This research is based on the number of optometrists per population (i.e. optometrist distribution ratio), eye services utilization (including optometrists and ophthalmologists), and population subgroups that may have higher health care needs (i.e. sociodemographic characteristics such as seniors population, low-income measures, and less education) (see Figure 1). [22, 26-28] The primary practice locations of optometrists in Canada were gathered from the provincial regulatory bodies. The Canadian Association of Optometrists (CAO) gathered primary practice information of optometrists for 2017 (i.e. six-digit postal codes) from seven provinces (British Columbia, Manitoba, New Brunswick, Ontario, Prince Edward Island, Quebec, Saskatchewan) whereas data from the remaining three provinces (i.e. Alberta, Newfoundland and Labrador, Nova Scotia) was downloaded directly from each of the provincial regulatory college’s website. CAO was unable to provide optometrist information for three Territories due to unavailability of any licensing bodies to provide data. A set of geographic coordinates for primary practice locations were generated using postal code geocoding and Google Maps. Next, in order to aggregate data in various geographic scales, supporting attributes from other layers such as health region boundaries, census subdivision (CSD) geographic units were assigned to each location.
Geographic proximity of optometry services was measured in terms of the number of optometrists per 10,000 population at health region levels (i.e. optometrist distribution ratio). The number of optometrists extracted from the provincial regulatory bodies (December 2017 to July 2018) from either the 2017 or 2018 registration year combined with Census derived population figures were used for estimating optometrist ratios at a health region level. Information about utilization of eye care services (i.e. combination of optometrist or ophthalmologist) is based on the CCHS 2013-2014 that was accessed via Ontario Data Documentation, Extraction Service and Infrastructure (odesi) web-based data exploration, extraction, and analysis tool (https://odesi.ca/). The CCHS is a cross-sectional, nationwide, and self-reported household survey that was collected from persons aged 12 and over living in Canadian health regions except those living on a reserve or as fulltime member of the Canadian Forces. [29] To get a fair sample distribution to the health regions and the provinces, the CCHS survey adopted a multi-stage sample allocation strategy including each provinces’ sample is allocated among its health regions as per their size of the population. [29] We used the following question to derive the information about utilization of eye care services at health regions: “CHP_Q06: [Not counting when you were an overnight patient, in the past 12 months/In the past 12 months], have you seen, or talked to: an eye specialist, such as an ophthalmologist or optometrist (about your physical, emotional or mental health)?” Unfortunately, the wording of the CHP question related to vision care services does not distinguish between optometrist or ophthalmologist use. Comparative analyses of ratio and utilization variables in association with population subgroups that usually have much higher health care needs was performed. We focused on the following three population subgroups with potentially higher needs: seniors (age 65 years and over), low-income, and lower educational attainment. Information about these three variables were extracted from 2016 Census and downloaded from the Statistics Canada website. The 2016 dissemination area (DA) census data were used to prepare the following HR level variables: population 65 years and over, low-income measures (after tax), and the population aged 15 and over with less than a high school diploma. These variables were expressed as percentages. Low-income measures is one of three measures of low income in Canada that calculates relative measures of low income based on the national income distribution where an adjustment of 50% median household income is set as a threshold. [30, 31] In this study, data were gathered from multiple sources at various geographic levels such as optometrist use at health regions, population subgroups at DA, and optometrist work location at locational scale, however, health regions are used as the unit of analysis.
Geospatial mapping methods that were used to analyze the patterns of optometrists per 10,000 population (i.e. ratio), self-reported eye care services utilization, and population subgroups can be divided into three ways. First, optometrist practice locations were associated with the different urban-rural classifications where we used statistical area classification to categorize census subdivisions (municipalities) into metropolitan and metropolitan influence zones (MIZs).[32] The MIZ classifies the CSDs outside census metropolitan areas (CMAs) and census agglomerations (CA) into four categories according to the degree of influence (strong, moderate, weak, or no influence) that the CMAs or CAs have on them. [32] These categories are based on the proportion of employed residents in a given CSD that commute to work in a CMA or CA (i.e. strong >30%, moderate 5-30%, weak< 5%, no influence 0 residents). [32] Second, optometrist ratios estimated at health regions levels were mapped. This was done after converting ratio values into five categories where a standard deviation (SD) classification approach was used (± 0.5 SD from the mean value were used as a cut-off for demonstrating distribution of optometrists across health regions).[33, 34] Third, a cross-classification technique was utilized to map the patterns of spatial distribution of optometrists in relation to self-reported use of vision care services and population subgroups. This was performed after separating each variable into three classes (i.e., low, moderate, and high). For this, a standard deviation classification scheme was followed where a ± 0.5 SD from the mean value was used as a cut-off for demonstrating distribution of each variable across health regions. For example, in case of optometrist ratio, the first two categories (< - 1.5 SD; -1.5 to - 0.50 SD) indicate poor distribution of optometrists (i.e., lower category), the third category (-0.5 to 0.5 SD) moderate, and the last two (0.5 to 1.5 SD, > 1.5 SD) indicate higher geographical availability of optometry services.
The following software were used for mapping and data analysis (spatial and nonspatial): ArcGIS Map, SPSS, and Microsoft Excel. A thematic mapping tool available in ArcGIS software (ArcGIS Desktop version 10.5, ESRI, Redlands, CA) was used to prepare a set of maps. Supporting datasets required for mapping were accessed by the research team through the Geographical Information System (GIS) Library Services at the University of Saskatchewan [35]. These datasets included a digital geographic boundary file for health regions, demographic data, digital geographic file of the 2016 Canadian Census at various geographic scales, and CanMap Postal Code Suite for geocoding purposes.