Effectiveness of Placebo Interventions for Patients With Nonspecific Low Back Pain

A Systematic Review and Meta-analysis

Rob H.W. Strijkers; Marco Schreijenberg; Heike Gerger; Bart W. Koes; Alessandro Chiarotto

Disclosures

Pain. 2021;162(12):2792-2804. 

In This Article

Results

Characteristics and Included Studies

The literature search revealed 7803 potentially relevant articles, and 21 RCTs were included in the review (Figure 1). Of the 21 studies, one reported on subjects with acute LBP,[58] one reported on subacute LBP,[16] and the remaining 19 studies reported on patients with chronic LBP.[2,4–7,14,35,37–43,45,46,50,54,62] Study characteristics of all 21 included RCTs are shown in Table 1.

Figure 1.

PRISMA 2009 flow diagram. LBP, lower back pain.

Risk of Bias Analysis

Three studies were identified as having an overall high risk of bias.[14,41,54] All of the studies had a high risk of bias on measurement of the outcome, which is inherent to the study setup. The risk of bias analysis for each individual study is presented in Figure 2.

Figure 2.

Cochrane risk of bias analysis summary of included studies.

Primary Endpoints

Chronic low Back Pain. Pain Intensity: There was moderate-quality evidence (19 RCTs, N = 1443) that placebo interventions reduced back pain intensity compared with no intervention at short-term follow-up (SMD −0.37, 95% CI −0.55 to −0.18, P < 0.0001, τ2 = 0.11) (refer to Figure 3). This between-group difference translates to approximately 8 points on a 0 to 100 pain scale, corresponding to approximately 14% major improvement in the placebo group, which did not reach the predefined 20% improvement for clinical relevance. Sensitivity analyses excluding RCTs with a high risk of bias showed a smaller but still statistically significant effect in favor of placebo interventions (SMD −0.27, 95% CI −0.47 to −0.06, P = 0.0003). There was low-quality evidence (3 RCTs, N = 154) that placebo interventions are not significantly better than no intervention for back pain reduction at medium-term follow-up (SMD −0.26, 95% CI −0.59–0.06, P = 0.11 τ2 = 0.00) (refer to Figure 3). Sensitivity analyses excluding high risk of bias studies confirmed this finding (SMD −0.19 95% CI −0.56 to 0.18, P = 0.31). No long-term follow-up data were available.

Figure 3.

Forest plots for pain intensity at short-term and medium-term follow-up. Risk of bias assessment: The red circle corresponds with a high risk of bias. The yellow circle corresponds with an unclear risk of bias. The green circle corresponds with a low risk of bias.

Physical Functioning: There was moderate-quality evidence (11 RCTs, N = 775) that placebo interventions improved physical functioning compared with no intervention at short-term follow-up (SMD −0.19, 95% CI −0.39 to −0.01, P = 0.07 τ2 = 0.06) (refer to Figure 4). This difference corresponds to approximately 1.3 points between-group difference on the 0 to 24 Roland–Morris Disability Questionnaire, which is not considered to be a clinically relevant effect. Sensitivity analysis excluding high risk of bias studies confirmed this finding (SMD −0.17 95% CI −0.32 to −0.02, P = 0.03). Only one study reported on medium-term outcomes for physical functioning, showing no statistical difference between placebo intervention and no intervention (MD −2.60 95% CI −9.09 to 3.89 P = 0.43 [scale 0–24]). No study reported on long-term outcomes.

Figure 4.

Forest plot for physical functioning at short-term follow-up. Risk of bias assessment: The red circle corresponds with a high risk of bias. The yellow circle corresponds with an unclear risk of bias. The green circle corresponds with a low risk of bias.

Health-related Quality of Life: There was high-quality evidence (3 RCTs, N = 267) that placebo interventions improved quality of life on the physical component summary score of the SF-36 compared with no intervention on short-term follow-up (MD 2.71, 95% CI 0.71–4.71, P = 0.008, τ2 = 0.34 [scale 0–100]) (refer to Figure 5). There was high-quality evidence (3 RCTs, N = 267) that placebo intervention does not improve quality of life on the mental component summary score of the SF-36 compared with no intervention on short-term follow-up (MD −0.49, 95% CI −3.49 to 2.51, P = 0.75, τ2 = 2.85 [scale 0–100]) (refer to Figure 5). No study data were available for medium-term or long-term follow-up.

Figure 5.

Forest plots of health-related quality of life at short-term follow-up. Risk of bias assessment: The red circle corresponds with a high risk of bias. The yellow circle corresponds with an unclear risk of bias. The green circle corresponds with a low risk of bias.

(sub)Acute low Back Pain. No meta-analysis for acute LBP was performed because of lack of studies on any of the primary endpoints. One study reported on physical functioning.[58] There was no statistical difference between placebo intervention and no treatment (MD 7.0; 95% CI −0.13 to 14.13, P = 0.05 [scale 0–24]). Another study reported on pain intensity in short-term follow-up.[16] Short-term pain intensity was not different between placebo intervention and no treatment (MD 7.0; 95% CI −6.02 to 20.02; P = 0.29 [scale 0–100]).

Subgroup Analysis

Subgroup analysis of open-label placebo RCTs as compared to placebo control intervention RCTs showed a similar effect on short-term pain intensity (SMD −0.33 vs SMD −0.37, appendix C, supplemental digital content available at https://links.lww.com/PAIN/B345, Figure 1). There was high-quality evidence (3 RCTs, N = 335) that placebo acupuncture intervention improved back pain intensity compared with no intervention at short-term follow-up (SMD −0.37, 95% CI −0.64 to −0.10, P = 0.007). There was very low-quality evidence (2 RCTs, N = 80) that placebo taping does not improve back pain compared with no intervention at short-term follow-up (SMD 0.09, 95% CI −0.52 to 0.70, P = 0.77). There was high-quality evidence (2 RCTs, N = 345) that placebo injection intervention does not improve back pain compared with no intervention at short-term follow-up (SMD 0.03, 95% CI −0.18 to 0.24, P = 0.77). There was very low-quality evidence (2 RCTs, N = 89) that placebo spinal manipulation does not improve back pain compared with no intervention at short-term follow-up (SMD −0.28, 95% CI −1.15 to 0.59, P = 0.53). There was low-quality evidence (2 RCTs, N = 80) that placebo taping therapy does not improve physical functioning compared with no intervention at short-term follow-up (SMD 0.15, 95% CI −0.29 to 0.59, P = 0.51). We found high-quality evidence (2 RCTs, N = 205) that open-label placebo pill treatment improves short-term pain intensity compared with no intervention (SMD −0.33 95% CI −0.61 to −0.05, P = 0.02) and also improves physical functioning (SMD −0.34 95% CI −0.62 to −0.07, P = 0.01). None of these between-group differences was clinically relevant. We found no difference in short-term pain intensity outcomes with exclusion of open-label placebo trials (SMD −0.437, 95% CI −0.5559 to −0.216, P = 0.20007) (appendix C, supplemental digital content available at https://links.lww.com/PAIN/B345).

processing....