Hospital performance standards and medical pricing: The impact of information disclosure in cardiac care

• DOI: http://doi.org/10.1111/jems.12381
• Author: Avi Dor,William Encinosa,Kathleen Carey
• Published date: 01 July 2020
• Date: 01 July 2020

J Econ Manage Strat. 2020;29:492–515.wileyonlinelibrary.com/journal/jems492

|

© 2020 Wiley Periodicals LLC

Received: 17 April 2018

|

Revised: 21 February 2020

|

Accepted: 6 May 2020

DOI: 10.1111/jems.12381

ORIGINAL ARTICLE

Hospital performance standards and medical pricing: The

impact of information disclosure in cardiac care

Avi Dor

1,2

|William Encinosa

3

|Kathleen Carey

4

1

Department of Health Policy and

Management, George Washington

University and NBER, Washington,

District of Columbia

2

NBER, Washington, District of Columbia

3

Center for Financing, Access, and Cost

Trends, Agency for Healthcare Research

and Quality and Georgetown University,

Rockville, Maryland

4

Department of Health Policy and

Management, Boston University, Boston,

Massachusetts

Correspondence

Avi Dor, Department of Health Policy and

Management,George Washington

University and NBER, Washington,

District of Columbia.

Email: avidor@gwu.edu

William Encinosa, Center for Financing,

Access, and Cost Trends, Agency for

Healthcare Research and Quality and

Georgetown University, Rockville,

Maryland.

Email: william.encinosa@ahrq.hhs.gov

Funding information

Agency for Healthcare Research and

Quality, Grant/Award Number:

R01 HS023610‐01

Abstract

A policy concern is that the initiation of Hospital Compare (HC) reporting in

Medicare provided leverage to insurers in price negotiations for lowering

private sector prices without regard to hospital performance. Using the se-

quential Nash bargaining framework we provide economic intuition to the

contrary: while average hospital prices decline under quality disclosures,

hospitals with above‐average quality are able to exert a stronger bargaining

position, consequently capturing prices above the market rate. To explore

this issue empirically we estimate variants of difference‐in‐difference

models, examining the effects of the three main scores (heart attack, heart

failure, and combined mortalities) on transaction prices of related hospital

procedures. States which had similar mandated reporting systems in place

before the initiation of HC form the control group. Analyzing claims data

of privately insured patients, we find that HC exerted downward pressure

on prices. However, hospitals rated “above‐average”captured higher

prices, thereby offsetting the overall policy effect fully or partially. Leads

andlagsanalysislendsfurther support for our difference‐in‐difference

approach. We find that highly ranked hospitals received a quality

premium of 8–14%, comparable to price effects found in other health care

markets. We conclude that HC was effective at constraining prices without

penalizing high performers.

KEYWORDS

difference‐in‐difference models, information rents, medical care, Nash bargaining, pricing and

product quality

1|INTRODUCTION

1.1 |Motivation

Consumer information in the form of hospital rating systems known as “report cards”have been in existence in various

forms as early as the 1990s. However, such report cards were made available in a small number of states that generated

them independently from each other, while employing a mix of measures and ranking methodologies. In 2005, the

Centers for Medicare and Medicaid Services (CMS) launched a uniform on‐line national rating system, known as

Hospital Compare (HC), aimed at informing consumers and promoting competition among hospitals. Initially, HC

consisted of a set of process measures of hospital performance based on general practices.

1

In 2008, these were

augmented with outcome‐based measures, namely mortality‐based hospital rankings which were deemed to be more

easily understood by patients and also more effective at motivating hospitals to engage in quality improvement practices

(Harris, 2007). Initially these outcome measures focused primarily on cardiac care. Outcomes for other medical con-

ditions were gradually phased in after 2012.

