Prevalence of visual impairment and refractive error in school-aged children in South Darfur State of Sudan

Global estimates indicate that there are around 19 million visually impaired children worldwide; of these, 1.4m are blind and 17.5m have low vision with many of them living in Africa.1 In a recent classification by the World Health Organization (WHO), the major cause of worldwide vision impairment was uncorrected refractive error (RE).1 Current worldwide estimate indicates that more than 90% of people with uncorrected RE live in rural and developing nations.2 WHO defines visual impairment (VI) in children as presenting visual acuity (VA) less than 6/12 in the better eye.3 However, uncorrected visual acuity (UVA) is defined as VA less than 6/12 in one or both eyes. Background: Global estimates suggest there are almost 19 million visually impaired children worldwide, the major cause being uncorrected refractive error (RE).


Introduction
Global estimates indicate that there are around 19 million visually impaired children worldwide; of these, 1.4m are blind and 17.5m have low vision with many of them living in Africa. 1 In a recent classification by the World Health Organization (WHO), the major cause of worldwide vision impairment was uncorrected refractive error (RE). 1 Current worldwide estimate indicates that more than 90% of people with uncorrected RE live in rural and developing nations. 2 WHO defines visual impairment (VI) in children as presenting visual acuity (VA) less than 6/12 in the better eye. 3 However, uncorrected visual acuity (UVA) is defined as VA less than 6/12 in one or both eyes.VI among children in poor countries is one of the priorities of many eye health programmes including the Vision 2020: the Right to Sight initiative. 3Children presenting with VI means that the impact of such impairment is perhaps much more critical because of the potentially longer duration of their lives as compared to older people; however, some studies have estimated that almost 60% of children die within 1 year of becoming blind. 3,4A population-based study on VI and RE in school-aged children that began in 1998, using the same protocol, Refractive Error Study in Children (RESC), was conducted on populations with different racial backgrounds and environments. 5This RESC protocol was developed by the WHO in collaboration with, and under financial support from the National Eye Institute, National Institutes of Health (NIH) and the United States. 6Published studies using this protocol have been conducted in different countries including South Africa, 7 China, 8 India, 9 Brazil 10 and Saudi Arabia. 11These studies have shown that the prevalence of VI among schoolaged children in South Africa is 2.1%, 7 in India 6.4% 9 and 2.6% in Brazil.The prevalence of myopia, hypermetropia and astigmatism among African school-aged children in South Africa were 4.0%, 2.6% and 14.6%, respectively. 7Uncorrected RE drives children and adults further into poverty by limiting their opportunities to education and employment and this could seriously affect their quality of life and productivity. 12fter the 1990s, two papers from Australia and India emphasised the fact that uncorrected RE was a major cause of VI in children. 13The WHO and the International Agency for Prevention of Blindness have included uncorrected RE to the prevention of blindness agenda and have developed strategies for the elimination of this simple yet avoidable cause of VI. 13 Sudan is the second largest country in Africa geographically covering an area of about 1.8 million square kilometres, with an estimated population of more than 39m inhabitants. 14udan recently reached 20% inflation with a shrinking economy and is facing American sanctions.In terms of the Human Development Index, Sudan is ranked as a lowincome country. 14Health indicators such as the under-5 mortality rate are 78 per 1000 live births and the average life expectancy at birth is predicted to be 60 years. 15The prevalence of VI and blindness among internally displaced children in Khartoum was 5.5% and 2.7%, respectively. 16ccording to the Vision 2020 programme in Sudan, 17 the prevalence of blindness in the Northern States, excluding Darfur, has reduced from 1.5% in 2003 to 1.0% in 2010. 17outh Darfur is located in the western part of Sudan and is one of the five states that comprise the region of Darfur. 18outh Darfur is the most populous state with a population of approximately 4.45m people. 19The majority (62.7%) of people in this state live below the poverty line according to the Household Survey 2009. 20No studies have been conducted to estimate the prevalence of VI and RE in school-aged children in South Darfur State of Sudan.The aim of the present study was therefore to determine the prevalence of VI and RE in this population group.The study objectives were to estimate the main causes of VI and types of RE and to determine the difference in prevalence between gender, age and school-grade levels.

