PDF(Pubmed)
Abstract：
BACKGROUND: Migraine affects 1.1 billion people globally and is the second leading cause of disability worldwide. In clinical trials, treatment efficacy is evaluated by comparing the differential responses in the treatment and placebo arms. Although placebo response in preventive migraine trials has been studied, there is limited research examining temporal trends. This study evaluates the trend of placebo response over thirty years in migraine prevention trials and investigates the association of potential confounders, such as patient, treatment, and study characteristics on placebo response using meta-analysis with regression.
METHODS: We conducted literature searches from January 1990 to August 2021 in bibliographical databases (PubMed, Cochrane Library, and EMBASE). Studies were selected according to PICOS criteria and included randomized, double-blind, placebo-controlled trials evaluating preventive migraine treatments in adult patients diagnosed with episodic or chronic migraine, with or without aura. The protocol was registered with PROSPERO (CRD42021271732). Migraine efficacy outcomes included were either continuous (e.g., monthly migraine days) or dichotomous (e.g., ≥ 50% responder rate (yes/no)). We assessed the correlation of the change in outcome from baseline in the placebo arm, with the year of publication. The relationship between placebo response and year of publication was also assessed after accounting to confounders.
RESULTS: A total of 907 studies were identified, and 83 were found eligible. For the continuous outcomes, the change from baseline in mean placebo response showed an increase over the years (rho = 0.32, p = 0.006). The multivariable regression analysis also showed an overall increase in placebo response over the years. The correlation analysis of dichotomous responses showed no significant linear trend between publication year and mean placebo response (rho = 0.08, p = 0.596). Placebo response also varied by route of administration.
CONCLUSIONS: Placebo response increased over the past 30 years in migraine preventive trials. This phenomenon should be considered when designing clinical trials and conducting meta-analyses.
METHODS: We conducted literature searches from January 1990 to August 2021 in bibliographical databases (PubMed, Cochrane Library, and EMBASE). Studies were selected according to PICOS criteria and included randomized, double-blind, placebo-controlled trials evaluating preventive migraine treatments in adult patients diagnosed with episodic or chronic migraine, with or without aura. The protocol was registered with PROSPERO (CRD42021271732). Migraine efficacy outcomes included were either continuous (e.g., monthly migraine days) or dichotomous (e.g., ≥ 50% responder rate (yes/no)). We assessed the correlation of the change in outcome from baseline in the placebo arm, with the year of publication. The relationship between placebo response and year of publication was also assessed after accounting to confounders.
RESULTS: A total of 907 studies were identified, and 83 were found eligible. For the continuous outcomes, the change from baseline in mean placebo response showed an increase over the years (rho = 0.32, p = 0.006). The multivariable regression analysis also showed an overall increase in placebo response over the years. The correlation analysis of dichotomous responses showed no significant linear trend between publication year and mean placebo response (rho = 0.08, p = 0.596). Placebo response also varied by route of administration.
CONCLUSIONS: Placebo response increased over the past 30 years in migraine preventive trials. This phenomenon should be considered when designing clinical trials and conducting meta-analyses.
摘要：
背景：偏头痛影响全球11亿人，是全球残疾的第二大原因。在临床试验中,通过比较治疗组和安慰剂组的差异反应来评估治疗功效。尽管已经研究了预防性偏头痛试验中的安慰剂反应，研究时间趋势的研究有限。这项研究评估了30年来偏头痛预防试验中安慰剂反应的趋势，并调查了潜在混杂因素的关联。如病人，治疗,并使用回归荟萃分析研究安慰剂反应的特征。
方法：我们从1990年1月到2021年8月在书目数据库中进行了文献检索(PubMed，科克伦图书馆,和EMBASE）。根据PICOS标准选择研究，并包括随机，双盲,安慰剂对照试验评估诊断为发作性或慢性偏头痛的成年患者的预防性偏头痛治疗，有或没有光环。该协议在PROSPERO（CRD42021271732）注册。包括偏头痛疗效结果是连续的(例如，每月偏头痛日)或二分法(例如，≥50%应答率(是/否))。我们评估了安慰剂组的基线结果变化的相关性，与出版之年。在考虑混杂因素后，还评估了安慰剂反应与发表年份之间的关系。
结果：共确定907项研究，83人符合条件。对于持续的结果，平均安慰剂反应较基线的变化显示多年来增加(rho=0.32,p=0.006).多变量回归分析还显示了多年来安慰剂反应的总体增加。二分反应的相关性分析显示，在发表年和平均安慰剂反应之间没有显着的线性趋势（rho=0.08，p=0.596）。安慰剂反应也因施用途径而异。
结论：在过去的30年中，偏头痛预防性试验中安慰剂反应增加。在设计临床试验和进行荟萃分析时应考虑这种现象。
方法：我们从1990年1月到2021年8月在书目数据库中进行了文献检索(PubMed，科克伦图书馆,和EMBASE）。根据PICOS标准选择研究，并包括随机，双盲,安慰剂对照试验评估诊断为发作性或慢性偏头痛的成年患者的预防性偏头痛治疗，有或没有光环。该协议在PROSPERO（CRD42021271732）注册。包括偏头痛疗效结果是连续的(例如，每月偏头痛日)或二分法(例如，≥50%应答率(是/否))。我们评估了安慰剂组的基线结果变化的相关性，与出版之年。在考虑混杂因素后，还评估了安慰剂反应与发表年份之间的关系。
结果：共确定907项研究，83人符合条件。对于持续的结果，平均安慰剂反应较基线的变化显示多年来增加(rho=0.32,p=0.006).多变量回归分析还显示了多年来安慰剂反应的总体增加。二分反应的相关性分析显示，在发表年和平均安慰剂反应之间没有显着的线性趋势（rho=0.08，p=0.596）。安慰剂反应也因施用途径而异。
结论：在过去的30年中，偏头痛预防性试验中安慰剂反应增加。在设计临床试验和进行荟萃分析时应考虑这种现象。