In practice, opinion surveys have shown that consumers were generally unaware or uninterested in these

rankings, even as they had become more accessible, and there is little evidence in the empirical literature to suggest

that hospital report cards had an impact on consumer choices of hospitals. Similarly, there is little evidence that

HC and other report cards have had a significant impact on patient outcomes. We briefly review the literature

below.

A less explored channel through which public reporting might impact health care markets, and ultimately

patients, is hospital pricing. In the private segment of the health care market, private insurers rather than in-

dividual patients incur the bulk of health care costs. As group purchasers, large insurance firms and managed care

organizations who engage in price negotiations with hospitals are more likely to incorporate information from

public reporting in their decision‐making (Dor, Encinosa, & Carey, 2015;Reinhardt,2006,2009). However, there is

a paucity of evidence on the impact of report cards on hospital pricing. Yet, in a well‐functioning market, a

hospital's ability to deliver better health outcomes could be expected to lead to greater demand and a more

advantageous bargaining position for the hospital, and hence be rewarded by higher prices. In this paper, we

provide economic intuition for this expectation using a simple hospital‐insurer bargaining model (Garmon, 2017;

Gowrisankaran, Nevo, & Town, 2015;Lewis&Pflum,2015), augmenting the sequential Nash Bargaining model to

account for consumer valuation of quality when there are third party payers (Dafny, Ho, & Lee, 2019). To address

the empirical gap we examine the impact of implementing quality reporting on inpatient prices, allowing for

hospital differentiation by relativerankings.Specifically,weestimatetheimpactoftheHCmeasures,asim-

plemented in 2008, on actual prices negotiated and transacted between private insurers and U.S. hospitals. These

HC measures are categorical rankings based on hospital risk‐adjusted mortality rates for the three medical con-

ditions made available on‐line as of 2008, which included heart attack, heart failure, and pneumonia. We focus on

prices of major cardiac procedures related to heart attacks and heart failure. Results suggest that HC exerts

downward pressure on prices, but that this effect is offset for hospitals ranked in the highest quality category.

These results differ from an older strand of literature in industrial organization theory. Matthews and Postlewaite

(1985)andSchlee(1996) show that quality information can harm consumers because the seller raises the price too

much after information disclosure. Our theory differs in that we model Nash bargaining of prices, rather than

monopoly or duopoly prices (see Ma & Mak, 2014, for a Bertrand competition case).

A main contribution of our paper is to bridge the gap between two important literatures to shed new light on the

impact of quality reports on prices. First, we use the contemporary literature on Nash bargaining, which has emerged as

the primary vehicle for modeling hospital‐insurer prices. While this has led to many new insights into the impact of

hospital mergers on prices (Garmon, 2017; Gowrisankaran et al., 2015; Ho & Lee, 2017; Lewis & Pflum, 2015). it has

been silent on the interaction between quality and price. Second, we incorporate the literature on the role of quality

disclosures which focused heavily on the impact of quality reports on consumers, while remaining relatively silent on

prices (for a review, see Dranove and Jin, 2010). For example, Jin (2005) and Jin, Zhe, and Sorensen (2006) find that

some health maintenance organizations (HMOs) voluntarily reported National Committee for Quality Assurance

(NCQA) quality measures to differentiate their product in the market, and that this was of high value to consumers.

2

Additionally, Jin, Zhe, and Leslie (2003) find that the public reporting of restaurant hygiene grade cards improve health

outcomes, by reducing hospitalizations from food‐borne diseases by 20%. We connect these two strands of literature by

introducing quality disclosure into the standard Nash bargaining model to explore the impact of quality disclosure on

price negotiations.

1.2 |Impacts of HC public reporting of hospital performance

Information about hospital quality performance began appearing in the public domain in the 1990s, albeit limited to

several states and using disparate methodologies. An early review of the gains from public disclosure of performance

data found that consumers showed little interest in or use of the available information due to various behavioral factors

(Marshall, Shekelle, Leatherman, & Brook, 2000). Anecdotal surveys of hospital administrators and executives have

DOR ET AL.

|

493