Measuring Primary Care and Its Benefits - … · Measuring Primary Care and Its Benefits Barbara Starfield, MD, MPH ... SWE JAP GER FR BEL US DK *Best ... Primary Care Score

Measuring Primary Care and Its Benefits

Barbara Starfield, MD, MPH

University of Hong KongHong Kong

October 15, 2009

Life Expectancy Compared with GDP per Capita for Selected Countries

Source: Economist Intelligence Unit. Healthcare International. 4th quarter 1999. London, UK: Economist Intelligence Unit, 1999.

Country codes:AG=ArgentinaAU=AustraliaBZ=BrazilCH=ChinaCN=CanadaFR=FranceGE=GermanyHU=HungaryIN=IndiaIS=IsraelIT=ItalyJA=JapanMA=MalaysiaME=Mexico

Starfield 11/06IC 3493 n

NE=NetherlandsPO=PolandRU=RussiaSA=South AfricaSI=SingaporeSK=South KoreaSP=SpainSW=SwedenSZ=SwitzerlandTK=TurkeyTW=TaiwanUK=United KingdomUS=United States

Why Is Primary Care Important?

Better health outcomes

Lower costs

Greater equity in health

Starfield 07/07PC 3757 n

Good Primary Care Requires• Health system POLICIES conducive to primary care

practice: What can we learn from other countries about the relative merits of direct provision of services rather than just financing of services?

• Health services delivery that achieves the important FUNCTIONS of primary care: What can be done to enhance practitioners’ recognition of and responsiveness to patients’ problems (patient-focus) rather than on the professional priorities of diagnoses (diagnosis-focus)?

Starfield 06/08PC 4042

PRIMARY HEALTH CARE “works”because it has definable system functions that provide the policy context for primary care.


PRIMARY CARE “works” because it has defined functions that include structural and process features of clinical health services that are known to improve outcomes of care.

Key system factors in achieving primary health care in both developing and industrialized countries are:• Universal financial coverage, under

governmental control or regulation• Efforts to distribute resources equitably

(according to degree of need)• No or low co-payments• Comprehensiveness of services

Starfield 07/07GH 3794 n

Sources: Starfield & Shi, Health Policy 2002; 60:201-18. Gilson et al, Challenging Inequity through Health Systems (http://www.who.int/social_determinants/resources/csdh_media/hskn_final_2007_en.pdf; accessed March 17, 2009).

0123456789

101112

0 1 2 3 4 5 6 7 8 9 10 11 12 13

System Characteristics (Rank*)

Prac

tice

Char

acte

ristic

s (R

ank*

)

UK

NTH

SP

FIN CANAUS

SWE JAP

GER FRBEL

US

DK

*Best level of health indicator is ranked 1; worst is ranked 13; thus, lower average ranks indicate better performance.Based on data in Starfield & Shi, Health Policy 2002; 60:201-18.

System (PHC) and Practice (PC) Characteristics Facilitating Primary Care, Early-Mid 1990s

Starfield 03/05ICTC 3099 n

At the clinical level,

• The critical structural features are Accessibility, mechanisms of Continuity/Information Systems, and the Range of Services available in primary care.

• The critical process features are Problem Recognition on the part of practitioners (both for initial problems and for reassessment), and Utilization of primary care services, both over time and for new problems as they arise.

Starfield 04/08EVAL 4018

Together, these features achieve the evidence-based FUNCTIONS of primary care: first contact, person-focused (not disease-focused) care over time, comprehensiveness, and coordination.

Primary Care Strength and Premature Mortality in 18 OECD Countries

*Predicted PYLL (both genders) estimated by fixed effects, using pooled cross-sectional time series design. Analysis controlled for GDP, percent elderly, doctors/capita, average income (ppp), alcohol and tobacco use. R2(within)=0.77.

Source: Macinko et al, Health Serv Res 2003; 38:831-65.

Year

High PC Countries*

Low PC Countries*

10000

PYLL

1970 1980 1990 20000

5000

Starfield 11/06IC 3496 n

Primary Care Oriented Countries Have

• Fewer low birth weight infants• Lower infant mortality, especially

postneonatal• Fewer years of life lost due to suicide• Fewer years of life lost due to “all except

external” causes• Higher life expectancy at all ages except

at age 80Starfield 07/07IC 3762 n

Sources: Starfield. Primary Care: Balancing Health Needs, Services, and Technology. Oxford U. Press, 1998. Starfield & Shi, Health Policy 2002; 60:201-18.

