Session: Hospital Report Cards
Room: Hollister McManus Lounge
Time: Mon 08:30-10:00

Presenter: Kyoungrae Jung (Pennsylvania State University. Health Policy and Administration)

Discussant: Vivian WuUniversity of Southern California

Abstract

Research Objective: This paper examines the effects of consumers’ perceptions about different dimensions of hospital quality in the context of future hospital choice.

Background and significance: While public reporting programs are expanding, consumers’ responses to those programs have been small. To devise effective approaches to increase consumer information, it is essential to clarify the quality aspects that consumers value and use for health care choices. While important, few studies examined this issue.

We contribute to this discussion by examining the contributions of consumers’ beliefs about quality to hospital choice. Utilizing data about stated preferences, we obtain estimates of consumers’ perceptions about several unobservable attributes (reputation, amenities, etc). This approach enables us to identify different dimensions of quality and examine their relative contributions to hospital choice. Further, since it allows us to include time-invariant hospital characteristics, we use unpublicized hospital clinical quality scores to examine whether consumers correctly inferred hospital quality prior to public disclosure and used it for hospital choices. Our study also extends the hospital choice literature by introducing individual-level variables: the inertia effect and satisfaction from experience.

Data: We use data from a survey of employees at a large employer. The survey collected information about hypothetical future hospital choice and “importance weights” about unobserved attributes. The survey had information about hospitalized employees’ satisfaction ratings with the hospital stay.

Methods: We capture consumers’ prior beliefs about unobserved attributes, utilizing the “importance weights” data. Following prior work by Harris and Keane (1999) and Harris, Schultz & Feldman (2002), we inter-act the importance weight variables with hospital-specific intercepts in a conditional logit model and interpret the coefficients as consumers’ perceptions about unobservable attributes. We then use these beliefs estimates, indicator of prior use, satisfaction ratings, and hospital clinical quality scores, to estimate a second conditional logit model for future hospital choices.

Results: The analysis suggests that consumers already know about hospital clinical quality prior to public reporting and use it for hospital choice. However, its contribution to choice was small (marginal effect: 0.04 for hospitals with 20% of market shares), compared with those of consumers’ perceptions about reputation or medical services (marginal effects of 0.11 and 0.13). We find that satisfaction ratings matter for future hospital choice (marginal effect: 0.13), and prior use has large effects (marginal effect: 0.64), implying strong persistency in hospital choice.

Discussion: As the role of consumers in achieving better-functioning health care markets is becoming important, efforts are made to increase consumer information. The study suggests that consumers may turn to public information, if it includes relevant indicators such as satisfaction ratings. It also suggests that other strategies are needed to increase consumers’ responses to public reporting considering the large effects of prior beliefs and inertia. Efforts by employers to evaluate hospital quality and provide employees with the information, along with financial incentives to choose high-quality providers, may improve access to and use of the information. Continuing exploration of issues related to consumer information is necessary to find ways to increase the role of consumers in decision-making and improve performance of health care markets.

Key Terms

public quality information, consumers’ perceptions, hospital choice

Authors:

Kyoungrae Jung (Penn State University. Health Policy and Administration) , Roger Feldman (University of Minnesota. Health Policy and Management) and Dennis Scanlon (Penn State University. Health Policy and Administration)