Agreement between high-quality clinical practice guidelines in their treatment recommendations for low back pain: a...

International and national clinical practice guidelines for low back pain often differ in their treatment recommendations. Previous studies have outlined disagreement between recommendations when comparing guidelines of varying quality, challenging the value of producing them. There is a need to quantify agreement between high-quality low back pain guidelines for monitoring recommendation changes over time and guiding future research for treatments with disagreement.

The aim of this review was to assess agreement between high-quality clinical practice guidelines in their treatment recommendations for low back pain published 1 January 2016 to 14 March 2024.

Systematic review.

Strength and direction of guideline recommendations were pooled for each low back treatment and the level of agreement or disagreement was reported as a percentage. The number of recommendations for individual treatments was also reported.

We searched 6 databases to identify high-quality clinical practice guidelines defined as those that scored ≥70% in 3 out of 6 AGREE II domains, including "Rigour of Development". Two authors independently screened titles and abstracts, then full texts to determine eligibility. Recommendations were stratified by low back pain chronicity and direction and strength of recommendation.

Of 4,920 publications screened, 22 high-quality clinical practice guidelines were included. Overall, there were 588 recommendations for 181 treatments (range 1-125 recommendations per CPG). In 154 instances clinical practice guidelines made no recommendation about a treatment although in 97 (63%) a recommendation was made by at least 1 other clinical practice guideline. Agreement could be assessed for 433 recommendations for 122 low back pain treatments. There was complete or partial agreement for almost 2-thirds of recommendations; complete agreement (ie, both direction and strength agreed) for 26%, and partial agreement (ie, same direction but different strength) for 39% across guidelines. Recommendations generally supported use of patient education, provision of a self-management plan and fear avoidance and physical therapy and did not support use of invasive treatments, electrotherapy or assistive devices. There was disagreement (ie, direction of a recommendation differed) for 35% of recommendations, mostly pharmacological interventions.

More research is needed to understand reasons for these disagreements as well as the variation across clinical practice guidelines in electing to make a recommendation.