A Review to Populate A Proposed Cost-Effectiveness Analysis of Glaucoma Screening in Sub-Saharan Africa

ABSTRACT

To populate a proposed cost-effectiveness analysis of glaucoma screening in Sub-Saharan Africa (SSA).

A complete search was conducted on PubMed, Medline and African Journals Online (AJOL) to obtain relevant published articles, which were included in this review. All relevant articles on prevalence of glaucoma in SSA and among other African-derived populations, severity of glaucoma, cost of diagnosis and management, clinical effectiveness of glaucoma screening and treatment and the different glaucoma screening strategies in SSA were reviewed.

Population screening interventions for glaucoma may be considered as follows: standalone screening for glaucoma, screening for glaucoma during cataract outreach, and screening incorporated with diabetic retinopathy image review using tele-ophthalmology. Our review suggests that cost of glaucoma treatment is relatively low with cost of medical treatment ranging from USD 273 to USD 480 per year/patient and surgical treatment cost of USD 283 per patient as with other developing countries. Compliance with medication is moderate to good in about 50% of glaucoma patients. Prevalence of glaucoma is much higher in SSA and almost 50% of glaucoma patients are blind in at least one eye at presentation in clinics (without outreach screening). Our review suggests a moderate sensitivity and specificity in identifying glaucoma with basic equipment (direct ophthalmoscope, contact tonometer and frequency doubling technology) during outreach screening although about a third or fewer take up glaucoma services in clinics.

Our review provides the necessary information to conduct a cost-effective analysis of glaucoma screening in SSA using the decision Markov model.

KEYWORDS:

• Glaucoma
• Screening
• Cost effectiveness analysis
• Markov model
• Sub- Saharan Africa

Introduction

Glaucoma is a progressive optic neuropathy with characteristic optic disc changes and functional visual field defects, for which elevated intraocular pressure (IOP) is one of the most important identified risk factors.¹ The disease is characterized by a distinctive pattern of optic nerve damage, loss of retinal ganglion cells, and thinning of the retinal nerve fibre layer.² Glaucoma can be categorized into two main types based on the morphology of the anterior chamber angle: open and closed-angle glaucoma. These two types of glaucoma differ with regard to risk factors, epidemiology, natural history and management. Glaucoma can also be classified into primary and secondary types, depending on the underlying etiology.

Primary open-angle glaucoma (POAG) is the most common form of glaucoma among persons in Africa or of African descent,^3–6 who have a higher prevalence of glaucoma, 4.32% over the age of 40 years as estimated by Quigley et al., compared to other racial groups.⁷ Tham et al⁸ reported a similar prevalence (4.20%) in Africa. There are an estimated 8.7 million people between the ages of 40 and 80 years living with POAG in Africa, and this number is expected to nearly double to 16.2 million by 2040.⁸ These estimates are based on population studies among African cohorts^{3,4, 5, 6,9} showing prevalence ranging from 4.2 to 8.8%. Population prevalence of blindness due to glaucoma in Africa (15–20% of affected persons) is double that of the rest of the world (8%).¹⁰ Africa has a disproportionate prevalence of glaucoma blindness relative to its age structure and population.¹¹

Glaucoma is a leading cause of irreversible blindness in Sub-Saharan Africa (SSA), and the proportion of blindness attributable to glaucoma is steadily rising as the population ages. Glaucoma blindness risk is influenced by the age of onset of the disease,¹² access to services,¹³ quality of care¹⁴ and adherence to treatment and follow-up.¹⁵ Other common factors underlying the high blindness prevalence among persons with glaucoma in Africa include socioeconomic deprivation, poor disease knowledge and awareness.^16–19 In Nigeria, Abdull et al. reported that up to 47% of glaucoma patients are blind in at least one eye at first presentation.²⁰

