DRG Workshop Belgrade, 18-22.November 2013. Data analysis by hospitals. Data analysis by payers. RIC

DRG Workshop Belgrade, 18-22.November 2013.

Data analysis by hospitals. Data analysis by payers.

RIC


ELEMENTS OF DATA MANAGEMENT

• Data reporting policies and specifications– Data dictionary– Data set specification (eg NMDS)– Data entry tools– Data extraction tools– Data edits and validation processes

• Data warehousing functions– (incl) metadata management, data access

• Data analysis reporting


INFORMATION SYSTEMS SCHEMA

SUPPLY HR FMIS PAS EHR

MPI

MHODS

AggregateAEDEMDESIS

MDSs

REPORTREPOSITORY

CUBESPIVOT

COGNOS,BRIO, EXCELLBUSINESS OBJECTSSAS ,SPSS, STATA, ACCESS

BUSINESS LAYER

DATAWAREHOUSE

REPORTING LAYER (MIS)

STATISTICAL ANALYSIS

MHCMI

MHDW

Oracle, SAP, etc


Hospital Information SystemsHospital Information SystemsHospital Financial DataHospital Financial Data

Minimum Basic Data Set-DRG Data Entry Tool- Clinical Data, DRG- Resource Consumption Data

(analysis tool)

Combo Software

Coding Analysis Reports

PICQ Software

Cost Reports

Cost Sheet-Allocation Statistics

(costfile)(volumefile)(CASES)

Data Analysis Reports

(preparecostsheet)

Diagnosis and procedures Analysis Reports


REPORTING FEEDBACK

• The key to – GOOD DATA– EFFICIENCY GAINS– PERFORMANCE IMPROVEMENT

• The basis for rational planning

• A primary mechanism to assess innovation and investment priorities.


Standard reports may include

• Productivity

• Complexity

• Allocative efficiency

• Technical efficiency

• Coding Completeness

• And various quality indicators– Outcome – Process

• Some examples follow …


Productivity

• Performance in total Weighted Episode (WEs) of all hospitals

• Overall performance in total WEs against target by hospital

• WEs by Major Diagnostic Categories (MDC) for Medical Diagnosis Related Group (DRGs) by hospital

• WEs by MDC for Procedural DRGs by hospital• Top 10 increment DRG families by hospital • Top 10 decrement DRG families by hospital


Complexity

– Casemix Index of Medical and Procedural DRGs of all hospitals

– Casemix Index (CMI) by MDC by hospital– Average WEs per patient by MDC in by

hospital– Average number of episodes per patient by

MDC by hospital– Casemix – “1 Year On”


Allocative Efficiency

• Percentage same day episodes by MDC of Medical and Procedural DRGs by hospital

• Change in same day episodes of Ambulatory Medical DRGs

• Percentage of episodes admitted via A&E Department by MDC by hospital


Technical Efficiency

• Average Length of Stay (ALOS) of Medical & Procedural DRGs of all hospitals

• Change in ALOS by MDC by hospital

• Cost per weighted DRG by hospital


Coding Completeness

– Percentage of total multi-day episodes for Medical and Procedural DRGs by severity level in all hospitals

– Percentage of total multi-day episodes with Major Co-morbidities and Complication (MCC) by MDC for Medical DRGs by hospital

– Percentage of total multi-day episodes with MCC by MDC for Procedural DRGs in all hospitals

– Percentage of total multi-day episodes with MCC by MDC by hospital

– WE / CMI by discharged specialty: 4-year trend analysis by hospital


WHY IS PATIENT LEVEL COSTING SO IMPORTANT?

12


The proposition

Tracking utilisation to patients is actually more important than simply accurate cost weights.


How can this be true?

