Reporting of Intimate Partner Violence among Male Couples: Cross-Sectional and Serial Dyadic Concordance

Purpose Intimate partner violence (IPV) in male couples is a public health concern, but the reliability of self-reported IPV data from gay, bisexual, and other men who have sex with men (GBMSM) is understudied. Research in heterosexual couples finds IPV-underreporting can be differential between victims/perpetrators and by gender; it is unknown if GBMSM-data has similar limitations. Methods This study assessed interpartner agreement (percent agreement; kappa statistics) between self- and partner-reported IPV, and compared agreement between 2 recall periods (1-year; 3-month), 3 granularities (any IPV; IPV domain; IPV behaviors), and across 3 surveys (BL; Wave 1 (W1); Wave 2 (W2)) in a sample of 404 male couples in the U.S. (2016-2017). Longitudinal dyadic concordance-trends were assessed to determine if couples had consistently reliable data over time. Results Past-year IPV-victimization prevalence was 66.41% and perpetration, 64.42%; past 3-month IPV-victimization prevalence was 47.30% (W1) and 52.57% (W2), and perpetration, 46.30% (W1) and 46.30% (W2). Interpartner agreement was consistently low across recall periods and granularities. Observed agreement was higher for those who did not report IPV, compared to those who reported experiencing IPV. At an individual-level, interpartner agreement on any given survey was not significantly predictive of subsequent data reliability. Conclusion Researchers should be cognizant of the potential for unpredictable and unreliable IPV reports from GBMSM. Further research on IPV data quality in male couples is needed.