In Burundi, only one study on this subject was conducted, and this was undertaken over a decade ago. It entailed an RAAB study carried out in 2012 among participants aged 50 years and older. Out of 3684 participants, a prevalence of VI of 1.7% was found based on presenting VA (PVA).¹⁴ This was relatively lower compared to other RAAB studies in East Africa, specifically Rwanda (4.4% and 5.3%)^15,16 Kandeke et al. suggested that the low prevalence could be because of positive health-seeking behaviour among Burundians and increased efforts by the government. They also speculated that the vulnerable, especially people with blindness and VI, may have succumbed to the conflicts that happened in Burundi in the past.¹⁴ More women were reported to have severe VI than men (0.7% compared to 0.5% for men). However, a higher prevalence of moderate VI was reported in men (2% compared to 1.4% in women).

An RAAB study conducted in Eritrea, a smaller East African country, sampled a total of 3163 participants who were 50 years and older, with 41.9% being male. The reported prevalence of moderate VI (PVA less than 6/18 but better than or equal to 6/60) was 10.5%, while the prevalence of severe VI (PVA less than 6/60 but better than or equal to 3/60) was 3%. In terms of gender, more men than women had VI. For example, 13.7% of men had moderate VI (PVA < 6/18 but better than or equal to 6/60), while 10.8% of women had moderate VI. Furthermore, 4.1% of men had severe VI compared with 2.7% of women.¹⁷

Eight of the 11 studies that reported on VI in Ethiopia were carried out on paediatric populations and reported a prevalence of VI between 1.8 and 11.7%. Notably, the age classification for the paediatric population differed across the studies. While some researchers limited their definition of the paediatric population to school-going children (5–18 years) as defined by the United Nations (UN),¹⁸ this review found that the age ranges considered included 5 years to 15 years,¹⁹ 5 years to 16 years,²⁰ 7 years to 17 years,²¹ 11 years to 16 years²² and 8 years to 18 years.²³ Asferaw²⁴ was relatively more versatile in his classification of a paediatric population as 6 years to 25 years. Comparatively higher prevalence of VI in adult populations (> 18 years) was reported by the remaining three studies, which ranged between 16.8% and 36.5%.^25,26,27 All studies based the diagnosis of VI on PVA. None of the studies in adult populations in Ethiopia reported on the prevalence of VI stratified by gender. In the paediatric populations, a higher prevalence of VI was reported in males (52% males compared to 48% females)²² (52.9% males compared to 47.1% females).²³

Two cross-sectional studies were carried out among paediatric populations in Kenya. Barasa et al.²⁸ reported that the prevalence of VI among 5- to 22-year-olds was 4.77%, and this is almost double that reported by Muma and Obonyo²⁹ of 2.4% in a slightly younger population (5- to 16-year-olds). The study by Barasa et al. was conducted in an economically less endowed location with limited access to healthcare. The authors speculated that this may have accounted for the higher prevalence of VI noted. In both studies, the prevalence of VI was noted to be higher in females than in males.^28,29 A higher prevalence of VI was recorded in the adult population, which could be as a result of a comparatively stricter visual acuity (VA) threshold for the diagnosis of VI, that is VA poorer than 6/12.^30,31 Mathenge et al. reported the prevalence of mild, moderate and severe vision impairment (MMSVI) among persons 50 years and older to be 5.1%, 8.1% and 0.4%, respectively.³⁰ Similarly, Signes-Soler et al. reported the prevalence of VI based on PVA to be 6.3% in an adult population older than 18 years of age.³⁰ In terms of gender, Mathenge et al. reported a higher prevalence of MSVI in men than in women (8.3% and 0.5% in men compared to 7.9% and 0.3% in women, respectively); however, a higher prevalence of mild VI was reported in women (5.2% compared to 5.0% in men).³⁰

Among adults, an RAAB study was conducted in the western province of Rwanda involving 2206 participants aged 50 years and above, with the prevalence of VI reported as 5.3%.¹⁶ A similar RAAB study involving all provinces of Rwanda, also by Mathenge et al., almost a decade later, among 5065 participants aged 50 years and above, reported a relatively reduced prevalence of VI of 4.4% and blindness of 1.1%.¹⁵ The reduction in VI, according to the study, was because of the increased efforts by the government and donor partners in tackling VI in the country. Lower values were also noted in the studies in younger populations. A prevalence of 1.6% was noted in children 7 years and younger,³² while more recently, El-Khoury et al. in a cross-sectional retrospective hospital-based study, reported the prevalence of severe VI in a sample younger than 18 years to be 4.2%.³³ In terms of gender, Mathenge et al.¹⁵ reported that more women than men had moderate VI (5.8% of women compared with 4.7% of men); however, the prevalence of severe VI was higher in men (1.5% of men compared with 1.1% of women). Almost a decade later, the gender trend of VI had not changed, although with reduced prevalence values. The prevalence of moderate VI was 2.1% among women and 2.0% among men. The prevalence of severe VI was 0.7% for both men and women. However, when all cases of VI were considered, the prevalence of VI was marginally higher in women (4.8%) than in men (4.0%).¹⁶

