Changes in the prevalence and causes of vision impairment and coverage of services between 2008 and 2024 in Koulikoro, Mali: results from two cross-sectional studies

Abstract

Background

While no national vision impairment (VI) data exist for Mali, population-based eye health surveys among people aged ≥50 y were conducted in several regions over the last two decades, including in the Koulikoro region in 2008. In 2024, a similar survey was repeated. This paper describes the changes in VI prevalence, causes and service coverage during the 16-y period.

Methods

The standard Rapid Assessment of Avoidable Blindness methodology was used with two-stage sampling to select a representative sample of 2600 adults (aged ≥50 y) in 2008 and 2040 in 2024. Ophthalmologists conducted a rapid vision examination on participants, assessing their presenting and best corrected visual acuity (VA), cause of any VI and cataract surgical status. Data were analysed using Stata.

Results

The age- and sex-adjusted prevalence of blindness decreased from 9.0% in 2008 to 4.4% in 2024, a −51.1% (p<0.001) decrease. Cataract surgical coverage at all levels of VA increased significantly from 2008 to 2024. The absolute increases were 29.3% (95% CI 26.5 to 32.1%; p<0.001), 28.5% (95% CI 25.6 to 31.4%; p<0.001) and 27.9% (95% CI 25.1 to 30.7%; p<0.001) for VA<3/60, VA<6/60 and VA<6/18, respectively.

Conclusion

The significant reduction in blindness and VI in Koulikoro from 2008 to 2024 was associated with improved cataract surgical services, which need to be further expanded.

Introduction

Vision impairment (VI) represents a major global public health concern and financial burden, particularly in low- and middle-income countries (LMICs).^1–4 The World Report on Vision 2019 estimated that at least 2.2 billion people worldwide have a vision impairment, of which at least 1 billion could have been prevented, or have not yet been addressed.⁵ Globally, approximately 43.3 million people are blind, and 295 million have a moderate to severe vision impairment.⁵ The VI burden has an annual cost of productivity loss estimated at US$411 billion.¹ At an individual level, VI perpetuates the cycle of poverty by substantially impacting quality of life, education, employment and overall well-being.⁵ The prevalence of VI in LMICs is estimated to be four times higher than in high-income regions.⁵ In 2020, the western sub-Saharan Africa region, where this study was set, was home to an estimated 48 million people living with a VI; the region has the highest age-standardised prevalence of blindness globally, estimated at 1.11% (95% CI 0.95 to 1.26%).¹

In Mali, there has never been a national study of blindness; however, several subnational studies focusing on specific eye diseases and population subgroups have been undertaken. In the late 1990s, the mapping of trachoma prevalence in some endemic areas of the country reported the prevalence of active trachoma as 23.1–46.7% and the prevalence of Trachomatous trichiasis (TT) at 25%.⁶ This evidence led to the implementation of the national trachoma control programme based on the SAFE (or Surgery, Antibiotics, Facial cleanliness and Environmental improvement) strategy and delivered in partnership with several non-governmental organisations (NGOs).⁶ In May 2023, the WHO validated elimination of trachoma as a public health problem in Mali.⁷

There have also been several VI surveys, which focused on a broader range of eye conditions conducted using a standard Rapid Assessment of Avoidable Blindness (RAAB) methodology.⁸ Two RAAB surveys in the Koulikoro region were conducted, in 2008⁹ and 2011.¹⁰ More recently, during 2022–2023, similar surveys were conducted in the Segou and Sikasso regions.¹¹

RAAB surveys are used by national and subnational ministries of health to plan eye care services, including the development of infrastructure and training of eye health personnel. In Mali, data from the two earlier surveys in Koulikoro, supplemented by the national eye health system assessment conducted in 2017,^12,13 guided the development of eye care services in the region. The work has been led by the Malian Ministry of Health with support from the international NGO, Sightsavers. The implemented activities aimed to increase coverage and improve the quality of cataract services and included training and deploying eye care technicians in primary and secondary health facilities; introducing free eye screening; and strengthening a community outreach programme with local volunteers providing community education and mobilisation.¹⁴

To understand how these interventions have affected VI prevalence in Koulikoro, and to realign services with needs, a third RAAB was conducted in the region in 2024. This survey also provided an opportunity to generate up-to-date data on cataract surgical coverage (CSC) and effective cataract surgical coverage (eCSC). eCSC is an indicator for monitoring progress towards universal eye health coverage, agreed at the 74th World Health Assembly in 2021.¹⁵

CSC is defined as the number of eyes operated upon for cataract expressed as a proportion of all eyes with operated upon or operable cataract; and eCSC is the number of eyes operated upon for cataract, with good visual outcomes expressed as a proportion of all eyes with operated upon or operable cataract.