Definitions of cost-effectiveness and clinical effectiveness

Cost-effectiveness analysis provides an evidence base for choosing interventions which deliver the greatest health gains for the available resources.²¹ It is calculated as the ratio of the net increase in health care cost to the net gain in health effects and can be expressed as the ratio of incremental costs to difference in effectiveness outcomes, also known as the incremental cost-effectiveness ratio (ICER).²²

Clinical effectiveness, on the other hand, is defined by the United Kingdom National Health Service as “the application of the best knowledge, derived from research, clinical experience, and patient preferences, to achieve optimum processes and outcomes of care for patients. The process involves a framework of informing, changing, and monitoring practice.”²³

The need for a cost-effectiveness study and the current review

The vast majority of persons in Africa with glaucoma are unaware of their disease, resulting in late presentation. In the Nigerian National Blindness Eye Survey, only 5.6% of participants with glaucoma knew of their diagnosis.³ A study in Ghana reported that persons with glaucoma in rural areas were twice as likely to be bilaterally blind (34.1%) from glaucoma than were urban dwellers (17.5%),²⁴ with the ratio of unilateral blindness particularly high (two times) among women in rural areas. Rural dwellers were also more likely to have glaucoma compared to urban-dwellers, who were more likely to be POAG suspects. It thus appears that glaucoma is more severe in rural settings, where early access to care is particularly poor.

Other studies confirm that the rate of undiagnosed glaucoma in developing countries is about 95%, compared to 50% in developed countries.^25,26 Glaucoma is often undiagnosed because it progresses unnoticed by affected persons until late in the disease course, when the central vision is affected. Diagnosing glaucoma in its early asymptomatic stage, necessary to reduce the risk of un-necessary blindness, requires access to eye services and often potentiated by screening outreach efforts.

Population-based screening for glaucoma has not been found to be cost-effective in developed countries.^27–29 However, Einarson et al.³⁰ in a study done in Canada reported that tele-glaucoma screening in a sparsely populated rural Canadian area was highly cost-effective and prevented 24% cases of blindness from glaucoma after 30 years. In addition, a recent report³¹ used a decision analytic Markov model to demonstrate that population screening for primary open-angle glaucoma (POAG) and primary angle-closure glaucoma (PACG) is cost-effective in both rural and urban China. The result of this study in China may be applicable to other settings with a highly blinding glaucoma and low labor costs such as in Sub-Saharan Africa. The report from China was consistent to an extent with a study in India by John et al.³² in a hypothetical screening program involving 10000 persons, which reported that community population screening might be cost-effective if targeted at the 40–69 age group and implemented in urban areas. The study estimated that community screening could prevent 2,190 person-years of blindness over ten years. Wittenbom et al.³³ reported that glaucoma screening was highly cost-effective in Barbados but not in Ghana, mainly due to the older population, higher life expectancy and higher per capita income in Barbados. This study, however, used a hypothetical model rather than actual health measures and was based entirely on a computer simulation model for the population of Ghana. These recent publications^30–32 suggest that glaucoma screening may be more cost-effective in poorer countries, presumably due to lower screening costs and greater risk of blindness due to very low rates of case detection in the absence of established eye health services or outreach programs.

Along similar lines, the American Academy of Ophthalmology Preferred Practice Pattern for open-angle glaucoma recommends the screening of high-risk populations such as persons of African descent (https://www.aao.org/preferred-practice-pattern/primary-open-angle-glaucoma-ppp2015). Ladapo et al.,³⁴ in a study to determine the potential benefit of universal glaucoma screening of African-Americans between the ages of 50 and 59 years, predicted that such a program could reduce the lifetime glaucoma blindness prevalence in this group by nearly half.

The high burden of glaucoma and glaucoma blindness and severity of glaucoma at presentation in Africa underscores the importance of better understanding the practicality of glaucoma screening outreach (glaucoma screening during cataract community screening outreaches, glaucoma as standalone and glaucoma screening during diabetic screening) through carefully modelling the cost-effectiveness of such efforts in African populations, which has not yet been done. There is currently no validated model for population-based glaucoma screening based on actual health measures to determine the cost-effectiveness of glaucoma screening among indigenous Africans. The current review aims to bring together the data necessary to populate such models.

