Experiences in the realisation of a research project on anthroposophical medicine in patients with advanced cancer

Questions under study: To date most of the published studies on the effectiveness of complementary therapies in cancer patients have yielded controversial results because of questionable methodology. Research strategies and methodologies acceptable to both conventional and unconventional medicine are difficult to find due to different belief systems. In this publication we describe the development and implementation of a project conducted as part of National Research Programme 34 (NFP 34). Detailed analysis of our experiences might provide some information on how to deal with practical difficulties in the planning and conduct of further research projects in this field. The project involved the anthroposophical Lukas Clinic in Arlesheim and the Institute of Medical Oncology of the University Hospital, Berne. This interdisciplinary research project was devised to study the relative merits of these two schools of medicine in the care of advanced cancer patients. The project was made up of three components: (1) a registration study aimed at comparing the case mix at the two institutions; (2) a three armed randomised study on the effectiveness of supportive therapy, comparing anthroposophy to psychosocial group therapy, and (3) a longitudinal study to monitor the evaluation of quality of life of patients at the anthroposophical clinic. Methods: After a brief review of the study protocol, which presents the theoretical framework of the project, problems of its implementation are described. Aspects of accrual, acceptance of randomisation and data availability are presented using simple descriptive statistics and logistic regression. Results: The registration study was duly completed with a total of 567 patients. For several reasons (not meeting inclusion requirements, high refusal rate) the accrual into the randomised study was slower than expected and required modification of the original design specifications with regard to inclusion criteria and data collection schedule. Additionally, a high dropout rate contributed to premature closure of this part of the project. The longitudinal study also suffered from low data availability at follow up. Conclusions: The study protocol constituted a major effort at compromise without loss of scientific rigour, and this effort demonstrates that it is possible to allow for different views on patients, on clinical interventions and on research strategies when establishing collaboration between different schools of medicine. Despite a theoretically sound framework, the randomised part of the project proved difficult in its practical execution. Some unexpected logistical constraints and some unmet expectations influenced the feasibility of this part of the project. Therefore, careful planning of research projects in this field of medicine should always include an extended analysis of various practical aspects of study implementation.