The prevalence of trachoma in Australia: the National Indigenous Eye Health Survey

The sampling method has been described in detail elsewhere.13 Briefly, we grouped Australian Bureau of Statistics (ABS) Indigenous Areas (IAREs) into five strata based on the Accessibility/Remoteness Index of Australia:14 “major city”, “inner regional”, “outer regional”, “remote”, and “very remote”. The very remote stratum was then subdivided into “very remote coastal” and “very remote inland”. Within each of these six strata, five sample areas or communities were randomly selected using 2006 ABS census data15 to identify geographical areas containing about 300 Indigenous people. The sample size of 1500 children aged 5 to 15 years and 1500 adults aged 40 years and older was selected based on the ability to detect changes in vision impairment;13 it also conformed to the World Health Organization’s recommendation for the assessment of trachoma of 50 children per cluster.16 Fieldwork was undertaken between January and December 2008.

An informal census was undertaken at each site using community data and local informants to establish the size of the eligible population of Indigenous children aged 5 to 15 years (inclusive) and Indigenous adults aged 40 years and older.

A standardised questionnaire collected data on demographics, general health, eye health and health service utilisation, and a standardised eye examination was performed.13 Trachoma was graded in each eye using a × 2.5 loupe according to the WHO simple grading system,17 where:

• trachomatous inflammation, follicular (TF) was defined as the presence of five or more follicles in the upper tarsal conjunctiva;

• trachomatous inflammation, intense (TI) as pronounced inflammatory thickening of the tarsal conjunctiva that obscures more than half of the normal deep tarsal vessels;

• trachomatous trichiasis (TT) as the presence of at least one eyelash rubbing on the eyeball;

• trachomatous scarring (TS) as the presence of scarring in the tarsal conjunctiva; and

• corneal opacity (CO) as easily visible corneal opacity over the pupil.

The trachoma graders were given self-directed trachoma grading training using an interactive CD-ROM.18 Photographs of the everted left tarsal plate were taken with a Nikon D40 digital camera using an 80 mm lens and extension rings (Nikon Corporation, Tokyo, Japan).

All trachoma photographs were graded twice by the same experienced grader using a fine grading system.19 Follicles 1 (F1) was defined as one or two small follicles in the upper tarsal conjunctiva; Follicles 2 (F2) as greater than two but fewer than five follicles of 0.5 mm in diameter in the upper tarsal conjunctiva (ie, less severe than TF); Scarring 1 (S1) as a small amount of early scarring apparent, but not clearly visible; and Scarring 2 (S2) as a moderate amount of early scarring apparent, but not clearly visible (ie, less severe than TS). Grading was performed on a laptop computer without knowledge of the clinical or photograding.

When the two photogradings disagreed on the presence or absence of a WHO-defined sign, the photograph was regraded for a third, definitive time.

In this analysis, the photograding was used to define trachoma (TF, TI and TS) except when photographs were not available, in which case the clinical grading was used. TT and CO were assessed clinically. Participants were considered to have active trachoma if TF and/or TI were present.

Data were entered into an electronic database using Access 2000 (Microsoft Corporation, Redmond, Wash, USA). Data were checked for missing values, and missing data were followed up where possible. Analyses were performed using Microsoft Access and Stata, version 10.2 (StataCorp, College Station, Tex, USA). Age-specific rates for adults are shown in 5-year age groups. A rolling 3-year and three-age-group (ie, 15 years) average was used to graph age-specific data for children and adults, respectively.

Primary ethics approval was obtained from the Human Research Ethics Committee of the Royal Victorian Eye and Ear Hospital. Further separate formal ethics approval was obtained from the Aboriginal Health and Medical Research Council of New South Wales, the Aboriginal Health Council of South Australia, the Menzies School of Health Research, the Central Australian Human Research Ethics Committee, the Western Australian Aboriginal Health Information and Ethics Committee, ACT Health (12 November 2007), the Tasmanian Scientific Research Advisory Committee, the Tasmanian Health and Medical Human Research Ethics Committee, and the Queensland Aboriginal and Islander Health Council.20 The protocol was also reviewed and approved by the Board of the National Aboriginal Community Controlled Health Organisation and by each community.

