Longitudinal Stability and Predictive Value of Socio/Emotional/Behavioral Screening in an Outpatient Pediatric Network.

Abstract

Background: In 2022, the American Academy of Pediatrics recommended annual social/emotional/behavioral (SEB) screening at preventive pediatric visits. Numerous SEB screeners have considerable empirical support for children of all ages. However, few studies inform the longitudinal use of SEB screeners in pediatrics.

Methods: Using electronic health records (EHR) from seven pediatric clinics in a large hospital system, we analyze the longitudinal stability of three SEB screeners among 9153 children and adolescents who were screened multiple times over a 3-year period, and we evaluate associations with International Classification of Disease (ICD) codes for SEB diagnoses.

Results: Moderate stability was documented for the Baby Pediatric Symptom Checklist (intraclass correlation coefficient [ICC]=.57), Preschool Pediatric Symptom Checklist (ICC=.60), and Pediatric Symptom Checklist-17 (ICC=.69). Stability of positive/negative results was sensitive to each screener's threshold. Across a range of thresholds that classify 7-20% of children as positive, all three screeners achieved positive predictive values (PPVs) of approximately 40-65% and diagnostic odds ratios ranging from 8 to 15. These findings compare favorably to published results for longer measures and structured interviews. Recommended thresholds varied markedly across screeners, and associations with ICD codes demonstrate that higher thresholds yield higher specificity but lower sensitivity (ie, a majority of ICD codes are not anticipated by positive screens).

Conclusions: Three common SEB screeners display longitudinal stability similar to longer, more established measures. Results demonstrate that recommended thresholds vary markedly across measures and that choice of threshold strongly influences sensitivity and specificity, underscoring the need for greater attention to selection of thresholds.