Aims and features of the current study, requirements for a Markov model

Markov models are useful when a decision problem involves managing an ongoing risk over time.³⁵ The specific times of occurrence of events of interest in such models are uncertain, and the utility of outcomes depends on the timing of these events. Additionally, it is possible that a given event may occur more than once. Markov models estimate such temporal information by modeling the transition probabilities between different stages of a disease. We propose to model discrete stages of glaucoma progression using the International Society for Geographical and Epidemiological Ophthalmology (ISGEO)³⁶ classification of POAG as follows: POAG suspect, definite POAG, POAG blind in one eye and POAG blind in both eyes. Briefly, the ISGEO classification³⁶ describes three levels of evidence for a diagnosis of glaucoma. Category one is based on both structural and functional evidence: eyes with a vertical cup-disc ratio (VCDR) above the 97.5^th percentile for the population in question and/or asymmetry between eyes ≥ 0.2, with a definite defect on the visual field test consistent with glaucoma. The second level of evidence comprises advanced structural damage with VCDR of ≥0.9 and/or VCDR asymmetry of ≥ 0.3 and no available visual field data, and the third level of evidence describes the scenario when the optic disc is not seen and visual field testing is impossible, but the eye has a visual acuity < 3/60, with an IOP of ≥ 26 mmHg and the presence of a relative afferent pupillary defect and/or evidence of glaucoma surgery or medical records confirming glaucomatous visual morbidity.

Prior studies have assessed the cost-effectiveness of glaucoma screening using Markov models.^26,30 We propose to use a decision-analytic Markov model to assess the cost-effectiveness of three different screening strategies for glaucoma in Africa: community eye screening of glaucoma as a standalone, glaucoma screening piggybacking on cataract screening outreach and glaucoma screening as a part of a diabetic retinopathy image review program.

The following types of data are needed to populate our proposed Markov model and are the principal focus of the current review: prevalence and incidence of glaucoma in Africa, including severity at the time of first presentation/diagnosis; the risk of glaucoma progression between the above-defined stages in persons of African descent, in the treated and untreated state; clinical effectiveness (sensitivity/specificity) and cost of the various tools utilized in glaucoma screening programs; compliance with screening referral, as well as with recommended treatments such as medication and surgery; and clinical effectiveness cost of medical and surgical treatments for glaucoma and cost of productivity losses related to blindness.

POAG is the most common type of glaucoma in Africa as reported in several population studies and systematic reviews.^37-100 Therefore, the current scoping review and subsequent cost-effectiveness model will focus on POAG.

Glaucoma prevalence in Sub Saharan Africa

A PubMed search using the following criteria ((Glaucoma) AND (Prevalence OR Epidemiology)) AND (Africa OR sub-Saharan Africa OR Africans) performed on 6 November 2019 identified 456 articles. Articles published prior to 2002 were eliminated to remove those not employing the International Society of Geographical and Epidemiological Ophthalmology (ISGEO) classification system,³⁶ after which 293 articles remained. Among these, 6 met the inclusion criteria of population-based prevalence studies conducted in Africa and used the ISGEO classification system. ^{3,4,5,39,40,41} Table 1 shows the population studies of glaucoma prevalence in Africa

Table 1. Population studies of glaucoma prevalence in Africa.

