Background: Male circumcision (MC) is a well-established component of HIV prevention in countries with high HIV prevalence and heterosexually driven epidemics. Delivery and monitoring of MC programs are reliant on good quality MC data. Such data are often generated through self-reported MC status surveys. This study examined self-reported MC status in comparison with genital photographs from men in Papua New Guinea (PNG).
Methods: This retrospective non-interventional study collated self-reported MC status data from the 'acceptability and feasibility of MC' study at 4 sites in PNG during 2010-2011. Participants reported their MC status based on an 8-category photographic classification covering the range of foreskin cutting practices in PNG. Genital photographs of 222 participants from this study were independently classified by 2 investigators. The 8-category photographic classification was simplified into a 3 category classification of 'no cut', 'straight cut' and 'round cut' before comparing for agreement between self-reporting and investigator assessment using Cohen's Kappa measure.
Results: Using the 3-category classification, there was 90.6% (201/222) agreement between self-assessment and investigator classification (κ value 0.805). Of the discordant 9.4% (21/222), 3.6% (8/222) self-classified as having a cut foreskin (5 straight cut; 3 round cut) while investigators classified as having no cut; 4.1% (9/222) self-classified as having no cut while investigators classified them as having had a cut (6 straight cut; 3 round cut) and 1.8% (4/222) self-classified as having a round cut while investigators classified as having a straight cut. Given the great variety of foreskin cutting practices and appearances, feasible explanations are suggested for two-thirds (13/21) of these discordant results.
Conclusions: This study demonstrates a high level of agreement between self-reporting and investigator assessment of MC status in PNG and suggests self-reporting of MC status to be highly reliable among men in PNG.