Abstract 23: Impact of Massachusetts Health Reform on Hospitalizations, Length of Stay and Costs of Inpatient Care: Does Safety-Net Status Matter?

Abstract

Objective: There is widespread concern that large-scale insurance expansion - such as that anticipated from the Affordable Care Act - has the potential to cause sharp increases in health care utilization and costs. In the setting of Massachusetts’ landmark 2006 health care reform, we estimated pre-reform to post-reform changes in inpatient care volumes and costs, contrasting the experience of safety-net hospitals (SNH) as the predominant providers of care for targeted reform beneficiaries, with that of non-SNH. Study Design: We analyzed MA inpatient discharge on all discharges from 2004-2010 for 2,636,326 non-elderly patients (age 18-64) across all 66 short-term acute care hospitals. Safety-net hospitals were identified as those in the top quartile of hospitals in the proportion of hospital admissions with Medicaid, Free Care and self-pay as the primary payer. Using linear regression models we estimated hospital-level post-reform changes in (a) # admissions, (b) length of stay (LOS; days), (c) charge per day ($) and (d) charge per stay ($), separately for SNH and non-SNH, adjusting for patient demographics and comorbidities. We also examined changes for subpopulations by race/ethnicity and socioeconomic status (SES; defined using patient zip code median income). To isolate the impact of reform from secular trends, we treated the elderly as the “control” population. Findings: There was no significant post-reform change in the number of admissions; average quarterly number of admissions per hospital were 1,480 pre-reform and 1,520 post-reform (p=0.68). A similar pattern was found for admissions by SNH status, and for minority and low-income subpopulations. Average LOS increased by a smaller amount among SNH (0.20 days; 95% CI=[0.15, 0.25]) than among non-SNH (0.30 days; 95% CI=[0.27, 0.33]). Average charges per day decreased among SNH ($-198; 95%=[$-251, $-145]) and increased among non-SNH ($249; 95%=[$215, $284]). A similar trend with a larger difference was found for average charges per stay (SNH=$-477; 95%CI=[$-768, $-187] and non-SNH=$1,442; 95%CI=[$1,248, $1,635]). Similar trends were found for both acute and non-acute admissions. Among blacks, Hispanics and low-income patients, none of the measures indicated larger increase in SNH compared to that in non-SNH. Conclusion: Following MA health reform, utilization of inpatient care did not increase at SNH, the predominant providers of inpatient care for populations targeted by the reform, compared to non-SNH. A similar trend was found for acute and non-acute admissions, and for minority and low-income subpopulations. Future analyses in the coming months will test robustness of these findings using the non-elderly patients from comparison states as the control population. Implications: Insurance coverage expansion in MA may not have increased overall use of inpatient use and costs.