The aim of this paper is to describe how the prevalence and causes of VI and the coverage of cataract surgical services, including outcomes of operated upon eyes, have changed from 2008 to 2024.

Materials and methods

Unlinked cross-sectional population-based surveys were conducted in the Koulikoro region, Mali, in 2008 and 2024. The surveys used the standard RAAB methodology described elsewhere,⁸ albeit with slight differences, as described below.

Study area

Both surveys aimed to be representative of the Koulikoro region; however, the study conducted in 2024 did not include all districts due to security issues and the inaccessibility of certain areas. The study investigators worked with the Koulikoro Regional Health Department and the chief medical officers at the region's referral health centres to conduct a security assessment, and as a result the districts of Banamba in the east and Nara in the north of the region were excluded from the sampling frame.

Sampling

The sampling strategy was similar in both surveys and followed the standard RAAB approach of two-stage sampling. First, probability proportional-to-size was used to select enumeration units from lists prepared by the National Population Directorate.¹⁶ Within each enumeration unit, a specific number of households were selected, with the field team moving from house to house to identify, enumerate and examine eligible individuals, which in RAAB surveys are residents aged ≥50 y. In large enumeration units (where the population was expected to exceed 500 individuals), field teams worked with local leaders to establish the village boundaries and size, and to use compact segment sampling to split the unit and choose a segment at random.¹⁷

In 2008 the sample size was calculated to be 2600 people, or 52 clusters of 50 people aged ≥50 y. This was based on an expected prevalence of blindness in adults aged ≥50 y of 5.7%, a precision of 20% (i.e. worst acceptable result of 4.6%), 95% CIs, a design effect of 1.5 and 10% non-response.

In 2024 the sample size was calculated to be 2040 people, or 51 clusters of 40 people aged ≥50 y. This was based on an expected prevalence of blindness of 7.1% from the 2011 RAAB,¹⁰ 95% CIs, a precision of 20% (i.e. worst acceptable result of 5.7%), a design effect of 1.4 and 10% non-response. A design effect of 1.4 with a cluster size of 40 was selected because of the security issues to avoid field teams being on the road too early or spending a lot of time in one place.

In both studies, the inclusion criteria were people aged ≥50 y who had been residents in the households for the past 6 mo. People normally eligible but absent on the day were enrolled but not examined and basic details of their visual status were gathered from their families or neighbours.

Data collection

All participants were examined at their homes by a study team, which included an ophthalmologist, who had been trained by a certified RAAB trainer and passed an intra-observer variation test. Visual acuity (VA) was measured and eyes with VA<3/60 were considered blind, eyes with VA<6/60–3/60 were severely visually impaired (SVI), eyes with VA<6/18–6/60 were moderately visually impaired (MVI) and eyes with VA≥6/18 were considered to have normal vision. In 2024, however, eyes with VA<6/12–6/18 were considered to have mild or early visual impairment, and only eyes with VA≥6/12 were considered to have normal vision.

Presenting visual acuity (PVA) was measured with available correction in daylight for each eye using the E-chart letter with optotype sizes 6/18 and 6/60 at 6 m, and those who could not see at 6 m were examined at 3 and 1 m. Eyes with PVA<6/18 in 2008 (VA<6/12 in 2024) were assessed for best corrected visual acuity with a pinhole. All eyes were examined with a direct ophthalmoscope in a dark room to understand whether the lens was normal, occluded, absent or if there was an intraocular lens. All eyes with PVA<6/18 (<6/12 in 2024) were assigned a cause for VI following the RAAB protocol. Eyes with VA<6/18 (<6/12 in 2024) and for which the causes could not be attributed to refractive error, cataract or corneal scarring were dilated with a short-acting mydriatic and underwent a posterior segment examination to determine the cause of vision impairment.

