|Home | About | Journals | Submit | Contact Us | Français|
To estimate the prevalence and causes of blindness and low vision in people aged 40 years in Timor‐Leste.
A population‐based cross‐sectional survey using multistage cluster random sampling to identify 50 clusters of 30 people. A cause of vision loss was determined for each eye presenting with visual acuity worse than 6/18.
Of 1470 people enumerated, 1414 (96.2%) were examined. The age, gender and domicile‐adjusted prevalence of functional blindness (presenting vision worse than 6/60 in the better eye) was 7.4% (95% CI 6.1 to 8.8), and for blindness at 3/60 was 4.1% (95% CI 3.1 to 5.1). The adjusted prevalence for low vision (better eye presenting vision of 6/60 or better, but worse than 6/18) was 17.7% (95% CI 15.7 to 19.7). Gender was not a risk factor for blindness or low vision, but increasing age, illiteracy, subsistence farming, unemployment and rural domicile were risk factors for both. Cataract was the commonest cause of blindness (72.9%) and an important cause of low vision (17.8%). Uncorrected refractive error caused 81.3% of low vision.
Strategies that make good‐quality cataract and refractive error services available, affordable and accessible, especially in rural areas, will have the greatest impact on vision impairment.
In 1999, after more than 400 years of colonisation, Timor‐Leste, northwest of Australia at the eastern end of the Indonesian archipelago, became an independent state. The health of the Timorese population was poor at that time. This was further compounded by the actions of anti‐independence forces, which resulted in population displacement, loss of skilled health personnel, and almost total destruction of social and physical infrastructure.1 Currently, Timor‐Leste is also one of the world's poorest nations. Its rates of infant and maternal mortality are among the highest. Malnutrition and infectious diseases, such as tuberculosis, malaria and leprosy, are rife.2
The World Health Organization estimates that there are currently 161 million people with vision impairment worldwide, of whom 37 million are blind.3 Approximately 80% of blind people live in the low‐resource countries of Asia and Africa, mostly in rural areas with few or underutilised eye‐care facilities.4,5,6 In many respects, Timor‐Leste is typical of these countries. With a predominantly rural population of 925000 people, Timor‐Leste has one referral hospital, in Dili, the capital. This has an eye clinic and offers cataract surgery and refraction services. Elsewhere, at the beginning of 2005, there were two provincial eye clinics, including one in the Bobonaro district. These provided primary eye‐care and refraction.
In Timor‐Leste's current state of emergent independence and reconstruction, its Ministry of Health is working to develop, establish and maintain an appropriate, cost‐effective and equitable health system. Despite limited health resources, eye health has been included in the Ministry's strategic planning framework. This is because of the significant personal and economic impact of blindness; most blindness is either preventable or treatable,3 and blindness prevention and treatment interventions are cost‐effective.7
As the Ministry prepared to develop a national eye health strategy and implementation plan in accordance with the World Health Organization's Global initiative for the elimination of avoidable blindness8 and the World Health Assembly's Resolution on the elimination of avoidable blindness,9 it became apparent that no population vision and eye health data were available in Timor‐Leste. Given the limited resource setting, a rapid assessment survey was identified as the most appropriate method to obtain an estimation of the prevalence rates and major causes of blindness and low vision in adults aged 40 years, which is the age group most at risk.3 Additional vision‐related quality of life and willingness‐to‐pay data relevant to the Timor‐Leste National Eye Health Strategy were also collected.
This paper reports the prevalence and causes of blindness and low vision in Timor‐Leste.
The study protocol was based on the Rapid Assessment of Cataract Surgical Services (RACSS).10,11,12 Among the modifications made, the entry age was reduced from 50 to 40 years, to permit exploration of aspects of refractive error and presbyopia correction.
In November 2004, a pilot study (30 participants, representative of the urban population to be screened in the main survey) was undertaken to refine and validate the questionnaire, including investigations of inter‐observer and intra‐observer variation, and test–retest reliability. These data were not included in the final survey analysis.
The sample frame for this study included an urban district, Dili, and one rural district, Bobonaro, 4½ h from Dili by road. By using an anticipated prevalence of vision impairment (presenting visual acuity worse than 6/18 in both eyes) of 11% in the target population, absolute precision of ±2.2% (20% relative), with 95% CI, and a design effect of 1.5 and a response rate of 80%, the required sample size was estimated as 1500 persons. From the sample frame, 50 clusters of 30 people were required (25 from all 6 subdistricts of both Dili and Bobonaro). Within each of the districts, the clusters were selected through probability proportionate to size sampling,11 using preliminary data from the 2004 national census.13
The survey was conducted from February to May 2005. Villages comprising each cluster were attended by a single survey team. Using a random process, the team leader (a population health trained optometrist) identified the first household to be targeted. Thereafter, consecutive households were approached according to the RACSS protocol, and eligible people were enumerated by trained Timorese fieldworkers until the 30 participants for that cluster were enrolled.
