Improving the success of mailed letter intervention programs to influence prescribing behaviors: a review.

摘要

Educational interventions have long been used as a means of influencing prescribing behavior. Various techniques including educational mailings, academic detailing, prescriber feedback with or without disclosing patient-identifying data, and supplemental patient information have been used to promote appropriate prescribing habits, reduce costs, and optimize patient care. While the effects of educational intervention programs are widely reported, little information is available regarding the effectiveness of various mailed intervention techniques.To review the effectiveness of mailed intervention programs and identify factors that may promote successful outcomes.A literature search was conducted via PubMed for reports of mailed intervention programs published through May 2012. Specific search terms included "drug utilization review," "drug utilization," "Medicaid," "prescribing feedback," "mailed physician intervention," and "mailed physician communications." Identified publications that met the following criteria were selected for inclusion: (a) evaluated printed educational materials disseminated via postal mail, (b) occurred in an outpatient setting, and (c) measured intervention impact on prescribing patterns, health care utilization, or economic outcomes. Publications that met all 3 criteria were abstracted for intervention strategy, follow-up period, data source, intervention target, prescriber acceptance of intervention, and effect on prescribing patterns, health care utilization, and economic outcomes.A total of 40 published reports regarding 39 unique interventions met inclusion criteria. The majority (34/39 [87.2%]) of studies were conducted in state or federally funded programs; only 5 programs involved private insurers. All programs used follow-up periods of ≤12 months after final intervention mailing. A total of 26 of the 39 unique interventions reported a positive impact on at least 1 target outcome. Programs that included a second recipient such as pharmacists (n = 4) reported a greater impact as compared with interventions mailed to prescribers alone. Programs that provided patient-identifying data had a higher success rate than those that supplied prescriber feedback and/or educational materials (21/25 [84.0%] vs. 5/14 [35.7%]); it should be noted that 2 of the 5 successful programs that provided nonpatient-identifying materials also used academic detailing. Programs that sent education material and/or prescriber feedback pertaining to multiple medication classes or disease states had minimal impact on prescribing patterns (n = 4). However, targeting 1 specific disease or medication supported by appropriate evidence resulted in favorable change in a short period of time. Additionally, providing recommendations that were supported by widely accepted clinical guidelines or literature were also associated with a high rate of success. A subset of programs that sought to evaluate health care utilization (n=5) and economic impact (n = 9) observed little change in measured outcomes. Evaluation of prescriber response forms conducted by 7 programs revealed that changes in therapy occurred in approximately 50% of patients with prescribers who intended to accept intervention recommendations.Though the degree of heterogeneity between articles prevents provision of definite results, it appears that a well-constructed mailed intervention program has the potential to evoke significant changes in prescribing patterns. Prescribers appear to be receptive to mailed interventions; however, there are limited data to determine the association between acceptance and actual prescribing change. Future research should focus on identifying barriers that may prohibit acceptance of recommendations from translating into changes in therapy. Additionally, future projects should include longer assessment periods to determine the duration of impact following final intervention mailing and potential effect on health care and economic outcomes.