Prevalence and risk factors for visual impairment among elderly patients attending the eye clinic at Mulago National Referral Hospital, Uganda: a cross-sectional study

Background The elderly have an increased risk of developing visual impairment (VI). Due to the increase in life expectancy of individuals in Sub-Saharan Africa, the population of the elderly is projected to increase. It is thus postulated that the prevalence of VI will increase which is currently unknown in Uganda. Objective To determine the prevalence and risk factors for VI among the elderly at Mulago National Referral Hospital eye clinic in Uganda. Methods This was a cross-sectional study carried out in 2020 with consecutive enrolment of patients aged 60 years and above. Obtaining history was followed by systemic and ocular examination. Statistical analysis was performed to determine the prevalence and factors associated with VI. Results Of 346 elderly participants examined, 174 (50.3%) were males and median age was 67 (IQR 63–74). Prevalence of VI was 32.1%. Cataract was the leading cause of blindness 54.1%, followed by refractive error (21.6%), glaucoma (11.7%), and corneal opacities (5.4%). Age (adjusted Prevalence Ratio (aPR): 1.05, 95% CI (1.02, 1.06)), history of diabetes mellitus (aPR 1.46, 95%CI (1.04, 2.05)), history of hypertension (aPR 1.46, 95%CI (1.10, 1.93)), having completed primary level of education (aPR 0.74, 95%CI (0.55, 0.98)) and secondary level of education (aPR 0.47, 95%CI (0.30,0.73)), presence of a cataract at examination (aPR: 2.28, 95%CI (1.66, 3.13)) were statistically significantly associated with VI. Conclusion In Mulago hospital, the prevalence of VI among the elderly is high with majority of the causes being correctable. We recommend that efforts towards early case identification of causes of VI among the elderly should be a priority.


Introduction
Visual impairment (VI) results in loss of economic and educational opportunities, reduced quality of life, and increased risk of falls and death 1,2 . People with VI suffer due to increased morbidity and mortality 3 . Globally, VI is caused by common eye diseases, such as glaucoma, cataracts, and age-related macular degeneration-are related to aging. 4-6 thus people aged 50 years and above contribute 82% of the global burden of blindness 1 . In Uganda, the population of older persons (60 and above) is estimated to be 1.6 million and is expected to increase to 5.5 million by 2050 7 . An increase in the elderly population will increase the burden of non-communicable diseases such as blindness 8 .
The Global Eye Health Plan action plan, 2014-2019 stresses the need to undertake epidemiological surveys on VI at regular intervals both nationally and sub-nationally to generate evidence on the magnitude and causes of VI 9 . Unfortunately, there is a scarcity of data on the prevalence and factors associated with VI among the elderly population of Uganda. This makes advocating and dethe prevalence of VI and the associated factors among the elderly attending Mulago National Referral Hospital eye clinics in Uganda.

Material and Methods Study Design
This was a cross-sectional study.

Study setting
This study was conducted among patients aged 60 years and above at Mulago National Referral Hospital in Uganda carried out during February and March 2020. Mulago Hospital is one of the six national referral hospitals in Uganda and has the largest ophthalmology department in terms of staff, equipment, and patient attendance. It also doubles as the teaching hospital for Makerere University. The ophthalmology department has two eye clinics: a ma level) and a consultation clinic ran by ophthalmologists. Both clinics run from Monday to Friday, with an estimated daily elderly patient attendance of 10 due to lockdown restrictions during the COVID-19 pandemic. This is half of the number of elderly patients seen before the pandemic. The study participants were recruited from both eye clinics.
ly 10 . We excluded patients with known allergies to drops (cyclopentolate and amethocaine) or Fluorescein stain used in the study for ocular examination and those who were too sick to withstand the rigor of a full ocular exam were excluded. Sample size and sampling of the study population We consecutively sampled patients attending the eye clinic. We used a sample size of 346 derived using a desired precision of 0.05 and assuming a VI prevalence of 34.2% 11 .

Data sources/ measurement
A pretested structured questionnaire administered by the researcher or trained research assistants (Ophthalmology Participants' demographic characteristics, social, ocular, and medical history were collected. A detailed ocular examination was done by the principal investigator starting with the right eye then the left eye. This included: distance visual acuity using a 6 m Snellen's chart or illiterate E chart; those with visual acuity (V/A) less than 6/6 were reassessed with a pinhole and then refracted with an autorefractometer. Near vision was then assessed using a Jaeger chart, and then refraction was done were assessed by the confrontational method compared assessed; the cover-uncover test was done to assess for phoria. Diplopia was sought for in all directions of gaze. Amsler grid was done in all subjects to assess macular function. Examination of the lids, conjunctiva, cornea, anterior chamber, pupil, and iris was done using a slit lamp. Tonometry using Perkin's applanation tonometer, after instilling an anesthetic drop (tetracaine) and staining all respondents. Dilating of the pupil was done using cyclopentolate eye drops and then indirect ophthalmoscopy was performed in study participants with a visual acuity less than 6/6. Investigations were determined on an individual basis to aid in achieving the study objectives and stain, X-ray of the orbit, ultra sound scan, and computerized tomography.
-ty in the better eye worse than 6/12 in the elderly. VI was further graded as mild, moderate, severe, and blindness according to the extent of visual acuity as follows: • Mild VI: Presenting visual acuity worse than 6/12 to 6/18 • Moderate VI: Presenting visual acuity worse than 6/18 to 6/60 • Severe VI: Presenting visual acuity worse than 6/60 to 3/60 • Blindness: Presenting visual acuity worse than 3/60 Any ocular anomaly detected during the patient assessment was both documented and managed where possible, or the relevant specialist was consulted on the course of management and referral.

