or

cardiologist. The patients randomized to the intervention were offered the CARIMAP. The

CARIMAP began four months after discharge and lasted for five months until follow-up at nine

months. The CARIMAP consisted of individual care plans and tailored interventions drafted in

cooperation with a team of cardiologists, nurses and physiotherapists, as well as medical

treatment optimisation, increased physical activity and health care education. At nine-months

follow-up, patients randomized to the CARIMAP showed significantly higher medication

adherence to beta-blockers, statins and ACE inhibitors and had significantly lower levels of total

cholesterol, LDL cholesterol, triglyceride, systolic BP and significantly fewer were smokers

compared to patients randomized to usual care (all P < 0.05). The content and the

implementation of the CARIMAP intervention was similar to our intervention even though the

CARIMAP was offered in phase III CR and not in phase II CR as in our study.


Page | 48

The follow-up time was short and follow-up was performed at the end of the intervention. The

significant effect could therefore be the result of patients not yet being on their own managing

the different CVD risk factors.51

An American RCT by Ho et al.99 from 2014 examined an intervention to improve medication

adherence compared to usual care in 253 military veterans admitted with ACS. The intervention

lasted for one year after discharge where follow-up was conducted. At follow-up, patients in the

intervention group had a significantly better composite medication adherence compared to

patients in the usual care group (89.3 % vs. 73.9 %, P = 0.003). No significant differences were

found in mean LDL-cholesterol (P = 0.37) or mean systolic or diastolic BP (both P = 0.50). Ho

et al.99 did not perform a socially differentiated CR intervention as we did but they made an

intervention in a study population of military veterans who could be suspected to be socially

vulnerable. The intervention lasted much longer than our intervention and follow-up was

conducted at the end of the intervention. It is uncertain what the content of usual care was and

thus it is not possible to clarify if the intervention and the usual care offered to the veterans can

be compared to our expanded and standard CR.

The association between SES and persistent medication adherence in patients with CAD were

examined by King et al.100 in a Welsh register- and population-based cohort study from 2018.

Patients with CAD who received primary or secondary prevention were followed for six years

from 2004 to 2010. SES was defined by a composite deprivation score consisting of the patients'

income, employment status, education, housing, health and geographical access to services. The

most deprived patients were compared to the least deprived patients. Persistence was measured

from time of initiation to the last prescription. No significant socioeconomic differences were

found in persistent medication adherence at follow-up (P > 0.05). The comparison of patients

according to deprivation score in King et al.100 could be compared to our additional analyses

where we compared socially vulnerable patients to non-socially vulnerable patients. The findings

concerning medication adherence are similar when the six-year follow-up in King et al.100 is

compared to our five-year follow-up, even though King et al.100 have used a more

comprehensive definition of SES than we did in our study.

A Norwegian cross-sectional study by Sverre et al.101 from 2017 examined the association

between a number of variables and achievement of BP targets (140/90 mm/Hg) in 1012 patients

hospitalised with MI or coronary revisualisation procedure and a part of the Norwegian Coronary

Prevention study (NOR-COR).


Page | 49

The study was conducted at follow-up and found no significant associations between the

achievement of BP targets and living alone (17.3% <140/90 mm/Hg vs. 18.4% >140/90 mm/Hg,

P > 0.05) or having a low educational level (67.9% <140/90 mm/Hg vs. 69.8 % >140/90

mm/Hg, P > 0.05).101 Another cross-sectional study by Bruthans et al.17 from 2016 used data

from the EUROASPIRE IV study to examine the association between educational level and

different risk factors in European patients (N = 7937) diagnosed with CHD. Education was

defined as primary, secondary and higher educational level. Control of risk factors was defined

by the Joint European Societies four and five guidelines. When using the patients in the higher

educational level as reference group, it was found that men with a primary educational level to a

significantly lesser extent reached the targets for smoking (OR 1.55, 95% CI: 1.20;1.99) and

HDL-cholesterol (OR 1.33, 95% CI: 1.09;1.62) and women with primary educational level to a

significantly lesser extent reached the targets for BP (OR 2.06, 95% CI: 1.46;2.90) and HDL-

cholesterol (OR 1.60, 95% CI: 1.13;2.26).17 The same association between increasing

educational level and successful smoking cessation in the EUROASPIRE IV study population

was found by Snaterse et al. in 2018.16 The percentage of successful quitters in the group of

patients with primary school education or less was 46 % compared to the 53 % of successful

quitters in the group of patients with a university education (P = 0.04).16 The detection of the

inverse association between SES and CVD risk factors is well established in these three studies

and agrees with our findings regarding smoking. However, the studies are cross-sectional and

thus there is no intervention and no delayed follow-up time. Therefore, it is not possible to

comment on potential causal relationships or development over time.

A systematic review of observational studies by Schneider et al. from 2018102 evaluated

medication non-adherence in patients with CVD. Data from 31 studies and 27,441 participants

were included in the review. It was found that medication non-adherence decreased if the

patients had access to insurance or another programme that could reduce their medical costs.102

Another systematic review of qualitative studies by Rashid et al. from 2014103 evaluated the

factors that promote medication adherence persistence in patients with CAD. Data from 17

studies and 391 patients were included. It was found that a good relationship between the patient

and the prescribing medical doctor was of critical importance. Another factor that promoted

medication adherence persistence was support from the patients' relatives whereas medication

costs could be a barrier.103 Our outcomes regarding medication adherence showed no descriptive

differences between socially vulnerable patients and non-socially vulnerable patients. This is in

line with the hypothesis in Schneider et al.102 and Rashid et al.103 that medication adherence


Page | 50

increased if the financial burden of medication purchase was reduced. We found that the quality

of the long-term secondary CVD prevention performed by GPs was high, supporting the findings

by Rashid et al.103regarding the importance of a good relationship between the patients and their

prescribing medical doctor, which most often will be the GPs in Denmark.

5.3 Use of health care services

Socially vulnerable patients who received expanded CR had a significantly higher proportion of

participation in annual chronic care consultations in general practice at two- and five-year

follow-up compared to socially vulnerable patients who received standard CR. No persistent

significant differences were seen concerning contacts to general practice, all-cause

hospitalisations or cardiovascular readmission. No major differences in the use of health care

services were observed when doing a ten-year descriptive evaluation of socially vulnerable

patients compared to non-socially vulnerable patients and any differences levelled off over the

years.

In Plüss et al.97,which has been reviewed previously, the effect of expanded CR was examined in

a RCT. At five-year follow-up patients randomized to expanded CR had a significantly lower

total number of hospitalisations (P < 0.01) and number of days at hospital (P = 0.02).97 In

Hedbäck et al.97, which was also reviewed previously, the long-term effect of CR was evaluated.

At ten-year follow-up patients receiving CR had significantly fewer readmissions and length of

admissions compared to patients who received usual care (P < 0.01).96 In comparison, Murphy et

al.98 did not find any significant differences in all-cause and cardiovascular readmissions when

comparing an individual complex intervention to usual care at six-year follow-up.98 Although

Murphy et al. performed their complex intervention in primary care and did not focus

specifically on socially vulnerable patients, the study is more similar to our study than Hedbäck

et al.96 and Plüss et al.97, which could also be indicated by the similar non-significant outcomes

in Murphy et al.98 and our study. This highlights the difficulties in developing and performing

complex and socially differentiated significant interventions.

The association between socioeconomic position and the use of health care contacts in patients

diagnosed with MI was examined in a Danish population-based cohort study by Nielsen et al.73

from 2015. The study included all patients admitted with MI in 2009 in Central Denmark Region

(N = 908). During the first six months after being admitted with MI, patients with less than ten


Page | 51

years of education (IRR 0.84 95% CI: 0.72;0.98) and patients who lived alone (IRR 0.87 95%

CI: 0.77;0.99) had significantly lower hospital contact rates. During the first two years after

being admitted with MI, the number of GP contacts was not significantly different between

patients with short (IRR 0.96 95% CI: 0.84;1.08) and long education or between patients who

did or did not (IRR 0.95 95% CI: 0.87;1.04) live with a partner.73 The study by Nielsen et al.73

was conducted in the same region in Denmark as our study and the study population was more

than twice as big as ours. Similar outcomes were observed when comparing the results of

Nielsen et al.73 to our additional analyses, indicating that there are no major differences in the use

of health care services in Denmark regardless of SES.

In an American prospective register-based study by Khawaja from 2012104, patients receiving

percutaneous coronary intervention (PCI) between 1998 and 2008 were identified (N = 15,498).

The association between different variables and 30-day readmission was examined. Patients with

less than a high-school education had significantly higher 30-day readmission rates (HR 1.35

95% CI: 1.17;1.55).104 Murphy et al.13 from 2008 examined the association between cohabitation

status and 30-day hospital readmission in patients undergoing CABG and found that patients

living alone were significantly more likely to be readmitted during the first month after surgery

(OR 3.42 95% CI: 1.38;8.48 P = 0.008).13 The follow-up time in Khawaja et al.104 and Murphy

et al.13 was very short compared to our long-term follow-up. This could partly explain their

significant results which are in contrast to our descriptive observations in the additional analyses.

Page | 52

6. Discussion of methods


Page | 53

6.1 Internal validity

6.1.1 Selection bias Selection bias is a systematic over- or underestimation of associations due to selection issues.105

Risk of selection bias may occur in comparative studies at three different stages: when patients

are excluded, when patients do not participate and when patients are lost to follow-up.106 If the

patients who are not included in the analyses due to exclusion, non-participation or loss to

follow-up up are different than the patients who are included in the analyses, there is a risk of

biased results. If exclusion, non-participation or loss to follow-up is only associated with

exposure or outcome, then the risk of bias is non-differentiated and will lead to the possible

incorrect assumption of no correlation (Type 2 error). If exclusion, non-participation or loss to

follow-up is associated with both exposure and outcome, the risk of bias is differentiated and can

lead to the possible incorrect assumption of a correlation or no correlation (Type 1 and Type 2

error).105-107

As seen in Figure 1, the proportion of patients excluded from 2000 to 2002 is 17 % and it is 11

% from 2002 to 2004. As the study is designed as a prospective cohort study, the exclusion

happened before exposure and follow-up. Thus, it is estimated that there is no risk of

differentiated selection bias. Figure 1 shows that non-participation is also higher from 2000 to

2002 (35 %) than from 2002 to 2004 (19 %). At this time the exposure but not the outcome was

well-known to the patients. Risk of selection bias pointing towards no correlation is therefore

likely present. For example, if the increased number of patients who participated from 2002 to

2004 were more socially vulnerable than the patients who participated from 2000 to 2002, the

possible effect of the intervention could have been lost. At follow-up, registers provided a

complete follow-up except regarding blood tests. It is not realistic to assume that the collection

of blood tests should be different between the compared groups neither according to exposure

nor according to outcome. At follow-up, the questionnaire survey in general practice achieved a

response rate of 79 %. The questionnaire was filled in by the patients' GPs. Also, here it does not

appear to be realistic to assume that there should be a systematic difference in which GPs who

responded or not according to the specific patient's exposure and outcome. It could be considered

to apply statistical imputation in order to address the issues concerning missing blood tests and

missing questionnaire data. The full follow-up when using registry-based data does only include

the patients who actually survived until follow-up. It could be considered to apply statistical


Page | 54

competing risk or pseudo values analyses to address this issue. However, no significant

differences in all-cause mortality was found in Paper I.80

6.1.2 Information bias Information bias is a systematic over- or underestimation of estimates due to information and

misclassification issues. Non-differentiated misclassification is where the degree of

misclassification in the groups compared is the same, is frequent and can result in possible

incorrect assumptions of no differences between the groups. Differentiated misclassification is

rare but will increase the risk of information bias.105 Baseline questionnaire data regarding

educational level, cohabitation status and smoking status was self-reported by the patients. It is

not assumed that the information quality between the groups was differentiated.

Misclassification in registry-based data would rarely occur and will most often be non-

differentiated. However, misclassification in the definition of cardiovascular mortality must be

considered. In Denmark, the validity of mortality statistics versus clinical records has been

evaluated and the sensitivity was found to be high. The lack of accuracy in IHD mortality is,

however, still an issue.89 It is not assumed that the potential misclassification of cardiovascular

mortality should be differentiated between the groups compared.

6.1.3 Confounding Confounding is defined as the mixing of causes and thereby misinterpretation of data. Three

conditions must be fulfilled for an exposure to be considered a confounder (Figure 8)105:

1. The exposure must be an independent risk factor for the given outcome105

2. The exposure must be unevenly divided between the two groups compared105

3. The exposure must not be part of the causal chain between exposure and outcome105

Figure 8.

One of t

when pa

identica

interven

unethica

When u

statistic

Paper II

actually

socially

socially

admitted

0.00), L

6.1.4 PThe pow

study co

2 errors

when a

significa

the orig

when pl

interven

Confounding

the ways to

articipants a

al at baselin

ntion.105 Wh

al to random

using the pro

al regressio

I81 or Paper

y an adjustm

y vulnerable

y vulnerable

d from 2002

LDL-cholest

Power wer of a stu

oncludes tha

s can be redu

study concl

ance level o

ginal socially

lanning the

ntion could

2

Exposu

interaction m

avoid risk

are randoml

e and any d

hen designin

mize the soc

ospective co

on analyses.

III82. After

ment for a co

e. Table 2 sh

e patients ad

2 to 2004 th

terol (P = 0

dy is 100 %

at there is n

uced by inc

ludes that th

of 0.05, the

y differentia

project. Ho

be explaine

2

ure

Co

model. From Ep

of confound

ly divided in

differences i

ng the origin

cially vulner

ohort study 105 Logistic

finishing P

omponent th

hows baselin

dmitted from

hey have ide

.02), triglyc

% minus the

no associatio

reasing the

here is an as

risk of Typ

ated CR pro

owever, it ca

ed by a too s

3

onfounde

Page | 55

Epidemiologi o

ding is to ap

nto interven

in outcomes

nal socially

rable patien

design, risk

c regression

Paper I, it w

hat was a pa

ne data for

m 2000 to 2

entical P-va

ceride (P =

risk of Typ

on when the

number of

ssociation w

pe 1 errors is

oject are poi

annot be rul

small study

1

Outc

er

og evidens105

pply the RC

ntion and co

s between th

y differentia

nts to either

k of confoun

analysis wa

as consider

art of the ca

the study po

002 and the

alues > 0.05

0.02) and st

pe 2 error. A

ere is in fact

study partic

when there i

s 5 %. Refle

intless as th

led out that

population

come

CT design. It

ontrol group

he groups de

ted CR proj

expanded o

nding can b

as applied in

ed that the a

ategorisation

opulation. W

e socially vu

5, except for

tatins prescr

A Type 2 err

t an associa

cipants. A T

is no associa

ections conc

hey should h

the non-sig

.105

6. Discussio

It is assumed

ps they coul

depend only

ject, it was

or standard

be addressed

n Paper I80

age adjustm

n as socially

When comp

ulnerable pa

r total chole

ribed (P = 0

ror occurs w

ation. The ri

Type 1 error

ation. When

cerning the

have been a

gnificant res

on of methods

d that

ld be

on the

considered

CR.25

d by using

but not in

ment was

y or non-

paring the

atients

esterol (P =

0.00).

when a

isk of Type

r occurs

n using a

power of

addressed

sults of the

s


Page | 56

6.1.5 Study population The study population was categorised as socially or non-socially vulnerable defined by

participants’ educational level, their cohabitation status and age. A Danish qualitative study

from 2016 performed at five hospitals and in six municipalities explored health professionals’

and patients’ experiences regarding socially differentiated CR. It is acknowledged by health

professionals that most research in socially differentiated CR is based on a division of the

patients according to educational level and social network. However, it was found that health

professionals in practice base CR on the individual patient’s needs and situation of life instead of

predefined differentiation criteria.108 Perhaps the health professionals have a point? Was the

categorisation applied in the original socially differentiated CR project too rigid and simplified?


Page | 57

Figure 9. Model for mechanisms causing social inequality in health and the associated interventions. From Challenging inequalities in health. From ethics to action.70,109

Figure 9 can be utilised when trying to understand the complexity of identifying what causes

social inequality in health and which interventions may reduce it. There are four mechanisms

that play a part in creating social inequality and for every mechanism there are policy entry

points for applying interventions.70,109


Page | 58

1) The first mechanism is social stratification, which is connected to genetics, gender, age,

ethnicity, education and health. The social stratification affects the social position that an

individual holds in society. Policies should focus on influencing this social stratification for

example by promoting policies that aim at decreasing gender differences.70,109

2) The second mechanism is differential exposure and is connected to the varying risk factors

that individuals are exposed to depending on their work and living conditions and their health

behaviour. Policies should focus on decreasing the specific exposures that will increase social

inequality in health. An example could be to offer free-of-charge smoking cessation courses for

all citizens.70,109

3) The third mechanism is differential vulnerability and is connected to the accumulation of

exposures which also accumulate risk of disease. Policies should focus on decreasing the

vulnerability for example by targeting interacting exposures instead of single exposures.70,109

4) The fourth mechanism is differential consequences and is connected to the fact that disease

can affect functioning and quality of life, which can have social consequences and lead to social

stratification as it can change the social position. Policies should focus on preventing unequal

consequences for example by increasing the allocation of resources towards rehabilitation.70,109

How to define and decrease social inequality in health is thus extremely complex. The

complexity is further highlighted by the fact that CR programmes must not only be based on the

individual patient’s needs and wishes but should also ensure that all patients receive the same

coherent and high quality treatment.55

6.1.6 Study design A limitation of the study is that patients from non-parallel time periods are compared. This is a

challenge because it can blur whether observed outcome differences are associated with the

intervention or if they are just a result of updated CR guidelines and other factors that change

over time. This could be suspected regarding medication adherence (decrease in the adherence to

beta-blockers and increase in the adherence to statins), blood tests (lower levels of cholesterol

and HbA1c) and participation in annual chronic care consultations in general practice (increasing

participation) when expanded CR to socially vulnerable patients from 2002 to 2004 was


Page | 59

compared to standard CR to socially vulnerable patients from 2000 to 2002 (Tables 4, 5 and 7).

Thus, all analyses have been performed in the non-socially vulnerable patients from 2002 to

2004 and compared to the non-socially vulnerable patients from 2000 to 2002, who all received

standard CR to evaluate if observed outcome differences among socially vulnerable patients

were also seen in the non-socially vulnerable patients. The analyses are available in Papers I80,

II81 and III82 and show that the outcome differences in socially vulnerable patients are also

present among the non-socially vulnerable patients. This implies that the observed changes are

caused by time and not the expanded CR intervention.

The additional analyses in which all socially vulnerable patients from 2000 to 2004 are

compared to all non-socially vulnerable patients from 2000 to 2004 were performed to create an

overview of the social inequality in long-term secondary CVD prevention in Denmark. The

analyses are shown solely as descriptive visualisations, as statistical analyses would violate some

of the assumptions that must be met when working with binary and numeric data. Two of these

assumptions are that the probability (binary data) should be the same for all patients and that the

observations should have the same mean and variance (numeric data) in all patients.110 When

comparing socially vulnerable patients to non-socially vulnerable patients these assumptions are

not met as the two groups most likely are different at baseline. However, the descriptive analyses

provide a picture of the level of social inequality in Denmark today.

6.1.7 Intervention The expanded CR intervention was based partly on CR guidelines applicable at the time84-86 and

partly on clinical knowledge obtained from the health professionals at Aarhus County Hospital /

Aarhus University Hospital. The clinical knowledge was based on the health professionals'

experiences from their daily work with CR. It included an assessment of which patients were

considered to be socially vulnerable. Moreover, the health professionals contributed when

designing the content of the expanded CR intervention regarding the challenges they observed

patients were struggling with. The expanded CR intervention was designed as a so-called

'realistic intervention'. The ulterior motive was that it should be an affordable intervention that

should be applicable for implementation in the daily clinical work with CR.

