Utilisation, Expenditure and Risk-Adjusted Trends in Eye-Care Services in the Private Sector, South Africa

 

Michael Mncedisi Willie^*

^*Policy Research and Monitoring, Council for Medical Schemes, Pretoria, South Africa

^*Corresponding Author:  Michael Mncedisi Willie, Policy Research and Monitoring, Council for Medical  Schemes,  Pretoria,  South  Africa. E-mail: m.willie@medicalschemes.co.za

Received: 10 April 2026; Revised: 03 July 2026; Accepted: 27 July 2026; Published: 31 August 2026

Citation: Michael Mncedisi Willie. “Utilisation, Expenditure and Risk-Adjusted Trends in Eye-Care Services in the Private Sector, South Africa.” Jour of Eye Dis Thera (2026): DOI: 10.59462/JEDT-1.1.102.

Copyright: © 2026 Michael Mncedisi Willie. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduc-tion in any medium, provided the original author and source are credited.

Abstract

Background: Eye disease is a leading and increasing cause of avoidable visual impairment, yet population-level data on eye-care utilisation and expenditure in the South African privately insured sector are limited. We characterised the delivery, cost and risk-adjusted trends of eye-care services across all registered medical schemes.
Methods: We analysed anonymised, aggregated statutory submissions covering approximately 9.1 million beneficiaries for financial years 2019–2024. Officially published benefits-paid statistics served as the control (denominator) file against which service-line data were reconciled. Eye-care activity was examined through two complementary lenses: professional services delivered by eye-care disciplines (ophthalmology, optometry, optical dispensing, orthoptics) and eye-related inpatient admissions (ICD-10 H00–H59). Utilisation was risk-adjusted by direct age standardisation to the pooled beneficiary structure, and stratified by age, province and scheme type.
Results: In 2024, eye-care professional benefits totalled R8.29 billion (3.2% of total scheme benefits of R259.3 billion), split almost equally between ophthalmology (R4.06 billion; 453 practices) and optometry (R4.23 billion; 3,186 practices). Ophthalmology benefits were 94.2% risk-funded versus 43.8% for optometry. Eye-related inpatient benefits rose from R1.13 billion (2019) to R2.82 billion (2024); cataract accounted for 69.9% of inpatient eye spend in 2024. The age-standardised eye admission rate rose from 10.0 to a peak of 29.7 per 1 000  beneficiary-years (2023) before falling to 20.7 per 1000 beneficiary-years (2024); age standardisation only modestly attenuated the crude trend, indicating the rise was not driven by population ageing alone. Two-thirds (66.9%) of eye admissions occurred in beneficiaries aged ≥ 60 years. Restricted schemes showed higher age-standardised admission rates than open schemes

Keywords: Eye care; Cataract; Optometry; Ophthalmology; Health expenditure; Medical schemes; Age standardisation; South Africa

Introduction

Visual impairment and blindness impose a large and inequitably distributed burden on individuals and health systems. In 2020 an estimated 43 million people were blind and more than a billion lived with a preventable or as-yet-uncorrected vision impairment, with cataract and uncorrected refractive error remaining the two leading causes of avoidable vision loss worldwide [1,2]. Both conditions are highly amenable to intervention: cataract surgery is among the most cost-effective procedures in all of medicine [5], and refractive correction requires only a spectacle prescription, yet global coverage of each remains incomplete [3,4]. As populations age and the prevalence of diabetes climbs, demand for cataract surgery, refractive services and the management of diabetic retinopathy is projected to rise steeply [5-8], placing sustained pressure on the arrangements that finance eye care [7,8].

In South Africa these pressures play out within a two-tiered health system. Roughly one in six people is covered by a registered medical scheme, financed through a combination of risk-pooled contributions and individual medical savings accounts and regulated under the Medical Schemes Act [9]. Schemes must fund a defined set of Prescribed Minimum Benefits (PMBs), which include cataract surgery for advanced visual impairment, but routine refractive and optical services are typically met from discretionary day-to-day benefits or savings [7,8]. The result is a distinctive financing landscape in which surgical (ophthalmological) and primary (optometric) eye care are funded through structurally different mechanisms, a distinction that, as we show, has real consequences for members. The public sector, by contrast, has long struggled to reach national cataract-surgery targets [10,11] and the wider sub-Saharan region carries a disproportionate share of avoidable blindness relative to its ophthalmic workforce [6,12].

