Hospital Utilization by Fee-for-Service and Medicare Advantage Enrollees

Select to access the PowerPoint® presentation (407 KB).

---

Slide 1

 

Hospital Utilization by Fee-for-Service and Medicare Advantage Enrollees

Lauren Hersch Nicholas
University of Michigan
September 15, 2009

 

Slide 2

 

Motivation

• Ongoing policy interest in expanding Medicare benefits while reducing spending
• Medicare Advantage plans provide a voluntary, managed care alternative to Fee-for-Service
• Payments to plans now exceed average FFS spending
• Little is known about quality or cost implications of increasing enrollment in Medicare Advantage plans

 

Slide 3

 

Research Questions

• Does managed care affect hospital utilization for Medicare beneficiaries?
  • Quality of outpatient care: Ambulatory Care Sensitive Admissions
  • Access to elective procedures: Referral-Sensitive Admissions
• Does managed care enrollment affect total Medicare spending?

 

Slide 4

 

Background

• Existing quality and utilization literature indicates quality problems in early Medicare managed care plans
• Yet managed care consistently better at preventive service use
• Cost spillovers from managed care believed to hold down FFS spending, but higher payments to plans raise total spending
• Managed care plans historically attract healthier enrollees
• Findings mostly from 1990s, don't identify casual effects

 

Slide 5

 

State Inpatient Database

• Discharge abstracts from hospitalizations in AZ, FL, NJ, and NY
  • 20% of Medicare beneficiaries and 25% of Medicare Advantage enrollees live in one of these 4 states
• All in-state hospitalizations from 1990-2005
• Include Medicare Advantage and Fee-for-Service beneficiaries
• ICD-9 diagnostics and procedure codes used to identify ambulatory care sensitive (AHRQ Prevention Quality Indicators) and referral-sensitive admissions
• Marker hospitalizations, which are not affected by medical care, provide comparison group
• Medicare enrollment date ? demographic information for all beneficiaries

 

Slide 6

 

Ambulatory Care Sensitive Admissions

• Potentially avoided with effective primary care

Hospitalizations per 1,000	MMC1	FFS
Acute
Dehydration	2.45	4.65
Pneumonia	7.47	13.58
Reptured Appendix	0.2	0.27
Urinary Tract Infection	3.15	5.69
Chronic
Angina	0.9	1.24
Asthma	1.32	2.15
Chronic Obstructive Pulmonary Disease	5.63	8.21
Congestive Heart Failure	12.51	19.5
Diabetes Short-term	0.28	0.39
Diabetes Long-term	2.27	3.25
Diabetes Uncontrolled	0.34	0.53
Diabetes Amputation	0.70	1
Hypertension	0.93	1.2

 

Slide 7

 

Referral-Sensitive Admissions

• Technology-intensive procedures, require referral
• Low rates of procedures may suggest barriers to service use

Hospitalizations per 1,000	MMC	FFS
Angioplasty	6.09	7.60
Coronary Artery Bypass	3.02	3.21
Elective Joint Replacement	5.42	8.16
Pacemaker Insertion	1.8	2.61

 

Slide 8

 

Marker Admissions

• Hospitalizations which are unrelated to recent medical care, reflect underlying health status, private information influencing insurance choice and utilization

Hospitalizations per 1,000	MMC	FFS
Appendicitis	0.26	0.34
Gastrointestinal Obstruction	2.38	3.81
Hip Fracture	3.95	6.53

 

Slide 9

 

Unadjusted Rates of Hospitalization for Medicare Advantage and Fee-for-Service Enrollees

Rate per 1,000

• ACS
  • Managed Care: 37.6
  • Fee for services: 60.9
• Referral Hospitalization Type
  • Managed Care: 16.3
  • Fee for services: 21.6
• Marker
  • Managed Care: 6.6
  • Fee for services: 10.7

 

Slide 10

 

Medicare Advantage and Fee-for-Service Enrollees are Demographically Similar

Variable	MMC	FFS
Black	11%	9%
Hispanic	4%	4%
Other Race	3%	4%
Female	58%	58%
Medicaid	8%	14%
Age	75.0	75.1
N	27,117,977	89,671,934

Source: Medicare Denominator File, 1999-2005

 

Slide 11

 

What explains differences in hospital utilization?

• Medicare Advantage plans attract healthier enrollees, otherwise provide the same care as Fee-for-Service
• Medicare Advantage plans manage care to limit utilization, ? reduce elective procedure use
• Medicare Advantage plans manage care to preserve beneficiary health, ? reduce potentially preventable admissions

 

Slide 12

 

Empirical Approach

• Insurance Type-Country-Year level regressions of rate of hospitalization on Medicare coverage type and demographics
• County and Year fixed effects
• Two-stage estimation procedure using ratio of observed to expected marker hospitalizations to control for unobserved health status differences
• Pairs-Cluster Bootstrap used to calculate standard errors

 

Slide 13

 

Effect of Managed Care on Rates of Hospitalization (1)

Difference in Rates of hospitalization per 1,000 Enrollees

	ACS	Acute ACS	Chronic ACS	Referral
MMC	-12.54*** (4.07)	-5.93*** (1.60)	-6.61** (2.64)	-4.06* (2.11)

Cluster robust standard errors in parentheses
* p < 0.10
** p < 0.05
*** p < 0.01

 

Slide 14

 

Effect of Managed Care on Rates of Hospitalization (2)

• Managed care significantly reduces potentially preventable hospitalizations
  • Acute reductions primarily from Pneumonia and Urinary Tract Infection? Earlier access to antibiotics?
• No overall managed care effect for referral-sensitive hospitalizations, but significant reduction in elective joint replacement (3.5 per 1,000 enrollees) and pacemaker insertion (0.9 per 1,000)
• Positive selection into Medicare Advantage plans accounts for between 25 and 35 percent of risk-adjusted differences

 

Slide 15

 

Trends in Ambulatory Care Sensitive Admissions in Medicare Advantage and Fee-for-Service

A chart showing the trends in Ambulatory Care Sensitive Admissions in Medicare Advantage and Fee-for-Service is shown.

 

Slide 16

 

Medicare Advantage and Medicare Spending

• Nationally, 1% increase in Medicare Advantage enrollment increases average Medicare spending between 0.3 and 1.1%
• Is extra spending on managed care cost-effective way to reduce ACS admissions?
• Increasing plan payment rates by $600 per enrollee per year would reduce ACS admissions rate by 1 per 1,000

 

Slide 17

 

Conclusions and Policy Implications

• Medicare Advantage plans have lower rates of ambulatory care sensitive admissions
• No overall difference in referral-sensitive admissions
• Both positive selection and true "managed care effect" explain observed differences in utilization
• Higher payments to plans concentrate enrollment on healthier enrollees, hospitalizations primarily reduced by low-cost interventions
• Potential to reduce total spending by improving access to acute care in FFS?