Placebo responses in non-surgical trials for knee osteoarthritis: A systematic review and meta-analysis

Placebo responses in non-surgical trials for knee osteoarthritis: A systematic review and meta-analysis | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Placebo responses in non-surgical trials for knee osteoarthritis: A systematic review and meta-analysis Robin Holtedahl, Alexandra Christina Hott, Inger Johanne Hansen, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7888576/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background Previous reviews have demonstrated notable placebo responses in placebo-controlled clinical trials assessing osteoarthritis treatments; however, there is limited consensus regarding the relative importance of factors influencing these responses. This review and meta-analysis aimed to (1) evaluate effect sizes observed in the placebo arms of knee osteoarthritis trials published since 2008, and (2) identify factors contributing to explained variance through meta-regression analysis. Methods Outcome data at baseline and follow-up were extracted, and Hedges’ g was calculated as a measure of effect size for each study using a random effects model. Heterogeneity was quantified using the I² statistic. Sources of variance were investigated through multivariate meta-regression involving 16 clinically relevant covariates. The Cochrane RoB-2 tool was used to assess risk of bias. Results Searches provided 179 trials that met inclusion criteria. Of these, 109 trials (60%) reported Hedges’ g values ≥ 0.5. Substantial heterogeneity was observed (I² = 89%, p < 0.0001), with a broad prediction interval (-0.30 to 1.63). In multivariate meta-regression two-thirds of the variance remained unexplained, while the covariates baseline pain, commercial funding, effect size in active treatment arms and geographical region together accounted for 33% of explained variance. Risk of bias, duration of follow-up, year of publication, and method of placebo administration were not found to significantly influence the outcomes. Conclusion : Most trials demonstrated moderate to large placebo effect sizes, but there was marked between-trial statistical heterogeneity and a substantial proportion of the variance remained unexplained. An important limitation is the lack of data regarding psychological variables, such as participant expectations and the patient–provider relationship, which are recognized as major determinants of placebo effects. Placebo Sham Knee osteoarthritis Non-surgical management Pain outcomes Systematic review Meta-analysis Figures Figure 1 Figure 2 Figure 3 Background The knee is the most common site of osteoarthritis (OA), with rising global prevalence [ 1 ]. OA is a degenerative joint disease, with gradual deterioration of articular cartilage, causing pain, stiffness, and reduced mobility. Non-surgical treatments focus on symptom relief and delaying progression, with treatment guidelines recommending education, exercise, and weight loss as initial care. Other treatments lack consistent evidence, whereas a patient-centred approach based on a biopsychosocial model is strongly advised [ 2 ]. Placebo or sham treatments are interventions that are presumed to have no specific therapeutic effects for the condition being treated. Placebo responses encompass all health-related changes resulting from administering a placebo and include placebo effects and as well as non-specific factors like regression to the mean and the natural course of the condition. The net benefit of a treatment is assessed by comparing the outcomes in the active treatment and placebo groups. To distinguish between the placebo effect and non-specific effects a no-treatment group is necessary [ 3 ]. Several factors have been identified as influencing the placebo effect, including the specific characteristics of the intervention, the patient’s beliefs and expectations, the quality of the patient–therapist relationship, and the level of empathy, confidence, and optimism communicated by the physician [ 4 ]. In the context of clinical trials placebo effects have been linked to the route of administration (invasiveness or complexity of the procedure) [ 5 , 6 ], trial duration [ 7 ], length of follow-up [ 8 , 9 ], sample size [ 10 ], period of study [ 11 ], and efficacy of the active treatment [ 12 , 13 ]. Subjective outcomes, including pain, are thought to be associated with larger placebo effects than objective outcomes [ 14 ]. The aims of this meta-analysis were 1) to analyse pain-related outcomes within the placebo arms of randomized controlled trials (RCTs) of non-surgical treatments for knee OA (KOA) published since 2007; 2) through meta-regression analysis to explore covariates potentially contributing to placebo responses, including those suggested in previous reviews, and 3) perform subgroup analyses based on relevant categorical covariates and risk of bias. Methods The protocol of this review was registered at PROSPERO (CRD 420251113096) and the review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) reporting checklist [ 15 ]. Data sources and eligibility criteria After registration of the protocol systematic searches were performed for eligible studies published between January 2008 and May 2025 in PubMED and EMBASE. The search terms were “knee osteoarthritis” AND (placebo OR sham) with the filters “randomized controlled trial” and English language. Two authors (RH and AH) assessed eligibility in two stages: 1) by screening each trial’s title and abstract and 2) by reviewing potentially eligible articles in full text. Each reviewer stated their reason for exclusion and consensus was reached in cases of disagreement. The inclusion criteria included 1) randomized, parallel-group trials of treatments for knee osteoarthritis that were double- or triple-blinded; 2) included a placebo (sham) arm (optionally also a no-treatment arm); 3) assessed the efficacy of any treatment except psychological or exercise-based interventions or anatomy-altering surgical procedures; 4) reported a pain-related knee-specific outcome measure on a continuous scale; and 5) enabled calculation of effect size, i.e.; provided standard deviation [SD], standard error [SE] or 95% confidence intervals [CI]. For more detailed eligibility criteria see Additional file 1: table S1 ). Data Extraction The main characteristics of each trial were entered by the first author into a standardized data extraction form using Microsoft Access v.2412. For a complete list of items, see Additional file 1: table S2. Outcome values in the placebo, active treatment and no-treatment control groups as well as registration and source of funding were extracted by the first author and confirmed by two independent authors. The WOMAC pain subscale was chosen as the preferred outcome measure; if not provided, total WOMAC or other knee-specific pain outcome measures. As radiological OA severity and duration of disease were not consistently described such data were not extracted. Pre-to-post change values were checked for direction. For metaregression analysis, 16 covariates judged to be clinically relevant were extracted (see Additional file 1: table S3). In trials that described ongoing treatments the outcome value at the end of the trial was extracted. In trials that reported outcome at later follow-up, including single-intervention trials, outcome value at the latest follow-up was extracted, with a 2-year upper limit. Because outcome measures and scales varied across studies, baseline values were converted to a 10 cm visual analogue scale (VAS), with higher scores indicating worse pain. The method of administering placebo was categorized into five groups, representing varying levels of perceived invasiveness or complexity: 1) physical (e.g., short-wave ultrasound, cryotherapy, taping), 2) oral (tablet, capsule, sachet, syrup), 3) intra-articular (in most cases normal saline), 4) topical (gel, ointment, liniment) and 5) intravenous, subcutaneous or intradermal (including acupuncture). Because interventions in the last group share the characteristics of penetration of skin and / or soft tissue it was judged reasonable to group them together. Risk of bias assessment Two reviewers independently assessed the risk of bias using the Revised Tool for Risk of Bias in Randomized Trials [ 16 ]. The categories were low, some concerns and high. Reporting bias was evaluated by checking if the trials were registered and whether their protocols were submitted before the trials began. Trial size (> 49 patients) and loss-to-follow-up (> 15%) were added as additional criteria. Consensus was reached by discussion. Sensitivity analysis was performed by removing trials that were at high risk of bias on any domain. Statistical analyses Calculations were performed using Comprehensive Meta-Analysis version 4.0.000 software (available from www.meta-analysis.com ), MedCalc (MedCalc® Statistical Software version 22.021 (MedCalc Software Ltd, Ostend, Belgium; https://www.medcalc.org ; 2024) and Microsoft Excel version 2503 (Microsoft, Redmond, WA, USA). The Hedges’ g value, which corrects for small sample sizes, was used as a measure of the effect size of change from baseline to follow-up. By convention an effect size of 0.8 is described as large; 0.5–0.8, moderate; and < 0.5, small ( www.cochrane.org/handbook ). Calculations were preferably based on each trial’s SD of change (alternatively SE, 95% CI or p-value) [ 17 ]; if not provided, on a recommended formula, with the pretest-posttest correlation set to 0.5 [ 18 ]. Similar calculations were performed for treatment and no-treatment arms. Two-sided p-values of < 0.05 were considered as statistically significant. A random effects model using the DerSimonian and Laird method was used because the studies were sampled from multiple populations. The 95% CIs were quantified in a weighted fashion using the inverse variance approach to incorporate both within- and between-trial variance. I 2 statistics were calculated to assess the degree of heterogeneity across studies, i.e., the proportion of variability not due to sampling error. Heterogeneity was considered high when presenting a value greater than or equal to 50%, as suggested elsewhere [ 19 ]. To assess whether future similar trials would be expected to provide comparable results, prediction intervals were also calculated [ 20 ]. Subgroup analyses were performed to compare effect sizes across geographical regions and across different routes of placebo delivery. In the univariate meta-regression analysis sixteen predefined trial-level covariates were entered into the model. In the next stage, a multivariate regression model was built by starting with the complete set of covariates and stepwise eliminating covariates with p-values ≥ 0.10. R 2 values of each of the remaining covariates as well as overall R 2 were estimated. Results Study selection After removal of duplicates searches identified a total of 549 trials. Of these, 295 trials were excluded after first screening and 75 after second screening, leaving 179 trials available for analysis. A flowchart of the selection process is shown in Fig. 1 , including reasons for exclusion. Trial characteristics The placebo arms comprised in total 13 031 patients at inclusion. The average loss to follow-up was 13%. Median sample size was 66 (interquartile range, IQR 46 to 150). Four trials with 141 participants in total also included a no-treatment arm. The average follow-up was 20 weeks (range 10 to 26) for single-procedure trials (n = 20) and 20 weeks (range 4 to104) for ongoing-treatment trials (n = 160). Table 1 provides a summary of some characteristics of the included trials. A list of the included trials and some of their baseline characteristics is presented in Additional file 1: table S4. No trials reported individual patient data. Table 1 Some characteristics of the included trials Geographical region, n, (%) Asia 82 (45) Europe 41 (23) North America 27 (15) Australia / New Zealand 9 (5) South America 6 (4) > 1 region/ other 14 (7) Multicenter trials, n (%) 55 (30) Commercial funding, n (%) 85 (47) Balanced randomization, n (%) 151 (84) Home exercise prescribed, n (%) 20 (11) Rescue medication allowed, n (%) 26 (14) Mean (range) age 60 (30 to 75) Mean (range) body mass index 28 (22 to 35) Mean (range) percentage male 30 (0 to 95) Table 2. Multivariate analysis Covariate Coefficient (SE) 2-sided p-value R 2 analogue Intercept 0.53 (0.18) Sponsorship (commercial vs non-commercial) 0.14 (0.08) < 0.0001 0.17 Mean pain level at baseline -0.11 (0.31) 0.03 0.09 Mean effect size in treatment arm 0.15 (0.03) < 0.0001 0.05 Region: N-America vs other -0.34 (0.24) 0.08 0.02 Overall R 2 0.33 Legend table 2. Multivariate analysis of trial level covariates that contributed to explained variance. The following covariates did not contribute: trial size, year of publication, duration of treatment, length of follow-up after last treatment, timing of follow-up, route of administration, mean age, sex distribution, mean body mass index. df = degrees of freedom. A modified version of Cochrane Risk of Bias tool (RoB-2) was used with the addition of trial size (> 49 participants overall) and loss to follow-up (> 15%). Selective reporting bias was based on compliance with requirements for registration of the trial in a public register; high risk: registered trials that submitted protocol after trial start; some concerns: no information about registration; low risk: registered trials that submitted protocol before start of trial. Risk of bias Of the 179 trials, 168 (93%) were judged to be at unclear or high risk of bias on at least one domain and 146 (81%) were at high risk on at least one domain; 132 (73%) trials adequately generated their randomization sequence, 112 (62%) appropriately concealed allocation, 84 (47%) adequately blinded patients and caregivers, and 88 (49%) adequately blinded outcome assessors (because only patient reported outcome measures were extracted