Authors	Kyari et al.^Citation3	Ashaye et al.^Citation39	Budenz et al^Citation5	Rotchford at al^Citation39	Rotchford et al^Citation40	Bastawrous et al,^Citation41
Country and year of publication	Nigeria 2015	Nigeria 2013	Ghana 2013	South Africa (Temba) 2003	South Africa (Zulus) 2002	Kenya 2018
Number of participants screened	13591	811	5603	839	1005	2171
Age screened	40 years and above	40 years and above	40 years and above	40 years and above	40 years and above	50 years and above
Number with glaucoma	682	59	362	55	51	88
Prevalence of glaucoma	5.02%	7.3% Adjusted for age:6.9%	6.8%	5.3%	4.5%	4.3%
ISGEO category 1	268 (39.3)	16(27.1)	316 (87.2%)	12(21.8%)	Not documented	64 (72.7)
ISGEO category 2	378 (55.4)	39(66.1)	23 (6.4%)	30(54.5%)		22(25.0%)
ISGEO category 3	36 (5.3)	4 (6.8)	23(6.4%	13(23.6%)		2 (2.3%)
Unilaterally blind, n (%)	365 (38%)	5(8.4%)	52 (14.4%)	32(58.2%)	12(31.7%) 2.1% age sex-adjusted of the prevalence of glaucoma	Visual Impairment = 21.6%
Bilaterally blind, n (%)	136 (20%)	3(5.1%)	9 (2.5%)	18(32.7%)	41(3.2%) of these were blind from glaucoma 0.9% of the prevalence of glaucoma	5.7%
POAG	208 (86.0%) This is a proportion of all the primary glaucoma	50(84.7%)	342 (94.5)	31(56.4%) (adjusted prevalence of 2.9%	28 (2.7%)	83(94.3)
PACG	35(14.0%)	2(3.4%)	9 (2.5%)	5(9.1%)	5(9.8%)	1 (1.14%)

Other systematic reviews and meta-analysis have also reported on the prevalence of glaucoma among persons of African descent. Tham et al.⁸ reported that the prevalence of primary open angle glaucoma (POAG) 4.20% (95% CrI, 2.08–7.35) was highest in Africa. This is similar to the report by Rudnicka et al.⁴² in another systematic review and meta-analysis who reported that populations of African descent have the highest prevalence of POAG, which was 4.20% (95% CI, 3.1%–5.8%).

Glaucoma in other African populations in the United States and the Caribbean

A PubMed search was conducted on prevalence of glaucoma in St Lucia (13 papers retrieved), Barbados (43 papers were retrieved), and African Americans (170 papers were retrieved). Only population-based studies on prevalence of glaucoma were included. One paper met our inclusion criteria in St Lucia, two in Barbados and one among African Americans in the United States. Tielsch et al.,⁴³ in a population study in Baltimore, reported that the prevalence of POAG was four times higher among persons of African descent than in Caucasians. Leske et al.⁴⁴ in the Barbados eye study reported a POAG prevalence of 7% in persons of African descent, 3.3% in persons of mixed race and 0.8% in whites. In a pilot study in the Barbados by Leske et al.,⁴⁵ a prevalence of 6% was reported, while prevalence of 12.8% was reported among blacks and other mixed persons older than 54 years. Mason et al.⁴⁶ in St Lucia reported an unadjusted prevalence of 10.2% (8.5–12.0) and age-adjusted prevalence of 9.6% (8.0–11.2%) in a predominantly African-descended population. These results among African-descended populations elsewhere are consistent with data from Africa, showing higher population prevalence of glaucoma than that reported for other global regions, especially Europe and European-descended persons.

Data on the severity of glaucoma at presentation in Africa

A literature search was done in PubMed and African journals online (AJOL) to review published papers on the severity of glaucoma at presentation in Africa.

In Nigeria, several clinic-based studies have reported blindness in one eye in > 44% of glaucoma patients at presentation. In a study done in Northern Nigeria,⁴⁷ 47% of persons presented with blindness in at least one eye, while in South-West Nigeria,⁴⁸ the corresponding figure was 44.6%, and in the South East, it was 49.2%.⁴⁹ All these studies were clinic based and therefore may be prone to bias. The majority of these studies were done in tertiary hospitals where patients are more likely to present with advanced glaucoma. These patients might have presented at an earlier stage to secondary eye care facilities.

