Interventions for treatment of COVID-19: A living systematic review with meta-analyses and trial sequential analyses (The LIVING Project)
Interventions for treatment of COVID-19: A living systematic review with meta-analyses and trial sequential analyses (The LIVING Project)
复制标题
DOI：
10.1371/journal.pmed.1003293
复制
发表时间：
2020-09-01
期刊：
影响因子：
15.8
通讯作者：
Jakobsen, Janus Christian
中科院分区：
文献类型：
--
作者：
Juul, Sophie;Nielsen, Emil Eik;Jakobsen, Janus Christian
作者摘要 为什么要做这项研究？严重急性呼吸综合征冠状病毒 2 (SARS-CoV-2) 感染已在全球范围内迅速传播，导致 2019 年冠状病毒病 (COVID-19) 在国际范围内暴发。需要进行实时系统评价，评估所有可能的治疗 COVID-19 干预措施的有益和有害影响。研究人员做了什么并发现了什么？我们通过荟萃分析和试验序贯分析进行了第一版实时系统评价，以比较所有治疗干预措施对 COVID-19 的效果。一项对 6,425 名患者进行随机分组的试验显示，与标准护理相比，地塞米松对全因死亡率和机械通气具有有益效果。两项试验的荟萃分析显示，瑞德西韦与安慰剂相比对严重不良事件具有有益作用，但我们发现对全因死亡率或不严重的不良事件没有影响。对 6 项试验的荟萃分析显示，有证据表明羟氯喹对不严重的不良事件有有害影响，但似乎对全因死亡率或严重不良事件没有任何影响。五项单独试验显示出具有统计学意义的结果，但不足以证实或拒绝实际的干预效果。其余的试验都没有证据表明实验干预措施对我们预定的结果有差异。这些发现意味着什么？目前尚无针对 COVID-19 患者的特定循证治疗方法。地塞米松可以降低 COVID-19 患者的全因死亡率和机械通气需求。瑞德西韦可能会减少 COVID-19 患者的严重不良事件，但临床效果和最佳治疗持续时间需要确认。迫切需要更多高质量、低偏倚风险的随机临床试验。 背景 2019 年冠状病毒病（COVID-19）是一种迅速传播的疾病，给个人、家庭、国家和世界造成了广泛的负担。迫切需要针对 COVID-19 的有效治疗。方法和结果 这是第一版随机临床试验的实时系统评价，比较了所有治疗干预措施对所有年龄组的 COVID-19 参与者的效果。我们计划进行汇总数据荟萃分析、试验序贯分析、网络荟萃分析和个体患者数据荟萃分析。我们的系统评价基于系统评价和荟萃分析的首选报告项目 (PRISMA) 和 Cochrane 指南，以及我们的 8 步程序，以更好地验证荟萃分析结果的临床意义。我们进行了固定效应和随机效应荟萃分析。主要结局是全因死亡率和严重不良事件。次要结局是入住重症监护、机械通气、肾脏替代治疗、生活质量和非严重不良事件。我们使用建议评估、制定和评估分级 (GRADE) 来评估证据的确定性。我们检索了相关数据库和网站，查找截至2020年8月7日已发表和未发表的试验。两名评审员独立提取数据并评估试验方法。我们纳入了 33 项随机临床试验，总共招募了 13,312 名参与者。所有试验总体上都存在较高的偏倚风险。我们确定了一项试验，将 6,425 名参与者随机分为地塞米松组和标准治疗组。该试验显示地塞米松对全因死亡率（比率 0.83；95% 置信区间 [CI] 0.75-0.93；p< 0.001；低确定性）和机械通气（风险比 [RR] 0.77；95% CI 0.62-0.95；p= 0.021；低确定性）具有有益作用。可以对 10 项比较进行荟萃分析。荟萃分析显示，没有证据表明瑞德西韦与安慰剂在全因死亡率（RR 0.74；95% CI 0.40-1.37；p= 0.34，I-2= 58%；2 项试验；极低确定性）或非严重不良事件（RR 0.94；95% CI 0.80-1.11；p= 0.48，I-2= 29%；2 次试验；低确定性）。荟萃分析显示，有证据表明，瑞德西韦与安慰剂相比，对严重不良事件（RR 0.77；95% CI 0.63-0.94；p= 0.009，I-2= 0%；2 项试验；极低确定性）具有有益作用，其中一项试验主要由呼吸衰竭引起。荟萃分析和试验序贯分析表明，我们可以排除羟氯喹与标准治疗相比可降低全因死亡率（RR 1.07；95% CI 0.97-1.19；p= 0.17；I-2= 0%；7 项试验；低确定性）和严重不良事件（RR 1.07；95% CI 0.96-1.18；p= 0.21；p= 0.21）风险的可能性。 I-2= 0%；7 项试验；低确定性）20% 或更多，荟萃分析显示对非严重不良事件有有害影响（RR 2.40；95% CI 2.01-2.87；p< 0.00001；I-2= 90%；6 项试验；非常低确定性）。荟萃分析显示，没有证据表明洛匹那韦-利托那韦与标准治疗在严重不良事件（RR 0.64；95% CI 0.39-1.04；p= 0.07，I-2= 0%；2 项试验；极低确定性）或非严重不良事件（RR 1.14；95% CI 0.85-1.53；p= 0.38，I-2= 75%；2 次试验；非常低的确定性）。荟萃分析显示，没有证据表明恢复期血浆与标准治疗在全因死亡率方面存在差异（RR 0.60；95% CI 0.33-1.10；p= 0.10，I-2= 0%；2 项试验；极低确定性）。五项单独试验显示出具有统计学意义的结果，但不足以证实或拒绝实际的干预效果。其余试验均未显示出与我们预定义结果存在差异的证据。由于缺乏相关数据，无法进行其他荟萃分析、网络荟萃分析或个体患者数据荟萃分析。这种实时回顾的主要限制是目前可用的数据很少。此外，纳入的试验均存在系统误差和随机误差的风险。结论 我们的结果表明地塞米松和瑞德西韦可能对 COVID-19 患者有益，但证据的质量低至极低，因此需要更多试验。我们可以排除羟氯喹与标准治疗相比将死亡和严重不良事件风险降低 20% 或更多的可能性。除此之外，目前尚无针对 COVID-19 的循证治疗方法。此次审查将持续为 COVID-19 治疗和临床研究的最佳实践提供信息。
