Feasibility study of surveying the adverse drug reaction surveillance systems in a large community of hospitals.

Abstract

To determine the feasibility of accurately assessing the types of hospital adverse drug reaction (ADR) surveillance systems. Cross-sectional survey by mailed, self-administered questionnaire followed by selected verification interviews. Harris County, Texas. All hospitals in the county with different pharmacy directors. Self description of surveillance system and number of ADRs reported. Forty-nine of 61 hospitals (80 percent) responded to a questionnaire. Forty-seven (96 percent) of the responding hospitals collected information on ADRs with 11 (22 percent) describing their surveillance system as active. Those individuals most often cited as responsible for ADR surveillance included pharmacists, quality assurance personnel, and nurses. Data were verified by personal interviews for 10 hospitals. The number of ADRs reported during the interviews was significantly lower than that reported in the questionnaires. Overall, the reporting of fatal and severe ADRs were more reliable than the reporting of moderate ADRs. These differences were the result of inadequate documentation and the lack of a uniform definition of ADRs. These data suggest that a large-scale ongoing survey of surveillance systems and reported adverse event rates has limitations and the reliability of data derived from a questionnaire should be verified. To improve the accuracy of surveys used to monitor hospital ADR surveillance systems, it is essential to develop reliable definitions for classifying ADRs and surveillance methods, as well as accurate measures of ADR documentation procedures.