Despite the scale of the privately insured sector, published analyses of eye-care utilisation and expenditure across this population are scarce; most South African and regional evidence concerns the public system or is derived from prevalence surveys rather than financing data [6,10,13]. Routine statutory returns submitted by schemes offer an unusually complete, industry-wide picture of who receives eye care, from which providers, for which conditions, and at what cost. Yet these returns are seldom exploited for condition-specific analysis, and still more rarely risk-adjusted to separate a genuine change in demand from a shift in the age structure of the covered population.

This study was designed to address that gap. Using anonymised, aggregated statutory data covering essentially the entire medical-scheme population over six financial years, we set out to (i) quantify the volume, cost and funding source of eye-care services delivered by ophthalmologists, optometrists and allied disciplines;

(ii) describe variation across the available demographic and structural dimensions, namely age, province and scheme type; (iii) characterise trends over time in eye-related inpatient admissions and expenditure; and (iv) risk-adjust utilisation by direct age standardisation, so that observed trends and between-group differences could be interpreted net of demographic change. Throughout, the officially published benefits-paid statistics were used as a control file to anchor and validate the analysis.

Materials and Methods

Study Design and Data Sources

We conducted a retrospective, whole-of-industry analysis of routine administrative data submitted by all registered South African medical schemes for financial years 2019 to 2024. Data were supplied as anonymised, pre-aggregated statutory extracts containing no patient identifiers; each record represents a cell of covered lives or claims cross-classified by scheme, demographic and service attributes. Six linked extracts were used: (i) a benefits-paid control extract giving total amounts claimed and paid, and average beneficiaries, by scheme, benefit option, financial year, age band, beneficiary type and province (approximately 0.82 million records; 2019–2025); (ii) a professional-services distribution extract giving practising-provider counts, visits, beneficiaries treated and benefits paid by discipline and province for 2024; (iii) an inpatient morbidity extract (statutory table B.4) giving admissions, inpatient days and benefits by diagnosis (International Shortlist for Hospital Morbidity Tabulation groupings), scheme type and age band (2019–2024; 1.03 million records); (iv) a discipline-by-diagnosis extract (statutory table B.14) for 2021–2024; (v) a beneficiary-count trend extract (2014–2024); and (vi) a diagnosis-group dictionary mapping ICD-10 ranges to reporting categories.

The Control (Denominator)File

The officially published benefits-paid extract was designated the control file. It constitutes the authoritative, reconciled statement of total scheme expenditure and covered lives and was used for three purposes: (a) to validate the analysis, in that aggregate benefits and beneficiary counts derived from all other extracts were reconciled against it before any rates were computed;

(b) to express eye-care spend as a share of total scheme benefits; and (c) as the source of exposure denominators. The field Average Beneficiaries, the mean number of lives covered over the reporting year, provided beneficiary-years of exposure by age and province and was used as the denominator throughout. The 2025 partial year present in the control file was excluded from all analyses. Rates were computed on beneficiary-years of exposure obtained from the linked service extracts. For 2024 this exposure base totalled approximately 9.11 million beneficiary-years, marginally below the control-file year total of 9.17 million beneficiaries, the difference arising from records that could not be assigned to a province or scheme type. Expenditure shares are expressed against the control-file total. Beneficiary counts reported in Table 1 are year totals rather than the average-beneficiary measure used by the Council for Medical Schemes in its published per-beneficiary statistics, which runs approximately 0.9% lower; benefits per beneficiary in Table 1 are correspondingly about 0.9% below the published figures. 

Definition of Eye-CareServices

Eye care was operationalised through two complementary and deliberately non-overlapping lenses. The professional-services lens captured all activity billed under the eye-care disciplines: ophthalmology (medical and surgical specialists), optometry (including supplementary optometry and ocular therapeutics), optical dispensing and orthoptics. The disease lens captured eye-related inpatient admissions classified to ICD-10 chapter VII, Diseases of the eye and adnexa (H00–H59), sub-divided into cataract (H25–H26, H28) and other eye disease. Because professional (largely ambulatory) services and inpatient admissions represent different settings, the two lenses are reported separately and are not summed, to avoid double counting.

