How is medication prescribing ceased? A systematic review.

BACKGROUND: Medication prescribing is a complex process where the focus tends to be on starting new medication, changing a drug regimen, and continuing a drug regimen. On occasion, a prudent approach to prescribing may necessitate ending an ongoing course of medication, either because it should not have been started in the first place; because its continued use would cause harm; or because the medication is no longer effective. OBJECTIVE: To identify effective strategies for stopping pre-existing prescribing in situations where continued prescribing may no longer be clinically warranted. RESEARCH DESIGN: Systematic searches for English-language reports of experimental and quasi-experimental research were conducted in PubMed (1951-November 2009), EMBASE (1966-September 2008), and International Pharmaceutical Abstract b (1970-September 2008). A manual search for relevant review articles and a keyword search of a local database produced by a previous systematic search for prescribing influence and intervention research were also conducted. STUDY SELECTION AND DATA EXTRACTION: Following initial title screening for relevance 2 reviewers, using formal assessment and data extraction tools, independently assessed abstracts for relevance and full studies for quality before extracting data from studies selected for inclusion. RESULTS: Of 1306 articles reviewed, 12 were assessed to be of relevant, high-quality research. A variety of drugs were examined in the included studies with benzodiazepines the most common. Studies included in the review tested 9 different types of interventions. Effective interventions included patient-mediated interventions, manual reminders to prescribers, educational materials given to patients, a face-to-face intervention with prescribers, and a case of regulatory intervention. Partially effective interventions included audit and feedback, electronic reminders, educational materials alone sent to prescribers, and distance communication combined with educational materials sent to prescribers. CONCLUSIONS: It appears possible to stop the prescribing of a variety of medications with a range of interventions. A common theme in effective interventions is the involvement of patients in the stopping process. However, prescribing at the level of individual patients was rarely reported, with data often aggregated to number of doses or number of drugs per unit population, attributing any reduction to cessation. Such studies are not measuring the actual required outcome (stopping prescribing), and this may reflect the broader ambiguity about when or why it might be important to end a prescription. Much more research is required into the process of stopping pre-existing prescribing, paying particular attention to improving the outcomes that are measured.