A practical model for effective eye care delivery in Southeast Asian rural communities: A proposal built based on experts' recommendations.

PURPOSE: To evaluate rural community-based eye care models from the perspective of community ophthalmology experts and suggest sustainable technological solutions for enhancing rural eye care delivery. METHODS: A semi-structured descriptive survey, using close-ended and open-ended questions, was administered to the experts in community ophthalmology sourced through purposive sampling. The survey was self-administered and was facilitated through online platforms or in-person meetings. Uniform questions were presented to all participants, irrespective of their roles. RESULTS: Surveyed participants (n = 22 with 15 from India and 7 from Nepal) in high-volume tertiary eye hospitals faced challenges with resources and rural outreach. Participants had mixed satisfaction with pre-operative screening and theatre resources. Delayed presentations and inexperienced surgeons contributed to the surgery complications. Barriers to rural eye care included resource scarcity, funding disparities, and limited infrastructure. In rural/peri-urban areas 87% of participants agreed with providing primary eye care services, with more than 60% of the experts not in agreement with the makeshift center model of eye care delivery. Key components for an effective eye care model are sustainability, accessibility, affordability, and quality. These can be bolstered through a healthcare management platform and a human-chain supply distribution system. CONCLUSION: Tailored interventions are crucial for rural eye care, emphasizing the need for stronger human resources, optimized funding, and community awareness. Addressing challenges pertinent to delayed presentation and surgical training is vital to minimizing complications, especially with advanced cataracts. Enhancements in rural eye care demand a comprehensive approach prioritizing accessibility, affordability, and consistent quality.

The Better Operative Outcomes Software Tool (BOOST) Prospective Study: Improving the Quality of Cataract Surgery Outcomes in Low-Resource Settings.

PURPOSE: Post-operative vision impairment is common among patients who have undergone cataract surgery in low-resource settings, impacting quality of clinical outcomes and patient experience. This prospective, multisite, single-armed, pragmatic validation study aimed to assess whether receiving tailored recommendations via the free Better Operative Outcomes Software Tool (BOOST) app improved surgical outcomes, as quantified by post-operative unaided distance visual acuity (UVA) measured 1-3 days after surgery. METHODS: During the baseline data collection round, surgeons in low and middle-income countries recorded clinical characteristics of 60 consecutive cataract cases in BOOST. Additional data on the causes of poor outcomes from 20 consecutive cases with post-operative UVA of <6/60 (4-12 weeks post-surgery) were entered to automatically generate tailored recommendations for improvement, before 60 additional consecutive cases were recorded during the follow-up study round. Average UVA was compared between cases recorded in the baseline study round and those recorded during follow-up. RESULTS: Among 4,233 cataract surgeries performed by 41 surgeons in 18 countries, only 2,002 (47.3%) had post-operative UVA 6/12 or better. Among the 14 surgeons (34.1%) who completed both rounds of the study (1,680 cases total), there was no clinically significant improvement in post-operative average UVA (logMAR units ±SD) between baseline (0.50 ± 0.37) and follow-up (0.47 ± 0.36) rounds (mean improvement 0.03, p = 0.486). CONCLUSIONS: Receiving BOOST-generated recommendations did not result in improved UVA beyond what could be expected from prospective monitoring of surgical outcomes alone. Additional research is required to assess whether targeted support to implement changes could potentiate the uptake of app-generated recommendations and improve outcomes.

Diagnostic Accuracy of Artificial Intelligence-Based Automated Diabetic Retinopathy Screening in Real-World Settings: A Systematic Review and Meta-Analysis.