Alrasheed et al. conducted the only study in Sudan, and reported the prevalence of VI in a paediatric population (6–15 years) to be 4.4%, 6.4% and 1.3%, respectively, based on PVA, uncorrected VA and best-corrected VA, in a school-based cross-sectional study among 1678 pupils. The study revealed that there was no statistically significant difference in the prevalence of VI based on gender.³⁴

In the period under review, two studies were published from Tanzania on VI: a community-based cross-sectional study³⁵ and an RAAB study.³⁶ Mashayo et al. reported the prevalence of VI in the general population (over 15 years of age) to be 10.4%,³⁵ while Habiyakire et al. reported the prevalence of VI in people 50 years and older to be 5.4%.³⁶ The significant difference between the prevalence reported by the two studies could be attributed to the age profiles of the study populations involved in each study. While Habiyakire et al.³⁶ reported a lower prevalence, they did not account for persons younger than 50 years. Also, according to Mashayo et al.,³⁵ the study was conducted in a ‘hard-to-reach’ district with less access to eye care. Habiyakire et al. reported almost similar estimates of VI between men and women (2.7% men versus 2.2% women).

In a cross-sectional study among 318 paediatric participants (< 18 years of age), and based in a children’s referral hospital in Uganda, Kinengyere et al. found a prevalence of VI of 42.1%.³⁷ All other relevant studies in Uganda were on adult populations and reported the prevalence of VI to range between 6.1% and 32.1%.^38,39,40,41 Generally, the prevalence of VI reported in Uganda is higher compared with the other East African countries. Furthermore, the aetiology of the study populations may be a contributing factor, as all prevalence studies in Uganda were hospital-based. This may limit the variety in the sample, as most people presenting to hospitals are usually aware they are visually impaired and in need of health care.

In the period under review, the major causes of VI reported in East Africa were URE, cataracts, glaucoma, corneal opacities, retinal diseases (including diabetic retinopathy, hypertensive retinopathy, age-related macular degeneration and macular holes) and amblyopia. Other causes of VI included nystagmus, cortical blindness, vitamin A deficiency and hereditary diseases.^{14,15,16,17,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43} Uncorrected refractive errors accounted for between 1.7%²⁴ and 94%²¹ of cases in this narrative review. Cataracts contributed between 2.7%⁴³ and 54.8%³⁶ of cases, while corneal opacity accounted for between 0.6%³¹ and 65%²⁴ of VI in East Africa.

In the paediatric populations included in the studies in East Africa, the major causes of VI reported were URE, amblyopia and cataracts. The prevalence of URE as the cause of VI in paediatric populations varied from as low as 1.7%²⁴ to 94%.²¹ Among those with VI in East Africa, the prevalence of amblyopia ranged from 2.2%²⁴ to 22%²⁹ and the prevalence of cataract reported ranged between 2.7%⁴³ and 18%.³³ In contrast, the underlying conditions that contributed to the prevalence of VI in the adult populations in East Africa were untreated cataracts, glaucoma, URE and retinal diseases. The prevalence of cataracts in the adult population in East Africa with VI ranged from 8.1%³¹ to 54.8%.³⁶ Glaucoma contributed between 0.9%^15,31 and 36.8%,²⁶ while the prevalence of uncorrected refractive error accounted for 7.95%²⁶ to 67%¹⁴ of VI in adults across East Africa.

Overall, the prevalence of VI in East Africa ranged from 1.6%³² to 42.1%³⁷ during the chosen study period. Compared to the crude global prevalence range of mild VI (0.88% – 4.77%) and moderate to severe VI (1.34% – 4.89%),⁴ and a similar prevalence of MSVI reported by Naidoo et al.⁴⁶ of 4.0% in sub-Saharan Africa in 2010, the prevalence of VI in East Africa had a relatively wider range (1.6% – 42.1%) and was higher. These differences could be because of the limited studies from East Africa in the period under review and varying criteria used by the studies included in the review. In a systematic review, Bourne et al.⁴ reviewed 512 studies from 98 countries (from January 1980 to October 2018), while the current review studied 29 studies from only eight countries and for a shorter review period. Similarly, Naidoo et al.⁴⁶ conducted a systematic review of 252 population-based studies from sub-Saharan Africa from January 1990 to January 2012. Furthermore, of the 29 studies in East Africa in the current review, 18 were published from 2017 onwards, compared with the earlier studies included in the reviews by Bourne et al. and Naidoo et al. The review by Naidoo et al. used stricter inclusion criteria by only including population-based studies. Studies that used hospital-based data, blindness registries, and self-reported vision status, which may create a bias when estimating prevalence, were excluded from their review. In their systematic review, Bourne et al.⁴ included only population-representative studies. The current review was more versatile and included all studies on VI published in East Africa in the given period. Hospital-based studies and some population-based studies have an inherent limitation in that participants are usually people with known VI seeking medical attention. This was noted in this review, where the hospital-based studies that reported the prevalence of VI from Uganda noted a higher prevalence of VI. Additionally, some population-based studies, which are not population-representative, usually carried out as a community outreach on a walk-in basis, will report a higher prevalence of VI.