MP Ann Clwyd wept in the House of Commons as she described how her husband [Owen] died 'like a battery hen'Labour MP Ann Clwyd says her biggest regret was that she didn't 'stand in the hospital corridor and scream'3 in protest of the 'callous lack of care' the NHS nurses showed her husband, as he lay dying of hospital acquired pneumonia.Ann describes her husband's final days, crushed 'like a battery hen' against the bars of his hospital bed, wearing an oxygen mask that was so tight it cut his face and pumped cold air in to his infected eye. Owen died after suffering 10 days of this treatment at the NHS University Hospital of Wales in Cardiff. Following her husband's death, the MP publicly spoke out on the 'normalisation of cruelty' that has become 'commonplace' in NHS hospitals.Owen had battled MS for 30 years, yet died of hospital-borne pneumonia. Shockingly, around 100,0004 patients catch an infection in UK hospitals every year, from which 5,000 people sadly die.http

://www.to

p10healthinsurance.com/artic

le/factsabouthealth

insurance?src=tab


KEY POINTS

1. Good data for clinical care monitoring (inputs utilization analysis) is also good costing data.

2. Utilization by case type gives accurate detailed costing – AND benchmarking patient level variability is vital for both

pricing and quality-of-care management.3. Clinicians should be the primary users of utilization

data and analyses – CLINICAL BEHAVIOURS << CLINICAL CULTURES <<

OPERATIONAL USE OF DATA4. Key immediate strategies.


PRICING HEALTHCARE

• 1. Improve data quality• 2. Increase comparability and consistency of data• 3. Improve transparency• 4. Develop the potential for new pricing

mechanisms• 5. Proportionate regulatory cost• 6. Improve the use of cost data by managers and

clinicians






21

Scenario A

Scenario B

Scenario C

Figures in £000s)

April 19, 2023

-40K

-20K

0K

20K

40K

60K

80K

12/13 Surplus/Deficit (£000)

12/13 Surplus/Deficit

Key12/13 Surplus/Deficit

The advent of the ‘S’ curve

2012/13 surplus/ deficit

2014/15 projected surplus/ deficit (expected case)

2014/15 projected surplus/ deficit with 1% annual efficiency shortfall

DRG Workshop Belgrade, 18-22.November 2013. 22

-60K

-50K

-40K

-30K

-20K

-10K

0K

10K

20K

30K

40K

50K

60K

70K

80K

90K

14/15 Sensitivity Case

-60K

-50K

-40K

-30K

-20K

-10K

0K

10K

20K

30K

40K

50K

60K

70K

80K

90K

14/15 Expected Case

14/15 Expected Case with 1 % Annual Efficiency Shortfall

Start in surplus but move into deficit

-60K

-50K

-40K

-30K

-20K

-10K

0K

10K

20K

30K

40K

50K

60K

70K

80K

90K

14/15 Sensitivity Case (£000)

-60K

-50K

-40K

-30K

-20K

-10K

0K

10K

20K

30K

40K

50K

60K

70K

80K

90K

14/15 Expected Case (£000)

14/15 Expected Case with 1 % Annual Efficiency Shortfall

208%

22992%

14/15 Expected Case Provider Split14/15 Expected Case

Deficit

Surplus

14458%

10542%

14/15 Sensitivity Case Provider Split14/15 Sensitivity Case

Deficit

Surplus

Remain in surplus

Start and remainin deficit

Start in surplus but move into deficit

Scenarios 2 and 3


The two key questions

1. How do we get the right data from costing systems into regular clinical review use?

2. How do we better enable clinicians to take the lead in performance improvement?


QUESTION 1.

• How do we get the right data from costing systems into regular clinical review use?


RELEVANCE

Kane Gorny, 22, left to die of thirst in an NHS London hospitalKane Gorny, an active 22-year-old, had already bravely beaten cancer the year before and needed a routine hip replacement due to the treatment. Whilst recovering from this simple operation, Kane suffered unthinkable neglect at the hands of his doctors and nurses in St. George's NHS hospital, Tooting.Having been denied his medication and refused his pleas for water; driven by thirst and desperation, Kane called 999 from his hospital bed. But when the police arrived to his aide, they were maliciously turned away by hospital staff who claimed he was delirious. Kane died of dehydration just hours later! Whilst his grieving mother took a moment to sit with her son's body, a nurse then heartlessly asked if she was finished so that they could 'bag him up'1.Disgustingly, Kane's case is not alone; 4 people die of dehydration every day in our hospitals2.

…Health insurance is one way to ensure you

receive the right treatment, when you need it.

