GPS-measured distance to clinic, but not self-reported transportation factors, are associated with missed HIV clinic visits in rural Uganda

Objective: Studies of the association between transportation barriers and HIV-related health outcomes have shown both positive and negative effects, possibly because a reliable, validated measure of transportation barriers has not been identified. Design: Prospective cohort study of HIV-infected patients in rural Uganda. Methods: Participants were enrolled from the HIV clinic at the regional referral hospital in Mbarara, Uganda as part of the Uganda AIDS Rural Treatment Outcomes (UARTO) Study. We collected the following measures of transportation barriers to HIV clinic: global positioning systems (GPS)-tracked distance measured by driving participants to their homes along their typical route; straight-line GPS distance from clinic to home, calculated with the Great Circle Formula; self-reported travel time; and self-reported travel cost. We assessed inter-measure agreement using linear regression, correlation coefficients and K statistics (by measure quartile) and validated measures by fitting linear regression models to estimate associations with days late for clinic visits. Results: One hundred and eighty-eight participants were tracked with GPS. Seventy-six percent were women, with a median age of 40 years and median CD4 cell count of 193 cells/μl. We found a high correlation between GPS-based distance measures (β = 0.74, P<0.001, R² = 0.92, K = 0.73), but little correlation between GPS-based and self-reported measures (all R²<0.4). GPS-based measures were associated with days late to clinic (P<0.001); but neither self-reported measure was associated (P> 0.85). Conclusion: GPS-measured distance to clinic is associated with HIV clinic absenteeism and should be prioritized over self-reported measures to optimally risk-stratify patients accessing care in rural, resource-limited settings.