In Ghana, Francis et al.²⁴ reported that 52.2% of glaucoma patients in rural areas presented with blindness in at least one eye, while in urban areas, the corresponding figure was 32.9%. This finding suggests that severity at first presentation may reflect a rural-urban disparity, which may be due to poor healthcare access in the countryside. In another study in Ghana, Gyasi et al.⁵⁰ reported that 24% of patients were blind in at least one eye at presentation

Figures for unilateral glaucoma blindness at the time of presentation in other African settings range from 29–44,% in Ethiopia^51,52 to 15% in Malawi,⁵³ while rates of bilateral blindness have been 18–56%.^50–58 In a very recent study, Jones et al.⁵⁸ reported that 44.7% of glaucoma patients presented with severe visual field loss in Tanzania. All these studies suggest that glaucoma in SSA is severe and often presents at an advanced stage. In addition, prevalence of glaucoma in the West African region^37,4 seems higher compared to the East⁶ and South African region.^39,40

Clinical effectiveness of glaucoma screening in Africa

Screening interventions and strategies: Several screening interventions have been described in the published literature, including case finding or opportunistic screening at the time of presentation for other ocular conditions.⁶⁰ Population screening for glaucoma may also be considered under three main headings: standalone screening for glaucoma,⁶¹ screening for glaucoma during cataract outreach,⁶² and screening incorporated with diabetic retinopathy image review using teleophthalmology.⁶³ Many of these studies have been done in China and India. Regarding standalone screening, Akaraiwe et al.⁶⁴ reported that 14.5% of patients who presented at an urban community outreach had glaucoma in South-East Nigeria.

In a recent study by Kiage et al. in rural Kenya,⁶⁵ web-based tele-ophthalmology assessment was compared with clinical slit-lamp examination to screen for glaucoma among diabetic patients. A moderate agreement between tele-ophthalmology and clinical assessment was reported, with a kappa value of 0.55 (95% CI = 0.48–0.62) and positive and negative predictive values of 77.5% and 82.2%, respectively, of tele-ophthalmology screening for glaucoma. The authors reported that media opacities, inadequate patient cooperation and poor-quality image captured together resulted in a 24% rate of unreadable images.

Screening of first-degree relatives (FDRs) of previously diagnosed glaucoma patients is another screening strategy, which has been used in parts of Africa. Munachonga et al.⁶⁶ conducted a randomized trial in which FDRs in the intervention arm were provided free examination, while those in the control arm were made to pay the standard fee. However, the intervention group were no more likely to take up services than controls (RR 1.87, 95% CI 0.94, 3.73, p > .05).

Compliance with follow up after glaucoma screening

Another index of the effectiveness of these screening strategies is the number of patients screening positive who eventually take up services. Very few studies have reported such data from programs in Africa. Compliance with referral to the base hospital during African glaucoma screening programs has been estimated between 17.6%⁶⁷ and 32.5%. (Olawoye in press).

Diagnostic accuracy of screening instruments in SSA

Glaucoma evaluation involves assessing structural changes at the optic nerve head, functional visual loss and intraocular pressure measurement. In many parts of Africa, direct ophthalmoscopy is frequently used to assess the optic nerve head during screening outreach, while intra-ocular pressure is assessed with Perkins, Tonopen, iCare or other portable tonometers. Functional damage may occasionally be tested with Frequency Doubling Technique visual field devices during outreach. More complex evaluation of the nerve fibre layer or other indicators of glaucoma damage are often not practical in Africa outside of major urban centers.

Ophthalmoscopy (Direct): Direct ophthalmoscope is often used to examine the optic nerve head with or without dilation of the pupil. An advantage of the direct ophthalmoscope is its portability and use of batteries instead of electrical current, access to which may be problematic in remote areas. The main disadvantage, however, is its inability to give a stereoscopic view of the nerve. Harper et al.⁶⁸ reported a sensitivity of 81% (95% CI 69–89%) and a specificity of 90% (95% CI 84–95%) with the use of the instrument, when compared to a gold standard of stereophotography in the evaluation of the narrowest rim width for glaucoma screening.⁶⁸ Cook et al.⁶⁹ reported that the use of the direct ophthalmoscope using a cut-off of 0.7 for the vertical cup to disc ratio, in addition to the testing for afferent pupil defect, could give an accuracy of up to 90% for the case detection of glaucoma.

