Yield of laboratory testing in pediatric ketogenic diet patients: Critical assessment of abnormal results and impact on clinical care.

OBJECTIVES: To retrospectively assess the incidence of high beta hydroxybutyrate, low bicarbonate (BIC), high acyl carnitine, low selenium, low magnesium, low zinc, low phosphorus, in a cohort of supplemented patients treated with the ketogenic diet (KD) for medically intractable epilepsy. To analyze effect of age, duration of exposure to KD, type of KD, and route of KD intake on lab abnormalities. To analyze the incidence of clinically actionable results, resulting in medical interventions based on abnormal results and to analyze costs of testing.

METHODS: Retrospective chart review and statistical analysis. Association between abnormal values (binary) and categorical variables was tested with Chi-square/Fisher's exact test. Associations between abnormal values (binary) and continuous variables were analyzed with logistic regression. Statistical analyses were performed in SAS 9.4.

RESULTS: We included 91 patients with average duration on diet of 46.73 months (IQR 18.8-75.5 months). Most patients were on the classic KD (81 KD- 59% on 4:1 ratio, 10 modified Atkins diet). 74% were orally fed and 70% completed lab visits to the 12-month mark. There was no significant association between abnormal laboratory parameters and duration of exposure, type of diet, route of administration. Younger children were more likely to have low BIC, high acyl carnitine. Older children were more likely to have low phosphorus. Less than 15% of patients reported clinical changes to suggest dietary deficiency in vitamins/ minerals and in < 11% of cases was an actionable laboratory parameter found.

SIGNIFICANCE: Our study is the first to document the real-life incidence of selected tests being abnormal when following consensus guidelines on lab testing. Elimination of tests with low yield will result in cost savings of up to $USD 185 per visit. Low phosphorus is frequently found in patients on KD.