When evaluating the expanded CR intervention, the bio-psychosocial model of illness could be

used (Figure 10). The model was introduced in 1977 as a contrast to the biomedical model. The


Page | 60

biomedical model focuses entirely on the disease and not the patients who are ill. The bio-

psychosocial model is a comprehensive model that acknowledges the importance of

psychological and social factors when trying to cure a disease and find the right approach to

decrease disability and increase the patients' functioning.111,112

Figure 10. The bio‐psychosocial model of illness. From Rehabilitation – a new approach. Part two: the underlying

theories.112

The model consists of four components concerning the patient (pathology, activities, social

participation and impairment) and four contextual components (personal, social, physical and

temporal contexts). To ensure that the rehabilitation process is as effective as possible, all

components must be identified and addressed in the individual patient.111,112

When re

CR inte

compon

As impl

vulnerab

publishe

2015.113

compon

pilot tes

and affe

Figure 11interventintervent

DeveloIdentifyIdentifyModellin

eviewing th

ervention ap

nents regard

lied in Figu

ble patients

ed their rece3 A complex

nents.114 The

sting, imple

ect each oth

1. Model of ketions. From Prtioner i medici

opment ing the evidening or developng process an

he social vul

pplied in the

ding the pati

ure 9 and Fig

s will most o

ent guidelin

x interventi

e MRC emp

mentation a

her.113-115

ey elements inrocess evaluatinsk forskning

nce base ping theory nd outcomes

lnerability,

e socially di

ient and the

gure 10, a c

often be ver

nes for deve

on is define

phasises fou

and evaluati

n the developmtion of comple.113‐115

FeasibiliTesting prEstimatingDetermini

ImplemeDisseminatSurveillancLong‐term

Page | 61

categorisati

ifferentiated

e context we

omprehensi

ry complex.

elopment an

ed as an inte

ur areas of i

ion (Figure

ment, pilot teex intervention

ity and pilorocedures g recruitmenting sample siz

entation tion ce and monito

m follow‐up

ion criteria

d CR projec

ere addresse

ive CR inte

. The Medic

nd evaluatio

ervention w

importance:

11). The fo

sting, evaluatns: Medical re

ting

t and retentioze

oring

and the con

t, it is estim

ed to a great

rvention de

cal Research

n of comple

with several i

developme

our areas are

ion and impleesearch counc

n

EvaluationAssessing efUnderstandiAssessing co

6. Discussio

ntent of the

mated that al

ter of lesser

esigned to so

h Council (M

ex intervent

interacting

ent, feasibili

e mutually d

ementation ofcil guideline an

n ffectiveness ing change prost effectivene

on of methods

expanded

ll of the

r extent.

ocially

MRC)

tions in

ity- and

dependent

f complex nd Komplekse

rocess ess

s

e


Page | 62

When designing the expanded CR intervention back in 1999, the overall aim of the project was

to produce evidence on whether it was possible to implement socially differentiated CR in

Denmark. The project was to a large extent explorative and based on clinical knowledge. If

designing the intervention today, the MRC guidelines could have provided a more standardised

process. However, it must be questioned if an intervention based on the individual patient can be

standardised? And if not, will complex interventions and evidence-based rehabilitation science

ever succeed to become an integrated part of the established medical science where interventions

are performed using blue and red pills?

6.2 External validity

External validity is often referred to as generalizability and describes the extent to which the

results of a study are valid if they were applied in another study population in another time

period. In other words, are the results in this PhD dissertation solely obtainable in the applied

study population with follow-up during the noughties or could they be generalised to a wider

population in the future?105-107 The Danish health care system provides free health care access for

all citizens funded through taxes.87 Thus, one must be careful to generalise the results to

countries where health care access is obtained through self-funded health insurances. However,

assessment of the external validity is not definitive. It relies to a large extent on professional

knowledge and not least on common sense.105-107

Page | 63

7. Conclusions

7. Conclusions

Page | 64

After having investigated the long-term effect of the expanded CR intervention compared to

standard CR in socially vulnerable patients admitted at Aarhus University Hospital from 2000 to

2004 and diagnosed with a first episode MI, the main conclusions of this PhD dissertation are:

No significant differences in all-cause or cardiovascular mortality, non-fatal recurrent events

or MACE were seen at ten-year follow-up.

No persistent significant differences in medication adherence, biological or lifestyle risk

factors were seen at two-, five- or ten-year follow-up. The observed significant differences in

medication adherence and the significant differences in blood tests were suspected to be

caused by time rather than the intervention. The mean levels of lipids, HbA1c and BP in

socially vulnerable patients were found to be in accordance with guidelines.

No persistent significant differences in the use of health care services in general practice and

at hospital were seen at two-, five- or ten-year follow-up. The observed significant

differences in participation in annual chronic care consultations in general practice were

suspected to be caused by time rather than the intervention.

When performing additional descriptive analyses of all the socially vulnerable patients compared

to all the non-socially vulnerable patients, the supplementary conclusions of this PhD

dissertation are:

No major visual differences were seen during the ten-year follow-up except for smoking. The

proportion of socially vulnerable smokers was notable when compared to the proportion of

non-socially vulnerable smokers. The socially vulnerable patients showed slightly higher

median blood test levels and mean BP levels; however, all within the ranges recommended in

guidelines.

Page | 65

8. Perspectives

8.Perspectives

Page | 66

Even though the burden of CVD mortality has decreased substantially during the last decades,

the European Society of Cardiology calls for action in an official consensus document39

concerning the continuing challenges in secondary prevention. The document identifies the gaps

in secondary prevention and proposes possible solutions. The gaps are related to the patients, the

health care providers and the health care system. Evidence shows that many patients do not

understand their disease and they feel that they have very little control over their lives. Also, they

do not adhere optimally to the official guidelines concerning lifestyle and medication. It is

questioned if the health care providers including cardiologists, GPs and other health care

professionals have in depth knowledge about secondary prevention to patients with CVD.

Moreover, it is questioned if knowledge is shared between the hospital specialists, GPs and

health care professionals in the municipalities. In many European countries, the health care

systems do still not refer all eligible patients with CVD to a secondary prevention programme.

Also, the lack of prevention centres is still a matter of concern. The possible solutions are

addressed using the division into the three phases applied in CR. There is a need for creating

awareness of how important the acute interventions in phase I are. They should be given as much

attention as the initial medical and invasive acute care. New innovative strategies in the early

outpatient prevention in phase II are needed to improve adherence. Finally, there is a need to

rethink the long-term prevention in phase III, to help as many patients with CVD as possible

with the limited resources available in the health care system.39

The Danish Health Authority also proposes that the interdisciplinary and cross-sectional

secondary prevention offered to Danish patients with IHD should be improved by applying both

more consistent high-quality services and by offering individually tailored patient care. It is

recommended that future CR should primarily be handled by the municipalities.22. However, a

new study by Egholm et al.116 examined the adherence to national clinical CR guidelines in

Danish hospitals and municipalities and found that there is a difference in content and quality

across settings between the offered CR at hospitals and in municipalities.116

We must emphasise the importance in strengthening the role even more of the GPs as the key

persons in the long-term secondary CVD prevention and to provide the relevant further

education for the GPs to update them on the newest in depth knowledge on secondary CVD

prevention. All of this must happen in close cooperation with the municipalities to prepare them

for their constantly growing responsibility concerning secondary CVD prevention.

8.Perspectives

Page | 67

When trying to comprehend the thinking behind and the initiatives launched in the Danish health

care system, it is important to understand the reality of the healthcare system being a part of a

non-static constantly changing political framework. The tax-funded free health care is organised

by politicians who are elected for a four-year period. In October 2018, the Danish Prime minister

announced an upcoming health care reform. At the time of writing, the content is not yet publicly

known. The government has published a new report stating that the Danes live longer and are

therefore more exposed to chronic diseases. The government's aim is to offer more high-quality

care close to the patients, implying that the health care reform will focus even more on the shift

from hospital-based to municipality-based care of chronic patients.117,118 The government has

also published a report that focuses on creating a better coherence between the different sectors

in the public system. The aim is to help the most socially vulnerable citizens who are often faced

with multiple challenges regarding health, social matters and employment issues and who often

get lost in the silo-divided public social and healthcare systems.119

The improvements in the field of cardiology and CR during the last half century are spectacular.

The results in this PhD dissertation show that social equality in health among the patients in the

study population is high regardless of SES. The PhD dissertation holds no information about

non-participating patients, and the promising results regarding social equality could partly be due

the most socially vulnerable patients not participating. Also, social equality was not reached

regarding smoking and the work must continue to try to improve the socially vulnerable patients'

health behaviour. Experiences from studies in England show that citizens with social needs and

mental health problems benefit from interventions with 'personal budgets' to increase their

independence.120-122 Perhaps these experiences could be used as inspiration when working with

socially differentiated and individually tailored CR in Denmark. Maybe it is time to think

different and alternative thoughts and focus on value for money instead of standardised offers?

To achieve the highest possible motivation in the socially vulnerable patients to increase the

likelihood of reducing social inequality in health, focus must be on empowering the patients and

giving them back the feeling of control over their lives. One thing is certain: The future

secondary CVD prevention must continue to address social inequality in health to improve the

health of all patients regardless of SES.

"It is common sense to take a method and try it. If it fails, admit it frankly and try another. But

above all, try something!"

Franklin D. Roosevelt

Page | 68

9. English summary

9.English summary

Page | 69

Background

CVD is still one of the leading causes of mortality and morbidity in Europe including Denmark

although notable improvements have been reached within cardiology. Cardiac rehabilitation

(CR) has a significant effect on modifiable risk factors in the secondary CVD prevention. The

general practitioner (GP) plays a key role in long-term CR. An inverse relationship is seen

between socioeconomic status (SES) and the incidence and mortality of CVD. One reason for

this is that patients are less likely to make behavioural changes if they have low SES. This social

inequality in health was addressed in a previous Danish study that examined the effect of a

socially differentiated project which offered socially vulnerable patients admitted with first

episode myocardial infarction (MI) an expanded CR intervention in addition to standard CR. The

study found a significant difference in modifiable risk factors at one year follow-up when the

intervention group was compared to a control group which consisted of socially vulnerable

patients admitted with first episode MI who received standard CR.

Aim

The aim of the PhD dissertation was to examine the long-term effect of the socially differentiated

project at two-, five- and ten-year follow-up in relation to mortality and morbidity, adherence to

risk factor management and use of health care services. Socially vulnerable patients admitted

with first episode MI who received expanded CR in addition to standard CR were compared to

socially vulnerable patients admitted with first episode MI who received standard CR.

Methods

All patients < 70 years admitted with first episode MI at Aarhus University Hospital, Denmark

from 2000 to 2004, who participated in CR and who gave informed consent were enrolled in the

long-term follow-up. Social vulnerability was defined according to educational level,

cohabitation status and age. The number of included patients was 379. Of these, 208 were

categorised as socially vulnerable of which 131 received expanded CR. The remaining 171

patients were categorised as non-socially vulnerable. The expanded CR intervention consisted of

more consultations, tailored rehabilitation plans, improved transition from hospital-based CR to

long-term CR in general practice and a telephone follow-up. Data were obtained from registers

and a questionnaire sent to the patients' GPs.

9.English summary

Page | 70

Results

Ten-years after baseline, no significant differences were seen between socially vulnerable

patients who received expanded CR and socially vulnerable patients who received standard CR

when they were compared in relation to all-cause mortality, cardiovascular mortality, non-fatal

recurrent events and major cardiovascular events. No long-lasting significant differences were

found during the ten-year follow-up on adherence to risk factor management when expanded CR

was compared to standard CR in socially vulnerable patients. At two- and five-year follow-up

the group which received expanded CR had a significantly lower medication adherence to beta-

blockers and at two-year follow-up a significantly higher medication adherence to statins. At

two- and five-year follow-up, levels of total cholesterol, LDL-cholesterol, triglyceride and

HbA1c in the group which received expanded CR were significantly lower and at two-year

follow-up, levels of HDL-cholesterol were significantly higher. A significant difference in the

use of health care services was found in the proportion of participation in annual chronic care

consultations in general practice which was significantly higher among socially vulnerable

patients who received expanded CR at two- and five-year follow-up when compared to socially

vulnerable patients who received standard CR.

In addition to the analyses based on the overall aim of the PhD dissertation, further descriptive

analyses comparing all socially vulnerable patients to all non-socially vulnerable patients were

performed. During the ten-year follow-up on mortality, morbidity, adherence to risk factor

management and use of health care services, no major differences between the patients were

found regardless of SES except when it came to smoking.

Conclusion and perspectives

This dissertation demonstrated that it was possible to reduce social inequality in health in the

long-term CR performed in primary care including general practice during the ten-year follow-

up of the study population. However, the results regarding especially the socially vulnerable

patients' smoking status indicate that there is still some way to go before socially equality in

health behaviour is reached. No persistent significant long-term effect was found when socially

vulnerable patients who received expanded CR were compared to socially vulnerable patients

who received standard CR. It is unknown if this could partly be explained by the high quality of

the standard CR offered. Alternative thinking concerning future socially differentiated CR is

needed and focus must be on empowering and motivating patients. The struggle to eradicate

social inequality in health must and should continue.

Page | 71

10. Dansk resumé

10. Dansk resumé

Page | 72

Baggrund

Hjertekarsygdom er fortsat en af de førende årsager til dødelighed og sygelighed i Europa,

herunder i Danmark, på trods af, at der er opnået markante forbedringer indenfor kardiologien.

Hjerterehabilitering (HR) har en signifikant effekt på modificerbare risikofaktorer i den

sekundære forebyggelse af hjertekarsygdom. Den praktiserende læge spiller en afgørende rolle i

den sene, livslange HR. Der ses et omvendt forhold mellem socioøkonomisk status og incidens

samt dødelighed som følge af hjertekarsygdom. En årsag til dette er, at patienter med lav

socioøkonomisk status er mindre tilbøjelige til at foretage livsstilsændringer. Et tidligere dansk

studie har undersøgt social ulighed i sundhed i et socialt differentieret projekt, hvori socialt

udsatte patienter indlagt med førstegangs akut myokardie infarkt (AMI) blev tilbudt udvidet HR i

tillæg til standard HR. Studiet fandt en signifikant forskel i modificerbare risikofaktorer ved et-

års opfølgning, da interventionsgruppen blev sammenlignet med en kontrolgruppe, der bestod af

socialt udsatte patienter indlagt med førstegangs AMI, som modtog standard HR.

Formål

Formålet med denne ph.d.-afhandling var at undersøge langtidseffekten af det socialt

differentierede projekt i forhold til dødelighed, sygelighed, adherence til risikohåndtering og

forbrug af sundhedsydelser. Socialt udsatte patienter indlagt med førstegangs AMI, og som

modtog udvidet HR, blev sammenlignet med socialt udsatte patienter indlagt med førstegangs

AMI, som modtog standard HR.

Metode

Alle patienter < 70 år, som blev indlagt med førstegangs AMI på Aarhus Universitetshospital i

Denmark fra 2000 til 2004, som deltog i HR, og som afgav informeret samtykke, deltog i

langtidsopfølgningen. Social udsathed blev defineret ud fra uddannelsesniveau, samlivsstatus og

alder. Antallet af inkluderede patienter var 379. Af disse blev 208 kategoriseret som socialt

udsatte, hvoraf 131 modtog udvidet HR. De resterende 171 patienter blev kategoriseret som

ikke-socialt udsatte. Den udvidede HR-intervention bestod af flere konsultationer,

skræddersyede rehabiliteringsplaner og fokuseret overdragelse til den sene, livslange HR i almen

praksis samt en telefonisk opfølgningssamtale. Data blev indhentet fra registre og via et

spørgeskema, som blev sendt til patienternes praktiserende læge.

10. Dansk resumé

Page | 73

Resultater

Ingen signifikante forskelle blev fundet ti år efter baseline, mellem socialt udsatte patienter som

modtog udvidet HR, og socialt udsatte patienter som modtog standard HR i forhold til

dødelighed, kardiovaskulær dødelighed, reinfarkter og større kardiovaskulære hændelser. Der

blev ikke fundet nogen langtidsholdbare signifikante forskelle i løbet af de ti års opfølgning i

forhold til adherence til risikohåndtering, når udvidet HR blev sammenlignet med standard HR i

gruppen af socialt udsatte patienter. Ved to- og femårs opfølgning havde gruppen, der modtog

udvidet HR, en signifikant lavere medicin adherence til beta-blokkere og ved toårs opfølgning en

signifikant højere medicin adherence til statiner. Ved to- og femårs opfølgning var niveauerne

for total kolesterol, LDL-kolesterol, triglycerid og HbA1c signifikant lavere i gruppen, der

modtog udvidet HR, og ved toårs opfølgning var niveauet for HDL-kolesterol signifikant højere.

Den eneste markante forskel i forbruget af sundhedsydelser blev fundet i forhold til deltagelse i

årskontroller i almen praksis, som var signifikant højere, blandt socialt udsatte patienter som

modtog udvidet HR i sammenligning med socialt udsatte patienter, som modtog standard HR

ved to- og femårs opfølgning. Ud over analyserne baseret på det overordnede formål for ph.d.-

afhandlingen, blev der gennemført yderligere beskrivende analyser, hvor alle socialt udsatte

patienter blev sammenlignet med alle ikke-socialt udsatte patienter. I løbet af den tiårs

opfølgning af dødelighed, sygelighed adherence til risikohåndtering og forbrug af

sundhedsydelser blev der ikke fundet markante forskelle mellem patienterne uanset deres

socioøkonomiske status med undtagelse af rygning.

Konklusion og perspektiver

Denne afhandling demonstrerede, at det er muligt at reducere den sociale ulighed i sundhed i den

sene, livslange HR i primærsektoren, herunder almen praksis, i løbet af den tiårs opfølgning i

studiepopulationen. Resultaterne vedrørende især de socialt udsatte patienters rygestatus viser

imidlertid, at der stadig er lang vej før lighed i sundhedsadfærd er opnået. Der blev ikke fundet

nogen vedvarende signifikant langtidseffekt, når socialt udsatte patienter, der modtog udvidet

HR, blev sammenlignet med socialt udsatte patienter, der modtog standard HR. Det er uklart om

dette delvist kan forklares ud fra kvaliteten af den tilbudte standard HR. Det er nødvendigt at

tænke alternativt i den fremtidige socialt differentierede HR, og fokus skal være på at styrke og

motivere patienterne. Kampen for at udrydde social ulighed i sundhed skal og bør fortsætte.

Page | 74

11. References

11. References

Page | 75

1. Mahmood SS, Levy D, Vasan S, Wang T. The Framingham Heart Study and the epidemiology of

cardiovascular disease: a historical perspective. Lancet 2014;383:999-1008.

2. Piepoli MF, Hoes AW, Agewall S, Albus C, Brotons C, Catapano AL et al. 2016 European guidelines on

cardiovascular disease prevention in clinical practice. The sixth joint task force of the European society of

cardiology and other societies on cardiovascular disease prevention in clinical practice (constituted by

representatives of 10 societies and by invited experts). Eur J Prev Cardiol 2016;23:1-96.

3. Wilkins E, Wilson L, Wickramasinghe K, Bhatnagar P, Leal J, Luengo-Fernandez R et al. European

cardiovascular disease statistics 2017. European Heart Network. 2017.

4. Anderson L, Taylor RS. Cardiac rehabilitation for people with heart disease: an overview of Cochrane systematic

reviews. Cochrane Database Syst Rev 2014;12:1-21.

5. Kachur S, Chongthammakun V, Lavie CJ, Schutter AD, Arena R, Milani RV et al. Impact of cardiac

rehabilitation and exercise training program in coronary heart disease. Prog Cardiovasc Dis 2017;60:103-14.

6. McMahon SR, Ades PA, Thompson PD. The role of cardiac rehabilitation in patients with heart disease. Trends

Cardiovasc Med 2017;27:420-25.

7. Rauch B, Davos CH, Doherty P, Saure D, Metzendorf MI, Salzwedel A. The prognostic effect of cardiac

rehabilitation in the era of acute revascularisation and statin therapy: A systematic review and meta-analysis of

randomized and non-randomized studies – The cardiac rehabilitation outcome study (CROS). Eur J Prev Cardiol

2016;23:1914-39.

8. Anderson L, Oldridge N, Thompson DR, Zwisler AD, Rees K, Martin N et al. Exercise-based cardiac

rehabilitation for coronary heart disease. J Am Coll Cardiol 2016;67:1-12.

9. Rasmussen JN, Rasmussen S, Gislason GH, Buch P, Abildstrom SZ, Køber L et al. Mortality after acute

myocardial infarction according to income and education. J Epidemiol Community Health 2006;60:351-56.

10. Klitkou ST, Wangen KR. Educational attainment and differences in relative survival after acute myocardial

infarction in Norway: a registry-based population study. BMJ Open 2017;7:1-8.

11. Bergström G, Redfors B, Angerås O, Dworeck C, Shao Y, Haraldsson I et al. Low socioeconomic status of a

patient's residential area is associated with worse prognosis after acute myocardial infarction in Sweden. Int J

Cardiol 2015;182:141-7.

11. References

Page | 76

12. Molshatzki N, Drory Y, Myers V, Goldbourt U, Benyamini Y, Steinberg DM et al. Role of socioeconomic status

measures in long-term mortality risk prediction after myocardial infarction. Med Care 2011;49:673-8.