Data Cleaning and De-Duplication

Categorical fields were standardised for leading and trailing whitespace, and age-band and province labels were harmonised. A material structural artefact was identified in the discipline extracts: several discipline codescarried multiple synonymous text labels (for example, code 70 appeared as “Optometrists”, “Supplementary Optometrists” and “Ocular Therapeutics”, and code 26 as both “Ophthalmology” and “Any Discipline”), with mirrored values that would triple- or double-count expenditure if naively summed. We therefore aggregated strictly at discipline-code level, retaining a single canonical label per code. This correction reduced the apparent 2024 eye-care professional benefit from an inflated R16.8 billion to a de-duplicated R8.29 billion and is essential for valid inference.

Risk Adjustment

Utilisation was risk-adjusted for the confounding effect of population age structure using direct standardisation. Age-specific eye-admission rates were computed as admissions divided by beneficiary-years of exposure within each of nineteen age bands. Age-standardised rates were obtained by applying these age-specific rates to a fixed standard population defined as the pooled 2019–2024 beneficiary age distribution, so that year-on-year and between-group comparisons are net of demographic drift. The same method was applied to compare open and restricted schemes, using scheme-type-specific exposure obtained by linking each scheme’s registration number to its scheme type.

Statistical Analysis andEthics

Analyses were descriptive and are reported as counts, expenditure in South African Rand (R), and rates per
1,000 beneficiary-years. Numerical reporting follows a single convention: percentages and rates per 1,000 to one decimal place, ratios to two, expenditure in Rand billions to three significant figures, and expenditure in Rand millions to one decimal place, except in Table 2, where two are retained so that the small allied-discipline values are not lost. Thousands are separated by a space throughout. Per-beneficiary expenditure was computed as benefits paid divided by exposure. Provider supply was expressed per 100,000 beneficiaries. Data were processed in Python (pandas). Because the study used only anonymised, aggregated data with no possibility of re-identification, it did not constitute research on human participants; no individual consent was required. The work adheres to the principles of the Declaration of Helsinki.

Results

Covered Population and Control Totals

The covered population was stable across the study period, ranging from 8.99 million beneficiaries in 2019 to 9.17 million in 2024. Total scheme benefits paid rose from R185.9 billion in 2019 to R259.3 billion in 2024, a nominal increase of 39.5% over five years (Table 1). Benefits dippedin 2020 to R178.0 billion, a decline of 4.2%, consistent with deferred and suppressed utilisation during the first year of the COVID-19 pandemic, before resuming growth. All service-line aggregates reconciled to within reporting tolerance against these control totals.

Financial year	Benefi-ciaries (million)	Total benefits (R billion)	Benefits per benefi-ciary (R)	Year- on-year change in total benefits paid (%)
2019	8.99	185.9	20 679	–
2020	8.89	178.0	20 022	–4.2
2021	8.94	205.2	22 953	15.3
2022	9.04	218.4	24 159	6.4
2023	9.13	239.0	26 177	9.4
2024	9.17	259.3	28 277	8.5

Table 1.Covered population and total scheme benefits, 2019–2024.

Eye-Care Professional Services, 2024

In 2024, professional eye-care services accounted for R8.29 billion in benefits paid, 3.2% of all scheme benefits, delivered by 3,659 practising provider entities across 2.27 million visits (Table 2). Expenditure was split almost equallybetween ophthalmology (R4.06 billion) and optometry (R4.23 billion), but the two disciplines differ markedly in structure. Optometry operated through seven times as many practices (3,186 vs 453) and served more than three times as many beneficiaries (1.37 million vs 0.41 million) at a lower benefit per beneficiary (R3,078 vs R9,919). The starkest contrast is in funding source: 94.2% of ophthalmology benefits were paid from the risk pool, consistent with the PMB status of cataract and other surgical eye conditions, whereas only 43.8% of optometry benefits were risk-funded, the balance being drawn from members’ savings accounts, reflecting the discretionary classification of routine refractive and optical care. Optical dispensing and orthoptics were negligible by comparison (together under 0.1% of eye-care benefits).

Discipline	Practices	Visits	Beneficiaries treated	Benefits (R million)	Benefit per beneficiary (R)	% from risk pool
Ophthalmology	453	794 871	409 225	4 059.00	9 919	94.2
Optometry	3 186	1 476 331	1 373 883	4 229.30	3 078	43.8
Optical dispensing	16	1 268	1 297	4.50	3 469	43.4
Orthoptics	4	71	73	0.03	411	38.8
All eye care	3 659	2 272 541	-	8 292.83	-	68.5

Table 2.Eye-care professional services by discipline, 2024 (de-duplicated at discipline-code level). Beneficiaries treated are counted within each discipline and overlap across disciplines, so the total is not additive and is shown as a dash.

