Algorithms for the identification of ophthalmic diseases in medico-administrative databases: A systematic review.

In 30 seconds

This systematic review evaluated algorithms for identifying key ophthalmic diseases in medico-administrative databases, analyzing 523 algorithms across various studies. The review highlighted significant heterogeneity among algorithms, with only 9% validated. The most common diseases studied were glaucoma (34%) and diabetic retinopathy (26%).

Key findings

• Of the 523 algorithms identified, only 9% were validated.
• Approximately 58% of algorithms used diagnosis codes exclusively.
• The majority of studies originated from Taiwan (36%) and Korea (27%).
• 48% of study objectives focused on factors associated with ophthalmic disease onset.

Why it matters

Understanding the variability and validation status of algorithms for ophthalmic disease identification is crucial for clinicians relying on medico-administrative databases for research and patient care decisions. Accurate identification can influence treatment outcomes and healthcare resource allocation.