0

0.5

1

1.5

2

1000 1500 2000 2500 3000 3500 4000

Per Capita Health Care Expenditures

Prim

ary

Care

Sco

re

Primary Care Score vs. Health Care Expenditures, 1997

US

NTH

CANAUS

SWE JAP

BEL FRGER

SP

DK

FIN

UK

Starfield 11/06ICTC 3495 n

• Have more equitable resource distributions• Have health insurance or services that are

provided by the government• Have little or no private health insurance• Have no or low co-payments for health services• Are rated as better by their populations• Have primary care that includes a wider range

of services and is family oriented• Have better health at lower costs

Primary health care oriented countries

Sources: Starfield and Shi, Health Policy 2002; 60:201-18. van Doorslaer et al, Health Econ 2004; 13:629-47. Schoen et al, Health Aff 2005; W5: 509-25.

Starfield 11/05IC 3326

Is Primary Care as important within

countries as it is among countries?

Starfield 07/07WC 3765 n

State Level Analysis:Primary Care and Life Expectancy

Source: Shi, Int J Health Serv 1994;24:431-58. Starfield 04/09WCUS 4178 n

71

72

73

74

75

76

77

78

4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5

Primary care physicians per 10,000 population

Life

exp

ecta

ncy

at b

irth

ME

NHVT

MA

RICT

NY

NJ

PAOH

IN

IL

MI

WI

MN

IA

MO

ND

SD

NE

KSDE MDVA

WV NC

SC

GA

FL

KYTN

AL

MS

AR

LA

OK TX MT

ID

WY

CO

NMAZ

UT

NV

WAOR

CA

AK

HI

Many other studies done WITHIN countries, both industrialized and developing, show that areas with better primary care have better health outcomes, including total mortality rates, heart disease mortality rates, and infant mortality, and earlier detection of cancers such as colorectal cancer, breast cancer, uterine/cervical cancer, and melanoma. The opposite is the case for higher specialist supply, which is associated with worse outcomes.

Starfield 09/0404-167

Sources: Starfield et al, Milbank Q 2005;83:457-502. Macinko et al, J Ambul Care Manage 2009;32:150-71.

Starfield 09/04WC 2957

Why Does Primary Care Enhance Effectiveness of Health Services?

• Greater accessibility• Better person-focused prevention• Better person-focused quality of clinical

care• Earlier management of problems (avoiding

hospitalizations)• The accumulated benefits of the four

features of primary careStarfield 05/09PC 4185Source: Starfield et al, Milbank Q 2005;83:457-502.

Does primary care reduce inequity in

health?

Starfield 07/07EQ 3769 n

Source: Shi et al, Soc Sci Med 2005; 61(1):65-75.

In the United States, an increase of 1 primary care doctor is associated with 1.44 fewer deaths per 10,000 population.

The association of primary care with decreased mortality is greater in the African-American population than in the white population.

Starfield 07/07WCUS 3770 n

A comparison of age-adjusted survival from breast cancer showed that

• Low SES is strongly associated with decreased survival in US, but not Canada.

• The survival advantage in Canada is present in low income areas only.

• The survival advantage in Canada is much larger at ages under 65.

• The Canadian survival advantage is larger for later stage diagnosis. That is, there is almost certainly a medical care benefit to equity in the Canadian context.

Source: Gorey, Breast cancer survival in Canada and the USA: meta-analytic evidence of a Canadian advantage in low-income areas, Int J Epidemiol 2009 forthcoming.

Starfield 08/09IC 4230

Why Does Primary Care Enhance Equity in Health?

• Greater comprehensiveness of services (especially important in the presence of multi-morbidity)

• Person-focused care over time (better knowledge of patient and better recognition of problems)

• Greater accessibility of services• Better coordination, thus facilitating care for

people of limited flexibility • Better person-focused prevention

Starfield 05/09PC 4184Source: Starfield et al, Milbank Q 2005;83:457-502.

Many other studies done WITHIN countries, both industrialized and developing, show that areas with better primary care have better health outcomes, including total mortality rates, heart disease mortality rates, and infant mortality, and earlier detection of cancers such as colorectal cancer, breast cancer, uterine/cervical cancer, and melanoma. The opposite is the case for higher specialist supply, which is associated with worse outcomes.

