Inpatient hospital care compared to outpatient or day care for people with eating disorders

Why is this review important?

International clinical practice guidelines recommend that overall, people with eating disorders should receive their treatment in an outpatient setting. Most people also prefer to avoid a hospital admission, as it takes more time and resources. However, it is not known if outpatient care is as effective as more intensive inpatient or partial (day) hospital care, or if it is more acceptable for people. Those who are at medical or psychiatric risk of harm or suicide, and those with anorexia nervosa who are severely underweight or rapidly losing weight may not be safe in an outpatient setting.

Who will be interested in this review?

People with lived experience of eating disorders and people who care for them will be interested in this review.

Which studies were included in the review?

We searched medical databases and trial registers to find randomised controlled studies that compared inpatient care to partial hospital care or outpatient care, alone or in combination, to July 2018. We included four trials that included 511 people with anorexia nervosa, and one trial of 55 people with bulimia nervosa.

What does the evidence from the review tell us?

There was not enough evidence from trials to support any one setting for people with anorexia nervosa, bulimia nervosa, or other eating disorders. There was no clear difference in weight gain for people with anorexia nervosa who were treated in different settings, but they seemed more likely to complete treatment when some or all of it was offered in settings outside the hospital. The evidence was low or very low-quality, so we are uncertain about these results.

What should happen next?

We need more trials comparing inpatient to outpatient or day care for people with anorexia nervosa and other eating disorders, when it is medically safe to consider less intensive care settings.

Authors' conclusions: 

There was insufficient evidence to conclude whether any treatment setting was superior for treating people with moderately severe (or less) anorexia nervosa, or other eating disorders.

More research is needed for all comparisons of inpatient care versus alternate care.

Read the full abstract...

Clinical guidelines recommend outpatient care for the majority of people with an eating disorder. The optimal use of inpatient treatment or combination of inpatient and partial hospital care is disputed and practice varies widely.


To assess the effects of treatment setting (inpatient, partial hospitalisation, or outpatient) on the reduction of symptoms and increase in remission rates in people with:

1. Anorexia nervosa and atypical anorexia nervosa;

2. Bulimia nervosa and other eating disorders.

Search strategy: 

We searched Ovid MEDLINE (1950- ), Embase (1974- ), PsycINFO (1967- ) and the Cochrane Central Register of Controlled Trials (CENTRAL) to 2 July 2018. An earlier search of these databases was conducted via the Cochrane Common Mental Disorders Controlled Trial Register (CCMD-CTR) (all years to 20 November 2015). We also searched the WHO International Clinical Trials Registry Platform and (6 July 2018). We ran a forward citation search on the Web of Science to identify additional reports citing any of the included studies, and screened reference lists of included studies and relevant reviews identified during our searches.

Selection criteria: 

We included randomised controlled trials that tested the efficacy of inpatient, outpatient, or partial hospital settings for treatment of eating disorder in adults, adolescents, and children, whose diagnoses were determined according to the DSM-5, or other internationally accepted diagnostic criteria. We excluded trials of treatment setting for medical or psychiatric complications or comorbidities (e.g. hypokalaemia, depression) of an eating disorder.

Data collection and analysis: 

We followed standard Cochrane procedures to select studies, extract and analyse data, and interpret and present results. We extracted data according to the DSM-5 criteria. We used the Cochrane tool to assess risk of bias. We used the mean (MD) or standardised mean difference (SMD) for continuous data outcomes, and the risk ratio (RR) for binary outcomes. We included the 95% confidence interval (CI) with each result. We presented the quality of the evidence and estimate of effect for weight or body mass index (BMI) and acceptability (number who completed treatment), in a 'Summary of findings' table for the comparison for which we had sufficient data to conduct a meta-analysis.

Main results: 

We included five trials in our review. Four trials included a total of 511 participants with anorexia nervosa, and one trial had 55 participants with bulimia nervosa. Three trials are awaiting classification, and may be included in future versions of this review. We assessed a risk of bias from lack of blinding of participants and therapists in all trials, and unclear risk for allocation concealment and randomisation in one study.

We had planned four comparisons, and had data for meta-analyses for one. For anorexia nervosa, there may be little or no difference between specialist inpatient care and active outpatient or combined brief hospital and outpatient care in weight gain at 12 months after the start of treatment (standardised mean difference (SMD) -0.22, 95% CI -0.49 to 0.05; 2 trials, 232 participants; low-quality evidence). People may be more likely to complete treatment when randomised to outpatient care settings, but this finding is very uncertain (risk ratio (RR) 0.75, 95% CI 0.64 to 0.88; 3 trials, 319 participants; very low-quality evidence). We downgraded the quality of the evidence for these outcomes because of risks of bias, small numbers of participants and events, and variable level of specialist expertise and intensity of treatment.

We had no data, or data from only one trial for the primary outcomes for each of the other three comparisons.

No trials measured weight or acceptance of treatment for anorexia nervosa, when comparing inpatient care provided by a specialist eating disorder service and health professionals and a waiting list, no active treatment, or treatment as usual.

There was no clear difference in weight gain between settings, and only slightly more acceptance for the partial hospital setting over specialist inpatient care for weight restoration in anorexia nervosa.

There was no clear difference in weight gain or acceptability of treatment between specialist inpatient care and partial hospital care for bulimia nervosa, and other binge eating disorders.