Operated upon eyes with VA≥6/12 are considered to have a ‘good’ outcome, those with VA<6/60–6/12 have a ‘borderline’ outcome and those with VA≤6/60 have a ‘poor’ outcome. VA was not measured to 6/12 in 2008 so for comparability we rely on the measure of ≥6/18.

Participants with unoperated cataract were asked their reasons for not having sought care. Participants who had been previously operated upon for cataract were asked for details of their operation(s) (age at the time of the operation, place of operation, surgical results).

After examination, all participants were given information about their eye health, and those who were diagnosed with conditions in need of treatment were given referrals to the nearest eye care centre.

Data management and analysis

In 2008, data were collected on paper forms that were double entered into a database and compared for accuracy. Data were initially analysed in the inbuilt RAAB software (Peek Vision Ltd, London, UK) analysis tool that produced standardised descriptive tables and were stored in .CSV files.

In 2024, data were captured directly on a tablet-based app built on the CommCare platform.¹⁸ Data were exported in .CSV format and analysed in Stata v18.

Estimates of prevalence of VI, causes of VI, CSC, eCSC and postsurgical visual outcomes are presented in this paper. Prevalence of VI and CSC/ eCSC estimates are adjusted for age and sex. The difference in prevalence of VI and CSC/eCSC between 2008 and 2024 was calculated as both absolute and relative change (in %). A two-sample test of proportions was used to determine if the change in VI and CSC/eCSC between 2008 and 2024 was statistically significant.

Both CSC and eCSC were calculated for three scenarios when operable cataract included eyes that were blind, had SVI or worse, and had MVI or worse.

Results

Participants’ characteristics

In 2008, 2438 out of 2600 recruited participants were examined, a response rate of 93.8%. In 2024, 1842 out of 2040 participants were examined, a response rate of 90.3% (Table 1).

Table 1.

Participant characteristics

	2008			2024
	Male, n (%)	Female, n (%)	Total, n (%)	Male, n (%)	Female, n (%)	Total, n (%)
Enumerated	1114 (42.8)	1486 (57.2)	2600 (100.0)	873 (42.8)	1167 (57.2)	2040 (100.0)
Examined	1064 (95.5)	1374 (92.5)	2438 (93.8)	802 (91.9)	1040 (89.1)	1842 (90.3)
Age (y)
50–59	370 (34.8)	557 (40.5)	927 (38.0)	324 (40.4)	477 (45.9)	801 (43.5)
60–69	357 (33.6)	469 (34.1)	826 (33.9)	249 (31.0)	330 (31.7)	579 (31.4)
70–79	249 (23.4)	265 (19.3)	514 (21.1)	145 (18.1)	169 (16.2)	314 (17.0)
>80	88 (8.3)	83 (6.0)	171 (7.0)	84 (10.5)	64 (6.2)	148 (8.0)
Total	1064 (43.6)	1374 (56.4)	2438 (100.0)	802 (43.5)	1040 (56.5)	1842 (100.0)
Population estimates
Age (y)	2009 Census			2023 Census
50–59	55 528 (51.2)	53 031 (48.8)	108 559 (48.8)	52 066 (48.7)	47 089 (49.0)	99 155 (48.8)
60–69	34 984 (32.2)	32 282 (29.7)	67 266 (30.2)	34 081 (31.9)	29 614 (30.8)	63 695 (31.4)
70–79	17 441 (16.1)	15 976 (14.7)	33 417 (15.0)	15 413 (14.4)	13 708 (14.3)	29 121 (14.3)
>80	6322 (5.8)	6853 (6.3)	13 175 (5.9)	5349 (5.0)	5683 (5.9)	11 032 (5.4)
Total	114 275 (51.4)	108 142 (48.6)	222 417 (100.0)	106 909 (52.7)	96 094 (47.3)	203 003 (100.0)

Women made up a higher proportion of examined participants in both studies: 56.4% in 2008 and 56.5% in 2024. The average age of the examined participants was 63.5 y in 2008 and 62.6 y in 2024.

Compared with population estimates, in both surveys people in the oldest age group (>80 y) were over-represented, and those in the youngest age group (aged 50–59 y) were under-represented (Table 1).