PURPOSE: To evaluate the diagnostic accuracy of artificial intelligence (AI)-based automated diabetic retinopathy (DR) screening in real-world settings. DESIGN: Systematic review and meta-analysis METHODS: We conducted a systematic review of relevant literature from January 2012 to August 2022 using databases including PubMed, Scopus and Web of Science. The quality of studies was evaluated using Quality Assessment for Diagnostic Accuracy Studies 2 (QUADAS-2) checklist. We calculated pooled accuracy, sensitivity, specificity, and diagnostic odds ratio (DOR) as summary measures. The study protocol was registered with the International Prospective Register of Systematic Reviews (PROSPERO - CRD42022367034). RESULTS: We included 34 studies which utilized AI algorithms for diagnosing DR based on real-world fundus images. Quality assessment of these studies indicated a low risk of bias and low applicability concern. Among gradable images, the overall pooled accuracy, sensitivity, specificity, and DOR were 81%, 94% (95% CI: 92.0-96.0), 89% (95% CI: 85.0-92.0) and 128 (95% CI: 80-204) respectively. Sub-group analysis showed that, when acceptable quality imaging could be obtained, non-mydriatic fundus images had a better DOR of 143 (95% CI: 82-251) and studies using 2 field images had a better DOR of 161 (95% CI 74-347). Our meta-regression analysis revealed a statistically significant association between DOR and variables such as the income status, and the type of fundus camera. CONCLUSION: Our findings indicate that AI algorithms have acceptable performance in screening for DR using fundus images compared to human graders. Implementing a fundus camera with AI-based software has the potential to assist ophthalmologists in reducing their workload and improving the accuracy of DR diagnosis.

Evaluation and model to achieve sex parity in cataract surgical coverage in Theni district, India.

BACKGROUND/AIMS: To propose an approach to determine the target ratio of cataract surgical rates (CSRs) of female to male subpopulations to increase sex parity in cataract surgical coverage (CSC), based on the sex gap in cataract burden and incidence, and demonstrate its application to Theni district, India. METHODS: A population-based longitudinal study between January 2016 and April 2018. We recruited 24 327 participants using random cluster sampling. We conducted detailed eye examinations of 7087 participants aged ≥40 years (4098 females, 2989 males). We fit exponential models to the age-specific and sex-specific cataract burden and estimated annual incidence rates. We developed a spreadsheet-based planning tool to compute the target CSR ratio of female to male subpopulations. RESULTS: Among those aged ≥40 years, cataract burden was 21.4% for females and 17.5% for males (p<0.05). CSC was 73.9% for females versus 78.6% for males (p<0.05), with an effective CSC of 52.6% for females versus 57.6% for males (p<0.05). Treating only incident cataracts each year requires a target CSR ratio of female to male subpopulations of 1.30, while addressing in addition 10% of the coverage backlog for females and 5% for males requires a target CSR ratio of 1.48. CONCLUSIONS: The female population in Theni district, as in many low-income and middle-income countries, bears a higher cataract burden and lower CSC. To enhance sex parity in coverage, both the higher number of annual incident cataracts and the larger backlog in females will need to be addressed.

Comparison of cataract surgery outcomes between a secondary and a tertiary eye hospital in Tamil Nadu, India.

PURPOSE: To examine the difference in post-operative visual outcomes of cataract surgeries between a tertiary and secondary centre at Aravind Eye Hospitals in Tamil Nadu, India. METHODS: Our retrospective cross-sectional study analysed cataract surgeries at a secondary centre and its associated tertiary centre in 2021. Our main outcome measures were postoperative best corrected visual acuity (BCVA), spherical equivalent, and intraoperative and postoperative complications. Two-sample proportion tests and logistic regression analyses were performed. RESULTS: The analysis of 32,302 cataract surgeries in 2021 of which 4357 were performed at the secondary centre and 27,945 were performed at the tertiary centre showed that the tertiary centre operated on more advanced cataract condition (p < 0.001). Intraoperative (p < 0.001) and post-operative complication rates (p < 0.001) were higher in the tertiary centre. The odds of effective outcomes (BCVA > = 6/12) controlling for all covariates are poorer (p < 0.0001) in the tertiary centre for both phacoemulsification (phaco) and manual small incision cataract surgeries (MSICS). CONCLUSION: World Health Organization recommendations for the effective outcome of cataract surgery are met by both the tertiary and secondary centres, but the odds of effective outcomes in the tertiary centre was lower after adjusting for all known factors. Further investigations of the causes of poor vision in both phaco and MSICS in the tertiary centre are needed to improve the situation.

Factors Affecting Follow-Up Adherence of Glaucoma Patients.