13. Murphy BM, Elliott PC, Grande MRL, Higgins RO, Ernest CS, Goble AJ et al. Living alone predicts 30-day

hospital readmission after coronary artery bypass graft surgery. Eur J Cardiovasc Prev Rehabil 2008;15:210-5.

14. Pogosova N, Saner H, Pedersen SS, Cupples ME, McGee H, Ho S et al. Psychosocial aspects in cardiac

rehabilitation: From theory to practice. A position paper from the cardiac rehabilitation section of the European

association of cardiovascular prevention and rehabilitation of the European society of cardiology. Eur J Prev Cardiol

2015;22:1290-310.

15. Dickens CM, McGowan L, Percival C, Douglas J, Tomenson B, Cotter L et al. Lack of a close confidant but not

depression, predicts further cardiac events after myocardial infarction. Heart 2004;90:518-22.

16. Snaterse M, Deckers JW, Lenzen MJ, Jorstad HT, Bacquer DD, Peters RJG et al. Smoking cessation in

European patients with coronary heart disease. Results from the EUROASPIRE IV survey: A registry from the

European society of cardiology. Int J Cardiol 2018;258:1-6.

17. Bruthans J, Mayer O, Bacquer DD, Smedt DD, Reiner Z, Kotseva K et al. Educational level and risk profile in

patients with coronary heart disease. Eur J Prev Cardiol 2016;23:881-90.

18. Korda RJ, Soga K, Joshy G, Calabria B, Attia J, Wong D et al. Socioeconomic variation in incidence of primary

and secondary major cardiovascular disease events: an Australian population-based prospective cohort study. Int J

Equity Health 2016;15:1-10.

19. Notara V, Panagiotakos DB, Kogias Y, Stravopodis P, Antonoulas A, Zombolos S et al. The Impact of

Educational Status on 10-Year (2004-2014) Cardiovascular Disease Prognosis and All-cause Mortality Among

Acute Coronary Syndrome Patients in the Greek Acute Coronary Syndrome (GREECS) Longitudinal Study. J Prev

Med Public Health 2016;49:220-9.

20. Beauchamp A, Peeters A, Tonkin A, Turrell G. Best practice for prevention and treatment of cardiovascular

disease through an equity lens: a review. Eur J Cardiovasc Prev Rehabil 2010;17:599-606.

21. Blas E, Kurup AS. Equity, social determinants and public health programmes. World Health Organization. 2010.

22. Sundhedsstyrelsen. Anbefalinger for tværsektorielle forløb for mennesker med hjertesygdom.

Sundhedsstyrelsen. 2018.

11. References

Page | 77

23. Regeringen. Sundere liv for alle. Nationale mål for danskernes sundhed de næste 10 år. Regeringen. 2014.

24. Sundhedsstyrelsen. EU's folkesundhedsprogram – 26. February 2018. EU's Folkesundhedsprogram er det tredje

af sin slags og løber i perioden 2014-2020. Available at https://www.sst.dk/da/planlaegning/eus-

folkesundhedsprogram (accessed 31. October 2018).

25. Meillier LK. I gang igen efter blodprop i hjertet – socialt differentieret hjerterehabilitering. Region Midtjylland.

2007.

26. Nielsen KM, Meillier LK, Larsen ML. Extended cardiac rehabilitation for socially vulnerable patients improves

attendance and outcome. Dan Med J 2013;60:1-5.

27. Sundhedsstyrelsen. Sygdomsbyrden i Danmark. Sundhedsstyrelsen. 2015.

28. Tsao CW, Vasan RS. Cohort Profile: The Framingham Heart Study (FHS): overview of milestones in

cardiovascular Epidemiology. Int J Epidemiol 2015;44:1800-13.

29. Menotti A, Lanti M, Kromhout D, Blackburn H, Jacobs D, Nissinen A et al. Homogeneity in the relationship of

serum cholesterol to coronary deaths across different cultures. 40-year follow-up of the seven countries study. Eur J

Cardiovasc Prev Rehabil 2008;15:719-25.

30. Menotti A, Puddu PE. How the Seven Countries Study contributed to the definition and

development of the Mediterranean diet concept: A 50-year journey. Nutr Metab Cardiovasc Dis 2015;25:245-52.

31. Yusuf S, Hawken S, Ôunpuu S, Dans T, Avezum A, Lanas F et al. Effect of potentially modifiable risk factors

associated with myocardial infarction in 52 countries (the INTERHEART study): case-control study. Lancet

2004;364:937-52.

32. World Health Organization. Prevention of cardiovascular disease. Guidelines for assessment and management of

cardiovascular risk. World Health Organization. 2007.

33. JBS3 Board. Joint British Societies' consensus recommendations for the prevention of cardiovascular disease

(JBS3). Heart 2014;100:1-67.

34. National institute for health and care excellence. Cardiovascular disease prevention. Public Health Guideline.

NICE. 2018.

35. Scottish intercollegiate guidelines network. Cardiac rehabilitation. A national clinical guideline. Health care

improvement Scotland. 2017.

11. References

Page | 78

36. Conroy RM, Pyörälä K, Fitzgerald AP, Sans S, Menotti A, Backer GD et al. Estimation of ten-year risk of fatal

cardiovascular disease in Europe: the SCORE project. Eur Heart J 2003;24:987-1003.

37. Boateng S, Sanborn T. Acute myocardial infarction. Dis Mon 2013;59:83-96.

38. Ibanez B, James S, Agewall S, Antunes MJ, Bucciarelli-Ducci C, Bueno Hector, Caforio ALP et al. 2017 ESC

Guidelines for the management of acute myocardial infarction in patients presenting with ST-segment elevation. Eur

Heart J 2018;39:119-77.

39. Piepoli MF, Corra U, Dendale P, Frederix I, Prescott E, Schmid JP et al. Challenges in secondary prevention

after acute myocardial infarction: A call for action. Eur Heart J Acute Cardiovasc Care 2017;6:299-310.

40. Kotseva K, Wood D, Bacquer DD. Determinants of participation and risk factor control according to attendance

in cardiac rehabilitation programmes in coronary patients in Europe: EUROASPIRE IV survey. Eur J Prev Cardiol

2018;25:1242-51.

41. World health organization. World report on disability. World health organization. 2011.

42. Maribo T, Nielsen CV. Rehabilitering – en grundbog. 2end ed. Gads Forlag; 2016.

43. Dansk cardiologisk selskab. Behandlingsvejledning – 19. April 2018. Hjerterehabilitering. Available at:

https://www.nbv.cardio.dk/hjerterehabilitering#292-indhold-af-rehabiliteringen (accessed 31. October 2018).

44. Dalal HM, Doherty P. Taylor RS. Cardiac rehabilitation. BMJ 2015;351:1-8.

45. Piepoli MF, Corra U, Adamopoulos S, Benzer W, Bjarnason-Wehrens B, Cupples M et al. Secondary prevention

in the clinical management of patients with cardiovascular diseases. Core components, standards and outcome

measures for referral and delivery. Eur J Prev Cardiol 2014;21:664-81.

46. Region Midtjylland. Forløbsprogram for hjertesygdom. Region Midtjylland. 2015.

47. Sundhedsstyrelsen. National klinisk retningslinje for hjerterehabilitering. Sundhedsstyrelsen. 2013.

48. Bjarnason-Wehrens B, McGee H, Zwisler AD, Piepoli MF, Benzer W, Schmid JP et al. Cardiac rehabilitation in

Europe: results from the European cardiac rehabilitation inventory survey. Eur J Cardiovasc Prev Rehabil

2010;17:410-8.

11. References

Page | 79

49. Anderson L, Brown JPR, Clark AM, Dalal H, Rossau HK, Bridges C et al. Patient education in the management

of coronary heart disease. Cochrane Database Syst Rev 2017;6:1-145.

50. Liu XL, Shi Y, Willis K, Wu CJ, Johnson M. Health education for patients with acute coronary syndrome and

type 2 diabetes mellitus: an umbrella review of systematic reviews and meta-analyses. BMJ Open 2017;7:1-30.

51. Sturchio A, Gianni AD, Campana B, Genua MP, Storti M, Di lasi G et al. Coronary artery risk management

programme (CARIMAP). Delivered by a rehabilitation day-hospital. Impact on patients with coronary artery

disease. J Cardiopulm Rehabil Prev 2012;32:386-93.

52. Saffi MAL, Polanczyk CA, Rabelo-Silva ER. Lifestyle interventions reduce cardiovascular risk in patients with

coronary artery disease: A randomized clinical trial. Eur J Cardiovasc Nurs 2014;13:436-43.

53. Lynggaard V, Nielsen CV, Zwisler AD, Taylor RS, May O. The patient education - Learning and Coping

Strategies - improves adherence in cardiac rehabilitation (LC-REHAB): A randomized controlled trial. Int J Cardiol

2017;236:65-70.

54. Retsinformation. Bekendtgørelse af sundhedsloven – 28. February 2018. Available at:

https://www.retsinformation.dk/Forms/R0710.aspx?id=199871#id9a9fd330-8258-4be9-a379-73e6e5c6596f

(accessed 31. October 2018).

55. Sundhedsstyrelsen. Sundhedsaftaler – 8. June 2018. Available at:

https://www.sst.dk/da/planlaegning/sundhedsaftaler (accessed 31. October 2018).

56. Sundhedsstyrelsen. Sundhedsaftaler 2015 – 2018 – 8. June 2018. Available at:

https://www.sst.dk/da/planlaegning/sundhedsaftaler/sundhedsaftaler-2015-2018 (accessed 31. October 2018).

57. Netværk af forebyggende sygehuse i Danmark & Hjerteforeningen. Hjerterehabilitering efter strukturreformen.

Netværk af forebyggende sygehuse i Danmark & Hjerteforeningen. 2007.

58. Piepoli MF, Hoes AW, Brotons C, Hobbs RFD, Corra U. Main messages for primary care from the 2016

European guidelines on cardiovascular disease prevention in clinical practice. Eur J Gen Pract 2018;24:51-6.

59. Pedersen KM, Andersen JS, Søndergaard J. General practice and primary health care in Denmark. J Am Board

Fam Med 2012;25:34-8.

60. Murphy E, Vellinga A, Murphy AW, Byrne M, Cupples ME, Buckley B et al. Primary care organizational

interventions for secondary prevention of ischaemic heart disease: a systematic review and meta-analysis. Br J Gen

Pract 2015;65:460-8.

11. References

Page | 80

61. Dansk selskab for almen medicin. Iskæmisk hjertekarsygdom. Årskontroller. Available at:

https://vejledninger.dsam.dk/hjerte/?mode=visKapitel&cid=1072&gotoChapter=1072#chapter_1166 (Accessed 31.

October 2018).

62. Marchal S, Hof AWJ, Hollander M. The new European guideline on cardiovascular disease prevention; how to

make progress in general practice? Eur J gen Pract 2018;24:57-9.

63. Christensen KS, Hansen DG. Den alment praktiserende læges rolle i forbindelse med rehabilitering. Ugeskr

Laeger 2014;12:940-3.

64. Hvass LR, Manghezi A, Folker AP, Sandø N. Social ulighed i sundhed – hvad kan kommunen gøre?

Sundhedsstyrelsen. 2012.

65. World Health Organization. Global strategy for health for all by the year 2000. World Health Organization.

1981.

66. World Health Organization. Targets for health for all. Targets in support of the European regional strategy for

health for all. World Health Organization. 1985.

67. World Health Organization. Health 21. The health for all policy framework for the WHO European region.

World Health Organization. 1999.

68. Vital. Folkesundhedsprogrammet 1999 - 2008. Sundhedsstyrelsen. 1999.

69. Regeringen. Sund hele livet - de nationale mål og strategier for folkesundheden 2002 - 10. Indenrigs- og

Sundhedsministeriet. 2002.

70. Diderichsen F, Andersen I, Manuel C. Ulighed i sundhed - årsager og indsatser. Sundhedsstyrelsen. 2011.

71. Bakah M, Raphael. New hypotheses regarding the Danish health puzzle. Scand J Public Health 2017;45:799-

808.

72. Diderichsen F, Scheele CE, Little IG. Tackling health inequalities locally: The Scandinavian experience.

University of Copenhagen. 2015.

11. References

Page | 81

73. Nielsen TJ, Vestergaard M, Fenger-Grøn M, Christensen B, Larsen KK. Healthcare contacts after myocardial

infarction according to mental health and socioeconomic position: A population-based cohort study. PLoS One

2015;10:1-3.

74. Gaalema D, Elliott RJ, Morford ZH, Higgins ST, Ades PA. Effect of socioeconomic status on propensity to

change risk behaviors following myocardial infarction: Implications for healthy lifestyle medicine. Prog Cardiovasc

Dis 2017;60:159-68.

75. Fors A, Gyllensten H, Swedberg K, Ekman I. Effectiveness of person-centred care after acute coronary

syndrome in relation to educational level: Subgroup analysis of a two-armed randomised controlled trial. Int J

Cardiol 2016;221:957-62.

76. Mayer-Berger W, Simic D, Mahmoodzad J, Burtscher R, Kohlmeyer, Schwitalla B et al. Efficacy of a long-term

secondary prevention programme following inpatient cardiovascular rehabilitation on risk and health-related quality

of life in a low-education cohort: a randomized controlled study. Eur J Prev Cardiol 2014;21:145-52.

77. Graversen CB, Eichhorst R, Ravn L, Christiansen SSR, Johansen MB, Larsen ML. Social inequality and barriers

to cardiac rehabilitation in the rehab-North register. Scan Cardiovasc J 2017;51:316-22.

78. Momsen AMH, Hald K, Nielsen CV, Larsen ML. Effectiveness of expanded cardiac rehabilitation on mortality

and lifestyle risk factors in patients diagnosed with coronary heart disease - a systematic review. J Card Pulm

Rehabil. 2018;2:1-18.

79. Videbæk J, Andersen LV, Bentzen J. Dansk hjertestatistik 2010. Hjerteforeningen & Syddansk Universitet.

2011.

80. Hald K, Nielsen KM, Nielsen CV, Meillier LK, Larsen FB, Christensen et al. Expanded cardiac rehabilitation in

socially vulnerable patients with myocardial infarction: a 10-year follow-up study focusing on mortality and

nonfatal events. BMJ Open 2018;8:1-8. (Appendix I)

81. Hald K, Larsen FB, Nielsen KM, Meillier LK, Johansen MB, Larsen ML, Christensen B, Nielsen CV.

Medication adherence, biological and lifestyle risk factors in patients with myocardial infarction: A ten-year follow-

up on socially differentiated cardiac rehabilitation. Accepted for publication in Scandinavian Journal of Primary

Health Care. (Appendix II)

82. Hald K, Meillier LK, Nielsen KM, Larsen FB, Johansen MB, Larsen ML, Nielsen CV, Christensen B. Does

socially differentiated cardiac rehabilitation affect the use of health care services after myocardial infarction? A ten-

year follow-up study. Submitted to Scandinavian Journal of Public Health. (Appendix III)

11. References

Page | 82

83. Danmarks Statistik. Alfabetisk liste over DUN-koder. Available at:

http://www.dst.dk/Site/Dst/Udgivelser/GetPubFile.aspx?id=4426&sid=alfabetisk (Accessed 31. October 2018).

84. Hjerteforeningen. Rehabilitering af hjertepatienter. Retningslinjer. Kerteminde Tryk A/S; 1997.

85. Balady GJ, Ades PA, Comoss P, Limacher M, Pina IL, Southard D et al. Core components of cardiac

rehabilitation / secondary prevention programs. A statement for healthcare professionals from the American heart

association and the American association of cardiovascular and pulmonary rehabilitation. Circulation

2000;102:1069-73.

86. Gohlke H, Gohlke-Bärwolf C. Cardiac rehabilitaiton: where are we going? Eur Heart J 1998;19:5-12.

87. Schmidt M, Pedersen L, Sørensen HT. The Danish Civil Registration System as a tool in epidemiology. Eur J

Epidemiol 2014;29:541-49.

88. Pedersen CB. The Danish civil registration system. Scand J Public Health 2011;39:22-5.

89. Larsen KH. The Danish register of causes of death. Scand J Public Health 2011;39:26-9.

90. Lynge E, Sandegaard JL, Rebolj M. The Danish national patient register. Scand J Public Health 2011;39:30-3.

91. Kildemoes HW, Sørensen HT, Hallas J. The Danish national prescription registry. Scand J Public Health

2011;39:38-41.

92. Grann AF, Erichsen R, Nielsen AG, Frøslev T, Thomsen RW. Existing data sources for clinical epidemiology:

The clinical laboratory information system (LABKA) research database at Aarhus University, Denmark. Clin

Epidemiol 2011;3:133-8.

93. Andersen JS, Olivarius NDF, Krasnik A. The Danish National Health Service Register. Scand J Public Health

2011;39:34-7.

94. Epidata Software. Available at: http://www.epidata.dk/credit.htm (Accessed 31. October 2018).

95. International Statistical Classification of Diseases and Related Health Problems 10th Revision. Available at: http://apps.who.int/classifications/icd10/browse/2016/en (Accessed 31. October 2018).

96. Hedbäck B, Perk J, Hörnblad M, Ohlsson U. Cardiac rehabilitation after coronary artery bypass surgery: 10-year

results on mortality, morbidity and readmissions to hospital. J Cardiovasc Risk 2001;8:153-8.

11. References

Page | 83

97. Plüss CE, Billing E, Held C, Henriksson P, Kiessling A, Karlsson MR et al. Long-term effects of an expanded

cardiac rehabilitation programme after myocardial infarction or coronary artery bypass surgery: a five-year follow-

up of a randomized controlled study. Clin Rehabil 2011;25:79-87.

98. Murphy AW, Cupples ME, Murphy E, Newell J, Scarrott CJ, Vellinga A et al. Six-year follow-up of the

SPHERE RCT: secondary prevention of heart disease in general practice. BMJ Open 2015;5:1-7.

99. Ho PM, Lambert-Kerzner A, Carey EP, Fahdi IE, Bryson CL, Melnyk SD et al. Multifaceted intervention to

improve medication adherence and secondary prevention measures after acute coronary syndrome hospital

discharge. A randomized clinical trial. JAMA Intern Med 2014;174:186-93.

100. King W, Lacey A, White J, farewell D, Dunstan F, Fone D. Socioeconomic inequality in medication

persistence in primary and secondary prevention of coronary heart disease - A population-wide electronic cohort

study. PLoS One 2018;13:1-12.

101. Sverre E, Otterstad JE, Gjertsen E, Gullestad L, Husebye E, Dammen T et al. Medical and sociodemographic

factors predict persistent smoking after coronary events. BMC Cardiovasc Disord 2017:17:1-9.

102. Schneider APH, Gaedke MA, Garcez A, Barcellos NT, Paniz VMV. Effect of characteristics of

pharmacotherapy on non-adherence in chronic cardiovascular disease: A systematic review and meta-analysis of

observational studies.

103. Rashid MA, Edwards D, Walter Fm, Mant J. Medication Taking in Coronary Artery Disease: A Systematic

Review and Qualitative Synthesis. Ann Fam Med 2014;12:224-32.

104. Khawaja FJ, Shah ND, Lennon RJ, Slusser JP, Alkatib AA, Rihal CS et al. Factors associated with 30-day

readmission rates after percutaneous coronoary intervention. Arch Intern Med 2012;172:112-7.

105. Juel S. Epidemiologi og evidens. 1st ed. København: Munksgaard Danmark; 2011.

106. Greenberg RS, Daniels SR, Flanders WD, Eley JW, Boring JR. Medical epidemiology. Fourth edition. The

McGraw-Hill Companies; 2005.

107. Fletcher RH, Fletcher SW, Fletcher GS. Clinical epidemiology. The essentials. Fifth edition. Wolters Kluwer.

Lippincott Williams & Wilkins; 2014.

108. Ingholt L, Rod MH, Holmberg T, Srivarathan A, Rossau HK, Tjørnhøj-Thomsen T. Socialdifferentieret

hjerterehabilitering i praksis. Syddansk Universitet & Hjerteforeningen. 2016.

11. References

Page | 84

109. Evans T, Whitehead M, Diderichsen F, Bhuiya A, Wirth M. Challenging inequalities in health. From ethics to

action. Oxford University Press; 2001.

110. Kirkwood BR, Sterne JAC. Essential Medical Statistics. 2nd ed. Blackwell Science; 2005.

111. Engel GL. The need for a new medical model: A challenge for biomedine. Science 1977;196:129-36.

112. Wade D. Rehabilitation - a new approach. Part two: the underlying theories. Clin Rehabil 2015;29:1145-54.

113. Moore GF, Audrey S, Barker M, Bond L, Bonell C, Hardeman W et al. Process evaluation of complex

interventions: Medical research council guidance. BMJ 2015;350:1-7.