Prevalence of vision impairment

The age- and sex-adjusted prevalence of blindness in the Koulikoro region decreased from 9.0% in 2008 to 4.4% in 2024. This represents an absolute change of −4.6% (95% CI -6.1 to −3.1%; p<0.001) and a relative change of −51.1% in the prevalence of blindness. The change was significant among both male (−3.8%, 95% CI −6.0 to −1.6%; p<0.001) and female (−5.3%, 95% CI −7.3 to −3.3%; p<0.001) participants, representing a relative change of −46.9% and −53.5%, respectively (Table 2). There was no significant change in the prevalence of SVI or MVI between 2008 and 2024, either overall or in sex-specific groups.

Table 2.

Age- and sex-adjusted estimate of the prevalence of VI in the 2008 and 2024 RAABs

	2008 RAAB, % [95% CI]			2024 RAAB, % [95% CI]			Change, % [95% CI]
Prevalence of VI	Male	Female	Total	Male	Female	Total	Male	p	Female	p	Total	p
Blindness	8.1 [5.6, 10.6]	9.9 [8.0, 11.9]	9.0 [7.4, 10.6]	4.3 [3.2, 5.9]	4.6 [3.5, 6.0]	4.4 [3.6, 5.5]	−3.8 [−6.0, −1.6]	<0.001	−5.3 [−7.3, −3.3]	<0.001	−4.6 [−6.1, −3.1]	<0.001
SVI	2.8 [1.4, 4.3]	4.3 [3.0, 5.6]	3.6 [2.5, 4.7]	2.5 [1.7, 3.7]	3.5 [2.5, 4.9]	3.0 [2.3, 3.8]	−0.3 [−1.8, 1.2]	0.69	−0.8 [−2.3, 0.7]	0.32	−0.6 [−1.7, 0.4]	0.28
MVI	9.7 [7.4, 12.0]	12.0 [9.9, 14.0]	10.8 [9.2, 12.4]	9.0 [7.1, 11.4]	10.0 [8.4, 11.9]	9.5 [8.2, 11.0]	−0.7 [−3.4, 2.0]	0.61	−2.0 [−4.5, 0.5]	0.12	−1.3 [−3.1, 0.5]	0.16

MVI: moderately visually impaired; RAAB: Rapid Assessment of Avoidable Blindness; SVI: severely visually impaired; VI: vision impairment.

Causes of vision impairment in the 2008 and 2024 RAABs

The proportion of blindness and lower levels of VI due to cataract has remained stable between the two studies. Untreated cataract was the main cause of 60.7% of blindness in 2008 and 59.4% in 2024 (Table 3). Similar patterns were observed for SVI and MVI. Glaucoma has increased as a cause of VI at all levels. It increased as a cause of blindness from 7.8% in 2008 to 15.8% in 2024, as a cause of SVI from 2.8% to 13.8% and of MVI from 3.2% to 11.5%.

Table 3.

Causes of visual impairment in the 2008 and 2024 RAABs

		2008 (%)			2024 (%)
		Blindness	SVI	MVI	Blindness	SVI	MVI
1	Untreated cataract	60.7	71.6	53.4	59.4	67.7	52
2	Glaucoma	7.8	2.8	3.2	15.8	13.8	11.5
3	Cataract surgical complications	9.3	3.7	4.2	11.9	10.8	1.5
4	Trachomatous corneal opacity	5.2	8.3	1.3	1	0	0
5	Uncorrected refractive error	0.7	3.7	32.8	2	3.1	27.5
6	Other causes*	16.3	10.1	5.1	9	4.5	7.5

CNS: central nervous system; MVI: moderately visually impaired; RAAB: Rapid Assessment of Avoidable Blindness; SVI: severely visually impaired.

^*Other causes refers to uncorrected aphakia, other corneal opacity, phthisis, onchocerciasis, diabetic retinopathy, age-related macular degeneration, other posterior segment disease and other globe or CNS abnormalities.

Cataract surgical coverage and effective cataract surgical coverage

CSC among persons at all levels of VA (3/60, 6/60 and 6/18) increased significantly between 2008 and 2024 (Table 4). The absolute changes were +29.3% (95% CI 26.5 to 32.1%; p<0.001), +28.5% (95% CI 25.6 to 31.4%; p<0.001) and +27.9% (95% CI 25.1 to 30.7%; p<0.001) for VA<3/60, VA<6/60 and VA<6/18, respectively. The change was significant among both men and women, and although the absolute changes were similar, the relative changes were much greater in women compared with men. For example, at VA<3/60, the absolute change in CSC was similar between men and women: +26.8% and +30.8%; however, these changes represented relative increases of 45.5% in men and 88.2% in women.

