|
nif:isString
|
-
Ethics Committee approval was obtained from the Royal Victorian Eye and Ear Hospital Human Research Ethics Committee (HREC-14/1199H). Approvals were obtained from the following state-level Indigenous ethics bodies to conduct research within Indigenous communities: Aboriginal Health and Medical Research Council of NSW (HREC-1079/15), the Menzies School of Health Research (HREC-2015-2360), the Aboriginal Health Council of Western Australia (HREC-622) and the Aboriginal Health Council of South Australia (HREC-04-15-604). Participants provided written informed consent. This study was conducted in accordance with the tenets of the Declaration of Helsinki.
The sampling methodology has been described in detail elsewhere [12]. In brief, data collection was conducted between March 2015 and April 2016. Probability proportional to size (PPS) multistage random cluster sampling was used to select a representative sample of Indigenous Australians aged 40 years and older and non-Indigenous Australians aged 50 years and older. Population data collected in the 2011 Australian Census were used to select 30 geographic areas across all geographic remoteness strata in Australia. Twelve Major City areas, six Inner Regional areas, six Outer Regional areas, 4 Remote areas and 2 Very Remote areas were selected. In total, 4,520 eligible non-Indigenous residents and 2240 eligible Indigenous residents were enumerated across all survey sites by trained recruiters. Of these, 3098 non-Indigenous participants (68.5% response rate) and 1738 Indigenous participants (77.6% response rate) were recruited to participate in the survey.
Recruited residents in each survey site attended a testing centre within 6km of the selected population cluster and provided written informed consent. Each participant underwent a standardised interviewer-administered questionnaire to collect key sociodemographic data, including gender, age, level of education and ethnicity. Participants were asked whether they were of Aboriginal or Torres Strait Islander origin (Indigenous Australians). The ethnicities of non-Indigenous Australians were categorised according to the Australian Standard Classification of Cultural and Ethnic Groups 2011 based on self-reported country of birth [13]. Stroke and diabetes histories were also recorded. Past ocular history was recorded for all participants, including history of diagnosis of refractive error, cataract, glaucoma, diabetic retinopathy, age-related macular degeneration, or other diseases. Participants were asked if they had ever undergone an eye examination, and if so, how recently. Those who had undergone an eye test were asked who had performed their most recent examination, and responses were recorded against a standardised list defined a priori: 1) Optometrist, 2) Ophthalmologist, 3) GP/local doctor, 4) Nurse, 5) Health worker, 6) Ophthalmic nurse/technician, 7) Other. Participants were asked whether they wore spectacles or contact lenses, and if so, whether their refractive correction was for distance or near correction or both.
A series of standardised eye examinations was performed by a trained examiner. Presenting distance visual acuity was measured in each eye separately using a logMAR chart (Brien Holden Vision Institute, Australia) at three metres in well-lit room conditions. If presenting visual acuity was worse than 6/12 in one or both eyes, pinhole testing was conducted. If vision improved to 6/12 or better in one or both eyes, auto-refraction was performed using a Nidek Ark-30 Type-R Hand-held auto-refractor/keratometer (Nidek Co. LTD, Japan). Trial lenses corresponding to auto-refraction measurements were placed in a trial frame and auto-refraction-corrected visual acuity was measured in each eye separately. Binocular presenting near vision was measured in well-lit room conditions using a CERA Vision Test E Chart (Centre for Eye Research Australia, Australia), held at the participant’s preferred reading distance. Anterior segment assessment, perimetry, intraocular pressure testing and two-field fundus photography were performed. Examiners provided participants with verbal feedback on their test results, and if abnormalities were detected, a referral letter was provided to be taken to a local doctor or optometrist.
Definitions of uncorrected or under-corrected refractive error and refractive error treatment coverage rates: This analysis focused on participants in whom uncorrected or under-corrected refractive error was the main cause of vision loss. Uncorrected or under-corrected refractive error was determined to be the main cause of vision loss if the distance visual acuity in one or both eyes improved to better than or equal to 6/12 (≥6/12) with pinhole testing or auto-refraction. The threshold of 6/12 was selected as this is considered the legal threshold for vision impairment in Australia [14]. Participants for whom visual acuity improved with pinhole or autorefraction, but their improvement was below the 6/12 threshold remained bilaterally vision impaired, and were not included as having uncorrected or under-corrected refractive error for the purpose of this analysis. Their vision loss was primarily caused by a different condition, and adequate visual function would not be restored with refractive correction. As participants with presenting visual acuity ≥6/12 did not undergo pinhole testing or autorefraction, those individuals with ≥6/12 vision and mild refractive error who may have improved further with refraction were not identified. These participants were also not considered to have uncorrected or under-corrected refractive error. Refractive error treatment coverage rates were calculated for participants whose vision loss was caused by uncorrected refractive error using the following formula: Refractive error treatment coverage rate = (n1n1+n2)×100. In this formula, n1 was the number of participants who reported that they wore spectacles and/or contact lenses for distance vision and achieved bilateral presenting distance visual acuity ≥6/12, and n2 was the number of participants who had refractive error as their main cause of bilateral vision loss (<6/12). Participants in n2 may have had no refractive correction (uncorrected refractive error) or they may have had refractive correction that was not sufficient to correct their visual acuity to better than 6/12 (under-corrected refractive error).
Means and standard deviations (SD) were calculated for normally-distributed continuous sociodemographic variables, and medians and inter-quartile ranges were calculated for skewed data. Counts and percentages were calculated for categorical sociodemographic variables. Refractive error treatment coverage rates were calculated for Indigenous and non-Indigenous participants. Sampling weight-adjusted coverage rates were calculated using logistic regression models. Treatment coverage rates were disaggregated by age, gender, ethnicity, language spoken at home, and geographic remoteness, and were tabulated as counts and percentages. Multivariable logistic regression analysis was used to examine the association between spectacle or contact lens correction and the following variables: Ethnicity, including Indigenous/non-Indigenous status, age (years), gender, number of years of education, main language spoken at home (English/other), geographic remoteness time since last eye examination and the type of eye health care professional who conducted the examination. Excluding ‘optometrist’ and ‘ophthalmologist’, the a priori list of eye health care providers was collapsed into the group ‘other’ due to a small sample size in each group and compared against ‘optometrist’ and ‘ophthalmologist’. Due to differences in inclusion criteria and sampling between Indigenous and non-Indigenous participants, regression analysis was performed separately for each group. For the final fitted logistic regression model, model residuals and delta beta values were examined to determine if potential outlying observations influenced results. The degree to which statistical assumptions were violated was also examined. Associations were considered statistically significant if p<0.05. All statistical analyses were undertaken using Stata version 14.2 (StataCorp, College Station, TX).
|
![[This page as RDF]](http://data.gesis.org/somesci/static/rdf-icon.gif){kind=icon}