the incidence of medicare payment reduction: evidence from the bba of 97 vivian y. wu university of...
TRANSCRIPT
The Incidence of The Incidence of Medicare Payment Medicare Payment Reduction: Evidence Reduction: Evidence from the BBA of 97from the BBA of 97
Vivian Y. WuVivian Y. WuUniversity of Southern CaliforniaUniversity of Southern California
AcademyHealth Annual Research Meeting, June 5, AcademyHealth Annual Research Meeting, June 5, 2007 2007
Responses to Medicare Responses to Medicare Payment CutsPayment Cuts
1.1. Charge private payer higher Charge private payer higher pricesprices
2.2. Improve efficiency: shorter Improve efficiency: shorter length of stay, less cares, …. length of stay, less cares, …. etc, without hurting qualityetc, without hurting quality
3.3. Lower quality of care Lower quality of care
Research questionResearch question
Who bears the burden of 1997 Who bears the burden of 1997 BBA Medicare reimbursement BBA Medicare reimbursement cuts? cuts?
Source: Dobson, A. et al., Health Affairs, Vol 25(1), 22-33
Prior ResearchPrior Research
1980’s: mixed1980’s: mixed Early 1990’s: cost-shifting diminishing Early 1990’s: cost-shifting diminishing Around BBA 97: Around BBA 97:
– Bernard, 2000 studied cross-subsidization Bernard, 2000 studied cross-subsidization between 1994-1998, elasticity was -0.5.between 1994-1998, elasticity was -0.5.
– Zwanziger and Bamezai, 2006, found cost-Zwanziger and Bamezai, 2006, found cost-shifting between 1993-2001 in CA was -shifting between 1993-2001 in CA was -0.17. 0.17.
Research QuestionsResearch Questions
Central questions: Central questions: – Do and can hospitals raise prices to Do and can hospitals raise prices to
private payers? private payers? – Does the behavior differ by Does the behavior differ by
ownership type?ownership type?– Does market environment Does market environment
(ownership composition and (ownership composition and managed care) have any impact on managed care) have any impact on this behavior? this behavior?
MethodMethod
Main Model: Main Model: – Long-difference model at hospital Long-difference model at hospital
level: level:
ΔΔ private price = private price = ΔΔ Medicare loss + Medicare loss + controlcontrol
Method Method
Key identifying variable: BBA “bite” Key identifying variable: BBA “bite” variablevariable
0.94
0.96
0.98
1
1.02
1.04
1.06
1.08
1.1
1.12
1996 1997 1998 1999 2000
BBA reduction
Market basket increase Bite
MethodMethod
Dependent variable: Dependent variable: – Private “price”: Private “price”:
Private revenue / private dischargesPrivate revenue / private discharges Private revenue / private daysPrivate revenue / private days
– Private LOSPrivate LOS
Method: Formal ModelMethod: Formal Model
PP((ii, t, t-1, t, t-1) = ) = ii + + Bite Bite((ii, t, t-1, t, t-1))
+ + Bite Bite((ii, t, t-1, t, t-1) * ownership) * ownership((ii, ,
t-1t-1))
+ + δ δ Bite Bite((ii, t, t-1, t, t-1) * HMO IV) * HMO IV((ii, t-1, t-1))
+ + ηη Bite Bite((ii, t, t-1, t, t-1) * FP Share) * FP Share((ii, t-, t-
11))
+ + λλ X X((ii, t, t-1, t, t-1)) + + X X((ii, t-1, t-1)) + + ((ii, t, t))
MethodMethod
Key independent variable: Key independent variable: – Ownership type: Teaching, NFP, Public Ownership type: Teaching, NFP, Public – FP market effect: % FP discharges in MSA FP market effect: % FP discharges in MSA – HMO effect: Instrument for HMO penetration HMO effect: Instrument for HMO penetration
(% in large firms, % white collar)(% in large firms, % white collar)
Other controls: Other controls: – ΔΔ case mix, size (beds), SNF, HH, and case mix, size (beds), SNF, HH, and
market dummies (HRR)market dummies (HRR)
ResultsResults
ΔΔ Private Rev per Private Rev per Private AdmissionPrivate Admission
ΔΔ Private Rev Private Rev per Private Dayper Private Day
ΔΔ Private LOS Private LOS
IP BiteIP Bite -.76**-.76**
[.14][.14]-.53**-.53**
[.12][.12]-.0008**-.0008**
[.00008][.00008]
SNF BiteSNF Bite .11.11
[.10][.10]-.01-.01
[.02][.02]-.00001-.00001
[.00008][.00008]
HH BiteHH Bite .03.03
[.04][.04]-.0003-.0003
[.009][.009].000007.000007
[.00003][.00003]
ΔΔ case mix case mix 482482
[556][556]4444
[124][124].29.29
[.41][.41]
TeachTeach 198198
[228][228]-28-28
[47][47].18.18
[.16][.16]
NFPNFP -381**-381**
[150][150]-36-36
[33][33]-.05-.05
[.11][.11]
PublicPublic -154-154
[188][188]-12-12
[41][41]-.13-.13
[.14][.14]
HMO IVHMO IV 2.692.69
[6.31][6.31]0.910.91
[1.34][1.34]-.01**-.01**
[.0005][.0005]
ResultsResultsΔΔ Private Rev per Private Rev per
Private AdmissionPrivate AdmissionΔΔ Private Rev per Private Rev per
Private AdmissionPrivate Admission
IP BiteIP Bite -1.06-1.06
[.25][.25]-.36*-.36*
[.19][.19]
SNF BiteSNF Bite -.10-.10
[.10][.10]-.09-.09
[.10][.10]
HH BiteHH Bite .03.03
[.04][.04].03.03
[.04][.04]
ΔΔ case mix case mix 486486
[557][557]413413
[556][556]
HMO IVHMO IV -2.45-2.45
[7.39][7.39]2.732.73
[6.31][6.31]
FP ShareFP Share 1.701.70
[5.03][5.03] 10.12*10.12*
[5.82][5.82]
Bite * HMO IVBite * HMO IV .008.008
[.006][.006] ----
Bite * FP shareBite * FP share -- -- -.015**-.015**
[.005][.005]
Key FindingsKey Findings
Overall cost-shifting: Overall cost-shifting: YesYes, 76%. , 76%. Ownership: Ownership: notnot by individual status by individual status
Market effect: Market effect: – ownership composition: ownership composition:
YesYes, , More FP enables more cost-shiftingMore FP enables more cost-shifting
– HMO penetration (IV): HMO penetration (IV): No effectNo effect
InterpretationsInterpretations
Large degree of cost-shifting comes Large degree of cost-shifting comes from higher prices. from higher prices. -> managed care may -> managed care may notnot be effective in be effective in
price bargaining in late 1990’s.price bargaining in late 1990’s. Price increases more when there’s Price increases more when there’s
more FP in the marketmore FP in the market-> there is NFP-FP difference-> there is NFP-FP difference-> cost-shifting depends on some joint -> cost-shifting depends on some joint
cost/quality function, which is determined cost/quality function, which is determined by market composition by market composition
Policy ImplicationsPolicy Implications
The majority of “savings” from The majority of “savings” from Medicare BBA cuts are financed Medicare BBA cuts are financed through a hidden “tax” on through a hidden “tax” on privately insured. privately insured.
Injecting “competition” (through Injecting “competition” (through managed care) may not prevent managed care) may not prevent hospital cost-shiftinghospital cost-shifting