Assessment of Underuse and Overuse of Screening Tests for Co-occurring Conditions Among Children With Obesity

Abstract

Key Points Question Among children aged 10 to 18 years diagnosed with obesity at well-child visits, what proportion received laboratory screening tests recommended by the American Academy of Pediatrics for obesity-related conditions, and what proportion received potentially unnecessary screening tests for hypothyroidism or hyperinsulinemia? Findings In this national cross-sectional analysis of 156 773 children in 2018-2019 commercial and Medicaid claims databases, 27.3% of children diagnosed with obesity at well-child visits received laboratory screening that was adherent to American Academy of Pediatrics guidelines, and 23.0% received potentially unnecessary thyroid or insulin testing. Meaning This study’s findings suggest that specific actionable guidelines and interventions are needed to prevent overuse and underuse of screening tests for children with obesity.

Electronic phenotypes to distinguish clinician attention to high body mass index, hypertension, lipid disorders, fatty liver and diabetes in pediatric primary care: Diagnostic accuracy of electronic phenotypes compared to masked comprehensive chart review

“…14,15 Electronic phenotypes with good sensitivity for identifying patients lacking clinician attention to a child's high body mass index (BMI) and risk for or presence of high blood pressure, diabetes, lipid disorders, or fatty liver may be useful for triggering EHR-enabled point-of-care clinical decision support. [16][17][18] In the limited age range of 6-12-year-old school-age children, we previously validated a computational approach that used numeric ICD codes to identify evidence that a clinician communicated that a child had a high body mass index (BMI) or communicated about, assessed, or treated one or more associated medical risks. 19 We then used the computational approach to identify that clinician attention to high BMI/medical risk was associated with a 20% increased likelihood that a 6-12-year-old child improved their weight.…”

“…Colleagues at an external institution used our first framework and observed a lower validity to detect attention to high BMI at pediatric annual/well child visits. 17 We examined the numeric ICD codes used by the first computational phenotype approach and learned they were returning text-strings unassociated with BMI/medical-risk attention (explained in further detail in Methods). In devising decision-support tools for high BMI and each comorbidity, we identified that guidelinebased BMI and risk-factor screenings are indicated for the expanded age range of 2-18 years, yet the first computable phenotype we developed used EHR data from 6 to 12-year-olds.…”

“…Electronic phenotypes with good specificity for ruling in clinician provision of guideline‐recommended care may support quality reporting efforts and allow testing the impact of such care on health outcomes 14,15 . Electronic phenotypes with good sensitivity for identifying patients lacking clinician attention to a child's high body mass index (BMI) and risk for or presence of high blood pressure, diabetes, lipid disorders, or fatty liver may be useful for triggering EHR‐enabled point‐of‐care clinical decision support 16–18 …”

“…In this study, there are five electronic phenotypes denoting that a patient received from a clinician distinct attention to (or care directed at) (1) child overweight/obesity, (2) high blood pressure, (3) prediabetes/diabetes, (4) fatty liver, or (5) lipid abnormalities. Colleagues at an external institution used our first framework and observed a lower validity to detect attention to high BMI at pediatric annual/well child visits 17 . We examined the numeric ICD codes used by the first computational phenotype approach and learned they were returning text‐strings unassociated with BMI/medical‐risk attention (explained in further detail in Methods).…”

Diabetes screening outcomes in youth presenting for paediatric weight management: A report of the Paediatric Obesity Weight Evaluation Registry

“…7 There is significant variation in clinical practices with regards to screening for T2DM in children with obesity among health care clinicians. [20][21][22][23] Many of the previous studies published on the utility of FPG or HbA1c are smaller studies, from single centres, or represent specific populations.…”