RESUMO
OBJECTIVE: To identify non-high-density lipoprotein cholesterol (HDL-C) and HDL-C thresholds for pediatric nonfasting lipid screens that are more predictive of the need for lipid-lowering pharmacotherapy and estimate numbers of potentially avoidable fasting lipid panels. STUDY DESIGN: In this retrospective review of children and youths aged 8-21 years presenting for preventive cardiology care, initial lipid results, recommendations for pharmacotherapy, and presence of additional cardiovascular risk factors were noted. Receiver operating characteristic curve analysis calculated threshold lipid values predicting the need for pharmacotherapy and were applied to 2 screening populations. Rates of potentially unnecessary fasting lipid panels were calculated. RESULTS: A non-HDL-C value >156 mg/dL for children with ≥1 cardiovascular risk factors and >199 mg/dL for children without risk factors conferred 95% or greater sensitivity in predicting a recommendation for pharmacotherapy with higher specificity, positive predictive value, and negative predictive value compared with current guidelines. HDL-C was a poor predictor of pharmacotherapy. Application of the current thresholds to screening populations indicated that 38.5%-92.3% of follow-up fasting lipid panels would not result in pharmacotherapy. CONCLUSION: Using higher non-HDL-C and lower HDL-C thresholds could prevent unnecessary follow-up lipid panels and reduce patient anxiety, cost, and time. This could improve compliance with universal pediatric lipid screening for both health care providers and families.