Background: Sciatica is an important cause of pain and disability but relatively few studies have looked at predictors of outcome in sciatica populations. Prognostic studies in non-specific low back pain are more common, but it has been suggested that the prognostic indicators for sciatica may be different. Our aim was to systematically review and describe the literature investigating the predictors of outcome in sciatica populations.

Methods: A systematic literature search of the databases (Medline, EMBASE, and CINAHL) and reference list of identified articles was conducted. Studies were included if they described subjects with sciatica, had a follow-up period of at least three months and measured outcomes including pain, disability, recovery, psychological outcomes or return to work. Methodological quality was assessed using a 15 item checklist.

Results: 596 papers were identified but only 12 met the criteria for inclusion. A combination of individual (e.g. gender, BMI), biomedical (e.g. size of disc prolapse, neurological deficit), social (e.g. job satisfaction, social status, manual labour) and psychological (distress, mental health) predictors of outcome were reported.

Conclusions: There are few high quality studies that have investigated prognostic factors associated with persistence of sciatic symptoms. Those identified explored a range of different factors, in a variety of settings and in subjects with variable duration of symptoms. Although the studies are difficult to interpret due to heterogeneity of the techniques used in analysis and presentation, they seem to suggest that clinical, occupational and individual factors might be more strongly associated with outcome than psychological factors in sciatica populations.

Conflicts of Interest: None

Funding: None

Background: Last year we presented the STarT Back Tool, which is validated for use in Primary Care. It subgroups patients into 3 categories (high, medium and low risk) on the basis of modifiable risk factors for chronicity. We are now piloting the feasibility of using the tool as part of a new approach to sub-grouping and targeting back pain in primary care.

Methods: The physiotherapy interventions for the 3 subgroups were developed after reviewing the literature, current guidelines, the content of existing targeted treatment programmes, and convening workshops with internationally recognised experts. Both the intervention training modules, and the targeted treatments were piloted. Consecutive back pain consulters were identified using GP electronic Read Codes (weekly downloads) and invited to attend the study’s back pain clinic. Consenting patients completed a baseline questionnaire and were classified by the tool into one of 3 sub-groups.

Results: 60 patients were recruited. 50 patients were allocated to receive treatment according to their subgroup allocation and 10 patients (control group) received a triage physiotherapy assessment (usual care) to decide if they needed further physiotherapy treatment. Primary outcomes include the Roland Morris Disability Questionnaire and the Pain Catastrophising Scale. Three-month follow-up postal questionnaires are currently being administered and outcomes will be presented at the conference. Clinicians involved (GPs, and physiotherapists) will be interviewed to identify the feasibility of this approach.

Conclusions: Once feasibility is established we will take this developmental work forwards into the clinical trial arena to investigate whether this novel “sub-grouping for targeted treatment” approach provides a cost effective way of reducing long-term risk of chronic disability in patients consulting their GP with back pain.