i Background and Purpose. Typically systematic reviews and guidelines consider chronic and acute/sub-acute back pain (< 3 months). However, LBP’s clinical course is often untidy, making this distinction difficult to apply in practice. We did a secondary analysis of the UK BEAM data-set to decide pain duration affected treatment outcome.

ii Method and results: We recruited 1,334 participants with four weeks or more of LBP to a trial of manipulation and exercise for low back pain. We found small sustained benefit from the manipulation package and a small short term benefit for the exercise package. One third of our participants reported that their pain had lasted for less than 90 days (range 4 weeks to 40 years). For this analysis the dependent variable was the Roland Morris Disability Questionnaire score at 12 months. Independent variables were treatment group and episode length, an interaction term between treatment (usual care, exercise, manipulation, and manipulation followed by exercise) and duration of current episode. Preliminary analysis indicates that the significance of the interaction terms for manipulation vs GP care, exercise vs. GP care and manipulation followed by exercise vs GP care are, 0.409, 0.037 and 0.889 respectively. The statistically significant result for the interaction between exercise and episode duration is a positive interaction, indicating that exercise may be more effective for those with shorter episodes of back pain and GP care more effective for longer episodes.

iii Conclusion: It is possible that duration of episode affects effectiveness of the UK BEAM exercise package, but not the UK BEAM manipulation package.

Systematic reviews show beneficial effects of spinal manipulation, general exercise, and ‘active management’. A national randomised factorial trial in primary care (UK BEAM trial) was designed to evaluate the effectiveness of these treatments for back pain. We will present the characteristics of participants recruited into the trial and preliminary health outcomes at one and three months.

Back pain patients, recruited from over 150 UK practices, were randomised to receive GP management, exercise classes, manipulation (either in NHS or private premises) or both manipulation and exercise classes. At one, three and twelve months, participants completed postal questionnaires which included questions about general health, experience of back pain, beliefs about back pain, psychological profile, functional disability and costs to both the NHS and the participants themselves.

The trial recruited 1334 participants, of which 84% and 77% completed one and three month questionnaires respectively. At randomisation, the mean Roland Disability Questionnaire (RDQ) score was 9.0 points (sd=4.0). This improved to 6.8 points (sd=4.8) at one month and to 5.5 points (sd=5.0) at three months.

Preliminary blinded results show an improvement in RDQ scores across all participants. The primary analysis, available late in 2002, will estimate the main effects of exercise and manipulation, each compared to GP care.