LAUSR

ABSTRACT Clinical relevance Homeless populations have lower health indicators, including in eye care. Few data exist on the levels and causes of visual impairment in Canadian homeless populations, and none in Montreal. Background This study aims to characterise the causes and levels of visual impairment, as well as eye care services utilisation among the Montreal homeless. Methods Using random sampling, five homeless shelters were selected. In each shelter, 20 participants were randomly selected. After obtaining informed consent, participants completed an ocular examination, which included: presenting visual acuity (pinhole as needed), intraocular pressure, confrontation visual field, dilated fundus examination, post-dilation autorefraction and questionnaire on social determinants of health. Results A total of 95 participants were examined, of which 97.9% were male. The median age was 49 years old (interquartile range 38–56.5). The age-adjusted prevalence of visual impairment (presenting visual acuity <6/12) was 23.6% (95% CI 15.1–32.9) compared to 6.0% in the Canadian population (Z = 77.9, p < 0.0001). With pinhole correction, the prevalence of visual impairment dropped to 5.8% (95% CI 1.7–11.8). Prevalence was 8.2% (95% CI 3.7–15.9) for cataracts, 11.4% (95% CI 5.9–19.7) for glaucoma or suspects and 4.7% (95% CI 1.7–11.9) for diabetic retinopathy. Lastly, 18.9% of participants had an ocular examination within the last year compared to 41.4% in Canada (Z = −4.5, p < 0.0001) and 13.7% had never had a comprehensive eye examination. Conclusions This sample population shows a prevalence of visual impairment which is four times that of the general Canadian population, with highly prevalent uncorrected refractive error, while accessing primary eye care twice less often.