Vase Lene, Vollert Jan, Finnerup Nanna B, Miao Xiaopeng, Atkinson Gary, Marshall Scott, Nemeth Robert, Lange Bernd, Liss Charlie, Price Donald D, Maier Christoph, Jensen Troels S, Segerdahl Märta
Department of Psychology and Behavioural Sciences, School of Business and Social Sciences, Aarhus University, Aarhus, Denmark Danish Pain Research Center, Aarhus University Hospital, Aarhus, Denmark Department of Pain Medicine, University Clinic Bergmannsheil, Bochum, Germany Department of Biometrics, Biogen Idec, Cambridge, MA, USA Pfizer Ltd, Global Innovative Pharma, Sandwich, United Kingdom Grünenthal GmbH, Aachen, Germany AstraZeneca Pharmaceuticals, Biostatistics and Informatics, Gaithersburg, MD, USA Division of Neuroscience, Department of Oral and Maxillofacial Surgery, University of Florida, Gainesville, FL, USA H. Lundbeck Pharma A/A, Valby, Denmark.
Pain. 2015 Sep;156(9):1795-1802. doi: 10.1097/j.pain.0000000000000217.
A large number of analgesics have failed to prove superiority over placebo in randomized controlled trials (RCTs), and as this has been related to increasing placebo responses, there is currently an interest in specifying predictors of the placebo response. The literature on placebo mechanisms suggests that factors related to patients' expectations of treatment efficacy are pivotal for the placebo response. Also, general characteristics of RCTs have been suggested to influence the placebo response. Yet, only few meta-analyses have directly tested these hypotheses. Placebo data from 9 industrially sponsored, randomized, double-blind, placebo-controlled, multicenter phase III trials in 2017 adult patients suffering from chronic painful osteoarthritis (hip or knee) or low back pain were included. The primary outcome was pain intensity. Based on previous studies, we chose 3 expectancy-related primary predictors: type of active medication, randomization ratio, and number of planned face-to-face visits. In addition, explorative analyses tested whether RCT and patients' characteristics predicted the placebo response. Opioid trials, a high number of planned face-to-face visits, and randomization ratio predicted the magnitude of the placebo response, thereby supporting the expectancy hypothesis. Exploratory models with baseline pain intensity, age, washout length, and discontinuation because of adverse events accounted for approximately 10% of the variability in the placebo response. Based on these results and previous mechanisms studies, we think that patients' perception of treatment allocation and expectations toward treatment efficacy could potently predict outcomes of RCTs.
在随机对照试验（RCT）中，大量镇痛药未能证明其优于安慰剂，鉴于这与安慰剂反应增加有关，目前人们对确定安慰剂反应的预测因素很感兴趣。关于安慰剂机制的文献表明，与患者对治疗效果的期望相关的因素对安慰剂反应至关重要。此外，有研究表明RCT的一般特征会影响安慰剂反应。然而，只有少数荟萃分析直接检验了这些假设。我们纳入了2017名患有慢性疼痛性骨关节炎（髋部或膝部）或腰痛的成年患者参与的9项由行业赞助的随机、双盲、安慰剂对照、多中心III期试验的安慰剂数据。主要结局是疼痛强度。基于先前的研究，我们选择了3个与期望相关的主要预测因素：活性药物类型、随机化比例和计划的面对面访视次数。此外，探索性分析检验了RCT和患者特征是否能预测安慰剂反应。阿片类药物试验、大量计划的面对面访视以及随机化比例预测了安慰剂反应的大小，从而支持了期望假设。包含基线疼痛强度、年龄、洗脱期长度以及因不良事件停药情况的探索性模型解释了安慰剂反应中约10%的变异性。基于这些结果和先前的机制研究，我们认为患者对治疗分配的认知以及对治疗效果的期望能够有力地预测RCT的结果。