Presenting visual acuity was measured in daylight for each eye separately using a simplified 3 m chart with tumbling E optotypes equal to standard Snellen sizes of 18, 60 and 120. Vision assessment began at the 6/18 level and progressed to the larger optotypes as required. A minimum of three out of five correctly identified optotypes at the 6/18 and 6/60 level, or two out of three at the 6/120 level, gave the final measurement. Presenting binocular near visual acuity for each participant was also measured. Any eye with visual acuity worse than 6/18 had pinhole acuity measured and underwent an examination to determine the cause of reduced vision.
Assessment of the anterior segment was made using loupe magnification and a torch. The status of the central lens was determined with a direct ophthalmoscope in a dimly lit room through an undilated pupil. An intact red reflex was considered indicative of a “normal” clear central lens. The presence of obvious red reflex dark shading, but transparent vitreous, was recorded as lens opacity. Where present, aphakia and pseudophakia with and without posterior capsule opacification were noted. The lens was determined to be not visible (“no view”) if there were dense corneal opacities or other ocular pathologies, such as phthisis bulbi, precluding any view of the lens. Thereafter, the posterior segment was examined using a direct ophthalmoscope.
Although any eye may have more than one condition contributing to vision reduction, for the purposes of this study, a single cause of vision loss was determined for each eye. In the absence of any other findings, uncorrected refractive error was the attributed cause of reduced vision if the acuity then improved to 6/18 with pinhole. Allocation of other causes, including corneal opacity and cataract, required clinical findings of sufficient magnitude to explain the vision loss. The attributed cause was always the condition most easily treated if each of the contributing conditions were individually treatable to a vision of 6/18 or better. Thus, for example, when uncorrected refractive error and lens opacity coexisted, refractive error, with its easier and less expensive treatment, was nominated as the cause. Where treatment of a condition present would not result in 6/18 or better acuity, it was determined to be the cause rather than any coincident or associated conditions amenable to treatment. Thus, for example, coincident retinal detachment and cataract would be categorised as posterior segment pathology.
Presenting visual acuity is the habitual (with correction if available) visual acuity of the individual. Blindness is reported as presenting visual acuity worse than 6/60 in the better eye (functional blindness), and also at the 3/60 level. Low vision is defined as presenting visual acuity of 6/60 or better, but worse than 6/18, in the better eye. Vision impairment is defined as presenting visual acuity worse than 6/18 in the better eye (the combination of low vision and functional blindness).
Data were entered into a specifically designed database daily during the survey. Prior to analysis, missing data and outliers were checked against the survey forms.
Point prevalence estimates of visual acuity and their 95% CIs were calculated. Definitive data from the 2004 national census became available after the survey.13 This was used to adjust the prevalence estimates for distribution of age and gender. These standardised estimates were then used to extrapolate to the entire country.
The strength of association of demographic factors with vision impairment was analysed and described using odds ratios (ORs) and their 95% CIs. Major causes of vision impairment were compared between demographic factors using Fisher's exact test. A p value 0.05, non‐overlapping 95% CIs or OR CIs without the value 1 were considered to be significant. SPSS V.12 was used for data analysis.
Approval for this study and its methodology was sought from the Timor‐Leste Ministry of Health and granted by the Minister of Health. Written consent was obtained from the village chiefs prior to the commencement of the survey in each cluster. Informed consent was obtained from each participant prior to all data collection and examinations. Communications occurred in Tetum, the local language or another, depending on the participant's preference.
All participants were advised of the availability of permanent eye‐care services in Dili and Bobonaro. Transport and accommodation arrangements were made for those with treatable causes of low vision and blindness who were willing to be referred to the Dili National Hospital.
One selected cluster proved to be inaccessible, with no ready adjacent substitute. Of the 1470 people enumerated, 1414 (96.2%) were examined. In all, 26 participants were not available during the survey and 30 refused participation. Data from these 56 people were not considered in the analysis.
The mean (SD) age of participants was 54.9 (20.1) years. In all, 49% were women. Participants were more likely to be younger, married, illiterate and not in paid employment (table 11).
The crude prevalence of functional blindness for those aged 40 years was 7.7% (109/1414), and of low vision was 15.5% (219/1414). Increased probabilities of both were associated with increasing age, illiteracy, not being married, farming and unemployment (compared with paid employment) and rural domicile (table 11).
The age, gender and domicile‐adjusted prevalence of functional blindness was 7.4% (95% CI 6.1 to 8.8). No gender difference was detected in the age‐adjusted prevalence of blindness within the urban, rural or overall samples. Nor were differences apparent in male, female and overall age‐adjusted prevalences between the urban, rural and overall populations (table 22).
The age, gender and domicile‐adjusted prevalence of low vision was 17.7% (95% CI 15.7 to 19.7). No gender difference was found within the urban, rural or overall groups. However, the age‐adjusted prevalence of low vision was significantly higher in the rural area compared with the urban area for men, women and both combined (table 22).
Using the definition of blindness as presenting acuity worse than 3/60, the crude prevalence was 4.7%, and the age, gender and domicile‐adjusted prevalence was 4.1% (95% CI 3.1 to 5.1).