Variation in Eye Admission Rates by Provinceand Age

Eye-care spend per beneficiary varied roughly two-fold across provinces (Figure 1; Table 3). The Eastern Cape (R1,144), Western Cape (R1,108), KwaZulu-Natal (R1,039) and Gauteng (R960) exceeded the industry average of about R910, while Mpumalanga (R561), North West (R622) and Limpopo (R681) fell well below it; the Northern Cape (R881) and the Free State (R907) sat below the average by smaller margins. Provider supply was uneven: Gauteng concentrated 34.9% of all eye-care practices, whereas per-capita supply (practices per 100,000 beneficiaries) was highest in Limpopo (64.5) and lowest in North West (29.5) and Northern Cape (29.7), indicating that a relatively dense optometric footprint in some under-spending provinces did not translate into higher expenditure, likely reflecting differences in case mix, referral to surgery and benefit take-up rather than access alone.

Figure 1.Eye-care benefits paid per beneficiary by province, 2024. Dashed line marks the industry mean of R910 per beneficiary. Bars are ordered by expenditure per beneficiary; the five highest-spending categories are shown in amber and the remainder in teal.

Province	Practices	Practices per 100 000	Benefits (R million)	Benefit per benefi-ciary (R)	Visits per benefi-ciary
Eastern Cape	225	34.3	749.6	1 144	0.34
Western Cape	509	38.6	1 460.8	1 108	0.28
KwaZulu-Natal	538	45.0	1 242.5	1 039	0.29
Gauteng	1 278	38.2	3 214.6	960	0.25
Free State	162	40.9	359.3	907	0.24
Northern Cape	57	29.7	169.2	881	0.24
Limpopo	310	64.5	326.9	681	0.24
North West	142	29.5	299.2	622	0.19
Mpumalanga	191	35.8	298.9	561	0.18
Nine provinces	3 412		8 121.0
Other or unspecified	247		171.8
All	3 659		8 292.8

Table 3.Eye-care professional services by province, 2024. Excludes the Other or Unspecified province category, which accounts for 247 practices and R171.8 million in benefits.

Figure 2.Age distribution of eye-related inpatient admissions, 2024, showing cataract as a share of the total. Cataract becomes the predominant admission type from the mid-50s onward.

Trends in Eye-Related Inpatient Care, 2019–2024

Eye-related inpatient benefits rose steadily from R1.13 billion in 2019 to R2.82 billion in 2024, a 149% nominal increase that substantially outpaced the 39.5% growth in total scheme benefits over the same period (Table 4). Cataract consistently accounted for approximately 70% of eye-related inpatient expenditure, with cataract benefits rising from R914.6 million to R1,974.9 million. Admission volumes were more variable: total eye admissions climbed from 84,758 (2019) to a peak of 274,141 (2023) before falling to 196,246 (2024). The exceptionally low 2019 figure and the 2023 peak are partly attributable to reporting completeness and to a post-pandemic catch-up in deferred elective surgery, respectively. The blended benefit per admission was R14,392 in 2024; cataract, a higher-cost procedure, averaged R17,315 per admission.

Figure 3.Eye-related inpatient admissions (stacked bars: cataract vs other eye disease, left axis) and total eye benefits paid (line, right axis), 2019–2024.

Financial year	Cataract ad-missions	Other eye ad-missions	Cataract bene- fits (R m)	Other eye ben-efits (R m)	Total eye bene-fits (R m)	Cataract share (%)
2019	68 548	16 210	914.6	219.0	1 133.6	80.7
2020	71 573	92 906	1 193.7	641.9	1 835.6	65.0
2021	83 955	73 144	1 423.2	632.0	2 055.2	69.2
2022	95 885	96 229	1 685.4	804.7	2 490.1	67.7
2023	119 257	154 884	1 753.8	896.0	2 649.8	66.2
2024	114 055	82 191	1 974.9	849.4	2 824.3	69.9

Table 4.Eye-related inpatient admissions and benefits by diagnosis group, 2019–2024.

