Does clinician treatment choice improve the outcomes of manual therapy for nonspecific low back pain? A metaanalysis

Peter Kent, Darryn Marks, Warrick Pearson, Jenny Keating

Research output: Contribution to journalReview articleResearchpeer-review

47 Citations (Scopus)


OBJECTIVE: The purpose of this study is to quantitatively compare outcomes for trials when treating clinicians did, or did not, have the discretion to decide on treatment technique.

METHODS: CINAHL, EMBASE, MEDLINE, the Physiotherapy Evidence Database, the Cochrane Controlled Trials register, reference list searching, and citation tracking were investigated. Ten randomized controlled trials (RCTs) of mobilization and manipulation for nonspecific low back pain (NSLBP) met the inclusion criteria. The effectiveness of manual therapy with and without clinician technique choice was assessed using descriptive statistics and metaanalysis for the outcomes of pain and activity limitation.

RESULTS: In approximately two thirds of the included RCTs, clinicians had choice of treatment technique. There were no systematic differences favoring results for RCTs that did allow clinician choice of treatment technique.

CONCLUSIONS: Few quality studies are available, and conclusions on the basis of these data need to be interpreted with caution. However, allowing clinicians to choose from a number of treatment techniques does not appear to have improved the outcomes of these RCTs that have investigated the effect of manual therapy for NSLBP. If tailoring manual therapy treatment to NSLBP patients does positively impact on patient outcomes, this is not yet systematically apparent.

Original languageEnglish
Pages (from-to)312-22
Number of pages11
JournalJournal of Manipulative and Physiological Therapeutics
Issue number5
Publication statusPublished - Jun 2005
Externally publishedYes


Dive into the research topics of 'Does clinician treatment choice improve the outcomes of manual therapy for nonspecific low back pain? A metaanalysis'. Together they form a unique fingerprint.

Cite this