The societal economic impact of vision impairment in adults 40 years and above: findings from the National Eye Survey of Trinidad and Tobago

The study addresses the question: What is the prevalent societal economic burden of presenting vision impairment among adults aged ≥40 years in Trinidad and Tobago, and how is it distributed across direct medical, direct non-medical, indirect, and intangible (wellbeing) costs? The context is the global imperative to reduce the burden of VI to advance Sustainable Development Goals, with prior global estimates indicating substantial costs but limited population-based, country-specific, bottom-up evidence. Existing literature shows large global productivity losses from VI and few observational cost-of-illness studies that capture individual-level utilisation and costs, with gaps in inclusion of informal care and a complete absence of Caribbean data. The purpose is to generate robust, locally-relevant estimates using data from a national eye survey and a contemporaneous health system survey, informing policy and resource allocation.

Prior work estimated the global cost of VI at approximately US$3 trillion in 2010 with increases projected, and global productivity losses of about US$410.7 billion (PPP) in 2018 for ages 15–64. Systematic reviews note a paucity of population-based observational cost studies, frequent omission of informal care productivity losses, heterogeneity in methods precluding meta-analysis, and no data from national eye surveys or the Caribbean. Consensus Guidelines (2010) outline cost categories but methodological variability persists, including the controversial valuation of intangible effects and use of disability weights, which vary widely across studies.

Design and data sources: Cross-sectional, bottom-up cost-of-illness analysis from a societal perspective using the 2014 National Eye Survey of Trinidad and Tobago (NESTT) and a contemporaneous national eyecare system survey. The NESTT used multistage, random cluster sampling with probability-proportional-to-size, sampling 9913 eligible individuals in 120 clusters (95.9% household coverage), including 4263 aged ≥40 years; 3589 (84.2%) had visual acuity measured. Clinic-based assessments included ophthalmic exams and multiple questionnaires (medical/ophthalmic and socioeconomic). Socioeconomic and clinical variables: A tailored socioeconomic questionnaire captured demographics, education, employment, income category, insurance, transport, and informal care. Clinical data included spectacle/contact lens ownership, surgical/laser history, and prescription topical ophthalmic medication use in the past 3 months. Those with BCVA worse than 6/18 in the better eye were queried about low-vision services and aids. Statistical analysis: Crude and survey-adjusted estimates used Stata svy commands accounting for design (island and cluster), selection probability weights, cluster response rates, and post-stratification to 2011 Census by municipality, sex, and 5-year age bands, with finite population corrections. Multilevel mixed-effects logistic regression (melogit) modeled odds of presenting VI (<6/12) by employment, education, literacy; eye care utilisation; insurance; and exclusive use of public eye care. Cases estimation: Adjusted prevalence by vision category applied to the 2014 mid-year population aged ≥40 years (541,894) to estimate case counts. Cost classification and perspective: Societal perspective included prevalent direct medical (eye care services, surgery, medications, lasers), direct non-medical (refractive correction, low-vision aids, transportation), indirect (productivity losses from unemployment/underemployment, sick leave, informal care), and intangible effects (loss of wellbeing via DALYs). Transfer payments and dead-weight losses were considered but excluded from final cost totals. Direct medical costs: Eye care utilisation (visits to optometrists, ophthalmologists, health centres, GPs, ED), hospital admissions, surgeries (e.g., cataract), laser therapy, and topical ophthalmic medications were identified via NESTT questionnaires. Unit costs (public/private) were obtained from the national eyecare system survey of all registered providers and ministry communications; missing public sector tariffs (e.g., day case/overnight admissions, public laser) were assumed equal to private tariffs. Ophthalmic drop use over 3 months was extrapolated to 12 months; unit prices were obtained from three private suppliers and public CDAP price lists; costs applied based on reported payment (free/paid/mixed). Direct non-medical costs: Refractive correction (spectacles/contact lenses) exclusively private in 2014; unit costs from 48 optometrist respondents; assumed 70% basic distance and 30% bi/tri/varifocal. Low vision: prevalence-based estimation of cases; crude proportions using different aids; unit costs drawn from European studies (2004) and RNIB shop, inflated to 2014; low vision assessment unit costs from local optometrists. Transport: mode-specific costs from a patient survey (n=450) at five public ophthalmology departments; applied to estimated return journeys based on utilisation. Indirect costs: Productivity losses estimated via a human capital approach. Employment rate (ER) for ages 40–64 computed by vision category; employment gap calculated as 1 minus (category ER / overall ER). Productivity loss per person equals employment gap × median annual income (TT$54,000 base; sensitivity: mean TT$66,960; elementary TT$40,704) × overall ER; multiplied by estimated cases (ages 40–64) per category. Part-time work loss assumed 50% hours reduction. Sick leave losses derived from self-reported sick days and lost earnings in the past year; friction period not applied. Informal care valued using the opportunity cost method based on reported hours in past month and hourly wage for elementary occupation (TT$21.20), multiplied by number needing care and average hours. Intangible effects: Prevalent YLDs for VI computed as cases × WHO disability weights (0.047 near VI, 0.005 mild VI, 0.089 moderate VI, 0.314 severe VI, 0.338 blindness), no discounting/age weighting; YLL assumed zero (VI not directly causing premature mortality). Monetary valuation used value of a statistical life (VSL)-derived value of a lost year of wellbeing for 2014 (US$42,483; TT$272,316) based on life expectancy 70.52 and prior VSL estimates, adjusted via GDP deflator. Sensitivity analyses explored alternative disability weights (GBD variants) and wide parameter ranges (±95% CI or ±50% where data sparse) across utilisation, unit costs, income, ER, cases, and informal care. Cost allocation and currency/inflation: Public/private unit costs matched to reported provider preference; mixed users assigned average. Government bore public sector and CDAP costs; individuals bore transportation, refractive correction, private services; insured individuals assumed 30% co-payment. Unit costs inflated to 2014 using World Bank GDP deflator; converted to TT$ and UK£ (TT$1=£0.0952).