Starfield 09/0404-167

Sources: Starfield et al, Milbank Q 2005;83:457-502. Macinko et al, J Ambul Care Manage 2009;32:150-71.

Starfield 09/04WC 2957

In 35 US analyses dealing with differences between types of areas (7) and 5 rates of mortality (total, heart, cancer, stroke, infant), the greater the primary care physician supply, the lower the mortality for 28. The higher the specialist ratio, the higher the mortality in 25.

Controlled only for income inequalitySource: Shi et al, J Am Board Fam Pract 2003; 16:412-22.

Starfield 11/06SP 3499 n

Above a certain level of specialist supply, the more specialists per population, the worse the outcomes.

Use of Specialists in the US• REFERRAL rates from primary care to

specialty care in the US are HIGH.• Between 1/3 and 3/4 (depending on

the type of specialist) of visits to specialists are for routine follow-up.

• The percentage of people SEEN BY a specialist in a year is high, especially in the presence of high morbidity burden.

Starfield 03/06SP 3396Sources: Forrest et al, BMJ 2002; 325:370-1. Valderas et al, Ann Fam Med 2009;7:104-11.

Resource Use, Controlling for Morbidity Burden*

• More DIFFERENT specialists seen: higher total costs, medical costs, diagnostic tests and interventions, and types of medication

• More DIFFERENT generalists seen: higher total costs, medical costs, diagnostic tests and interventions

• More generalists seen (LESS CONTINUITY): more DIFFERENT specialists seen among patients with high morbidity burdens. The effect is independent of the number of generalist visits. That is, the benefits of primary care are greatest for people with the greatest burden of illness.

Starfield 09/07CMOS 3854

*Using the Johns Hopkins Adjusted Clinical Groups (ACGs)

Source: Starfield et al, Ambulatory specialist use by patients in US health plans: correlates and consequences. J Ambul Care Manage 2009 forthcoming.

Family Physicians, General Internists, and Pediatricians

A nationally representative study showed that adults and children with a family physician (rather than a general internist, pediatrician, or sub-specialist) as their regular source of care had lower annual cost of care, made fewer visits, had 25% fewer prescriptions, and reported less difficulty in accessing care, even after controlling for case-mix, demographic characteristics (age, gender, income, race, region, and self-reported health status). Half of the excess is in hospital and ER spending; one-fifth is in physician payments; and one-third is for medications.

Starfield 03/09 PC 4162Source: Phillips et al, Health Aff 2009;28:567-77.

We know that1. Inappropriate referrals to specialists lead to

greater frequency of tests and more false positive results than appropriate referrals to specialists.

2. Inappropriate referrals to specialists lead to poorer outcomes than appropriate referrals.

3. The socially advantaged have higher rates of visits to specialists than the socially disadvantaged.

4. The more the training of MDs, the more the referrals.

Source: Starfield et al, Health Aff 2005; W5:97-107 (http://content.healthaffairs.org/cgi/reprint/hlthaff.w5.97v1). van Doorslaer et al, Health Econ 2004; 13:629-47;

Starfield 08/05SP 3241

A MAJOR ROLE OF PRIMARY CARE IS TO ASSURE THAT SPECIALTY CARE IS MORE APPROPRIATE AND, THEREFORE, MORE EFFECTIVE.

Joint Principles of the Patient-centered Medical Home

• Personal physician: ongoing relationship for first contact, continuous, comprehensive care

• Physician directed medical practice• Whole person oriented• Coordinated and/or integrated care• Quality and safety• Enhanced access• Added value payment

Starfield 03/08MH 4005Source: AAFP/AAP/ACP/AOA. Joint Principles. March 2007.

Proposed PC/MH (Patient-centered Medical Home) Criteria

• Electronic health record• Teams• Chronic care guidelines

Starfield 06/08MH 4043

Question: Do these “enhancements”improve primary care?

This requires evaluation.

Is a Focus on Chronic Disease Compatible with the Patient-

Centered Medical Home?In Pennsylvania, the Governor’s Office of Health Care Reform convened several health plans and physician societies in the southeastern part of the state to “institute a PCMH approach to manage the care of chronically ill patients”.

