Feasibility and validity of telephone triage for adverse events during a voluntary medical male circumcision campaign in Swaziland

Background: Voluntary medical male circumcision (VMMC) reduces HIV acquisition among heterosexual men by approximately 60%. VMMC is a surgical procedure and some adverse events (AEs) are expected. Swaziland's Ministry of Health established a toll-free hotline to provide general information about VMMC and to manage post-operative clinical AEs through telephone triage. Methods: We retrospectively analyzed a dataset of telephone calls logged by the VMMC hotline during a VMMC campaign. The objectives were to determine reasons clients called the VMMC hotline and to ascertain the accuracy of telephone-based triage for VMMC AEs. We then analyzed VMMC service delivery data that included date of surgery, AE type and severity, as diagnosed by a VMMC clinician as part of routine post-operative follow-up. Both datasets were de-identified and did not contain any personal identifiers. Proportions of AEs were calculated from the call data and from VMMC service delivery data recorded by health facilities. Sensitivity analyses were performed to assess the accuracy of phone-based triage compared to clinically confirmed AEs. Results: A total of 17,059 calls were registered by the triage nurses from April to December 2011. Calls requesting VMMC education and counseling totaled 12,492 (73.2%) and were most common. Triage nurses diagnosed 384 clients with 420 (2.5%) AEs. According to the predefined clinical algorithms, all moderate and severe AEs (153) diagnosed through telephone-triage were referred for clinical management at a health facility. Clinicians at the VMMC sites diagnosed 341 (4.1%) total clients as having a mild (46.0%), moderate (47.8%), or severe (6.2%) AE. Eighty-nine (26%) of the 341 clients who were diagnosed with AEs by clinicians at a VMMC site had initially called the VMMC hotline. The telephone-based triage system had a sensitivity of 69%, a positive predictive value of 83%, and a negative predictive value of 48% for screening moderate or severe AEs of all the AEs. Conclusions: The use of a telephone-based triage system may be an appropriate first step to identify life-threatening and urgent complications following VMMC surgery.