PURPOSE: The objective of our study is to understand the factors associated with non-adherence to the physician's follow-up advice by persons with glaucoma. METHODS: We use a four-year panel dataset containing demographic, clinical, and intervention details and doctor's advised follow-up date for a random sample of 2,622 glaucoma patients from an Indian tertiary eye care hospital. We model this unique "advised follow-up date" in multivariate logistic regressions to identify factors associated with patients' coming early or late as against on-time. An OLS regression also examines an association between delayed follow-up and declining visual acuity. RESULTS: Demographic variables such as age, gender, distance to hospital, IOP, and visual acuity were not associated with delayed adherence to follow-up. Patients advised to review within 2 months (p < .001), paying patients (p < .001), and those habituated to routine follow-up (diabetes patients) (p < .01) are less likely to delay care-seeking. Patients are more likely to come on-time in visits immediately after clinical interventions relative to later ones (p < .001). Our second analysis reveals the presence of cataract and the very first post-surgery visit as factors influencing patients to come early. Our third analysis reveals that a higher proportion of delayed visits had resulted in worsening of vision in the glaucoma-affected eye. CONCLUSION: Our results suggest that active counselling by doctors, especially for routine follow-up visits, can help in better follow-up adherence and prevention of glaucoma-related visual impairment.

Cataract surgery workload estimates in Theni district, India.

BACKGROUND/AIMS: To estimate the annual cataract surgery workload in Theni district, Tamil Nadu, India based on current utilisation of cataract services, prevalence of blindness and vision impairment (VI), and cataract burden-reduction goals. METHODS: We conducted a population-based longitudinal study between January 2016 and April 2018. We recruited 24 327 participants based on a random cluster sampling method; 7127 participants were ≥40 years. During the year following initial enrolment, we tracked utilisation of eye care services; and at the end of the 1-year period, we conducted a detailed eye examination of participants age ≥40. RESULTS: In the sample age ≥40 years, 13.0% had a visually significant cataract, and 17.8% had prior cataract surgery in at least one eye. The prevalence of cataract blindness based on presenting visual acuity in the better eye (PVABE)<3/60 was 0.34% and VI (PVABE<6/12) was 9.92%. 3.10% of the study population had obtained cataract surgery during 1 year, resulting in a cataract surgical rate of 9085. We estimated the effective cataract surgical coverage (eCSC) to be 54.5% and the CSC to be 75.7%, implying a sizeable quality gap. Prevalence, utilisation and coverage varied by age and gender. We estimated that a goal of eliminating the backlog of VI (PVABE<6/12) in 5 years would increase the annual cataract surgery workload by 11.5% from the current level. CONCLUSIONS: Our estimates of cataract surgery workloads under different scenarios can provide a useful input into planning of eye health services in Theni district.

Impact of practicing internal benchmarking on continuous improvement of cataract surgery outcomes: a retrospective observational study at Aravind Eye Hospitals, India.

OBJECTIVE: We aim to assess the effectiveness of a cataract surgery outcome monitoring tool used for continuous quality improvement. The objectives are to study: (1) the quality parameters, (2) the monitoring process followed and (3) the impact on outcomes. DESIGN AND PROCEDURES: In this retrospective observational study we evaluated a quality improvement (QI) method which has been practiced at the focal institution since 2012: internal benchmarking of cataract surgery outcomes (CATQA). We evaluated quality parameters, procedures followed and clinical outcomes. We created tables and line charts to examine trends in key outcomes. SETTING: Aravind Eye Care System, India. PARTICIPANTS: Phacoemulsification surgeries performed on 718 120 eyes at 10 centres (five tertiary and five secondary eye centres) from 2012 to 2020 were included. INTERVENTIONS: An internal benchmarking of surgery outcome parameters, to assess variations among the hospitals and compare with the best hospital. OUTCOME MEASURES: Intraoperative complications, unaided visual acuity (VA) at postoperative follow-up visit and residual postoperative refractive error (within ±0.5D). RESULTS: Over the study period the intraoperative complication rate decreased from 1.2% to 0.6%, surgeries with uncorrected VA of 6/12 or better increased from 80.8% to 89.8%, and surgeries with postoperative refractive error within ±0.5D increased from 76.3% to 87.3%. Variability in outcome measures across hospitals declined. Additionally, benchmarking was associated with improvements in facilities, protocols and processes. CONCLUSION: Internal benchmarking was found to be an effective QI method that enabled the practice of evidence-based management and allowed for harnessing the available information. Continuous improvement in clinical outcomes requires systematic and regular review of results, identifying gaps between hospitals, comparisons with the best hospital and implementing lessons learnt from peers.

Cataract services for all: Strategies for equitable access from a global modified Delphi process.