Study design
This was a cross-sectional, school-based study of VI and RE among children from the South Darfur State of Sudan.The research focused on primary schoolchildren in the 6-to 15year age range.According to the Ministry of Education, the overall number of learners enrolled in primary schools in South Darfur during 2014-2015 was 338 068 school-aged children consisting of 187 270 boys and 150 798 girls, from public and private schools. 21An RESC protocol was implemented in order to determine the prevalence of VI and RE in these children.The protocol defines RE when carried out by cycloplegic auorefractometer as follows: (1) myopia at least -0.5 D in one or both eyes, (2) hypermetropia at least 2 D in one or both eyes and (3) astigmatism at 0.75 D or more cylindrical refraction. 22

Inclusion criteria
Children aged 6-15 years and who are able to provide parental consent and all children attending the school on the examination days were included in the study.

Exclusion criteria
The exclusion criteria included those whose age did not correspond to the defined age group and those unable to provide parental consent.

Sample selection
The study sample was selected through stratified multistage cluster sampling.Assuming a prevalence of RE (P) = 5% and the worst acceptable prevalence (P) = 4%, according to the estimated prevalence of RE in Africa (5%), South Africa (4%) 7 and Kenya (5.1%). 23The sample size was calculated using the sample size calculation for eye survey formula 24 : and where n = minimum sample size required (approximate); N = total number of schoolchildren in the South Darfur State, in the academic year 2015 = 338 068 with Z = value of z statistic at 95% confidence level = 1.96;P = assumed prevalence of children with RE = 5%; Q = 100% -P = 95%; E = maximum acceptable random sampling error = 1.5%;W = the likely design effect = 2 was considered as we employed the cluster random sampling method.The value 3.8416 is 1.96² corresponding to Z at 95 probability of not exceeding E. In addition, the sample size was adjusted for an anticipated 10% absenteeism and non-participation rate.The final sample size was 1775 schoolchildren where eight schools (four for males and four for females) were randomly selected from the sampling frame of 21 districts of the South Darfur State.The study sample consisted of six public schools and two private schools randomly selected according to the list obtained from Ministry of Education.Within each school, the class levels represented the third stage of the sampling unit.One class from each grade (1-8) with a minimum cluster size of 28 children was randomly chosen.

Recruitment of the schoolchildren
Each school selected for the study sample was visited by the principal investigator to explain the purpose of the study to the school administration, and schoolchildren were given requests for consent and permission forms for their parents to allow them to participate in the study.Parents were asked to sign the consent letters if they agreed to allow their children to participate in the study; thereafter, such children were invited to take part in the study.

Training of the assistant researchers
Optometric research assistants with at least 3 years of experience in clinical optometry were recruited to assist with data collection.The research assistants underwent intensive training in the study protocol procedures.The principal investigator explained and demonstrated these methods in five sessions.

Pilot study
A pilot study was conducted outside the main area of study on 100 children not included in the study sample.The data collected were captured and analysed using the Statistical Package for Social Science (SPSS) Version 22.The results indicated that, some children in the vision examination reported that they had a problem in their vision but after detailed assessment of refraction, together with outer eye and inner eye examination, the results revealed that there was no abnormality in their eyes.In others, their vision examination result was normal after examination of refraction, but the outer eye and/or inner eye results demonstrated that they did in fact have eye problems.This was because of a misunderstanding of the vision examination by schoolchildren or some of them did this to get free medication and spectacles.Because the principal investigator indicated in the child's consent form that any participants, with eye problems, would be provided free spectacles and eye drops, this could have contributed to this anomaly.Therefore, to overcome this problem in the main study, the principal investigator rechecked the VA for each child after examinations to ensure the accuracy of the measurement of VA.

Ethical and legal consideration
Ethical permission for conducting the study was obtained from the University of KwaZulu-Natal's Biomedical Research Ethics committee (ref: BE247/14) and the National Research Ethics Review Committee in Sudan.Permission was also obtained from the South Darfur authorities in Sudan to undertake the research at their facilities.Informed consent was obtained from all participants included in the sample study to facilitate a better understanding of conditions of involvement in the study.The research was conducted in accordance with the Declaration of Helsinki.

