Chenyang Zhang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Chenyang Zhang has a total of 6 co-authors. Here are the top co-authors by number of shared publications:
- Guosheng Yin
Department of Statistics and Actuarial Science, The University of Hong Kong
7 shared publications - Huaqing Jin
Department of Statistics and Data Science, Tsinghua University , Qinghuayuan , Beijing 100084
4 shared publications Yuanshan Wu
School of Statistics and Mathematics, Zhongnan University of Economics and Law
2 shared publicationsGuosheng Yin
Department of Statistics and Actuarial Science, University of Hong Kong
1 shared publicationJiaqi Gu
Department of Mathematics and Statistics, University of South Florida
1 shared publicationYifeng Wen
Key Laboratory of Shaanxi Province for Craniofacial Precision Medicine Research, College of Stomatology, Xi’an Jiaotong University, Xi’an, China
1 shared publication
Latest Publications
- 1 December 2025
- Published by Institute of Mathematical Statistics in The Annals of Applied Statistics
- Vol. 19 (4) , 3023-3044
Normalized area above quantile: An alternative measure to restricted mean survival time
The restricted mean survival time (RMST), which quantifies the average survival time within an investigational time interval, has been widely used to evaluate treatment effects, especially when comparing treatment survival profiles across treatments where the proportional hazards assumption does not hold. However, in the ENACT trial, we observed significant differences in follow-up periods across treatment groups (34 months vs. 20 months), and analysis relying on the RMST difference could result in a substantial loss of useful survival information. As an alternative, we propose integrating the survival curve along the quantile axis instead of the time axis to retain more information, which leads to a new clinical measure called the normalized area above quantile (NAAQ). The NAAQ discards survival information below a chosen quantile τ and assesses the conditional expected survival time among patients with 100(1−τ)% shortest survival time (corresponding to the worst (1−τ) survivors). For right-censored survival data, we propose a nonparametric NAAQ estimator from the Kaplan–Meier curve and prove its asymptotic properties under mild conditions. Simulation studies demonstrate the promising performance of the NAAQ estimator and the NAAQ-based test, which are further applied to two real clinical trials to illustrate the broad practical value of NAAQ.
- 1 December 2025
- Published by Institute of Mathematical Statistics in The Annals of Applied Statistics
- Vol. 19 (4) , 3023-3044
Normalized area above quantile: An alternative measure to restricted mean survival time
The restricted mean survival time (RMST), which quantifies the average survival time within an investigational time interval, has been widely used to evaluate treatment effects, especially when comparing treatment survival profiles across treatments where the proportional hazards assumption does not hold. However, in the ENACT trial, we observed significant differences in follow-up periods across treatment groups (34 months vs. 20 months), and analysis relying on the RMST difference could result in a substantial loss of useful survival information. As an alternative, we propose integrating the survival curve along the quantile axis instead of the time axis to retain more information, which leads to a new clinical measure called the normalized area above quantile (NAAQ). The NAAQ discards survival information below a chosen quantile τ and assesses the conditional expected survival time among patients with 100(1−τ)% shortest survival time (corresponding to the worst (1−τ) survivors). For right-censored survival data, we propose a nonparametric NAAQ estimator from the Kaplan–Meier curve and prove its asymptotic properties under mild conditions. Simulation studies demonstrate the promising performance of the NAAQ estimator and the NAAQ-based test, which are further applied to two real clinical trials to illustrate the broad practical value of NAAQ.
