Treating Severe Anorexia Nervosa in the Community

Can treatment for severe anorexia nervosa be delivered safely in a community setting? According to a recent paper by Calum Munro and colleagues (2014, open access), the answer is yes.

In 2001, a systematic review by Meads, Gold, and Burls found that inpatient treatment is not more or less effective than outpatient treatment for individuals with AN. Of course there will always be patients who will require inpatient care, but given the high cost, lack of clear efficacy, and known risks, it is important to ask if there are better options, particularly for a subgroup of individuals who may not need or may not benefit from inpatient care.

In their paper, Munro et al. describe a program that they’ve developed for treating individuals with severe AN in the community. The program is called the Anorexia Nervosa Intensive Treatment Team (ANITT) service. It is one component of a four-tiered system–it is a step below specialist inpatient care and a step above outpatient therapy that includes “group and individual therapy, dietetic and psychiatric treatment.”

This tiered system matches individuals to a specific tier based on their needs, which improves access to intensive services for those who need it most.

The following graphic explains the criteria for ANITT and the treatment services and care plan offered:




The ANITT model appeals to me on many levels. Overall, there seems to be a genuine interest in listening to the patient, working with (and not against) the patientand treating the patient as a whole person. There’s also a lot of focus on openness, transparency, and trust.

Assessment, Engagement & Stabilization

First, the team, particularly the dietitian, works with the patient to achieve a stable and “safe enough” (their term) weight. This is done by “fine tuning of nutritional intake and energy use around changes that the patient is willing and able to make.” Physical risk and safety are assessed by more than just a BMI value. The goal of this phase is to stabilize and build a trusting relationship with the patient.

Intensive Therapy 

This phase entails therapy twice a week for 18 months. Twice a week. Eighteen months. Following the initial 18 months, there’s a “reduced-intensity” treatment for 6 months. After the two years, patients can be discharged (if they’ve achieved good recovery), continue on with another 6-12 months of intensive treatment, or proceed with supportive treatment services.

Supportive Treatment Service

The authors describe supportive treatment services as “more flexible and varied,” with the focus on improving quality of life, physical and psychological risk management, and maintenance of a good relationship with the treatment team.

Psychological Therapy

A few things stood out to me with regard to psychological therapy. First, the emphasis on individual needs. Second:

Our core conceptualisation of anorexia nervosa is to consider the control of eating and weight as coping strategies which give our patients relief from overwhelming feelings of vulnerability. We therefore do not seek to threaten this by removing their main coping strategies until they begin to establish other means of meeting their needs. We work around the control of eating and weight as long as a minimum level of physical safety can be maintained.

While I’m not so sure about the whole “relief from overwhelming feelings of vulnerability” thing (why not just “overwhelming feelings/emotions” or “difficulties with emotional awareness/processing/regulation,” but maybe those all mean very similar things, I don’t know), I do very much like the part about not taking away the person’s main coping strategy without first establishing another strategy in its place, or at least developing ways for individuals to deal with their feelings and emotions in, well, more constructive ways.

The psychotherapy relies on the schema therapy mode model. I am not familiar with it, but here’s an open access paper that describes the framework in some detail.

This is a part of Munro et al.’s paper that I really liked:

Contrary to accepted wisdom, we find the majority of our patients, once engaged in a sufficiently trusting therapeutic relationship, able to engage in therapy even when at low BMIs.

The authors expand on this a bit more in the discussion,

In our opinion . . . there is insufficient evidence of a causal relationship between starvation, cognitive abnormalities and a failure to engage in psychotherapy . . . An alternative psychological hypothesis to explain what is often labelled as starvation-related ‘cognitive deficit’ is that this is the emotional detachment and defensiveness of an individual feeling emotionally vulnerable within an insufficiently trusting therapeutic relationshipOur clinical experience is that emotionally engaged, trusting therapeutic relationships can be established with the majority of patients with severe anorexia nervosa, even at very low BMIs.

