The impact of glaucoma on quality of life in Ethiopia: a case-control study

This study was reviewed and approved by the Ethics Committee of the College of Medicine and Health Sciences, University of Gondar, Ethiopia. The nature of the study was explained to potential participants. Individuals were enrolled into the study if they gave informed verbal consent. No names or identifying information were recorded on the questionnaires. Interviews were conducted in private, and all subjects were assured of confidentiality.

Participants were recruited at the Ophthalmology Outpatient Department, University of Gondar Eye Care and Training Centre, Amhara Region, Ethiopia. In this study the cases were consecutive individuals with a diagnosis of glaucoma attending the department between April 24 and May 27, 2015. We excluded individuals with glaucoma who also had (i) high refractive errors (greater than ±5 Diopters), (ii) visually significant cataracts (greater than Stage 2 LOCS III classification) [11], (iii) any other visually significant ocular pathology, (iv) a history of intraocular surgery (incisional or laser) in the preceding 3 months. We also excluded individuals with a major systemic illness or physical disabilities that could affect response to GQoL-15 questionnaire.

The control participants were also recruited from patients attending the same Ophthalmology Outpatient Department. The controls had a visual acuity of 6/9 or better (+/− correction) and otherwise healthy eyes. Pseuodophakic individuals with a visual acuity of 6/9 or better were eligible. We excluded individuals who had (i) high refractive errors (greater than ±5 Diopters), (ii) visually significant lens opacities (greater than Stage 2 LOCS III classification), (iii) suspicious optic discs, or (iv) corrected visual acuity of less than 6/9. We also excluded individuals with a major systemic illness or physical disabilities that could affect response to GQoL-15 questionnaire. The controls were recruited after the case, between February 1 and May 31, 2016. We recruited one control for every four cases; they were frequency matched to the age and sex profile of the cases.

Cases and controls underwent a standard ophthalmic examination as part of their routine clinical care at the department. The presenting distance visual acuity (VA), with spectacle correction if available, was measured for each eye separately using a tumbling-E Snellen visual acuity chart at six metres. Pinhole test was done for those with visual acuity 6/18 or worse. If participants were unable to see this at six metres, the testing distance was reduced to three metres. If they failed to see any letters at three metres, vision was assessed for “counting fingers” (CF), “Hand Movement” (HM) or “Perception of Light” (POL). Presenting distance visual acuity was defined by the revised WHO classification,; patients were categorized according to the vision in the better seeing eye [12]:

The eyes were examined by an ophthalmologist specialised in glaucoma (2 years fellowship training in glaucoma). The anterior segment was assessed using a slitlamp bio-microscope, intraocular pressure (IOP) measured by Goldman tonometry, the anterior chamber angle was examined by gonioscopy and stereoscopic optic nerve head evaluation performed using a 90D (Volk Optical Inc.) indirect lens.

Glaucomatous Optic Neuropathy (GON) was defined based on the scheme developed by the International Society of Geographical and Epidemiological Ophthalmology [13]. Cases were diagnosed if one or more of the following optic nerve head changes were observed in either eye: (1) vertical cup to disc ratio (VCDR) of ≥0.7 or (2) a difference in the VCDR of ≥0.2 between the 2 eyes or (3) the narrowest remaining vertical neuro-retinal rim of ≤0.1 disc diameter. We did not include visual field (VF) assessment in the diagnosis and staging of glaucoma in this study as the test was only performed by a minority of our patients.

The following operational definitions for the sub-types of glaucoma were used [13]:

The severity of glaucoma was staged using the Canadian Ophthalmological Society evidence-based clinical practice guidelines for the management of glaucoma in the adult eye [14]:

A structured, pre-tested questionnaire that consisted of questions on socio-demographic characteristics, clinical assessment of glaucoma and the Glaucoma Quality of Life-15 (GQL-15) questionnaire was used to collect data [15]. The questionnaire was translated into the local language (Amharic) and then back translated into English by expertise and 5% of the questionnaire was pre-tested. Data were collected through face to face interviews (most study participants were illiterate) and medical chart review.

