There are high health and economic costs associated with inappropriate prescribing for chronic disease. This thesis explores whether it is possible to broaden and increase the rigour of therapeutic decision-making strategies in clinical practice. An argument is advanced that builds a case for using n-of-1 trials as a diagnostic tool in routine clinical practice. N-of-1 trials are within patient randomised, double-blind, cross-over comparisons of drug with placebo or alternative drug using multiple pairs of treatment periods. They refine the targeting strategy, trial of therapy, normally used by doctors. If implemented on a routine basis in clinical practice, they could reasonably be expected to improve health outcomes and reduce prescribing costs for suitable chronic diseases. A service which provided ready-made n-of-1 trials had not been available in Australia before. Therefore, we embarked on a program of research to study whether this was feasible and whether it was useful to patients and clinicians.
First, n-of-1 trials (as diagnostic tests of drug effectiveness) for osteoarthritis and attention deficit hyperactivity disorder (ADHD) were established in highly selected academic general practices. These pilot studies showed that the trials were feasible. Therefore, the service was broadened to become a diagnostic service conducting n-of-1 trials in a similar way to pathology testing, delivered by post and telephone to doctors and patients Australia-wide. This innovative method of delivering the service proved feasible in clinical practice.
Second, the clinical usefulness of the postal n-of-1 service, with respect to short term impact on management, was assessed. Two series of n-of-1 trials (for chronic pain and ADHD) were evaluated in terms of whether they changed drug use immediately afterwards. The trials were associated with a high proportion of management changes, and about a third of patients on nonxvi stopped them afterwards.
Third, patient perspectives are described in a qualitative study examining patients' (and their parents') experiences of n-of-1 trials. Patients and parents felt empowered by their active participation, which led to increased knowledge, awareness and understanding of their condition, their bodies' response to it, and its drug therapy.
Finally the clinical usefulness of the service was assessed with respect to longterm impact on management in two studies. During the 12 months after the trial, for chronic pain, persistence with patients' clinical decision decreased substantially, but for ADHD, persistence remained high. Over half of the ADHD doctor-family groups made management decisions congruent with the trial results. In summary, our post and telephone service allowed access to n-of-1 trials in routine clinical practice. The trials were favorably regarded by most patients, who found them educational and empowering. Overall, we concluded that n-of-1 trials may result in important changes in both short and long-term care. To establish causality of these changes in management, comparison with anon n-of-1 trial group in a randomised controlled trial is required.
However, the n-of-1 trials were expensive. There were difficulties with patient recruitment, medication, withdrawal, and study design, although we addressed and overcame most of these. Awareness among clinicians and policy makers remains a real barrier to widespread implementation.