Best Practice Physiotherapy for Patients with Low Back Pain in Primary Care : Clinical Outcomes and Explanatory Factors

Sammanfattning: Background: Clinical practice guidelines provide general recommendations informing best practice physiotherapy for low back pain (LBP). Despite increased research on LBP, the recommendations have barely changed over the last few decades and the burden of LBP remains. New research strategies have been suggested to further understand the complexity of factors influencing recovery of LBP. Several mechanisms are expected to underpin the benefits of physiotherapy care. How different factors influence and explain treatment outcome in patients in different phases of primary care pathway needs to be better understood for the development of better targeted LBP interventions.The overall aim of this thesis was to investigate if and how best practice physiotherapy primary care for patients with LBP can improve clinical outcome.Methods: The four papers in the thesis are based on two randomised controlled trials (RCT) evaluating best practice physiotherapy for patients with LBP in different phases of the primary care pathway reflecting the variation of patients’ needs. Two papers investigated pre-surgery physiotherapy compared to being on a waiting-list in patients with degenerative lumbar spine disorders who are surgical candidates (n = 197). Effects on walking ability and quadriceps femoris strength were evaluated. Associations between the pre-surgery physical factors with 1-year post-surgery physical activity (PA) level were analysed using multiple linear regression. The dose-response relationship was investigated comparing the effects of attending ≤11 treatment sessions with ≥12 treatment sessions. Multiple mediation analyses and conditional process analyses were used to explore physical and psychosocial factors as mediators and patients’ treatment expectations as a moderator of the treatment’s effects on disability, back pain intensity, health-related quality of life (HRQoL), and self-rated health.Two papers were based on a stepped cluster RCT, where a physiotherapy primary healthcare model for LBP (the BetterBack MoC) was regionally implemented. Patients seeking care for LBP were allocated to either the BetterBack MoC (after implementation) or to routine care (before implementation) (n = 467). In a prospective cohort study within the RCT, the associations between patients’ initial illness perceptions and outcomes in disability, back pain intensity, HRQoL, and self-care enablement (i.e., perceived ability to understand and cope with LBP) after 3 and 12 months were explored using stepwise linear regression. Single mediation analyses were used to test whether a priori hypothesised patients’ illness perceptions and self-care enablement at 3 months mediated effects in disability and pain at 6 months of care according to the BetterBack MoC compared to routine care. Exploratory mediation analyses were also used to compare guideline-adherent care with non-adherent care. Guideline-adherent care was defined as care that included education and exercise interventions and did not include non-evidence-based interventions, referral to specialist care, or imaging.Results: Small positive effects from pre-surgery physiotherapy were seen in walking ability and quadriceps femoris strength. No clear dose-response relationship could be demonstrated when comparing the effects of ≤11 treatment sessions with ≥12. Pre-surgery physical outcome measures together explained 27.5% of the variation in PA level 1-year post-surgery, mainly explained by the pre-surgery PA level. The effect of the pre-surgery physiotherapy on patients’ PA level partly explained the treatment’s effect on self-rated health. Furthermore, among biopsychosocial factors, self-efficacy related to activities of daily living (ADL) partly explained the effect on all outcomes. PA related fear avoidance beliefs partly explained the effects on pain and self-rated health. Patients’ treatment expectations moderated the effect in all outcomes. High expectations had a positive moderating effect, while expectation of full recovery had a suppressive effect. In patients seeking physiotherapy for LBP, negative initial prognosis and treatment expectations were associated with worse scores in several outcomes at 3- and 12-month follow-ups. Patients’ illness perceptions and self-care enablement did not explain the effects of care after implementing the BetterBack MoC. This was mainly due to the BetterBack MoC not having superior effects over routine care on the hypothesised mediators. Illness perceptions and self-care enablement at 3 months were associated with disability and pain at 6 months. Further, these factors partly explained the effects of guideline-adherent care in disability and pain.Conclusions: Patients’ illness perceptions, self-care enablement, ADL self-efficacy, and PA related fear-avoidance beliefs were supported to be potential factors explaining the effect on clinical outcomes of best practice physiotherapy for LBP in primary care. Patients’ initial expectations regarding the prognosis and treatment may influence prospective outcomes including patients’ self-care enablement. More specifically, pain and disability outcomes of best practice physiotherapy for patients seeking care for LBP may improve by targeting patients’ illness perceptions and self-care enablement. Patients who are surgical candidates can increase their physical capacity and walking related performance through best practice physiotherapy. In this pre-surgery phase, the treatment should include targeting patients’ ADL self-efficacy, PA related fear-avoidance beliefs, and PA level for improved disability, pain, and HRQoL.