- Open Access
- 13 October 2022
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 17 (10) , e0275888
Safety of paclitaxel-coated devices in the femoropopliteal arteries: A systematic review and meta-analysis
Background: Clinical benefit of paclitaxel-coated devices for patients with peripheral arterial disease has been confirmed in randomized controlled trials (RCTs). A meta-analysis published in 2018 identified late mortality risk over a long follow-up period due to use of paclitaxel-coated devices in the femoropopliteal arteries, which caused enormous controversy and debates globally. This study aims to further evaluate the safety of paclitaxel-coated devices by incorporating the most recently published data. Methods: We searched for candidate studies in PubMed (MEDLINE), Scopus, EMBASE (Ovid) online databases, government web archives and international cardiovascular conferences. Safety endpoints of interest included all-cause mortality rates at one, two and five years and the risk ratio (RR) was used as the summary measure. The primary analysis was performed using random-effects models to account for potential clinical heterogeneity. Findings: Thirty-nine RCTs including 9164 patients were identified. At one year, the random-effects model yielded a pooled RR of 1.06 (95% CI [0.87, 1.29]) indicating no difference in short-term all-cause deaths between the paclitaxel and control groups (crude mortality, 4.3%, 214/5025 versus 4.5%, 177/3965). Two-year mortality was reported in 26 RCTs with 382 deaths out of 3788 patients (10.1%) in the paclitaxel arm and 299 out of 2955 patients (10.1%) in the control arm and no association was found between increased risk of death and usage of paclitaxel-coated devices (RR 1.08, 95% CI [0.93, 1.25]). Eight RCTs recorded all-cause deaths up to five years and a pooled RR of 1.18 (95% CI [0.92, 1.51]) demonstrated no late mortality risk due to use of paclitaxel-coated devices (crude mortality, paclitaxel 18.2%, 247/1360 versus control 15.2%, 122/805). Conclusions: We found no significant difference in either short- or long-term all-cause mortalities between patients receiving paclitaxel-coated and uncoated devices. Further research on the longer-term safety of paclitaxel usage (e.g., 8- or 10-year) is warranted. Registration: PROSPERO, CRD42021246291.
- 13 October 2022
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 17 (10) , e0275888
Open AccessSafety of paclitaxel-coated devices in the femoropopliteal arteries: A systematic review and meta-analysis
Background: Clinical benefit of paclitaxel-coated devices for patients with peripheral arterial disease has been confirmed in randomized controlled trials (RCTs). A meta-analysis published in 2018 identified late mortality risk over a long follow-up period due to use of paclitaxel-coated devices in the femoropopliteal arteries, which caused enormous controversy and debates globally. This study aims to further evaluate the safety of paclitaxel-coated devices by incorporating the most recently published data. Methods: We searched for candidate studies in PubMed (MEDLINE), Scopus, EMBASE (Ovid) online databases, government web archives and international cardiovascular conferences. Safety endpoints of interest included all-cause mortality rates at one, two and five years and the risk ratio (RR) was used as the summary measure. The primary analysis was performed using random-effects models to account for potential clinical heterogeneity. Findings: Thirty-nine RCTs including 9164 patients were identified. At one year, the random-effects model yielded a pooled RR of 1.06 (95% CI [0.87, 1.29]) indicating no difference in short-term all-cause deaths between the paclitaxel and control groups (crude mortality, 4.3%, 214/5025 versus 4.5%, 177/3965). Two-year mortality was reported in 26 RCTs with 382 deaths out of 3788 patients (10.1%) in the paclitaxel arm and 299 out of 2955 patients (10.1%) in the control arm and no association was found between increased risk of death and usage of paclitaxel-coated devices (RR 1.08, 95% CI [0.93, 1.25]). Eight RCTs recorded all-cause deaths up to five years and a pooled RR of 1.18 (95% CI [0.92, 1.51]) demonstrated no late mortality risk due to use of paclitaxel-coated devices (crude mortality, paclitaxel 18.2%, 247/1360 versus control 15.2%, 122/805). Conclusions: We found no significant difference in either short- or long-term all-cause mortalities between patients receiving paclitaxel-coated and uncoated devices. Further research on the longer-term safety of paclitaxel usage (e.g., 8- or 10-year) is warranted. Registration: PROSPERO, CRD42021246291.
