A study of the utility of a computer-aided adverse reaction surveillance (CAARS)

Lee Bowman, Purdue University

Abstract

The occurrence of adverse drug reactions (ADRs) in the hospital has been cause for significant morbidity and mortality affecting one in five patients. Individual hospitals, the Joint Commission for Accreditation of Healthcare Organizations as well as governmental regulatory agencies have pressed health care providers to better identify and monitor these reactions. Although a number of ADR monitoring systems appear in the literature, none has documented sensitive, specific and time efficient methods for continuous monitoring of drug therapy. The method studied in this thesis (CAARS) utilizes a computerized medical record which includes patient specific information from the clinical laboratory, pharmacy and radiology. Decision rules based on a drug's known adverse effect profile were created that would identify potential ADRs. These rules would document a potential APR when a prescribed drug and a clinical event consistent with the decision rules occurred simultaneously. These potential ADRs were subsequently compared to a comprehensive monitoring scheme (which included chart review) and the utility of this computerized scheme was evaluated. This study found that the CAARS was comparably sensitive and specific to identify patients with ADRs. In addition these ADRs tended to be more severe than those not identified with CAARS. The overall usefulness of this system would be enhanced by monitoring, the total number of drug exposures, all patients with decreased renal function as well as maintaining voluntary spontaneous reporting by satellite pharmacists. The response by nursing may be of significant value to compliment CAARS but much greater education and information feedback mechanisms would likely need to be implemented.

Degree

Ph.D.

Advisors

Carlstedt, Purdue University.

Subject Area

Pharmaceuticals

COinS