This study adhered to the tenets of the Declaration of Helsinki and ethics approval was granted by the University of Western Australia Human Research Ethics Committee (2024/ET000213).
Study sample
The Modified Monash Model (MMM) 2019 [20] rural classification system was chosen for this study as it measures remoteness and population size, and is the model that the Australian Government is transitioning towards for workforce programs [21]. MM 1 classification refers to metropolitan centres with increasing rurality as the number increases. Service locations were included in this study if they were located in MM 2 (regional centres), MM 3 (large rural towns), MM 4 (medium rural towns), MM 5 (small rural towns), MM 6 (remote communities), or MM 7 (very remote communities).
Practice locations
Ophthalmology and optometry practice addresses in Western Australia were identified using open access sources including the National Health Services Directory, Google Maps, and the Optometry Australia directory. All addresses were entered into the Australian Government Health Workforce Locator [22] to confirm they were located within MM 2 to MM 7.
Identified practices were contacted by telephone to check they were operational, and the clinic manager, optometrist, or ophthalmologist was invited to complete an online survey. Survey questions were designed to understand the distribution and availability of services, including the total full-time equivalent (FTE) at each practice and the availability of equipment at the practice. Ophthalmology hours included consulting and surgical time. Questions were modified for each profession; for example, optometry practices were asked about where patients were referred to for ophthalmology care, and ophthalmology practices were asked about subspecialty services.
An online survey was piloted with 12 participants to check the survey clarity and ease of use. Pilot participants shared a similar background to those in the implemented study but were based in a different geographic location. Questions were revised in response to feedback and pilot responses were not included in the final study. The final surveys were made available online using Qualtrics and included 23 questions for the optometry survey and 16 questions for the ophthalmology survey (Supplementary file 1). The survey link was emailed to practices that expressed interest over the phone, and a reminder email was sent after a few weeks. The survey was administered from May 2024 to August 2024. Responses were exported and collated in Microsoft Excel (version 2402). Follow up calls and online open-source information were used to complete data that were missing.
Visiting services
A combination of sources was used to discover the location of visiting services. Visiting services were defined in this study as outreach services where practitioners travel to deliver care to a location where permanent eye care services did not exist. A question in the survey asked whether practitioners delivered outreach services. If the response was affirmative, respondents were asked about the frequency and location of the services. Rural Health West, the state rural workforce agency, and Lions Outback Vision, an organisation that delivers outreach services, provided additional data on visiting services for both optometry and ophthalmology. Data on the type of outreach, frequency, and duration of service were collected from each source.
Population and boundary data
Population data from the most recent Australian Census in 2021 and geographical boundary shapefiles were obtained from the Australian Bureau of Statistics (ABS) website [5]. Greater Capital City Statistical Areas data from the ABS [23] were used to remove metropolitan areas of Western Australia from the map. Statistical Areas One (SA1) are the smallest unit of Census of Population and Housing data from the ABS [23] and were used to create high-acuity distance-based population analyses. It should be noted that the MMM 2019 classification system is based on the 2016 Census data and new SA1 areas were introduced in the 2021 Census. Therefore, some SA1 areas had no MMM classification. For FTE mapping of service-to-population ratios, SA3 population data were used. This was deemed more appropriate than SA1 as the geographic distribution of the population in remote Australia means that communities often have vast distances between them [21]. Using SA1 to visually display FTE mapping would have resulted in expansive areas where no services appeared available. Furthermore, SA3 boundaries share similarities with the regions defined by the Western Australia State Planning Commission [24] which may facilitate more meaningful comparisons in the context of service availability interpretation and alignment with policymakers.
Data analysis – mapping
All identified service location addresses were geocoded using Google Maps API into latitude and longitude coordinates [25]. Survey responses and visiting service information from organisations were cross-referenced to ensure that no duplication occurred. Where there was a discrepancy of hours reported between the eye care service provider and the funding organisation, the survey response from the provider was used. Quantum GIS (QGIS [version 3.34.3]) software was used to layer population and boundary data with practice coordinates to map services.
For distance-based analyses, SA1 centroids and Euclidean buffers of 50 km and 100 km were used. Euclidean distances are commonly used in the assessment of geographic availability of healthcare services and have strong correlation with network distances [26]. A distance of 100 km was chosen as this has been previously used in studies to measure healthcare access in rural and remote Western Australia [27] and 50 km was chosen to account for the different distances that patients are willing to travel depending on rurality [28]. Centroid position relative to buffers were extracted and analysed with population data in Excel. Using QGIS, population data were randomly distributed across each SA1 for the purpose of visually representing the relative population density but were not used in data analysis. The mapping of services used ABS data and survey responses on service location only, and survey responses regarding equipment and nearby services were not integrated into the GIS analysis.
Data analysis – survey
For this study, FTE per 100,000 was chosen as the metric to calculate service-to-population ratios as this facilitates comparison with AIHW workforce reports. Vector analysis tools were used to identify the SA3 that each service was located in. Then, the FTE per 100,000 of the population was calculated for each SA3, assuming 1 FTE to be 38 h for optometrists and 40 h for ophthalmologists in line with AIHW calculations [29]. For the six practices (9.1%) that did not respond and where no open access information was available, the average FTE from all other respondents was used. Other survey responses, including equipment, were collated and descriptively analysed in Microsoft Excel.
link