- 16 January 2022
- Published by Oxford University Press (OUP) in Biometrics
- Vol. 79 (2) , 1383-1396
Bayesian Nonparametric Analysis of Restricted Mean Survival Time
The restricted mean survival time (RMST) evaluates the expectation of survival time truncated by a prespecified time point, because the mean survival time in the presence of censoring is typically not estimable. The frequentist inference procedure for RMST has been widely advocated for comparison of two survival curves, while research from the Bayesian perspective is rather limited. For the RMST of both right- and interval-censored data, we propose Bayesian nonparametric estimation and inference procedures. By assigning a mixture of Dirichlet processes (MDP) prior to the distribution function, we can estimate the posterior distribution of RMST. We also explore another Bayesian nonparametric approach using the Dirichlet process mixture model and make comparisons with the frequentist nonparametric method. Simulation studies demonstrate that the Bayesian nonparametric RMST under diffuse MDP priors leads to robust estimation and under informative priors it can incorporate prior knowledge into the nonparametric estimator. Analysis of real trial examples demonstrates the flexibility and interpretability of the Bayesian nonparametric RMST for both right- and interval-censored data.
- 16 January 2022
- Published by Oxford University Press (OUP) in Biometrics
- Vol. 79 (2) , 1383-1396
Bayesian Nonparametric Analysis of Restricted Mean Survival Time
The restricted mean survival time (RMST) evaluates the expectation of survival time truncated by a prespecified time point, because the mean survival time in the presence of censoring is typically not estimable. The frequentist inference procedure for RMST has been widely advocated for comparison of two survival curves, while research from the Bayesian perspective is rather limited. For the RMST of both right- and interval-censored data, we propose Bayesian nonparametric estimation and inference procedures. By assigning a mixture of Dirichlet processes (MDP) prior to the distribution function, we can estimate the posterior distribution of RMST. We also explore another Bayesian nonparametric approach using the Dirichlet process mixture model and make comparisons with the frequentist nonparametric method. Simulation studies demonstrate that the Bayesian nonparametric RMST under diffuse MDP priors leads to robust estimation and under informative priors it can incorporate prior knowledge into the nonparametric estimator. Analysis of real trial examples demonstrates the flexibility and interpretability of the Bayesian nonparametric RMST for both right- and interval-censored data.
- Open Access
A Systematic Review and Network Meta-Analysis for COVID-19 Treatments
Background Numerous interventions for coronavirus disease 2019 (COVID-19) have been investigated by randomized controlled trials (RCTs). This systematic review and Bayesian network meta-analysis (NMA) aim to provide a comprehensive evaluation of efficacy of available treatments for COVID-19. Methods We searched for candidate COVID-19 studies in WHO COVID-19 Global Research Database, PubMed, PubMed Central, LitCovid, Proquest Central and Ovid up to December 19, 2020. RCTs for suspected or confirmed COVID-19 patients were included, regardless of publication status or demographic characteristics. Bayesian NMA with fixed effects was conducted to estimate the effect sizes using posterior means and 95% equal-tailed credible intervals (CrIs), while that with random effects was carried out as well for sensitivity analysis. Bayesian hierarchical models were used to estimate effect sizes of treatments grouped by their drug classifications. Results We identified 96 eligible RCTs with a total of 51187 patients. Compared with the standard of care (SOC), this NMA showed that dexamethasone led to lower risk of mortality with an odds ratio (OR) of 0.85 (95% CrI [0.76, 0.95]; moderate certainty) and lower risk of mechanical ventilation (MV) with an OR of 0.68 (95% CrI [0.56, 0.83]; low certainty). For hospital discharge, remdesivir (OR 1.37, 95% CrI [1.15, 1.64]; moderate certainty), dexamethasone (OR 1.20, 95% CrI [1.08, 1.34]; low certainty), interferon beta (OR 2.15, 95% CrI [1.26, 3.74]; moderate certainty), tocilizumab (OR 1.40, 95% CrI [1.05, 1.89]; moderate certainty) and baricitinib plus remdesivir (OR 1.75, 95% CrI [1.28, 2.39]; moderate certainty) could all increase the discharge rate respectively. Recombinant human granulocyte colony-stimulating factor indicated lower risk of MV (OR 0.20, 95% CrI [0.10, 0.40]; moderate certainty); and patients receiving convalescent plasma resulted in better viral clearance (OR 2.28, 95% CrI [1.57, 3.34]; low certainty). About two-thirds of the studies included in this NMA were rated as high risk of bias, and the certainty of evidence was either low or very low for most of the comparisons. Conclusion The Bayesian NMA identified superiority of several COVID-19 treatments over SOC in terms of mortality, requirement of MV, hospital discharge and viral clearance. These results provide a comprehensive comparison of current COVID-19 treatments and shed new light on further research and discovery of potential COVID-19 treatments.