the participant was in all trials designated as the assessor). Only nine (5%) trials assessed the success of participant blinding. In 53 trials (29%), the total sample size was < 50, while 53 trials (29%) reported ≥ 15% missing outcome data. High risk or some concerns of bias due to selective reporting was found in 68 (38%) and 54 (30%) trials, respectively. A summary of risk of bias in each domain is shown in Fig. 2 , while assessments of risk of bias in each trial are provided in Additional file 1: fig. S1 . Clinical outcomes Effect sizes Analysis of effect sizes (Hedges’ g) in clinical outcomes in the placebo arms showed substantial heterogeneity, with an overall I 2 of 88% (p < 0.0001), i.e., 88% of the variance in observed effects reflected variance in true effects rather than sampling error. Effect sizes ranged from − 0.94 to 3.80 (interquartile range, IQR, 0.33 to 0.97), while the pooled effect size was 0.64 (95% CI 0.56 to 0.72). The prediction interval was − 0.31 to 1.60, i.e., the true effect size in 95% of all comparable populations would fall in this interval. Moderate or large effect sizes (Hedges g ≥ 0.5) were described in 109 (60%) trials. Fifteen (8%) trials described worsened pain compared to baseline. Pooled effect size in the no-treatment arms of four trials was 0.14 (95% CI = -0.03 to 0.31). Regression analysis In univariate meta-regression analysis, six of 16 predefined covariates were significantly (p ≤ 0.05) associated with effect size (Additional file 1: table S5). In multivariate analysis twelve covariates had p-values of ≥ 0.10 and were removed from the model. In the final model four covariates remained: effect size in the active treatment arms, geographic region, pain level at baseline, and source of funding, together explaining 33% of the variance (Additional file 1: table S2, figs. S2a-d) Subgroup and sensitivity analyses Subgroup analysis failed to reveal significant differences in pooled effect size according to the route of delivery (p = 0.65) (Fig. 3 ). Forest plots (Additional file 1: figs. S3a-e) show effect sizes per trial in each of the five subgroups. I-squared values remained high in all subgroups. Subgroup analysis revealed significant differences between geographic regions (p = 0.006); trials performed in North America (all but two from the USA) had the highest pooled effect size and those in Australia the lowest (Hedges’ g 0.89 and 0.50, respectively) (Additional file 1: fig. S2a). Pooled effect size did not differ significantly based on risk of bias in any of the seven domains. After removing trials with high risk of bias in at least one domain, overall pooled effect size remained virtually unchanged (0.65, 95% CI 0.52 to 0.78) (Additional file 1: tables S6a and S6b). Correlation analysis A moderately strong correlation was found between effect size in the placebo and treatment groups, with Pearson's r of 0.40 (P < 0.001) (Additional file 1: Fig. 2 d). Discussion In this meta-analysis of placebo responses in randomized controlled trials evaluating non-surgical interventions for KOA, improvements in pain were reported in more than 90% of the trials, with effect sizes ranging from moderate to large in six out of ten trials. The substantial between-trial variance and wide prediction interval indicates uncertainty; therefore, pooled effect size estimates were intentionally not emphasized and examination of sources of heterogeneity were prioritized [ 21 , 22 ]. Based on findings in previous reviews as well as clinical judgement, sixteen clinically relevant trial-level covariates were incorporated into a univariate metaregression model, six of which were significantly associated with effect size. After stepwise backward multivariate regression analysis four covariates remained statistically significant, together explaining 33% of the variance. Strengths and limitations Strengths of this meta-analysis include the use of a rigorous and comprehensive search strategy to update previous reviews; clear and transparent eligibility criteria; calculations based on each trial’s SD of change; incorporating both uni- and multivariate regression models to investigate the role of potential covariates; and a standardized critical appraisal of studies for quality. Extraction of outcome data (means, SDs, participant numbers, registration and funding) was duplicated, thus reducing the risk of errors or bias. Finally, the calculation of effect sizes allowed for combining studies using different outcome instruments, thus expanding the range of eligible trials. A major limitation of this meta-analysis is the inability, due to a lack of information in the included trials, to evaluate the role of psychological factors that are known to influence placebo effects. While the personality traits of participants (such as optimism or suggestibility) are unlikely to have differed much across the included trials, it is reasonable to assume that differences in context-related elements, including the patient-provider relationship, contributed substantially to the observed differences in placebo responses between trials. A description of the core elements involved in such relationships is outside the scope of this review, but competence and warmth conveyed by the health-care provider have been suggested as two key dimensions [ 23 ]. In turn, such elements may contribute to positive expectations, which are thought to play a role in shaping placebo effects and can lead to clinically relevant and enduring improvements, even on physical outcomes [ 24 ]. For example, in a double-blind sham-controlled surgical trial on patients with advanced Parkinson’s disease, perceived treatment was more strongly related to outcome than was the actual treatment received during the 12-month follow-up [ 25 ]. Incorporating treatment outcome expectations as a covariate may improve assay sensitivity in future placebo-controlled pain trials [ 26 , 27 ]. An additional limitation is that the analysis relies on trial-level aggregate data rather than individual participant data (IPD). While IPD could offer greater sensitivity and might reveal clinically significant associations [ 38 ], only a small number of published OA trials currently report such data [54]. Based on trials with a no-treatment arm, previous reviews have recognised non-specific factors as important contributors to placebo responses [ 3 ]. In our analysis, pooled effect size in the four trials with a no-treatment group was small (0.14); however, the wide 95% confidence intervals limits certainty regarding the role of non-specific effects. Symptoms associated with KOA are known to naturally fluctuate; duration of symptoms prior to enrolment is therefore a potential confounder by affecting the likelihood of spontaneous improvement [ 28 ]. In a prospective cohort of patients with KOA, regression to the mean was found to be responsible for as much as one point on a 0–10-point pain scale [ 29 ]. Although we were prevented from including disease duration as a covariate in our analysis because of insufficient or unclear data, it is possible that between-trial differences in disease duration could have contributed to explained variance. Sources of explained variance In multivariate analysis type of sponsorship contributed 17% to explained variance, with commercially funded trials reporting higher effect sizes. To our knowledge only two meta-analyses have specifically examined the influence of funding on placebo responses, with differing conclusions [ 30 , 31 ]. It has been suggested that commercially sponsored trials more frequently test novel agents or interventions, leading to greater expectations of efficacy among participants and outcome assessors [ 30 , 32 ]. In our analysis pain level at baseline was identified as a predictor of improved outcomes, consistent with findings from other reviews, including in the musculoskeletal domain [ 13 , 33 , 34 ]. This may suggest that individuals presenting with more severe knee symptoms and underlying pathology derive greater benefit from treatment. Alternatively, it could reflect statistical regression to the mean, whereby extreme baseline scores tend to approach the mean at subsequent follow-up assessments [ 35 ]. Effect size in the active treatment arms accounted for five percent of explained variance, and there was a correlation of 0.40 between effect sizes in the active treatment and placebo arms. These findings are not unexpected and are likely due to underlying contextual and non-specific factors common to both groups. Ideally, with an equal number of visits, appropriate blinding, randomization etc., such factors would be expected to be comparable across groups. Substantially stronger associations have been reported in other reviews [ 36 , 37 ] and in our previous review of placebo responses in trials of minimally invasive interventions we estimated an R 2 of 0.78 for primary endpoints [ 38 ]. Trials permitting the use of rescue medication demonstrated significantly improved outcomes in univariate, but not in multivariate, analysis. Trials prescribing home exercise as an adjunct to placebo reported slightly higher effect sizes, though not significant. Effect sizes exhibited significant variation across geographic regions, with studies conducted in North America reporting the highest and those from Asia and Australia the lowest effect sizes both in uni- and multivariate analysis. A prior systematic review of OA trials reported lower placebo responses in developed countries compared to developing countries; however, the specific criteria employed in their analysis were not described [ 5 ]. To our knowledge, no other reviews have specifically examined the influence of geographic region on placebo responses in OA trials. Comparison with other trials and reviews The observation that effect sizes in placebo arms are typically in the moderate to high range aligns with findings from several systematic reviews, demonstrating that the placebo effect constitutes a substantial proportion of the overall efficacy observed with active treatments [ 5 , 39 , 40 ]. Due to substantial statistical heterogeneity concerns may be raised regarding the validity of treatment effects, such as in research examining the efficacy of NSAIDs. However, variability observed in placebo groups is to be anticipated because of differences in demographic factors (e.g., age, sex distribution), study design elements (e.g., outcome measures, follow-up duration), and clinical characteristics (e.g., participant expectancy, disease duration). For example, even patients with similar radiographic features of knee OA have been shown to be extremely diverse in their pain pathophysiology, with the relative contribution of nociceptive and neuropathic mechanisms varying across studies [ 41 ]. Later publication year and length of treatment or follow-up period have been linked to increased placebo responses in other reviews [ 5 , 34 , 42 ]. Though such associations were not replicated in the present analysis it should be noted that only the outcome value at one time point per trial was extracted; therefore within-trial longitudinal changes over time cannot be excluded. Interestingly, previous analyses have described persisting effects on pain and function after both sham surgery [ 43 – 46 ] and intraarticular injections [ 8 ], and patients with advanced Parkinson’s disease that underwent sham surgery described persistent improvements in physical function 12 months after the procedure [ 25 ]. Route of placebo delivery In our analysis, the method of placebo administration was not significantly associated with outcome, contradicting a common assertion that more invasive, complex, or time-intensive interventions tend to elicit stronger placebo effects [ 5 , 6 , 41 , 47 , 48 ]. Like the current analysis, these reviews were based on non-randomised comparisons. The inability to control for confounding variables in this design introduces potential bias, thereby making the concept of "enhanced placebos" uncertain. The most reliable approach for evaluating the comparative effectiveness of different placebo interventions involves directly comparing delivery methods within a single clinical trial. This requires a four-arm design, incorporating two distinct active treatments alongside their respective placebos. In a meta-analysis based on twelve studies conforming to this trial design the authors found no consistent differences in effect size between placebo trials with varying “intensiveness” [ 49 ]. Two of the RCTs included in that analysis comprised patients with musculoskeletal disorders. One of them compared a placebo pill with sham acupuncture for arm pain [ 50 ]. The sham group reported significantly greater improvements in pain, but no changes in arm function or symptom severity. In a trial that included patients with acute musculoskeletal pain secondary to trauma, there were no differences in pain response between parenteral and oral placebos measured up to two hours after administration [ 51 ]. A more recent RCT of ketoprofen for KOA compared topical and oral placebos; at week 12 improvements were significantly greater in the topical group [ 52 ]. Given the limited number of published RCTs that have incorporated more than one type of placebo within the same study, we contend that the concept of "enhanced placebo" lacks sufficient empirical support, and further research employing this type of design is necessary to substantiate this hypothesis. Conclusion More than half of the included trials described moderate to large effect sizes in their placebo arms, although significant statistical heterogeneity was noted across studies. Multivariate analysis revealed that about a third of the observed variance could be explained by four covariates: type of sponsorship, baseline pain level, effect size in the active treatment arm and geographic region. The factors underlying the remaining unexplained variance are undetermined, though psychological variables—such as patient expectations and provider interactions—are likely to have contributed. Due to the lack of data within the trials analysed, the potential impact of such factors could not be evaluated. Incorporating psychological assessments in future OA trials could improve our understanding of their role in placebo responses. Declarations Ethics approval and consent to participate Not applicable Consent for publication Not applicable Availability of data and materials The datasets used and analysed in this current study are available from the corresponding author upon reasonable request. Competing interests The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Funding No funding was involved in this review Not applicable Acknowledgements Not applicable CRediT authorship contribution statement Robin Holtedahl : Conceptualization, Methodology, Formal analysis, Investigation, Data curation, Writing – original draft, Writing – review & editing, Visualization. Alexandra Christina Hott : Conceptual­ isation, Methodology, Data curation, Writing – review & editing. Jens Ivar Brox : Methodology, Data curation, Writing – review & editing. Inger Johanne Hansen : Data curation. Johan Bjørneboe : Data curation, Writing – review & editing. All authors read and approved the final manuscript. Artificial Intelligence Artificial Intelligence was used to improve grammar. Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research. Patient consent for publication Not applicable. References Long H, Liu Q, Yin H, Wang K, Diao N, Zhang Y, et al. Arthritis Rheumatol. 2022;74:1172–83. https://doi.org/10.1002/art.42089 . Prevalence Trends of Site-Specific Osteoarthritis From 1990 to 2019: Findings From the Global Burden of Disease Study 2019. Gibbs AJ, Gray B, Wallis JA, Taylor NF, Kemp JL, Hunter DJ, et al. Recommendations for the management of hip and knee osteoarthritis: A systematic review of clinical practice guidelines. Osteoarthritis Cartilage. 2023;31:1280–92. https://doi.org/10.1016/j.joca.2023.05.015 . Krogsbøll LT, Hróbjartsson A, Gøtzsche PC. Spontaneous improvement in randomised clinical trials: meta-analysis of three-armed trials comparing no treatment, placebo and active intervention. BMC Med Res Methodol. 2009;9:1. https://doi.org/10.1186/1471-2288-9-1 . Benedetti F. Placebo and the new physiology of the doctor-patient relationship. Physiol Rev. 2013;93:1207–46. https://doi.org/10.1152/physrev.00043.2012 . Zou K, Wong J, Abdullah N, Chen X, Smith T, Doherty M, et al. Examination of overall treatment effect and the proportion attributable to contextual effect in osteoarthritis: Meta-Analysis of randomised controlled trials. Ann Rheum Dis. 2016;75:1964–70. https://doi.org/10.1136/annrheumdis-2015-208387 . Zhang W, Robertson J, Jones AC, Dieppe PA, Doherty M. The placebo effect and its determinants in osteoarthritis: Meta-analysis of randomised controlled trials. Ann Rheum Dis. 2008;67:1716–23. https://doi.org/10.1136/ard.2008.092015 . Undurraga J, Baldessarini RJ, Randomized. Placebo-Controlled Trials of Antidepressants for Acute Major Depression: Thirty-Year Meta-Analytic Review. Neuropsychopharmacology. 2012;37:851–64. https://doi.org/10.1038/npp.2011.306 . Previtali D, Merli G, Di Laura Frattura G, Candrian C, Zaffagnini S, Filardo G. The Long-Lasting Effects of Placebo Injections in Knee Osteoarthritis: A Meta-Analysis. Cartilage. 2021;13:S185–96. https://doi.org/10.1177/1947603520906597 . Wartolowska KA, Gerry S, Feakins BG, Collins GS, Cook J, Judge A, et al. A meta-analysis of temporal changes of response in the placebo arm of surgical randomized controlled trials: An update. Trials. 2017;18:12–9. https://doi.org/10.1186/s13063-017-2070-9 . Hróbjartsson A, Gøtzsche PC. Placebo interventions for all clinical conditions. Cochrane Database Syst Reviews 2010;2010. https://doi.org/10.1002/14651858.CD003974.pub3 Tuttle AH, Tohyama S, Ramsay T, Kimmelman J, Schweinhardt P, Bennett GJ, et al. Increasing placebo responses over time in U.S. clinical trials of neuropathic pain. Pain. 2015;156:2616–26. https://doi.org/10.1097/j.pain.0000000000000333 . Walach H, Maidhof C. Is the placebo effect dependent on time? A meta-analysis. How expectancies shape experience. American Psychological Association; 2004. pp. 321–32. https://doi.org/10.1037/10332-013 . Häuser W, Bartram-Wunn E, Bartram C, Reinecke H, Tölle T. Systematic review: Placebo response in drug trials of fibromyalgia syndrome and painful peripheral diabetic neuropathy - Magnitude and patient-related predictors. Pain. 2011;152:1709–17. https://doi.org/10.1016/j.pain.2011.01.050 . Hróbjartsson A, Gøtzsche PC. Is the placebo powerless? Update of a systematic review with 52 new randomized trials comparing placebo with no treatment. J Intern Med. 2004;256:19–100. https://doi.org/10.1111/j.1365-2796.2004.01355.x . Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD et al. The PRISMA 2020 statement: An updated guideline for reporting systematic reviews. BMJ 2021;372. https://doi.org/10.1136/bmj.n71 Sterne JA, Hernán MA, Reeves BC, Savović J, Berkman ND, Viswanathan M, et al. ROBINS-I: A tool for assessing risk of bias in non-randomised studies of interventions. BMJ (Online). 2016;355:1–7. https://doi.org/10.1136/bmj.i4919 . Dankel SJ, Loenneke JP. Effect sizes for paired data should use the change score variability rather than the pretest variability. J Strength Cond Res. 2021;35:1773–8. https://doi.org/10.1519/JSC.0000000000002946 . Sterne JAC, Savović J, Page MJ, Elbers RG, Blencowe NS, Boutron I, Cates CJ, Cheng H-Y, Corbett MS, Eldridge SM, Hernán MA, Hopewell S, Hróbjartsson A, Junqueira DR, Jüni P, Kirkham JJ, Lasserson T, Li T, McAleenan A, Reeves BC, Shepperd S, Shrier I, Stew HJ. RoB 2: A revised Cochrane risk-of-bias tool for randomized trials. BMJ 2019:1–24. Higgins JPT, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. Br Med J. 2003;327:557–60. Inthout J, Ioannidis JPA, Rovers MM, Goeman JJ. Plea for routinely presenting prediction intervals in meta-analysis. BMJ Open. 2016;6:10247. https://doi.org/10.1136/bmjopen-2015 . Nordmann AJ, Kasenda B, Briel M. Meta-analyses: What they can and cannot do. Swiss Med Wkly. 2012;142. https://doi.org/10.4414/smw.2012.13518 . Cote MP, Lubowitz JH, Rossi MJ, Brand JC. Reviews Pooling Heterogeneous, Low-Evidence, High-Bias Data Result in Incorrect Conclusions: But Heterogeneity is an Opportunity to Explore. Arthrosc - J Arthroscopic Relat Surg. 2018;34:3126–8. https://doi.org/10.1016/j.arthro.2018.10.005 . Howe LC, Leibowitz KA, Crum AJ. When your doctor Gets it and Gets you: The critical role of competence and warmth in the patient-provider interaction. Front Psychiatry. 2019;10. https://doi.org/10.3389/fpsyt.2019.00475 . Rosenkjær S, Lunde SJ, Kirsch I, Vase L, Expectations. How and when do they contribute to placebo analgesia? Front Psychiatry 2022;13. https://doi.org/10.3389/fpsyt.2022.817179 Mcrae C, Cherin E, Yamazaki M, TG, Diem G, Vo AH, Russell D et al. Effects of Perceived Treatment on Quality of Life and Medical Outcomes in a Double-blind Placebo Surgery Trial. vol. 61. 2004. Frisaldi E, Shaibani A, Benedetti F. Why We should Assess Patients’ Expectations in Clinical Trials. Pain Ther. 2017;6:107–10. https://doi.org/10.1007/s40122-017-0071-8 . Langford DJ, Lou R, Sheen S, Amtmann D, Colloca L, Edwards RR, et al. Expectations for Improvement: A Neglected but Potentially Important Covariate or Moderator for Chronic Pain Clinical Trials. J Pain. 2023;24:575–81. https://doi.org/10.1016/j.jpain.2022.12.011 . Glasziou P, Chalmers L, Rawlins M, McCulloch P. When are randomised trials unnecessary? Picking signal from noise Paul Glasziou and colleagues. Br Med J. 2007;334:349–51. Englund M, Turkiewicz A. Pain in clinical trials for knee osteoarthritis: estimation of regression to the mean. Lancet Rheumatol. 2023;5:e309–11. https://doi.org/10.1016/S2665-9913(23)00090-5 . Jones BDM, Razza LB, Weissman CR, Karbi J, Vine T, Mulsant LS, et al. Magnitude of the Placebo Response across Treatment Modalities Used for Treatment-Resistant Depression in Adults: A Systematic Review and Meta-analysis. JAMA Netw Open. 2021;4. https://doi.org/10.1001/jamanetworkopen.2021.25531 . Ioannidis K, Huneke NTM, Solly JE, Fusetto Veronesi G, Tzagarakis C, Parlatini V, et al. Placebo and nocebo effects in gambling disorder pharmacological trials: A meta-analysis. Acta Neuropsychiatr. 2024. https://doi.org/10.1017/neu.2024.52 . Dunlop BW, Thase ME, Wun C-C, Fayyad R, Guico-Pabia CJ, Musgnung J, et al. A meta-analysis of factors impacting detection of antidepressant efficacy in clinical trials: the importance of academic sites. Neuropsychopharmacology. 2012;37:2830–6. https://doi.org/10.1038/npp.2012.153 . Vase L, Vollert J, Finnerup NB, Miao X, Atkinson G, Marshall S, et al. Predictors of the placebo analgesia response in randomized controlled trials of chronic pain: A meta-analysis of the individual data from nine industrially sponsored trials. Pain. 2015;156:1795–802. https://doi.org/10.1097/j.pain.0000000000000217 . Wen X, Luo J, Mai Y, Li Y, Cao Y, Li Z, et al. Placebo Response to Oral Administration in Osteoarthritis Clinical Trials and Its Associated Factors: A Model-Based Meta-analysis. JAMA Netw Open. 2022;E2235060. https://doi.org/10.1001/jamanetworkopen.2022.35060 . Vickers AJ, Altman DG. Analysing controlled trials with baseline and follow up measurements. Br Med J. 2001;323:1123–4. https://doi.org/10.1136/bmj.323.7321.1123 . Kirsch I, Sapirstein G. Listening to Prozac but hearing placebo: A meta-analysis of antidepressant medication. Prev Treat. 1998;1:1–16. https://doi.org/10.1037/1522-3736.1.1.12a . Walach H, Sadaghiani C, Dehm C, Bierman D. The therapeutic effect of clinical trials: understanding placebo response rates in clinical trials-A secondary analysis 2005. https://doi.org/10.1186/1471-2288-5 Holtedahl R, Brox JI, Tjomsland O. Placebo effects in trials evaluating 12 selected minimally invasive interventions: a systematic review and meta-analysis. BMJ Open. 2015;5:e007331. https://doi.org/10.1136/bmjopen-2014-007331 . Huang Z, Chen J, Hu QS, Huang Q, Ma J, Pei FX, et al. Meta-analysis of pain and function placebo responses in pharmacological osteoarthritis trials. Arthritis Res Ther. 2019;21. https://doi.org/10.1186/s13075-019-1951-6 . Chen AT, Shrestha S, Collins JE, Sullivan JK, Losina E, Katz JN. Estimating contextual effect in nonpharmacological therapies for pain in knee osteoarthritis: a systematic analytic review. Osteoarthritis Cartilage. 2020;28:1154–69. https://doi.org/10.1016/j.joca.2020.05.007 . Mandl LA, Losina E. Relative Efficacy of Knee Osteoarthritis Treatments: Are All Placebos Created Equal? Ann Intern Med. 2015;162:71–2. https://doi.org/10.7326/M14-2636 . Reiter-Niesert S, Boers M, Detert J. Short-term placebo response in trials of patients with symptomatic osteoarthritis: Differences between hip and knee. Osteoarthritis Cartilage. 2016;24:1007–11. https://doi.org/10.1016/j.joca.2016.01.002 . Sihvonen R, Paavola M, Malmivaara A, Itälä A, Joukainen A, Kalske J, et al. Arthroscopic partial meniscectomy for a degenerative meniscus tear: a 5 year follow-up of the placebo-surgery controlled FIDELITY (Finnish Degenerative Meniscus Lesion Study) trial. Br J Sports Med. 2020;54:1332–9. https://doi.org/10.1136/bjsports-2020-102813 . Schrøder CP, Skare Ø, Reikerås O, Mowinckel P, Brox JI. Sham surgery versus labral repair or biceps tenodesis for type II SLAP lesions of the shoulder: A three-armed randomised clinical trial. Br J Sports Med. 2017;51:1759–66. https://doi.org/10.1136/bjsports-2016-097098 . Moseley JB, O’Malley K, Peterson NJ, Menke TJ, Brody BA, Kuykendall DH, et al. A controlled trial of arthroscopic surgery for osteoarthritis of the knee. N Engl J Med. 2002;347:81–2. Beard DJ, Rees JL, Cook JA, Rombach I, Cooper C, Merritt N, et al. Arthroscopic subacromial decompression for subacromial shoulder pain (CSAW): a multicentre, pragmatic, parallel group, placebo-controlled, three-group, randomised surgical trial. Lancet. 2018;391:329–38. https://doi.org/10.1016/S0140-6736(17)32457-1 . Abhishek A, Doherty M. Mechanisms of the placebo response in pain in osteoarthritis. Osteoarthritis Cartilage. 2013;21:1229–35. https://doi.org/10.1016/j.joca.2013.04.018 . Bannuru RR, McAlindon TE, Sullivan MC, Wong JB, Kent DM, Schmid CH. Effectiveness and implications of alternative placebo treatments: A systematic review and network meta-analysis of osteoarthritis trials. Ann Intern Med. 2015;163:365–92. https://doi.org/10.7326/M15-0623 . Fässler M, Meissner K, Kleijnen J, Hróbjartsson A, Linde K. A systematic review found no consistent difference in effect between more and less intensive placebo interventions. J Clin Epidemiol. 2015;68:442–51. https://doi.org/10.1016/j.jclinepi.2014.11.018 . Kaptchuk TJ, Stason WB, Davis RB, Legedza ATR, Schnyer RN, Kerr CE, et al. Sham device versus inert pill: Randomised controlled trial of two placebo treatments. Br Med J. 2006;332:391–4. https://doi.org/10.1136/bmj.38726.603310.55 . Schwartz NA, Turturro MA, Istvan DJ, Larkin GL. Patients’ Perceptions of Route of Nonsteroidal Anti-inflammatory Drug Administration and Its Effect on Analgesia. vol. 7. 2000. Conaghan PG, Dickson J, Bolten W, Cevc G, Rother M. A multicentre, randomized, placebo- and active-controlled trial comparing the efficacy and safety of topical ketoprofen in Transfersome gel (IDEA-033) with ketoprofen-free vehicle (TDT 064) and oral celecoxib for knee pain associated with osteoarthritis. Rheumatol (United Kingdom). 