Disc photos using smartphones: The Portable Eye Examination Kit (PEEK) has been described by PEEK vision in the United Kingdom. It is low cost and holds a lot of promise for glaucoma screening in low-resource settings.

Intraocular pressure measurement

Intraocular pressure is a high-risk factor for the development and progression of glaucoma and the only modifiable risk factor.^70–73 The slit lamp mounted Goldmann’s applanation tonometer (GATT) is regarded as the gold standard for the measurement of intraocular pressure.⁷⁴ However, it is not often available for community outreaches in SSA. Therefore, hand-held portable tonometers are often used.

Contact handheld tonometers: A commonly used device is the Perkins applanation tonometer (PAT), which is portable and similar in optics to Goldmann applanation tonometry (GATT).⁷⁵ Arora et al.⁷⁵ reported the 95% limit of agreement between GATT and PAT to be −0.64–+1.08 mmHg with a mean difference of 0.22 mmHg in 100 patients measured with the two instruments. The Tonopen, another frequently used device, appears to give slightly less accurate measurement compared to GATT. In a study by Horowitz et al,⁷⁶ the Tonopen was found to underestimate IOP higher than 20 mmHg. The authors reported 95% limits of agreement as −5.5 to 4.7 mmHg.⁷⁶ In another study comparing the Tonopen with the gold standard of GATT, Billy et al.⁷⁷ reported a sensitivity of 56.3%, a specificity of 97.8%, a positive predictive value of 69.2% and a negative predictive value of 96.2% in the measurement of intraocular pressure using a cut off of 21 mmHg. Portability of this device is an advantage; however, and it measures with greater precision and reliability when IOP is <20 mmHg.⁷⁶

Non-contact tonometers: Recently, the I-care tonometer was introduced into the market. It has the advantage of being portable and can be used quickly and safely without anaesthetic. I-care device appears to be useful in routine clinical settings. Brusini et al.,⁷⁸ however, reported that the deviations of IOP using the I-care were highly correlated with central corneal thickness values (r = 0.63, P < .01. The regression model showed that for every CCT change of 10 microns, there was a deviation of 0.7 mmHg in I-care reading. It must be noted therefore that measured IOP with I-care is lower compared to GATT in eyes with thinner central cornea, smaller corneal curvature and glaucoma.⁷⁹ Therefore, although this instrument can be used to measure IOP at screening centres, definitive IOP measurements would need to be confirmed with GATT at main hospitals.

Frequency doubling technology (FDT) for evaluation of the visual field: In contrast to gold standard automated perimetry devices, the FDT measures fewer (19) visual field locations in the central 20 degrees of the visual field. It can be used in supra-threshold and threshold screening modes and is portable, user-friendly and rapid: testing in the suprathreshold mode can be completed in 1 minute per eye. In a study by Cello et al.,⁸⁰ values of 100% sensitivity and specificity for detecting advanced glaucoma, 96% sensitivity and 96% specificity for moderate and 85% sensitivity and 90% specificity for early glaucomatous visual field loss were reported.

Clinical Effectiveness of glaucoma treatment

The African race is a risk factor for development and progression of glaucoma, and medical and surgical treatment failure.⁸¹ In much of Africa, patients usually pay out of pocket for medical service at hospitals.⁸² There are basically three broad categories of treatment for glaucoma after a diagnosis has been made: topical drop therapy, laser and incisional surgery. Smith et al.,⁸³ in a meta-analysis, reported a pooled effect size of 0.66 (95% CI of 0.61–0.71) for medical intervention in SSA, although only two papers were used in the meta-analysis. The pooled effect size for surgical intervention (mainly trabeculectomy) was 0.73 (95% CI 0.65–0.80). Other results in the surgical intervention meta-analysis include one paper on deep sclerectomy,⁸⁴ two papers on combined trabeculectomy and cataract,^85,86 one on viscodilation and canaloplasty⁸⁷ and one paper on viscocanalostomy.⁸⁸ These studies demonstrated a high degree of heterogeneity.⁸³ The pooled effect size for laser was reported as 0.39 (95% CI 0.27–0.54) with a high heterogeneity between studies,⁸³ while that for surgically implantable devices (Ahmed valved implants) was 0.56 (95% CI 0.23–0.84, Q = 14.02, P = .001)⁸³