Clinical examinations
Examination procedures employed a modified RESC protocol.Distance VA was assessed using Snellen Tumbling E-chart with E's of standard size at a 6-meter distance.
Children with VA ≤ 6/12 were tested by pinhole and if their VA improved, they underwent cycloplegic refractions.All the schoolchildren underwent a penlight and low-power hand magnifier examination to rule out any anteriorsegment abnormalities in the following parts of the eye: eyelids, conjunctiva, cornea, the pupil and pupillary reflex reaction.A cover test was conducted for phorias or tropias and deviations were measured using the corneal light reflex (Hirschberg test) and the Prism Cover Test at distance and near fixation, respectively.The learners underwent motility tests to assess eye muscle function.Subjective refractions were performed by achieving best-corrected VA; while cycloplegic refractions were done for the learners where vision improved with pinhole.First one drop of ophthalmic topical anaesthetic was instilled in each eye, followed by a waiting period of 2 min to achieve ocular surface anaesthesia.Thereafter, children received two drops of 1% cyclopentolate administrated 5 min apart to each eye.If cycloplegia was not completed, then a third drop was administered as required.Cycloplegia was considered to be completed if the pupil dilated to 6 mm or greater and light reflex was absent.
After cycloplegia, optometrists refracted learners, regardless of VA: using a Hand Held Auto Refractor/Keratometer (Retinomax K-plus 3).The children with VA ≤ 6/12 whose vision did not improve by pinhole test had ocular and fundus examination by direct ophthalmoscopy and any abnormal findings were recorded.Learners presenting with eye organic defects were referred for further treatment (Figure 1).

Data analysis
Data forms were reviewed for accuracy and completeness before data capture.The data entry was performed by the principal investigator using SPSS (Version 22), the data were checked for data entry errors and/or missing values before data analysis.The data for each subject were analysed descriptively using standard deviations, modes and percentages.The relationship between measures was determined using correlation, cross-tabulations and chisquared analysis.The analysis of variance (ANOVA) was used for statistical comparison of means.For all statistical determinations, significance levels were established at p = 0.05.

Study population
Of the 1775 school-aged children selected to participate in the study, 1678 children presented at schools on the examination days resulting in a participation rate of 94.5%.The records of 12 students who were actually older than 15 years were eventually excluded from the study; thus, the results of 1666 children were analysed as discussed below.

Socio-demographic characteristics of participants
The age of the children ranged from 6 to 15 years with mean age of 10.8 ± 2.8 years (s.d.) while the median age was 11.00 years and modal age was 14 years.There were 839 female children representing 50.4% and 827 male children representing 49.6% of the sample.The mean ages and standard deviations of female and male children were 11.00 ± 2.7 years (s.d.) and 11.00 ± 2.9 years (s.d.), respectively.The age groups most represented were 14 and 12 years at 12.4% and 11.5%, respectively.The groups least represented were 6 years at 5.6%.The differences between the mean ages of male and female children was significant (ANOVA: F = 5.71, p = 0.017).Grades 1 and 2 had the highest proportion of respondents at 14.5% and 13.7%, respectively.This was followed by grades 5 and 7 at 13.20% and 12.30%, respectively, and the lowest proportion was grade 6 at 10.7%.The differences in the mean ages of the children in the classrooms were significant (ANOVA: F = 1219.092,p = 0.00).

Distribution of signs and symptoms among school-aged children
A total of 1137 (68.2%) of respondents did not complain of any ocular symptoms.In all, 180 (10.8%) complained of blurred vision, followed by 125 (  1).

Prevalence of visual impairment
The findings of VI among 1666 school-aged children are shown in Table 2.The prevalence of VI with the uncorrected VA in the better eye ≤ 6/12 was 74 (4.4% [95% CI, 2.9-5.9]).None of these children were found to be wearing eye glasses.The prevalence of VI increased according to age.Younger ages were associated with lower prevalence: 3.0% (95% CI, 1.5-4.5) in children aged 6-7 years, 5.2% (95% CI, 3.7-6.7) in those aged 10-11 years and 6.4% (95% CI, 4.9-7.9) in those aged 12-13 years.However, there was no marked difference

Binocular motor function
Heterotropia or heterophoria were found in 852 (51.1%) children at the 50-cm fixation distance and in 95 children (5.7%) at the 6-m fixation distance.Most of the children had exophoria at near 781 (46.9%).Tropia was observed in five (0.30%) children; three children had esotropia and two had exotropia.3).

