Public-health-center HIV screening coverage and AIDS-stage diagnosis in Japan before and during the COVID-19 pandemic: a nationwide prefecture-level panel study

Public health centers provide free anonymous HIV testing in Japan, but represent only one component of the national testing system. We examined whether public-health-center HIV screening coverage was associated with AIDS-stage diagnosis and whether a pandemic-period association remained after accounting for regional differences. We conducted a nationwide ecological panel study using annual data from all 47 prefectures between 2015 and 2023. The outcome was the number of AIDS-stage diagnoses among newly reported HIV/AIDS cases. The primary exposure was the one-year-lagged public-health-center HIV screening rate per 100,000 residents, expressed per doubling. Binomial regression models included calendar-year trend and a pandemic-period indicator for 2020–2023; the principal model additionally included prefecture fixed effects. Standard errors were clustered by prefecture. The analysis included 412 prefecture-year observations. During 2015–2023, 10,820 HIV/AIDS cases were newly reported. Public-health-center HIV screening decreased by 65.0%, from 103,082 tests in 2019 to 36,056 in 2020, while the national proportion of AIDS-stage diagnoses increased from 26.9% to 31.5%. In the pooled model, a doubling of the lagged screening rate was associated with lower odds of AIDS-stage diagnosis (odds ratio [OR] 0.829, 95% confidence interval [CI] 0.773–0.889). This association was attenuated and was not statistically significant after prefecture fixed-effects adjustment (OR 0.948, 95% CI 0.866–1.038). The pandemic-period estimate in the fixed-effects model narrowly reached statistical significance (OR 1.210, 95% CI 1.003–1.458) but was not robust when the period was divided into 2020–2021 and 2022–2023. The inverse association between public-health-center HIV screening and AIDS-stage diagnosis was largely attenuated after accounting for time-invariant prefecture characteristics. The pandemic-period association was sensitive to model specification. These findings suggest that stable regional characteristics and broader pandemic-related changes may contribute to differences in AIDS-stage diagnosis, but do not support a causal effect of public-health-center screening or the COVID-19 pandemic.