Capacity for glaucoma care in the hospital: Kyari et al.⁸⁹ in Nigeria recently reported that the number of Ophthalmologists per million population was 3.2 per million, which was just below the 4.0 per million recommended by Vision 2020, although there was a mal-distribution of the ophthalmologists to the urban centres. This suggests that there may be some capacity to manage glaucoma patients referred from screening outreaches at least in urban settings in SSA.

Progression Rates with treatment

No studies in Africa have investigated the progression of glaucoma with and without treatment over time. Yoshikawa et al.⁹⁰ proposed that when IOP reduction of >25% is achieved, over 5 years follow-up 20%, of patients will move from stage 1 (mild) to stage 2 (moderate) and 8.3% will move from stage 2 to 3 (severe), while 6.7% will move directly from stage 1 to stage 3. In the Barbados eye study,^91.92 the 4-year risk of open-angle glaucoma was 2.2% (CI = 1.7–2.8%). There were 67 new cases of glaucoma (out of 2989 at risk) with half just diagnosed and another half already receiving treatment. Age specific incidence increased from 1.2% in persons 40–49 years to 4.2% in those 70 years and above. In the Early Manifest Glaucoma Treatment (EMGTS) study,⁷² 59% of treated patients compared to 76% of untreated patients progressed at 8 years follow-up based on standard automated perimetry study despite an average reduction of 25%.

Cost of screening and treatment

The cost of managing glaucoma includes the cost of screening (equipment, costs of health personnel) and the cost of the treatment. The cost of treatment in the first year and in subsequent years shall be used to populate the cost-effectiveness model. Often, at screening centres, simple equipment such as the direct ophthalmoscope and the hand-held tonometers such as the Perkins or Tonopen is used. There are no studies in SSA that have evaluated the cost of screening at the community outreach. However, the model shall be populated with real-time data at a regular community screening outreach in Nigeria. Definitive investigations such as optical coherence tomogram (OCT) and visual field test are often done at base hospital. Again, real-time data shall be obtained at the tertiary hospital in Ibadan, Nigeria. The cost of treating glaucoma medically has been reported in several studies in Africa. Adio et al.⁹³ reported that the average monthly cost of glaucoma medications for one person was USD 40 in Nigeria (480 USD per year). Ocansey et al.⁹⁴ also reported a similar cost of medications in Ghana (USD 429 per person per year) and direct monthly non-medication costs of USD 54 per patient per year (USD 483 per person per year). Omoti et al.⁹⁵ in Nigeria reported that the cost of medical treatment per patient per year was USD 273. A meta-analysis,⁸³ which included three published papers in SSA (two in Nigeria and one in Ghana), reported a mean medication cost per patient per year of USD 393. Omoti⁹⁵ also reported the cost of surgery per patient to be USD283.

Indirect productivity loss: Indirect productivity loss (waiting time to see the doctor) per visit to the hospital is estimated to be 30 USD per visit (150 USD for a total of 5 visits per year) based on the study by Adio et al.⁹³ using the human per capital method. It is estimated that patients will spend an average of 3 − 8 hours per visit.