Open AccessA Systematic Review and Network Meta-Analysis for COVID-19 Treatments
Background Numerous interventions for coronavirus disease 2019 (COVID-19) have been investigated by randomized controlled trials (RCTs). This systematic review and Bayesian network meta-analysis (NMA) aim to provide a comprehensive evaluation of efficacy of available treatments for COVID-19. Methods We searched for candidate COVID-19 studies in WHO COVID-19 Global Research Database, PubMed, PubMed Central, LitCovid, Proquest Central and Ovid up to December 19, 2020. RCTs for suspected or confirmed COVID-19 patients were included, regardless of publication status or demographic characteristics. Bayesian NMA with fixed effects was conducted to estimate the effect sizes using posterior means and 95% equal-tailed credible intervals (CrIs), while that with random effects was carried out as well for sensitivity analysis. Bayesian hierarchical models were used to estimate effect sizes of treatments grouped by their drug classifications. Results We identified 96 eligible RCTs with a total of 51187 patients. Compared with the standard of care (SOC), this NMA showed that dexamethasone led to lower risk of mortality with an odds ratio (OR) of 0.85 (95% CrI [0.76, 0.95]; moderate certainty) and lower risk of mechanical ventilation (MV) with an OR of 0.68 (95% CrI [0.56, 0.83]; low certainty). For hospital discharge, remdesivir (OR 1.37, 95% CrI [1.15, 1.64]; moderate certainty), dexamethasone (OR 1.20, 95% CrI [1.08, 1.34]; low certainty), interferon beta (OR 2.15, 95% CrI [1.26, 3.74]; moderate certainty), tocilizumab (OR 1.40, 95% CrI [1.05, 1.89]; moderate certainty) and baricitinib plus remdesivir (OR 1.75, 95% CrI [1.28, 2.39]; moderate certainty) could all increase the discharge rate respectively. Recombinant human granulocyte colony-stimulating factor indicated lower risk of MV (OR 0.20, 95% CrI [0.10, 0.40]; moderate certainty); and patients receiving convalescent plasma resulted in better viral clearance (OR 2.28, 95% CrI [1.57, 3.34]; low certainty). About two-thirds of the studies included in this NMA were rated as high risk of bias, and the certainty of evidence was either low or very low for most of the comparisons. Conclusion The Bayesian NMA identified superiority of several COVID-19 treatments over SOC in terms of mortality, requirement of MV, hospital discharge and viral clearance. These results provide a comprehensive comparison of current COVID-19 treatments and shed new light on further research and discovery of potential COVID-19 treatments.