Vision loss from cataract is unequally distributed, and there is very little evidence on how to overcome this inequity. This project aimed to engage multiple stakeholder groups to identify and prioritise (1) delivery strategies that improve access to cataract services for under-served groups and (2) population groups to target with these strategies across world regions. We recruited panellists knowledgeable about cataract services from eight world regions to complete a two-round online modified Delphi process. In Round 1, panellists answered open-ended questions about strategies to improve access to screening and surgery for cataract, and which population groups to target with these strategies. In Round 2, panellists ranked the strategies and groups to arrive at the final lists regionally and globally. 183 people completed both rounds (46% women). In total, 22 distinct population groups were identified. At the global level the priority groups for improving access to cataract services were people in rural/remote areas, with low socioeconomic status and low social support. South Asia and Sub-Saharan Africa were the only regions in which panellists ranked women in the top 5 priority groups. Panellists identified 16 and 19 discreet strategies to improve access to screening and surgical services, respectively. These mostly addressed health system/supply side factors, including policy, human resources, financing and service delivery. We believe these results can serve eye health decision-makers, researchers and funders as a starting point for coordinated action to improve access to cataract services, particularly among population groups who have historically been left behind.

Effective cataract surgical coverage in adults aged 50 years and older: estimates from population-based surveys in 55 countries.

BACKGROUND: Cataract is the leading cause of blindness globally. Effective cataract surgical coverage (eCSC) measures the number of people in a population who have been operated on for cataract, and had a good outcome, as a proportion of all people operated on or requiring surgery. Therefore, eCSC describes service access (ie, cataract surgical coverage, [CSC]) adjusted for quality. The 74th World Health Assembly endorsed a global target for eCSC of a 30-percentage point increase by 2030. To enable monitoring of progress towards this target, we analysed Rapid Assessment of Avoidable Blindness (RAAB) survey data to establish baseline estimates of eCSC and CSC. METHODS: In this secondary analysis, we used data from 148 RAAB surveys undertaken in 55 countries (2003-21) to calculate eCSC, CSC, and the relative quality gap (% difference between eCSC and CSC). Eligible studies were any version of the RAAB survey conducted since 2000 with individual participant survey data and census population data for people aged 50 years or older in the sampling area and permission from the study's principal investigator for use of data. We compared median eCSC between WHO regions and World Bank income strata and calculated the pooled risk difference and risk ratio comparing eCSC in men and women. FINDINGS: Country eCSC estimates ranged from 3·8% (95% CI 2·1-5·5) in Guinea Bissau, 2010, to 70·3% (95% CI 65·8-74·9) in Hungary, 2015, and the relative quality gap from 10·8% (CSC: 65·7%, eCSC: 58·6%) in Argentina, 2013, to 73·4% (CSC: 14·3%, eCSC: 3·8%) in Guinea Bissau, 2010. Median eCSC was highest among high-income countries (60·5% [IQR 55·6-65·4]; n=2 surveys; 2011-15) and lowest among low-income countries (14·8%; [IQR 8·3-20·7]; n=14 surveys; 2005-21). eCSC was higher in men than women (148 studies pooled risk difference 3·2% [95% CI 2·3-4·1] and pooled risk ratio of 1·20 [95% CI 1·15-1·25]). INTERPRETATION: eCSC varies widely between countries, increases with greater income level, and is higher in men. In pursuit of 2030 targets, many countries, particularly in lower-resource settings, should emphasise quality improvement before increasing access to surgery. Equity must be embedded in efforts to improve access to surgery, with a focus on underserved groups. FUNDING: Indigo Trust, Peek Vision, and Wellcome Trust.

Strategies for cataract and uncorrected refractive error case finding in India: Costs and cost-effectiveness at scale.

