Are within and/or between session improvements in pain and function prognostic of medium and long-term improvements in musculoskeletal problems? A systematic review

Key points

1. 0113 studies met inclusion criteria, covering mostly low back pain (10 of 13 studies)
2. 02Of 10 outcomes examined, only 1 supported early within-session changes as prognostic, 3 did not, and 5 were inconclusive
3. 03Evidence quality was rated moderate for 1 outcome, low for 1, and very low for 8 outcomes
4. 04Only 1 study evaluated between-session changes and found very low-quality evidence for prognostic value
5. 05No studies examined whether early improvements in function predicted later outcomes

How it was conducted

Design

Systematic review following PRISMA guidelines

Databases searched

PubMed/Medline, Cochrane, CINAHL, AMed from inception to December 2018

Included studies

13 studies (7 prospective cohort studies, 6 secondary analyses of RCTs)

Populations

Predominantly low back pain (10 studies); 1 each for neck, shoulder, and hip pain

Primary prognostic factors

Within-session and between-session changes in pain location and pain intensity

Outcomes assessed

Pain, disability, return to work, and global effects at medium-term (4 weeks to 6 months) and long-term (greater than 6 months)

What they found

• Of 9 outcomes examining within-session pain location or intensity changes, 5 were inconclusive, 3 were negative (no prognostic value), and 1 was positive
• Within-session changes in pain location for long-term disability: 5 of 6 follow-up measurements did not support prognostic value; rated moderate quality evidence for no effect
• Within-session changes in pain location for medium-term pain: 5 of 10 follow-up measurements supported prognostic value, 5 did not; rated low-quality and inconclusive
• Within-session changes in pain location for long-term pain: 4 of 7 follow-up measurements did not support prognostic value, 3 did; rated very low-quality and inconclusive
• Within-session changes in pain intensity for medium-term disability: 2 studies, both found no prognostic value; rated very low-quality
• Within-session changes in pain intensity for global effects at medium-term: 2 studies with mixed results; rated very low-quality and inconclusive
• Between-session changes in pain intensity for medium-term disability: 1 study found an association but CIs were very wide; rated very low-quality
• Inter-rater agreement on data extraction and risk of bias assessment was 90% (kappa 0.82)

Limitations

• Only 13 studies were included, reflecting the scarcity of research on this specific topic
• Most included studies had moderate to high risk of bias, particularly in study attrition and analysis/reporting
• The majority of primary studies did not report confidence intervals, making it impossible to fully assess imprecision using GRADE
• No studies examined early improvements in function as a prognostic factor, leaving a major gap in the evidence

Why it matters

For patients

Patients should not assume that feeling better after their first treatment session guarantees a good long-term recovery, nor should a lack of immediate improvement discourage them from continuing appropriate care.

For clinicians

Clinicians using immediate symptomatic response (such as centralization or pain reduction in the first session) as the sole guide for continuing or discontinuing treatment should be aware there is little high-quality evidence supporting this approach as a reliable long-term predictor.

For readers

This review highlights a significant evidence gap: the widely used clinical practice of using early session improvements to guide ongoing management lacks robust prognostic support, and more rigorous research with adequate follow-up and reported effect sizes is needed.