Risk-adjusted utilisation trends

Because eye disease is heavily concentrated in older beneficiaries, part of the observed rise in admissions could reflect gradual ageing of the covered population. Direct age standardisation to the pooled 2019–2024 age structure showed that this was not the principal driver (Figure 4). The age-standardised eye admission rate rose from 10.0 per 1 000 beneficiary-years in 2019 to a peak of 29.7 per 1,000 in 2023, then eased to 20.7 per 1,000 in 2024. Standardised and crude rates tracked closely throughout: the standardised 2024 rate (20.7) was only marginally below the crude rate (21.5), confirming that demographic drift explained little of the trend and that the growth reflected genuine changes in the propensity to receive inpatient eye care, including post-pandemic recovery of deferred surgery.

Figure 4.Crude and age-standardised eye admission rates per 1,000 beneficiaries, 2019–2024

Note.The close alignment between the crude and age-standardised ratessuggests that changes in the age composition of beneficiaries explain only a limited proportion of the observed trend in eye-related admissions. Age standardisation facilitates comparison over time by accounting for differences in population age structure.

Age-standardised rates differed consistently by scheme type (Table 5). Restricted schemes, typically employer-or industry-based, recorded higher standardised eye admission rates than open schemes in every year from 2020 onwards, reaching 24.3 versus 17.7 per 1 000 in 2024. In 2019, the first year of the series, the position was reversed, with restricted schemes at 9.4 and open schemesat 10.3 per 1 000; this coincides with the anomalously low 2019 admission counts noted under Strengths and Limitations and should be read with that caution. Restricted-scheme membership is younger than that of open schemes, so a crude comparison would understate the difference; standardisation therefore reveals a gap that the unadjusted rates conceal, pointing to differences in benefit richness, member composition and care-seeking. Both scheme types showed the same 2023 peak and 2024 moderation.

The 2023 values should not be interpreted. The open-scheme rate rose by 67.9% between 2022 and 2023 and fell by 34.9% the following year, against movements of +14.2% and –24.3% in restricted schemes, compressing the restricted-to-open ratio to 1.18 from 1.73 in 2022. Benefits per admission in 2023 fell to R14 706 for cataract and R5 785 for other eye disease, against a stable trend of approximately R17 400 and approximately R9 350 respectively, while expenditure itself remained on trend. The admission counts for that year therefore appear inflated relative to the expenditure they generated, in a pattern consistent with duplicated records rather than with a rise in activity, and the excess appears concentrated in open-scheme submissions, which accounts for the compression of the ratio. Scheme-type comparisons should rest on 2020 to 2022 and 2024; the 2019 values are separately affected by known under-reporting.

Financial year	Open schemes	Restricted schemes	Restricted: open ratio
2019	10.3	9.4	0.91
2020	15.0	24.3	1.62
2021	15.1	21.4	1.42
2022	16.2	28.1	1.73
2023	27.2	32.1	1.18
2024	17.7	24.3	1.37

Table 5.Age-standardised Eye Admission Rates per 1,000 Beneficiary-Years by Scheme Type, 2019–2024

Note:The 2023 estimates are affected by an apparent artefact in the admission counts and should therefore not be interpreted. The 2019 estimates are also affected by known under-reporting. Comparisons across scheme types should consequently focus on the more reliable estimates for 2020–2022 and 2024.

Reconciliation with the control file

Expressed against the control totals, eye care is a small but non-trivial and disproportionately growing component of scheme expenditure. Professional eye-care services represented 3.2% of total 2024 benefits, while eye-related inpatient benefits grew at close to four times the rate of overall benefits between 2019 and 2024 (149% vs 39.5%). Combining the two lenses, and acknowledging partial overlap in ophthalmology fees, total identifiable eye-related expenditure approached R9–10 billion in 2024, of the order of 3.5 to 4% of scheme benefits, a figure consistent with the ageing, diabetes-driven demand profile documented above.

Discussion

Eye care sits at the intersection of an ageing population, a rising diabetes burden and a financing system that treats surgical and primary services very differently, yet the privately insured segment of that system has rarely been examined in its own right [7,10]. This study set out to fill that gap: to quantify the volume, cost and funding source of eye care across all registered South African medical schemes, to map its variation by age, province and schemetype, to trace its trajectory over six financial years, and to establish, through direct age standardisation, whether any observed growth reflected genuine change in demand or merely the ageing of the covered population. Anchoring every service-line figure to the published benefits-paid control totals allowed these questions to be answered across the entire covered population with an unusual degree of confidence. Four findings emerged, and each speaks to the wider literature in an instructive way.

