The Effect of a Tiered Hospital Benefit and Safety Incentive on Hospital Admissions: Evidence from Stated and Revealed Preference Data
Posted: 15 Jun 2007
Date Written: March 13, 2007
Objective: This paper examines the effect of a tiered hospital benefit and safety incentive in a commercially insured population, and examines whether estimates from the hospital choice model are improved when stated preference data from a consumer survey is added to revealed preference data from claims.
Study Setting: The Hospital Safety Incentive (HSI) applies to union employees enrolled in the Boeing Company's Traditional Medical Plan effective July 2004. The HSI gives patients a financial incentive to choose hospitals that meet the Leapfrog Group's three patient safety leaps. While the TMP's standard coverage for hospital care became 95% of allowed hospital charges after July 1, 2004 (from 100% previously), union beneficiaries enrolled in the TMP can achieve a benefit of 100% for hospital care if admitted to a hospital that meets the Leapfrog standards.
Data Sources: Claims and enrollment data identify the subset of employees with a hospitalization and are used to define the hospital choice set and the preferred hospital. Claims data formed the sampling frame for a twenty minute telephone survey of union and non-union employees (n = 1,200 in both the pre and post periods with a 70% participation rate). The survey collected information about preferences for various hospital attributes and the degree to which hospitalized patients were involved in the choice of hospital or deferred entirely to their physician's preference. Data on hospital service offerings, financial status and quality of clinical care were obtained from the AHA annual survey and from the JCAHO "Quality Check" data. Each patient's choice set included all hospitals located in the Seattle metropolitan area offering treatment for the diagnosis or procedure for which the patient was hospitalized. GIS software was used to compute the distance between the patient's residence and each hospital in the choice set.
Analytic Methods: We estimate a random utility model to express the probability that individual i chooses hospital j from a set of k hospital alternatives. An individual's utility for a given hospital is specified as a function of patient characteristics Xi and hospital level characteristics Hj. The vector of patient characteristics include both demographic variables as well as the employee's response to specific survey questions about the importance of various hospital attributes such as distance, quality and reputation. The hospital level characteristics include measures from both the AHA and the Quality Check data. We also include interaction terms between certain X and H variables that match, for example between the patients survey response about the importance of distance in selecting a hospital and the actual distance variable computed using the GIS software.
Principal Findings: The HSI did not influence union employees' choice of hospitals or change their preferences for factors important in selecting a hospital, including the relative importance of quality and safety information or out-of-pocket costs. Additional results are forthcoming, including an estimate of the dollar value of the HSI needed to move a significant share of the commercial market from non-compliant to Leapfrog compliant facilities.
Keywords: Tiered hospital benefits, Patient safety, Hospital choice
JEL Classification: I11
Suggested Citation: Suggested Citation