ICER = incremental cost effectiveness ratio; n/a = not available; NR = not reported; RCT = randomised clinical trial; QALY = quality-adjusted life year; Prob. CE= Probability intervention is cost-effective at a £20,000/£30,000 threshold.

Cost/effect in order of least to most costly intervention.

Full incremental analysis of available strategies: first strategies are ruled out that are dominated (another strategy is more effective and has lower costs) or subject to extended dominance (the strategy is more effective and more costly but the incremental cost effectiveness ratio is higher than the next most effective option and so it would never be the most cost effective option); incremental costs, incremental effects and incremental cost effectiveness ratios are calculated for the remaining strategies by comparing each to the next most effective option.

Resource use data (2002-2005) and unit costs (2003/3) may not reflect the current NHS context. EQ-5D tariff used is not stated (although as UK study judged likely to be UK tariff). Study does not include all non-invasive treatment options.

Time horizon may not be sufficient to capture all benefits and costs if benefits persist beyond 18 months. Within-trial analysis and so does not reflect full body of available evidence for this comparison.

Cost components incorporated: interventions, primary care contacts (GP, practice nurse, physiotherapist, other), secondary care contacts (hospital admissions and outpatient appointments).