N-of-1 Methods and their Contribution to Systematic Reviews and Meta-analyses

Abstract

Background: N-of-1 trials are prospectively planned, multiple crossover evaluations, conducted in individual patients. Evidence shows that a range of designs and statistical methods have been applied to N-of-1 trials. This thesis helps to provide a comprehensive understanding about the methodology and reporting of N-of-1 trials by synthesizing all published evidence. Furthermore, while the primary objective of N-of-1 trials is to assess treatment response in individual patients, this thesis explores whether any secondary benefits can be derived from N-of-1 trials and the data they generate. Given the number of N-of-1 trials conducted in the area of attention deficit/hyperactivity disorder, this condition was chosen as the clinical model explored in the thesis. Objectives: i) To provide a systematic overview of published N-of-1 trials; ii) To assess how N-of-1 trials that have been conducted to assess the same interventions for the same conditions, using identical outcome measures can be aggregated in order to yield group estimates of treatment effect; and iii) To assess how N-of-1 trials can be combined with RCT data into a single meta-analysis. Methods: A series of systematic reviews were conducted in which each review consisted of a thorough search strategy, an assessment of inclusion of primary studies, a risk of bias assessment and either a qualitative or quantitative synthesis of data. A second reviewer was involved in all reviews. Results: This thesis found that N-of-1 trials have been conducted in over 50 conditions, and that the majority of published N-of-1 trials are published as a series. Our findings also indicate that N-of-1 trials can be meta-analyzed across participants in order to yield population treatment effect estimates. Furthermore, we found that combining N-of-1 trials with RCT data into a single meta-analysis, impacts both the magnitude and precision of overall treatment effect estimates. Conclusions: This thesis examined the potential for N-of-1 trials beyond their primary purpose of providing estimates of individual treatment effectiveness and demonstrates a method of aggregating N-of-1 trials across participants as well as with RCT evidence. Clinical and research recommendations on how to move this field forward have been provided.