Author summaryWhy was this study done? Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection has spread rapidly worldwide, causing an international outbreak of the coronavirus disease 2019 (COVID-19). There is a need for a living systematic review evaluating the beneficial and harmful effects of all possible interventions for treatment of COVID-19. What did the researchers do and find? We conducted the first edition of a living systematic review with meta-analyses and trial sequential analyses to compare the effects of all treatment interventions for COVID-19. One single trial randomizing 6,425 patients showed evidence of a beneficial effect of dexamethasone versus standard care on all-cause mortality and mechanical ventilation. Meta-analysis of 2 trials showed evidence of a beneficial effect of remdesivir versus placebo on serious adverse events, but we found no effects on all-cause mortality or adverse events considered not serious. Meta-analysis of 6 trials showed evidence of a harmful effect of hydroxychloroquine on adverse events considered nonserious and did not seem to have any effect on all-cause mortality or serious adverse events. Five single trials showed statistically significant results but were underpowered to confirm or reject realistic intervention effects. None of the remaining trials showed evidence of a difference of the experimental interventions on our predefined outcomes. What do these findings mean? No certain evidence-based treatment is currently available for COVID-19 patients. Dexamethasone may reduce all-cause mortality and need for mechanical ventilation in COVID-19 patients. Remdesivir may reduce serious adverse events in COVID-19 patients, but the clinical effects and the optimal treatment duration need confirmation. More high-quality, low risk of bias randomized clinical trials are urgently needed.Background Coronavirus disease 2019 (COVID-19) is a rapidly spreading disease that has caused extensive burden to individuals, families, countries, and the world. Effective treatments of COVID-19 are urgently needed. Methods and findings This is the first edition of a living systematic review of randomized clinical trials comparing the effects of all treatment interventions for participants in all age groups with COVID-19. We planned to conduct aggregate data meta-analyses, trial sequential analyses, network meta-analysis, and individual patient data meta-analyses. Our systematic review is based on Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) and Cochrane guidelines, and our 8-step procedure for better validation of clinical significance of meta-analysis results. We performed both fixed-effect and random-effects meta-analyses. Primary outcomes were all-cause mortality and serious adverse events. Secondary outcomes were admission to intensive care, mechanical ventilation, renal replacement therapy, quality of life, and nonserious adverse events. We used Grading of Recommendations Assessment, Development and Evaluation (GRADE) to assess the certainty of evidence. We searched relevant databases and websites for published and unpublished trials until August 7, 2020. Two reviewers independently extracted data and assessed trial methodology. We included 33 randomized clinical trials enrolling a total of 13,312 participants. All trials were