I like this because it echoes a lot of frustration I know has been experienced by individuals with AN who have been denied therapy or been told that it wouldn’t be beneficial because their weight is too low. But they want therapy, isn’t that enough? In any case, the time could be spent developing a trusting therapeutic relationship.

I find this blanket statement that therapy is ineffective until a patient reaches some BMI point to be incredibly patronizing and incredibly dismissive of many patients’ needs and wants, and ultimately, counterproductive to treatment in those cases.

To deny AN patients therapy when they are gaining weight and to tell them they will not benefit because their weight is low is counterproductive and may even be harmful. Some may not benefit, of course (but that is true for individuals at normal weights, too), but if a patient wants therapy, I think they should have the opportunity to receive it.

Dietetic Services

Here, too, dietitians work with the patient to make realistic goals while “supporting the patient to make their own decisions about nutritional change” and helping them tackle barriers to change.

Patient Risk Management

Mortality rates are high among ED patients; however, it is important to not overstate the risks. For one, I do not think scaring patients (or parents) into recovery is the way to go. Secondly:

Many patients describe experiences of being told they are at high risk of dying, do not become seriously unwell and therefore lose trust in clinicians. We think it is important not to exaggerate the risks but to make risk assessment as transparent and objective as possible.

(I feel similarly about the claims that purging is not an effective way of getting rid of calories; when I’ve been told that, knowing full well from experience that it is an effective way of getting rid of calories, I felt that the person telling me this either didn’t know what they were talking about or were lying to me. Regardless, it did not engender trust.)


The authors provide only a glimpse of the results from a survey of current or recently discharged patients (72% of whom completed the survey). It would have been nice to see the full results of the survey, but this is preliminary data. The authors state that they “aim to publish data on quality of life and symptomatic outcomes for a cohort of ANITT patients in the near future.”

On a scale of 1 (not satisfied at all) to 5 (extremely satisfied), patients had an overall satisfaction rating of 4.0. Pretty good! The patients felt the staff was supportive and genuine, that the care they received was individualised, and that the treatment approach was holistic, focusing on more than just weight gain.

Between 2009-2011, only two patients dropped out and only five had to be detained under the Mental Health Act.


Recall that patients entering the ANITT service had to have a BMI <13 OR BMI <15 and losing >1 kg/week. Of the 101 patients treated (over an 8 year period), 4 had passed away (crude mortality rate of 4.0% over an 8-year period); all had been ill for more than 10 years.

One patient died of starvation-related causes while an in-patient; one died of an overdose after weight restoration; the third patient died of post-operative septicaemia following acute bowel obstruction having dropped out of ANITT care; and the fourth patient died of pneumonia and heart failure with comorbid insulin-dependent diabetes, in the context of a starvation state.

Several other studies have found a similar crude mortality rate (see herehere, and here), although some have reported higher rates (11.5%). Importantly, BMI (unsurprisingly) appears to be an important risk factor for mortality:

The authors found that SMRs were not substantially increased across the BMI spectrum from 11.5 to 17.5, but increased significantly at the lowest BMIs <11.5. This evidence is consistent with our relatively low mortality rate despite regularly managing patients in the community with BMIs in the range 11.5-13.5.

And this, this, is what I wish more physicians would realize and/or admit to their patients (and their families):

… managed and stabilised low weight may allow for adequate physiological adaptation, substantially reducing the risk of serious medical complications or death.

While some may argue stating this promotes ED behaviours, I think that telling patients the truth promotes trust and openness, which I think are crucial for therapeutic alliance and, ultimately, successful treatment and recovery.


The authors conclude the paper with this:

We therefore argue that in-patient care is overused because of fears about physical safety rather than objective evidence of acute risk and because of beliefs about starvation-related cognitive deficits preventing engagement in therapy.

Honestly, I am excited for treatment programs like this. Of course, I do not know how well what is on paper translates into practice, but I look forward to the possibility of more treatment services being offered in the community setting–without unnecessarily taking patients away from their families, friends, work and/or school, and community. I think there should be many treatment options to serve the diverse needs of this patient population, and I hope this is one step toward that.