Evaluating Geographic Access to Radiotherapy Services in Canada

Abstract

To identify jurisdictions and demographic factors associated with increased driving times to the nearest RT facility in Canada. RT facilities in Canada were identified using information obtained from the Canadian Association of Radiation Oncology and the IAEA Directory of Radiotherapy Centers. The 2021 Canadian census was used to determine demographic profiles for each dissemination area, the smallest geographic unit (target size 400-700 people) on which all census data is reported. Poor geographic access was defined as a driving time between a dissemination area and the closest RT facility exceeding 90 minutes as computed using the OpenStreetMap database. The association between province and poor geographic access was assessed using multivariable logistic regression adjusting for population density by decile. Demographic characteristics were subsequently introduced into the logistic regression to identify factors associated with poor access. A total of 51 radiotherapy centers and 57,937 dissemination areas were mapped. Median driving time to the closest RT facility was 23 minutes, and 13% of the population had poor access to RT. RT was least accessible in the three territories, whose residents had a median driving time of 1096 minutes. Among the provinces, Newfoundland had the longest median driving time of 91 minutes while Manitoba had the shortest median driving time of 20 minutes. On logistic regression with population density as a covariate, dissemination areas in Newfoundland (OR, 9.64 compared to Ontario; 95% CI, 8.3 to 11.2) and British Columbia (OR, 5.46; 95% CI, 5.03 to 5.93) had the highest odds of poor geographic access, whereas Ontario (reference) and PEI (OR, 0.49; 95% CI, 0.34 to 0.69) had the lowest odds. Inclusion of demographic variables in the logistic regression revealed that dissemination areas with higher median age (OR, 1.035 per year; 95% CI, 1.030 to 1.039), more males (OR, 1.033 per %; 95% CI, 1.021 to 1.045), lower income (OR, 1.030 per % low income status; 95% CI, 1.024 to 1.035), increased aboriginal representation (OR, 1.050 per %; 95% CI, 1.045 to 1.054), fewer visible minorities (OR, 0.979 per %; 95% CI, 0.973 to 0.985), fewer immigrants (OR, 0.934 per %; 95% CI, 0.927 to 0.941), and lower attainment of postsecondary education (OR, 0.976 per %; 95% CI, 0.972 to 0.979) were associated with poorer access to RT. Approximately 13% of Canadians have a driving time exceeding 90 minutes to the closest RT facility. Geographic access to RT varied by province and dissemination areas with higher median age, low socioeconomic status, a greater proportion of males, and higher aboriginal populations were associated with poorer access. Our study identified deficiencies in access to existing RT facilities, and its results may inform planning efforts for locations of new facilities in Canada.