114. Bjerre ED, Hansen ABG. Komplekse interventioner i medicinsk forskning. Ugeskr Laeger 2018;19:2-6.

115. Craig P, Dieppe P, Macintyre S, Michie S, Nazareth I, Petticrew M. Developing and evaluating complex

interventions: the new medical research council guideline. Int J Nurs Stud 2013;50:587-92.

116. Egholm CL, Rossau HK, Nilsen P, Bunkenborg G, Rod MH, Doherty P et al. Implementation of a politically

initiated national clinical guideline for cardiac rehabilitation in hospitals and municipalities in Denmark. Health

Policy 2018;122:1043-51.

117. Sundheds- og ældreministeriet. Ny rapport giver overblik inden sundhedsreformen. 5. October 2018. Available

at: https://www.sum.dk/Aktuelt/Nyheder/Sundhedspolitik/2018/Oktober/Ny-rapport-giver-overblik-inden-

sundhedsreformen.aspx (Accesssed 31. October 2018).

118. Sundheds- og ældreministeriet. Sundheds- og ældreøkonomisk analyse. Kontaktmønstre på tværs af sektorer

blandt befolkningen, kronikere og ældre medicinske patienter. 2018.

119. Regeringen. Sammen med borgeren. En helhedsorienteret indsats. Finansministeriet. 2018.

120. Larkin M. Developing the knowledge base about carers and personalisation: Contributions made by an

exploration of carers' perspectives on personal budgets and the carer-service user relationship. Health Soc Care

Community 2015;23:33-41.

121. Hamilton S, Szymczynska P, Clewett N, Manthorpe J, Tew J, Larsen J et al. The role of family carers in the use

of personal budgets by people with mental health problems. Health Soc Care Community 2017;25:158-66.

11. References

Page | 85

122. Larsen J, Tew J, Hamilton S, Manthorpe J, Pinfold V, Szymczynska P, Clewett N. Outcomes from personal

budgets in mental health: Service users' experiences in three English local authorities. J Ment Health

2015;24:219-24.

Page | 86

12. Appendices

12. Appendices

Appendix I

Paper I

Hald K, Nielsen KM, Nielsen CV, Meillier LK, Larsen FB, Christensen B, Larsen ML.

Expanded cardiac rehabilitation in socially vulnerable patients with myocardial infarction:

a 10-year follow-up study focusing on mortality and non-fatal events.

BMJ Open. 2018;8:1-8.

Appendix II

Paper II

Hald K, Larsen FB, Nielsen KM, Meillier LK, Johansen MB, Larsen ML, Christensen B,

Nielsen CV. Medication adherence, biological and lifestyle risk factors in patients with


rehabilitation.

Accepted for publication in Scandinavian Journal of Primary Health Care.

Appendix III

Paper III

Hald K, Meillier LK, Nielsen KM, Larsen FB, Johansen MB, Larsen ML, Nielsen CV,

Christensen B. Does socially differentiated cardiac rehabilitation affect the use of health

care services after myocardial infarction? A ten-year follow-up study.

Submitted to Scandinavian Journal of Public Health.

Appendix IV

Invitation letter and GP questionnaire

Paper I

1Hald K, et al. BMJ Open 2018;8:e019307. doi:10.1136/bmjopen-2017-019307

Open Access

Expanded cardiac rehabilitation in socially vulnerable patients with myocardial infarction: a 10-year follow-up study focusing on mortality and non-fatal events

Kathrine Hald,1 Kirsten Melgaard Nielsen,2 Claus Vinther Nielsen,1 Lucette Kirsten Meillier,3 Finn Breinholt Larsen,3 Bo Christensen,4 Mogens Lytken Larsen5

To cite: Hald K, Nielsen KM, Nielsen CV, et al. Expanded cardiac rehabilitation in socially vulnerable patients with myocardial infarction: a 10-year follow-up study focusing on mortality and non-fatal events. BMJ Open 2018;8:e019307. doi:10.1136/bmjopen-2017-019307

► Prepublication history for this paper is available online. To view these files, please visit the journal online (http:// dx. doi. org/ 10. 1136/ bmjopen- 2017- 019307).

Received 4 September 2017Revised 27 October 2017Accepted 8 December 2017

1Department of Public Health, Section for Clinical Social Medicine and Rehabilitation, Aarhus University, Aarhus, Denmark2Department of Cardiology, Aarhus University Hospital, Aarhus, Denmark3DEFACTUM, Social and Health Services and Labour Market, Central Denmark Region, Aarhus, Denmark4Department of Public Health, Section for General Medical Practice, Aarhus University, Aarhus, Denmark5Department of Cardiology, Danish Centre for Inequality in Health, Aalborg University Hospital, Aalborg, Denmark

Correspondence toKathrine Hald; kathrine. hald@ stab. rm. dk

Research

AbstrACtObjective Cardiac rehabilitation (CR) has been shown to reduce cardiovascular risk. A research project performed at a university hospital in Denmark offered an expanded CR intervention to socially vulnerable patients. One-year follow-up showed significant improvements concerning medicine compliance, lipid profile, blood pressure and body mass index when compared with socially vulnerable patients receiving standard CR. The aim of the study was to perform a long-term follow-up on the socially differentiated CR intervention and examine the impact of the intervention on all-cause mortality, cardiovascular mortality, non-fatal recurrent events and major cardiac events (MACE) 10 years after.Design Prospective cohort study.setting The cardiac ward at a university hospital in Denmark from 2000 to 2004.Participants 379 patients aged <70 years admitted with first episode myocardial infarction (MI). The patients were defined as socially vulnerable or non-socially vulnerable according to their educational level and their social network. A complete follow-up was achieved.Intervention A socially differentiated CR intervention. The intervention consisted of standard CR and additionally a longer phase II course, more consultations, telephone follow-up and a better handover to phase III CR in the municipal sector, in general practice and in the patient association.Main outcome measures All-cause mortality, cardiovascular mortality, non-fatal recurrent events and MACE.results There was no significant difference in all-cause mortality (OR: 1.29, 95% CI 0.58 to 2,89), cardiovascular mortality (OR: 0.80, 95% CI 0.31 to 2.09), non-fatal recurrent events (OR:1.62, 95% CI 0.67 to 3.92) or MACE (OR: 1.31, 95% CI 0.53 to 2.42) measured at 10-year follow-up when comparing the expanded CR intervention to standard CR.Conclusions Despite the significant results of the socially differentiated CR intervention at 1-year follow-up, no long-term effects were seen regarding the main outcome measures at 10-year follow-up. Future research should

focus on why it is not possible to lower the mortality and morbidity significantly among socially vulnerable patients admitted with first episode MI.

IntrODuCtIOn According to the European Association for Cardiovascular Prevention and Rehabilita-tion, cardiovascular disease (CVD) remains a leading cause of mortality and morbidity, although CVD mortality has declined

strengths and limitations of this study

► This is the first longitudinal study to analyse the long-term effects of a socially differentiated cardiac rehabilitation intervention given to patients admitted with first episode myocardial infarction, which provide knowledge in better understanding how to reduce social inequalities in health.

► Highly valid Danish register data were used that combined with a unique personal 10-digit civil registration number that is given to all citizens living in Denmark provides the study with a complete follow-up.

► The study was not carried out as a randomised controlled trial. To minimise potential confounding, regression analysis was used. Moreover, the patients were almost similar at baseline.

► The intervention given in the study was designed as a ‘realistic intervention’. The aim was to create an intervention that would be affordable and applicable to most rehabilitation centres if proven effective.

► Patients from non-parallel time periods were being compared. All analyses were performed on both the socially and non-socially vulnerable patients. A difference between the non-socially vulnerable patients could have indicated that any changes among the socially vulnerable patients were just a general development in risk management and secondary prevention.

on 19 Decem

ber 2018 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2017-019307 on 23 January 2018. Dow

nloaded from

http://bmjopen.bmj.com/

http://dx.doi.org/10.1136/bmjopen-2017-019307



http://crossmark.crossref.org/dialog/?doi=10.1136/bmjopen-2017-019307&domain=pdf&date_stamp=2018-01-22


2 Hald K, et al. BMJ Open 2018;8:e019307. doi:10.1136/bmjopen-2017-019307

Open Access

considerably in the past 20 years.1 However, the 1-year mortality rate is around 20% in patients with myocardial infarction (MI). Among the patients who survive, 20% will experience a recurrent MI within 1 year.2 It is estimated that recurrent events are caused by progression of coro-nary and systemic atherosclerosis.2 Secondary prevention including cardiac rehabilitation (CR) is therefore essen-tial to improve the long-term prognosis of patients with MI and to improve their quality of life and functional capacity.2 3 CR consists of multidisciplinary interventions with focus on risk assessment and management.2

A recent Cochrane meta-analysis and a review exam-ining the effect of exercise-based CR with at least 6-month follow-up found that CR significantly improved psycho-logical function and reduced cardiovascular mortality.4 5 Another recent meta-analysis reported that CR containing lifestyle modification programmes significantly reduced recurrent events, all-cause mortality and cardiovascular mortality if CR combined goal setting, self-monitoring, planning and feedback.6 Two randomised controlled trials (RCTs) examined the effect of an expanded CR intervention. One of the interventions consisted of different lifestyle modification activities as well as stress management therapy. The other of the interventions consisted of exercise-based CR. At 3-year and 5-year follow-ups, the patients randomised to receive expanded CR experienced fewer non-fatal recurrent events and a lower cardiovascular mortality compared with patients receiving standard CR.7 8

Patients with low socioeconomic status, defined by their social class, educational level, income, occupation and marital status, are less likely to participate in and complete CR.9–11 This is also seen in patients with MI when focusing on mortality and non-fatal recurrent events.12–15 Patients with a low educational level have a significantly higher long-term mortality than patients with a high educational level.16 Likewise, patients living alone have a significantly higher long-term mortality risk compared with patients living with a partner.17

On a cardiac ward at a university hospital in Denmark, a socially differentiated CR intervention was performed from 2000 to 2004. The aim of the intervention was to target the social groups at highest risk of not partic-ipating in CR, not completing CR and who have the poorest long-term outcomes. The intervention was designed as a ‘realistic intervention’ based on the health professionals’ experiences. The idea of the ‘realistic intervention’ was that it should be affordable and prac-tical to implement if proven effective. Patients defined as socially vulnerable received expanded CR, and outcome was compared with socially vulnerable patients receiving standard CR according to international guidelines. At 1-year follow-up, patients in the intervention group had significantly better results in relation to medicine compliance, lipid profile, blood pressure and body mass index.18

The aim of the present study was to perform a long-term follow-up on the socially differentiated CR intervention

and examine the impact of the intervention on mortality and non-fatal recurrent events 10 years after.

MethODsstudy designThis is a prospective cohort study. Patients were followed from baseline, defined as time of admission with first episode of MI, and during the next 10 years. Follow-up was performed at the exact day 10 years after their admission.

The 4-year socially differentiated CR intervention was carried out on a cardiac ward at a university hospital in Denmark between 2000 and 2004.

This study focuses on the socially vulnerable patients who received expanded CR compared with those who received standard CR.

Patient populationFrom 1 April 2000 to 31 March 2002, all patients aged <70 years admitted with first episode of MI were systematically identified. Of the 205 patients with MI, 171 were referred to standard CR; 133 patients gave informed consent to participate. Of these, 78 patients were categorised as socially vulnerable and 55 were categorised as non-so-cially vulnerable. All of the 133 patients received standard CR according to international guidelines.

From 1 September 2002 to 31 December 2004, all patients aged <70 years admitted with first episode of MI were assessed by a project nurse and referred to either standard CR or expanded CR. A total of 303 patients were admitted; 270 patients were referred to CR of whom 246 patients gave informed consent to participate. Of these, 130 patients were categorised as socially vulnerable and received expanded CR, and the remaining 116 patients were categorised as non-socially vulnerable and received standard CR.

Patients were defined as socially vulnerable if they had: (1) low educational level (education classified 1–4 in the Danish Educational Nomenclature if age <55 years and 1–3 if age >55 years) and/or (2) if they lived alone. Patients were defined as non-socially vulnerable if they did not meet the criteria above.

Patients were excluded if they suffered from severe comorbidities such as stroke, dementia, mental disorders, retardation or severe alcohol abuse. Patients suffering from depression or anxiety were not excluded.

The study population, categorisation and CR character-istics are described in detail in figure 1.

exposureThe expanded CR intervention consisted of standard CR and a longer phase II course, more consultations, tele-phone follow-up and a better handover to phase III CR in the municipal sector, in general practice and in the patient association.

The standard CR intervention was consistent with inter-national guidelines.

on 19 Decem


http://bmjopen.bm

j.com/

BM



nloaded from



Open Access

The differences between the two CR interventions are described in detail in table 1.

study outcomesThe main outcome measures in the present study were all-cause mortality, cardiovascular mortality, non-fatal recurrent events (MI and unstable angina pectoris) and major cardiac events (MACE) defined as cardiovascular mortality and non-fatal recurrent events. The endpoints were adjusted for gender, age, diabetes and smoking status at baseline.

Data sourcesBaseline patient data were collected at admission from clinical databases and from questionnaires filled in by the patients. In 1968, The Danish Civil Registration System was introduced. The system provides all persons living in Denmark with a unique personal 10-digit civil regis-tration number. This number was used to link the study

population to different registers ensuring a high validity and completeness. Endpoint data concerning mortality were collected from The Danish Cause of Death Register established in 1970. Cardiovascular mortality was defined using the International Classification of Diseases (ICD-10). Data on non-fatal recurrent events were retrieved using the ICD-10 from The Danish National Patient Registry established in 1977.

statisticsCategorical variables are presented as numbers and percentages. Continuous variables are presented as mean with SD. The Kaplan-Meier estimate plots were used to evaluate survival probability and event-free proba-bility. Logistic regression was applied when performing adjusted analyses. All endpoints are presented as ORs with 95% (CIs) and P values. A significance level of 0.05 was applied. When performing the adjusted analyses, the

Figure 1 Flow chart of study participants.

on 19 Decem


http://bmjopen.bm

j.com/

BM



nloaded from



Open Access

rule of 10 was used. All statistical analyses were carried out using the statistics software program Stata V.14.1.

resultsbaseline characteristicsFrom 1 April 2000 to 31 December 2004, 379 patients were referred to and participated in a socially differentiated CR intervention receiving either a standard or expanded CR intervention (figure 1). Baseline characteristics of the

patients are given in table 2. A complete follow-up after 10 years was achieved.

All-cause mortalityA total of 17% of the vulnerable patients died during the 10-year follow-up period; 18% of these patients had received expanded CR and 15% had received standard CR. No significant differences were found between the two groups as an OR of 1.29 (95% CI 0.58 to 2.89), and a P value of 0.53 was obtained (table 3). As indicated in

Table 1 Content of the socially differentiated cardiac rehabilitation intervention

Standard cardiac rehabilitation Expanded cardiac rehabilitation

Phase IAcute treatment until discharge

► Start of medical and acute surgical treatment. ► Start of secondary prevention concerning medication,smoking, diet and exercise. ► Psychological and social support to patients and relatives.

Like standard cardiac rehabilitation

Phase IIDischarge from hospital until return to vocational activities

► 5–6 weeks of cardiac rehabilitation. ► Three consultations with medical doctor. ► Four consultations with nurse. ► Two consultations with dietitian. ► 6–12 weeks of exercise course. ► Screening for depression and anxiety.

Like standard cardiac rehabilitation and: ► Extra 2 weeks of cardiac rehabilitation. ► One extra consultation with nurse. ► Sharing of patient’s own rehabilitation plan with general practice.

Phase IIIFurther course after phase II

► Referral to general practice. ► Information about activities in the municipal sector and in The Danish Heart Association.

Like standard cardiac rehabilitation and: ► Referral to half hour of preventive consultation in general practice. ► Referral to activities in the municipal sector and in The Danish Heart Association. ► Telephone follow-up 2 months after completion of phase II.

Table 2 Baseline characteristics at patient admission with first episode myocardial infarction (n=379)

Socially vulnerable patients Non-socially vulnerable patients


Time period


Time period

Standardn=782000–2002(%/SD)

Expandedn=1302002–2004(%/SD)

Standardn=552000–2002(%/SD)

Standardn=1162002–2004(%/SD)

Age, years 56 (8.15) 55 (8.53) 60 (7.56) 57 (8.50)

Gender, male 57 (73) 93 (71) 42 (76) 94 (81)

Educational level, The Danish Educational Nomenclature

3.18 (1.19) 3.26 (1.39) 4.80 (1.08) 4.75 (1.19)

Living alone 27 (35) 51 (39) 0 0

Current smoker 59 (76) 83 (64) 34 (62) 60 (52)

Body mass index 27.26 (4.35) 26.26 (4.08) 26.37 (3.99) 26.54 (3.12)

Hypertension 18 (23) 28 (22) 11 (20) 23 (20)

Hyperlipidaemia 20 (26) 37 (28) 13 (24) 44 (38)

Diabetes mellitus 10 (13) 16 (12) 6 (11) 10 (9)

on 19 Decem


http://bmjopen.bm

j.com/

BM



nloaded from



Open Access

figure 2, no significant associations were found at 10-year follow-up among the non-socially vulnerable patients receiving standard CR.

Cardiovascular mortalityAmong the vulnerable patients, 9% suffered from cardio-vascular mortality. Of the patients receiving expanded CR, 8% died compared with 10% among patients receiving standard CR. No significant differences were found at 10-year follow-up; OR 0.80 (95% CI 0.31 to 2.09) and P value 0.65 (table 3). As indicated in figure 2, no signif-icant associations were found at 10-year follow-up among the non-socially vulnerable patients receiving standard CR.

non-fatal recurrent eventsOnly patients who did not experience a non-fatal recur-rent event during the first 30 days after admission were included in the analysis. A total of 17% of the vulner-able patients experienced a non-fatal recurrent event during the 10-year follow-up; among these, 19% received expanded CR and 13% received standard CR. No signifi-cant differences were found between the two groups; OR 1.62 (95% CI 0.67 to 3.92) and a P value of 0.29 (table 3). As indicated in figure 2, no significant associations were found at 10-year follow-up among the non-socially vulner-able patients receiving standard CR.

Major cardiac eventsThe percentage of vulnerable patients who either experi-enced cardiovascular mortality or experienced a non-fatal recurrent event within 30 days after admission until 10-year follow-up was 23% in total and in each group. No significant differences were seen between the two groups; OR 1.31 (95% CI 0.53 to 2.42) and a P value of 0.63 (table 3). As indicated in figure 2, no significant associa-tions were found at 10-year follow-up among the non-so-cially vulnerable patients receiving standard CR.

DIsCussIOnstudy findingsThere were no significant differences between socially vulnerable patients admitted with first episode MI receiving expanded CR and socially vulnerable patients receiving standard CR concerning the four endpoints: all-cause mortality, cardiovascular mortality, non-fatal recurrent events and MACE at 10-year follow-up (table 3). Moreover, no significant results were found at 10-year follow-up among the non-socially vulnerable patients who all received standard CR.

Comparison with other studiesTwo studies have examined the effect of an expanded CR intervention. In a Swedish RCT by Plüss et al,7 224 patients aged <75 years with recent MI and/or coronary artery bypass grafting (CABG) were randomised to either expanded CR or standard CR between 1999 and 2002 and followed for 5 years. Patients were excluded if suffering from a significant psychiatric disease or alcohol abuse. All patients received 3 months of standard CR including consultations with health professionals and a social worker, physical exercise, patient education and advice on smoking cessation. The patients receiving the expanded intervention also stayed 5 days at a patient hotel after discharge, where they participated in a cooking school for 3 weeks and attended a stress management course for 1 year. The study had an almost complete follow-up and a significantly lower number of the patients in the intervention group suffered a non-fatal recurrent event at 5-year follow-up (hazard rate 0.47, 95% CI 0.21 to 0.97, P value 0.04). No significant results were found regarding all-cause and cardiovascular mortality.7

The study by Plüss et al7 has many similarities with the present study. Sweden and Denmark have similar welfare states with the same access to free healthcare and social services. The patients in the two studies were recruited in the same time period and had comparable characteristics

Table 3 Endpoints at 10-year follow-up among socially vulnerable patients admitted with first episode myocardial infarction and participating in socially differentiated cardiac rehabilitation in the period from 2000 to 2004

Total(n=208)

Expanded cardiac rehabilitation(n=130)

Standard cardiac rehabilitation(n=78)

OR(95% CI) P value

All-cause mortality* 35 (17) 23 (18) 12 (15) 1.29 (0.58 to 2.89) 0.53Cardiovascular† 19 (9) 11 (8) 8 (10) 0.80 (0.31 to 2.09) 0.65

Total(n=176‡)

Expanded cardiac rehabilitation(n=115‡)

Standard cardiac rehabilitation(n=61‡)

OR(95% CI) P value

Non-fatal recurrent events* 30 (17) 22 (19) 8 (13) 1.62 (0.67 to 3.92) 0.29Major cardiac events§ 41 (23) 27 (23) 14 (23) 1.31 (0.53 to 2.42) 0.75

Data are given as numbers (percentage).*Adusted for gender, age and diabetes mellitus.†Adjusted for gender.‡Only patients who did not suffer from a recurrent event during the first month after admission were included in the analysis.§Adjusted for gender, age, diabetes and smoking status.

on 19 Decem


http://bmjopen.bm

j.com/

BM



nloaded from



Open Access

concerning disease and age. Furthermore, exclusion criteria were the same. However, the Swedish in contrast to the present study found significant results. This could be explained by the Swedish intervention being more pervasive and lasting a whole year and thereby consti-tuting a major part of the long-term secondary preven-tion. Furthermore, the Swedish intervention was not socially differentiated. It could thus be speculated that the patients who profited the most from the intervention were the patients who were not socially vulnerable.

