Cost-sharing for Emergency Care and Unfavorable Clinical Events: Findings from the Safety And Financial Ramifications of ED Copayments (SAFE) Study AcademyHealth.

Cost-sharing for Emergency Care and

Unfavorable Clinical Events: Findings from the

Safety And Financial Ramifications of ED Copayments (SAFE) Study

AcademyHealth Annual Conference7 June 2004

SAFE Study Team

Joseph P. Newhouse, PhDMaggie Price, MA

Richard Brand, PhDTom Ray, MBA

Bruce Fireman, MAJoseph V. Selby, MD, MPH John Hsu, MD, MBA, MSCE

Harvard UniversityKaiser Foundation Research Institute

University of California, San Francisco

Funding Support:

Agency for Healthcare Research and QualityNo other relevant financial relationships to disclose

Background

• Health Care Costs Are Increasing Each Year

• Millions of Americans Face Increasing Levels of Cost-sharing

– Both Higher Levels and Differential/Tiered Copays

– A Common Practice Is Higher ED Copays

• Clinical Impact of Higher Cost-Sharing for Emergency Services Is Unclear, Especially in Managed Care Environment

Cost-Sharing Evidence

• Important to Establish Outcome Effects

– No Insurance (Full Cost to Patient) Is Associated With Decreased Use of Medical Care and Worse Clinical Outcomes

– RAND HIE Showed ED Cost-sharing (Partial Cost to Patient) Is Associated With a Reduction in Use of Emergency Care in General Population; No Apparent Outcome Effect

– Entire ED Effect Within Lacerations on Non-Sutured Lacerations

• Comparison with the RAND HIE Results

– Cost-Sharing the Same for ED and Office Visits

– Did Not Sample Among Elderly

– Did Not Study Cost Sharing In Managed Care Settings

– Small Sample Meant Did Not Establish Effect on Mortality

Objective

To Investigate the Impact of Cost-sharing for Emergency Care

on Emergency Department (ED) Visits,

Deaths, Hospitalizations, and ICU Admissions

Methods• Design: Quasi-experimental Study with

Concurrent Controls (Diff-in-Diff)

• Natural Experiment: Increase in ED Copayment Levels for Over Half the Population

• Population: – 2,257,445 Patients with Commercial Insurance

– 261,091 Patients with Medicare Insurance

• Setting: Prepaid, Integrated Delivery System

• Time Period: 1999 - 2001

• Excluded: Patients With Medicaid

Cost-Sharing Levels• Commercial Insurance:

– Free Care: No Cost-sharing for ED care– $1 – 5 Copayments– $10 –15 Copayments– $20 – 35 Copayments– $50+ Copayments

• Medicare Insurance:– Free Care: No Cost-sharing for ED care– $1 – 15 Copayments– $20+ Copayments

There were no copayments other than the listed amounts during the study period.

Statistical Analysis• Poisson Random Effects Model

• Propensity Score for Covariates: Age, Gender, Comorbidity (DxCG-based),

Prior Utilization, SES (2000 US Census- based), Having a Regular Provider, Pharmacy Copayments, Medical Center

• Adjusted for Year and Month

Clinical Events

• ED Visits: In-system and Out-of-system

– All ED Visits

• Hospitalizations: In-system and Out-of-

system– All Non-elective Hospitalizations

– In-system Hospitalizations with ICU Admissions

• Deaths: In-system and Out-of-system

– All-cause Mortality

Results• 2,257,445 Subjects With Commercial Insurance in 1999

– 61% experienced increased cost-sharing during the study– 52% experienced increased cost-sharing in 2000– 21% experienced increased cost-sharing in 2001

• 261,091 Subjects With Medicare Insurance in 1999– 68% experienced increased cost-sharing during the study– 60% experienced increased cost-sharing in 2000– 13% experienced increased cost-sharing in 2001

• Mean ED Visits: – 18.4 Visits Per 100 Person-years (Commercial)– 52.0 Visits Per 100 Person-years (Medicare)

• Mean Hospitalizations: – 2.4 Hospitalizations Per 100 Person-years (Commercial)– 17.9 Hospitalizations Per 100 Person-years (Medicare)

Changes in ED Copayment Levels: Commercially Insured Subjects (1999-2001)

0

10

20

30

40

50

60

70

1999 2000 2001

Year

Pe

rce

nt

of

Co

ho

rt (

%)

$0 $1 - $5$10 - $15$20 - $35$50 - $100

n = 2,257,445 n = 1,989,248 n = 1,817,246

Commercial Insurance

ED Copayment Level

Changes in ED Copayment Levels: Medicare Insured Subjects (1999-2001)

