Trends in incidence and medical costs of human papillomavirus-related diseases in Korea: evidence from National Health Insurance claims.
OBJECTIVE Human papillomavirus (HPV) infection is a major cause of cervical, anal and oropharyngeal cancers. Although Korea has implemented a national HPV vaccination programme since 2016, nationwide evidence on incidence trends and medical costs of HPV-related diseases remains limited. METHODS This retrospective population-based study analysed Korean National Health Insurance Service claims data. Age-standardised incidence (ASI) rates of cervical intraepithelial neoplasia (CIN1 and CIN2/3), genital warts (GWs), and cervical, oropharyngeal and anal cancers from 2015 to 2022 were calculated using the 2022 Korean population as the standard. Trends in ASI were analysed using Joinpoint regression with a piecewise log-linear (Poisson) model. For diseases with and without statistically significant Joinpoints, trends were reported using pre-identified and post-identified segment-specific annual percentage changes (APCs) and the average annual percentage change (AAPC), respectively. Trends in total medical costs from 2015 to 2021 were evaluated using linear regression. RESULTS ASI trends differed by sex and disease type. Cervical cancer incidence among females declined significantly during 2015-2022 (AAPC; -4.2%, p<0.001). Joinpoint regression identified significant inflection points in 2019 for both female and male GWs. Female GW incidence increased during 2015-2019 (APC; 3.7%, p=0.010) and then decreased significantly during 2019-2022 (APC; -14.0%, p<0.001). Male GW incidence increased significantly during 2015-2019 (APC; 11.3%, p<0.001) and then showed no statistically significant change during 2019-2022 (APC; 1.4%, p=0.290). CONCLUSIONS These descriptive findings provide population-level evidence on HPV-associated disease burden in Korea but cannot establish causal attribution to vaccination, screening or other preventive interventions, necessitating continued surveillance with individual-level data linkage.