2013;52:1303–12. https://doi.org/10.1093/rheumatology/ket133 . Additional Declarations No competing interests reported. Supplementary Files Binder6.pdf Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7888576","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":545780734,"identity":"b8bcbc63-3d5a-4997-b524-f6a82b94baea","order_by":0,"name":"Robin Holtedahl","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Robin","middleName":"","lastName":"Holtedahl","suffix":""},{"id":545780735,"identity":"bf0d58ba-91b9-4e1d-b99a-e6df1bbb56d4","order_by":1,"name":"Alexandra Christina Hott","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA7UlEQVRIiWNgGAWjYDACdjDJDMQ8IIYNA4ME4wP8WpiRtRxISANqYTYgScthwlr4m5mPSfzcYZ3P38B78PHHH+ft5Wc3MzAXVODWInGYLdmw90y65YwDfMkGBxJuJ264c5iBecYZPNYc5jF8wNt22IDhAI+ZBFBLgoFE/gFm3jbcOuQP8384+BeoRR6i5Zy9/IxkBrxaDA7zMD4G2WIA0XKAseEGAS2Gh9mMjWXb0g0MD/MYG5xJS07cANRymAePX+SONz+TfNtmbSB3vMfwQYWNHchhjI958IQYAjAjsQ8Qo2EUjIJRMApGAW4AAMy2TMZUpZ15AAAAAElFTkSuQmCC","orcid":"","institution":"Sørlandet Hospital Kristiansand","correspondingAuthor":true,"prefix":"","firstName":"Alexandra","middleName":"Christina","lastName":"Hott","suffix":""},{"id":545780736,"identity":"28b1c262-3241-4837-a56b-bd9e48ad0c6b","order_by":2,"name":"Inger Johanne Hansen","email":"","orcid":"","institution":"Sørlandet Hospital Kristiansand","correspondingAuthor":false,"prefix":"","firstName":"Inger","middleName":"Johanne","lastName":"Hansen","suffix":""},{"id":545780737,"identity":"75d1daa1-f0c0-44eb-bc75-829f1dde8876","order_by":3,"name":"John Bjørneboe","email":"","orcid":"","institution":"Oslo University Hospital","correspondingAuthor":false,"prefix":"","firstName":"John","middleName":"","lastName":"Bjørneboe","suffix":""},{"id":545780738,"identity":"6f805c95-1321-40a3-83a4-635d69862432","order_by":4,"name":"Jens Ivar Brox","email":"","orcid":"","institution":"Oslo University Hospital","correspondingAuthor":false,"prefix":"","firstName":"Jens","middleName":"Ivar","lastName":"Brox","suffix":""}],"badges":[],"createdAt":"2025-10-17 16:38:30","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7888576/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7888576/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":96205698,"identity":"ebcfd9e6-4474-4e5b-bdab-491ed5f41653","added_by":"auto","created_at":"2025-11-18 17:12:11","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1074622,"visible":true,"origin":"","legend":"","description":"","filename":"Advancesinrheum.docx","url":"https://assets-eu.researchsquare.com/files/rs-7888576/v1/dcd1d041a467f406bbb82558.docx"},{"id":96205700,"identity":"1a650096-eaa2-4f6c-8c8a-86e77e62f3e7","added_by":"auto","created_at":"2025-11-18 17:12:11","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":7316,"visible":true,"origin":"","legend":"","description":"","filename":"a4fed49abf864a39ad0f12f1d60873f5.json","url":"https://assets-eu.researchsquare.com/files/rs-7888576/v1/71f094b305f4a4f12715d089.json"},{"id":96205702,"identity":"177f5c1c-868e-416c-aa0b-450cff7f27a2","added_by":"auto","created_at":"2025-11-18 17:12:12","extension":"pdf","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2251957,"visible":true,"origin":"","legend":"","description":"","filename":"Binder6.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7888576/v1/5c5a29aa08bb85a79218fdfc.pdf"},{"id":96205703,"identity":"8e741d7e-316f-4022-a350-5fb0de17c234","added_by":"auto","created_at":"2025-11-18 17:12:12","extension":"xml","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":135709,"visible":true,"origin":"","legend":"","description":"","filename":"a4fed49abf864a39ad0f12f1d60873f51enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7888576/v1/f2d36984975eb8b3c5ee6ed6.xml"},{"id":96205706,"identity":"7a666b41-63c3-4cdd-8a40-35c9e4135fd3","added_by":"auto","created_at":"2025-11-18 17:12:12","extension":"emf","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":531884,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.emf","url":"https://assets-eu.researchsquare.com/files/rs-7888576/v1/709a822a3b7b22db6dd18e42.emf"},{"id":96205710,"identity":"796d565f-b9cb-4bf7-97c2-779126ff9e22","added_by":"auto","created_at":"2025-11-18 17:12:12","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":28356,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7888576/v1/d3984474165662d82196a3a9.png"},{"id":96205707,"identity":"a2eb45ce-792d-479b-bac9-b175ea81407e","added_by":"auto","created_at":"2025-11-18 17:12:12","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":87583,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7888576/v1/59fa0a4dab44ad3e1647bd41.png"},{"id":96252260,"identity":"f3ec5141-aaaa-4dd8-a5c9-1fb0e6d4a26b","added_by":"auto","created_at":"2025-11-19 07:40:43","extension":"xml","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":133210,"visible":true,"origin":"","legend":"","description":"","filename":"a4fed49abf864a39ad0f12f1d60873f51structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7888576/v1/f36cfb856838a526c4a55fe4.xml"},{"id":96205711,"identity":"3dfee6b1-acab-409e-95df-b05a98183670","added_by":"auto","created_at":"2025-11-18 17:12:12","extension":"html","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":148716,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7888576/v1/d7f7bb8a258fc08dac5eb767.html"},{"id":96250544,"identity":"b499e390-eda1-465b-bd51-e0a944677883","added_by":"auto","created_at":"2025-11-19 07:38:36","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":11514,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePRISMA flow diagram showing the screening process and search results.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7888576/v1/8d1250793267f9b89906fca0.png"},{"id":96205705,"identity":"30688856-3698-4a3a-adc2-8dbc513b8899","added_by":"auto","created_at":"2025-11-18 17:12:12","extension":"jpeg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":135331,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSummary of risk of bias.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA modified version of Cochrane Risk of Bias tool (RoB-2) was used with the addition of trial size (\u0026gt;49 participants overall) and loss to follow-up (\u0026gt;15%). Selective reporting bias was based on compliance with requirements for registration of the trial in a public register; high risk: registered trials that submitted protocol after trial start; some concerns: no information about registration; low risk: registered trials that submitted protocol before start of trial.\u003c/p\u003e","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7888576/v1/83f3af83f9f1b3b6102088bc.jpeg"},{"id":96205709,"identity":"81485ef7-d556-4880-b00a-14575ebd1198","added_by":"auto","created_at":"2025-11-18 17:12:12","extension":"jpeg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":443126,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eEffects size (Hedges’g) by route of placebo delivery\u003c/strong\u003e.\u003c/p\u003e\n\u003cp\u003eSubgroup analysis, random effects model. 95% confidence intervals (black diamonds); pooled effect size (red diamond). SC, subcutaneous; IV, intravenous.\u003c/p\u003e","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7888576/v1/3c00c1b491ab3bb28cebd142.jpeg"},{"id":99788030,"identity":"790c698f-4bb6-44ac-9701-237a4dffc75e","added_by":"auto","created_at":"2026-01-08 12:43:33","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1621886,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7888576/v1/a5d7e6bd-6f01-4b77-8346-e6813589928a.pdf"},{"id":96252482,"identity":"9510b03c-bcfc-45c9-8122-cdfa94d56188","added_by":"auto","created_at":"2025-11-19 07:41:02","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":2251957,"visible":true,"origin":"","legend":"","description":"","filename":"Binder6.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7888576/v1/f77b765bfc326d50f8d32221.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Placebo responses in non-surgical trials for knee osteoarthritis: A systematic review and meta-analysis","fulltext":[{"header":"Background","content":"\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eThe knee is the most common site of osteoarthritis (OA), with rising global prevalence [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. OA is a degenerative joint disease, with gradual deterioration of articular cartilage, causing pain, stiffness, and reduced mobility. Non-surgical treatments focus on symptom relief and delaying progression, with treatment guidelines recommending education, exercise, and weight loss as initial care. Other treatments lack consistent evidence, whereas a patient-centred approach based on a biopsychosocial model is strongly advised [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e].\u003c/p\u003e\u003cp\u003ePlacebo or sham treatments are interventions that are presumed to have no specific therapeutic effects for the condition being treated. Placebo responses encompass all health-related changes resulting from administering a placebo and include placebo effects and as well as non-specific factors like regression to the mean and the natural course of the condition. The net benefit of a treatment is assessed by comparing the outcomes in the active treatment and placebo groups. To distinguish between the placebo effect and non-specific effects a no-treatment group is necessary [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eSeveral factors have been identified as influencing the placebo effect, including the specific characteristics of the intervention, the patient\u0026rsquo;s beliefs and expectations, the quality of the patient\u0026ndash;therapist relationship, and the level of empathy, confidence, and optimism communicated by the physician [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. In the context of clinical trials placebo effects have been linked to the route of administration (invasiveness or complexity of the procedure) [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e], trial duration [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e], length of follow-up [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e], sample size [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e], period of study [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e], and efficacy of the active treatment [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. Subjective outcomes, including pain, are thought to be associated with larger placebo effects than objective outcomes [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e].\u003c/p\u003e\u003cp\u003e The aims of this meta-analysis were 1) to analyse pain-related outcomes within the placebo arms of randomized controlled trials (RCTs) of non-surgical treatments for knee OA (KOA) published since 2007; 2) through meta-regression analysis to explore covariates potentially contributing to placebo responses, including those suggested in previous reviews, and 3) perform subgroup analyses based on relevant categorical covariates and risk of bias.\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eThe protocol of this review was registered at PROSPERO (CRD 420251113096) and the review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) reporting checklist [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e].\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003eData sources and eligibility criteria\u003c/h2\u003e\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eAfter registration of the protocol systematic searches were performed for eligible studies published between January 2008 and May 2025 in PubMED and EMBASE. The search terms were \u0026ldquo;knee osteoarthritis\u0026rdquo; AND (placebo OR sham) with the filters \u0026ldquo;randomized controlled trial\u0026rdquo; and English language.\u003c/p\u003e\u003cp\u003eTwo authors (RH and AH) assessed eligibility in two stages: 1) by screening each trial\u0026rsquo;s title and abstract and 2) by reviewing potentially eligible articles in full text. Each reviewer stated their reason for exclusion and consensus was reached in cases of disagreement. The inclusion criteria included 1) randomized, parallel-group trials of treatments for knee osteoarthritis that were double- or triple-blinded; 2) included a placebo (sham) arm (optionally also a no-treatment arm); 3) assessed the efficacy of any treatment except psychological or exercise-based interventions or anatomy-altering surgical procedures; 4) reported a pain-related knee-specific outcome measure on a continuous scale; and 5) enabled calculation of effect size, i.e.; provided standard deviation [SD], standard error [SE] or 95% confidence intervals [CI]. For more detailed eligibility criteria see Additional file 1: table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e).\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eData Extraction\u003c/h3\u003e\n\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eThe main characteristics of each trial were entered by the first author into a standardized data extraction form using Microsoft Access v.2412. For a complete list of items, see Additional file 1: table S2. Outcome values in the placebo, active treatment and no-treatment control groups as well as registration and source of funding were extracted by the first author and confirmed by two independent authors. The WOMAC pain subscale was chosen as the preferred outcome measure; if not provided, total WOMAC or other knee-specific pain outcome measures. As radiological OA severity and duration of disease were not consistently described such data were not extracted. Pre-to-post change values were checked for direction. For metaregression analysis, 16 covariates judged to be clinically relevant were extracted (see Additional file 1: table S3). In trials that described ongoing treatments the outcome value at the end of the trial was extracted. In trials that reported outcome at later follow-up, including single-intervention trials, outcome value at the latest follow-up was extracted, with a 2-year upper limit. Because outcome measures and scales varied across studies, baseline values were converted to a 10 cm visual analogue scale (VAS), with higher scores indicating worse pain.\u003c/p\u003e\u003cp\u003eThe method of administering placebo was categorized into five groups, representing varying levels of perceived invasiveness or complexity: 1) physical (e.g., short-wave ultrasound, cryotherapy, taping), 2) oral (tablet, capsule, sachet, syrup), 3) intra-articular (in most cases normal saline), 4) topical (gel, ointment, liniment) and 5) intravenous, subcutaneous or intradermal (including acupuncture). Because interventions in the last group share the characteristics of penetration of skin and / or soft tissue it was judged reasonable to group them together.