Does outcome feedback make you a better emergency physician? A systematic review and research framework proposal

The organization of emergency medical care limits the ability of emergency physicians to know the outcomes of most of their patients after the patients leave the emergency department. This lack of outcome feedback may hinder the practice of emergency medicine (EM) by preventing "calibration" of the decision tools of practitioners. We sought to determine what is currently known about outcome feedback in EM, including its incidence, impact and modifiers.

We searched the following databases: PreMEDLINE, MEDLINE, EMBASE, CINAHL, CENTRAL, PsycINFO, DARE, Health Technology Assessment Database and AMED. We performed manual searches on abstract databases, reference lists, various health information and research websites, and nonindexed journals.

Selection entailed a 2-step screening pro cess to exclude articles not pertaining to outcome feedback in EM.

Our search yielded 1128 bibliographic records, from which screening identified 7 relevant reports: 5 surveys, 1 system level evaluation and 1 intervention trial.

All studies were found to have "inadequate" or "unable to assess" reporting and study quality. Systems for outcome feedback to EM residents have been increasingly available since 1984, though they are perceived to be inadequate. Commonly used mechanisms for outcome feedback include automatic routing of discharge summaries, case conferences for admitted patients and telephone calls to patients or families for discharged patients. With respect to attending emergency physicians, no conclusions or clinical recommendations can be made given the level of available evidence.

The potential importance of outcome feedback remains, at this time, underevaluated. We propose a research framework, and hypothesize that increasing outcome feedback would increase emergency physician diagnostic accuracy, therapeutic outcomes, clinical efficiency and job satisfaction. Future research in this area should include surveys and focus groups, as well as simulated or real-world intervention trials.