A geostatistical study of socioeconomic status (SES), rurality, seasonality and index test results as drivers of free private groundwater testing in southern Ontario, 2012–2016

Approximately 12% of the Canadian population uses private wells for daily water consumption; however, well water testing rates are on the decline, resulting in an increased risk of waterborne acute gastrointestinal illness. To date, limited research has explored the determinants influencing well testing practices. Accordingly, the current study sought to investigate the drivers of “one-off” and repeat well water testing in southern Ontario during the 5-year period 2012–2016, using the worlds largest private groundwater testing data-frame. Data from >400,000 wells were geospatially integrated with all tests conducted by the provincial laboratory in southern Ontario. The Ontario Marginalization Index (ON-Marg) was used as a proxy measure of socioeconomic status (SES), with rurality, based on population density, season, and index (1st) test results assessed as effect modifiers. Multivariate analysis was undertaken using log-binomial regression. Approximately 27.5% of wells (n = 417,406) were tested during the study period, 66.7% of which were sampled more than once; 3% of all samples tested positive for E. coli (>0 colony forming unit/100 mL). In rural regions (<150 people/km²), wells located in low SES areas were 13% more likely to be tested compared to high SES areas (95% CI: 1.11, 1.15). In urban (>400 people/km²) and peri-urban regions (>150 and <400 people/km²), wells located in low SES areas were 14% (95% CI: 0.78, 0.95) and 15% (95% CI: 0.76, 0.94) less likely to be tested compared to high SES areas. Wells located in low SES areas were 6% more likely to be re-tested (95% CI: 1.04, 1.07). Positive index tests were associated with a 17% increased likelihood of repeat testing (95% CI: 1.16, 1.18). Accordingly, the authors conclude that location and SES are significant predictors of well water testing, with index test status being the most influential predictor of repeat well testing.