2008 HSR&D National Meeting Abstract
3056 — Predictors of physician's intention to investigate Adverse Drug Events
Weir CR (SLC TREP), Nebeker J
Physicians do not recognize between 75 to 25% of moderate to severe ADEs. Lack of detection and reporting is related to physician’s attitudes, beliefs and knowledge. This study reports on the development and validation of a survey instrument designed to assess the beliefs and attitudes associated with physician’s intention to investigate ADEs.
The Theory of Planned Behavior guided instrument development, which consisted of two phases. First, physicians from the specialties of medicine, surgery and psychiatry were interviewed regarding the process of ADE investigation. Responses were transcribed and qualitative methodology used to identify key themes: 1) uncertainty about cause, 2) effort; 3) belief that ADEs rarely causes harm, 4) self-efficacy, 5) time pressures, and 6) information availability. Item creation and instrument validation constituted the second phase with 120 items reduced to 75 after piloting. Ninety-two of 300 residents completed the questionnaire as part of a larger study.
Scales were constructed for intention, attitudes, normative beliefs, self-efficacy, outcome beliefs and controllability (alpha > =0.60 for all scales). Each correlated significantly with intentions to investigate possible ADEs and were regressed on intentions (entered simultaneously) with the overall model significant at R2 = 0.42, (F7,90 = 9.48; p=0.00). Controllability were highly significant (t=4.98;p=.000) as were normative beliefs (2.77; p=.007). Self-efficacy and overall attitudes were not significant after controlling for other variables. Specific controllability factors were: patient symptoms (r=.35), working in a team (r=.33), reviewing medications on rounds (r=.29), pharmacist on the team (r=.51) and writing out medications on a card (r=.70). Using an algorithm, information from nurses, being too busy and attending supervision were not associated with intentions. Specific normative beliefs significantly correlating with intention to investigate ADEs were the expectations of: attendings (r=.51), housestaff leadership (r=.40), and the behavior of other housestaff (r=.34). Perceived expectations of hospital administration were not correlated with intentions.
Investigating ADEs may be related mainly to the use of personal strategies and normative support.
Interventions to increase physician’s likelihood of regularly investigating ADEs should focus on making normative expectations salient and increasing support for specific clinical strategies.