0

10

20

30

40

50

60

70

80

1999 2000 2001

Year

Pe

rce

nt

of

Co

ho

rt (

%)

$0 $1 - $15$20 - $50

n = 261,091 n = 242,602 n = 228,046

Medicare Insurance

ED

Baseline Characteristics: Commercial

n (%)

Total 2,257,445 (100)

Age

<15yrs 480,188 (21.3)

15-17yrs 111,607 (4.9)

18-29yrs 375,413 (16.6)

30-39yrs 394,481 (17.5)

40-49yrs 413,975 (18.3)

50-64yrs 433,680 (19.2)

65-74yrs 36,106 (1.6)

75-84yrs 10,022 (0.4)

85+yrs 1,973 (0.1)

Female 1,146,478 (50.8)

Low SES Neighborhood 508,861 (22.5)

n (%)

Chronic Disease Status in 1998

Asthma 211,684 (9.4)

Diabetes 69,065 (3.1)

Heart Failure 6,936 (0.3)

Coronary Artery Disease 25,194 (1.1)

Hypertension 177,524 (7.9)

ED Visits in 1998

No Visits1,947,13

6 (86.3)

Office Visits in 1998

No Visits 555,516 (24.6)

Hospitalizations in 1998

No Hospitalizations2,222,70

4 (98.5)

Drug Cost-sharing

$0-$5 Copayments1,536,89

1 (68.1)

$7-$15 Copayments 556,085 (24.6)

Co-insurance 164,469 (7.3)

Baseline Characteristics: Medicare

  n (%)

Total261,09

1 (100)

Age

<15yrs 26 (0)

15-17yrs 0 (0)

18-29yrs 298 (0.1)

30-39yrs 1,309 (0.5)

40-49yrs 2,646 (1)

50-64yrs 9,924 (3.8)

65-74yrs 142,44

7 (54.6)

75-84yrs 85,074 (32.6)

85+yrs 19,367 (7.4)

Female 145,81

0 (55.8)

Low SES Neighborhood 49,251 (18.9)

n (%)

Chronic Disease Status in 1998

Asthma 24,646 (9.4)

Diabetes 35,750(13.7

)

Heart Failure 14,380 (5.5)

Coronary Artery Disease 36,600 (14)

Hypertension 118,204(45.3

)

ED Visits in 1998

No Visits 194,327(74.4

)

Office Visits in 1998

No Visits 22,146 (8.5)

Hospitalizations in 1998

No Hospitalizations 237,142(90.8

)

Drug Cost-sharing

$0-$5 Copayments 119,672(45.8

)

$7-$15 Copayments 130,166(49.9

)

Co-insurance 11,253 (4.3)

Unadjusted Rates of Clinical Events by ED Copayment Level Across All Years (1999-2001)

Commercial Insurance Population

EDCopaymentLevel

MeanOffice

Copayment

Person-years ED Visits Hospitalizations ICU

Admissions Deaths

Visits Rate* Admits Rate* Admits Rate* Counts Rate* $0 $0 1,077,940 219,349 20.35 26,419 2.45 7,331 0.68 1,940 0.18

$1-5 $5 1,102,233 216,740 19.66 26,445 2.40 6,914 0.63 2,236 0.20

$10-15 $12 600,072 110,254 18.37 10,852 1.81 2,392 0.40 640 0.11

$20-35 $8 2,490,448 441,951 17.75 64,556 2.59 17,454 0.70 5,453 0.22

$50-100 $11 377,448 53,425 14.15 7,248 1.92 1,969 0.52 531 0.14

Total 5,648,1421,041,71918.44 135,520 2.40 36,060 0.64 10,800 0.19

*Rates reported as events per 100 person-years. Patients with time-varying ED copaymentscontribute person-time to multiple copayment-levels.

Unadjusted Rates of Clinical Events by ED Copayment Level Across All Years (1999-2001)

Medicare Insurance Population

EDCopaymentLevel

Mean OfficeCopayment

Person-years ED Visits Hospitalizations ICU

Admissions Deaths

Visits Rate* Admits Rate* Admits Rate* Counts Rate* $0 $0 140,558 74,367 52.91 25,595 18.21 8,220 5.85 5,025 3.58

$1-15 $4 241,198 125,599 52.07 41,742 17.31 13,670 5.67 8,287 3.44

$20-50 $9 324,614 167,641 51.64 58,829 18.12 19,075 5.88 12,527 3.86

Total 706,369 367,607 52.04 126,166 17.86 40,965 5.80 25,839 3.66

*Rates reported as events per 100 person-years. Patients with time-varying ED copayments contribute person-time to multiple copayment-levels.