First, eye care proved to be a material and fast-growing cost centre. Professional eye-care benefits reached R8.29 billion in 2024, 3.2% of all scheme benefits, while eye-related inpatient spending grew by 149%, or roughly two-and-a-half-fold, over the study period, outpacing the 39.5% growth in total scheme expenditure by a wide margin. This trajectory is consistent with the long-standing projection that demand for eye services would rise as populations age and diabetes becomes more prevalent [13-16], and it lends empirical, financing-side weight to global burden-of-disease estimates that have hitherto rested largely on prevalence modelling [1,2,4]. That cataract surgery, repeatedly shown to be among the most cost-effective of all health interventions [17-19], accounts for roughly 70% of inpatient eye spend suggests that this growth, while substantial, is buying high-value care rather than low-value activity.

Second, the sector was structurally bifurcated in a way that carries direct consequences for members. Ophthalmology and optometry commanded almost identical total budgets, yet sat on opposite sides of the risk–savings divide: 94.2% of ophthalmology benefits were paid from the risk pool, reflecting the PMB status of cataract and other surgical eye disease, whereas fewer than half of optometry benefits were, the balance being drawn from discretionary savings. Routine refractive and optical care may remain more exposed to affordability constraints and out-of pocket expenditure, whereas cataract surgery is more often associated with treatment of an established clinical condition and the potential for substantial improvement in vision and quality of life [20]. This matters because uncorrected refractive error remains one of the largest and most avoidable causes of vision impairment globally [3,21], and because the prevalence of both myopia and presbyopia is rising [13,14]. A benefit structure that leaves spectacle correction dependent on the state of a member’s savings account risks reproducing, within an insured population, the very coverage gaps that population surveys document in the wider region [6,17]. The finding gives a concrete, financing-level mechanism for a barrier that the access literature has generally framed in terms of cost and geography [22,23].

Third, the rise in surgical demand was genuine rather than demographic. Age standardisation to the pooled beneficiary structure barely attenuated the upward trend: the standardised 2024 admission rate (20.7 per 1,000) sat only marginally below the crude rate (21.5), so the growth cannot be explained by an ageing membership alone. Part of the increase plainly reflects the post-pandemic recovery of deferred elective surgery, a pattern visible in the 2020 dip and the 2023 peak, but the sustained level shift is more consistent with rising underlying activity, improving surgical throughput and the accumulating ophthalmic consequences of diabetes [15,16]. This is an important corrective: without standardisation, an observer might reasonably have attributed the whole trend to demographicdrift and concluded that no change in provision was warranted. The analysis shows the opposite, and in doing so illustrates why routine utilisation reporting in this sector should be risk-adjusted as a matter of course, if progress against the global eye-health targets is to be measured meaningfully [23,24].

Fourth, the geographic and scheme-type variation had clear equity implications. Per-beneficiary eye-care spend varied roughly two-fold across provinces and, tellingly, tracked provider density only weakly: Limpopo combined the densest optometric footprint with below-average expenditure. Access to eye care thus appears to be shapedless by the raw availability of practitioners than by case mix, referral pathways into surgery and the architecture of benefits, echoing evidence from sub-Saharan Africa and from the global eye-health literature that workforce supply is necessary but far from sufficient for effective coverage [6,12,22,23]. The persistently higher age-standardised admission rates in restricted schemes, whose membershipis younger than that of open schemes, are consistent with richer benefit designs and distinct member populations rather than with greater clinical need, and they deserve scrutiny from a fairness perspective. Read alongside the concept of effective cataract surgical coverage [22], these gradients suggest that measuring inputs, whether spend or practitioner counts, is a poor proxy for whether beneficiaries are actually receiving the sight-restoring care they need.

Taken together, the results argue for closer monitoring of optometric savings exposure, for network and theatre-capacity planning calibrated to a cataract-heavy, elderly demand profile, and for standardised, risk-adjusted reporting of eye-care utilisation that would let schemes and the regulator separate a real change in need from a demographic or reporting artefact: the kind of integrated, people-centred information base that underpins the global eye-health targets adopted by the World Health Assembly [24]. One methodological finding reinforces the last point: de-duplicating the mirrored discipline labels in the raw returns halved the apparent size of the sector, from an inflated R16.8 billion to the reconciled R8.29 billion. That such a large distortion could pass undetected without validation against the published control totals is a caution to anyone analysing these statutory extracts, and a reminder that data provenance deserves the same rigour as the analysis built upon it.

