Abstract
Objective. Data entry errors may occur in body weights and heights assessed during routine medical care. These errors may affect data quality markedly and create a large number of biologically implausible values. To address this issue, we evaluated the quality of body weight and height measures for children based on sequential health care encounters. Methods. We evaluated the weight and height data of children aged 0–18 years receiving care at Kaiser Permanente Southern California medical centers. Error rates were calculated before and after excluding implausible values for height and weight as recorded in the electronic medical chart reviews. Results. The error rates in weight and height data of children aged <2, 2–5, 6–9, 10–13, 14–18 years were 0.4%, 0.7%, 1.0%, 1.0% and 0.7%, respectively. The most frequently identified errors were implausibly low values for height and implausibly high values for weight. After excluding implausible values, the error rates were 0.4%, 0.4%, 0.6%, 0.4% and 0.1%, respectively. The sensitivity of our approach to detect errors was 10.9%, 36.6%, 32.9%, 59.2%, and 82.5%, respectively. Conclusions. Error rates in weight and height recorded in the electronic medical record during routine medical care are low, raising the potential for this information to be used for population care management. With little effort and with the recording of this information at each encounter, error rates can be further lowered to avoid misclassification of children as obese.
Acknowledgements
This research was supported by Kaiser Permanente Direct Community Benefit Funds. We gratefully thank Julie Stern, Margaret Chi and Britta Amundsen for the careful review of the electronic medical records.
Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.