No abstract

No abstract

This study was designed to investigate the impact of initial Body Mass Index (BMI) and motivation for recovery on a 12-week day therapy programme for Eating Disorders. Outcome was assessed by changes in eating disorder symptomology, mood and self-esteem. A battery of assessment questionnaires was given to 34 eating disorder clients on admission to and completion of the programme. This study found that BMI at admission did not affect treatment outcome directly. Only patient self-rated motivation directly affected treatment outcome with more highly motivated clients making more significant changes to their eating disorder symptomology across the programme. However, motivation, BMI and duration of illness influenced how likely a patient was to complete the 12 weeks and completion of the programme had a direct effect on eating disorder symptomology, mood and self-esteem. BMI and motivation can therefore be seen to indirectly effect the outcome of treatment via the ability to complete the programme.

After inadvertently making an unauthorised protocol deviation, two researchers were left with a weakened study and feeling disillusionedResearch governance is designed to ensure that "health and social care research is conducted to high scientific and ethical standards." 1 Currently the same process is applied to all breaches, regardless of their severity or likely implications. Although we do not deny the importance and relevance of research governance, our experience leads us to question how it is applied. What we didOur project, funded through a small grant from the trust, explored the effect of several variables on outcome in a day therapy service for eating disorders. Our outcome measures comprised several questionnaires administered at three monthly intervals to clients with eating disorders. As a result of advice from our project steering group (a necessary requirement for such projects), we agreed to introduce a simple qualitative measure to balance the fact that our original protocol used only quantitative measures. We used an interview based on a standard questionnaire (the Morgan and Russell scale 2 ) but adapted to form a semistructured interview covering quality of life areas such as social contacts, relationships, family, and employment.The reason for the study becoming subject to the research governance process was that we incorporated this improvement into the protocol without informing the assistant research and development director or local research ethics committee and without adding it to the patient information and consent forms. We were unaware of the requirement to do this and were not told that it was necessary. However, we have been told that in future researchers will be formally notified of this requirement. What happenedThe transgression emerged during a routine telephone conversation with the assistant research and development director. Immediately, we were asked to stop all our research activities while our protocol deviation was subjected to the research governance procedures. This involved three months of formal meetings with the assistant research and development director, resubmission of the protocol and associated patient information and consent forms, and amendments to the application to the ethics committee (15 copies required). We also had to write letters to the ethics committee and the assistant research and development director explaining where we had gone wrong and the amendments made. Furthermore, all other projects in the unit were subjected to a lengthy audit process. To restart the project we had to formally request permission from the assistant director. After effectsThis process had a major effect on the study. The research was frozen for two months, during which time patients left the service and could not be followed up and other patients joined the service and could not be incorporated into our project. Moreover, ongoing monitoring of patients in the project (weekly measures of self rated motivation) could not be obtained. This has left us with incomplete datase...