Back to workshop overview

W2.5

Title:  How should main results of randomised controlled trials of treatments for low back pain be presented?

Authors and affiliation: Underwood MR 1, Lamb SE 2, Lall R 2

1 Barts and The London, London UK 
2 Warwick Clinical Trials Unit, University of Warwick, UK

Email address: m.underwood@qmul.ac.uk

Introduction: There many randomised controlled trials of different treatments for low back pain.  Some consensus has been achieved on the primary outcomes used in these trials.  Few of the larger high quality studies have shown a substantial sustained benefit in the primary outcome measure from adding additional treatment to ‘usual care’.  Interpreting the clinical importance of positive clinical trials is problematic.  Simply reporting a statistically significant difference in mean values in at one follow-up time may not be meaningful to health care purchasers, health professionals, or indeed patients.  This approach may not make it clear how many individual participants have a clinically important improvement/deterioration; how much additional improvement participants gain when compared to the natural history of the back pain, or benefits accrued over the whole follow up period.  Using one or more alternative approach may allow clinical trial data to more directly inform service provision for patients with low back pain.

Objective: To share experience of interpreting trial results to produce a critical summary of the alternative approaches to presenting the main outcome data.

Methods / content: A summary of different presentation approaches identified from the literature will be presented.  This will be supplemented by illustrations of the consequences of different presentational approaches in a sample dataset.  Interactive group discussion will be used to develop consensus the merits and demerits of the approaches used in the example and any alternative approaches identified by the group.The results of this workshop will prepared as a paper for publication.