Background: India has the largest number of individuals suffering from visual impairment and blindness in the world. Recent surveys indicate that demand-based factors prevent more than 80% of people from seeking appropriate eye services, suggesting the need to scale up cost-effective case finding strategies. We assessed total costs and cost-effectiveness of multiple strategies to identify and encourage people to initiate corrective eye services. Methods: Using administrative and financial data from six Indian eye health providers, we conduct a retrospective micro-costing analysis of five case finding interventions that covered 1·4 million people served at primary eye care facilities (vision centers), 330,000 children screened at school, 310,000 people screened at eye camps and 290,000 people screened via door-to-door campaigns over one year. For four interventions, we estimate total provider costs, provider costs attributable to case finding and treatment initiation for uncorrected refractive error (URE) and cataracts, and the societal cost per DALY averted. We also estimate provider costs of deploying teleophthalmology capability within vision centers. Point estimates were calculated from provided data with confidence intervals determined by varying parameters probabilistically across 10,000 Monte Carlo simulations. Findings: Case finding and treatment initiation costs are lowest for eye camps (URE: $8·0 per case, 95% CI: 3·4-14·4; cataracts: $13·7 per case, 95% CI: 5·6-27·0) and vision centers (URE: $10·8 per case, 95% CI: 8·0-14·4; cataracts: $11·9 per case, 95% CI: 8·8-15·9). Door-to-door screening is as cost-effective for identifying and encouraging surgery for cataracts albeit with large uncertainty ($11·3 per case, 95% CI: 2·2 to 56·2), and more costly for initiating spectacles for URE ($25·8 per case, 95% CI: 24·1 to 30·7). School screening has the highest case finding and treatment initiation costs for URE ($29·3 per case, 95% CI: 15·5 to 49·6) due to the lower prevalence of eye problems in school aged children. The annualized cost of operating a vision center, excluding procurement of spectacles, is estimated at $11,707 (95% CI: 8,722-15,492). Adding teleophthalmology capability increases annualized costs by $1,271 per facility (95% CI: 181 to 3,340). Compared to baseline care, eye camps have an incremental cost-effectiveness ratio (ICER) of $143 per DALY (95% CI: 93-251). Vision centers have an ICER of $262 per DALY (95% CI: 175-431) and were able to reach substantially more patients than any other strategy. Interpretation: Policy makers are expected to consider cost-effective case finding strategies when budgeting for eye health in India. Screening camps and vision centers are the most cost-effective strategies for identifying and encouraging individuals to undertake corrective eye services, with vision centers likely to be most cost-effective at greater scale. Investment in eye health continues to be very cost-effective in India. Funding: The study was funded by the Seva Foundation.

Use of predictive models to identify patients who are likely to benefit from refraction at a follow-up visit after cataract surgery.

PURPOSE: To develop predictive models to identify cataract surgery patients who are more likely to benefit from refraction at a four-week postoperative exam. METHODS: In this retrospective study, we used data of all 86,776 cataract surgeries performed in 2015 at a large tertiary-care eye hospital in India. The outcome variable was a binary indicator of whether the difference between corrected distance visual acuity and uncorrected visual acuity at the four-week postoperative exam was at least two lines on the Snellen chart. We examined the following statistical models: logistic regression, decision tree, pruned decision tree, random forest, weighted k-nearest neighbor, and a neural network. Predictor variables included in each model were patient sex and age, source eye (left or right), preoperative visual acuity, first-day postoperative visual acuity, intraoperative and immediate postoperative complications, and combined surgeries. We compared the predictive performance of models and assessed their clinical impact in test samples. RESULTS: All models demonstrated predictive accuracy better than chance based on area under the receiver operating characteristic curve. In a targeting exercise with a fixed intervention budget, we found that gains from predictive models in identifying patients who would benefit from refraction ranged from 7.8% (increase from 1500 to 1617 patients) to 74% (increase from 250 to 435 patients). CONCLUSION: The use of predictive statistical models to identify patients who are likely to benefit from refraction at follow-up can improve the economic efficiency of interventions. Simpler models like logistic regression perform almost as well as more complex machine-learning models, but are easier to implement.

Keeping an eye on eye care: monitoring progress towards effective coverage.

The eye care sector is well positioned to contribute to the advancement of universal health coverage within countries. Given the large unmet need for care associated with cataract and refractive error, coupled with the fact that highly cost-effective interventions exist, we propose that effective cataract surgery coverage (eCSC) and effective refractive error coverage (eREC) serve as ideal indicators to track progress in the uptake and quality of eye care services at the global level, and to monitor progress towards universal health coverage in general. Global targets for 2030 for these two indicators were endorsed by WHO Member States at the 74th World Health Assembly in May, 2021. To develop consensus on the data requirements and methods of calculating eCSC and eREC, WHO convened a series of expert consultations to make recommendations for standardising the definitions and measurement approaches for eCSC and eREC and to identify areas in which future work is required.

Differences Between Male and Female Residents in Case Volumes and Learning in Cataract Surgery.