Statistical analysis
Descriptive statistical measures such as means, standard deviations and medians, interquartile range, frequencies, proportions, and percentages for continuous and categorical variables wherever appropriate, were computed. The prevalence of VI among the elderly attending the eye clinic at Mulago National Referral Hospital was calculated as a proportion of the number of elderly with VI over the total number of elderly enrolled in the study. -son regression model with robust variance estimation were used to estimate the prevalence risk ratios and their 95% CI. The outcome was dichotomized as yes = 1, if one had any degree of VI, and no = 0, if one had no VI.
A forward stepwise multivariate model was constructed -0.05 and a 95% CI that did not cross the null value. Statistical analysis was performed using STATA 15.0 (College Station, Texas, USA).

Characteristics of the study population
A total of 346 elderly participated in the study with an equal distribution of the sexes. The median age was 67 years (inter-quartile range [IQR] of 63-74) with most (57.2%) of the elderly falling in the 60-69 age group. About 4% of the elderly reported a history of trauma while 28% had a history of an eye operation. The commonly reported chronic illnesses were Diabetes (11.3%), (HIV) (2.9%). This is summarized in Table 1 below The most common ocular morbidities were; cataracts (24.82%), refractive error (22.61%), pseudophakia (13.42%) as summarized in Table 2 below.  The common diseases found among the elderly with VI were cataracts at 54.05%, refractive error at 21.62%, glaucoma at 11.71%, corneal opacities at 5.4%, age-related macular degeneration at 1.8% and hypertensive retinopathy at 1.8%. The common diseases found among the elderly with VI were cataracts at 54.05%, refractive error at 21.62%, glaucoma at 11.71%, corneal opacities at 5.4%, age-related macular degeneration at 1.8% and hypertensive retinopathy at 1.8%. The common diseases found among the elderly with VI were cataracts at 54.05%, refractive error at 21.62%, glaucoma at 11.71%, corneal opacities at 5.4%, age-related macular degeneration at 1.8% and hypertensive retinopathy at 1.8%. With every added year, the prevalence of VI increased by 5%. The odds of VI was 44% higher among elders with a history of DM compared to those without. Those with hypertension had 42% higher odds of having VI compared to those without. The elders who had primary education and those with secondary education were 26% and 53%, respectively, less likely to have VI. Presence of cataracts on examination caused an increase in the odds of having VI by 2.3 times.

Discussion
We assessed the prevalence and factors associated with visual impairment (VI) among the elderly attending the eye clinic at Mulago National Referral Hospital. About a third of the elderly had VI with half of these having cataracts. Other conditions found in this study in order of frequency were refractive errors, glaucoma, corneal opacities, age-related macular degeneration, and hypertensive retinopathy. Increasing age, history of diabetes mellitus, history of hypertension, and primary and secondary edu- The prevalence found in our study was higher than studies done in Taiwan (17.7%), Delhi (24.5%), and Afghanistan (22.6%) [12][13][14] . However, these studies were population-based studies. Furthermore, with cataracts being the main cause of VI, the cataract surgical rate in these countries is higher than in Uganda 15,16 . In sub Saharan Africa, it is estimated that only one out of ten cataracts ever gets operated 17 . The prevalence was comparable to a population-based study done in Nigeria that showed the prevalence of VI among pensioners to be 34.2% 11 . In this study, presence of cataract at examination, age, history of diabetes mellitus, history of hypertension, and VI. Similar to our study, several studies have found cataracts to be associated with VI among elders 1,13,[18][19][20] . In low and middle income countries (LMICs), the cataract surgical rate is low thus many people especially the elderly with visually impairing cataracts live with VI as they await cataract surgery 15,16 . This study also found increasing age to be associated with VI which has been documented in several other studies. Increasing age increases ones risk to most of the leading blinding conditions like cataracts, glaucoma, age related macular degeneration and diabetic retinopathy 2, 12, 21-23 . Furthermore, the elders with history of either diabetes or hypertension were more likely to be visually impaired compared to those without. These systemic diseases affect the eyes causing different disorders that affect vision like early onset of cataracts, diabetic retinopathy, hypertensive retinopathy, retinal vascular occlusions [24][25][26] . Studies done among the elderly in other LMICs 21,22 . However, the study participants who were educated were less likely to have Northern Indian 27 . It has been noted that VI prevention and its correction are not frequent in subjects with low education levels and there is also poor compliance with and severity 28 . Though level of income and marital status have been this was not found in this study 18, 29, 30. These studies that probable lack of a support system and thus reduced acbe due to the strong support systems in the African culture where elders are economically and socially taken care of by their children or grandchildren, so the elders' level ence their access to care 31 to a study done among an elderly population in Taiwan where marital status was not statistically associated with VI 12, 32 .
The study has some limitations. This was a cross-sectional study that we couldn't establish temporal associations. The study relied on self-report of information from the study participants, hence there could be information bias since some elders may not honestly disclose some information, for example, level of education, marital status. not be generalizable to the entire general population.

Conclusion
The prevalence of VI among the elderly in Mulago hospital is high with the commonest causes being treatable conditions which include cataracts, refractive errors, glaucoma, corneal opacities and over 76% of the causes can be treated by cataract surgery and correction of refractive errors. Age, history of diabetes mellitus, history of hypertension, education, presence of cataract at examination were found to be associated with VI. We recommend that the elderly with diabetes and hypertension should undergo regular eye examination to detect and manage the causes of VI. Measures to improve the cataract surgical rate in Uganda are needed to address the high prevalence of cataracts.

List of abbreviations aPR adjusted Prevalence Ratio NCDs
Non-communicable diseases COVID Corona virus disease VI Visual Impairment V/A Visual Acuity