This revised application proposes a competing continuation of HS 06512, """"""""Screening Quality of Care Using Administrative Data- (9/30/90-9/29/92). Using readily available discharge abstract data, this research derived the Complications Screening Program (CSP), to screen hospitals for complications potentially resulting from substandard care, such as post- operative pneumonia, post-procedural hemorrhage, wound infection, and medication incidents.
While aimi ng to have good face validity as quality indicators, the CSP is nonetheless limited by important reservations about administrative data. The initial proposal acknowledged these drawbacks, indicating that it would be important to validate the CSP before using it to screen hospitals for in-depth quality reviews. This application proposes that validation. This validation examines four questions addressing different dimensions of validity. (1) Diagnosis and procedure codes from computerized hospital discharge abstract data are used by the CSP to indicate potential complications of care. Do hospitals code these diagnoses and procedures completely and accurately? (2) The CSP makes assumptions from discharge abstract data - primarily about timing of clinical events - to produce its judgments about potential quality problems. Are these assumptions correct? (3) The CSP """"""""flags"""""""" individual cases as potentially problematic. Are individual cases flagged by the CSP more likely to have quality problems than other cases? (4) The ultimate goal of the CSP is to examine rates of potential quality problems at the hospital level, not to focus on individual cases. Do hospitals with higher-than-expected rates of flagged cases have systematic problems with quality of care? This study involves detailed medical record reviews and a pilot test of hospital site visits to examine these aspects of validity using methods developed by the investigators. Peer Review Organizations (PROs) in California and Connecticut will examine a total of 2,490 medical records, focusing on coding completeness and accuracy, validity of CSP assumptions, and key processes of care relating to each type of complication identified by the CSP. Site visits to six hospitals in Connecticut will collect data on clinical care unit organizational effectiveness and on system-related factors affecting key process of care steps. This revised proposal is submitted in response to a Program Announcement (PA-93-084), """"""""Health Care Quality Improvement and Quality Assurance Research,' from the Agency for Health Care Policy and Research (AHCPR). Among other topics, PA-93-084 calls for timely studies of cost effective quality of care measures and their validity.
