BackgroundButaro Cancer Center of Excellence (BCCOE) was founded to serve Rwanda’s rural low-income population, providing subsidized cancer diagnosis and treatment with transport stipends for the lowest-income patients. We examined whether travel distance to BCCOE was associated with advanced-stage diagnoses and treatment completion.MethodsWe conducted a retrospective cohort study using medical record data from BCCOE patients with pathologically-confirmed breast cancer from 2012–2016. Women with no prior surgery were included in the stage analysis; those with non-metastatic disease were included in the treatment analysis. We calculated travel distances using spatial analytic software and used multivariable logistic regression to examine the association of distance and other patient characteristics with late-stage diagnoses and treatment completion within one year of diagnosis.ResultsThe analytic cohort for stage included 426 patients; 75.1% had late-stage (stage 3 or 4) disease. In univariable analyses, patients residing in BCCOE’s surrounding district had a lower proportion of late-stage diagnoses compared to those residing outside the district (57.9% v 76.8%, p = 0.02). In adjusted analyses, odds of late-stage diagnosis were 2.46 (95% CI:1.21–5.12) times higher among those in distance quartile 4 (> 135.8 km) versus 1 (< 55.7 km); the effect of distance was less strong in sensitivity analyses excluding patients from BCCOE’s surrounding district. Patients from sectors with > 50% poverty had 2.33 times higher odds of late-stage diagnoses (95% CI:1.07–5.26) relative to those with poverty < 30%. In the treatment completion cohort (n = 348), 49.1% of patients completed surgery and chemotherapy within a year. In adjusted analyses, travel distance and poverty were not linearly associated with treatment completion.ConclusionsAt Rwanda’s first public cancer facility, sector-level poverty and longer travel distances were associated with late-stage breast cancer diagnoses, but less clearly associated with treatment completion, perhaps partly due to travel stipends provided to the lowest-income individuals undergoing treatment. Our findings support further investigation into wider use of travel stipends to facilitate early diagnosis and treatment completion.
Read full abstract