- Total societal cost of VI (≥40 years) in 2014: TT$3,842,324,655 (UK£365,650,241); per-capita (≥40 years) £674.76. - Intangible effects (loss of wellbeing) comprised 73.3% of total (TT$2,817,279,256). Excluding wellbeing loss, economic cost: TT$1,025,045,399 (100%), with indirect costs 70.5% (TT$722,379,355), direct medical 17.9% (TT$183,303,734), direct non-medical 11.6% (TT$119,362,310). Per-capita excluding wellbeing loss: £180.01 (direct medical £32.19; direct non-medical £20.96; indirect £126.86). - Major indirect cost components: productivity loss from VI TT$630,859,320; part-time work loss TT$80,276,228; sick leave TT$6,211,776; informal care TT$5,032,032. - Direct medical cost components included: ophthalmologist services TT$53,339,693; topical ophthalmic medications TT$66,612,229; day-case cataract surgery TT$25,841,007; overnight admissions TT$14,203,343; laser therapy TT$7,963,698. - Direct non-medical costs: spectacles/contact lenses TT$70,391,870; low vision aids/assessments TT$37,576,201; transportation TT$11,394,239. - Estimated cases (≥40 years, 2014): distance VI 64,431 (95% CI 54,623–74,077); 86.1% (55,481; 95% CI 53,401–57,221) potentially avoidable. Near VI from uncorrected presbyopia: 120,842 (95% CI 112,715–128,971), avoidable. - Employment rate by vision: normal distance & near 73.2% (95% CI 70.9–75.4), MSVI 41.4% (95% CI 30.6–52.2), blind 0%; associated individual productivity losses ranged from TT$2,367 (near VI only) to TT$36,869 (blindness); normal vision showed TT$2,639 productivity gain vs median. - Socioeconomic gradients: higher odds of presenting VI among service industry workers (OR 2.7, 95% CI 1.3–5.4) and elementary jobs (OR 2.9, 95% CI 1.5–5.8) vs professionals. Lower odds of VI with higher education: secondary (OR 0.43, 95% CI 0.31–0.59), post-secondary (OR 0.44, 95% CI 0.25–0.77), university (OR 0.30, 95% CI 0.15–0.62) vs primary only; literacy associated with reduced odds (OR 0.26, 95% CI 0.12–0.58). - Eye care utilisation (past 12 months): ~30% in normal/mild VI/MSVI; 18% near VI; 7% blind. - Sensitivity analysis: Disability weights strongly influenced intangible cost estimates (GBD 2013 weights reduced wellbeing cost to 3,485 DALYs ≈ TT$941.7m; original GBD increased to 12,259 DALYs ≈ TT$3,338.4m). Tornado analysis showed productivity loss parameters and case counts were major drivers of uncertainty.

The findings quantify, for the first time in the Caribbean, the broad societal burden of VI, demonstrating that intangible wellbeing losses dominate total costs, followed by substantial productivity losses. This addresses the research question by providing comprehensive, population-based, bottom-up estimates that incorporate direct, indirect, and intangible components aligned with consensus guidelines. Comparisons with high-income countries show similar dominance of wellbeing and productivity costs, yet Trinidad and Tobago exhibited lower-than-expected utilisation of advanced treatments and diagnostics in 2014, suggesting underinvestment and service gaps relative to need. High topical medication spending and glaucoma’s prominence as a cause of VI indicate potential efficiency gains through procurement strategies. The economic burden appears to fall disproportionately on lower-income and less-educated households, with marked employment disparities by vision status, underscoring socioeconomic inequities. Policy implications include prioritizing interventions that improve access to effective refractive services, cataract surgery, diabetic retinopathy screening, glaucoma management, low-vision services, and workplace enablement. Given the sensitivity of wellbeing estimates to disability weights, standardization in cost-of-illness methodologies and disability valuations is needed to improve cross-country comparability and guide resource allocation.

Vision impairment in Trinidad and Tobago imposes a large societal cost, largely borne by affected individuals and families through wellbeing loss and productivity impacts. At least 86% of distance VI is potentially avoidable, highlighting opportunities for prevention and treatment. The study contributes a detailed, bottom-up, nationally representative cost estimate and a reproducible methodological framework integrating health system and survey data. Future work should include standardized cost-of-illness tools, refine valuation of intangible effects, and conduct cost-effectiveness analyses of scaling low-vision aids, workforce enablement, refractive services, cataract surgical coverage, and diabetic retinopathy screening to inform policy and investments.

- Age and population scope: limited to community-dwelling adults aged ≥40 years; institutionalized populations and long-term care costs related to VI were not included. - Response and selection bias: moderate response rates (59–66%); blind individuals may have been underrepresented due to access barriers; those who attended clinics may differ from housebound individuals. - Recall bias: self-reported utilisation and costs over 12 months may be misremembered. - Cost scope: transfer payments and dead-weight losses excluded; opportunity costs for carers not fully captured; did not explicitly measure direct health costs from falls, fractures, accidents, diabetic self-management complications, or depression linked to VI. - Data and assumptions: some unit costs imputed (e.g., public sector laser/admissions assumed equal to private); productivity losses estimated via human capital approach using median income and employment gaps; friction period not applied to sick-leave estimates; disability weight choice substantially affects intangible cost estimates. - Service utilisation limits: rare high-cost interventions (e.g., anti-VEGF) not captured; study not powered for uncommon events.