at overall high risk of bias. We identified one trial randomizing 6,425 participants to dexamethasone versus standard care. This trial showed evidence of a beneficial effect of dexamethasone on all-cause mortality (rate ratio 0.83; 95% confidence interval [CI] 0.75-0.93;p< 0.001; low certainty) and on mechanical ventilation (risk ratio [RR] 0.77; 95% CI 0.62-0.95;p= 0.021; low certainty). It was possible to perform meta-analysis of 10 comparisons. Meta-analysis showed no evidence of a difference between remdesivir versus placebo on all-cause mortality (RR 0.74; 95% CI 0.40-1.37;p= 0.34, I-2= 58%; 2 trials; very low certainty) or nonserious adverse events (RR 0.94; 95% CI 0.80-1.11;p= 0.48, I-2= 29%; 2 trials; low certainty). Meta-analysis showed evidence of a beneficial effect of remdesivir versus placebo on serious adverse events (RR 0.77; 95% CI 0.63-0.94;p= 0.009, I-2= 0%; 2 trials; very low certainty) mainly driven by respiratory failure in one trial. Meta-analyses and trial sequential analyses showed that we could exclude the possibility that hydroxychloroquine versus standard care reduced the risk of all-cause mortality (RR 1.07; 95% CI 0.97-1.19;p= 0.17; I-2= 0%; 7 trials; low certainty) and serious adverse events (RR 1.07; 95% CI 0.96-1.18;p= 0.21; I-2= 0%; 7 trials; low certainty) by 20% or more, and meta-analysis showed evidence of a harmful effect on nonserious adverse events (RR 2.40; 95% CI 2.01-2.87;p< 0.00001; I-2= 90%; 6 trials; very low certainty). Meta-analysis showed no evidence of a difference between lopinavir-ritonavir versus standard care on serious adverse events (RR 0.64; 95% CI 0.39-1.04;p= 0.07, I-2= 0%; 2 trials; very low certainty) or nonserious adverse events (RR 1.14; 95% CI 0.85-1.53;p= 0.38, I-2= 75%; 2 trials; very low certainty). Meta-analysis showed no evidence of a difference between convalescent plasma versus standard care on all-cause mortality (RR 0.60; 95% CI 0.33-1.10;p= 0.10, I-2= 0%; 2 trials; very low certainty). Five single trials showed statistically significant results but were underpowered to confirm or reject realistic intervention effects. None of the remaining trials showed evidence of a difference on our predefined outcomes. Because of the lack of relevant data, it was not possible to perform other meta-analyses, network meta-analysis, or individual patient data meta-analyses. The main limitation of this living review is the paucity of data currently available. Furthermore, the included trials were all at risks of systematic errors and random errors. Conclusions Our results show that dexamethasone and remdesivir might be beneficial for COVID-19 patients, but the certainty of the evidence was low to very low, so more trials are needed. We can exclude the possibility of hydroxychloroquine versus standard care reducing the risk of death and serious adverse events by 20% or more. Otherwise, no evidence-based treatment for COVID-19 currently exists. This review will continuously inform best practice in treatment and clinical research of COVID-19.