OBJECTIVE: It is important that ophthalmology residency programs are structured to provide opportunities for all residents to gain adequate surgical experience. The aim of this study is to examine differences between female and male residents in number of cataract surgeries performed, surgical outcomes, and rate of learning, and to draw implications for the training of ophthalmology residents. DESIGN: In this hospital-based retrospective cohort study we examined the cumulative number of surgeries performed, and the intraoperative complications rate, of male and female residents over time. We used multivariable regression to measure the association between cataract surgery volumes and resident gender, controlling for covariates. We compared the intraoperative complications rates of female and male residents in each quarter of the program and overall, and estimated a multivariable logistic model of intraoperative complications with resident gender, surgical experience, their interaction, and other predictors. SETTING: Three-year residency programs at 5 Aravind Eye Hospitals in India. PARTICIPANTS: Residents entering in 2012, 2013, 2014, or 2015. RESULTS: In our sample of 95 residents (64 females, 31 males), a male resident on average performed 662.2 surgeries, which was 26.9% (p < 0.05) more than the 521.8 surgeries performed by a female resident. The difference in cumulative case volume grew steadily during the residency, and was not explained by the 29 more days of leave taken by female residents. On average a male resident performed 21% more surgeries per surgical turn than a female resident. In each quarter female residents had a lower cumulative rate of intraoperative complications than males, but the differences were not statistically significant. Learning curve analysis showed that the effect of experience on intraoperative complications flattened at about 300 surgeries, and that the curves for male and female residents were not statistically different. CONCLUSION: Female residents performed substantially fewer surgeries than their male counterparts in our data. However, surgical outcomes were comparable between groups with complication rates flattening after 300 surgeries for all trainees.

Seven-year trends in cataract surgery indications and quality of outcomes at Aravind Eye Hospitals, India.

PURPOSE: To report trends in cataract surgery indications, visual acuity outcomes, complication rates and reoperation rates at ten Aravind Eye Hospitals in Tamil Nadu, India. METHODS: In this retrospective database study we analyzed seven years of data of cataract surgeries at Aravind during January 1, 2012 to December 31, 2018. Our main outcome measures were preoperative and 1-month postoperative uncorrected visual acuity (UCVA) and best-corrected visual acuity (BCVA); and rates of intraoperative complications and reoperations. We performed Cochran-Armitage trend tests. RESULTS: In 1.86 million eyes that underwent cataract surgery, the percentage of eyes undergoing phacoemulsification (PE) showed an increasing trend (p < 0.0001), from 26.8% in 2012 to 33.5% in 2018. The percentage of eyes undergoing manual small incision cataract surgery (SICS) showed a decreasing trend (p < 0.0001), from 70.0% in 2012 to 65.1% in 2018. For SICS, the percentages of eyes with good postoperative UCVA and BCVA showed increasing trends (p < 0.0001 for each), and the percentages of eyes with poor postoperative UCVA and BCVA showed decreasing trends (p < 0.0001 for each). Results for PE surgeries were mixed. For both PE and SICS rates of intraoperative complications showed a declining trend (p < 0.0001 for each) and a slight increasing trend for reoperations (PE p < 0.02; SICS p < 0.0001). CONCLUSIONS: During this seven-year period there were noticeable trends in cataract surgeries performed at Aravind in terms of the mix of surgical procedures, preoperative vision, and cataract surgical outcomes.

Predictors of patient compliance with follow-up visits after cataract surgery.

PURPOSE: To identify patient characteristics associated with follow-up compliance and to draw implications for better cataract treatment services. SETTING: Aravind Eye Hospital, Madurai, India. DESIGN: Retrospective case study. METHODS: The data of all cataract surgeries performed in 2015 were analyzed. After each surgery, patients were asked to return for follow-up after 1 month. The follow-up rates were compared between patients with different demographic characteristics, surgical factors, and preoperative and discharge visual acuities. The behaviors of patients who complied with the follow-up advice were analyzed, including the number of days from surgery to follow-up and number of follow-up visits. Multivariate regression models were used to identify predictors associated with these behaviors. RESULTS: The overall follow-up rate for the 86 776 surgeries analyzed was 85.6%. Patients more likely to follow-up were women, younger than 70 years, and paying (versus subsidized or free) and had phacoemulsification rather than manual small-incision cataract surgery (all P < .001). Patients who had complications, reoperations, or poorer visual acuity at discharge were less likely to comply with the follow-up advice (P < .001). CONCLUSIONS: Targeted interventions to boost follow-up rates should be directed to patients with the characteristics found in this study. Furthermore, average measures of visual outcomes at 4 weeks are likely to be overstated relative to the truth.