A systematic review of N-of-1 methods, analysis and meta-analysis

Tags: Poster
Vohra S, Shamseer L, Bukuta C, Sampson M, Barrowman N, Moher D

Background: An N-of-1 trial is a randomized, multiple cross-over evaluation performed in a single subject. While many N-of-1 designs and statistical methods have been reported, the most efficient and rigorous methods for conducting and analyzing N-of-1 trial results have yet to be identified. Objectives: To systematically review: 1) N-of-1 trial designs; 2) N-of-1 trial statistical analysis; and 3) methods for combining data from a series of N-of-1 trials. Methods: Three systematic reviews on N-of-1 methods, analysis and meta-analysis are being completed. A common search strategy has been employed for all three reviews. Inclusion criteria: (i) primary studies (ABA or ABAB design) of a health intervention and/or clinical population or meta-analysis thereof; (ii) novel methodological papers; and (iii) methodological reviews. Exclusion criteria: multiple baseline/probe design, AB design, non-single-subject design. Clinical and methodological aspects of N-of-1 reports are being examined. Results: The searches identified 4716 articles, of which 468 met inclusion criteria. At the time of abstract submission, 217 articles await screening. Approximately 79% are reports of single N-of-1 trials or methodological aspects of study design, 70% describe statistical aspects of such studies, and 26% either combine or discuss meta-analysis of data from multiple N-of-1 trials. At least 25 different statistical analysis techniques were employed. Diverse conditions such as rheumatologic diseases, cardiac or respiratory diseases and neurological conditions have been included; interventions ranged from conventional pharmaceuticals to unconventional therapies and complementary and alternative therapies. Studies in at least eight different languages were included, illustrating the international relevance of N-of-1 trials. Conclusion: N-of-1 trials have the potential to transform knowledge translation of research evidence (i.e. to bridge the evidence-practice gap). N-of-1 trials will directly improve research capacity, rigorous evaluation and evidence-based decision making in health care, as they can be conducted by community-based providers and directly enhance their ability to achieve evidence–based practice. These systematic reviews will inform development of the CONSORT Extension for N-of-1 Trials (CENT).