In an Italian RCT by Giannuzzi et al,8 3241 patients aged <75 years with recent MI were randomised to either

expanded CR or usual care. At first, all patients received the same standard CR for 1 month consisting of phys-ical training, lifestyle consultations and medical therapy. Hereafter 1621 patients continued in usual care, and 1620 patients received an expanded CR intervention. The expanded CR intervention consisted of 2 hours of counselling and physical training every month for half a year and thereafter every 6 months for 3 years. Compared with usual care the expanded CR intervention showed significant improvements concerning cardiovascular mortality and recurrent events. The study by Giannuzzi et al8 differs from the present study regarding to the time

Figure 2 Kaplan-Meier estimates of the probability of all-cause mortality, cardiovascular mortality, non-fatal recurrent events and major cardiac events.

on 19 Decem


http://bmjopen.bm

j.com/

BM



nloaded from



Open Access

frame of the intervention. The intervention lasted for 3 years, and thus it was an important part of the long-term secondary prevention like Plüss et al.7 Also, the outcomes was collected at the end of the 3-year intervention and do not hold any information about the long-term effects.8

strengths, limitations and external value of the studyOne of the strengths of the present study is the complete follow-up. This is partly because the patients were iden-tified by their unique personal 10-digit civil registration number and partly because of the use of highly valid Danish register data. The information concerning mortality and morbidity were registered by health professionals using ICD-10 and did thus not rely on the memory of patients or relatives. Another strength is that the patients were almost similar at baseline. The only variables with consid-erable variation were educational level and whether the patients lived alone. This could be explained by these variables defining whether patients were socially vulner-able or not. It should, however, be noted that smoking status and the presence of hyperlipidaemia also varied.

The fact that patients from non-parallel time periods were being compared raises some methodological issues. All analyses were performed on both the socially and non-socially vulnerable patients. A difference between the non-socially vulnerable patients could have indicated that any changes among the socially vulnerable patients were just a general development in risk management and secondary prevention. However, no significant differ-ences were found.

The present study was carried out as a prospective cohort study and not as an RCT, thus there is a risk of confounding and bias. An attempt to minimise poten-tial confounding was made by using logistic regression analysis. Potential information bias cannot be ruled out concerning the self-reported questionnaires. However, it must be expected that potential bias must be non-dif-ferentiated and thereby changing the results towards the null hypothesis. A risk of selection bias could occur as attendance rates were significantly higher in the time period of the intervention than in the period where the control group received standard CR. If more highly socially vulnerable patients participated in the inter-vention, then it could be difficult to see any significant results of the intervention if they were compared with the low-risk part of the socially vulnerable patients in the group receiving standard CR.

A reason that no significant changes were found between the socially vulnerable patients receiving expanded CR and the ones receiving standard CR could be that standard CR is an evidence-based, structured and multidisciplinary intervention of high quality that any significant changes due to the expanded CR would be hard to detect. The mean age of the patients were around late 50s. Any changes in hard endpoints such as mortality and non-fatal recurrent events could be lacking, because it must be expected that the patients have had an unhealthy life style for many years resulting in severe

irreversible atherosclerosis. Also, the non-significant results could indicate the importance of phase III CR. More focus should be placed on supporting the patients in the long-term CR similar to the study by Plüss et al7 and trying to maintain and strengthen the knowledge that the patients obtain during phase II CR.

The external validity of the present study could be applied to CR in a hospital setting in most western coun-tries, especially countries with free healthcare and a wide access to social services.

Future researchFuture research should focus on why it was not possible to lower the mortality and morbidity significantly among socially vulnerable patients admitted with first episode MI. The authors suggest at least three plausible expla-nations that could be helpful when designing new inter-ventions. (1) Maybe it is not possible to lower social inequality in mortality and morbidity by using socially differentiated interventions. (2) Maybe the expanded CR should have focused on other things such as stress reduction, mindfulness or coping like it was the case in Plüss et al7 and in another recently published RCT focusing on stress management training.19 (3) Perhaps the intensity and the time frame were wrong. In Plüss et al,7 the expanded intervention lasted 1 year, and the patients therefore received support in phase II and in phase III as a part of the long-term secondary preven-tion.7 In order to minimise the costs and maximise the benefit of a more intense and longer CR programme, alternate low-resource settings and interventions such as digital devices and home-based CR must be consid-ered as well as a focus on those patients who will benefit mostly on participation.20 21

COnClusIOnDespite the significantly improved results of the socially differentiated CR intervention at 1-year follow-up, no long-term significant effects were seen regarding mortality and non-fatal recurrent events at follow-up after 10 years.

Contributors All authors contributed to the conception and design of the work. All authors contributed to acquisition, analysis and interpretation of data. KH and MLL drafted the manuscript. KMN, LKM, FBL, BC and CVN critically revised the manuscript. All authors approved the final version and agree to be accountable for all aspects of work ensuring integrity and accuracy.

Funding The authors disclosed received financial support for the research, authorship and/or publication of this article by: Aarhus University (17117581), Central Denmark Region (A-111 and 1-15-1-72-13-09), The Health Foundation (16-13-0098), The Committee of Multipractice Studies in General Practice (16-1461) and Trygfonden (119795).

Competing interests None declared.

Patient consent Obtained.

ethics approval The study was approved by The Danish Data Protection Agency (case number: 1-16-02-684-14). Ethical approval is not required for register-based studies in Denmark.

Provenance and peer review Not commissioned; externally peer reviewed.

Data sharing statement No additional data available.

on 19 Decem


http://bmjopen.bm

j.com/

BM



nloaded from



Open Access

Open Access This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http:// creativecommons. org/ licenses/ by- nc/ 4. 0/

© Article author(s) (or their employer(s) unless otherwise stated in the text of the article) 2018. All rights reserved. No commercial use is permitted unless otherwise expressly granted.

reFerenCes 1. Piepoli MF, Hoes AW, Agewall S, et al. European guidelines on

cardiovascular disease prevention in clinical practice. Rev Esp Cardiol 2016;2016:1–78.

2. Piepoli MF, Corrà U, Dendale P, et al. Challenges in secondary prevention after acute myocardial infarction: A call for action. Eur J Prev Cardiol 2016;23:1994–2006.

3. Smolina K, Wright FL, Rayner M, et al. Long-term survival and recurrence after acute myocardial infarction in England, 2004 to 2010. Circ Cardiovasc Qual Outcomes 2012;5:532–40.

4. Anderson L, Oldridge N, Thompson DR, et al. Exercise-based cardiac rehabilitation for coronary heart disease: Cochrane systematic review and meta-analysis. J Am Coll Cardiol 2016;67:1–12.

5. Lavie CJ, Menezes AR, De Schutter A, et al. Impact of cardiac rehabilitation and exercise training on psychological risk factors and subsequent prognosis in patients with cardiovascular disease. Can J Cardiol 2016;32:S365–373.

6. Janssen V, De Gucht V, Dusseldorp E, et al. Lifestyle modification programmes for patients with coronary heart disease: a systematic review and meta-analysis of randomized controlled trials. Eur J Prev Cardiol 2013;20:620–40.

7. Plüss CE, Billing E, Held C, et al. Long-term effects of an expanded cardiac rehabilitation programme after myocardial infarction or coronary artery bypass surgery: a five-year follow-up of a randomized controlled study. Clin Rehabil 2011;25:79–87.

8. Giannuzzi P, Temporelli PL, Marchioli R, et al. Global secondary prevention strategies to limit event recurrence after myocardial infarction: results of the GOSPEL study, a multicenter, randomized controlled trial from the Italian cardiac rehabilitation network. Arch Intern Med 2008;168:2194–204.

9. Mikkelsen T, Korsgaard Thomsen K, Tchijevitch O. Non-attendance and drop-out in cardiac rehabilitation among patients with ischaemic heart disease. Dan Med J 2014;61:1–5.

10. Beauchamp A, Peeters A, Tonkin A, et al. Best practice for prevention and treatment of cardiovascular disease through an equity lens: a review. Eur J Cardiovasc Prev Rehabil 2010;17:599–606.

11. Laustsen S, Hjortdal VE, Petersen AK. Predictors for not completing exercise-based rehabilitation following cardiac surgery. Scand Cardiovasc J 2013;47:344–51.

12. Mårtensson S, Gyrd-Hansen D, Prescott E, et al. Does access to invasive examination and treatment influence socioeconomic differences in case fatality for patients admitted for the first time with non-ST-elavation myocardial infarction or unstable angina? EuroIntervention 2015;11:1–10.

13. Molshatzki N, Drory Y, Myers V, et al. Role of socioeconomic status measures in long-term mortality risk prediction after myocardial infarction. Med Care 2011;49:673–8.

14. Korda RJ, Soga K, Joshy G, et al. Socioeconomic variation in incidence of primary and secondary major cardiovascular disease events: an Australian population-based prospective cohort study. Int J Equity Health 2016;15:1–10.

15. Stirbu I, Looman C, Nijhof GJ, et al. Income inequalities in case death of ischaemic heart disease in the Netherlands: a national record-linked study. J Epidemiol Community Health 2012;66:1159–66.

16. Rasmussen JN, Rasmussen S, Gislason GH, et al. Mortality after acute myocardial infarction according to income and education. J Epidemiol Community Health 2006;60:351–6.

17. Notara V, Panagiotakos DB, Papataxiarchis E, et al. Depression and marital status determine the 10-year (2004-2014) prognosis in patients with acute coronary syndrome: The GREECS study. Psychol Health 2015;30:1116–27.

18. Nielsen KM, Meillier LK, Larsen ML. Extended cardiac rehabilitation for socially vulnerable patients improves attendance and outcome. Dan Med J 2013;60:1–5.

19. Blumenthal JA, Sherwood A, Smith PJ, et al. Enhancing cardiac rehabilitation with stress Management training. A randomized, clinical efficacy trial. Circulation 2016;133:1341–50.

20. Grace SL, Turk-Adawi KI, Contractor A, et al. Cardiac rehabilitation delivery model for low-ressource settings: an international council of cardiovascular prevention and rehabilitation consensus statement. Prog Cardiovasc Dis 2016;59:303–22.

21. Kachur S, Chongthammakun V, Lavie CJ, et al. Impact of cardiac rehabilitation and exercise training programs in coronary heart disease. Prog Cardiovasc Dis 2017;60:103–14.

on 19 Decem


http://bmjopen.bm

j.com/

BM



nloaded from

http://creativecommons.org/licenses/by-nc/4.0/

http://creativecommons.org/licenses/by-nc/4.0/

http://dx.doi.org/10.1177/2047487316663873

http://dx.doi.org/10.1177/2047487316663873

http://dx.doi.org/10.1161/CIRCOUTCOMES.111.964700

http://dx.doi.org/10.1016/j.jacc.2015.10.044

http://dx.doi.org/10.1016/j.cjca.2016.07.508

http://dx.doi.org/10.1016/j.cjca.2016.07.508

http://dx.doi.org/10.1177/2047487312462824

http://dx.doi.org/10.1177/2047487312462824

http://dx.doi.org/10.1177/0269215510376006

http://dx.doi.org/10.1001/archinte.168.20.2194

http://dx.doi.org/10.1001/archinte.168.20.2194

http://dx.doi.org/10.1097/HJR.0b013e328339cc99

http://dx.doi.org/10.3109/14017431.2013.859295

http://dx.doi.org/10.3109/14017431.2013.859295

http://dx.doi.org/10.1097/MLR.0b013e318222a508

http://dx.doi.org/10.1186/s12939-016-0471-0

http://dx.doi.org/10.1186/s12939-016-0471-0

http://dx.doi.org/10.1136/jech-2011-200924

http://dx.doi.org/10.1136/jech.200X.040972

http://dx.doi.org/10.1136/jech.200X.040972

http://dx.doi.org/10.1080/08870446.2015.1034720

http://dx.doi.org/10.1080/08870446.2015.1034720

http://dx.doi.org/10.1161/CIRCULATIONAHA.115.018926

http://dx.doi.org/10.1016/j.pcad.2016.08.004

http://dx.doi.org/10.1016/j.pcad.2017.07.002


Paper II

1

Medication adherence, biological and lifestyle risk factors in patients with


rehabilitation

Accepted for publication December 2018: Scandinavian Journal of Primary Health Care

Kathrine Hald1, Finn Breinholt Larsen

2, Kirsten Melgaard Nielsen

3, Lucette Kirsten Meillier

2,

Martin Berg Johansen4, Mogens Lytken Larsen

5, Bo Christensen

6 & Claus Vinther Nielsen

1.

1Section for Clinical Social Medicine and Rehabilitation, Department of Public Health, Aarhus

University, Aarhus, Denmark

2 DEFACTUM, Social and Health Services and Labour Market, Central Denmark Region,

Aarhus, Denmark

3Department of Cardiology, Aarhus University Hospital, Aarhus, Denmark

4 Unit of Clinical Biostatistics, Aalborg University Hospital, Aalborg, Denmark

5 Danish Centre for Inequality in Health, Department of Cardiology, Aalborg University

Hospital, Aalborg, Denmark

6Section for General Medical Practice, Department of Public Health, Aarhus University, Aarhus,

Denmark

Corresponding author: Kathrine Hald, Section for Clinical Social Medicine and Rehabilitation,

Department of Public Health, Aarhus University, Aarhus, P. P. Oerums Gade 9-11, Building 1B,

8000 Aarhus C, Denmark.

Email: [email protected]

Words: 3168. 2 tables and 3 figures.

Short running title: Socially differentiated cardiac rehabilitation

2

Key points

The socially differentiated intervention did not significantly improve medication adherence

or biological and lifestyle risk factors.

Despite the non-significant effect of the intervention, equality in health was improved except

concerning smoking.

General practitioners managed to support the long-term secondary cardiovascular disease

prevention in all patients regardless of social status.

3

Abstract

Objective

There is strong evidence that medication adherence and lifestyle changes are essential in patients

undergoing secondary cardiovascular disease prevention. Cardiac rehabilitation (CR) increases

medication adherence and improves lifestyle changes. Patients with cardiac diseases and a low

educational level and patients with little social support are less responsive to improve medication

adherence and to adapt lifestyle changes. The aim of the present study was to investigate the

long-term effects of a socially differentiated CR intervention on medication adherence as well as

changes in biological and lifestyle risk factors at two- five- and ten-year follow-up.

Design

A prospective cohort study.

Setting

The cardiac ward at Aarhus University Hospital, Denmark.

Intervention

A socially differentiated CR intervention in addition to the standard CR program.

Subjects

Patients admitted with first-episode myocardial infarction between 2000 and 2004, N=379.

Patients were defined as socially vulnerable or non-socially vulnerable according to their

educational level and extent of social network.

Main outcome measures

Primary outcome was medication adherence to antithrombotics, beta-blockers, statins and

angiotensin-converting enzyme inhibitors. Secondary outcomes were biological and lifestyle risk

factors defined as; total cholesterol, low-density lipoprotein cholesterol, high-density lipoprotein

cholesterol, glycated hemoglobin, blood pressure and smoking status.

Results

No significant long-term effect of the intervention was found.

Conclusions

The results indicate a non-significant effect of the intervention. However, it was found that

equality in health was improved in the study population except concerning smoking. General

practitioners manage to support the long-term secondary cardiovascular disease prevention in all

patients regardless of social status.

4

Introduction

The '2016 European Guidelines on cardiovascular disease prevention in clinical practice'

emphasizes that medication adherence and lifestyle changes are essential in secondary

cardiovascular disease prevention [1]. A core component of this is cardiac rehabilitation (CR),

which is a structured and multidisciplinary intervention [2]. Several studies have shown that CR

increases medication adherence and adherence to lifestyle changes [3-8]. Long-term adherence

to medication and lifestyle is crucial and associated with a lower risk of mortality and recurrent

events [1,2,9,10]. Key strategies to improve adherence include a strong relationship between

clinician and patient, a safe transition from hospital to primary care and communication skills to

ensure the patient’s understanding of the risks of non-adherence [11].

Patients with a low educational level and little social support are less responsive to achieving

medication and lifestyle changes when diagnosed with cardiovascular disease (CVD) [12-16].

Even in countries with equal and free access to health care, social inequalities are observed,

implying that new initiatives are needed in secondary cardiovascular disease prevention [17].

In a Danish socially differentiated CR intervention performed from 2000 to 2004, the focus was

on minimizing social inequality in patients diagnosed with myocardial infarction (MI). At one-

year follow-up, socially vulnerable patients receiving the intervention had a significantly better

medication adherence, a significantly better lipid profile and a significantly lower systolic blood

pressure (BP) when compared to socially vulnerable patients receiving standard CR [18].

The aim of the present study was to investigate the long-term effects of the above mentioned

socially differentiated CR intervention on medication adherence as well as changes in biological

and lifestyle risk factors at two- five- and ten-year follow-up.

5

Material and methods

Design and study participants

The study was designed as a prospective cohort study and conducted from 1 April 2000 to 31

December 2004. The study population was enrolled at admission to hospital and follow-up was

conducted at two, five and ten years. The patients entering the study were all < 70 years,

admitted at Aarhus University Hospital, Denmark and diagnosed with first episode of MI.

Patients were excluded if they suffered from severe comorbidities such as stroke, dementia,

mental disorders (not depression or anxiety), retardation or severe alcohol abuse [18,19].

The study was divided into two phases which consisted of two years observation of clinical

practice regarding standard CR from 2000 to 2002 and followed by a two-year intervention from

2002 to 2004. A total of 508 patients were admitted with first episode MI and of these 379

patients (75 %) who were offered and attended CR were included in the study and provided

written informed consent [18,19].

When entering the study, patients were defined as socially vulnerable or non-socially vulnerable.

If patients had a low educational level or lived alone they were defined as socially vulnerable. 78

patients admitted between 2000 and 2002 and 130 patients admitted between 2002 and 2004 met

the above criteria and were categorized as socially vulnerable. The remaining 171 patients, 55

patients admitted between 2000 and 2002 and 116 patients admitted between 2002 and 2004,

were categorized as non-socially vulnerable. Further details of the definition have been described

previously [18,19].

Intervention

All 379 patients received standard CR in accordance with international guidelines [1]. In CR

phase I from hospital admission to discharge, patients received medical and acute surgical

treatment. In CR phase II from discharge and the next 12 weeks, patients had three consultations

with a doctor, four consultations with a nurse, two consultations with a dietician and participated

in 12-week exercise program. In CR phase III, patients were referred to general practice and

informed about relevant activities in The Danish Heart Association and in the municipal sector

[18,19].

6

The 130 patients categorized as socially vulnerable and admitted between 2002 and 2004

received an expanded CR intervention in addition to the standard program. The expanded CR

intervention was two week longer in phase II and contained an extra consultation with a nurse.

The patients played an active part in designing their own rehabilitation plan which was sent to

their general practitioner. The patients were referred to a phase III start-up consultation in

general practice which was based on their individual rehabilitation plan. The patients were

referred to activities in The Danish Heart Association and the municipal sector. Moreover,

patients received a follow-up telephone call from a nurse three months after completing phase II

CR [18,19].

Outcomes

Primary outcome was medication adherence to antithrombotics, beta-blockers, statins and

angiotensin-converting enzyme inhibitors (ACE inhibitors). Secondary outcomes were biological

and lifestyle risk factors defined as; total cholesterol, low-density lipoprotein (LDL) cholesterol,

high-density lipoprotein (HDL) cholesterol, glycated hemoglobin (HbA1c), blood pressure and

smoking status.