Table 4.

Age- and sex-adjusted cataract surgical coverage (persons) and effective cataract surgical coverage in the 2008 and 2024 RAABs

	2008 RAAB, % [95% CI]			2024 RAAB, % [95% CI]			Change, % [95% CI]
	Male (%)	Female (%)	Total (%)	Male (%)	Female (%)	Total (%)	Male	p	Female	p	Total	p
CSC
VA<3/60	58.9 [55.9, 61.9]	34.9 [32.4, 37.4]	46.3 [44.3, 48.3]	85.7 [83.3, 88.1]	65.7 [62.8, 68.6]	75.6 [73.6, 77.6]	26.8 [23.0, 30.6]	<0.001	30.8 [27.0, 34.6]	<0.001	29.3 [26.5, 32.1]	<0.001
VA<6/60	49.2 [46.2, 52.2]	28.7 [26.3, 31.1]	38.3 [36.4, 40.2]	77.8 [74.9, 80.7]	56.8 [53.8, 59.8]	66.8 [64.6, 69.0]	28.6 [24.4, 32.8]	<0.001	28.1 [24.3, 31.9]	<0.001	28.5 [25.6, 31.4]	<0.001
VA<6/18	30.5 [27.7, 33.3]	18.5 [16.4, 20.6]	24.0 [22.3, 25.7]	62.0 [58.6, 65.4]	42.3 [39.3, 45.3]	51.9 [49.6, 54.2]	31.5 [27.1, 35.9]	<0.001	23.8 [20.2, 27.4]	<0.001	27.9 [25.1, 30.7]	<0.001
eCSC
VA<3/60	28.5 [25.8, 31.2]	13.9 [12.1, 15.7]	20.8 [19.2, 22.4]	34.6 [31.3, 37.9]	38.0 [35.1, 40.9]	36.3 [34.1, 38.5]	6.1 [1.8, 10.4]	<0.01	24.1 [20.6, 27.6]	<0.001	15.1 [12.8, 18.2]	<0.001
VA<6/60	23.4 [20.9, 25.9]	11.2 [9.5, 12.9]	16.9 [15.4, 18.4]	31.4 [28.2, 34.6]	31.8 [29.0, 34.6]	31.6 [29.5, 33.7]	8.0 [3,9, 12.1]	<0.001	20.6 [17.3, 23.9]	<0.001	14.7 [12.1, 17.3]	<0.001
VA<6/18	14.8 [12.7, 16.9]	7.4 [6.0, 8.8]	10.8 [9.6, 12.0]	23.5 [20.6, 26.4]	21.8 [19.3, 24.3]	22.6 [20.7, 24.5]	8.7 [5.1, 12.3]	<0.001	14.4 [11.5, 17.3]	<0.001	11.8 [9.5, 14.1]	<0.001

CSC: cataract surgical coverage; eCSC: effective cataract surgical coverage; RAAB: Rapid Assessment of Avoidable Blindness; VA: visual acuity.

eCSC also increased significantly at all levels of VA. Both the absolute and relative increase in eCSC was substantially higher among female participants.

Visual outcomes among eyes operated upon for cataract

Visual outcomes among eyes operated upon for cataract improved between the two surveys. In 2008, 26.6% of operated upon eyes had good presenting vision, that is, they could see 6/18, increasing to 37.1% with a pinhole. In 2024, 51.5% of eyes had good presenting vision, increasing to 65.2% using a pinhole (Figure 1).

Figure 1.

Visual outcomes after cataract surgery in 2008 and 2024.

Discussion

The comparative analysis of the findings from the 2008 and 2024 surveys suggests that there has been a significant reduction in the prevalence of blindness in the Koulikoro region. Positive changes have been observed in both the coverage of cataract services and the postsurgery visual outcomes, resulting in a significant increase of both CSC and eCSC. This is one of the most marked improvements reported in recently published repeated RAABs in sub-Saharan Africa. Two other areas, which showed a statistically significant reduction in blindness in recent years, were the Kaolack region in Senegal, where the prevalence decreased from 6.5% in 2010 to 3.0% in 2022,¹⁹ and the Nampula region in Mozambique, where it decreased from 6.2% in 2011 to 4.5% in 2018.²⁰ The prevalence of blindness in Sierra Leone²¹ and in the Fatick region of Senegal¹⁹ stayed statistically unchanged over the past decade.

