ABSTRACT
Timely identification of individuals "at-risk" for myopia progression is the leading requisite for myopia practice as it aids in the decision of appropriate management. This study aimed to develop 'myopia progression risk assessment score' (MPRAS) based on multiple risk factors (10) to determine whether a myope is "at-risk" or "low-risk" for myopia progression. Two risk-score models (model-1: non-weightage, model-2: weightage) were developed. Ability of MPRAS to diagnose individual "at-risk" for myopia progression was compared against decision of five clinicians in 149 myopes, aged 6-29 years. Using model-1 (no-weightage), further 7 sub-models were created with varying number of risk factors in decreasing step-wise manner (1a: 10 factors to 1g: 4 factors). In random eye analysis for model-1, the highest Youden's J-index (0.63-0.65) led to the MPRAS cut-off score of 41.50-43.50 for 5 clinicians with a sensitivity ranging from 78 to 85% and specificity ranging from 79 to 87%. For this cut-off score, the mean area under the curve (AUC) between clinicians and the MPRAS model ranged from 0.89 to 0.90. Model-2 (weighted for few risk-factors) provided similar sensitivity, specificity, and AUC. Sub-model analysis revealed greater AUC with high sensitivity (89%) and specificity (94%) in model-1g that has 4 risk factors compared to other sub-models (1a-1f). All the MPRAS models showed good agreement with the clinician's decision in identifying individuals "at-risk" for myopia progression.
Subject(s)
Myopia , Humans , Myopia/diagnosis , Risk Factors , Risk AssessmentABSTRACT
PURPOSE: This study aimed to predict myopia prevalence in urban Indian children and to describe the generational effect of myopia in different age groups over the next three decades from the year 2020. METHODS: A systematic review of myopia prevalence in India was performed using the Preferred Reporting Items for the Systematic Reviews and Meta-analysis (PRISMA) guidelines and included eight studies with 28 600 participants, which were published in the period 1 January 1999 to 31 December 2019. The best fit for the prediction model was assessed with the baseline prevalence data plotted against different years and fitted with multiple mathematical regressions (linear, second-order polynomial, third order polynomial and exponential). Based on the quality of the fit assessed by the coefficient of determination (R2 ) values, the sum of squared residuals and statistical significance, final predictions for myopia prevalence in the 5 to 15-year-old urban Indian children was estimated using the aptly suited linear regression model. To describe the generational effect on myopia prevalence over the next three decades, the prevalence of myopia in both children and adults, based on the available literature (1999 to 2020) was plotted against age, as the baseline. RESULTS: The prevalence of myopia in 5 to 15-year-old urban children increased from 4.44% in 1999 to 21.15% in 2019. Our predictions, based on the slope of 0.8% every year (4.05% for every 5 years) indicate that the prevalence of myopia will increase to 31.89% in 2030, 40.01% in 2040 and 48.14% in 2050. Due to the generational effect (caused by the nature of the condition lasting a lifetime once developed), there will be an overall increase in myopia prevalence across all age groups of 10.53% in the next three decades (2020 to 2050). CONCLUSION: The estimates of myopia prevalence across all age groups indicate the possible future epidemic of myopia in India within a few decades, similar to the situation in East Asian countries, unless active intervention to prevent myopia and changes in lifestyle are instigated to counteract myopia in India. Meticulously designed eye care services with focussed anti-myopia strategies are needed to control the rising myopia prevalence in India.