Prevalence of refractive error
The prevalence of RE was estimated based on presenting VA (UVA of 6/9 or worse in one or both eyes).Myopia was defined as the spherical equivalent of at least |-0.http://www.avehjournal.orgOpen Access higher among 13 year olds (4.2%) followed by 15 year olds (2.8%) and lower for the ages 10, 8, and 6 years, which was 0.6%, 0.6% and 1.1%, respectively.The prevalence of myopia compared with emmetropia was statistical significant (p = 0.023) and associated with increasing grade level, whereas it was not significant (p = 0.073) with the children's age.
The prevalence of the children who needed refractive correction in one or both eyes was 187 (11.2%) but no child presented with spectacles.The need for spectacles was the highest for both genders in the higher-grade levels.

Posterior-segment examination
Posterior

The principal causes of visual impairment
The causes of UVA of 6/12 or worse at least in one eye are presented in Table 5. RE was the main cause of VI in 61 (57%) affected children.Retinal disorders (retinal degeneration and retinal detachment) were the cause of VI in 14 (13.1%)children.Amblyopia accounted for 6 (5.6%) children.Corneal opacity was the cause of vision impairment in one child (0.9%) and cataract was present in four (3.7%) children.
Reduced vision was unexplained in 11 (10.3%)children and other causes of VI accounted for 10 (9.3%) children.

Schoolchildren who received spectacles, eye drops or were referred
About 1410 (84.6%) school-aged children had normal vision and no ocular abnormalities and were not referred.The distributions of the remaining schoolchildren who received eye care or who were referred for further treatment were as follows: 216 (13.0%) schoolchildren had uncorrected RE, inflammatory and allergic eye conditions.The learners with uncorrected RE were given spectacles and the others were provided with eye drops.Forty (2.4%) learners presented with other eye diseases and were referred to an eye hospital for further treatment.