However, it is important to consider the 2009 and 2023 population variation when interpreting these changes. There was a modest relative increase in the proportion of individuals aged 60–79 y in the study area in 2023 compared with 2009, despite a slight overall population decline. Specifically, the combined proportion of individuals in the 60–69 and 70–79 y age groups increased from 48.3% in 2009 to 49.7% in 2023 (Table 1). This demographic shift is particularly relevant given the age-related nature of many causes of VI, such as cataract and age-related macular degeneration.

While the study design does not allow us to attribute causes, there are a number of factors that may have contributed to the observed changes in the Koulikoro region. First, the baseline prevalence of blindness was very high, and it is known that once eye care services become available, the prevalence of blindness falls more rapidly in the context of very high initial levels. Second, Mali has worked intensely to eliminate trachoma, which includes large-scale TT surgery campaigns and, as a result, many cases of irreversible blindness due to TT have been prevented. Third, significant investments have been made in training and deploying eye care technicians in primary and secondary health facilities, introducing free eye screening and strengthening a community outreach programme with local volunteers providing community education and mobilisation.¹⁴ The relationship between higher CSC and the prevalence of blindness is well established. Over the past 16 y, CSC in Koulikoro increased by 25%. In Kaolack in Senegal,¹⁹ where the reduction in blindness was also statistically significant, the increase was 17% over a 12-y period. In Sierra Leone and the Fatick region of Senegal, the increase in CSC was less substantial, about 10%, resulting in a more limited impact on the prevalence of blindness.^19,21

Although the prevalence of blindness decreased significantly from 2008 to 2024, the prevalence of SVI and MVI has remained relatively stable. One potential reason for this is that eye care programmes have typically concentrated on individuals with eye health issues that result in vision loss.²² Furthermore, the stable prevalence of SVI and MVI may indicate a growing burden of vision impairment caused by factors other than cataracts, such as diabetic retinopathy, glaucoma and age-related macular degeneration. These conditions are more challenging to detect and treat, particularly in areas with limited resources.²³ These conditions often progress slowly and are not always prioritised in outreach or surgical programmes that are mainly focused on cataracts.

In light of these findings, it is recommended that the country programme should reconsider the current intervention thresholds. Lowering the visual acuity threshold for cataract surgery and expanding screening and treatment for non-cataract causes of VI could help prevent progression to blindness. This approach is consistent with the WHO’s recommendation of integrated, patient-centred eye care, and with its call to address all levels of vision impairment and not just blindness as part of universal health coverage.⁵

The 2018 eye health system assessment in Mali documented a number of positive developments in the eye health sector, including the national policy for promoting quality in healthcare and the establishment of the National Agency for the Assessment of Hospitals to monitor quality, costs and patient satisfaction. It also noted strengths in terms of the integration of eye care into primary healthcare, fee exemptions for indigent and vulnerable populations. Additional highlights included eye health education initiatives by the Malian Union of the Blind and the sustained provision of financial and technical assistance by international donors.¹³

One very significant positive finding reported in this study was a substantial increase in CSC and eCSC among women, providing clear evidence of the improvements in the equity of service delivery. Very few studies that used the RAAB methodology in low-income African contexts reported gender equity in access to cataract surgery. The 2015 Rwanda RAAB showed no statistically significant differences in CSC between men and women.²⁴ In most other African settings, CSC is heavily skewed towards men and the changes over time often increase the gap. In Kaolack in Senegal, while the change in CSC increased from 68% to 78% for men, it stayed unchanged at 51% for women.¹⁹ In Sierra Leone, the CSC for men increased from 49% to 68%, but decreased for women from 34% to 33%.²¹ The improvements in gender equity in Mali have been attributed to specific programmatic activities targeting women and implemented in partnerships of organisations of persons with disabilities, women's associations and patient groups.¹³ A more detailed analysis of such interventions and how they were integrated within the existing health system will be of critical importance to share learnings globally.