A significant aspect of VI is visual field loss, as defined by the ICD-11. However, all the studies reviewed reported on the prevalence of VI based solely on VA, without reflection on visual field (VF). Thus, reporting on the prevalence of VI based on VA only may not represent a comprehensive picture of the prevalence of VI. The seeming neglect of consideration of VF by all studies in reporting VI could be linked to logistical constraints. Automated VF tests (VFT) (perimetry) are usually bulky and require electricity, making them cumbersome for use in mobile clinics. Additionally, it is comparatively more time-consuming than recording VA, which reduces the feasibility of the study. Furthermore, visual field loss is usually secondary to other primary conditions such as glaucoma; most researchers therefore rely on the primary diagnosis, namely, VA measurements, which are relatively less cumbersome to identify and enumerate. We recommend the inclusion of confrontational VFT (CVFT) in the definition and classification of VI. The CVFT, even though a gross test, will contribute to a better understanding of the extent of VI. Automated confrontation testing devices, which are portable and have higher sensitivity, can be considered for hard-to-reach areas without access to standard VFTs.^47,48 It is simple, easy to administer and can be administered with little training.

Globally, the leading causes of VI have been identified as URE, cataracts, glaucoma, age-related macular degeneration, myopic macular degeneration and diabetic retinopathy.^3,4,49 These findings are supported by this narrative review, where URE, untreated cataracts and corneal opacity were also noted to be the major causes of VI reported in those with VI in East Africa. Alrasheed et al. suggest that the increased use of technology (urbanisation) and being indoors may be contributing to the rise in the prevalence of refractive error, especially in children.³⁴ Bourne et al., agree with that assertion and add that an increasing population and an ageing population have increased the need for eye care services, which have been insufficient to cater for the rising demand.⁴ Therefore, URE as the second leading cause of VI in adult populations may be a spillover from unmet demands in childhood populations. The finding of cataracts and corneal opacities as major causes of VI in East Africa is also similar to that noted by Naidoo et al. in sub-Saharan Africa, which may relate to certain conditions like trachoma being endemic in the continent, as well as limited access to health care, including cataract surgery, in developing countries.^46,50

One noteworthy observation from this review is the higher prevalence of VI noted in paediatric populations, usually driven by a higher prevalence of URE. The reason that may account for the difference in the causes of VI between paediatric and adult populations could be summarised as follows: children may have limited awareness of poor vision, coupled with less access to eye care, consequently, most cases of URE will go unnoticed in children because they are unaware that their vision is impaired.⁵¹ On the other hand, cataracts and retinal diseases are most prevalent in the adult population, probably because of metabolic changes that occur in the body as a result of ageing.⁵²

Of the causes of VI reported, the two most prevalent underlying conditions, URE and cataracts, are treatable.^4,15 Uncorrected refractive error can be diagnosed efficiently and often treated by dispensing spectacles or contact lenses or via refractive surgery. Cataracts can be managed surgically. These simple, yet effective treatment interventions offer great opportunities to reduce the global burden of VI. However, access to and affordability of these interventions often pose the greatest challenge, particularly in developing countries like those in East Africa⁴. Thus, increased vision screening to identify URE and treatment using spectacles are necessary to stem VI. Additionally, improved cataract surgical rates (CSR) are important to address the reversible prevalence of cataracts, with alleviation campaigns such as community outreaches and eye camps to tackle untreated cataracts in East Africa.

Analysis of VI trends based on gender in this review revealed that VI is generally noted to be more common in women than in men.^15,19,37,42 This finding is consistent with that reported by Naidoo et al. in sub-Saharan Africa (3.8% in men [3.1% – 4.7%] and 4.2% for women [3.6% – 5.3%]).⁴⁶ The reasons attributed to this include: women live longer than men and therefore have a higher risk of age-related diseases than men; inequalities in accessing health care, as men have a higher capacity to access health care than women; or a general reflection of global demography, where there are more women than men in the world.^9,44,53 However, two studies in the current review reported contrary results with a comparatively higher prevalence of VI in men than in women, which may be attributed to the demography of the study populations involved in those studies.^20,23 Both studies were carried out in basic schools in Ethiopia, and a basic school enrolment report revealed more men than women enrolled in schools in Ethiopia in the period under review.⁵⁴