\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\n\u003ch3\u003eRisk of bias assessment\u003c/h3\u003e\n\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eTwo reviewers independently assessed the risk of bias using the Revised Tool for Risk of Bias in Randomized Trials [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. The categories were low, some concerns and high. Reporting bias was evaluated by checking if the trials were registered and whether their protocols were submitted before the trials began. Trial size (\u0026gt;\u0026thinsp;49 patients) and loss-to-follow-up (\u0026gt;\u0026thinsp;15%) were added as additional criteria. Consensus was reached by discussion. Sensitivity analysis was performed by removing trials that were at high risk of bias on any domain.\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\n\u003ch3\u003eStatistical analyses\u003c/h3\u003e\n\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eCalculations were performed using Comprehensive Meta-Analysis version 4.0.000 software (available from \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e\u003ca href=\"http://www.meta-analysis.com\" target=\"_blank\"\u003ewww.meta-analysis.com\u003c/a\u003e\u003c/span\u003e\u003cspan address=\"http://www.meta-analysis.com\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), MedCalc (MedCalc\u0026reg; Statistical Software version 22.021 (MedCalc Software Ltd, Ostend, Belgium; \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.medcalc.org\u003c/span\u003e\u003cspan address=\"https://www.medcalc.org\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e; 2024) and Microsoft Excel version 2503 (Microsoft, Redmond, WA, USA). The Hedges\u0026rsquo; \u003cem\u003eg\u003c/em\u003e value, which corrects for small sample sizes, was used as a measure of the effect size of change from baseline to follow-up. By convention an effect size of 0.8 is described as large; 0.5\u0026ndash;0.8, moderate; and \u0026lt;\u0026thinsp;0.5, small (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e\u003ca href=\"http://www.meta-analysis.com\" target=\"_blank\"\u003ewww.cochrane.org/handbook\u003c/a\u003e\u003c/span\u003e\u003cspan address=\"http://www.cochrane.org/handbook\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). Calculations were preferably based on each trial\u0026rsquo;s SD of change (alternatively SE, 95% CI or p-value) [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]; if not provided, on a recommended formula, with the pretest-posttest correlation set to 0.5 [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. Similar calculations were performed for treatment and no-treatment arms. Two-sided p-values of \u0026lt;\u0026thinsp;0.05 were considered as statistically significant.\u003c/p\u003e\u003cp\u003eA random effects model using the DerSimonian and Laird method was used because the studies were sampled from multiple populations. The 95% CIs were quantified in a weighted fashion using the inverse variance approach to incorporate both within- and between-trial variance. I\u003csup\u003e2\u003c/sup\u003e statistics were calculated to assess the degree of heterogeneity across studies, i.e., the proportion of variability not due to sampling error. Heterogeneity was considered high when presenting a value greater than or equal to 50%, as suggested elsewhere [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. To assess whether future similar trials would be expected to provide comparable results, prediction intervals were also calculated [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eSubgroup analyses were performed to compare effect sizes across geographical regions and across different routes of placebo delivery. In the univariate meta-regression analysis sixteen predefined trial-level covariates were entered into the model. In the next stage, a multivariate regression model was built by starting with the complete set of covariates and stepwise eliminating covariates with p-values\u0026thinsp;\u0026ge;\u0026thinsp;0.10. R\u003csup\u003e2\u003c/sup\u003e values of each of the remaining covariates as well as overall R\u003csup\u003e2\u003c/sup\u003e were estimated.\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\n \u003ch2\u003eStudy selection\u003c/h2\u003e\n \u003cdiv class=\"BlockQuote\"\u003e\n \u003cp\u003eAfter removal of duplicates searches identified a total of 549 trials. Of these, 295 trials were excluded after first screening and 75 after second screening, leaving 179 trials available for analysis. A flowchart of the selection process is shown in Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e, including reasons for exclusion.\u003c/p\u003e\n \u003c/div\u003e\n\u003c/div\u003e\n\u003ch3\u003eTrial characteristics\u003c/h3\u003e\n\u003cdiv class=\"BlockQuote\"\u003e\n \u003cp\u003eThe placebo arms comprised in total 13 031 patients at inclusion. The average loss to follow-up was 13%. Median sample size was 66 (interquartile range, IQR 46 to 150). Four trials with 141 participants in total also included a no-treatment arm. The average follow-up was 20 weeks (range 10 to 26) for single-procedure trials (n\u0026thinsp;=\u0026thinsp;20) and 20 weeks (range 4 to104) for ongoing-treatment trials (n\u0026thinsp;=\u0026thinsp;160). Table \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e provides a summary of some characteristics of the included trials. A list of the included trials and some of their baseline characteristics is presented in Additional file 1: table S4. No trials reported individual patient data.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv class=\"gridtable\"\u003e\n \u003cdiv align=\"left\" class=\"colspec\"\u003e\u003cbr\u003e\u003c/div\u003e\n \u003ctable id=\"Tab1\" border=\"1\" class=\"fr-table-selection-hover\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eSome characteristics of the included trials\u003c/span\u003e\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\" style=\"width: 20.7128%;\"\u003e\n \u003cp\u003eGeographical region, n, (%)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" style=\"width: 15.4188%;\"\u003e\n \u003cp\u003eAsia\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" style=\"width: 2.8952%;\"\u003e\n \u003cp\u003e82 (45)\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 20.7128%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 15.4188%;\"\u003e\n \u003cp\u003eEurope\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 4.4999%;\"\u003e\n \u003cp\u003e41 (23)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 20.7128%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 15.4188%;\"\u003e\n \u003cp\u003eNorth America\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 4.4999%;\"\u003e\n \u003cp\u003e27 (15)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 20.7128%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 15.4188%;\"\u003e\n \u003cp\u003eAustralia / New Zealand\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 4.4999%;\"\u003e\n \u003cp\u003e9 (5)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 20.7128%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 15.4188%;\"\u003e\n \u003cp\u003eSouth America\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 4.4999%;\"\u003e\n \u003cp\u003e6 (4)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 20.7128%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 15.4188%;\"\u003e\n \u003cp\u003e\u0026gt;\u0026thinsp;1 region/ other\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 4.4999%;\"\u003e\n \u003cp\u003e14 (7)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 28.1244%;\"\u003e\n \u003cp\u003eMulticenter trials, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 8.0072%;\"\u003e\n \u003cp\u003e55 (30)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 8.8675%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 28.1244%;\"\u003e\n \u003cp\u003eCommercial funding, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 8.0072%;\"\u003e\n \u003cp\u003e85 (47)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 8.8675%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 28.1244%;\"\u003e\n \u003cp\u003eBalanced randomization, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 8.0072%;\"\u003e\n \u003cp\u003e151 (84)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 8.8675%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 28.1244%;\"\u003e\n \u003cp\u003eHome exercise prescribed, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 8.0072%;\"\u003e\n \u003cp\u003e20 (11)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 8.8675%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 28.1244%;\"\u003e\n \u003cp\u003eRescue medication allowed, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 8.0072%;\"\u003e\n \u003cp\u003e26 (14)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 8.8675%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 20.7128%;\"\u003e\n \u003cp\u003eMean (range) age\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 15.4188%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 4.4999%;\"\u003e\n \u003cp\u003e60 (30 to 75)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 28.1244%;\"\u003e\n \u003cp\u003eMean (range) body mass index\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 8.0072%;\"\u003e\n \u003cp\u003e28 (22 to 35)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 8.8675%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 28.1244%;\"\u003e\n \u003cp\u003eMean (range) percentage male\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 8.0072%;\"\u003e\n \u003cp\u003e30 (0 to 95)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 8.8675%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003cdiv class=\"gridtable\"\u003e\n \u003ctable id=\"Tabb\" border=\"1\"\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eTable\u0026nbsp;2. Multivariate analysis\u003c/span\u003e\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eCovariate\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eCoefficient (SE)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e2-sided p-value\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eR\u003csup\u003e2\u003c/sup\u003e analogue\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eIntercept\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e0.53 (0.18)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eSponsorship (commercial vs non-commercial)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e0.14 (0.08)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e0.17\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eMean pain level at baseline\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e-0.11 (0.31)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e0.03\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e0.09\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eMean effect size in treatment arm\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e0.15 (0.03)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e0.05\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eRegion: N-America vs other\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e-0.34 (0.24)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e0.08\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e0.02\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eOverall R\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e0.33\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e\u003cstrong\u003eLegend table 2.\u003c/strong\u003e Multivariate analysis of trial level covariates that contributed to explained variance. The following covariates did not contribute: trial size, year of publication, duration of treatment, length of follow-up after last treatment, timing of follow-up, route of administration, mean age, sex distribution, mean body mass index. df\u0026thinsp;=\u0026thinsp;degrees of freedom. A modified version of Cochrane Risk of Bias tool (RoB-2) was used with the addition of trial size (\u0026gt;\u0026thinsp;49 participants overall) and loss to follow-up (\u0026gt;\u0026thinsp;15%). Selective reporting bias was based on compliance with requirements for registration of the trial in a public register; high risk: registered trials that submitted protocol after trial start; some concerns: no information about registration; low risk: registered trials that submitted protocol before start of trial.\u003c/p\u003e\n\u003ch3\u003eRisk of bias\u003c/h3\u003e\n\u003cdiv class=\"BlockQuote\"\u003e\n \u003cp\u003eOf the 179 trials, 168 (93%) were judged to be at unclear or high risk of bias on at least one domain and 146 (81%) were at high risk on at least one domain; 132 (73%) trials adequately generated their randomization sequence, 112 (62%) appropriately concealed allocation, 84 (47%) adequately blinded patients and caregivers, and 88 (49%) adequately blinded outcome assessors (because only patient reported outcome measures were extracted the participant was in all trials designated as the assessor). Only nine (5%) trials assessed the success of participant blinding. In 53 trials (29%), the total sample size was \u0026lt;\u0026thinsp;50, while 53 trials (29%) reported\u0026thinsp;\u0026ge;\u0026thinsp;15% missing outcome data. High risk or some concerns of bias due to selective reporting was found in 68 (38%) and 54 (30%) trials, respectively. A summary of risk of bias in each domain is shown in Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e, while assessments of risk of bias in each trial are provided in Additional file 1: fig. \u003cspan class=\"InternalRef\"\u003eS1\u003c/span\u003e.