[!!!!!?????]http

://www.top10healthinsurance.com/article/factsabouthealthinsurance?src=ta

b


The locus of efficiency and effectiveness

“EVERY CLINICAL DECISION IS AN EXPENDITURE DECISION”

Don Hindle


But patient 15644 was different from normal

Look! You spent too much on patient 15644

Don Hindle


Sorry, I'll try not to do it again

I've checked, and patient 15644 wasn't different at all!

Don Hindle


Mean cost/case

NursingMedicalAllied healthORPathologyImagingPharmacyConsumablesOverheadOther

Total

Our hospital

750500170420

9065

12075

450200

$2850

Other hospitals

550420180340

8050

11090

380200

$2400

Benchmarking by item cost only ..FINANCIAL VIEW …

Don Hindle


Elements of cPaths

Admitted how long before procedure?

Post-op hours in CCU

Review of discharge plan at admission?

Discharge OK delegated?

Our hospital

9 hours

6 hours

No

No

Other hospitals

6 hours

Nil

Yes

Yes

…and cost/utilisation benchmarking .. CLINICIAN VIEW

financial data and clinical pathwaysFrom Don Hindle


TIMELINESS

• We need to be able to review whether changes are working

• We need to be able to correct and steer regularly – monthly – weekly

• We need to be able to check immediate activity against norms – agreed protocols - to guide decisions and choices


EXAMPLE DATA FLOW SCHEMA

SUPPLY HR FMIS PAS EHR

MPI

MHODS

AggregateAEDEMDESIS

MDSs

REPORTREPOSITORY

CUBESPIVOT

eg. COGNOS,BRIO, EXCELLeg. BUSINESS OBJECTSeg. SAS ,SPSS, STATA, ACCESS

BUSINESS LAYER

DATAWAREHOUSE

REPORTING LAYER (MIS)

STATISTICAL ANALYSIS

MHCMI

MHDW

eg. Oracle, SAP, etc


GRANULARITY

• Why are the costs higher / lower?• What are we doing differently?• Are our patient-type ratios different?• Are we differentiating – tailoring the pathways

to different patients – is our variability matching standard patterns


QUESTION 2.

• How do we better enable clinicians to take the lead in performance improvement?


Clinical process review

• Patient journey• Clinical pathway• Evidence based practice• Variance analysis


QUESTION 3

• How can clinicians match best practice and properly accommodate natural patient variability?


Culture eats reform for breakfast

• “I am a doctor/nurse not a manager – I care for people not ‘cases’. Everyone is different.”

• Is the care team or the individual responsible• “Why are we doing this type of case

differently?”• What is the evidence that it works better?• “How can we redeploy the savings if we can

achieve less costly practice?”


So ..how well are we doing?

Elderly and frail 'Mrs H'5 bruised, soaked in her own urine and heavily dishevelled under NHS 'care'Mrs H was an independent 88-year-old, living in her own home - until she had a fall. Admitted to the Heart of England NHS Trust, Mrs H suffered further disturbing ailments, including a broken collar bone. Upon her transferral to a new care home, the manager voiced concern over her bruised, soiled and distressed state. She arrived in clothing that was not her own, held in place by large paper clips, along with several bags of dirty clothes. Sadly Mrs H died before the investigation into her cruel treatment was complete and the NHS Trust admitted their failings.An astonishing 3,0006 complaints are made about the NHS every single week, illustrating that the core 'care' principle of nursing seems to be disappearing rapidly from its practice. NHS nurses even claim the new generation of university-trained nurses believe basic care tasks, such as bathing their patients, are 'beneath them'7.

If you were to fall il

l, it's only natural th

at you

would want to be tre

ated and cared for to a high

standard: Private faciliti

es have strong

reputations to uphold and less time pressures,

meaning consultants can dedicate more attention

to individual patients.

http://www.top10healthinsurance.com/artic

le/factsabouthealthinsurance?src=tab


Definition of Clinical Pathways• Systematically developed written statements that:

• Are evidence based

• Incorporate the views of clinicians (medical, nursing, allied health) patients and managers

• Prospectively describe how care will provided and how effectiveness, quality and efficiency will be monitored and assessed.

