Purpose: To investigate possible causes for, and implications of, variations in coding of glaucoma for patients with uveitis in the Medicare database.
Patients and methods: We identified 6,391 patients coded with one of several ICD-9 codes designating a form of uveitis from a 5% sample of the 1999 Medicare database. The proportion of patients with uveitis codes who also had codes for uveitic glaucoma, primary open-angle glaucoma (POAG), or unspecified open-angle glaucoma (OAG) were determined. The proportion with codes for argon laser trabeculoplasty (ALT), trabeculectomy, visual field examinations, fundus photographs, and gonioscopy were calculated in three groups: patients with both uveitis and uveitic glaucoma; those with both uveitis and POAG; and those with both uveitis and unspecified open-angle glaucoma. Summaries of the total allowed reimbursement related to eye visits and to eye codes were determined.
Results: Of the 6,391 patients with uveitis who were studied, 1,260 (19.7%) also had the code for POAG; 412 (6.4%) had the code for OAG; and 117 (1.8%) had the code for uveitic glaucoma. Visual field testing was less common among patients with uveitic glaucoma (37%) than among those with POAG/uveitis (46%) or OAG/uveitis (44%) (P < 0.001). There were no statistically significant differences among the three groups in the amount of total reimbursement related to eye services or eye codes.
Conclusions: The code for POAG appears to be used in lieu of uveitic glaucoma in many cases. Discrepancies are probably not influenced by expected reimbursements, as much as by a lack of understanding about the available ICD-9 codes.