Electronic Health Record Classification of Tobacco Smoke Exposure and Cotinine Levels in Hospitalized Pediatric Patients.

Objectives: Documentation of children's tobacco smoke exposure (TSE) in the electronic health record (EHR) can have important implications for clinical care. However, it may not be accurate if it is not based on biochemical assessment, the most reliable method of verifying TSE. Our objectives were to compare the accuracy of EHR classification of TSE with cotinine verification and to explore parent and child variables associated with biochemically verified TSE.

Methods: Participants were 171 hospitalized pediatric patients (ages 0-17 years; mean age 5.1 [SD 3.7] years) who had EHR documentation of TSE and measured salivary cotinine. Children with cotinine levels >1 ng/mL were classified as having biochemical verification of TSE. Parents reported sociodemographic characteristics, and children's EHRs were abstracted for TSE status, past medical history, and diagnoses. We conducted χ tests to assess the agreement between EHR classification of TSE status and cotinine levels. Then, we assessed the relationship between sociodemographic and clinical variables and cotinine using crude and adjusted logistic regression models.

Results: Overall, 71% (121 of 171) of EHR classifications were correct on the basis of cotinine levels. Specificity analyses showed that 77% (53 of 69) were correctly identified as exposed to tobacco smoke. Sensitivity analyses showed that 67% (68 of 102) were correctly identified as unexposed. The negative predictive value was 0.61 (53 of 87); 39% (34 of 87) were misclassified as unexposed. The positive predictive value was 0.81 (68 of 84); 19% (16 of 84) were misclassified as exposed.

Conclusions: Almost 40% of children were misclassified in the EHR as unexposed to tobacco smoke. Biochemical verification should be used as part of universal TSE screening during pediatric hospitalizations.