[What factors explain the length of hospitalization for anorexia nervosa: A systematic review]

Abstract

Objectives: The cost of hospital treatment for anorexia nervosa (AN) is very high given its duration. Identifying factors related to length of hospital stay (LOS) would make it possible to consider targeted therapeutic strategies that, by optimizing care, would reduce their duration and costs. The objective of this work is to identify the factors related (predictive and associated) to LOS for AN.

Methods: Systematic review of existing literature up to October 2020, based on Pubmed, according to PRISMA recommendations (Preferred Reported Items for Systematic reviews and Meta-Analysis). Factors related to LOS have been described in two categories: factors related to clinical aspects ; and factors related to therapeutic aspects and management modalities. We distinguished predictive factors (identified as pre-hospitalization or contemporaneous with hospital admission) and associated factors (observed during hospitalization) for each category.

Results: Thirteen articles were selected. Samples from the selected studies ranged from 35 to 381 subjects, mostly women with restrictive type AN (R-AN), but some samples included all types of AN, or focused on purging-type forms. The mean age at admission ranged from 13.6 years (Standard Deviation-SD:±1.6) to 30.3 years (SD :±13.9), corresponding to adolescent, adult or mixed samples. Mean body mass indices at admission ranged from 12.3 (SD±1.4) to 16.6 (SD:±2.1). The duration of disease progression ranged from 11.7 months±2.2 to 9.7 years. Mean LOS are short for studies conducted in pediatrics or in medical services (ranging from 13.0 days [SD±7.3] to 22.1 days [SD±9.4]); they are more variable for studies conducted in psychiatry: from 15.6 days (SD±1.0) to 150.2 days (SD±80.8). Among the factors related to an increase in LOS, clinical predictors included: older age at onset or admission; longer duration of the disorder; low minimum body weight during AN; low BMI at admission; purgative form of anorexia nervosa; and high levels of dietary symptoms (asceticism and ineffectiveness dimensions on Eating Disorder Inventory-2). Therapeutic and management modality predictive factors were: a higher number of hospitalizations for AN; the use of enteral nutrition (nasogastric or percutaneous gastric tube) on admission or during hospitalization; the use of intravenous renutrition coupled with oral renutrition; hospitalization far from the patient's home; absence of hospital care in psychiatry after medical stabilization in a somatic unit; compulsory hospitalization. Associated factors were: the presence of psychiatric comorbidities; greater weight gain during hospitalization. Among the factors related to a decrease in LOS, the clinical predictive factor were: greater self-confidence at admission (measured by the Eating Disorder Recovery Self-Efficacy Questionnaire). Therapeutic and management modality predictors included: increased caloric intake of oral renutrition on admission; intake of oral nutritional supplements on admission; and hospitalization in urban areas. The associated factor was: compliance with the weight contract in the adolescent population.

Discussion: Factors related to an increase in LOS are explained by: higher resistance to treatment, higher severity of the disease, the time required for weight gain in services using cognitive-behavioural therapy, complications associated with renutrition modalities such as parenteral renutrition, difficulties in organising outpatient follow-up which require better consolidation of inpatient treatment and the lack of multidisciplinary care in medical services. Factors related to a decrease in LOS are due to: faster weight gain, the presence of a greater number of outpatient follow-up structures close to the hospital and better adherence to treatment to complete the weight contract.

Conclusions: Taking these factors into account during hospitalization for AN would help optimize care, duration and costs. This situation therefore requires the development of therapeutic trials targeting the identified factors in order to reduce LOS in the treatment of AN.

Keywords: Anorexia nervosa; Anorexie mentale; Durée de séjour; Hospitalisation; Hospitalization; Inpatient treatment; Length of stay; Predictors; Prédicteurs; Traitement hospitalier.