Despite these significant improvements, the eye health system in Koulikoro requires further investments and strengthening. Cataract remains a very important cause of all VI and the leading cause of blindness responsible for 59% of the burden. Our data also show that, similar to other countries in the region, glaucoma is also becoming more important as a cause of VI. Glaucoma, if undiagnosed and untreated, can lead to irreversible blindness. The disease is also more difficult and expensive to manage, which can create challenges for the Malian health system over time. In addition, we have observed an increase in the proportion of cataract surgical complications reported as the principal cause of SVI and blindness. This highlights the need not only to expand access to cataract surgery, but also to ensure consistent surgical quality and postoperative care. A systematic review of population based studies conducted in 2023 found that although the number of cataract surgeries had increased globally, the quality of these surgeries, as measured by postoperative visual acuity, varied significantly, particularly in LMICs. Surgical complications including the development of posterior capsule opacification were among the main causes of poor visual outcomes.²⁵ These findings suggest that, although surgical coverage is improving, the quality assurance mechanisms, including surgeon training, equipment standards and postoperative follow-up, must be strengthened to prevent avoidable vision loss due to surgical complications.

Further, while human resources for eye health in Mali have strengthened and the number of ophthalmologists increased from 35 in 2011²⁶ to 52 in 2017, the ratio of ophthalmologists to population in 2017 was around 2.9 per million inhabitants,¹³ below the level recommended by the International Agency for Prevention of Blindness for Africa of 4 per million.²⁶ There is also a need to improve the system of supervision and mentoring of human resources, including those at the primary care level and to strengthen the Health Information Management System.²⁷ User fees for eye health consultations, surgery and spectacles are very high and create barriers to the uptake of services.^13,28 Opportunities for more affordable eye care and prepaid financing mechanisms need to be explored.

The study is subject to several limitations. First, due to security concerns, the 2024 survey did not include all areas in the Koulikoro region. This limits the generalisability of the results to those locations deemed relatively unsafe at the time of the fieldwork and the prevalence of blindness and CSC may be different in these locations. Second, the RAAB protocol allows only a single cause to be attributed to each eye or person, with preference given to those that are simple to treat; for example, unaddressed refractive error and cataract. This means that individuals with multiple morbidities will have only their simplest-to-treat condition recorded, and thus the prevalence of conditions such as glaucoma and other posterior segment diseases may be significantly underestimated. Third, comparisons between groups should be made with caution as the results do not reflect the full distribution of causes of VI within the study population. Finally, while the population-extrapolated estimates presented are based on the most reliable population data available, their accuracy is limited by any limitations inherent to the source data.

In conclusion, our study highlights that the prevalence of blindness in people aged 50 years and above (≥50 y) in the Koulikoro region has decreased significantly. It also provides evidence of improvement in both CSC and eCSC in the study population, with a significant improvement in women. The results show that investment in the Koulikoro eye care programme over the past 16 y is associated with the availability, quality and equity of eye care services. However, further investments in the system are required to further improve the volume and quality of surgical services and advance Mali's progress towards universal eye health coverage. Lessons from positive experiences in the Koulikoro region need to be disseminated widely throughout Mali and other similar settings. The findings also highlight the importance of investment in RAAB surveys for accurate monitoring of trends in access to eye care services and guiding further investments in strengthening the eye health system.

Authors’ contributions

VP, EJ, JO, DK, BD, NI and LT conceived the study; VP, JO, DK, SP EJ and LT led the data collection; SP analysed the data; VP led the drafting of the manuscript and commented on the results, tables and the figure. EJ, SB, JO, NI, SP and ES critically revised the manuscript. All authors contributed to the interpretation of results, provided critical revisions to the manuscript, read and approved the final version of the manuscript.

Funding

This study was supported by Sightsavers.

Competing interests

None declared.

Ethical approval

Ethical approval for both studies was obtained from the ‘Comité national d’éthique des sciences de la santé’, with the following reference number for 2024: decision N 2024/007/MSDS-CNESS of 10 June 2024 of Mali. However, the ethical approval reference number for the 2008 RAAB could not be found. Prior to enumeration and examination, the purpose of the study was explained to each eligible participant, and an opportunity was given to them to ask questions. Written consent was obtained from each participant, and the right of the individual to participate voluntarily, to refuse or to withdraw without providing an explanation, was explained.

Data availability

The data underlying this article are available on the RAAB World Survey website.²⁹ On this website, Koulikoro, Mali 2008 RAAB data⁹ are also available.