Physiotherapists' attitudes and beliefs about low back pain influence their clinical decisions and advice

Key points

1. 01Mean Back-PAQ score of 82.7 (out of 170) indicated generally helpful beliefs, but protection and special-nature themes scored highest for unhelpful beliefs
2. 02Unhelpful beliefs were significantly associated with clinical decisions toward back protection and movement avoidance (r = -0.47, p < 0.001)
3. 03Physiotherapists with more experience, current disabling low back pain, or no MSK specialty had more unhelpful beliefs
4. 04Only 23.3% of physiotherapists gave fully guideline-consistent advice about loading the back without adding protective caveats
5. 05Educational strategies targeting physiotherapists' own misconceptions are needed before guideline implementation can succeed

How it was conducted

Design

Cross-sectional survey

Participants

288 French-speaking Swiss physiotherapists, recruited via snowball sampling (Dec 2018 - Mar 2019)

Beliefs measure

Back Pain Attitudes Questionnaire (Back-PAQ): 34 items, score 34-170, higher = more unhelpful beliefs

Clinical decisions measure

Clinical vignette depicting a young woman with acute low back pain; 5 scored questions on activity, work, bed-rest, treatment, and protective advice

Analysis

Spearman correlations, independent t-tests, and multiple linear regression; significance level alpha < 0.05

What they found

• Mean Back-PAQ score: 82.7 (SD 17.2); general population mean ~113, patients mean ~120
• Individual characteristics explained 18.2% of variance in Back-PAQ score (F(7,280) = 10.109, p < 0.001)
• Physiotherapists with current disabling low back pain had mean Back-PAQ score of 96.82 vs 81.18 for those without (p < 0.001)
• MSK specialty associated with lower (more helpful) Back-PAQ score: 81.29 vs 85.91 (p = 0.03)
• Total vignette score (mean 2.8 out of 5, SD 1.3) was negatively correlated with Back-PAQ score (r = -0.47, p < 0.001)
• 53.1% recommended work limitation toward avoidance; Back-PAQ mean 87.23 vs 77.57 for no limitation (difference 9.66, 95% CI 5.81-13.5, p < 0.001)
• 49.0% recommended bed rest toward avoidance; Back-PAQ mean 86.59 vs 78.97 (difference 7.62, 95% CI 3.72-11.51, p < 0.001)
• 68.8% gave protective posture advice; associated Back-PAQ mean 89.09 vs 68.64 for those who did not (difference 20.45, 95% CI 16.85-24.04, p < 0.001)
• Only 23.3% selected the two fully guideline-consistent back-loading advice statements without adding any protective advice
• 85.1% chose progressive exercises as a first treatment; not choosing exercises was associated with higher (more unhelpful) Back-PAQ score (mean 89.95 vs 81.43, p = 0.003)
• Electrotherapy use (19.8% of respondents) associated with Back-PAQ mean 94.25 vs 79.85 for non-users (difference 14.39, 95% CI 9.66-19.12, p < 0.001)

Limitations

• Non-probabilistic snowball sampling may introduce self-selection bias and limits generalisability to all Swiss physiotherapists
• Response rate cannot be calculated because the total reach of the invitation chain is unknown
• Clinical vignettes measure intended rather than actual clinical behaviour
• Two of the five vignette questions were newly developed for this study and their measurement properties have not been independently validated

Why it matters

For patients

Patients may receive inconsistent advice, hearing both encouragement to stay active and warnings to protect the back, depending on their physiotherapist's own beliefs.

For clinicians

Physiotherapists should reflect on their own beliefs about back vulnerability and protection, as these beliefs demonstrably shape the advice they give and may reinforce unhelpful fear-avoidance patterns in patients.

For readers

This study highlights that guideline uptake for low back pain management requires directly addressing clinician beliefs, not just distributing guidelines.