Does variation in the quality of care in minority serving hospitals explain cancer disparities? Colorectal cancer is the third most common cancer in the US and the second most common cause of cancer death. Racial and ethnic disparities in cancer outcomes are well documented in the literature, but there is little about the mechanisms that support and maintain disparities. More recent explanations suggest that the quality of care in hospitals where minorities cluster for treatment may play a role.
Aims and Hypotheses: The overall objective of the current study is to assess clinical performance in the institutions where minorities cluster for colorectal cancer (CRC) care and elucidate how the quality of care in these settings correlate with outcome disparities. The central hypothesis is that minorities cluster for CRC care in hospitals with low compliance with evidence based care and higher complication rates;and that use of these hospitals, as well as the quality of care delivered therein, will explin some of the observed disparities in CRC. Methods: Retrospective data analysis of a large, all-state, all-age, all-payer administrative data set comprised a linkage between the California Cancer Registry, the California Patient Discharge and Hospital Annual Financial Data. Minority-serving hospitals (MSH) will be defined by density of black and Hispanic patients. These hospitals will be further characterized as to Medicaid utilization rates, annual surgical volume, National Cancer Institute (NCI) cancer center designation;and performance on evidence based measures of care (12 lymph node examination, chemotherapy for stage III disease and preoperative staging for rectal cancer) and Patient safety indicator (PSI) rates (postoperative sepsis and respiratory failure). The AHRQ software will be used to calculate PSI rates. Analysis: A combination of Pearson chi-squared analysis will be used to define and characterize MSH and compare performance of MSH against other hospitals on quality measures and PSI's. Cox proportional hazard and multi- level (hierarchical) modeling will be used to assess the correlation of MSH use and performance on 5-year mortality. Multi-level modeling will be used to distinguish the effects of the hospital from the effects of patient characteristics on mortality
Does variation in the quality of care in minority serving hospitals explain cancer disparities? The current proposal investigating variation in the quality o care in minority serving hospitals has public health relevance because these hospitals not only serve minorities, but also serve the poor and uninsured. As the rate of un- and underinsurance increases across the country, and as the implementation of health care reform begins, the results of our study may inform policy to improve cancer care and reduce cancer disparities.