- 7 August 2020
- Published by Wiley in Statistics in Medicine
- Vol. 39 (26) , 3879-3895
Restricted mean survival time for interval‐censored data
Restricted mean survival time (RMST) evaluates the mean event‐free survival time up to a prespecified time point. It has been used as an alternative measure of treatment effect owing to its model‐free structure and clinically meaningful interpretation of treatment benefit for right‐censored data. In clinical trials, another type of censoring called interval censoring may occur if subjects are examined at several discrete time points and the survival time falls into an interval rather than being exactly observed. The missingness of exact observations under interval‐censored cases makes the nonparametric measure of treatment effect more challenging. Employing the linear smoothing technique to overcome the ambiguity, we propose a new model‐free measure for the interval‐censored RMST. As an alternative to the commonly used log‐rank test, we further construct a hypothesis testing procedure to assess the survival difference between two groups. Simulation studies show that the bias of our proposed interval‐censored RMST estimator is negligible and the testing procedure delivers promising performance in detecting between‐group difference with regard to size and power under various configurations of survival curves. The proposed method is illustrated by reanalyzing two real datasets containing interval‐censored observations.
- 7 August 2020
- Published by Wiley in Statistics in Medicine
- Vol. 39 (26) , 3879-3895
Restricted mean survival time for interval‐censored data
Restricted mean survival time (RMST) evaluates the mean event‐free survival time up to a prespecified time point. It has been used as an alternative measure of treatment effect owing to its model‐free structure and clinically meaningful interpretation of treatment benefit for right‐censored data. In clinical trials, another type of censoring called interval censoring may occur if subjects are examined at several discrete time points and the survival time falls into an interval rather than being exactly observed. The missingness of exact observations under interval‐censored cases makes the nonparametric measure of treatment effect more challenging. Employing the linear smoothing technique to overcome the ambiguity, we propose a new model‐free measure for the interval‐censored RMST. As an alternative to the commonly used log‐rank test, we further construct a hypothesis testing procedure to assess the survival difference between two groups. Simulation studies show that the bias of our proposed interval‐censored RMST estimator is negligible and the testing procedure delivers promising performance in detecting between‐group difference with regard to size and power under various configurations of survival curves. The proposed method is illustrated by reanalyzing two real datasets containing interval‐censored observations.
- Open Access
- 21 July 2020
- Published by JMIR Publications Inc. in JMIR Preprints
Current Status on Treatments and Clinical Trials for COVID-19 (Preprint)
UNSTRUCTURED With the worldwide rapidly growing number of patients infected with the novel coronavirus (COVID-19), the death toll has also been climbing up at a fast speed. There is an urgent need to search for cures for COVID-19 patients. A large number of clinical trials have been launched to test some existing or new antiviral therapeutics and vaccines. In contrast to starting from the scratch, many trials are initiated directly in phase II or III with the hope to expedite the developmental process. We summarize the information on the registered trials for the top ten COVID-19 drugs, and give an overview on the current situation and trend of treatments and clinical trials. In particular, we review those trials that have already been finished and discuss lessons that can be learned from them.
- 21 July 2020
- Published by JMIR Publications Inc. in JMIR Preprints
Open AccessCurrent Status on Treatments and Clinical Trials for COVID-19 (Preprint)
UNSTRUCTURED With the worldwide rapidly growing number of patients infected with the novel coronavirus (COVID-19), the death toll has also been climbing up at a fast speed. There is an urgent need to search for cures for COVID-19 patients. A large number of clinical trials have been launched to test some existing or new antiviral therapeutics and vaccines. In contrast to starting from the scratch, many trials are initiated directly in phase II or III with the hope to expedite the developmental process. We summarize the information on the registered trials for the top ten COVID-19 drugs, and give an overview on the current situation and trend of treatments and clinical trials. In particular, we review those trials that have already been finished and discuss lessons that can be learned from them.