Compliance with treatment

Five studies have reported compliance with medications in sub-Saharan Africa.^96–100 Mehari et al.⁹⁶ reported 42.6% compliance with anti-glaucoma medications over a minimum of six months in urban Ethiopia, using the Morisky Medication Adherence-8 scale, a validated tool for assessing compliance. Another study done in a tertiary hospital in Ethiopia reported a higher rate of adherence, 61.4%, with medical drop therapy based on self-report.¹⁰⁰ This study, however, recruited patients with varying lengths of follow-up, which ranged from 2 weeks to 168 months. Tamrat et al.⁹⁷ reported a rate of 32.5% in South-West Ethiopia using a combination of physician notes and self-report in patients who had been on medications for at least six months. In Togo, West Africa, Santos et al.⁹⁸ reported that 10.3% self-reported good adherence, while 65.4% had fair adherence with glaucoma medications. The study was done in four clinics in Lomé Togo. Different methods were used in these studies to determine adherence to medications resulting in a wide range of values. The study by Mehari et al.⁹⁶ shall be used to populate the cost-effectiveness model since a validated questionnaire was used.

Utilities and quality-adjusted life-years (QALY): QALYs shall be calculated using utilities for each glaucoma stage. Mild POAG shall be assumed to be 0.80 anf 0.75 for moderate glaucoma and 0.71 for severe glaucoma based on the work of Burr and colleagues.²⁸

Outcome: Incremental cost utility ratios (ICURs) and incremental cost-effectiveness ratios (ICERs) are the main outcome measures. The ICUR is the difference in the total QALYs between the two cohorts of the screened and the unscreened, while the ICER is the difference in the years of blindness avoided by the two cohorts. The outcome shall be calculated as the difference in total costs between the screened and the unscreened cohorts divided by the difference in the total QALYs between the two cohorts (ICUR) or by the years of blindness avoided between the two cohorts (ICER) from the societal and health care perspective.

Table 2 shows the summary of data necessary to populate a cost effeciveness model

Table 2. Summary of reviewed data necessary to populate a cost-effectiveness model.

Data to populate Model	Review questions and Objectives	Findings and Interpretation
Prevalence of glaucoma in Sub- Saharan Africa (SSA)	What is the prevalence of glaucoma in population based studies in Africa?	Prevalence of glaucoma in SSA is very high compared to other populations. It ranges from 4.3% – 7.2%
Prevalence of glaucoma among African descents	What is the prevalence of glaucoma in population based studies among African descents?	Prevalence range of 7.0–9.6%
Severity of glaucoma at presentation	What is the severity of glaucoma at presentation in SSA?	Unilateral blindness 15.0–52.2% Bilateral blindness: 18.0–56.0%.
Compliance with follow up after screening	What proportion of patients have follow up care after testing positive for glaucoma during screening outreach?	17.6% and 32.5%.*
Frequency Doubling Technique	How does FDT perform in early and moderate glaucoma?	96% sensitivity and 96% specificity for moderate glaucoma and 85% sensitivity and 90% specificity for early glaucomatous visual field loss
Clinical effectiveness of glaucoma treatment	How effectively can glaucoma be managed in SSA?	0.66 (95% CI of 0.61–0.71) pooled estimates for medical treatment. Surgical treatment (trabeculectomy): 0.73 (95%CI 0.65–0.80). Lasers: 0.39 (95% CI 0.27–0.54)
Cost of treatment	What is the cost of treating glaucoma across SSA?	Medical: A range of USD 273–483. Surgical: USD283.

In press

There are limitations to this review. This was not a systematic review and therefore, we could not perform a meta-analysis nor calculate estimates and biases. The authors, however, included all available published data that met the inclusion criteria and were relevant to populate the cost-effectiveness screening model

Conclusion: There is a gap in the understanding of the cost-effectiveness of glaucoma in SSA. This review has been conducted to populate a proposed cost-effectiveness analysis model with the aim of determining the cost effectiveness and the cost utility of glaucoma screening in sub-Saharan Africa.

Disclosure statement

Professor Congdon is Director of Research for Orbis International, a non-governmental organization working to promote eye health, including programs for glaucoma control in Africa and elsewhere. The other authors have no conflicts of interest

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.