To what extent is this approach consistent with the principles of population-oriented primary care and the patient-centered medical home? Who is left out?

Starfield 03/09 D 4163

There is more variability in disease manifestations and persistence within diseases than across diseases because:• diseases are not necessarily unique

pathophysiological entities• variability in diagnostic styles and

practices• presence of co-morbidity

Starfield 10/01D 3887

Co- and Multi-morbidity(Morbidity Burden)

Starfield 09/07CM 3864 n

Co-morbidity is the concurrent existence of one or more unrelated conditions in an individual with any given condition. Multi-morbidity is the co-occurrence of biologically unrelated illnesses.

Starfield 03/06CM 3375

For convenience and by common terminology, we use co-morbidity to represent both co- and multi-morbidity.

Morbidity Burdens of Socially Disadvantaged and Socially Advantaged People

0 5 10 15 20 25 30

10+ Morbidity-Types

6-9 MorbidityTypes

% of Total Population

HMO

CHC (Disadvantaged)


The high frequency of

Co-morbidity

Multi-morbidity

Morbidity burden

makes it inappropriate to focus on single diseases


Co-morbidity, Inpatient Hospitalization, Avoidable Events, and Costs*

11

362

296

267

216

169

119

74

40

208

1734

57

86

119

152

182

233

84

211

1154

2394

4701

13,973

0

50

100

150

200

250

300

350

400

0 1 2 3 4 5 6 7 8 9 10+

Number of types of conditions

Rate

per

100

0 be

nefic

iari

es

0

2000

4000

6000

8000

10000

12000

14000

16000

Cos

ts

ACSC Complications Costs

(4 or moreconditions)

Source: Wolff et al, Arch Intern Med 2002; 162:2269-76. *ages 65+, chronic conditions only


Expected Resource Use (Relative to Adult Population Average) by Level of Co-Morbidity, British Columbia, 1997-98


None Low Medium HighVery High

Acute conditions only

0.1 0.4 1.2 3.3 9.5

Chronic condition 0.2 0.5 1.3 3.5 9.8High impact chronic condition

0.2 0.5 1.3 3.6 9.9

Source: Broemeling et al. Chronic Conditions and Co-morbidity among Residents of British Columbia. Vancouver, BC: University of British Columbia, 2005.

Thus, it is co-morbidity, rather than presence or impact of chronic conditions, that generates resource use.

As thresholds for diagnosing disease are lowered over time, the variability within “diseases”will increase even further, as will the prevalence of multiple simultaneous or sequential diseases.


Increase in Treated Prevalence: Selected Conditions, US, People with Private

Insurance, 1987-2002Treated Prevalence

Percentage Change, 1987-2002Hyperlipidemia 437(Heart disease 9) Bone disorders 227Upper GI problems 169Cerebrovascular disease 161Mental problems 136Diabetes 64Endocrine disorders 24Hypertension 17Bronchitis 13

Source: Thorp et al, Health Affairs 2005; W5:317-25, 2005.Starfield 09/06D 3858 n

What is needed is person-focused care over time, NOT disease-focused care.


When people (not diseases) are the focus of attention

• Outcomes are better• Side effects are fewer• Costs are lower• Population health is greater

Starfield 09/07PC 3868 nSource: Starfield et al, Health Aff 2005; W5:97-107.

What Is the Appropriate Care Model?

• Primary care that meets primary care (not disease-specific) standards*

• Specialty referrals that are appropriate, i.e., evidence-based**

• Specialty care that meets specialty care standards**


*exist**do not exist

Modern medicine is largely atheoretical. With the exception of a few rare genetic conditions, we do not understand why some people have greater susceptibility to disease and, particularly why some people are more prone to multimorbidity than other people. On the other hand, some people seem to be more resilient to health problems. Why?


Diseases• are professional constructs• can be and are artificially created to suit

special interests; the sum of deaths attributed to diseases exceeds the number of deaths

• do not exist in isolation from other diseases and are, therefore, not an independent representation of illness

• are but one manifestation of ill health


Sources: Chin. The AIDS Pandemic: the Collision of Epidemiology with Political Correctness. Radcliffe Publishing, 2007. De Maeseneer et al. Primary Health Care as a Strategy for Achieving Equitable Care: a Literature Review Commissioned by the Health Systems Knowledge Network. WHO Health Systems Knowledge Network, 2007. Available at: http://www.wits.ac.za/chp/kn/De%20Maeseneer%202007%20PHC%20as%20strategy.pdf. Mangin et al, BMJ 2007; 335:285-7. Murray et al, BMJ 2004; 329:1096-1100. Tinetti & Fried, Am J Med 2004; 116:179-85. Walker et al, Lancet 2007; 369:956-963. Rosenberg, Milbank Q 2002;80:237-60. Moynihan & Henry, PLoS Med 2006;3:e191.

