Abstract
Purpose: :
To determine whether insurance status or geographic location affects the likelihood of receiving a diagnosis of cataract.
Methods: :
Data from the National Ambulatory Medical Care Survey (NAMCS) 2002-2007 was utilized to determine the association of cataract prevalence with a number of assumed risk factors. A multivariate regression analysis was performed for the prevalence of cataract while controlling for the following variables: age, gender, race, diabetes status, geographic location, tobacco use, and insurance status. Regression models performed with statistical analysis software included logit, probit, and linear probability models. Marginal effects were calculated for each variable to determine the magnitude of impact on the prevalence of cataract. Those variables with statistically significant correlations were re-analyzed in a more concise regression model.
Results: :
The 70-79yr age group had greater odds of receiving a diagnosis of cataract compared to the 90yr and over group (OR = 4.22, 95% CI [3.53 - 4.94]). A diagnosis of diabetes increased the likelihood of obtaining a concomitant diagnosis of cataract in comparison to office visits where no diagnosis of diabetes was given (OR = 3.24, 95% CI [1.23 - 8.52]). Insurance status and geographic location did not demonstrate a statistically significant correlation with the prevalence of cataract.
Conclusions: :
Diabetic status and age group play a statistically significant role in predicting the increased likelihood of cataract prevalence based on physician office visits in this nationally-representative cohort. Insurance status and geographical location did not demonstrate a statistically significant correlation. Increased public health strategies specifically targeting patients with increased age and diabetes should be made in an effort to screen, prevent, and treat patients from potentially developing cataracts.
Keywords: cataract • clinical (human) or epidemiologic studies: prevalence/incidence • clinical (human) or epidemiologic studies: biostatistics/epidemiology methodology