Bower & Gilbody (Reference Bower and Gilbody2005) raised questions about stepped care in psychological therapies to which I would like to provide some answers. Our treatment evaluation study of guided self-change for bulimia nervosa incorporated a self-care manual and investigated acceptability, drop-out rate, ‘extra treatment’ and longer-term outcome (Thiels et al, Reference Thiels, Schmidt and Treasure1998a , Reference Thiels, Schmidt and Treasure2003). We did not dare to offer the manual only as a first step to Germans spoilt by a healthcare system with excellent provision of long-term psychotherapy. Instead we compared 8 fortnightly sessions plus a self-care manual (guided self-change) with 16 weekly individual sessions of cognitive-behavioural therapy (CBT). There were no significant differences between therapies regarding drop-out rate, general satisfaction with treatment and views regarding the usefulness of the therapies.

A journalist who wanted to write about our study in the weekly newspaper Die Zeit met with strong resistance. Although we published the 6-month follow-up results in the American Journal of Psychiatry (Reference Thiels, Schmidt and TreasureThiels et al, 1998b , quoted by Reference Bower and GilbodyBower & Gilbody, 2005) the head of the science section of Die Zeit did not believe that such low therapist input could work. Some psychiatrists in the hospital where the journalist's husband worked feared for their jobs.

Bower & Gilbody (Reference Bower and Gilbody2005) state that ‘research on the patient acceptability assumption would need to access the views of a number of different populations…’. We invited family physicians, psychiatrists, gynaecologists and various counselling services to refer those who complained of symptoms suggestive of bulimia nervosa (the clinical picture of which was briefly described) (Reference Thiels, Schmidt and TreasureThiels et al, 1998b ). An article about the service in a local newspaper led to several self-referrals. Unfortunately, we did not compare acceptability according to the source of referral and thus can only report the overall acceptability.

Follow-up by personal interview using expert and self-rated instruments was carried out a mean of 54.2 months (s.d.=5.8) after the end of therapy. Significant improvements were achieved or maintained in both groups on the main outcome measures. These included eating disorder symptoms from expert ratings (Eating Disorder Examination sub-scores: overeating, vomiting, dietary restraint, shape and weight concern), self-report (Bulimic Investigatory Test, Edinburgh) and a global five-point severity scale. Improvement was also seen on the subsidiary outcome variables Beck Depression Inventory, Self-Concept Questionnaire and knowledge of nutrition, weight and shape. During the week before follow-up 66.7% of the guided self-change group and 61.5% of the CBT group had not binged, vomited or misused laxatives.

A considerable proportion in both groups had extensive further psychological treatment, mainly for their eating disorder. The majority of these had not done well during initial treatment. An analysis of covariance showed a significant interaction between treatment group and additional treatment between the 6-month and the 4-year follow-up. Cross-tabulation showed that this difference was due to the fact that three of the five in the CBT group with additional treatment between the two follow-up assessments improved more than the eight in the CBT group without additional therapy (χ²=6.24; d.f.=1; P<0.035). However, the six in the guided self-change group who received additional treatment made as little progress as the seven who did not (χ²=0.26; d.f.=1; P<1.00). Four out of 12 individuals received additional treatment from their study therapist, the other eight received additional treatment elsewhere. The question is whether those who were allocated to guided self-change would not have recovered in the course of 4 years with any therapy or whether they might have done better with more therapist contact from the beginning of treatment.