Objective: Eating disorders (EDs) are some of the most common chronic disorders in adolescent girls, and have some of the worst prognoses among psychiatric diseases. However, reported data on mortality and morbidity of ED patients are scarce, and no previous studies have compared the short-term outcomes of enteral nutrition (EN) and intravenous hyperalimentation (IVH) in patients with EDs.
Method: Using the Diagnostic Procedure Combination database, a national inpatient database in Japan, we searched for ED patients who received EN or IVH. We investigated the backgrounds, complications, and in-hospital mortality for all ED patients. We compared the length of stay between the EN and IVH groups using the Cox regression model. In-hospital mortality was compared between the groups using propensity score matching and inverse probability weighting.
Results: We identified 3,611 patients with EDs from 540 hospitals. The mean body mass index was 13.1 ± 1.9 kg/m(2); 41 (1.1 %) patients died. The mean length of stay was 61.7 days. Compared with the EN-alone group (n = 634), the IVH-alone group (n = 278) showed significantly higher proportions of sepsis (0.5 vs. 5.8 %; p < 0.001) and disseminated intravascular coagulation (0.5 vs. 2.9 %; p = 0.005). The Cox regression showed no significant difference in hospital discharge between the two groups. Propensity-matched analysis evidenced lower in-hospital mortality in the EN group than the IVH group (0.4 vs. 3.0 %; p = 0.019).
Discussion: ED patients treated with IVH were significantly more likely to have higher in-hospital mortality and morbidity than those receiving EN.