Data collection and measures

In Denmark, citizens are assigned a unique personal 10-digit number, which was used in the data

collection. Information on medication adherence was obtained from The Danish National

Prescription Register [20]. The database contains information on all prescription drugs sold in

Denmark. Medication adherence was defined as the purchase of at least one prescription every

year of the follow-up period with the specific ATC-codes B01AC04, B01AC06, B01AC56 for

anti-thrombotics, C07AA, C07AB for beta-blockers, C10AA, C10AB, C10AC, C10AD,

C10AX09, C10BA for statins and C09AA, C09BB, C09CA, C09D for ACE inhibitors.

Information on total cholesterol, LDL-cholesterol, HDL-cholesterol, triglyceride and HbA1c was

obtained from a local laboratory database 'LABKA' containing information on results of all

blood tests performed in both the primary and secondary sector in the region where Aarhus

University Hospital is located [21]. Cholesterol and triglyceride were measured as mmol/L and

HbA1c was measured as mmol/mol. The value for each of the laboratory outcomes at the time of

follow-up was an average value computed from all the values obtained in the follow-up period.

7

Information on blood pressure and smoking status was collected through a questionnaire sent to

the patient's general practitioner. In Denmark, patients diagnosed with MI are entitled to an

annual chronic care consultation in general practice [22]. The general practitioner was informed

about the patient's admission date and provided information on blood pressure and smoking

status at the annual consultation. Blood pressure was measured as mm/Hg and smoking status

was reported as smoker / non-smoker. The questionnaire data were typed into a data

documentation program by two different evaluators. All answers were assessed and if any

dissimilarities occurred, the questionnaires were reevaluated by both evaluators.

Statistical analysis

The baseline characteristics of patients in each group are described using frequencies and

percentages or means and standard deviations as appropriate. Evaluations of the primary and

secondary outcomes were performed as visualizations of the outcome measures during follow-

up, and as statistical tests at follow-up between socially vulnerable patients receiving the

standard CR versus the expanded CR. To evaluate the potential differences between the two

calendar periods, a supplemental analysis of the non-socially vulnerable patients in the two

corresponding time periods was also conducted. All data were based on yearly survivors in the

study population. Data on medication adherence were shown as yearly proportions of patients

who had redeemed at least one prescription for each drug. The 95% confidence intervals (CIs) of

the proportions at follow-up were also calculated. Comparisons were evaluated using a chi-

square test of independence. The biomarker data were based on all blood samples collected

through LABKA and shown as medians in visualizations and compared by calculating

differences in means with 95% CIs and using a t test to test for equality of means. The

questionnaire data were summarized and evaluated by calculating means and proportions with

the corresponding tests as described for the medication and biomarker data. All data

management and analyses were performed using Stata/MP 14.2, and P-values less than 0.05

were considered statistically significant.

8

Results

Baseline characteristics of the study population can be seen in Table 1. The mean age was 57

years and approximately three out of four were male. As a result of the definition of socially

vulnerable patients had a lower educational level and a higher fraction was living alone. Total

cholesterol, fasting blood glucose, body mass index and smoking status were quite similar in all

groups, except in the group of socially vulnerable patients admitted between 2000 and 2002 who

had higher values and smoked more. Patients admitted between 2000 and 2002 were prescribed

ACE inhibitors and statins less often than patients admitted between 2002 and 2004 regardless of

social status.

Medication adherence

As indicated in Figure 1, adherence to anti-thrombotics during the ten-year follow-up was higher

than 80 %. This was also the case in relation to statins with one exception, as the non-socially

vulnerable patients admitted between 2000 and 2002 showed a steady above 60 % adherence

during the ten-year follow-up. Adherence to beta-blockers was higher in the groups admitted

between 2000 and 2002 than in the groups admitted between 2002 and 2004 irrespective of

whether the patients were categorized as socially vulnerable or not. Adherence to ACE inhibitors

was around 40-60 % in all groups throughout the 10-year follow-up.

In relation to anti-thrombotics and ACE-inhibitors, no significant differences in adherence were

seen between the socially vulnerable patients at two-, five- and ten-year follow-up. In relation to

beta-blockers, a significant difference was seen at two- and five-year follow-up (P-values 0.00

and 0.02), showing that significantly more socially vulnerable patients receiving standard CR

redeemed at least one prescription on beta-blockers in each of the follow-up years. No significant

difference was seen at ten-year follow-up. In relation to statins, a significant difference was seen

at two-year follow-up (P-value 0.04). Significantly more socially vulnerable patients receiving

the expanded CR purchased at least one prescription on statins in the follow-up year. No

significant differences were seen at five- or ten-year follow-up (Table 2). As visualized in Figure

1, some of the same tendencies were observed in the group of non-socially vulnerable patients.

9

Biological and lifestyle risk factors

As seen in Figure 2, the levels of the blood tests were acceptable. In general, the patients

admitted between 2000 and 2002 had less acceptable blood test levels in the first years after

baseline regardless of being categorized as socially vulnerable or not. This effect was less clear

during the last part of the follow-up.

Socially vulnerable patients receiving the expanded CR intervention had significantly lower total

cholesterol (P-values 0.00, 0.00), LDL cholesterol (P-values 0.00,0.00), triglyceride (P-values

0.00,0.04) and HbA1c levels (P-values 0.00,0.01) at two- and five- year follow-up. No

significant differences were seen at ten-year follow-up. Socially vulnerable patients receiving the

expanded CR intervention showed significantly higher HDL cholesterol at two-year follow-up

(P-value 0.01). No significant differences were seen at five- and ten-year follow-up (Table 2). As

visualized in Figure 2, some of the same tendencies were present in the group of non-socially

vulnerable patients. However, no significant differences were seen in relation to triglyceride.

Of the 379 forwarded questionnaires, 301 were returned by the general practitioners (response

rate 79 %). Not all returned questionnaires were filled out completely. The response rate in the

group of socially vulnerable patients admitted between 2000 and 2002 was 77 % and it was 81 %

in the group of socially vulnerable patients admitted between 2002 and 2004 (P-value 0.50). The

response rate in the group of non-socially vulnerable patients admitted between 2000 and 2002

was 69 % and it was 84 % in the group of non-socially vulnerable patients admitted between

2002 and 2004 (P-value 0.02).

As seen in Figure 3, patients had acceptable mean blood pressure levels regardless of being

categorized as socially vulnerable or not. No mean blood pressure levels of more than 135/85

mm/Hg were seen. No significant differences in systolic or diastolic blood pressure were seen

between socially vulnerable patients receiving the standard CR and socially vulnerable patients

receiving the expanded CR during the ten-year follow-up (Table 2). As visualized in Figure 3,

some of the same tendencies were seen in the group of non-socially vulnerable patients.

Figure 3 shows that a higher proportion of socially vulnerable patients than non-socially

vulnerable patients smoked. The proportion of smokers decreased during the first two years after

admission in all groups when compared to baseline data (Table 1). The proportion of smokers

increased from two to five years after admission and declined again in each of the groups during

10

the remaining part of the ten-year follow-up. No significant differences in smoking status were

seen between socially vulnerable patients receiving the standard CR and socially vulnerable

patients receiving the expanded CR during the ten-year follow-up (Table 2). As visualized in

Figure 3, the same tendency was present in the group of non-socially vulnerable patients.

11

Discussion

The present study investigated the long-term effects of a socially differentiated CR intervention

on medication adherence, biological and lifestyle risk factors at two- five- and ten-year follow-

up. Overall, no significant long-term effects were found. However, significantly more patients

categorized as socially vulnerable receiving the expanded CR redeemed at least one prescription

on statins at the two-year follow-up. Also, socially vulnerable patients receiving the expanded

CR had significantly lower levels of total cholesterol, LDL cholesterol, triglyceride and HbA1c

at two- and five-year follow-up and a significantly higher level of HDL cholesterol at two-year

follow-up.

Strengths and weaknesses

A strength of the study is that the majority of the data are retrieved from Danish registers, which

can be assumed to provide an almost complete follow-up. The risk of selection bias in data

collected through questionnaires should, however, always be considered. As the questionnaires

were filled out by the patients' general practitioners we have no reasons to believe that selection

bias was present. Also, no significant differences were seen in the response rate between the

socially vulnerable patients receiving either the standard or expanded CR.

The ten-year follow-up period is a major strength. To our knowledge, no previous studies have

investigated the ten-year effect of a socially differentiated CR intervention. However, it must be

considered that the effect of the intervention perhaps was watered down over the years and that

other non-measurable competing reasons could have influenced the long-term effect.

Another strength is that even though we do not show any convincing effect of the intervention,

we do show that it is possible to tailor a long-term secondary cardiovascular disease prevention

which improves equality in health. In Denmark, the general practitioner is the main care provider

in long-term secondary cardiovascular disease prevention. Our results show that general

practitioners manage to successfully support the socially vulnerable patients except when it

comes to smoking. However, the proportion of smokers increased from two to five years after

admission regardless of social status indicating that some lifestyle changes are more difficult to

maintain in the long-term secondary prevention.

12

A weakness of the study is the external validity. It can be difficult to apply the results to

countries without free access to health care and countries which do not offer reimbursement of

medicine costs. However, all CR programs offered in Europe to patients suffering from CVD

must be expected to comply with the recommendations by 'The European Society of Cardiology'

[1].

It is also a weakness that data are based on yearly survivors if there was any selection bias in the

deaths. However, it has earlier been established that no significant differences in all-cause

mortality were seen at ten-year follow-up between the socially vulnerable patients receiving the

standard CR and the socially vulnerable patients receiving the expanded CR [19]. Moreover, it is

a weakness if the socially vulnerable patients included in the study are different than the socially

vulnerable patients who were not referred to or did not participate in CR. If the socially

vulnerable patients who participated were the ones with most resources and the most healthy,

there is a risk of selection bias and type-2-errors.

In addition, it seems that time plays a part in the results. One could suspect that beta-blockers

were more frequently used between 2000 and 2002 and that statins were more frequently used

between 2002 and 2004. Also, it seems that the development in the blood test levels could be a

sign that the threshold values have been lowered over the years. Reflections concerning the non-

parallel time periods in this study and the non-randomized study design have been discussed in

detail elsewhere [19].

Findings in relation to other studies

In an Italian randomized controlled trial from 2012 by Sturchio et al, the aim was to evaluate the

effectiveness of an individualized management program to modify the risk profile in patients

with coronary artery disease [8]. At nine-month follow-up a significant difference was seen

between patients randomized to the intervention and patients randomized to standard care in

relation to total – and LDL cholesterol, triglyceride, systolic blood pressure, number of smokers

and adherence to beta-blockers, ACE inhibitors and statins. The intervention by Sturchio et al [8]

was significantly more effective than standard care. This could indicate that the intervention

offered was more effective than the socially differentiated intervention in the present study.

However, it must be emphasized that the follow-up time in the study by Sturchio et al [8] was

short. Also, there is no information in the study about the patients' social status. If only non-

13

socially vulnerable patients participated it could cause selection bias and thus reduce the validity

of the study. Moreover, it must be considered that the reason why we do not see any significant

differences is that standard CR ('usual care') in Denmark has such a high quality that it may be

difficult to detect any differences between a standard and an expanded intervention.

In a British systematic review and qualitative synthesis from 2014 by Rashid et al [11], the aim

was to understand the factors that promote medication persistence seen from a patient

perspective. It was found that the support of family members is important, which goes well in

line with the definition in this study of being socially vulnerable with little social support or not.

Moreover, it was found that a good relationship between the patient and the prescribing clinician

is very important, which in Denmark almost always will be the general practitioner. This

supports our previous statement that general practitioners seemingly manage to support the

socially vulnerable patients so well that equality in medication adherence and biological and

lifestyle risk factors were improved. Furthermore, it was stated by Rashid et al [11] that patients

believe that medicine is more powerful than life-style changes. This is applicable to the results in

the present study concerning smoking.

14

Conclusion and implications

The present study did not find any significant long-term effects of the socially differentiated CR

intervention. Overall, it was found that patients regardless of being categorized as socially

vulnerable or not were adherent to prescribed medicine and that their biological and lifestyle risk

factors were acceptable at follow-up. The authors acknowledge the general practitioners' effort in

supporting the patients in the long-term secondary prevention and equality in health was

improved in the socially vulnerable part of the study population compared to existing literature

where socially vulnerable patients to a lesser extent achieve medication adherence and life style

changes [12-17].A major challenge is how to reduce the proportion of smokers in the long-term

secondary prevention where equality in health was not improved.

15

Acknowledgements

We highly acknowledge the participation of general practitioners for providing the data from

questionnaires to this study.

Ethical approval

The Danish Data Protection Agency approved the study (Case number: 1-16-02-684-14). No

ethical approval was required.

Disclosure statement

The authors declare that they have no competing interests.

Funding

This work was supported by: Aarhus University (Grant number: 17117581), Central Denmark

Region (Grant number: A-111, 1-15-1-72-13-09), The Health Foundation (Grant number: 16-13-

0098), The Committee of Multipractice Studies in General Practice (Grant number: 16-1461) and

TrygFonden (Grant number: 119795).

16

References

[1] Piepoli MF, Hoes AW, Agewall S, Albus C, Brotons C, Catapano et al. 2016 European

guidelines on cardiovascular disease prevention in clinical practive. G Ital Cardiol. 2017;18:547-

612.

[2] Piepoli MF, Corra U, Dendale P, Frederix I, Prescott E, Schmid JP. Challenges in secondary

prevention after acute myocardial infarction: A call for action. Eur Heart J Acute Cardiovasc

Care. 2017;6:299-310.

[3] Sharma KK, Gupta R, Mathur M, Natani V, Lodha S, Roy S et al. Non-physician health

workers for improving adherence to medications and healthy lifestyle following acute coronary

syndrome: 24-month follow-up study. Indian Heart J. 2016;68:832-40.

[4] Anderson L, Oldridge N, Thompson DR, Zwisler AD, Rees K, Martin N et al. Exercise-based

cardiac rehabilitation for coronary heart disease. J Am Coll Cardiol. 2016;67:1-12.

[5] McMahon SR, Ades PA, Thompson PD. The role of cardiac rehabilitation in patients with

heart disease. Trends Cardiovasc Med. 2017;27:420-25.

[6] Ho PM, Lambert-Kerzner A, Carey EP, Fahdi IE, Bryson CL, Melnyk SD et al. Multifaceted

intervention to improve medication adherence and secondary prevention measures after acute

coronary syndrome hospital discharge: A randomized clinical trial. JAMA Intern Med.

2014;174:186-93.

[7] Choudhry NK, Glynn RJ, Avorn J, Lee JL, Brennan TA, Reisman L et al. Untangling the

relationship between medication adherence and post-myocardial infarction outcomes:

Medication adherence and clinical outcomes. AM Heart J. 2014;167:51-58.

[8] Sturchio A, Gianni AD, Campana B, Genua M, Storti M, Lasi GD et al. Coronary artery risk

management programme delivered by a rehabilitation day-hospital. J Cardiopulm rehabil Prev.

2012;32:386-93.

[9] Rea TD, Heckbert SR, Kaplan RC, Smith NL, Lemaitre RN, Psaty BM. Smoking status and

risk for recurrent coronary events after myocardial infarction. Ann Intern Med. 2002;137:494-

500.

[10] Fanari Z, Malodiya A, Weiss SA, Hammami S, Kolm P, Weintraub WS. Long-term use of

dual antiplatelet therapy for the secondary prevention of atherothrombotic events: Meta-analysis

of randomized controlled trials. Cardiovasc Revasc Med. 2017;18:10-15.

17

[11] Rashid MA, Edwards D, Walter FM, Mant J. Medication taking in coronary artery disease:

A systematic review and qualitative synthesis. Ann Fam Med. 2014;12:224-32.

[12] Chan RHM, Gordon NF, Chong A, Alter DA. Influence of socioeconomic status on lifestyle

behavior modifications among survivors of acute myocardial infarction. Am J Cardiol.

2008;102:1583-8.

[13] Castellano JM, Sanz G, Penalvo JL, Bansilal S, Fernandez-Ortiz A, Alvarez L et al. A

polypill strategy to improve adherence: Results from the FOCUS project. J Am Coll Cardiol.

2014;64:2071-82.

[14] Bruthans J, Mayer O, Bacquer DD, Smedt DD, Reiner Z, Kotseva K et al. Educational level

and risk profile and risk control in patients with coronary heart disease. Eur J Prev Cardiol.

2016;23:881-90.

[15] Berndt N, Bolman C, Mudde A, Verheugt F, Vries HD, Lechner L. Risk groups and

predictors of short-term abstinence from smoking in patients with coronary heart disease. Heart

Lung. 2012;41:332-43.

[16] Gaalema DE, Elliot RJ, Morford ZH, Higgins ST, Ades PA. Effect of socioeconomic status

on propensity to change risk behaviors following myocardial infarction: Implications for healthy

lifestyle medicine. Prog Cardiovasc Dis. 2017;60:159-68.

[17] Nielsen JB, Leppin A, Gyrd-Hansen D, Jarbøl DE, Søndergaard J, Larsen PV. Barriers to

lifestyle changes for prevention of cardiovascular disease – a survey among 40-60-year old

Danes. BMC Cardiovasc Disord. 2017;17:1-8.

[18] Nielsen KM, Meillier LK, Larsen ML. Extended cardiac rehabilitation for socially

vulnerable patients improves attendance and outcome. Dan Med J. 2013;60:1-5.

[19] Hald K, Nielsen KM, Nielsen CV, Meillier LK, Larsen FB, Christensen et al. Expanded

cardiac rehabilitation in socially vulnerable patients with myocardial infarction: A 10-year

follow-up study focusing on mortality and non-fatal events. BMJ Open. 2018;8:1-8.

[20] Kildemoes HW, Sørensen HT, Hallas J. The Danish National Prescription Registry. Scand J

Public Health. 2011;39:38-41.

[21] Grann AF, Erichsen R, Nielsen AG, Frøslev T, Thomsen RW. Existing data sources for

clinical epidemiology: The clinical laboratory information system (LABKA) research database at

Aarhus University, Denmark. Clin Epidemiol. 2011;3:133-138.

[22] Dansk Selskab for Almen Medicin. Årskontroller. Indhold og praktisk vejledning. Available

at: https://vejledninger.dsam.dk/hjerte/?mode=visKapitel&cid=1072#chapter_1072 (accessed 3

October 2018.

18

Legends

Table 1. Baseline characteristics of 379 patients admitted with first-episode myocardial

infarction receiving socially differentiated cardiac rehabilitation. Patients are divided into groups

based on social vulnerability and time period of admission.

Table 2. Assessment of medication adherence and biological and lifestyle risk factors among

socially vulnerable patients admitted between 2000 and 2002 (N=78) and between 2002 and

2004 (N=130) at Aarhus University Hospital, Denmark with first-episode myocardial infarction

who participated in socially differentiated cardiac rehabilitation intervention and who were

followed-up at two, five and ten years. Values are based on yearly survivors and on available

data from registers and questionnaires.

Figure 1. Proportions of patients redeeming at least one prescription for anti-thrombotics, beta-

blockers, statins and ACE inhibitors each year after first-episode myocardial infarction

admission by groups of social vulnerability and calendar period of admission. Proportions are

based on all patients with a first admission at Aarhus University Hospital, Denmark between

2000 and 2004 (N=379) who survived each year of follow-up.

Figure 2. Median values of blood test results among patients each year after first-episode

myocardial infarction by groups of social vulnerability and calendar period of admission.

Medians are based on data from all patients with a first admission at Aarhus University Hospital,

Denmark between 2000 and 2004 (N=379) who have a valid register-based value of each type of

blood test within each year of follow-up.

Figure 3. Mean values and proportions of blood pressure and smoking status among patients

with first-episode myocardial infarction admission by groups of social vulnerability and calendar

period of admission. Values are based on questionnaire data from general practitioners of all

patients with a first admission at Aarhus University Hospital, Denmark between 2000 and 2004

(N=379) with a valid questionnaire response at each time of follow-up.

19


infarction receiving socially differentiated cardiac rehabilitation. Patients are divided into

groups based on social vulnerability and time period of admission.

