Methodology of neuropsychological research in multicentre randomized clinical trials: a model derived from the International Subarachnoid Aneurysm Trial.

As advances in medicine and surgery lead to reductions in mortality rates for life-threatening conditions, it has become increasingly important to refine the methodology of auditing long-term morbidity. The inclusion of appropriate neuropsychological outcomes in a large multicentre randomized clinical trial poses considerable methodological and logistical difficulties. This paper presents a model developed to implement such a multicentre neuropsychological and quality of life audit for a subset of patients within the International Subarachnoid Aneurysm Trial (ISAT), the largest ever randomized trial in the treatment of subarachnoid haemorrhage. Based on our experience of collecting quality of life and neuropsychological outcomes from more than 550 patients, data are presented on the relative cost and efficacy of different organizational strategies, methods of canvassing patients and associated response rates. On the basis of this experience, we estimate a potential recruitment pool of 135 cases would be required to obtain some neuropsychological data on 100 cases. The design of any similar trial would therefore need to accommodate a loss to follow-up of approximately one third of the sample. In addition, our experience suggests that for a trial of this size and complexity, the deployment of centrally-based co-ordinators travelling to satellite centres is more cost-effective than employing co-ordinators based at those centres. Extrapolations from the observations and calculations reported here can be employed as an evidence base to inform the design of neuropsychological outcome studies in large multicentre trials.