The impact of therapeutic alliance in physical therapy for chronic musculoskeletal pain: a systematic review of the literature

Key points

1. 013 of 7 studies found a significant positive association between stronger therapeutic alliance and reduced chronic musculoskeletal pain
2. 02One RCT showed sham electrical therapy plus enhanced alliance reduced pain by 54.5%, matching active therapy with limited alliance (45.6% reduction); combining both produced 77.4% reduction
3. 03Trust and individualized treatment planning were the most consistently identified positive influences on alliance strength across multiple studies
4. 04Patient anger expression and hostility negatively affected patient-rated alliance; combined high anger and depression negatively affected therapist-rated alliance
5. 05Patients and therapists do not always perceive alliance strength in the same way, suggesting physical therapists should actively seek patient feedback

How it was conducted

Design

Systematic review following PRISMA guidelines (PROSPERO registration CRD42016042587)

Databases searched

PubMed, CINAHL, and Embase, from inception to January 2017

Studies included

7 studies: 1 RCT (low risk of bias), 2 secondary analyses of RCTs, and 4 non-randomized or qualitative observational studies

Population

Adults with chronic musculoskeletal pain (pain lasting more than 3 months), predominantly chronic low back pain

Primary outcome

Pain intensity outcomes and factors influencing therapeutic alliance strength

Risk of bias

Six of seven studies had significant risk of bias; only one RCT (Fuentes et al., 2014) was rated low risk

What they found

• Cheing et al. (2014): 0.23 cm improvement in VAS pain intensity for every 1-point increase in working alliance score (beta = -0.23, p = .002)
• Ferreira et al. (2013): 0.044 cm improvement on VAS for every 1-point increase on WATOCI (beta = -0.044, 95% CI = -0.070 to -0.017, p = .001)
• Fuentes et al. (2014): enhanced TA group showed significantly greater pain reduction vs limited TA (p < .01)
• Fuentes et al. (2014): sham therapy plus enhanced TA reduced pain by 54.5% (PI-NRS baseline 4.10 cm, post-treatment 1.88 cm)
• Fuentes et al. (2014): active therapy plus limited TA reduced pain by 45.6% (PI-NRS baseline 4.01 cm, post-treatment 2.18 cm)
• Fuentes et al. (2014): active therapy plus enhanced TA produced the greatest reduction of 77.4% (PI-NRS baseline 4.03 cm, post-treatment 0.89 cm)
• Burns et al. (1999): only moderate correlation found between patient-rated and therapist-rated alliance scores
• Agreement between independent reviewers in full-text selection was 46% initially, rising to 87% after structured discussion

Limitations

• Only 7 studies met inclusion criteria and 6 carried significant risk of bias, preventing definitive conclusions about effect size
• Only one study (Fuentes et al., 2014) directly manipulated therapeutic alliance as an independent variable; most others measured it observationally
• Qualitative studies relied on self-selected or treatment-responder samples, introducing substantial selection bias
• Therapeutic alliance is difficult to isolate as a construct because it may be confounded by outcome expectancy and patient beliefs about pain

Why it matters

For patients

Patients with chronic musculoskeletal pain may get better results from physical therapy when they feel genuinely heard, trust their therapist, and are actively involved in setting their own goals.

For clinicians

Physical therapists should prioritize building a trusting, individualized relationship alongside exercise and manual therapy, and consider communication skills training to address patient anger or hostility that can weaken the alliance.

For readers

This review provides preliminary but low-certainty evidence that the interpersonal quality of physical therapy matters for pain outcomes, and calls for better-designed trials that experimentally manipulate alliance as a treatment component.