We are often asked how long should treatment take. At the recent ICED in New York there was a plenary on shorter treatments where the panel seemed to feel that short treatments were equivalent to longer treatments. Put another way, most change in eating disorder symptoms occurs early in treatment. This of course is true in some cases and predicts a better outcome. If all patients could turn up highly motivated to change and successfully manage the difficulty of change (and preferably be normal weight with mild symptoms) this would be great. There is a risk of the idea that all change occurs in the first 4-8 sessions leading to clinician and patient hopelessness and decreases in self efficacy (which is correlated with outcome) if change is not forthcoming early.

I looked at one of the papers on which these conclusions were based. Rose & Waller (2017) concluded that in a primary care setting (so not a specialist eating disorder service but a clinic of outpatient therapists trained in CBT) that eating disorder behaviours can show “substantial improvement”. They further concluded that improvement peaked between 8-12 sessions and that there was no significant improvement in adding sessions. The authors do point out the limitations of the study (including both AN and BN patients, it is a single unreplicated study). However, what is not emphasised is important. Their remission criteria is BMI > 18.5, absence of binge/purge symptoms for a month and an EDE-Q global score within one standard deviation of normal. The mean BMI of the sample at admission to the trial was 22.5 (way above one of the criteria for remission). The lowest BMI in the study was 17 – one wonders how many patients had a BMI of less than 18.5 on entry to the study. When people have a BMI above your remission criteria it is hard to know how to take your results, particularly for patients with AN.

Patients were binging and purging around twice a week at admission. The study reported that there were small to medium changes for binge eating and medium changes to purging and to eating attitudes.  Looking at the bingeing data, on average, patients were bingeing  once a week in the month prior to the end of treatment. Diagnositic criteria for BN is bingeing once a week for three months. This means that on average, this group continued to binge at a rate that was significantly less than admission (treatment helped) but that still made criteria for BN. The on average is important as it means some people did well with the treatment and some people did not.

Overall, the results indicated that 23.4% of patients made full remission at the end of therapy. So 76.6% did not make a full remission.

It’s hard to make any definitive conclusions on the basis of the above. We could conclude that some people will do well with a shorter treatment and some, perhaps the majority, will not. The article points to the importance of early behaviour change. No arguments there. I am concerned that therapists were advised to discontinue treatment after 8 sessions if significant progress was not being made. Not to look at what the treatment may not be offering to the patient, but to discontinue treatment. There is a significant tension between having a time limited therapy to focus both therapist and client on behaviour change and the need for formulation as to why a therapy protocol may not be serving the client’s needs. I am not sure it is the client that should be discontinued. Maybe it is the protocol? Maybe it is the protocol for certain patients, at certain times?

Article – https://www.ncbi.nlm.nih.gov/pubmed/28884837