Application of data mining algorithms to study data trends for corneal transplantation.

Area of Science:

• Ophthalmology
• Health Services Research
• Data Mining

Background:

• Corneal transplantation (CT) is a critical procedure for vision restoration.
• Understanding demographic and geographic trends in CT is essential for equitable healthcare delivery.
• Data mining techniques can reveal patterns in large healthcare datasets.

Purpose of the Study:

• To analyze corneal transplantation (CT) data in Florida from 2005-2014.
• To segment CT by patient demographics, geographic location, and surgical technique.
• To identify trends and disparities in corneal transplantation procedures.

Main Methods:

• Retrospective database study using Healthcare and Cost Utilization Project data.
• Analysis of Current Procedural Terminology codes for lamellar keratoplasty (ALK), endothelial keratoplasty (EK), and penetrating keratoplasty (PKP).
• Extracted and analyzed payer status, ethnicity, age, gender, and geography (urban/rural).

Main Results:

• CT constituted less than 1% of Florida ambulatory surgeries (2005-2014).
• Endothelial keratoplasty (EK) volume increased, while penetrating keratoplasty (PKP) and ALK decreased.
• Significant disparities found in transplantation technique by sex and ethnic group, with EK more common in females and White patients, while PKP was more common in Black patients.

Conclusions:

• Corneal transplantation rates in Florida appear higher in females and lower in ethnic minorities.
• Specific techniques showed propensities: PKP with African Americans, EK with females and Medicare reimbursement.
• Data mining is valuable for ophthalmology insights but presents challenges with large datasets.