The effect of registry-based performance feedback via short text messages and traditional postal letters on prescribing parenteral steroids by general practitioners—A randomized controlled trial

• Sending feedback about recent steroid prescription to general practitioners reduced parenteral steroids prescription rate by about one-third.
• Feedback in the form of short text messages was demonstrated to be at least as effective as postal letters, and much cheaper.
• Reduction in this outcome was unrelated to gender and age but it was associated with the number of feedback messages sent.

BackgroundIt is conjectured that providing feedback on physicians’ prescribing behavior improves quality of drug prescriptions. However, the effectiveness of feedback provision and mode of feedback delivery is not well understood. The objective of this study was to assess and compare the effect of traditional paper letters (TPL) and short text message (STM) feedback on general practitioners’ prescribing behavior of parenteral steroids (PSs).MethodsIn a single-blind randomized controlled trial, 906 general practitioners (GPs) having at least 10 monthly prescriptions were randomly recruited into two interventions and one control study arms with 1:1 allocation, stratified by percentage of prescriptions. The intervention was the provision of 3 feedback messages containing prescribing indices in TPL and STM (in the first two arms) versus the control arm (CG) with an interval of 3 months between these messages. We calculated the PS Defined Daily Dose (DDD) for every GP, every month, and compared between the 3 arms, before and after the interventions. The expected primary outcome was to reduce prescription of parenteral steroids by participants. The study was performed in the Kerman Social Security Organization in Iran.ResultsA total of 906 GPs were selected for the trial, but only 721 of them (TPL = 191, STM = 228, CG = 302) were recruited for the 1st feedback. The mean age of GPs was 44 and 59% of them were male. The prescribed parenteral steroid DDDs at baseline were similar (TPL = 121.62, STM = 127.49, CG = 115.68, P > 0.5). At the end of the study, DDDs in the TPL and STM arms were similar (TPL = 104.38, STM = 101.90, P > 0.9) but DDDs in each intervention arm was statistically significantly lower than in CG (CG = 156.17, P < 0.0001). Being in TPL and STM arms resulted in 36.1 and 41.7 units of decrease in DDD respectively, compared to the control arm (P < 0.02 and P < 0.005) after the one-year duration of the study.ConclusionFeedback by TPLs and STMs on prescribing performance effectively reduced prescribing PSs by GPs. STM, being a cheap and fast tool, is potentially powerful and efficient for drug prescription rationalization.