From Pieter Degeling


A clinical pathway enables a clinical team to:

prospectively describe the sequence of diagnostic and therapeutic

events whose occurrence or non-occurrence will significantly affect quality, cost and outcome.

prospectively define the quality and outcome indicators that it will use to judge its performance.

prospectively cost the care that it provides and in so doing defines the cost indicators that it will use to judge its efficiency.

routinely and consistently monitor its performance on quality and cost

alter its method for treating a condition, in light of what it [systematically] has learned about treating the condition.



Quality indicators

Outcome Indicators

Routine Review of Variance

Prospectively Costed

Clinical Pathway

Characteristics of Integrated care pathways



Fin

al Pro

duct

s

Intermediate Products

Hip ICP 1

Hip ICP 2

Knee ICP 1

# ICP 1

Imaging Pharmacy Path. Hotel services

Pathway Based Clinical ProductionPathway Based Clinical Production



Important provisos

• CPs are not immutable documents setting out inviolable treatment regimens.

• The existence of a pathway does not obviate clinicians’ responsibility to make clinical judgements and to tailor care according to their assessment of the clinical needs of individual patients.

• Thus clinical variation remains a ‘to be expected’ (in the sense of an often required) feature of clinical practice.

• The matter at issue is what a clinical team can learn from these variations and how they can systematize this learning.

• Accordingly, when the care process varies from that described in the pathway, the reasons for the variance are recorded and become the focus of structured across-profession conversations described above.



High volume case types –

how many pathways? Within a 600 -700 bed DGH in the NHS

Emergency admissions account for 53% of all inpatient episodes and 83% of all bed days consumed within the Trusts

40 HRGs (of 603) account for 46% of emergency admissions and 42% emergency generated bed days

40 HRGs account for 60% of in patient elective episodes and 40% of elective bed-days, and

40 HRGs account for 84% of day elective episodes

10 HRGs account for 98% of all maternity and births admissions and 97% of maternity and birth bed days



Year to Date February 2013

A-DRG Elective Inpatients No of Adm % Cumm %

V60 Alcohol Intoxication and Withdrawal 158 5.24% 5.24%

F15 Interventional Coronary Procedures W/O AMI W Stent Implantation 115 3.81% 9.05%

G10 Hernia Procedures 113 3.74% 12.79%

F14 Vascular Procedures Except Major Reconstruction W/O CPB Pump 107 3.55% 16.34%

D11 Tonsillectomy and/or Adenoidectomy 105 3.48% 19.81%

L04 Kidney, Ureter and Major Bladder Procedures for Non-Neoplasm 68 2.25% 22.07%

H08 Laparoscopic Cholecystectomy 60 1.99% 24.06%

O01 Caesarean Delivery 50 1.66% 25.71%

F06 Coronary Bypass W/O Invasive Cardiac Investigation 45 1.49% 27.20%

F08 Major Reconstructive Vascular Procedures W/O CPB Pump 44 1.46% 28.66%

E01 Major Chest Procedures 43 1.42% 30.09%

I03 Hip Replacement 43 1.42% 31.51%

I30 Hand Procedures 43 1.42% 32.94%

L07 Transurethral Procedures Except Prostatectomy 43 1.42% 34.36%

V61 Drug Intoxication and Withdrawal 42 1.39% 35.75%

F12 Implantation or Replacement of Pacemaker, Total System 42 1.39% 37.14%

G02 Major Small and Large Bowel Procedures 41 1.36% 38.50%

I04 Knee Replacement 39 1.29% 39.79%

B02 Cranial Procedures 38 1.26% 41.05%

E02 Other Respiratory System OR Procedures 36 1.19% 42.25%


Year to Date February 2013

A-DRG Day Cases No of Adm % Cumm %

L61 Haemodialysis 14656 58.97% 58.97%

F74 Chest Pain 515 2.07% 61.04%

F42 Circulatory Disorders W/O AMI W Invasive Cardiac Investigative Procedures 448 1.80% 62.84%

