Classification and rates of adverse events in a Malawi male circumcision program: impact of quality improvement training

BACKGROUND: Assessing safety outcomes is critical to inform optimal scale-up of voluntary medical male circumcision (VMMC) programs. Clinical trials demonstrated adverse event (AE) rates from 1.5 to 8 %, but we have limited data on AEs from VMMC programs. METHODS: A group problem-solving, quality improvement (QI) project involving retrospective chart audits, case-conference AE classification, and provider training was conducted at a VMMC clinic in Malawi. For each identified potential AE, the timing, assessment, treatment, and resolution was recorded, then a clinical team classified each event for type and severity. During group discussions, VMMC providers were queried regarding lessons learned and challenges in providing care. After baseline evaluation, clinicians and managers initiated a QI plan to improve AE assessment and management. A repeat audit 6 months later used similar methods to assess the proportions and severity of AEs after the QI intervention. RESULTS: Baseline audits of 3000 charts identified 418 possible AEs (13.9 %), including 152 (5.1 %) excluded after determination of provider misclassification. Of the 266 remaining AEs, the team concluded that 257 were procedure-related (8.6 AEs per 100 VMMC procedures), including 6 (0.2 %) classified as mild, 218 (7.3 %) moderate, and 33 (1.1 %) severe. Structural factors found to contribute to AE rates and misclassification included: provider management of post-operative inflammation was consistent with national guidelines for urethral discharge; available antibiotics were from the STI formulary; providers felt well-trained in surgical skills but insecure in post-operative assessment and care. After implementation of the QI plan, a repeat process evaluating 2540 cases identified 115 procedure-related AEs (4.5 AEs per 100 VMMC procedures), including 67 (2.6 %) classified as mild, 28 (1.1 %) moderate, and 20 (0.8 %) severe. Reports of AEs decreased by 48 % (from 8.6 to 4.5 per 100 VMMC procedures, p < 0.001). Reports of moderate-plus-severe (program-reportable) AEs decreased by 75 % (from 8.4 to 1.9 per 100 VMMC procedures, p < 0.001). CONCLUSIONS: AE rates from our VMMC program implementation site were within the range of clinical trial experiences. A group problem-solving QI intervention improved post-operative assessment, clinical management, and AE reporting. Our QI process significantly improved clinical outcomes and led to more accurate reporting of overall and program-reportable AEs.