The GQL-15 is a user friendly 15-item questionnaire made up of four domains of perceived visual disability: (1) central/near vision, (2) outdoor mobility, (3) peripheral vision and (4) glare / dark adaptation [16]. There are different numbers of question items in the four domains: central/near vision (2 items), outdoor mobility (1 item), peripheral vision (6 items) and glare / dark adaptation (6 items). Therefore the maximum potential score the domains also vary. It is a well validated tool with high internal consistency and test-retest reliability [15, 17]. The GQL-15 has a scoring scale ranging from 1 to 5, where 1 stands for “no difficulty in performing the activity” and 5 for “severe difficulty due to visual reasons”. If patients do not perform a specific activity for reasons other than impaired vision, a score of 0 is given for that task. This can yield a total score of between 0 and 75. Poorer GQL-15 and increasing difficulty with vision-related activities were associated with higher subscale scores [17‐20].

The coded data were checked, cleaned and entered into Epi Info version 7 and then exported into SPSS Version 20 (SPSS, Chicago III) for analysis. Frequencies, percentage, mean and standard deviation were used to describe study results as required. To calculate the subscale scores for the four domains of the GQL-15, the item level responses were scored on a numerical interval scale ranging from 0, indicating no difficulty, to 5, indicating severe difficulty. The subscale score for each domain was calculated using an average of the scores generated for the component item-level responses. Comparisons of cases and controls were done after adjusting variables for age and sex. Linear regression models and the t-test were employed to compare significant differences in GQL-15 scores and to generate mean and mean differences between cases and controls in each GQL-15 subscale and domain, respectively. Univariable analysis was first conducted for each variable. Variables that satisfied p-value <0.2 were selected for further analysis. The strength of association was interpreted using confidence interval. P-value <0.05 was accepted as statistically significant in this study.

We recruited 307 glaucoma patients (cases) and 76 controls. The socio-demographic characteristics of the study participants are presented in Table 1. The age (p = 0.62) and sex (p = 0.87) distribution of the cases and controls were highly comparable. The mean age of the cases was 60.3 years (SD ± 14.0) and controls was 59.5 years (SD ±12.5). The large majority were married, with no evidence of a difference in marital status between groups (p = 0.06). Most were of Amhara ethnicity. Overall, the majority of participants lived in rural areas, however, there were proportionately more people from rural areas among the controls (75.0%) compared to the cases (55.4%) (p = 0.02). Educational level and occupation types were comparable between cases and controls. There was some evidence of a difference in income, with the controls earning slightly more than the cases (p = <0.0001).

The clinical characteristics of the cases are shown in Table 2. The majority had been diagnosed for between 1 and 5 years (194, 63.2%) and had bilateral disease (279, 90.9%). One hundred and thirteen (36.8%) had advanced glaucoma (determined by the extent of optic nerve damage) and 70 (22.8%) cases were either severely visual impaired or blind. The most frequent types of glaucoma were primary open angle glaucoma (164, 53.4%) and pseudoexfoliative glaucoma (111, 36.2%). Most of the control groups had a visual acuity of 6/6 (42, 55.3%) while the rest had 6/9. The main clinical diagnoses in the control groups include early cataract (24, 31.6%), corrected refractive errors (21, 27.6%) and pseudophakia with good outcome (16, 21.1%).

The mean GQOL-15 scores of glaucoma cases and the controls are presented in Table 3. The mean GQL-15 score was substantially higher (indicating poorer quality of life) among the glaucoma cases than the controls (46.3 vs 18.6, p < 0.0001). The glaucoma cases had higher scores than the controls for all four domains (p < 0.0001).

The relationship between GQL-15 scores and demographic characteristics among individuals with glaucoma are presented in Table 4. A higher total GQL-15 and domain scores (poorer QOL) were found for men, older people, illiterate individuals, rural residents, and people with lower monthly income. These same factors remained significant after adjustment in a multivariable linear regression model. The relationship between GQL-15 scores and clinical characteristics among the glaucoma cases are presented in Table 5. Increasing severity of glaucoma, longer duration of glaucoma and decreasing visual acuity were all associated with higher GQL-15 scores.

We performed a sub group analysis comparing cases with no or only mild VI to the controls [Table 6]. The mean GQL-15 scores for this sub-group of glaucoma cases still had worse QOL than the controls. Similarly, we compared the sub-group of cases with early glaucoma to the controls [Table 7]. This sub-group of cases had worse GQL-15 scores than the controls.