@article{14788, author = {Olga A Vsevolozhskaya and Karina C Manz and Pierre M Zephyr and Teresa M. Waters}, title = {Measurement matters: changing penalty calculations under the hospital acquired condition reduction program (HACRP) cost hospitals millions}, abstract = {Abstract Background Since October 2014, the Centers for Medicare and Medicaid Services has penalized 25% of U.S. hospitals with the highest rates of hospital-acquired conditions under the Hospital Acquired Conditions Reduction Program (HACRP). While early evaluations of the HACRP program reported cumulative reductions in hospital-acquired conditions, more recent studies have not found a clear association between receipt of the HACRP penalty and hospital quality of care. We posit that some of this disconnect may be driven by frequent scoring updates. The sensitivity of the HACRP penalties to updates in the program’s scoring methodology has not been independently evaluated. Methods We used hospital discharge records from 14 states to evaluate the association between changes in HACRP scoring methodology and corresponding shifts in penalty status. To isolate the impact of changes in scoring methods over time, we used FY2018 hospital performance data to calculate total HAC scores using FY2015 through FY2018 CMS scoring methodologies. Results Comparing hospital penalty status based on various HACRP scoring methodologies over time, we found a significant overlap between penalized hospitals when using FY 2015 and 2016 scoring methodologies (95%) and between FY 2017 and 2018 methodologies (46%), but substantial differences across early vs later years. Only 15% of hospitals were eligible for penalties across all four years. We also found significant changes in a hospital’s (relative) ranking across the various years, indicating that shifts in penalty status were not driven by small changes in HAC scores clustered around the penalty threshold. Conclusions HACRP penalties have been highly sensitive to program updates, which are generally announced after performance periods are concluded. This disconnect between performance and penalties calls into question the ability of the HACRP to improve patient safety as intended. }, year = {2021}, journal = {BMC Health Serv Res}, volume = {21}, pages = {131}, month = {02/2021}, issn = {1472-6963}, doi = {10.1186/s12913-021-06108-w}, }