CHOOSING WISELY: REFEEDING HYPOPHOSPHATEMIA AND THE VALUE OF DAILY LABORATORY MONITORING IN HOSPITALIZED ADOLESCENTS WITH RESTRICTIVE EATING DISORDERS

Purpose: Youth with eating disorders (EDs) who require inpatient hospitalization for medical stabilization and weight restoration are at risk for refeeding syndrome (RS). RS is characterized by fluid and electrolyte shifts including hypophosphatemia. In this context, hypophosphatemia refers to refeeding hypophosphatemia (RH). One study showed that 27% of adolescents admitted to an inpatient ED unit developed RH within 5 days of hospitalization. Consequently, many ED programs monitor daily serum phosphorous levels during the first week of hospitalization. The risk factors for RH in youth with restrictive EDs are unknown. The objectives of this study are to determine the incidence of RH in hospitalized youth with restrictive EDs; to identify risk factors for RH compared to those without RH (non-RH); and to complete a cost analysis of daily laboratory monitoring in hospitalized youth with restrictive EDs vs. those at risk for RH. Methods: A retrospective chart review was performed on all hospitalized youth aged 8-18 years with a DSM-5 diagnosis of anorexia nervosa (AN), atypical anorexia nervosa (ANN), and avoidant restrictive food intake disorder (ARFID) between 2014- 2017. Data collection included demographics, anthropometric measurements, duration of illness, rate of weight loss, target goal weight (TGW), vital sign stability, menstrual history, reported caloric intake, nutritional protocol modifications, clinical signs of RS, admission hemoglobin, liver function tests, albumin, and electrolytes during the first week of hospitalization. Abnormal laboratory values were defined by the hospital's reference ranges. Data was analyzed using continuous variables to describe means +/- SD, and categorical variables were described using percentages. Results: Of the 150 subjects, 88%, 7.3%, and 4.7% had a diagnosis of AN, ARFID and AAN, respectively. Mean age was 14.2 +/- 2; 88.7% were female. RH was identified in 25.3% of patients; 94.7% had AN and 5.3% had ARFID, mean age was 13.2+/-1.9, and 71% had a %TGW < 80%. Admission BMI and caloric intake prior to admission were significantly lower for those with RH vs. non-RH (15.24 kg/m2 +/- 1.9 vs. 16.05 kg/m2 +/- 2.2, p=0.043; 770.9 kcal/day +/- 565.9 vs 1004.3 kcal/day +/- 558.8, p=0.035, respectively). Patients with RH had a higher mean hemoglobin on admission vs. non-RH (138.5 g/dL +/- 8.7 vs. 132.1 g/dL +/- 15.4, p=0.021). 47.4% were identified with RH on Day 1 of hospitalization. 65.8% with RH were started on oral p