Permission was obtained to examine children in schools. This research was conducted in accordance with the tenets of the Declaration of Helsinki as revised in 2000. Written, informed consent was obtained for all participants before examination.

The unadjusted rate of active trachoma (TF) in children was 3.8%; after adjustment for sampling, the rate was 2.8%. All the children with TI also had TF, so rates for TF and TI are not presented separately (Box 1). Two adults, both women, had TF; both came from very remote communities, and one also had TI. The prevalence of TF among children in these two communities was 10.1% and 23.0%, respectively. In 14 communities, no children had TF; sporadic single cases of TF occurred in one major city, and one inner regional, two remote, and two very remote coastal sites (Box 2). In 10 communities, two or more children had TF, but the prevalence in five of these communities was less than 5%. However, in five communities, all in the two very remote strata, the prevalence of TF was 5% or greater.

The age-specific rates show a peak prevalence of TF among the youngest children and a progressive overall decrease with increasing age (Box 3). In Australia, the prevalence of TF is usually presented for 5 to 9 year olds;9 862 children aged 5 to 9 years were examined for our study. The overall prevalence of TF in these children mirrored that found among all children 5 to 15 years. However, the prevalence of TF among children aged 5 to 9 years was higher in several communities — six of the 10 very remote communities had a prevalence greater than 5%; the highest was 25.7%. The overall prevalence of TF among children aged 5 to 9 years was 5.2%.

Scarring was uncommon among children, even in communities with a higher prevalence of TF (Box 2). Conversely, TS was common among adults and widely spread, occurring in all strata; in three very remote communities, over half the adults were affected (Box 2). The age-specific rates of scarring increased with age, although in most strata, rates fell in the oldest age groups (Box 4).

Seven men and nine women had TT. Their ages ranged from 51 to 88 years. Three of these participants also had CO — two women aged 64 and 88 years and a man aged 83 years. One had visual acuity of < 6/18 in one eye and < 6/60 in the other, the second had visual acuity of < 6/60 in both eyes, and the third had light perception in one eye and no light perception in the other. A fourth participant had visual acuity of < 6/12 associated with TT. An additional participant with TT was blind from cataract. The age-specific rates show a gradual increase of TT in the older age groups (Box 5). No child was found to have TT or CO.

Overall, 6.2% of children had F1, 4.8% had F2 and 3.8% had TF. None of these signs were seen in six communities, and F1 alone was detected in three communities (Box 6). F2 occurred in five communities without TF, although in the remaining 16 communities TF was accompanied by F1 and F2 (Box 6). There was a significant correlation between the presence of F2 and TF (r = 0.72; P < 0.001). There was only a weak, non-significant correlation between F1 and TF (r = 0.34; P = 0.07).

Overall, 3.6% of adults had S1, 8.5% had S2 and 15.7% had TS. In three communities, no adults had any signs of scarring (Box 7). S1 alone was found in one community, S2 occurred alone in one community and together with S1 in another, and TS was present in 24 (Box 7). These fine grading signs were clustered strongly. The correlation between the presence of S2 and TS was strong (r = 0.49; P = 0.006), but there was no correlation between S1 and TS (r = − 0.002; P = 0.99).

In addition to the study population, 375 Indigenous children and 404 adults who lived outside the defined study area presented for examination. Ineligible individuals were examined in 21 of the 30 sites. The prevalence of TS in ineligible adults was 6.7%, and its distribution mirrored that in eligible adults. TT was found in one ineligible adult in a community where TT also occurred in eligible adults. The rate of TF was 4.0% in ineligible children, and its distribution mirrored that seen in eligible children, with one exception. In one outer regional site in Queensland, four of 37 ineligible children (11%) and three of 85 adults (4%) were found to have TF, even though TF had not been found in the eligible participants. Most of those with TF were members of one extended family.