Adjusted Relative Rates of Clinical Events by ED Copayment Level:

Commercial Insurance Population

EDCopaymentLevel ED Visits Hospitalizations ICU Admissions Deaths

RR (95% CI) RR (95% CI) RR (95% CI) RR (95% CI) $0 1.0 1.0 1.0 1.0

$1-5 0.962 (0.955, 0.970) 0.999 (0.978, 1.020) 1.010 (0.972, 1.050) 1.086 (1.021, 1.156)

$10-15 0.932 (0.922, 0.941) 0.932 (0.905, 0.960) 0.895 (0.845, 0.947) 0.909 (0.822, 1.006)

$20-35 0.879 (0.873, 0.886) 0.961 (0.943, 0.980) 0.954 (0.922, 0.988) 0.857 (0.810, 0.907)

$50-100 0.765 (0.756, 0.774) 0.902 (0.873, 0.932) 0.946 (0.891, 1.004) 0.903 (0.818, 0.998)

Adjusted Relative Rates of Clinical Events by ED Copayment Level:

Medicare Insurance Population

EDCopaymentLevel ED Visits Hospitalizations ICU Admissions Deaths

RR (95% CI) RR (95% CI) RR (95% CI) RR (95% CI) $0 1.0 1.0 1.0 1.0

$1-15 0.973 (0.956, 0.989) 0.987 (0.962, 1.013) 0.990 (0.949, 1.031) 1.007 (0.966, 1.050)

$20-50 0.956 (0.939, 0.973) 0.994 (0.967, 1.022) 1.003 (0.959, 1.049) 0.873 (0.834, 0.913)

Adjusted Relative Rates of Clinical Events by ED Copayment Level in Subjects Living in Low SES Neighborhoods*:

Commercial Insurance Population

EDCopaymentLevel ED Visit Hospitalization ICU Admission Deaths

RR (95% CI) RR (95% CI) RR (95% CI) RR (95% CI) $0 1.0 1.0 1.0 1.0

$1-5 0.930 (0.916, 0.944) 0.955 (0.915, 0.996) 0.971 (0.898, 1.050) 1.057 (0.932, 1.199)

$10-15 0.909 (0.891, 0.928) 0.919 (0.866, 0.976) 0.876 (0.779, 0.985) 0.996 (0.810, 1.226)

$20-35 0.847 (0.835, 0.859) 0.947 (0.911, 0.984) 0.931 (0.866, 1.000) 0.922 (0.823, 1.033)

$50-100 0.737 (0.719, 0.756) 0.842 (0.787, 0.900) 0.826 (0.727, 0.938) 1.084 (0.885, 1.327)

* We defined low SES neighborhoods as areas with low income (≥20% of residents have household incomes below the federal poverty level) or low educational attainment (≥25% of residents ≥25 years old have less than a high-school education).

Adjusted Relative Rates of Clinical Events by ED Copayment Level in Subjects Living in Low SES Neighborhoods*:

Medicare Insurance Population

* We defined low SES neighborhoods as areas with low income (≥20% of residents have household incomes below the federal poverty level) or low educational attainment (≥25% of residents ≥25 years old have less than a high-school education).

EDCopaymentLevel ED Visit Hospitalization ICU Admission Deaths

RR (95% CI) RR (95% CI) RR (95% CI) RR (95% CI) $0 1.0 1.0 1.0 1.0

$1-15 0.924 (0.889, 0.961) 0.953 (0.897, 1.013) 0.959 (0.872, 1.055) 0.894 (0.814, 0.983)

$20-50 0.929 (0.893, 0.967) 0.961 (0.904, 1.022) 0.942 (0.852, 1.041) 0.846 (0.765, 0.937)

Conclusions

In This Population of Patients in a Prepaid, Integrated Delivery System:

• Having to Pay a Portion of ED Costs Reduced ED visits, and by Roughly the Same Amount as in the RAND HIE

• There was No Evidence of Clinical Harm Associated with Having to Pay Higher ED Costs, i.e. Higher Cost-Sharing Did Not Result in More Hospitalizations or Deaths.

Limitations

• Range of ED Cost-sharing Levels: Free Care to $100 Copayments Per Visit

– But Copays Above $100 per Visit Are Rare

• No Measures of Patient Awareness– May Have Understated Steady State Effects

• Single Integrated Delivery System

Implications

• Moderate Levels of Cost-Sharing for Emergency Services Appear Save Money With No Evidence of Adverse Effects