Discussion
Cycloplegic refractions were used to assess the RE of schoolchildren in this study as for previous studies by Naidoo et al. 7 in South Africa as well as Pi et al. 8 in China, Aldebasi 25 in Saudi Arabia and Paudel et al. 26 in Vietnam.The prevalence of presenting VI (UVA ≤ 6/12 in the better eye) of 4.4% is lower than 5.5% found in RESC studies by Zeidan et al. 16 in Khartoum, 6.9% found by Megbelayin and Asana 27 in Nigeria, 9.5% found by Mehari and Yimer 28 in rural central Ethiopia and 7.7% found by Pi et al. 29 in Western Chinese children aged 6-15 years.However, the prevalence from this study is higher compared to figures reported from other regions, which used the same protocol.Naidoo et al. 7 in southern Africa found a prevalence of 2.1% and Salomao et al. 10 in southern Latin America found a prevalence of 2.67%.However, the result of this study is similar to the prevalence of vision impairment among children in Western Africa at 4.5%. 30These results indicate that VI among school-aged children is a concern requiring urgent intervention.Thus, it is evident that lack of primary eye care to deal with childhood VI as well as lack of health education programmes to raise the awareness of the community about the consequences of childhood VI are the primary reasons for the prevalence of VI in schoolchildren.The study findings revealed no significant association between gender and prevalence of VI (p = 0.83), which was similar to studies by Megbelayin and Asana 28 in Nigeria and Pi et al. 29 in Western China who found no significant correlation between sex and prevalence of VI.However, our findings showed the prevalence of VI becomes higher with increasing age and grades (p = 0.00), in line with studies in Nigeria 29 and Western China. 29e prevalence of RE in either eye was 178 (11.2%), which is lower than that found by school-based studies conducted in similar age groups such as those by Paudel et al. 26 in Vietnam (21.4%),El Bayoumy et al. 31 in Egypt (22.1%) and Al Wadaani et al. 11 in Saudi Arabia (13.7%), but higher than other reports of similar age groups in Ethiopia at 9.0% by Yared et al. 32 and in China at 6.3% by Li et al. 33 The prevalence of RE in one or both eyes was found to increase from 6.6% in grade 2 to 12.7% in grade 8, which was statistically significant (p = 0.00).
The prevalence of RE among schoolchildren in this report was 11.2%, which is statistically significant for age and grade levels of the children (p = 0.00).However, it was lower than that reported from similar age groups in the Cape Coast of Ghana (25.6% prevalence 30 ), Vietnam (16.3% 26 ), Egypt (22.1% 31 ) and Qassim Province primary schoolchildren in Saudi Arabia (at 16.3% 25 ).However, the prevalence was higher than that found in RESC studies in South Africa at 2.4%, 7 Iran at 3.5%, 34 Nepal at 8.6% 35 and Nigeria at 2.2%. 36On the other hand, the prevalence of RE in this study was similar to the 11.6% 37 for children in Uganda as well as 13.3% 30 among schoolchildren in the central region of Ghana and 13.7% 25 for primary schoolchildren in Saudi Arabia.The prevalence of RE in children in the present study is also higher than that reported from other studies conducted in Africa that used the RESC protocol.A possible reason for this was that this study used the definition of RE, as per the original protocol, which defined RE's detected when performed by cycloplegic autorefractometer. 5However, most studies conducted in African countries did not follow the RESC definition for RE and cycloplegic autorefractions were not used in the majority of the studies, except those conducted by Naidoo et al. 7 The prevalence of myopia (spherical equivalent RE of -0.5 D or more in one or both eyes) was found in this study to be 6.8%, which was similar to that of 6.8% 38 in Chile, 6.0% 28 in Ethiopia and 5.8% 25 in Qassim Province, Saudi Arabia.However, it was relatively lower than that found in Ethiopia and Ghana, where the prevalence was 13.6% 32 and 14.1% 30 , respectively.Other studies reported a lower prevalence of myopia among school-aged children in Iran at 3.4% 34  The results of this study showed that schoolchildren in older ages and higher grades had a higher prevalence of myopia, which was similar to studies conducted in Vietnam 26 and Ethiopia. 32his associated increase in the prevalence of myopia with ages and grade levels may be because of decreased outdoor activity of many children and this has been reported as an issue in other papers.This study found that there was no significant difference in the prevalence of myopia between male and female children.
The prevalence of hypermetropia in this study of 1.9% was significantly lower than that reported in other studies of similar school-aged children conducted in other countries; for instance, 26.4% 32 in Ethiopia, 6.9% 40 in Ghana and 9.9% 11 in Saudi Arabia.However, it was similar to results found in a South African eye survey of 1.8% 7 and of school-aged children in rural Northern China at 1.6%. 33Our result of a trend of increasing prevalence of hypermetropia among children with increasing age is also in line with results obtained from South Africa 7 and China. 32The increase in prevalence of hypermetropia with increasing age might be because of a decrease in power of crystalline lens or loss of residual accommodation.The prevalence of hypermetropia was not significantly different between male and female children as also found, for example, by Shrestha et al. 35 among Nepalese school-aged children.
The prevalence of astigmatism in the current study was 2.5%.This is much lower than the findings by Ogbomo and Assien among Ghanaian children at 6.6% 30 , Khalaj et al. in Iran at 16.1% 41 and by Naidoo et al. 7 14.6% in South Africa.However, the prevalence of astigmatism found in this study was similar This study revealed that the prevalence of myopia in schoolaged children was slightly higher than that found in previous studies conducted among African school-aged children.However, the present study agrees with recent studies that there is a gradual increase in myopia over the last decade and this might be because of increased time spent indoors and lack of time outdoors as children are increasingly engaging in tasks using technology such as computers and mobile phones rather than outdoor activities.This trend of increasing myopia has been stated by Holden et al., 45 who reported that the prevalence of myopia worldwide was 22.9%.According to a systemic review of data from 145 studies, they predicted that this will increase to 49.8% by 2050.Our study revealed that almost two-thirds of the visually impaired children had uncorrected RE and had never received refractive correction.Naidoo et al. 46 have identified that uncorrected refractive error is a major challenge globally with 108m people worldwide suffering from VI because of uncorrected RE.Our study identified this as a major public health challenge in Sudan as well.Despite the presence of significant uncorrected RE of 187 (11.2%) among school-aged children in Sudan, they did not receive RE correction, suggesting that there are significant barriers for accessing RE correction as well as lack of knowledge about the effects of uncorrected RE.Therefore, health education programmes about the use of spectacles and consequences of uncorrected RE are important as well as developing a comprehensive child eye care plan to reduce the risk of uncorrected RE.