Socially vulnerable participants

Non-socially vulnerable participants


Time period


Time period

Standard Rehabilitation

N=78

2000-2002

Expanded Rehabilitation

N=130

2002-2004

P

Value Standard

Rehabilitation N=55

2000-2002

Standard

Rehabilitation N=116

2002-2004

P

Value


56 (8.2) 55 (8.5) 0.65 60 (7.6) 57 (73) 0.02

Gender, male 57 (73) 93 (71) 0.81 42 (76) 94 (81)

0.48 Education level

(DUN) 3.2 (1.2) 3.3 (1.4) 0.66 4.8 (1.1) 4.8 (1.2) 0.79

Living alone 27 (35) 51 (39) 0.51 0 0 -


5.7 (1.5) 5.2 (1.0) 0.00 5.2 (0.9) 5.2 (0.9) 0.88


3.5 (1.0) 3.2 (0.9) 0.02 3.3 (0.9) 3.2 (0.8) 0.58


1.2 (0.3) 1.2 (0.3) 0.13 1.3 (0.3) 1.3 (0.4) 0.89

Triglyceride, mmol/L 2.0 (1.1) 1.7 (1.0) 0.02 1.5 (1.0) 1.7 (0.9)

0.25

Fasting blood glucose, mmol/L

7.5 (4.6) 6.9 (2.8) 0.25 6.8 (3.3) 6.7 (2.0) 0.69

Body Mass Index 27.3 (4.4) 26.3 (4.1) 0.10 26.4 (4.0) 26.5 (3.1)

0.77 Prescribed beta-blocker

71 (91) 116 (89) 0.67 49 (89) 107 (92) 0.50

Prescribed ACE-inhibitor

24 (31) 55 (42) 0.09 20 (36) 49 (42) 0.46

Prescribed statin 20 (26) 99 (76) 0.00 10 (18) 104 (90) 0.00

Prescribed anti-thrombotics

72 (92) 126 (97) 0.13 47 (86) 112 (97) 0.01

Current smoker 59 (76) 83 (64) 0.28 34 (62) 60 (52) 0.29

20

Table 2. Assessment of medication adherence and biological and lifestyle risk factors

among socially vulnerable patients admitted between 2000 and 2002 (N=78) and between

2002 and 2004 (N=130) at Aarhus University Hospital, Denmark with first-episode

myocardial infarction who participated in socially differentiated cardiac rehabilitation

intervention and who were followed-up at two, five and ten years. Values are based on

yearly survivors and on available data from registers and questionnaires

Socially vulnerable participants 2000-2002 2002-2004

Year of follow-up

Proportion* Mean**

N Proportion*

Mean** N

Ratio* Difference**

P value

Anti-thrombotic

2 0.90* 78 0.95* 128 1.1* (0.9 - 1.1) 0.37 5 0.92* 75 0.89* 125 1.0* (0.9 - 1.1) 0.46

10 0.94* 66 0.90* 107 1.0* (0.9 - 1.0) 0.33

Beta-blockers

2 0.82* 78 0.55* 128 0.7* (0.6 - 0.8) 0.00

5 0.68* 75 0.52* 125 0.8 *(0.6 – 1.0) 0.02

10 0.58* 66 0.47* 107 0.8* (0.6 - 1.1) 0.16

Statins

2 0,82* 78 0.91* 128 1.1* (1.0 - 1.3) 0.04

5 0.88* 75 0.86* 125 1.0* (0.9 - 1.1) 0.74

10 0.91* 66 0.90* 107 1.0* (0.9 - 1.1) 0.79

ACE inhibitors

2 0.38* 78 0.47* 128 1.2* (0.9 - 1.7) 0.23

5 0.44* 75 0.53* 125 1.2* (0.9 - 1.6) 0.22

10 0.50* 66 0.56* 107 1.1* (0.8 - 1.5) 0.43

Total Cholesterol

2 5.1** 78 4.4** 130 -0.6 **(-0.8- -0.4) 0.00

5 4.7** 74 4.3** 116 -0.4 **(-0.6 - -0.2) 0.00

10 4.3** 68 4.2** 105 -0.1** (-0.3 - 0.2) 0.51

HDL Cholesterol

2 1.2** 78 1.3** 130 0.1** (0.1 - 0.2) 0.01

5 1.3** 74 1.3** 114 0.0 **(-0.1 - 0.1) 0.95

10 1.2** 68 1.3** 103 0.1 **(-0.0 - 0.2) 0.11

LDL Cholesterol

2 3.0** 77 2.5** 130 -0.5 **(-0.7 - -0.4) 0.00

5 2.6** 74 2.3** 111 -0.3 **(-0.4 - -0.1) 0.00

10 2.4** 68 2.3** 102 -0.1** (-0.3 - 0.1) 0.37

Triglyceride 2 2.0** 78 1.5** 130 -0.4 **(-0.7 - -0.2) 0.00

5 1.8** 74 1.6** 112 -0.2 **(-0.4 - -0.0) 0.04

10 1.7** 68 1.6** 103 -0.03** (-0.3 - 0.2) 0.80

HbA1c 2 50.7** 30 42.3** 117 -8.4 **(-12.6 - -4.2) 0.00

5 52.1** 40 45.7** 71 -6.4 **(-11.3 - -1.4) 0.01

10 48.9** 43 44.9** 93 -4.0** (-8.1 - 0.1) 0.06

Systolic blood pressure

2 131.8** 53 131.4** 93 -0.4** (-6.0 - 5.1) 0.89

5 132.7** 54 132.6** 96 -0.1** (-6.8 - 6.5) 0.97

10 134.1** 55 132.4** 92 -1.7** (-7.5 - 4.2) 0.57

Diastolic blood pressure

2 79.5** 53 79.0** 93 -0.5** (-3.9 - 2.9) 0.77

5 78.3** 54 79.0** 96 0.7** (-2.7 - 4.1) 0.68

10 80.0** 55 79.4** 92 -0.5** (-4.2 - 3.2) 0.78

Smoking status

2 0.52* 33 0.46* 70 0.9* (0.6 - 1.4) 0.58

5 0.59* 39 0.47* 68 0.8* (0.6 - 1.2) 0.23

10 0.37* 43 0.37* 75 1.0* (0.6 - 1.6) 0.98

21

Figure 1. Proportions of patients redeeming at least one prescription for anti-thrombotics,

beta-blockers, statins and ACE inhibitors each year after first-episode myocardial infarction

admission by groups of social vulnerability and calendar period of admission. Proportions

are based on all patients with a first admission at Aarhus University Hospital, Denmark

between 2000 and 2004 (N=379) who survived each year of follow-up.

22

Figure 2. Median values of blood test results among patients each year after first-episode

myocardial infarction by groups of social vulnerability and calendar period of admission.

Medians are based on data from all patients with a first admission at Aarhus University

Hospital, Denmark between 2000 and 2004 (N=379) who have a valid register-based value

of each type of blood test within each year of follow-up.

23

Figure 3. Mean values and proportions of blood pressure and smoking status among

patients with first-episode myocardial infarction admission by groups of social vulnerability

and calendar period of admission. Values are based on questionnaire data from general

practitioners of all patients with a first admission at Aarhus University Hospital, Denmark

between 2000 and 2004 (N=379) with a valid questionnaire response at each time of

follow-up.

Paper III

For Peer Review Only

Does socially differentiated cardiac rehabilitation affect the use of health care services after myocardial infarction? A

ten-year follow-up study

Journal: Scandinavian Journal of Public Health

Manuscript ID Draft

Manuscript Type: Original Article

Date Submitted by the Author: n/a

Complete List of Authors: Hald, Kathrine; Section for Clinical Social Medicine and Rehabilitation, Department of Public Health, Aarhus Universitymeillier, lucette; DEFACTUM, Social and Health Services and Labour Market, Central Denmark RegionNielsen, Kirsten; Department of Cardiology, Aarhus University HospitalLarsen, Finn; DEFACTUM, Social and Health Services and Labour Market, Central Denmark RegionJohansen, Martin; Unit of Clinical Biostatistics, Aalborg University HospitalLarsen, Mogens ; Danish Centre for Inequality in Health, Department of Cardiology, Aalborg University HospitalNielsen, Claus; Section for Clinical Social Medicine and Rehabilitation, Department of Public Health, Aarhus UniversityChristensen, Bo; Section for General Medical Practice, Department of Public Health, Aarhus University

Discipline: Social medicine

Problem Areas and Research: Cardiovascular diseases and their risk factors

Methodology: Quantitative longitudinal

http://mc.manuscriptcentral.com/spub E-mail: [email protected]

Scandinavian Journal of Public Health


Page 1 of 27



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


0.2

.4.6

.81

Prop

ortio

n pa

rtici

patin

g

0 2 4 6 8 10Year of follow-up

Based on yearly survivors

Participation in annual chronic careconsultations in general practice

05

1015

2025

30M

ean

num

ber o

f con

tact

s to

GP


Based on yearly survivorsContacts to general practice

Socially vulnerableStandard rehabilitation 2000-2002

Socially vulnerableExpanded rehabilitation 2002-2004

Non-socially vulnerableStandard rehabilitation 2000-2002


Figure 1. Use of health care services in general practice among 379 patients receiving socially differentiated cardiac rehabilitation after first-episode myocardial infarction admission by groups of social vulnerability and calendar period. Proportions and means are based on yearly survivors each year of follow-up.

Page 2 of 27



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


0.5

11.

5M

ean

num

ber o

f adm

issi

ons


Based on yearly survivorsAll-cause hospitalizations

0.2

.4.6

Mea

n nu

mbe

r of r

eadm

issi

ons


Based on yearly survivorsCardiac readmissions





Figure 2. Use of health care services in hospital among 379 patients receiving socially differentiated cardiac rehabilitation after first-episode myocardial infarction admission by groups of social vulnerability and calendar period. Means are based on yearly survivors each year of follow-up.

Page 3 of 27



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


Table 1. Baseline characteristics of 379 patients admitted with first-episode myocardial infarction receiving socially differentiated cardiac rehabilitation. Patients are divided into groups based on social vulnerability and time of admission.




Time period


Time period

Standard rehabilitation

N=782000-2002

Expanded rehabilitation

N=1302002-2004


N=552000-2002


N=1162002-2004


56 (8.15) 55 (8.53) 60 (7.56) 57 (8.50)

Gender, male 57 (73) 93 (71) 42 (76) 94 (81)

Educational level, DUN 3.18 (1.19) 3.26 (1.39) 4.80 (1.08) 4.75 (1.19)Living alone 27 (35) 51 (39) 0 0Other diseases 39 (50) 82 (64) 27 (49) 68 (59)Diabetes mellitus 10 (13) 16 (12) 6 (11) 10 (9)Hyperlipidemia 20 (26) 37 (28) 13 (24) 44 (38)Hypertension 18 (23) 28 (22) 11 (20) 23 (20)Body mass index 27.26 (4.35) 26.26 (4.08) 26.37 (3.99) 26.54 (3.12)Current smoker 59 (76) 83 (64) 34 (62) 60 (52)

Data are given as numbers (percentage) or as means (standard deviation).

Page 4 of 27



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


Table 2. Assessment of use of health care services among socially vulnerable patients admitted from 2000-2002 (N=78) and 2002-2004 (N=130) at Aarhus University Hospital, Denmark with first-episode myocardial infarction who participated in socially differentiated cardiac rehabilitation and who were evaluated at two-, five- and ten-year follow-up. Values are based on yearly survivors and on available data from registers.


Year of follow-up

Proportion*Mean** N Proportion*

Mean** N Ratio*Difference** P-value

2 0.01* 78 0.09* 128 7.3* (1.0;55.2) 0.025 0.15* 75 0.38* 125 2.6* (1.4;4.6) 0.00

Participation in annual chronic care consultations in general practice 10 0.47* 66 0.36* 107 0.8* (0.5;1.1) 0.13

2 19.0** 78 20.3** 128 1.3** (-3.6;6.1) 0.605 21.5** 75 18.3** 125 -3.2** (-8.1;1.8) 0.21Number of contacts

to general practice10 25.0** 66 19.2** 107 -5.8** (-11.0;-0.6) 0.032 0.8** 78 0.6** 128 -0.2** (-0.6; 0.3) 0.465 0.7** 75 0.5** 125 -0.2** (-0.6; 0.2) 0.30Number of all-cause

hospitalizations10 0.5** 66 0.4** 107 -0.1** (-0.3;0.2) 0.812 0.3** 78 0.2** 128 -0.1** (-0.29;0.07) 0.245 0.1** 75 0.2** 125 0.1** (-0.1;0.2) 0.43Number of cardiac

readmissions 10 0.1** 66 0.1** 107 0.0** (-0.1;0.1) 0.98

Page 5 of 27



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


1

Title page

Title

Does socially differentiated cardiac rehabilitation affect the use of health care services

after myocardial infarction? A ten-year follow-up study

Author names

Kathrine Hald1, Lucette Kirsten Meillier2, Kirsten Melgaard Nielsen3, Finn Breinholt

Larsen2, Martin Berg Johansen4, Mogens Lytken Larsen5, Claus Vinther Nielsen1 & Bo

Christensen6.

Affiliations

1Section for Clinical Social Medicine and Rehabilitation, Department of Public Health,

Aarhus University, Aarhus, Denmark

2DEFACTUM, Social and Health Services and Labour Market, Central Denmark Region,

Aarhus, Denmark

3Department of Cardiology, Aarhus University Hospital, Aarhus, Denmark

4Unit of Clinical Biostatistics, Aalborg University Hospital, Aalborg, Denmark

5Danish Centre for Inequality in Health, Department of Cardiology, Aalborg University

Hospital, Aalborg, Denmark

6Section for General Medical Practice, Department of Public Health, Aarhus University,

Aarhus, Denmark

Corresponding author

Kathrine Hald, Section for Clinical Social Medicine and Rehabilitation, Department of

Public Health, Aarhus University, Aarhus, P. P. Oerums Gade 9-11, Building 1B, 8000

Aarhus C, Denmark.

Email: [email protected]

Word count: 2716.

Page 6 of 27



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


2

Abstract

Aim: To examine the long-term effect of a socially differentiated cardiac rehabilitation

(CR) intervention tailored to reduce social inequalities in health regarding use of health

care services in general practice and hospital among patients admitted with first-episode

myocardial infarction (MI).

Methods: A prospective cohort study was conducted at Department of Cardiology at

Aarhus University Hospital, Denmark between 2000 and 2004. The intervention consisted

of socially differentiated CR including standard CR and additional rehabilitation with focus

on cross-sectional collaboration. The study population was all patients <70 years

admitted with first-episode MI categorized as socially vulnerable (N=208) or non-socially

vulnerable (N=171) based on educational level and social network. The outcome

measures were participation in annual chronic care consultations in general practice,

contacts to general practice, all-cause hospitalizations and cardiovascular readmissions.

Results: At ten-year follow-up, there were no significant differences in the proportion of

socially vulnerable patients receiving expanded CR who participated in annual chronic

care consultations or the mean number of all-cause hospitalizations and cardiovascular

readmissions. At ten-year follow-up, socially vulnerable patients receiving standard CR

had significantly more contacts to general practice.

Conclusions: The present study found no persistent association between the socially

differentiated CR intervention and long-term use of health care services in general

practice and hospital in patients admitted with first-episode MI during a ten-year follow-

up. The long-term follow-up showed that general practice was able to maintain a regular

contact with all patients regardless of social status in the long-term secondary

cardiovascular disease prevention.

Keywords: Cardiovascular diseases; Rehabilitation; Socioeconomic factors; Health

services; General practice.

Page 7 of 27



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


3

Background

The '2016 European Guidelines on cardiovascular disease prevention in clinical practice'

define cardiovascular disease (CVD) prevention as a coordinated set of actions at

population or individual level aiming at eliminating or minimizing the impact of CVD and

related disabilities. The guidelines emphasize that the general practitioners play a key

role in initiating, coordinating and providing long-term follow-up including preventive

care and chronic disease monitoring in patients diagnosed with CVD [1]. In a recent

Cochrane publication including six Cochrane reviews, cardiac rehabilitation (CR) was

found to be effective in secondary cardiovascular disease prevention. The reviews

included 148 randomized, controlled trials and 98,093 patients with CVD. Patients

participating in exercise-based CR improved their health-related quality of life and

decreased their risk of hospital admission compared to patients not participating in CR

[2]. Whereas CR decreased the risk of hospital admission, readmission and in-patient

days [2-4], low educational level and living alone were associated with a higher

probability of readmission and emergency department use in patients already diagnosed

with CVD [5-10]. However, a Danish study found that patients with a low educational

level and patients living alone had a lower rate of hospital contacts during the first six

months after being admitted with first-episode myocardial infarction (MI) [11]. A CR

intervention focusing on reducing social inequalities in health was carried out in Denmark

between 2000 and 2004. A group of socially vulnerable patients received an expanded

CR intervention and was compared to a group of socially vulnerable patients receiving

standard CR. The intervention group showed significant results regarding adherence to

secondary prevention guidelines at one-year follow-up [12]. However, no long-term

effects were seen regarding mortality and morbidity at 10-year follow-up [13].

Page 8 of 27



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


4

Aim

The aim of the present study was to examine the long-term effect of a socially

differentiated CR intervention tailored to reduce social inequalities in health on the use of

health care services in general practice and hospitals among patients admitted with first-

episode MI.

Page 9 of 27



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


5

Methods

Design and study participants

The study was conducted as a prospective register-based cohort study. The 379 study

participants were < 70 years old and followed from baseline defined as admission with

first-episode MI; follow-up was conducted at two, five and ten years. The setting was

Department of Cardiology at Aarhus University Hospital, Denmark. The study was

conducted between 2000 and 2004. Patients were categorized as socially vulnerable if

they had a lower educational level (education classified 1-4 in The Danish Educational

Nomenclature - DUN if age < 55 years and 1-3 if age > 55 years) and / or if they lived

alone. According to these criteria, 78 patients admitted between 2000 and 2002 and 130

patients admitted between 2002 and 2004 were categorized as socially vulnerable.

Patients were excluded from the study if they had severe comorbidities or abuses

[12,13].

Intervention

All 130 socially vulnerable patients admitted between 2002 and 2004 received an

expanded CR intervention in addition to standard CR. During the acute treatment from

admission to discharge (Phase I CR), the intervention was identical to standard CR. From

discharge until return to vocational activities (Phase II CR), expanded CR lasted two

weeks longer and included one extra consultation with a nurse. The patients played an

active role in defining an individual rehabilitation plan, which was shared with their

general practitioner. In the long-term secondary prevention (Phase III CR), the patients

receiving the intervention were referred to a preventive consultation in general practice

and to rehabilitation activities in primary care and The Danish Heart Association.

Furthermore, the patients in the intervention group had a telephone follow-up

consultation two months after completing Phase II CR. All 78 socially vulnerable patients

admitted between 2000 and 2002 received standard CR and served as controls. All 171

non-socially vulnerable patients in the study population (55 admitted between 2000 and

2002 and 116 admitted between 2002 and 2004) also received standard CR [12,13].

Page 10 of 27



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


6

Outcomes

The outcome measure of the study was use of health care services. In the present study,

the health care services of interest were participation in annual chronic care consultations

in general practice, contacts to general practice, all-cause hospitalizations and

cardiovascular readmissions.

Data collection and measures

Data were collected from Danish registers using civil registration (CPR) numbers of the

patients. A CPR-number is a unique 10-digit identification number given to all citizens

with a residence permit in Denmark. The first six digits indicate the citizen's birthday and

the last four digits is a combination of digits to distinguish between citizens born in the

same year and on the same day. Each CPR-number is unique and will follow the citizen

forever. The CPR-number is used in all Danish registers and ensures complete linkage

and accurate follow-up when using register-based data [14]. Data on participation in

annual chronic care consultations in general practice and contacts to general practice

were retrieved from The Danish National Health Service Register [15]. In Denmark,

general practice is the corner stone of primary health care. General practitioners are

similar to family physicians and act as gatekeepers in referring patients to medical

specialists as well as in- and out-patient hospital examinations and treatment [16].

General practice in Denmark has the medical responsibility for secondary CVD

prevention. Patients are invited to an annual chronic care consultations in general

practice where medication, treatment goals, lifestyle, mental well-being, compliance and

motivation are assessed and discussed. Participation in annual chronic care consultations

in general practice was measured dichotomously (yes / no) for every year in the follow-

up period. Number of contacts to general practice was assessed for every year in the

follow-up period. A 'contact' was defined as any kind of contact involving the clinic of the

general practitioner, including telephone and email consultations, home visits and

services provided by other health care professionals in the clinic. Data on all-cause

hospitalizations and cardiovascular readmissions were retrieved from the Danish national

Page 11 of 27



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


7

patient register [17]. The International Classification of Diseases (ICD-10) was used to

define 'cardiovascular readmission'. All-cause hospitalizations and cardiovascular

readmissions were measured as number of hospital stays for every year in the follow-up

period.

Statistical analysis

Baseline characteristics of the study population were described using either frequencies

and percentages or means and standard deviations within groups. Use of health care

services was compared between socially vulnerable patients receiving expanded CR in

addition to standard CR and socially vulnerable patients receiving standard CR only. An

additional comparison was made between non-socially vulnerable patients who all

received standard CR to evaluate potential differences between the two calendar periods.