Strengths and Limitations

The principal strength of this study is its coverage: essentially the entire medical-scheme population, reconciled against officially published control totals, over six years and across all major eye-care disciplines and diagnoses. Several limitations should be borne in mind. The data are aggregated administrative submissions, not clinical records, so we cannot verify diagnoses, visual-acuity outcomes or the appropriateness of care, and refractive error is only partially observable because much optometric activity is coded to examination rather than to a specific ICD chapter. Expenditure is reported in nominal Rand and has not been deflated; part of the year-on-year growth reflects medical inflation.

On that last point, a constant-Rand comparison is instructive. Expressed in 2023 prices, total scheme benefits moved from approximately R228.6 billion in 2019 to R239.0 billion in 2023, a real increase of about 4.6%, against a nominal increase of some 29% over the same period. Almost all of the nominal growth in scheme expenditure was therefore price rather than volume, which makes the real growth in eye-related inpatient expenditure a more striking outlier than the nominal comparison alone suggests. A full real-terms analysis, with the deflator stated, is a natural extension of this work.

Reporting completeness varies across years: the anomalously low 2019 inpatient counts and the 2023 peak are at least partly artefacts of submission timing and post-pandemic catch-up, and should not be over-interpreted as a smooth epidemiological trend. Finally, the professional-services extract was cross-sectional (2024 only), so the discipline-level split could not itself be trended; the longitudinal analysis rests on the inpatient extract. The de-duplication decision, although conservative and validated against the control file, could in principle under-or over-state activity for the small allied disciplines. A residual difference of approximately 0.9% between the beneficiary counts used here and the average-beneficiary measure published by the regulator means that per-beneficiary values reported in Table 1 sit marginally below the corresponding published figures; the direction and magnitude of every trend are unaffected.

A further limitation concerns compositional change in the covered population. Direct age standardisation removes the confounding effect of the age distribution and no other dimension of composition. Over the study period the scheme population was reshaped by selective exit as well as by ageing. Membership of open schemes declined while the total covered population grew only marginally, from 8.99 million in 2019 to 9.17 million in 2024, and demarcation-exempt health insurance products, which provide limited benefits and carry no Prescribed Minimum Benefits, have expanded rapidly over the same period on industry estimates. Because members who leave comprehensive cover for limited-benefit products are likely to be younger and in better health than those who remain, the residual scheme population may be sicker at any given age. A rising age-standardised admission rate is therefore consistent with two explanations that the present design cannot separate: genuine growth in the propensity to receive inpatient eye care, which we favour on the evidence of the sustained level shift and the concurrent growth in cataract expenditure, and progressive risk selection into the insured pool. Distinguishing them would require linkage of scheme-level entry and exit records to admission data, which the aggregated extracts used here do not support, and is a priority for further work.

Conclusion

This industry-wide analysis established that eye care is a substantial, structurally distinctive and fast-growing component of privately financed health care in South Africa. It was dominated by cataract surgery in an elderly population and by optometric refractive services funded largely from discretionary savings, and it was marked by pronounced provincial and scheme-type variation that persisted after risk adjustment. Crucially, age standardisation showed that the rise in surgical eye demand reflected genuine change rather than population ageing alone. Routine, risk-adjusted surveillance of eye-care utilisation anchored to the published benefits-paid statistics used here as a control would strengthen benefit design, provider-network planning and equity monitoring, and help ensure that a growing and largely avoidable burden of visual impairment is met both efficiently and fairly.

Declarations

Ethics approval and consent to participate. This study analysed only anonymised, aggregated administrative data with no possibility of individual re-identification and did not constitute research on human participants; ethics approval and individual consent were therefore not required. The work adheres to the principles of the Declaration of Helsinki.

Consent for publication:Not applicable; no individual-level or identifiable data are reported.

Availability of data and materials:The analysis draws on statutory returns submitted by registered medical schemes to the Council for Medical Schemes (CMS). The aggregated benefits-paid statistics used as the control file are published by the regulator and are publicly available; the granular record-level extracts are held under the Council’s data-governance framework and may be made available on reasonable request, subject to the applicable permissions and confidentiality safeguards.

Competing Interests:The author is employed by the CMS, the regulator that collects and curates the statutory data analysed in this study. The author declares no financial competing interests. The views expressed are those of the author and do not necessarily represent the official position of the CMS.

Funding:This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

Author’s Contributions:MMW conceived and designed the study, curated and analysed the data, and wrote the manuscript. The author read and approved the final version.

