Improving the rate of preschool vision screening: an interrupted time-series analysis

To implement a 6-month quality improvement project in 13 clinics in North Carolina to improve vision screening rates for children 3 through 5 years of age.

We trained each clinic in approaches to vision screening and selected champions to provide feedback based on a 3-month baseline chart audit of up to 90 charts in each clinic and then 60 monthly chart audits in each clinic.

Overall, the baseline rate of distance vision testing (92%) and stereopsis testing (80%) was high. By the end of the project, there were increases in both the rate of distance vision testing (97%; P < .001) and stereopsis testing (89%; P < .001). Initially, there were many different tests used to assess distance visual acuity and some variation in the thresholds used for referral for eye care. Tests were standardized across clinics by the end of the project. The proportion of all children who were untestable was high throughout the project, including 45% among 3-year-olds by the end of the project. Follow-up rescreening was rarely documented. By the end of the project, only 48% of children with an abnormal screen result were documented to be referred. Within each clinic, concerns about the accuracy of testing persisted throughout the project.

We were successful in standardizing vision testing. Even with training, the proportion of untestable children was high. Rates of documented referral were low, which reflects provider concerns about testing accuracy. New strategies are needed to improve testability and ensure timely referral and follow-up after an abnormal vision screen result.