Purpose
1) To estimate compliance with annual eye screening guidelines among patients with diabetes mellitus (DM), and 2) to evaluate system-and individual-level factors associated compliance.
Methods
Utilizing computerized billing records we selected Bascom Palmer Eye Institute patients who were initially seen with ICD-9 diagnoses of DM without complications and without diabetic retinopathy (DR) or other eye disease. Medical records of patients first seen in 2007-08 (n=248) were reviewed for demographic and ocular variables, including all clinic visits through 2010. Compliance was defined by follow-up exam frequency and permitted follow-up intervals as large as 15 months to accommodate exam scheduling challenges: (1) Never returned (2) non-compliant but seen within the last 15 months (3) non-compliant and not seen within the last 15 months (4) fully compliant. We geocoded addresses of participants using the Google API. Local Kriging, an optimal interpolation method, assessed participants’ neighborhood socio-economic status (SES) using 8 indicators from the 2000 US Census data. An effort was made to survey by telephone patients who never returned.
Results
203 (82%) medical records were available for review with only three patients excluded given evidence of a DR diagnosis. The N (%) of patients in the four compliance groups were: 108 (54%), 30 (15%), 37 (19%), and 25 (12%), respectively. Compliance was not significantly associated with gender, type of insurance, or years since DM diagnosis (p>0.1). Fully compliant patients were more likely to be younger (p=0.036) and African American patients were more likely to return for follow-up exams than Hispanic or non-Hispanic White patients (p=0.038). None of the neighborhood SES variables studied differentiated patients by compliance group (all p>0.1). 69 (66%) of patients not returning for follow-up were telephoned but only 8 could be contacted for the survey, a sample size too small from which to draw inferences.
Conclusions
Less than 15% of patients in our study were fully compliant with exam guidelines and >50% never returned for a follow-up exam. Compliance was associated with both age and ethnicity; however, these relationships were not explained by SES variables.
Keywords: 460 clinical (human) or epidemiologic studies: health care delivery/economics/manpower