Limitations
The current study had some limitations.Firstly, not all the schools in the South Darfur State were included in the sampling frame and the schools in the camp of internally displaced people were not included in this study.Thus, our findings reflect only the prevalence of VI and RE among school-aged children studying at public and private schools in the South Darfur State.Secondly, we conducted all the clinical eye examinations in the schools to improve the participation rate.However, conditions such as illumination, ventilation and comfort were different from school to school.However, we did ensure that we chose similar environments to conduct the study.Thirdly, the number of schoolchildren decreased in higher-grade levels, which may introduce bias in the results of VI and RE, as the numbers were not uniform across all school grades.Finally, our study employed a modified RESC protocol.This involved minor changes in the instruments that were used for collecting the data as recommended by the original protocol.We used Snellen (Tumbling E-optotype), torch and magnifier as opposed to LogMAR charts and slit lamp as well as ophthalmoscopy as opposed to fundus biomicroscopy.

Conclusion
To our knowledge, no similar study has been conducted in this region; thus, the data obtained makes a valuable contribution to the understanding of eye health challenges in the South Darfur State of Sudan.The study concluded that uncorrected RE was a major cause of VI among school-aged children in this region.These findings indicate an urgent need for developing a comprehensive childhood eye care plan for delivering eye care services to school-aged children, through collaboration between government, private sectors, stakeholders and non-government organisations working in preventing avoidable childhood blindness and VI.This supports the need for regular vision screening programmes for the prevention of avoidable causes of vision.

FIGURE 1 :
FIGURE 1: Chart illustrating the procedures for assessing vision and detecting vision impairment.

TABLE 1 :
Distribution of uncorrected visual acuity for right, left and better eye and best-corrected visual acuity by percentage and confidence interval.

TABLE 2 :
Prevalence of visual impairment among school-aged children by age groups, gender and school grade.

TABLE 3 :
Distribution of normal and external eye disease by percentage and confidence interval.

TABLE 4 :
The prevalence of refractive error in one or both eyes by age, gender and school grade.
by Fotouhi et al. as well as in South Africa at 4.0% 7 by Naidoo et al. and in Southern India at 4.1% 39 by Dandona et al.

VA 6/12 or worse Children with VA 6/12 or worse in one or both eyes Prevalence in the population one or both eyes, % (95% confidence interval) Right eye Left eye
28 the 2.5%42reported in Riyadh by Al-Rowaily and 2.17%28reported in Ethiopia by Mehari and Yimer.Of the examined children, 0.3% had manifest strabismus (three children had esotropia and two had exotropia), and this was similar to findings by Wedner et al.43among schoolchildren in rural Tanzania, where the prevalence of strabismus was 0.5%.This is lower than the 1.3% found in South African children by Naidoo et al.7and 1.2% found among Iranian children by Jamali et al.44In this study, uncorrected RE was the most common cause of VI among school-aged children, which accounted for 57% of cases.This result is similar to that found in many studies, which used the RESC protocol in African school-aged children such as in South Africa (66.4%7) and in Ethiopia (65.9%28).However, this result is lower than that found in Asian children at 87.3%34reported from Iran by Fotouhi et al. and at 86.08%29inWesternChina by Pi et al.The findings are lower than those in Asian children.This could be because of genetic differences as well as the different lifestyles in terms of outdoor activities of the two groups.Thsecond cause of VI among children was retinal disorders at 13.1%, comparable to that found among South African children (10.9% 7 ) but lower than that reported from Ethiopia (3.5%28) and Iran (0.5%34). Thvalence of amblyopia was 5.6%, which was lower than that found among Ethiopian and South African children at 9.6% 28 and 9.4% 7 , respectively, and similar to that found in Ghana at 6.0%30.The prevalence of cataract was 3.7% relatively similar to that found in Ethiopia at 4.2%28and South Africa at 2.3% 7 by Naidoo et al.The high prevalence of the anterior-and posterior-segment abnormalities found among school-aged children in this study is a further reflection of the inadequacy of childhood eye care services in South Darfur State of Sudan.