Despite large improvements in behavior of populations from the 1970s to the 19990s (e.g., a very large decrease in consumption of fats as a proportion of energy), the prevalence of obesity and diabetes increased greatly. Researchers have linked a growing number of chronic diseases (type 2 diabetes, coronary artery disease, stroke, non-alcoholic liver disease, fatty liver disease, polycystic ovary syndrome, asthma, some cancers, and Alzheimer’s disease) to the metabolic disorder known as insulin resistance, with widely divergent views of the pathogenesis by which it occurs.


Sources: Willett & Leibel, Am J Med 2002;113 Suppl 9B:47S-59S. Taubes, Science 2009;325:256-60

What characterizes illness is its variability, not its average manifestations. Virtually all of the conclusions of randomized controlled clinical trials are based on the average response. Variability, which underlies the genesis of illness, the role of risk factors, and the impact of interventions, goes unrecognized.


If many people with relatively low risk are subjected to the same interventions (e.g., the polypill) as those with high risk, two of the three principles of prevention (avoiding unintended effects and low cost-effectiveness) will be violated.

Starfield 09/09PREV 4271

Ambulatory Diagnostic Groups (ADGs)

Time limited (4)AllergiesAsthma

Likely to recur (3)Malignancy

Chronic medical (2)Chronic specialty (6)

DermatologicInjuries (2)Psychosocial/psychophysiologic (3)Signs/symptoms (3)

DiscretionarySee and reassurePreventive/administrativePregnancyDental

Total number of ADGs = 32Starfield 199797-018

Starfield 12/97AC 1260

Going from ADGs to ACGs• During a single year, a patient’s diagnoses may fall into as many as 32

distinct ADGs. The potential permutations are vast. For practicality, a case-mix system must have a manageable number of mutually-exclusive categories

• Clinically similar ADGs are combined into CADGs (collapsed ADGs).• Individual CADGs and the most common combinations are designated

as MACs (Major Ambulatory Categories) with one additional MAC for “all other combinations”

• ACGs are formed from the MACs, based upon relative contributions to resources use

• Some ACGs are subgroups of a MAC based on– age and/or sex– total number of ADGs– total number of major ADGs

Starfield 199797-093


Entire Population

Age >= 1

Age < 1 MAC-26

MACMAC--11

MACMAC--22ACG 0400

MACMAC--44ACG 0700

MACMAC--55ACG 0800

MACMAC--33 MACMAC--2525MACMAC--1919

MACMAC--2020ACG 3400





MACMAC--1414 MACMAC--1616ACG2400

MACMAC--1717MACMAC--1515ACG2300

MACMAC--77ACG 1000

MACMAC--88ACG 1100

MACMAC--1010


MACMAC--99ACG 1200

MACMAC--66ACG 0900

MACMAC--1212 MACMAC--1818ACG 2800

MACMAC--2424

To MAC 26

tree

Split into MACs,Based on CADGs

To MAC 24

tree

To MAC 12

tree

Age

1

2-5

6 +

ACG 0100

ACG 0200

ACG 0300

ADG05 ?

Yes

No

ACG 0600

ACG 0500

ADG25 ?

ADG24?

NoYes

Yes

No

ACG 1300

ACG 1400

ACG 1500

Age

1

2-5

6 + ADG05 ?

Yes No

ACG 2200

ACG 2000

ACG 2100

ADG25?

ADG24?

NoYes

Yes

NoACG 2600

ACG 2500

ACG 2700

1 or 2 input files?

Age

1

2-5

6 +

ADG05 ?

Yes No

12 +

ACG 3100

ACG 3000

ACG 2900

ACG 3200ACG 3300

2

Claims info?