G64 Inflammatory Bowel Disease 447 1.80% 64.64%

I66 Inflammatory Musculoskeletal Disorders 419 1.69% 66.32%

Z60 Rehabilitation 410 1.65% 67.97%

Q60 Reticuloendothelial and Immunity Disorders 408 1.64% 69.62%

Z64 Other Factors Influencing Health Status 336 1.35% 70.97%

I74 Injury to Forearm, Wrist, Hand or Foot 320 1.29% 72.26%

I30 Hand Procedures 307 1.24% 73.49%

X60 Injuries 232 0.93% 74.42%

G66 Abdominal Pain or Mesenteric Adenitis 218 0.88% 75.30%

O60 Vaginal Delivery 206 0.83% 76.13%

B68 Multiple Sclerosis and Cerebellar Ataxia 168 0.68% 76.81%

Q61 Red Blood Cell Disorders 156 0.63% 77.43%

G70 Other Digestive System Diagnoses 147 0.59% 78.02%

G67 Oesophagitis and Gastroenteritis 144 0.58% 78.60%

I75 Injury to Shoulder, Arm, Elbow, Knee, Leg or Ankle 139 0.56% 79.16%

B77 Headache 134 0.54% 79.70%

F76 Arrhythmia, Cardiac Arrest and Conduction Disorders 121 0.49% 80.19%


Why high volume case types?

High volume case types are those for which:

• Pathways will produce the biggest ‘bang for the buck’ as we strive to improve:

• Efficiency• Effectiveness• Quality

• We can generate valid data for statistical analysis of variance



Variance Analysis

For each high volume case type collect, collate and analyze data on:

- the occurrence and non occurrence of the events described in a pathway

- quality indicators, clinical outcomes indicators and resource usage

These data are then used to:- profile and compare performance within and between hospitals over

time

- identify changes in practice that will lift performance to the benefit of improvements in quality, appropriateness, safety and efficiency including changing the pathway



QUESTION 4

• Which data on their caseload patterns and peer norms are most useful?


MACRO TO MICRO DRILLDOWNS

• Totals to clinical units – individual clinical teams

• Totals to particular DRGs – patient types – • Inputs across peers for sub sets of case

categories


VARIANCES FROM TARGETS

• Monthly patterns – modelled utilisation• Modelled seasonal variations• Modelled patient type demand/supply

projections– Population factors – Supply factors – eg leave – theatre rostering –

training schedules –


QUESTION 5

• How can quality and safety risk processes/practices be identified before the scandal / disaster / avoidable mortality rates / avoidable complications rates arise?


Quality and efficiency improvement oriented clinical unit management

Requires condition specific multidisciplinary conversations on questions such as:Are we doing the right things?

In light of assessed health needs and existing resource constraints, are we delivering value for money?

How appropriate and effective is the composite of activities that we undertake in treating this condition?

Are we doing things right? How systematized and reliable are our care processes for this condition? Are we clear about who should do what and when? How are we performing on risk, safety, quality, patient evaluation, clinical outcomes and

efficiency?

What can and must we do to improve quality and efficiency?



The evidence shows that:

The improvement focus of funding and clinical unit management systems are sharpened greatly with the development and implementation of clinical pathways that for each high volume case type:

– describe the detailed composition of the care process, – define cost, – define quality and outcome measures , – provide clinical teams with a basis for monitoring and improving

their performance



USING DRG DATA TO ASK THE QUESTIONS THAT STIMULATE CHANGE.

55


Warning flags – monthly variations on the big picture


Statistical and clinicial review



High volume DRG cost pattern query


Surplus DRG price to cost pattern query


Deficit cost to price DRG cost pattern query


So now to the next generation of casemix - caremix

• Eg MODELS OF CARE• Includes efficiency – that incorporates value – the

‘right thing’ and the ‘thing right’• So new types of episodes – integrated plans

across settings/providers – eg “year of care”• How do we cost these cross setting/provider

episodes and find a like with like benchmarking mechanism?

• The answer is we have to do both – not either-or.


Example of clinician feedback and data validation dynamics?

Documents

DRG Workshop Belgrade, 18-22.November 2013. Data analysis by hospitals. Data analysis by payers. RIC