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\n \u003ch2\u003eClinical outcomes\u003c/h2\u003e\n \u003cdiv id=\"Sec12\" class=\"Section3\"\u003e\n \u003ch2\u003eEffect sizes\u003c/h2\u003e\n \u003cdiv class=\"BlockQuote\"\u003e\n \u003cp\u003eAnalysis of effect sizes (Hedges\u0026rsquo; g) in clinical outcomes in the placebo arms showed substantial heterogeneity, with an overall I\u003csup\u003e2\u003c/sup\u003e of 88% (p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001), i.e., 88% of the variance in observed effects reflected variance in true effects rather than sampling error. Effect sizes ranged from \u0026minus;\u0026thinsp;0.94 to 3.80 (interquartile range, IQR, 0.33 to 0.97), while the pooled effect size was 0.64 (95% CI 0.56 to 0.72). The prediction interval was \u0026minus;\u0026thinsp;0.31 to 1.60, i.e., the true effect size in 95% of all comparable populations would fall in this interval. Moderate or large effect sizes (Hedges g\u0026thinsp;\u0026ge;\u0026thinsp;0.5) were described in 109 (60%) trials. Fifteen (8%) trials described worsened pain compared to baseline. Pooled effect size in the no-treatment arms of four trials was 0.14 (95% CI = -0.03 to 0.31).\u003c/p\u003e\n \u003c/div\u003e\n \u003c/div\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\n \u003ch2\u003eRegression analysis\u003c/h2\u003e\n \u003cdiv class=\"BlockQuote\"\u003e\n \u003cp\u003eIn univariate meta-regression analysis, six of 16 predefined covariates were significantly (p\u0026thinsp;\u0026le;\u0026thinsp;0.05) associated with effect size (Additional file 1: table S5). In multivariate analysis twelve covariates had p-values of \u0026ge;\u0026thinsp;0.10 and were removed from the model. In the final model four covariates remained: effect size in the active treatment arms, geographic region, pain level at baseline, and source of funding, together explaining 33% of the variance (Additional file 1: table S2, figs. S2a-d)\u003c/p\u003e\n \u003c/div\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\n \u003ch2\u003eSubgroup and sensitivity analyses\u003c/h2\u003e\n \u003cdiv class=\"BlockQuote\"\u003e\n \u003cp\u003eSubgroup analysis failed to reveal significant differences in pooled effect size according to the route of delivery (p\u0026thinsp;=\u0026thinsp;0.65) (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e). Forest plots (Additional file 1: figs. S3a-e) show effect sizes per trial in each of the five subgroups. I-squared values remained high in all subgroups.\u003c/p\u003e\n \u003c/div\u003e\n \u003cp\u003eSubgroup analysis revealed significant differences between geographic regions (p\u0026thinsp;=\u0026thinsp;0.006); trials performed in North America (all but two from the USA) had the highest pooled effect size and those in Australia the lowest (Hedges\u0026rsquo; g 0.89 and 0.50, respectively) (Additional file 1: fig. S2a).\u003c/p\u003e\n \u003cp\u003ePooled effect size did not differ significantly based on risk of bias in any of the seven domains. After removing trials with high risk of bias in at least one domain, overall pooled effect size remained virtually unchanged (0.65, 95% CI 0.52 to 0.78) (Additional file 1: tables S6a and S6b).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\n \u003ch2\u003eCorrelation analysis\u003c/h2\u003e\n \u003cdiv class=\"BlockQuote\"\u003e\n \u003cp\u003eA moderately strong correlation was found between effect size in the placebo and treatment groups, with Pearson\u0026apos;s \u003cem\u003er\u003c/em\u003e of 0.40 (P\u0026thinsp;\u0026lt;\u0026thinsp;0.001) (Additional file 1: Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ed).\u003c/p\u003e\n \u003c/div\u003e\n\u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eIn this meta-analysis of placebo responses in randomized controlled trials evaluating non-surgical interventions for KOA, improvements in pain were reported in more than 90% of the trials, with effect sizes ranging from moderate to large in six out of ten trials. The substantial between-trial variance and wide prediction interval indicates uncertainty; therefore, pooled effect size estimates were intentionally not emphasized and examination of sources of heterogeneity were prioritized [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. Based on findings in previous reviews as well as clinical judgement, sixteen clinically relevant trial-level covariates were incorporated into a univariate metaregression model, six of which were significantly associated with effect size. After stepwise backward multivariate regression analysis four covariates remained statistically significant, together explaining 33% of the variance.\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e\u003ch2\u003eStrengths and limitations\u003c/h2\u003e\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eStrengths of this meta-analysis include the use of a rigorous and comprehensive search strategy to update previous reviews; clear and transparent eligibility criteria; calculations based on each trial\u0026rsquo;s SD of change; incorporating both uni- and multivariate regression models to investigate the role of potential covariates; and a standardized critical appraisal of studies for quality. Extraction of outcome data (means, SDs, participant numbers, registration and funding) was duplicated, thus reducing the risk of errors or bias. Finally, the calculation of effect sizes allowed for combining studies using different outcome instruments, thus expanding the range of eligible trials.\u003c/p\u003e\u003cp\u003eA major limitation of this meta-analysis is the inability, due to a lack of information in the included trials, to evaluate the role of psychological factors that are known to influence placebo effects. While the personality traits of participants (such as optimism or suggestibility) are unlikely to have differed much across the included trials, it is reasonable to assume that differences in context-related elements, including the patient-provider relationship, contributed substantially to the observed differences in placebo responses between trials. A description of the core elements involved in such relationships is outside the scope of this review, but competence and warmth conveyed by the health-care provider have been suggested as two key dimensions [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. In turn, such elements may contribute to positive expectations, which are thought to play a role in shaping placebo effects and can lead to clinically relevant and enduring improvements, even on physical outcomes [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. For example, in a double-blind sham-controlled surgical trial on patients with advanced Parkinson\u0026rsquo;s disease, perceived treatment was more strongly related to outcome than was the actual treatment received during the 12-month follow-up [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. Incorporating treatment outcome expectations as a covariate may improve assay sensitivity in future placebo-controlled pain trials [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eAn additional limitation is that the analysis relies on trial-level aggregate data rather than individual participant data (IPD). While IPD could offer greater sensitivity and might reveal clinically significant associations [\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e], only a small number of published OA trials currently report such data [54].\u003c/p\u003e\u003cp\u003eBased on trials with a no-treatment arm, previous reviews have recognised non-specific factors as important contributors to placebo responses [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. In our analysis, pooled effect size in the four trials with a no-treatment group was small (0.14); however, the wide 95% confidence intervals limits certainty regarding the role of non-specific effects. Symptoms associated with KOA are known to naturally fluctuate; duration of symptoms prior to enrolment is therefore a potential confounder by affecting the likelihood of spontaneous improvement [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]. In a prospective cohort of patients with KOA, regression to the mean was found to be responsible for as much as one point on a 0\u0026ndash;10-point pain scale [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]. Although we were prevented from including disease duration as a covariate in our analysis because of insufficient or unclear data, it is possible that between-trial differences in disease duration could have contributed to explained variance.\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec18\" class=\"Section2\"\u003e\u003ch2\u003eSources of explained variance\u003c/h2\u003e\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eIn multivariate analysis type of sponsorship contributed 17% to explained variance, with commercially funded trials reporting higher effect sizes. To our knowledge only two meta-analyses have specifically examined the influence of funding on placebo responses, with differing conclusions [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. It has been suggested that commercially sponsored trials more frequently test novel agents or interventions, leading to greater expectations of efficacy among participants and outcome assessors [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eIn our analysis pain level at baseline was identified as a predictor of improved outcomes, consistent with findings from other reviews, including in the musculoskeletal domain [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]. This may suggest that individuals presenting with more severe knee symptoms and underlying pathology derive greater benefit from treatment. Alternatively, it could reflect statistical regression to the mean, whereby extreme baseline scores tend to approach the mean at subsequent follow-up assessments [\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eEffect size in the active treatment arms accounted for five percent of explained variance, and there was a correlation of 0.40 between effect sizes in the active treatment and placebo arms. These findings are not unexpected and are likely due to underlying contextual and non-specific factors common to both groups. Ideally, with an equal number of visits, appropriate blinding, randomization etc., such factors would be expected to be comparable across groups. Substantially stronger associations have been reported in other reviews [\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e] and in our previous review of placebo responses in trials of minimally invasive interventions we estimated an R\u003csup\u003e2\u003c/sup\u003e of 0.78 for primary endpoints [\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eTrials permitting the use of rescue medication demonstrated significantly improved outcomes in univariate, but not in multivariate, analysis. Trials prescribing home exercise as an adjunct to placebo reported slightly higher effect sizes, though not significant. Effect sizes exhibited significant variation across geographic regions, with studies conducted in North America reporting the highest and those from Asia and Australia the lowest effect sizes both in uni- and multivariate analysis. A prior systematic review of OA trials reported lower placebo responses in developed countries compared to developing countries; however, the specific criteria employed in their analysis were not described [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. To our knowledge, no other reviews have specifically examined the influence of geographic region on placebo responses in OA trials.\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec19\" class=\"Section2\"\u003e\u003ch2\u003eComparison with other trials and reviews\u003c/h2\u003e\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eThe observation that effect sizes in placebo arms are typically in the moderate to high range aligns with findings from several systematic reviews, demonstrating that the placebo effect constitutes a substantial proportion of the overall efficacy observed with active treatments [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e]. Due to substantial statistical heterogeneity concerns may be raised regarding the validity of treatment effects, such as in research examining the efficacy of NSAIDs. However, variability observed in placebo groups is to be anticipated because of differences in demographic factors (e.g., age, sex distribution), study design elements (e.g., outcome measures, follow-up duration), and clinical characteristics (e.g., participant expectancy, disease duration). For example, even patients with similar radiographic features of knee OA have been shown to be extremely diverse in their pain pathophysiology, with the relative contribution of nociceptive and neuropathic mechanisms varying across studies [\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eLater publication year and length of treatment or follow-up period have been linked to increased placebo responses in other reviews [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e]. Though such associations were not replicated in the present analysis it should be noted that only the outcome value at one time point per trial was extracted; therefore within-trial longitudinal changes over time cannot be excluded. Interestingly, previous analyses have described persisting effects on pain and function after both sham surgery [\u003cspan additionalcitationids=\"CR44 CR45\" citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e] and intraarticular injections [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e], and patients with advanced Parkinson\u0026rsquo;s disease that underwent sham surgery described persistent improvements in physical function 12 months after the procedure [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e].