Yes

No

1

ACG 5100

ACG 5200

ACG 5110

ACG 1900

ACG 9900Missing Age

MAC Major Ambulatory CategoryADG Ambulatory Diagnostic GroupCADG Collapsed ADGACG Ambulatory Care Group

Key


Decision Tree for ACGs

Source: JHU ACG Case Mix Adjustment System, V. 4.0, 1997.Starfield 04/97AC 1091

MAC-12

# of ADGs

Provides warning if :Gender = Male,Age < 10 or age > 55

0 -1 2-3

# of MajorADGs

ACG 1730

0 1+

4-5

# of MajorADGs

0 1+

6+

# of MajorADGs

ACG 1731

ACG 1732

Yes

No

Delivered?

ACG 1741

ACG 1742

Yes

No

Delivered?

ACG 1751

ACG 1752

Yes

No

Delivered?

ACG 1761

ACG 1762

Yes

No

Delivered?

ACG 1771

ACG 1772

Yes

No

Delivered?

0 1+

ACG 1721

ACG 1722

Yes

No

ACG 1720

Delivered?

ACG 1740 ACG 1750 ACG 1760 ACG 1770

MAC Major Ambulatory CategoryADG Ambulatory Diagnostic GroupACG Ambulatory Care Group

Key

ACG 1711

ACG 1712

Yes

No

ACG 1710

Delivered?Note: This level of branching is optional


Decision Tree for MAC 12 – Pregnant Women


# of ADGs

0-5 6+

Note: This level ofbranching is optional

* Low birth weight refers to infants less than 2500 grams

# of MajorADGs

1+0

Yes

No

Yes

No

Yes

No

Yes

No

0 1+


Key

ACG 5310 ACG 5320 ACG 5330 ACG 5340

Lowbirth weight *

Lowbirth weight *

Lowbirth weight *

Lowbirth weight *

ACG 5312

ACG 5311

ACG 5322

ACG 5321

ACG 5332

ACG 5331

ACG 5342

ACG 5341

MAC-26

# of MajorADGs


Decision Tree for MAC 26 – Infants


MAC-24MAC-24

# of ADGs

# of ADGs

ACG 38001-16

35 +

17-34 SexSex

ACG 3900

ACG 4000

ACG 4100

Female

Male

AgeAge

2-3

AgeAge

1-16 # ofMajorADGs

17 + # ofMajorADGs

0

ACG 5020

ACG 50101

ACG 50302 +

0 - 1

ACG 5050

ACG 50402

ACG 50603

ACG 50704 +

10 +

17-44 # of Major ADGs

0

ACG 4320

ACG 43101

ACG 43302 +

1 - 16 # of Major ADGs

45 + # ofMajorADGs

0

ACG 4420

ACG 44101

ACG 44302 +

AgeAge

4-5

0ACG 4210

ACG 42201 +

35 +# of

MajorADGs

ACG 4920

ACG 49102

ACG 49303

ACG 49404 +

AgeAge

6-9

17-34 SexSex

6- 16 # of Major ADGs

1 - 5 # of Major ADGs

0 - 1

M

F

# ofMajorADGs

# ofMajorADGs

0

ACG 4720

ACG 47101

ACG 47302 +

0

ACG 4820

ACG 48101

ACG 48302 +

ACG 4510

ACG 45201 +

0

ACG 4610

ACG 46201 +

0


Key


Decision Tree for MAC 24 – Multiple ADG Categories


AC 1095Figure 2: Decision Tree for MAC 12 -- Pregnant women

MAC-12

# of ADG s

Provides warning if : Gender = Male, Age < 10 or age > 55

0 -1 2-3

# of MajorADG s

ACG 1730

0 1+

4-5

# of MajorADG s

0 1+

6+

# of MajorADG s

ACG 1731

ACG 1732

Yes

No

Delivered?

ACG 1741

ACG 1742

Yes

No

Delivered?

ACG 1751

ACG 1752

Yes

No

Delivered?

ACG 1761

ACG 1762

Yes

No

Delivered?

ACG 1771

ACG 1772

Yes

No

Delivered?

0 1+

ACG 1721

ACG 1722

Yes

No

ACG 1720

Delivered?

