To the Editor Instrument-based vision screening represents a paradigm shift in the approach to identify vision-threatening conditions in young children. Stults et al 1 describe trends in device use, ophthalmology referral, and amblyopia diagnosis following the expansion of photoscreening in a multispecialty practice in Northern California. The electronic health record (EHR) data used in this study provide novel insight into the downstream outcomes of vision screening devices. However, several aspects of the study could be clarified that might help readers interpret the results and guide further research into the impact of instrument-based screening on vision care.It might be helpful to know the types of screening devices that were used across the health care system. The sensitivity and specificity of the screening device, screening approach, and referral thresholds may greatly influence referral rates in a population. The study's findings appear to be for instrument-based screening only, but it might be informative to understand overall trends in vision screening (including traditional chart-based screening) to provide insight into the number of additional children screened due to photoscreening expansion.Eye care is a leading specialty referral in pediatric primary care, 2 and many children may be referred for reasons other than failed vision screening. It might be valuable to know the percentage of ophthalmology referrals after instrumentbased screening and how many of the referred children had completed a visit with an ophthalmologist-an additional area where potential racial, ethnic, and socioeconomic disparities may exist. 3 The counterintuitive decrease in amblyopia diagnosis following photoscreener expansion may benefit from further exploration. It might help to clarify the source of the diagnosis codes (primary care vs eye specialist) and the time frame over which amblyopia diagnoses were identified. We believe that children should be followed up for a specified time to identify incident amblyopia diagnoses.We applaud the authors for pursuing this important area, making use of the rich EHR data from a large health care system, and raising questions about health care equity. Future studies may be strengthened by using rigorous quasiexperimental health services research designs (eg, differencein-differences and interrupted time series). Given changes to screening patterns for young children, 4 it may be important to follow children up over longer, clearly defined periods and compare their outcomes with those of contemporaneous control groups. We believe there is great need to identify children at risk of preventable vision loss without overwhelming the health care system, particularly given projected shortages of pediatric specialists. 5