October 21, 2009
By Lisa Cowan Ph.D. and Asher Kimchi M.D.
Los Angeles, CA – In the midst of a tough economy and ongoing discussions over health care reform, a study published in the early online edition of the journal Circulation: Cardiovascular Quality and Outcomes brings to light important associations between mortality and resource use in hospitals when treating patients with heart failure. “Looking Forward, Looking Back: Assessing Variations in Hospital Resource Use and Outcomes for Elderly Patients With Heart Failure" concludes that California teaching hospitals with higher resource use in caring for hospitalized heart failure patients had lower mortality rates.
Previous studies aimed at assessing how resources are used for expired elderly Medicare beneficiaries with heart failure, as a consequence of study design, have not been able to draw conclusions on health outcomes. The inhomogeneity of disease severity is minimized by analyzing patients who have died; however, in only studying deceased patients, no conclusions can be drawn between mortality and resource use. These prior studies, through their design, suggested that hospitals with the lowest resource use establish benchmarks, and higher resource utilization is identified as inefficient. From these studies it seems that reducing hospital resource use to align with performance benchmarks would result in substantial savings in health care spending without adversely affecting patient outcomes.
Dr. Michael Ong et al’s study, "Looking Forward, Looking Back” departs from assessing expenditures on deceased patients exclusively in order to circumvent limitations of previous studies, namely: inability to identify differences in health outcomes across different hospitals and assuming patterns of resource use among all patients are well represented by resource use in expired patients. Two cohorts of elderly Medicare beneficiaries hospitalized for heart failure at one of 6 nonprofit academic hospitals in California (UC Davis, UC Irvine, UC Los Angeles, UC San Diego, UC San Francisco, or Cedars-Sinai Medical Center) between 2001 and 2005 were formed through analysis of administrative data. The “Looking Forward” cohort (n = 3999) was comprised of all patients hospitalized during the study period regardless of patient outcome (excluding cardiac, renal, or hepatic transplant patients, or those admitted for transplant evaluation, patients transferred from another facility, or patients assigned to surgical Diagnostic Related Group classification - with the exception of valve replacements, and pacemaker/defibrillator placements which were included in the study). The “Looking Back” cohort (n = 1639) was a subset of the first group and included patients who expired during the study period.
Two resource use outcomes were generated from administrative data: total hospital days and indexed total direct costs. Mortality data within 180-days after initial admission were obtained from administrative data and the National Death Index. The “Looking Forward” multivariate risk-adjusted hospital means resulted in a range of 17 – 26% for mortality, 7.8 – 14.9 days for total hospital days, and 0.66 – 1.30 times the mean value for indexed total direct costs. The Spearman rank correlation coefficients were determined to be -0.68 between mortality and hospital days and -0.93 between indexed total direct costs and mortality. The “Looking Back” cohort resulted in risk-adjusted hospital means that ranged from 9.1 – 21.7 days for total hospital days and 0.91 – 1.79 times the average value for indexed total direct costs.
This study analyzed expenditures on Medicare beneficiaries hospitalized for heart failure at one of 6 nonprofit academic hospitals in California and was able to make correlations between mortality and resource use. In light of pressures to reduce spending, Ong et al’s study on heart failure patients clearly indicates that teaching hospitals that used more resources had lower mortality rates. Savings from reducing resource use may adversely affect health outcomes and conclusions based solely on deceased patients overlook valuable correlations between mortality and resource use.
Authors of the original article: Michael K. Ong, Carol M. Mangione, Patrick S. Romano, Qiong Zhou, Andrew D. Auerbach, Alein Chun, Bruce Davidson, Theodore G. Ganiats, Sheldon Greenfield, Michael A. Gropper, Shaista Malik, J. Thomas Rosenthal, and Jose J. Escarce