Because of uniform availability, hospital administrative data are appealing for surveillance of adverse drug events (ADEs). Expert-generated surveillance rules that rely on the presence of International Classification of Diseases, 9th Revision Clinical Modification (ICD-9-CM) codes have limited accuracy. Rules based on nonlinear associations among all types of available administrative data may be more accurate.
By applying hierarchically optimal classification tree analysis (HOCTA) to administrative data, derive and validate surveillance rules for bleeding/anticoagulation problems and delirium/psychosis.
Retrospective cohort design.
A random sample of 3987 admissions drawn from all 41 Utah acute-care hospitals in 2001 and 2003.
Professional nurse reviewers identified ADEs using implicit chart review. Pharmacists assigned Medical Dictionary for Regulatory Activities codes to ADE descriptions for identification of clinical groups of events. Hospitals provided patient demographic, admission, and ICD9-CM data.
Incidence proportions were 0.8% for drug-induced bleeding/anticoagulation problems and 1.0% for drug-induced delirium/psychosis. The model for bleeding had very good discrimination and sensitivity at 0.87 and 86% and fair positive predictive value (PPV) at 12%. The model for delirium had excellent sensitivity at 94%, good discrimination at 0.83, but low PPV at 3%. Poisoning and adverse event codes designed for the targeted ADEs had low sensitivities and, when forced in, degraded model accuracy.
Hierarchically optimal classification tree analysis is a promising method for rapidly developing clinically meaningful surveillance rules for administrative data. The resultant model for drug-induced bleeding and anticoagulation problems may be useful for retrospective ADE screening and rate estimation.
From the *VA Salt Lake City Geriatrics, Research, Education, and Clinical Center (GRECC); †VA Salt Lake City Informatics Decision Enhancement And Surveillance (IDEAS) Center; ‡University of Utah, Department of Medicine, Salt Lake City, Utah; §Intermountain Healthcare Institute for Healthcare Delivery Research, Salt Lake City, Utah; ¶Northwestern University, Department of Emergency Medicine, Chicago, Illinois; ∥Jesse Brown VA Medical Center, Chicago, Illinois; **University of Utah, Department of Biomedical Informatics, Salt Lake City, Utah; ††VA Cooperative Studies Program, Clinical Research Pharmacy Coordinating Center, Albuquerque, New Mexico; ‡‡University of North Carolina, Chapel Hill, North Carolina; ¶¶University of Maryland, Department of Health Services Administration, College Park, Maryland; ∥∥Research Triangle International, Research Triangle Park, North Carolina; and ***Utah Department of Health, Salt Lake City, Utah.
Supported by VA HSR&D grants TRP-02-147-2, RCD-02-176-2; RTI contract R4605.001; AHRQ grant RO1 HS11885.
This study was conducted under institutional review board number IRB00011614 at the University of Utah and the Veterans Administration's George E. Whalen Medical Center.
Reprints: Jonathan R. Nebeker, MS, MD, Department of Medicine, University of Utah, GRECC 182, 500 Foothill Drive, Salt Lake City, UT 84148. E-mail: Jonathan.Nebeker@hsc.utah.edu.