In the survey sample, cataract was the most common cause of blindness, accounting for 76.1% (95% CI 68.1 to 84.1) of cases at the 6/60 level (table 33)) and 78.8% (95% CI 68.9 to 88.7) using the 3/60 definition. Of the 83 cases worse than 6/60, 4.8% had no perception of light, 16.9% perceived light, 41.0% counted fingers and 37.3% read 3/60. Surgery‐related and posterior segment causes were the next most prevalent causes of blindness (both 6.4%). No difference was demonstrated between urban and rural residence for cataract (p=0.818) or refractive error (p=1.0) as a cause of blindness.
Uncorrected refractive error was the most frequent cause of low vision in the survey sample (table 44).). It was responsible for 73.1% (95% CI 67.2 to 78.9) of cases overall, and was a more probable cause among rural residents than among urban residents (p<0.001). Cataract accounted for 25.1% (95% CI 19.4 to 30.9) of low vision, and was a more common cause for urban dwellers than for rural dwellers (p<0.001).
The age, gender and domicile‐adjusted prevalence of cataract‐induced functional blindness was 5.4% (95% CI 4.2 to 6.6). For low vision and vision impairment caused by refractive error, this was 14.4% (95% CI 12.5 to 16.2) and 14.9% (95% CI 13.1 to 16.8), respectively.
The age, gender and domicile‐adjusted prevalence of functional blindness in the sample population was 7.4%. That for low vision was 17.7%. Direct comparison with other prevalence surveys from the region is difficult because of differences in definitions and methodology, including age stratification. However, a similar or higher prevalence of blindness has been reported for a similar age group in Indonesia,14 Papua New Guinea15 and parts of India.16,17,18,19 A lower prevalence has been described for Bangladesh,20 China,21,22 Malaysia,23 Mongolia,24 Nepal25,26 and Tibet.27 The overall estimate for South East Asia is also lower.3 The relatively high prevalence of blindness in Timor‐Leste may, in part, be due to the historical lack of permanent eye‐care services, and low utilisation where they have been available.28
The risk factors associated with blindness and low vision in the current survey were in accordance with those in other studies: increasing age,16,18,19,20,22,26 illiteracy or no formal education16,18,20,22,26 and rural residence.16,19,20 However, despite estimates that women are 1.4–2.2 times more likely to be vision impaired in all regions of the world,3,29 this study did not identify female gender as a risk factor. This apparent absence of such gender disparity in Timor‐Leste may be because of the insufficient power of the sample size. Or, perhaps the past profound lack of services has equally disadvantaged both men and women. Future gender analyses will be telling, given that the current (2004/5) male to female eye‐care service utilisation ratio is 60:40.2
As with other studies from other countries in the region,14,15,16,17,18,19,20,21,22,23,25,26,27 cataract was found to be the most prevalent cause of blindness in Timor‐Leste. Similarly, other studies using presenting rather than best‐corrected vision14,16,17,18,19,20,21,22,23,26 also found refractive error to be the most frequent cause of low vision.
Extrapolating the age, gender and domicile‐adjusted blindness and low vision prevalences to the estimated 20% of Timor‐Leste's population aged 40 years13 yielded a mean (SD) of 46500 (6500) people vision impaired. Of these, 13500 (2500) are functionally blind. Uncorrected refractive error is responsible for 27500 (3500) people with vision impairment, and cataract for 16000 (3000). Cataract accounts for 10000 (2000) of those who are functionally blind. However, caution should be exercised because, owing to the resource and logistic restrictions of the survey, the rural sample may not be representative of the entire rural population. A greater prevalence of vision impairment may be expected in the more remote communities in the eastern part of the country.
Use of a standardised validated population‐based methodology, together with good observer agreement and a high response rate, is the strength of this study.11 However, pupil dilatation and more sophisticated examination techniques and grading criteria could have resulted in a more accurate diagnosis of the cause of reduced vision. This is particularly relevant to the potential overdiagnosis of cataract at 6/24, 6/36 and even 6/60 levels, and needs to be considered when these data are used for planning. This potential loss of accuracy was accepted because the examination was shorter, with reduced participant discomfort. Also, the utility of the results may have been improved by recording all conditions contributing to vision loss in a single eye, rather than just the validated RACSS convention of using the single most treatable condition.
Although a more rigorous, comprehensive and detailed survey may have provided more accurate data, this simple rapid assessment methodology has produced adequate information with which to start planning eye‐care services. It would seem that a national eye health strategy initially concentrating on high volume, high‐quality cataract surgery and the correction of refractive error would go far to relieve the burden of vision impairment in Timor‐Leste.
We thank the Timor‐Leste Ministry of Health, District Health Management Bobonaro, and the survey teams of the Dili and Bobonaro districts for their help.
RACSS - Rapid Assessment of Cataract Surgical Services
Funding: This work was, in part, financially supported by the Australian Federal Government through the Co‐operative Research Centres Program (Vision CRC).
Competing interests: The International Centre for Eyecare Education distributes and receives financial benefit from the sale of ready‐made spectacles in developing countries. However, the authors, who were previously employees of this organisation, have no personal pecuniary interest in the manufacture, distribution or sale of spectacles. The authors have no pecuniary interest in any product mentioned or in the outcome of this survey.