HIV testing history and positivity among repeat testers and first-time testers in older adults: a descriptive study from Hejiang county, Southwest China, 2018-2025.

BACKGROUND

Understanding HIV testing patterns among older adults is critical for optimizing screening strategies, yet detailed descriptions of testing history remain scarce. We leveraged data from a county-wide screening program to characterize testing patterns and HIV positivity among adults aged ≥50 years in southwestern China.

METHODS

We used data from a 2025 county-wide HIV screening program in Hejiang County. Records were linked to a provincial historical database (2018-2024) to classify participants as first-time or repeat testers. We compared demographics and HIV positivity rates between the two groups, and estimated the annual HIV detection rate (per 1,000 person-years) among repeat testers from 2018 to 2025. Multivariable logistic regression assessed the association between testing history and HIV positivity, adjusting for age, gender, education, and marital status. Multivariable Cox regression examined factors associated with HIV detection, with a counting process model as main analysis and a midpoint-based model as sensitivity analysis.

RESULTS

Among 235,771 tests, overall HIV positivity was 0.50% (n = 1,176). First-time testers (n = 20,438) were younger (60.2% aged 50-59 years) and more often male (57.8%), whereas repeat testers (n = 215,333) were older (41.5% aged ≥70 years). Positivity rates were higher among repeat testers than first-time testers (0.51 vs. 0.38%), especially in males. In multivariable logistic regression, repeat testers had higher odds of HIV positivity (aOR = 1.53, 95% CI: 1.39-1.64, P < 0.001). Among repeat testers, the annual detection rate rose from 0.19 (2018) to a peak of 1.32 (2021), then fell to 0.55 (2025). In multivariable Cox regression, male (HR = 2.61), age 60-69 years (HR = 1.54), unmarried (HR = 2.39), divorced (HR = 2.06), and widowed (HR = 5.65) were associated with higher hazard, while high school education or above was protective (HR = 0.36). Sensitivity analysis using a midpoint-based Cox model yielded consistent estimates (e.g., HR for male: 2.61 in both models), with all P < 0.05 except age ≥80 and unknown marital status.

CONCLUSION

This descriptive study reveals marked demographic differences and HIV positivity between first-time and repeat testers aged ≥50 years, and identifies demographic factors associated with HIV detection. Linking screening data to historical records offers a practical method for characterizing testing histories and may serve as a reference for similar settings.