Abstract
Purpose: :
To determine an Achievable Benchmark of Care ( ABC) using objective, practice-based measures of Preschool Vision Screening (PVS). The ABC shows the rates achieved by a subgroup of the top performing providers using measures obtained in real world, primary care settings.
Methods: :
We obtained all claims for Well Child Check-ups (WCC) and Vision Screening (VS) filed with Alabama Medicaid during fiscal year 2003. AL Medicaid reimburses for VS using a separate billing code, which we used to estimate individual provider's VS rate for children aged 3, 4, and 5 years. At each year of age, >98% of VS were filed with the WCC; thus the number of WCCs provided a valid denominator for the VS rate. We classified each provider as pediatrician, family physician, other MD, non-MD or clinic. After adjusting the data for small denominators, we ordered each provider sequentially in descending order until the subset of providers included at least 10% of children. The ABC was calculated as the total number of patients in the subset with a vision screen divided by the total number of patients in the subset with a WCC. The Chi Square Test of Equal Proportions was used to evaluate whether provider type influenced achieving the benchmark. Cochran Armitage Trend Test were used to determine whether patient age influenced the benchmark or the proportion of non-screening providers.
Results: :
Approximately 800 providers filed claims at each year of age. A significant increase in the ABC was found with age (p<0.001), with top performing providers screening 69.6% of children at age 3, 82.0% at age 4 and 95.3% at age 5 years. The average VS rate for all providers was 9.8% at age 3, 19.1% at age 4 and 36.0% at age 5 years. Excluding non-screening providers, average VS rates were 30.7% at age 3, 42.2% at age 4 and 62.2% at age 5 years. Non-MDs were more likely to achieve the ABC (p=0.007). The proportions of non-screeners decreased significantly as age increased (p < 0.001).
Conclusions: :
We found a large variation in the rates of PVS among primary care providers, with many providers not screening any of the younger children. However, a subset of primary care providers met or exceeded the ABC at each age. Further work should evaluate office practices of top performing providers of PVS.
Keywords: clinical (human) or epidemiologic studies: health care delivery/economics/manpower • screening for ambylopia and strabismus • clinical (human) or epidemiologic studies: biostatistics/epidemiology methodology