Cataract surgery risk stratification models: a systematic review

Risk stratification models can assist cataract surgeons in clinical decision-making by categorizing patients into distinct groups based on their likelihood of complications. In this systematic review, we assess the characteristics of cataract surgery risk stratification models.

Using the Preferred Reporting Items for Systematic reviews and Meta-Analyses guidelines, we searched six databases (PubMed, OVID, Embase, CINAHL, Cochrane Trials, and Web of Science) in January 2024. We included peer-reviewed, full-text, English-language studies describing models used preoperatively to assess the likelihood of complications in cataract surgery. We excluded letters, editorials, and non-peer-reviewed publications, such as conference abstracts and studies describing predictive models that did not group the patients into distinct risk categories. We constructed a checklist from three frameworks to critically appraise the participants, predictors, and risk of bias in the models.

Of 4192 articles, eight met the inclusion criteria. Most models were designed for attending surgeons only and for phacoemulsification to predict zonular complications and posterior capsule rupture. The most common risk factors identified in the models were poor patient positioning, advanced age, small pupils, and pseudoexfoliation syndrome. Methodological limitations included the lack of multivariable modeling, standardized outcome measures, and external validation.

Cataract surgeons should understand the limitations of cataract surgery risk stratification models. Existing models can be improved with more robust methods, the use of standardized metrics, and external validation.