RESUMEN
Objective: To determine whether “interactional group discussions” could reduce prescriptions of injections by physicians. Study design: Randomized controlled trial. Setting: Rural public health care facilities, North 24 Parganas district, West Bengal, India. Subjects: 72 medical officers, 36 each in intervention and control groups. Intervention: Interactional group discussions. Outcome measure: Proportion of prescriptions including at least one injection. Results: In the intervention group, 249 of 1,080 prescriptions (23%) included at least one injection compared with 79 of 1,080 prescriptions (7%) before and after the intervention, respectively. (RR: 0.32, 95% CI: 0.25-0.40). In the control group, 231 of 1,080 prescriptions (21%) included at least one injection before the intervention vs 178 of 1,080 prescriptions (16%) after the intervention (RR 0.77, 95% CI: 0.65-0.92). Conclusion: Interactional group discussions reduce prescription of injections.