All data were based on yearly survivors. Participation in annual chronic care consultations

in general practice was assessed as a dichotomous outcome for each year of follow-up

and reported as proportion of patients participating with 95%-confidence intervals (CI).

These proportions were compared by calculating ratios and using chi-squared tests.

Contacts to general practice were compared using mean number of contacts during each

year of follow-up and compared by calculating the difference in mean number of contacts

with 95%-CI and using t-tests. All-cause hospitalizations and cardiovascular readmissions

were shown as mean number of admissions during each year of follow-up and compared

by calculating the difference in mean number of admissions with 95%-CI and using t-

tests. All data management and analyses were performed using Stata/MP 14.2, and P-

values below 0.05 were considered statistically significant.

Page 12 of 27



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


8

Results

Table 1 shows baseline characteristics of the study population. The mean age in the

study population was 57 years and three out of four were males. In accordance with the

criteria of defining the study population as socially vulnerable or not, the socially

vulnerable patients had a lower educational level and were more likely to live alone.

Patients diagnosed with comorbidities varied from 49 % to 64 %. Socially vulnerable

patients admitted between 2000 and 2002 had a higher body mass index and were more

likely to be smokers than the rest of the study population.

Use of health care services in general practice

Participation in annual chronic care consultations

In general, the proportion of patients participating in the annual chronic care

consultations in general practice increased over the ten-year follow-up. There was a

tendency for the patients enrolled between 2000 and 2002 to participate to a lesser

extent in the annual chronic care consultations than the patients enrolled between 2002

and 2004 regardless of social status. At the end of the ten-year follow-up period,

approximately 40 % of the study population participated in the annual chronic care

consultations (Figure 1). At two- (P-value=0.02) and five-year (P-value=0.00) follow-up,

socially vulnerable patients receiving expanded CR participated significantly more often in

the annual chronic care consultations than socially vulnerable patients receiving standard

CR. At ten-year follow-up, no significant differences were seen among the socially

vulnerable patients (P-value=0.13) (Table 2). The same tendencies were found among

the non-socially vulnerable patients (Figure 1).

Contacts to general practice

The mean number of contacts to general practice remained stable (15-25 contacts per

year) during the ten-year follow-up period regardless of social status (Figure 1). No

significant differences in mean number of contacts to general practice were seen between

socially vulnerable patients receiving expanded CR and socially vulnerable patients

Page 13 of 27



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


9

receiving standard CR at two- (P-value=0.60) and five-year follow-up (P-value= 0.21).

At ten-year follow-up, socially vulnerable patients receiving standard CR had a

significantly higher mean number of contacts to general practice (P-value=0.03) than

socially vulnerable patients receiving expanded CR (Table 2). No significant differences

were found among the non-socially vulnerable patients (Figure 1).

Use of health care services in hospital

All-cause hospitalizations

The mean number of all-cause hospitalizations declined during the ten-year follow-up in

all groups. During the first years of follow-up, socially vulnerable patients experienced

more admissions but at ten-year follow-up the mean number of admissions regardless of

social status equalized (Figure 2). No significant differences in the mean number of

hospital admissions were seen between socially vulnerable patients receiving expanded

CR and socially vulnerable patients receiving standard CR at two- (P-value=0.46), five-

(P-value= 0.30) or ten-year (P-value=0.81) follow-up (Table 2). No significant

differences were found among the non-socially vulnerable patients (Figure 2).

Cardiovascular readmissions

During the first two-years of follow-up, the mean number of cardiovascular readmissions

was 0.4 to 0.5 per year regardless of social status. From two- to ten-year follow-up, the

mean number of cardiovascular readmissions declined to a mean number of 0.1 annual

cardiovascular readmissions per patient (Figure 2). No significant differences in the mean

number of cardiac readmissions were seen between socially vulnerable patients receiving

expanded CR and socially vulnerable patients receiving standard CR at two- (P-

value=0.24), five- (P-value= 0.43) or ten-year (P-value=0.98) follow-up (Table 2). No

significant differences were found among the non-socially vulnerable patients (Figure 2).

Page 14 of 27



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


10

Discussion

In this prospective cohort study, we examined the long-term effects of a socially

differentiated CR intervention in relation to use of health care services. At ten-year

follow-up, there were no significant differences in the proportion of socially vulnerable

patients receiving expanded CR and participating in annual chronic care consultations

compared to socially vulnerable patients receiving standard CR. The same associations

were seen for the mean number of all-cause hospitalizations and cardiovascular

readmissions. At ten-year follow-up, socially vulnerable patients receiving standard CR

had significantly more contacts to general practice compared to socially vulnerable

patients receiving expanded CR.

Strengths and weaknesses of the study

One of the strengths of the present study is the use of highly valid register-based data as

well as the follow-up period of ten years. Register-based data provide an almost

complete follow-up of all yearly survivors in the follow-up year in question. Risk of

selection bias in relation to all-cause mortality when using data on yearly survivors is

low, as established in a previous study that the all-cause mortality in the study

population was not associated with the exposure [13]. As the present study is designed

as a prospective cohort study and not as a randomized controlled trial, the risk of

confounding must be considered. The homogeneous study population indicates a small

likelihood of confounding (Table 1). However, the risk of residual confounding cannot be

ruled out. Prospective data collection has been conducted using registers. Thus, no risk

of information bias can be expected as data do not depend on the memory of the study

participants. In relation to the register-based data extraction, specific codes were applied

in agreement with experts within general practice and cardiology when defining

participation in annual chronic care consultations and cardiovascular readmissions. This

provided a consistent data extraction and thus a low risk of information bias. The

proportion of patients participating in the annual chronic care consultations increased

during follow-up. Also, the part of the study population admitted between 2002 and 2004

Page 15 of 27



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


11

participated more than the study population admitted between 2000 and 2002. During

the noughties, annual chronic care consultations were put on the agenda in Denmark and

were given priority. The increase in the proportion of patients participating and also the

significant differences between the socially vulnerable patients receiving expanded CR

and the socially vulnerable patients receiving standard CR could be explained by this,

thus constituting a study weakness.

Findings in relation to other studies

To our knowledge, no previous studies have examined the long-term effect of a socially

differentiated CR intervention in relation to the MI patients' use of health care services.

In 2016, Fors et al investigated the effect of person-centered care (PCC) following MI

with focus on the patients' educational level in Sweden [5]. A cohort of 199 patients < 75

years received either usual care and PCC or usual care only. Patients receiving PCC

played an active role in designing the rehabilitation plan in corporation with the PCC team

and general practice. All patients were followed for six months and evaluated using a

composite score consisting of self-efficacy, return to work, re-hospitalization and

mortality. Patients with low education receiving the PCC intervention had a significantly

higher composite score than patients with a low education receiving usual care (P-value

0.04) [5]. Another Swedish RCT study of 131 patients by Andersson et al from 2010

compared a five-year intensive lifestyle program including stress management among

younger women in comparison to standard care [4]. Patients in the intervention group

had significantly fewer emergency visits and number of inpatient days compared to the

group receiving standard care [4]. The health care structures in Sweden and Denmark

are comparable. The study population and intervention in Fors et al [5] is also quite

comparable to the study population in the present study. Fors et al [5] managed to

create an invention that resulted in a significantly better outcome for patients with a low

educational level. An explanation for this could be that the results were collected at the

end of the intervention where patients were still receiving the intervention [5]. The

intervention in Andersson et al [4] lasted for five years and showed significant results in

relation to emergency visits and number of in-patient days at follow-up at the end of the

Page 16 of 27



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


12

study. This indicates the importance of a consistent long-term secondary intervention.

Even though no major significant long-term effects of socially differentiated CR

intervention was seen in this study, it is worth mentioning that the present study found

that the mean number of all contacts to general practice remained stable and no major

differences between the socially and non-socially vulnerable patients were seen during

the ten-year follow-up period. This could indicate that general practice has succeeded in

maintaining a regular contact with patients regardless of the patients' social status.

Compared to the present study, the study population in Andersson et al [4] was younger

and consisted only of women. Moreover, the intervention was not socially differentiated,

and it is unclear if the patients improving significantly were equally divided between

socially vulnerable and non-socially vulnerable patients [4]. Additionally, the results were

collected at the end of the five-year intervention where patients were still part of the

intervention; this was also the case in Fors et al [4,5].

Page 17 of 27



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


13

Conclusions

The present study found no persistent association between the socially differentiated CR

intervention and the long-term use of health care services in general practice and

hospital among patients admitted with first-episode MI during a ten-year follow-up

period. However, the long-term follow-up showed that general practice was able to

maintain a regular contact with all patients regardless of social status in the long-term

secondary cardiovascular disease prevention, thus contributing to reduce inequality in

primary health care.

Page 18 of 27



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


14

Acknowledgements

The authors wish to thank the patients who participated in the socially differentiated CR

intervention.

Page 19 of 27



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


15

Ethical approval

The Danish Data Protection Agency approved the study (Case number: 1-16-02-684-14).

No ethical approval was required.

Page 20 of 27



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


16

Declaration of conflicting interest

The authors declare that they have no competing interests.

Page 21 of 27



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


17

Funding

This work was supported by: Aarhus University (Grant number: 17117581), Central

Denmark Region (Grant number: A-111, 1-15-1-72-13-09), The Health Foundation

(Grant number: 16-13-0098), The Committee of Multipractice Studies in General Practice

(Grant number: 16-1461) and TrygFonden (Grant number: 119795).

Page 22 of 27



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


18

References

[1] Piepoli MF, Hoes AW, Agewall S, Albus C, Brotons C, Catapano et al. 2016 European

guidelines on cardiovascular disease prevention in clinical practive. G Ital Cardiol

2017;18:547-612.

[2] Anderson L, Taylor RS. Cardiac rehabilitation for people with heart disease: An

overview of Cochrane systematic reviews (Review). Cochrane Database Syst Rev

2014;12:1-49.

[3] Kachur S, Chongthammakun V, Lavie CJ, Schutter AD, Arena R, Milani RV et al.

Impact of cardiac rehabilitation and exercise training programs in coronary heart disease.

Prog Cardiovasc Dis 2017;60:103-114.

[4] Andersson A, Sundel KL, Unden AL, Schenck-Gustafsson K, Eriksson I. A five-year

rehabilitation programme for younger women after a coronary event reduces the need

for hospital care. Scand J Public Health 2010;38:566-73.

[5] Fors A, Gyllensten H, Swedberg K, Ekman I. Effectiveness of person-centred care

after acute coronary syndrome in relation to educational level: Subgroup analysis of a

two-armed randomised controlled trial. Int J Cardiol 2016;221:957-62.

[6] Khawaja FJ, Shah ND, Lennon RJ, Slusser JP, Alkatib AA, Rihal CS et al. Factors

associated with 30-Day readmission rates after percutaneous coronary intervention. Arch

Intern Med 2012;172:112-7.

[7] Murphy BM, Elliot PC, Grande MRL, Higgins RO, Ernest CS, Goble AJ et al. Living

alone predicts 30-day hospital readmission after coronary artery bypass graft surgery.

Eur J Cardiovasc Prev Rehabil 2008;15:210-5.

[8] Roe-Prior P. Sociodemografic variables predicting poor post-discharge outcomes for

hospitalized elders with heart failure. Medsurg Nurs 2007;16:317-21.

[9] Howie-Esquivel J, Spicer JG. Association of partner status and disposition with

rehospitalization in heart failure patients. Am J Crit Care 2012;21:65-73.

[10] Schwarz KA, Elman CS. Identification of factors predictive of hospital readmissions

for patients with heart failure. Heart Lung 2003;32:88-99.

Page 23 of 27



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


19

[11] Nielsen TJ, Vestergaard M, Fenger-Grøn M, Christensen B, Larsen KK. Healthcare

contacts after myocardial infarction according to mental health and socioeconomic

position: A population-based cohort study. PloS One 2015;30:1-13.

[12] Nielsen KM, Meillier LK, Larsen ML. Extended cardiac rehabilitation for socially

vulnerable patients improves attendance and outcome. Dan Med J 2013;60:1-5.

[13] Hald K, Nielsen KM, Nielsen CV, Meillier LK, Larsen FB, Christensen et al. Expanded

cardiac rehabilitation in socially vulnerable patients with myocardial infarction: A 10-year

follow-up study focusing on mortality and non-fatal events. BMJ Open 2018;8:1-8.

[14] Pedersen CB. The Danish civil registration system. Scan J Public Health 2011;39:22-

25.

[15] Andersen JS, Olivarius NDF, Krasnik A. The Danish national health service register.

Scan J Public Health 2011;39:34-37.

[16] Pedersen KM, Andersen JS, Søndergaard J. General practice and primary health care

in Denmark. J Am Board Fam Med 2012;25:34-38.

[17] Lynge E, Sandegaard JL, Rebolj M. The Danish national patient register. Scand J

Public Health 2011;39(7):30-3.

Page 24 of 27



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


20

Figures

0.2

.4.6

.81

Prop

ortio

n pa

rtici

patin

g


Based on yearly survivors

Participation in annual chronic careconsultations in general practice

05

1015

2025

30M

ean

num

ber o

f con

tact

s to

GP


Based on yearly survivorsContacts to general practice





Figure 1. Use of health care services in general practice among 379 patients receiving

socially differentiated cardiac rehabilitation after first-episode myocardial infarction

admission by groups of social vulnerability and calendar period. Proportions and means

are based on yearly survivors each year of follow-up.

Page 25 of 27



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


21

0.5

11.

5M

ean

num

ber o

f adm

issi

ons


Based on yearly survivorsAll-cause hospitalizations

0.2

.4.6

Mea

n nu

mbe

r of r

eadm

issi

ons


Based on yearly survivorsCardiac readmissions





Figure 2. Use of health care services in hospital among 379 patients receiving socially

differentiated cardiac rehabilitation after first-episode myocardial infarction admission by

groups of social vulnerability and calendar period. Means are based on yearly survivors

each year of follow-up.

Page 26 of 27



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


22

Tables


infarction receiving socially differentiated cardiac rehabilitation. Patients are divided into

groups based on social vulnerability and time of admission.




Time period


Time period


N=782000-2002

Expanded rehabilitation

N=1302002-2004


N=552000-2002


N=1162002-2004


56 (8.15) 55 (8.53) 60 (7.56) 57 (8.50)

Gender, male 57 (73) 93 (71) 42 (76) 94 (81)

Educational level, DUN 3.18 (1.19) 3.26 (1.39) 4.80 (1.08) 4.75 (1.19)Living alone 27 (35) 51 (39) 0 0Other diseases 39 (50) 82 (64) 27 (49) 68 (59)Diabetes mellitus 10 (13) 16 (12) 6 (11) 10 (9)Hyperlipidemia 20 (26) 37 (28) 13 (24) 44 (38)Hypertension 18 (23) 28 (22) 11 (20) 23 (20)Body mass index 27.26 (4.35) 26.26 (4.08) 26.37 (3.99) 26.54 (3.12)Current smoker 59 (76) 83 (64) 34 (62) 60 (52)

Data are given as numbers (percentage) or as means (standard deviation).

Page 27 of 27



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


23

Table 2. Assessment of use of health care services among socially vulnerable patients

admitted from 2000-2002 (N=78) and 2002-2004 (N=130) at Aarhus University

Hospital, Denmark with first-episode myocardial infarction who participated in socially

differentiated cardiac rehabilitation and who were evaluated at two-, five- and ten-year

follow-up. Values are based on yearly survivors and on available data from registers.


Year of follow-

up

Proportion*Mean** N Proportion*

Mean** N Ratio*Difference** P-value

2 0.01* 78 0.09* 128 7.3* (1.0;55.2) 0.025 0.15* 75 0.38* 125 2.6* (1.4;4.6) 0.00

Participation in annual chronic care consultations in general practice 10 0.47* 66 0.36* 107 0.8* (0.5;1.1) 0.13

2 19.0** 78 20.3** 128 1.3** (-3.6;6.1) 0.605 21.5** 75 18.3** 125 -3.2** (-8.1;1.8) 0.21

Number of contacts to general practice 10 25.0** 66 19.2** 107 -5.8** (-11.0;-0.6) 0.03

2 0.8** 78 0.6** 128 -0.2** (-0.6; 0.3) 0.465 0.7** 75 0.5** 125 -0.2** (-0.6; 0.2) 0.30

Number of all-cause hospitalizations 10 0.5** 66 0.4** 107 -0.1** (-0.3;0.2) 0.81

2 0.3** 78 0.2** 128 -0.1** (-0.29;0.07) 0.245 0.1** 75 0.2** 125 0.1** (-0.1;0.2) 0.43Number of cardiac

readmissions 10 0.1** 66 0.1** 107 0.0** (-0.1;0.1) 0.98

Page 28 of 27



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Praksisnavn Appendix IV Adresse Postnummer og by

Aarhus C, den XX. XX 2017

Kære praktiserende læge

Almen praksis har en central rolle i langtidsopfølgningen af patienter med hjertekarsygdom. Med denne undersøgelse ønsker vi at klarlægge, hvorledes hjertepatienter har det 10 år efter deres AMI. Hvordan ser deres risikoprofil ud, og hvordan er deres helbred?

I undersøgelsen indgår 379 patienter, som alle blev indlagt med førstegangs AMI på Aarhus Universitetshospital i perioden 2000-2004, og som deltog i en socialt differentieret hjerterehabiliteringsintervention.

Din patient var med i undersøgelsen, og vi sender derfor et spørgeskema til dig, idet vi i Sygesikringsregistret kan se, at du er læge for nedenstående patient. Hvis patienten er død, er du den sidst registrerede læge for patienten. For at belyse dele af patientens livsstil vil vi gerne bede dig udfylde det vedlagte spørgeskema.

Det vedlagte spørgeskema handler altså om en specifik patient, og du kan derfor modtage flere spørgeskemaer. Dette spørgeskema vedrører:

CPR: XXXXXX-XXXX

som deltog i den socialt differentierede hjerterehabiliteringsintervention i perioden 2000-2004.

Vi håber, du vil udfylde spørgeskemaet og returnere det i den vedlagte svarkuvert.

Honorering og godkendelse Hvert udfyldt og returneret spørgeskema honoreres med 253,30 kr. svarende til honorar for 2 moduler. Honoraret udbetales efter returnering af spørgeskemaet og den vedlagte honorarblanket. Projektet er godkendt af Datatilsynet og Multipraksisudvalget, som anbefaler de praktiserende læger at deltage. Projektet er blandt andre støttet af Region Midtjyllands Praksisforskningsfond.

På forhånd tak for hjælpen og vi står naturligvis til disposition, hvis yderligere oplysninger ønskes.

Med venlig hilsen

Kathrine Hald Ph.d.-studerende og projektleder Cand.scient.san., Klinisk diætist DEFACTUM, Region Midtjylland og Aarhus Universitet. P. P. Ørums Gade 9-11, Bygning 1B 8000 Aarhus C. Tlf. 78 41 44 40.

Bo Christensen Praktiserende læge Professor, Ph.d. Afdeling for Almen Medicin, Aarhus Universitet. Bartholin Allé 2 8000 Aarhus C. Tlf. 21 76 66 95.

Appendix IV

Ph.d.-Projekt: Langtidsopfølgning på effekt af socialt differentieret hjerterehabilitering

Vi ønsker oplysninger om blodtryk og rygestatus 2, 5 og 10 år efter patienterne blev indlagt med førstegangs AMI på Aarhus Universitetshospital.

Patienten blev indlagt den:

XX-XX-XXXX

1. Angiv oplysninger om blodtryk der er målt tættest på 2, 5 og 10 år efter indlæggelsesdatoen:

2 år Blodtryk _ _ _ / _ _ _ mmHg Det blev målt den: _ _ - _ _ - 20_ _

Det blev målt _ I konsultationen/_ Hjemme (incl. døgnblodtryk)


Det blev målt _ I konsultationen/_ Hjemme (incl. døgnblodtryk)


Det blev målt _ I konsultationen/ _ Hjemme (incl. døgnblodtryk)

2. Angiv oplysninger om rygestatus noteret tættest på 2, 5 og 10 år efter indlæggelsesdatoen:

2 år Ryger _ Ikke-ryger _

Oplysningen om rygestatus blev angivet den: _ _ - _ _ - 20_ _





Mange tak for hjælpen.

Løbenr. X

Documents

foll soc entiated cardiac rehabilita tion...soc - e Lon ially valuated by g-term differ clinical ou foll entia tcomes, se PhD Ka Aa low-up ted ca econdary pre dissert thrine H Health