\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec20\" class=\"Section2\"\u003e\u003ch2\u003eRoute of placebo delivery\u003c/h2\u003e\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eIn our analysis, the method of placebo administration was not significantly associated with outcome, contradicting a common assertion that more invasive, complex, or time-intensive interventions tend to elicit stronger placebo effects [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e, \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e, \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e]. Like the current analysis, these reviews were based on non-randomised comparisons. The inability to control for confounding variables in this design introduces potential bias, thereby making the concept of \"enhanced placebos\" uncertain. The most reliable approach for evaluating the comparative effectiveness of different placebo interventions involves directly comparing delivery methods within a single clinical trial. This requires a four-arm design, incorporating two distinct active treatments alongside their respective placebos.\u003c/p\u003e\u003cp\u003eIn a meta-analysis based on twelve studies conforming to this trial design the authors found no consistent differences in effect size between placebo trials with varying \u0026ldquo;intensiveness\u0026rdquo; [\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e]. Two of the RCTs included in that analysis comprised patients with musculoskeletal disorders. One of them compared a placebo pill with sham acupuncture for arm pain [\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e]. The sham group reported significantly greater improvements in pain, but no changes in arm function or symptom severity. In a trial that included patients with acute musculoskeletal pain secondary to trauma, there were no differences in pain response between parenteral and oral placebos measured up to two hours after administration [\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e]. A more recent RCT of ketoprofen for KOA compared topical and oral placebos; at week 12 improvements were significantly greater in the topical group [\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eGiven the limited number of published RCTs that have incorporated more than one type of placebo within the same study, we contend that the concept of \"enhanced placebo\" lacks sufficient empirical support, and further research employing this type of design is necessary to substantiate this hypothesis.\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e"},{"header":"Conclusion","content":"\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eMore than half of the included trials described moderate to large effect sizes in their placebo arms, although significant statistical heterogeneity was noted across studies. Multivariate analysis revealed that about a third of the observed variance could be explained by four covariates: type of sponsorship, baseline pain level, effect size in the active treatment arm and geographic region. The factors underlying the remaining unexplained variance are undetermined, though psychological variables\u0026mdash;such as patient expectations and provider interactions\u0026mdash;are likely to have contributed. Due to the lack of data within the trials analysed, the potential impact of such factors could not be evaluated. Incorporating psychological assessments in future OA trials could improve our understanding of their role in placebo responses.\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets used and analysed in this current study are available from the\u003c/p\u003e\n\u003cp\u003ecorresponding author upon reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNo funding was involved in this review\u003c/p\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCRediT authorship contribution statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRobin Holtedahl\u003c/strong\u003e: Conceptualization, Methodology, Formal\u003c/p\u003e\n\u003cp\u003eanalysis, Investigation, Data curation, Writing \u0026ndash; original draft, Writing \u0026ndash;\u003c/p\u003e\n\u003cp\u003ereview \u0026amp; editing, Visualization. \u003cstrong\u003eAlexandra Christina Hott\u003c/strong\u003e: Conceptual\u0026shy;\u003c/p\u003e\n\u003cp\u003eisation, Methodology, Data curation, Writing \u0026ndash; review \u0026amp; editing. \u003cstrong\u003eJens Ivar Brox\u003c/strong\u003e: Methodology, Data curation, Writing \u0026ndash; review \u0026amp; editing. \u003cstrong\u003eInger Johanne Hansen\u003c/strong\u003e:\u003c/p\u003e\n\u003cp\u003eData curation. \u003cstrong\u003eJohan Bj\u0026oslash;rneboe\u003c/strong\u003e: Data curation, Writing \u0026ndash; review \u0026amp; editing. All authors read and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eArtificial Intelligence\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eArtificial Intelligence was used to improve grammar.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePatient and public involvement\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePatients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePatient consent for publication\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eLong H, Liu Q, Yin H, Wang K, Diao N, Zhang Y, et al. Arthritis Rheumatol. 2022;74:1172\u0026ndash;83. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/art.42089\u003c/span\u003e\u003cspan address=\"10.1002/art.42089\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Prevalence Trends of Site-Specific Osteoarthritis From 1990 to 2019: Findings From the Global Burden of Disease Study 2019.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGibbs AJ, Gray B, Wallis JA, Taylor NF, Kemp JL, Hunter DJ, et al. Recommendations for the management of hip and knee osteoarthritis: A systematic review of clinical practice guidelines. Osteoarthritis Cartilage. 2023;31:1280\u0026ndash;92. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.joca.2023.05.015\u003c/span\u003e\u003cspan address=\"10.1016/j.joca.2023.05.015\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKrogsb\u0026oslash;ll LT, Hr\u0026oacute;bjartsson A, G\u0026oslash;tzsche PC. Spontaneous improvement in randomised clinical trials: meta-analysis of three-armed trials comparing no treatment, placebo and active intervention. BMC Med Res Methodol. 2009;9:1. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/1471-2288-9-1\u003c/span\u003e\u003cspan address=\"10.1186/1471-2288-9-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBenedetti F. Placebo and the new physiology of the doctor-patient relationship. Physiol Rev. 2013;93:1207\u0026ndash;46. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1152/physrev.00043.2012\u003c/span\u003e\u003cspan address=\"10.1152/physrev.00043.2012\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZou K, Wong J, Abdullah N, Chen X, Smith T, Doherty M, et al. Examination of overall treatment effect and the proportion attributable to contextual effect in osteoarthritis: Meta-Analysis of randomised controlled trials. Ann Rheum Dis. 2016;75:1964\u0026ndash;70. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1136/annrheumdis-2015-208387\u003c/span\u003e\u003cspan address=\"10.1136/annrheumdis-2015-208387\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhang W, Robertson J, Jones AC, Dieppe PA, Doherty M. The placebo effect and its determinants in osteoarthritis: Meta-analysis of randomised controlled trials. Ann Rheum Dis. 2008;67:1716\u0026ndash;23. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1136/ard.2008.092015\u003c/span\u003e\u003cspan address=\"10.1136/ard.2008.092015\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eUndurraga J, Baldessarini RJ, Randomized. Placebo-Controlled Trials of Antidepressants for Acute Major Depression: Thirty-Year Meta-Analytic Review. Neuropsychopharmacology. 2012;37:851\u0026ndash;64. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/npp.2011.306\u003c/span\u003e\u003cspan address=\"10.1038/npp.2011.306\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePrevitali D, Merli G, Di Laura Frattura G, Candrian C, Zaffagnini S, Filardo G. The Long-Lasting Effects of Placebo Injections in Knee Osteoarthritis: A Meta-Analysis. Cartilage. 2021;13:S185\u0026ndash;96. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/1947603520906597\u003c/span\u003e\u003cspan address=\"10.1177/1947603520906597\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWartolowska KA, Gerry S, Feakins BG, Collins GS, Cook J, Judge A, et al. A meta-analysis of temporal changes of response in the placebo arm of surgical randomized controlled trials: An update. Trials. 2017;18:12\u0026ndash;9. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s13063-017-2070-9\u003c/span\u003e\u003cspan address=\"10.1186/s13063-017-2070-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHr\u0026oacute;bjartsson A, G\u0026oslash;tzsche PC. Placebo interventions for all clinical conditions. Cochrane Database Syst Reviews 2010;2010. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/14651858.CD003974.pub3\u003c/span\u003e\u003cspan address=\"10.1002/14651858.CD003974.pub3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTuttle AH, Tohyama S, Ramsay T, Kimmelman J, Schweinhardt P, Bennett GJ, et al. Increasing placebo responses over time in U.S. clinical trials of neuropathic pain. Pain. 2015;156:2616\u0026ndash;26. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1097/j.pain.0000000000000333\u003c/span\u003e\u003cspan address=\"10.1097/j.pain.0000000000000333\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWalach H, Maidhof C. Is the placebo effect dependent on time? A meta-analysis. How expectancies shape experience. American Psychological Association; 2004. pp. 321\u0026ndash;32. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1037/10332-013\u003c/span\u003e\u003cspan address=\"10.1037/10332-013\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eH\u0026auml;user W, Bartram-Wunn E, Bartram C, Reinecke H, T\u0026ouml;lle T. Systematic review: Placebo response in drug trials of fibromyalgia syndrome and painful peripheral diabetic neuropathy - Magnitude and patient-related predictors. Pain. 2011;152:1709\u0026ndash;17. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.pain.2011.01.050\u003c/span\u003e\u003cspan address=\"10.1016/j.pain.2011.01.050\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHr\u0026oacute;bjartsson A, G\u0026oslash;tzsche PC. Is the placebo powerless? Update of a systematic review with 52 new randomized trials comparing placebo with no treatment. J Intern Med. 2004;256:19\u0026ndash;100. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.1365-2796.2004.01355.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1365-2796.2004.01355.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePage MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD et al. The PRISMA 2020 statement: An updated guideline for reporting systematic reviews. BMJ 2021;372. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1136/bmj.n71\u003c/span\u003e\u003cspan address=\"10.1136/bmj.n71\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSterne JA, Hern\u0026aacute;n MA, Reeves BC, Savović J, Berkman ND, Viswanathan M, et al. ROBINS-I: A tool for assessing risk of bias in non-randomised studies of interventions. BMJ (Online). 2016;355:1\u0026ndash;7. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1136/bmj.i4919\u003c/span\u003e\u003cspan address=\"10.1136/bmj.i4919\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDankel SJ, Loenneke JP. Effect sizes for paired data should use the change score variability rather than the pretest variability. J Strength Cond Res. 2021;35:1773\u0026ndash;8. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1519/JSC.0000000000002946\u003c/span\u003e\u003cspan address=\"10.1519/JSC.0000000000002946\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSterne JAC, Savović J, Page MJ, Elbers RG, Blencowe NS, Boutron I, Cates CJ, Cheng H-Y, Corbett MS, Eldridge SM, Hern\u0026aacute;n MA, Hopewell S, Hr\u0026oacute;bjartsson A, Junqueira DR, J\u0026uuml;ni P, Kirkham JJ, Lasserson T, Li T, McAleenan A, Reeves BC, Shepperd S, Shrier I, Stew HJ. RoB 2: A revised Cochrane risk-of-bias tool for randomized trials. BMJ 2019:1\u0026ndash;24.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHiggins JPT, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. Br Med J. 2003;327:557\u0026ndash;60.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eInthout J, Ioannidis JPA, Rovers MM, Goeman JJ. Plea for routinely presenting prediction intervals in meta-analysis. BMJ Open. 2016;6:10247. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1136/bmjopen-2015\u003c/span\u003e\u003cspan address=\"10.1136/bmjopen-2015\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eNordmann AJ, Kasenda B, Briel M. Meta-analyses: What they can and cannot do. Swiss Med Wkly. 2012;142. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.4414/smw.2012.13518\u003c/span\u003e\u003cspan address=\"10.4414/smw.2012.13518\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCote MP, Lubowitz JH, Rossi MJ, Brand JC. Reviews Pooling Heterogeneous, Low-Evidence, High-Bias Data Result in Incorrect Conclusions: But Heterogeneity is an Opportunity to Explore. Arthrosc - J Arthroscopic Relat Surg. 2018;34:3126\u0026ndash;8. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.arthro.2018.10.005\u003c/span\u003e\u003cspan address=\"10.1016/j.arthro.2018.10.005\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHowe LC, Leibowitz KA, Crum AJ. When your doctor Gets it and Gets you: The critical role of competence and warmth in the patient-provider interaction. Front Psychiatry. 2019;10. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fpsyt.2019.00475\u003c/span\u003e\u003cspan address=\"10.3389/fpsyt.2019.00475\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRosenkj\u0026aelig;r S, Lunde SJ, Kirsch I, Vase L, Expectations. How and when do they contribute to placebo analgesia? Front Psychiatry 2022;13. