Sa 407 Resource Allocation And Risk Adjustment

Front 1

Outline - Short

Back 1

Introduction
Resource allocation to health plan
Methodological Issues
Conclusion

Front 2

Outline – Long

Back 2

Introduction
Resource allocation to health plan
- Equity
- Efficiency
- Risk adjustment
Risk adjustment - Methodological Issues
- needs factors: legitimate and illegitimate, challenges
- Matrix approach
- Index approach with multilevel modeling
- Hybrid approach
Empirical data

Front 3

Introduction

Back 3

Resource allocation in health systems
Revenue raising
- Tax financed: hierarchical – central, regional, and local gov
- SHI – open enrollment: competition
Purchasing health services
- Budget
- Community-rated premiums
Providing health services
- Prospective:: budgets, capitation
- Retrospective: FFS, DRGs

Front 4

Resource allocation to health plan - Equity

Back 4

horizontal equity – equal access (opportunity) for equal need (risk)
- tax – equity between regions
- SHI – risk selection

Front 5

Resource allocation to health plan - Efficiency

Back 5

- scarcity of resources = fixed budget constraint – prospective payment has stronger incentives for cost-control by shifting risk to purchaser → provider
- maximize population health improvement by stopping point of equal marginal need met
- prospective payment – match need to access
o surplus – not technically efficient
o deficit – opportunity cost of health improvement foregone – quality skimping
- risk-adjusted capitation – unbiased estimate of a plan member’s expected cots based on members’ personal characteristics

Front 6

Risk-adjustment Methodological issues – Short

Back 6

Needs factors
- legitimate
- illegitimate
- challenges in selecting needs factors
Matrix Approach – individual data – Sweden
Index Approach with Mulitlevel modeling - UK
Hybrid approach
Further Issues

Front 7

Methodological issues – Needs factors - legitimate

Back 7

a. age, sex, race
b. employment/disability
c. geographical location
d. M & M
e. social factors – deprivation/SES
f. supply side – variations in unit cost eg MFF (Market forces factor in England, hospital and PCT – costs of staff, land, buildings); centers of excellence

Front 8

Methodological issues – Needs factors - illegitimate

Back 8

2. Illegimate = supply side
a. England - supplier induced demand – purchasers are not compensated for physician supply

Front 9

Methodological issues – Needs factors - challenges

Back 9

a. limited data – requires reliable and comprehensive data
b. limited research evidence on predictive
c. multicollinearity – double counting - unless redistributive objective – address inequitable utilization in resources
d. disentangle legitimate needs factors from illegitimate needs factors
- hospitalization is a good predictor of future HC utilization, but providers could manipulate
e. political process influenced by providers

Front 10

Methodological issues – Matrix approach

Back 10

1. Individual level data – claims data, administrative data – SHI
2. Preferable because unit of capitation is patient
3. Stockholm County
a. comprehensive records of social circumstances and health care utilization
b. age, sex, marital status, housing tenure, employment, previous inpatient diagnosis – may be manipulated by providers

Front 11

Methodological issues – Index approach with multilevel modeling

Back 11

1. Aggregate data at level of health plan/provider in tax-financed system
- EMR individual level data
2. Mortality is a poor proxy for morbidity
a. chronic diseases → sig HC utilization without mortality
b. eg depression, arthritis, asthma
3. Ecological fallacy – possibility of identifying a putative relationship between a putative needs factor and health care expenditure that does not hold on the individual level due to aggregation of two distinct populations → solution: small geographic areas eg England
4. Belgium – prior to 2002 – competitive sickness funds - series of indices on age, sex, unemployment, disability, mortality, urbanization

Front 12

Methodological issues – Hybrid approach

Back 12

1. Rudimentary matrix approach + index approach
2. England – geographic health authorities – age → mortality, morbidity, unemployment, elderly living alone, ethnicity, SES, cost variations (MFF)

Front 13

Further issues - Dynamic

Back 13

– should consider population migration: rural → urban
- update regularly to reflect rapid changes in HC technology

Front 14

legitimate needs factors

Back 14

Biological: Age, sex, ethnicity, disability, M&M
Social: SES/deprivation, employment, geographical
Supply: variations in unit cost – MFF

Front 15

Sweden needs factors

Back 15

Bio: age, sex, prior inpatient diagnosis
Soc: marital status, housing tenure, employment

Front 16

Belgium needs factors

Back 16

Bio: age, sex, disability, mortality
Soc: unemployment, urbanization

Front 17

UK: needs factors

Back 17

Bio: age, ethnicity, morbidity, mortality
Soc: SES, elderly living alone, unemployment
Supply: MFF