ACG 1740 ACG 1750 ACG 1760 ACG 1770

MAC Major Am bulatory CategoryADG Ambulatory Diagnost ic GroupACG Ambulatory Care Group

Key

ACG 1711

ACG 1712

Yes

No

ACG 1710

Delivered?Note: This l evel of branchi ng is opt ional

Entire Population

Age >= 1

Age < 1 MAC-26

MAC-1MAC-1

MAC-2MAC-2ACG 0400

MAC-4MAC-4ACG 0700

MAC-5MAC-5ACG 0800

MAC-3MAC-3 MAC-25MAC-25MAC-19MAC-19

MAC-20MAC-20ACG 3400





MAC-14MAC-14 MAC-16MAC-16ACG 2400

MAC-17MAC-17MAC-15MAC-15ACG 2300

MAC-7MAC-7ACG 1000

MAC-8MAC-8ACG 1100

MAC-10MAC-10


MAC-9MAC-9ACG 1200

MAC-6MAC-6ACG 0900

MAC-12MAC-12 MAC-18MAC-18ACG 2800

MAC-24MAC-24

To MAC 26

tree

Spl it into MACs,Based on CADGs

To MAC 24

tree

To MAC 12

tree

Figure 1: Decision Tree for ACGs

Age

1

2-5

6 +

ACG 0100

ACG 0200

ACG 0300

ADG05 ?

Yes

No

ACG 0600

ACG 0500

ADG25 ?

ADG24?

NoYes

Yes

No

ACG 1300

ACG 1400

ACG 1500

Age

1

2-5

6 + ADG05 ?

Yes No

ACG 2200

ACG 2000

ACG 2100

ADG25?

ADG24?

NoYes

Yes

NoACG 2600

ACG 2500

ACG 2700

1 or 2 input f iles?

Age

1

2-5

6 +

ADG05 ?

Yes No

12 +

ACG 3100

ACG 3000

ACG 2900

ACG 3200ACG 3300

2

Claim s info?

Yes

No

1

ACG 5100

ACG 5200

ACG 5110

ACG 1900

ACG 9900Mi ssing Age

MAC Major Am bulatory CategoryADG Ambulatory Diagnost ic GroupCADG Coll apsed ADGACG Ambulatory Care Group

Key

Figure 3: Decision Tree for MAC 26 -- Infants

# ofADG s

0-5 6+

Note: This l evel ofbranchi ng is opt ional

* Low Birthweight refers to infants less than 2500 grams

#of Maj orADG s

#of Maj orADG s

1+0

Yes

No

Yes

No

Yes

No

Yes

No

0 1+


Key

ACG 5310 ACG 5320 ACG 5330 ACG 5340

LowB irthweight *

LowB irthweight *

LowB irthweight *

LowB irthweight *

ACG 5312

ACG 5311

ACG 5322

ACG 5321

ACG 5332

ACG 5331

ACG 5342

ACG 5341

MAC-26

MAC-24

# of ADG s

ACG 3800 1-16

35 +

17-34 Sex

ACG 3900

ACG 4000

ACG 4100

Female

Male

Age

2-3

Age

1-16 # ofMaj orADG s

17 + # ofMaj orADG s

0

ACG 5020

ACG 50101

ACG 50302 +

0 - 1

ACG 5050

ACG 50402

ACG 50603

ACG 50704 +

10 +

Figure 4: Decision Tree for MAC 24 -- Multiple ADG categories

17-44 # of Maj or ADG s

0

ACG 4320

ACG 43101

ACG 43302 +

1 - 16 # of Maj or ADG s

45 + # ofMaj orADG s

0

ACG 4420

ACG 44101

ACG 44302 +

Age

4-5

0ACG 4210

ACG 42201 +

35 +# of

Maj orADG s

ACG 4920

ACG 49102

ACG 49303

ACG 49404 +

Age

6-9

17-34 Sex

6- 16 # of Maj or ADG s

1 - 5 # of Maj or ADG s

0 - 1

M

F

# ofMaj orADG s

# ofMaj orADG s

0

ACG 4720

ACG 47101

ACG 47302 +

0

ACG 4820

ACG 48101

ACG 48302 +

ACG 4510

ACG 45201 +

0

ACG 4610

ACG 46201 +

0


KeyStarfield 04/97AC 1095

Documents

Measuring Primary Care and Its Benefits - … · Measuring Primary Care and Its Benefits Barbara Starfield, MD, MPH ... SWE JAP GER FR BEL US DK *Best ... Primary Care Score