Acknowledgements:The author thanks the data custodians for access to the statutory submissions and colleagues who reviewed the analytical approach.

 

References

1. GBD 2019 Blindness and Vision Impairment Collaborators. Causes of blindness and vision impairment in 2020 and trends over 30 years. Lancet Glob Health. 2021;9(2):e144–e160.

2. Bourne, Rupert, Jaimie D. Steinmetz, Seth Flaxman, Hugh R. Taylor, Serge Resnikoff, Robert James Casson et al. “Trends in prevalence of blindness and distance and near vision impairment over 30 years: an analysis for the Global Burden of Disease Study.” The Lancet global health 9, no. 2 (2021): e130-e143.

3. World Health Organization. World report on vision. Geneva: WHO; 2019.

4. Flaxman SR, Bourne RRA, Resnikoff S, Ackland P, Braithwaite T, Cicinelli MV, et al. Global causes of blindness and distance vision impairment 1990–2020: a systematic review and meta-analysis. Lancet Glob Health. 2017;5(12):e1221–e1234.

5. Lansingh VC, Carter MJ, Martens M. Global cost-effectiveness of cataract surgery. Ophthalmology. 2007;114(9):1670–1678.

6. Naidoo K, Gichuhi S, Basanez MG, Flaxman SR, Jonas JB, et al. Prevalence and causes of vision loss in sub-Saharan Africa. Br J Ophthalmol. 2014;98(5):612–618.

7. Council for Medical Schemes. Annual Report 2023/24. Pretoria: CMS; 2024.

8. Council for Medical Schemes. Industry statistics and prescribed minimum benefits review. Pretoria: CMS; 2023.

9. Republic of South Africa. Medical Schemes Act 131 of 1998 and Regulations. Government Gazette; 1998.

10. Cook C, Cleary K. Eye health in South Africa. S Afr Med J. 2018;108(11):880–882.

11. Lecuona K, Cook C. South Africa’s cataract surgery rates: why are we not meeting our targets? S Afr Med J. 2011;101(8):510–512.

12. Resnikoff S, Lansingh VC, Washburn L, Felch W, Gauthier TM, et al. Estimated number of ophthalmologists worldwide. Br J Ophthalmol. 2020;104(4):588–592.

13. Holden BA, Fricke TR, Wilson DA, Jong M, Naidoo KS, et al. Global prevalence of myopia and high myopia. Ophthalmology. 2016;123(5):1036–1042.

14. Fricke TR, Tahhan N, Resnikoff S, Papas E, Burnett A, et al. Global prevalence of presbyopia and uncorrected near vision impairment. Ophthalmology. 2018;125(10):1492–1499.

15. Yau JWY, Rogers SL, Kawasaki R, Lamoureux EL, Kowalski JW, et al. Global prevalence and major risk factors of diabetic retinopathy. Diabetes Care. 2012;35(3):556–564.

16. International Diabetes Federation. IDF Diabetes Atlas, 10th ed. Brussels: IDF; 2021.

17. Bright T, Kuper H, Macleod D, et al. Barriers to accessing eye care services in low- and middle-income countries. Trop Med Int Health. 2018;23(3):234–247.

18. Tabin G, Chen M, Espandar L. Cataract surgery for the developing world. Curr Opin Ophthalmol. 2008;19(1):55–59.

19. Frick KD, Foster A. The magnitude and cost of global blindness. Am J Ophthalmol. 2003;135(4):471–476.

20. Kirby M. Understanding cataract in the ageing eye. Community Eye Health. 2020;33(109):20–21.

21. Naidoo KS, Jaggernath J. Uncorrected refractive errors. Indian J Ophthalmol. 2012;60(5):432–437.

22. Ramke J, Gilbert CE, Lee AC, Ackland P, Limburg H, et al. Effective cataract surgical coverage: an indicator for measuring quality-of-care. PLoS One. 2017;12(3):e0172342. 

23. Aghaji, Ada, Helen ED Burchett, Wanjiku Mathenge, Hannah Bassey Faal, Rich Umeh, Felix Ezepue, Sunday Isiyaku et al. “Technical capacities needed to implement the WHO’s primary eye care package for Africa: results of a Delphi process.” BMJ open 11, no. 3 (2021): e042979.

24. World Health Organization. Integrated people-centred eye care: a global framework. Geneva: WHO; 2020.