Keratoconus prevalence estimates in heterogeneous Iraqi clinical populations: A systematic review and exploratory meta-analysis
Hassan A. Aljaberi, Saeed RahmaniBackground
Available keratoconus (KC) estimates in Iraq are derived mainly from clinic-based or selected populations and vary across sampling frameworks, diagnostic definitions, and units of analysis.
Objectives
To summarize study-specific KC and KC-spectrum estimates, demographic and risk-factor evidence, and major sources of heterogeneity without inferring national prevalence.
Design
Systematic review with primary study-specific and population-stratified synthesis and a secondary exploratory meta-analysis.
Data Sources and Methods
PubMed, Scopus, Embase, Google Scholar, and the Cochrane Library were searched through December 2025 for English-language observational studies of Iraqi populations published between 2010 and 2025. Two reviewers independently selected studies and extracted data. Methodological quality was assessed using design-appropriate Joanna Briggs Institute checklists. Studies without an eligible prevalence denominator were retained for descriptive or risk-factor synthesis but excluded from quantitative prevalence analysis. The overall six-study model was treated as secondary and exploratory. Sensitivity analyses examined participant/patient-level denominators, clinically established KC, direct topographic/tomographic assessment, and comparable sampling frameworks. Restricted maximum likelihood with Hartung–Knapp adjustment was used for the revised overall model, with DerSimonian–Laird retained as a sensitivity comparator; a 95% prediction interval was also estimated.
Results
Nine studies were included; six contributed to quantitative synthesis. Study-reported KC/KC-spectrum estimates ranged from 4.1% to 30.2%. Population-stratified estimates were 15.26% (95% CI, 3.74%–45.52%; I 2 =98.89%) in adult ophthalmology/refractive-problem clinic populations and 10.97% (95% CI, 7.50%–15.78%; I 2 =63.51%) in selected pediatric refractive/high-risk populations; one refractive-surgery study reported 4.08% (95% CI, 3.10%–5.36%). Participant/patient-level sensitivity analysis remained highly heterogeneous (I 2 =98.59%). The secondary overall model was not interpreted as a common or national prevalence. One matched case-control study reported associations with eye rubbing, family history, and parental consanguinity.
Conclusion
Study-specific and population-stratified findings are more informative than a single combined estimate. Representative population-based studies using standardized diagnostic criteria and participant-level reporting are needed.