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fpsyt.2022.817179\u003c/span\u003e\u003cspan address=\"10.3389/fpsyt.2022.817179\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMcrae C, Cherin E, Yamazaki M, TG, Diem G, Vo AH, Russell D et al. Effects of Perceived Treatment on Quality of Life and Medical Outcomes in a Double-blind Placebo Surgery Trial. vol. 61. 2004.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFrisaldi E, Shaibani A, Benedetti F. Why We should Assess Patients\u0026rsquo; Expectations in Clinical Trials. Pain Ther. 2017;6:107\u0026ndash;10. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s40122-017-0071-8\u003c/span\u003e\u003cspan address=\"10.1007/s40122-017-0071-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLangford DJ, Lou R, Sheen S, Amtmann D, Colloca L, Edwards RR, et al. Expectations for Improvement: A Neglected but Potentially Important Covariate or Moderator for Chronic Pain Clinical Trials. J Pain. 2023;24:575\u0026ndash;81. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jpain.2022.12.011\u003c/span\u003e\u003cspan address=\"10.1016/j.jpain.2022.12.011\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGlasziou P, Chalmers L, Rawlins M, McCulloch P. When are randomised trials unnecessary? Picking signal from noise Paul Glasziou and colleagues. Br Med J. 2007;334:349\u0026ndash;51.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eEnglund M, Turkiewicz A. Pain in clinical trials for knee osteoarthritis: estimation of regression to the mean. Lancet Rheumatol. 2023;5:e309\u0026ndash;11. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S2665-9913(23)00090-5\u003c/span\u003e\u003cspan address=\"10.1016/S2665-9913(23)00090-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJones BDM, Razza LB, Weissman CR, Karbi J, Vine T, Mulsant LS, et al. Magnitude of the Placebo Response across Treatment Modalities Used for Treatment-Resistant Depression in Adults: A Systematic Review and Meta-analysis. JAMA Netw Open. 2021;4. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1001/jamanetworkopen.2021.25531\u003c/span\u003e\u003cspan address=\"10.1001/jamanetworkopen.2021.25531\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eIoannidis K, Huneke NTM, Solly JE, Fusetto Veronesi G, Tzagarakis C, Parlatini V, et al. Placebo and nocebo effects in gambling disorder pharmacological trials: A meta-analysis. Acta Neuropsychiatr. 2024. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1017/neu.2024.52\u003c/span\u003e\u003cspan address=\"10.1017/neu.2024.52\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDunlop BW, Thase ME, Wun C-C, Fayyad R, Guico-Pabia CJ, Musgnung J, et al. A meta-analysis of factors impacting detection of antidepressant efficacy in clinical trials: the importance of academic sites. Neuropsychopharmacology. 2012;37:2830\u0026ndash;6. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/npp.2012.153\u003c/span\u003e\u003cspan address=\"10.1038/npp.2012.153\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eVase L, Vollert J, Finnerup NB, Miao X, Atkinson G, Marshall S, et al. Predictors of the placebo analgesia response in randomized controlled trials of chronic pain: A meta-analysis of the individual data from nine industrially sponsored trials. Pain. 2015;156:1795\u0026ndash;802. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1097/j.pain.0000000000000217\u003c/span\u003e\u003cspan address=\"10.1097/j.pain.0000000000000217\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWen X, Luo J, Mai Y, Li Y, Cao Y, Li Z, et al. Placebo Response to Oral Administration in Osteoarthritis Clinical Trials and Its Associated Factors: A Model-Based Meta-analysis. JAMA Netw Open. 2022;E2235060. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1001/jamanetworkopen.2022.35060\u003c/span\u003e\u003cspan address=\"10.1001/jamanetworkopen.2022.35060\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eVickers AJ, Altman DG. Analysing controlled trials with baseline and follow up measurements. Br Med J. 2001;323:1123\u0026ndash;4. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1136/bmj.323.7321.1123\u003c/span\u003e\u003cspan address=\"10.1136/bmj.323.7321.1123\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKirsch I, Sapirstein G. Listening to Prozac but hearing placebo: A meta-analysis of antidepressant medication. Prev Treat. 1998;1:1\u0026ndash;16. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1037/1522-3736.1.1.12a\u003c/span\u003e\u003cspan address=\"10.1037/1522-3736.1.1.12a\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWalach H, Sadaghiani C, Dehm C, Bierman D. The therapeutic effect of clinical trials: understanding placebo response rates in clinical trials-A secondary analysis 2005. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/1471-2288-5\u003c/span\u003e\u003cspan address=\"10.1186/1471-2288-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHoltedahl R, Brox JI, Tjomsland O. Placebo effects in trials evaluating 12 selected minimally invasive interventions: a systematic review and meta-analysis. BMJ Open. 2015;5:e007331. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1136/bmjopen-2014-007331\u003c/span\u003e\u003cspan address=\"10.1136/bmjopen-2014-007331\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHuang Z, Chen J, Hu QS, Huang Q, Ma J, Pei FX, et al. Meta-analysis of pain and function placebo responses in pharmacological osteoarthritis trials. Arthritis Res Ther. 2019;21. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s13075-019-1951-6\u003c/span\u003e\u003cspan address=\"10.1186/s13075-019-1951-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChen AT, Shrestha S, Collins JE, Sullivan JK, Losina E, Katz JN. Estimating contextual effect in nonpharmacological therapies for pain in knee osteoarthritis: a systematic analytic review. Osteoarthritis Cartilage. 2020;28:1154\u0026ndash;69. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.joca.2020.05.007\u003c/span\u003e\u003cspan address=\"10.1016/j.joca.2020.05.007\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMandl LA, Losina E. Relative Efficacy of Knee Osteoarthritis Treatments: Are All Placebos Created Equal? Ann Intern Med. 2015;162:71\u0026ndash;2. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.7326/M14-2636\u003c/span\u003e\u003cspan address=\"10.7326/M14-2636\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eReiter-Niesert S, Boers M, Detert J. Short-term placebo response in trials of patients with symptomatic osteoarthritis: Differences between hip and knee. Osteoarthritis Cartilage. 2016;24:1007\u0026ndash;11. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.joca.2016.01.002\u003c/span\u003e\u003cspan address=\"10.1016/j.joca.2016.01.002\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSihvonen R, Paavola M, Malmivaara A, It\u0026auml;l\u0026auml; A, Joukainen A, Kalske J, et al. Arthroscopic partial meniscectomy for a degenerative meniscus tear: a 5 year follow-up of the placebo-surgery controlled FIDELITY (Finnish Degenerative Meniscus Lesion Study) trial. Br J Sports Med. 2020;54:1332\u0026ndash;9. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1136/bjsports-2020-102813\u003c/span\u003e\u003cspan address=\"10.1136/bjsports-2020-102813\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSchr\u0026oslash;der CP, Skare \u0026Oslash;, Reiker\u0026aring;s O, Mowinckel P, Brox JI. Sham surgery versus labral repair or biceps tenodesis for type II SLAP lesions of the shoulder: A three-armed randomised clinical trial. Br J Sports Med. 2017;51:1759\u0026ndash;66. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1136/bjsports-2016-097098\u003c/span\u003e\u003cspan address=\"10.1136/bjsports-2016-097098\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMoseley JB, O\u0026rsquo;Malley K, Peterson NJ, Menke TJ, Brody BA, Kuykendall DH, et al. A controlled trial of arthroscopic surgery for osteoarthritis of the knee. N Engl J Med. 2002;347:81\u0026ndash;2.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBeard DJ, Rees JL, Cook JA, Rombach I, Cooper C, Merritt N, et al. Arthroscopic subacromial decompression for subacromial shoulder pain (CSAW): a multicentre, pragmatic, parallel group, placebo-controlled, three-group, randomised surgical trial. Lancet. 2018;391:329\u0026ndash;38. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S0140-6736(17)32457-1\u003c/span\u003e\u003cspan address=\"10.1016/S0140-6736(17)32457-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAbhishek A, Doherty M. Mechanisms of the placebo response in pain in osteoarthritis. Osteoarthritis Cartilage. 2013;21:1229\u0026ndash;35. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.joca.2013.04.018\u003c/span\u003e\u003cspan address=\"10.1016/j.joca.2013.04.018\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBannuru RR, McAlindon TE, Sullivan MC, Wong JB, Kent DM, Schmid CH. Effectiveness and implications of alternative placebo treatments: A systematic review and network meta-analysis of osteoarthritis trials. Ann Intern Med. 2015;163:365\u0026ndash;92. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.7326/M15-0623\u003c/span\u003e\u003cspan address=\"10.7326/M15-0623\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eF\u0026auml;ssler M, Meissner K, Kleijnen J, Hr\u0026oacute;bjartsson A, Linde K. A systematic review found no consistent difference in effect between more and less intensive placebo interventions. J Clin Epidemiol. 2015;68:442\u0026ndash;51. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jclinepi.2014.11.018\u003c/span\u003e\u003cspan address=\"10.1016/j.jclinepi.2014.11.018\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKaptchuk TJ, Stason WB, Davis RB, Legedza ATR, Schnyer RN, Kerr CE, et al. Sham device versus inert pill: Randomised controlled trial of two placebo treatments. Br Med J. 2006;332:391\u0026ndash;4. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1136/bmj.38726.603310.55\u003c/span\u003e\u003cspan address=\"10.1136/bmj.38726.603310.55\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSchwartz NA, Turturro MA, Istvan DJ, Larkin GL. Patients\u0026rsquo; Perceptions of Route of Nonsteroidal Anti-inflammatory Drug Administration and Its Effect on Analgesia. vol. 7. 2000.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eConaghan PG, Dickson J, Bolten W, Cevc G, Rother M. A multicentre, randomized, placebo- and active-controlled trial comparing the efficacy and safety of topical ketoprofen in Transfersome gel (IDEA-033) with ketoprofen-free vehicle (TDT 064) and oral celecoxib for knee pain associated with osteoarthritis. Rheumatol (United Kingdom). 2013;52:1303\u0026ndash;12. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/rheumatology/ket133\u003c/span\u003e\u003cspan address=\"10.1093/rheumatology/ket133\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Placebo, Sham, Knee osteoarthritis, Non-surgical management, Pain outcomes, Systematic review, Meta-analysis","lastPublishedDoi":"10.21203/rs.3.rs-7888576/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7888576/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cb\u003eBackground\u003c/b\u003e Previous reviews have demonstrated notable placebo responses in placebo-controlled clinical trials assessing osteoarthritis treatments; however, there is limited consensus regarding the relative importance of factors influencing these responses. This review and meta-analysis aimed to (1) evaluate effect sizes observed in the placebo arms of knee osteoarthritis trials published since 2008, and (2) identify factors contributing to explained variance through meta-regression analysis.\u003c/p\u003e\u003cp\u003e\u003cb\u003eMethods\u003c/b\u003e Outcome data at baseline and follow-up were extracted, and Hedges\u0026rsquo; g was calculated as a measure of effect size for each study using a random effects model. Heterogeneity was quantified using the I\u0026sup2; statistic. Sources of variance were investigated through multivariate meta-regression involving 16 clinically relevant covariates. The Cochrane RoB-2 tool was used to assess risk of bias.\u003c/p\u003e\u003cp\u003e\u003cb\u003eResults\u003c/b\u003e Searches provided 179 trials that met inclusion criteria. Of these, 109 trials (60%) reported Hedges\u0026rsquo; g values\u0026thinsp;\u0026ge;\u0026thinsp;0.5. Substantial heterogeneity was observed (I\u0026sup2; = 89%, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001), with a broad prediction interval (-0.30 to 1.63). In multivariate meta-regression two-thirds of the variance remained unexplained, while the covariates baseline pain, commercial funding, effect size in active treatment arms and geographical region together accounted for 33% of explained variance. Risk of bias, duration of follow-up, year of publication, and method of placebo administration were not found to significantly influence the outcomes.\u003c/p\u003e\u003cp\u003e\u003cb\u003eConclusion\u003c/b\u003e: Most trials demonstrated moderate to large placebo effect sizes, but there was marked between-trial statistical heterogeneity and a substantial proportion of the variance remained unexplained. An important limitation is the lack of data regarding psychological variables, such as participant expectations and the patient\u0026ndash;provider relationship, which are recognized as major determinants of placebo effects.\u003c/p\u003e","manuscriptTitle":"Placebo responses in non-surgical trials for knee osteoarthritis: A systematic review and meta-analysis","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-11-18 17:12:07","doi":"10.21203/rs.3.rs-7888576/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"0e14e3cf-d7bb-4061-91b2-013bff5887a5","owner":[],"postedDate":"November 18th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-12-27T02:53:51+00:00","versionOfRecord":[],"versionCreatedAt":"2025-11-18 17:12:07","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7888576","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7888576","identity":"rs-7888576","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}