This study complements the existing literature on disparities associated with Indigenous and non-Indigenous small drinking water systems. The team took a quantitative approach and assessed relationships between seasonality, location, and type of community against the number of drinking water advisories in Saskatchewan for a 4-year period from 2012 to 2016. Generalised estimating equations were used to determine significant factors contributing to the likelihood of drinking water advisories comparing Indigenous to non-Indigenous communities of similar sizes. Results indicated that the season and the interaction between community type and region (north vs. south) were significant in the model for counts of advisories. Reserve communities in the north had a drinking water advisory count that was 5.19 times greater than those of reserves in the south, 2.63 times greater than counts for towns in the south and 4.94 times greater than those of villages in the south. Additional comparisons indicated that reserves in the north had 2.43 times as many advisories as villages in the north, but towns situated in the south part of the province had 1.98 times as many advisories as reserves in the south, and 1.88 times as many advisories as villages in the south. The work confirms heightened risk among northern Indigenous communities and suggests that increased attention to, and investment in, securing water resources is necessary in rural Saskatchewan and globally.

Rural communities dependent on unregulated drinking water are potentially at increased health risk from exposure to contaminants. Perception of drinking water safety influences water consumption, exposure, and health risk. A community-based participatory approach and probabilistic Bayesian methods were applied to integrate risk perception in a holistic human health risk assessment. Tap water arsenic concentrations and risk perception data were collected from two Saskatchewan communities. Drinking water health standards were exceeded in 67% (51/76) of households in Rural Municipality #184 (RM184) and 56% (25/45) in Beardy's and Okemasis First Nation (BOFN). There was no association between the presence of a health exceedance and risk perception. Households in RM184 or with an annual income >$50,000 were most likely to have in-house water treatment. The probability of consuming tap water perceived as safe (92%) or not safe (0%) suggested that households in RM184 were unlikely to drink water perceived as not safe. The probability of drinking tap water perceived as safe (77%) or as not safe (11%) suggested households in BOFN contradicted their perception and consumed water perceived as unsafe. Integration of risk perception lowered the adult incremental lifetime cancer risk by 3% to 1.3 × 10-5 (95% CI 8.4 × 10-8 to 9.0 × 10-5 ) for RM184 and by 8.9 × 10-6 (95% CI 2.2 × 10-7 to 5.9 × 10-5 ) for BOFN. Probability of exposure to arsenic concentrations >1:100,000, negligible cancer risk, was 23% for RM184 and 22% for BOFN.