- Open Access
- 10 July 2020
- Published by JMIR Publications Inc. in JMIR Public Health and Surveillance
- Vol. 6 (3) , e19538-241
Statistical Issues and Lessons Learned From COVID-19 Clinical Trials With Lopinavir-Ritonavir and Remdesivir
Background: Recently, three randomized clinical trials on coronavirus disease (COVID-19) treatments were completed: one for lopinavir-ritonavir and two for remdesivir. One trial reported that remdesivir was superior to placebo in shortening the time to recovery, while the other two showed no benefit of the treatment under investigation. Objective: The aim of this paper is to, from a statistical perspective, identify several key issues in the design and analysis of three COVID-19 trials and reanalyze the data from the cumulative incidence curves in the three trials using more appropriate statistical methods. Methods: The lopinavir-ritonavir trial enrolled 39 additional patients due to insignificant results after the sample size reached the planned number, which led to inflation of the type I error rate. The remdesivir trial of Wang et al failed to reach the planned sample size due to a lack of eligible patients, and the bootstrap method was used to predict the quantity of clinical interest conditionally and unconditionally if the trial had continued to reach the originally planned sample size. Moreover, we used a terminal (or cure) rate model and a model-free metric known as the restricted mean survival time or the restricted mean time to improvement (RMTI) to analyze the reconstructed data. The remdesivir trial of Beigel et al reported the median recovery time of the remdesivir and placebo groups, and the rate ratio for recovery, while both quantities depend on a particular time point representing local information. We use the restricted mean time to recovery (RMTR) as a global and robust measure for efficacy. Results: For the lopinavir-ritonavir trial, with the increase of sample size from 160 to 199, the type I error rate was inflated from 0.05 to 0.071. The difference of RMTIs between the two groups evaluated at day 28 was –1.67 days (95% CI –3.62 to 0.28; P=.09) in favor of lopinavir-ritonavir but not statistically significant. For the remdesivir trial of Wang et al, the difference of RMTIs at day 28 was –0.89 days (95% CI –2.84 to 1.06; P=.37). The planned sample size was 453, yet only 236 patients were enrolled. The conditional prediction shows that the hazard ratio estimates would reach statistical significance if the target sample size had been maintained. For the remdesivir trial of Beigel et al, the difference of RMTRs between the remdesivir and placebo groups at day 30 was –2.7 days (95% CI –4.0 to –1.2; P<.001), confirming the superiority of remdesivir. The difference in the recovery time at the 25th percentile (95% CI –3 to 0; P=.65) was insignificant, while the differences became more statistically significant at larger percentiles. Conclusions: Based on the statistical issues and lessons learned from the recent three clinical trials on COVID-19 treatments, we suggest more appropriate approaches for the design and analysis of ongoing and future COVID-19 trials.
- 10 July 2020
- Published by JMIR Publications Inc. in JMIR Public Health and Surveillance
- Vol. 6 (3) , e19538-241
Open AccessStatistical Issues and Lessons Learned From COVID-19 Clinical Trials With Lopinavir-Ritonavir and Remdesivir
Background: Recently, three randomized clinical trials on coronavirus disease (COVID-19) treatments were completed: one for lopinavir-ritonavir and two for remdesivir. One trial reported that remdesivir was superior to placebo in shortening the time to recovery, while the other two showed no benefit of the treatment under investigation. Objective: The aim of this paper is to, from a statistical perspective, identify several key issues in the design and analysis of three COVID-19 trials and reanalyze the data from the cumulative incidence curves in the three trials using more appropriate statistical methods. Methods: The lopinavir-ritonavir trial enrolled 39 additional patients due to insignificant results after the sample size reached the planned number, which led to inflation of the type I error rate. The remdesivir trial of Wang et al failed to reach the planned sample size due to a lack of eligible patients, and the bootstrap method was used to predict the quantity of clinical interest conditionally and unconditionally if the trial had continued to reach the originally planned sample size. Moreover, we used a terminal (or cure) rate model and a model-free metric known as the restricted mean survival time or the restricted mean time to improvement (RMTI) to analyze the reconstructed data. The remdesivir trial of Beigel et al reported the median recovery time of the remdesivir and placebo groups, and the rate ratio for recovery, while both quantities depend on a particular time point representing local information. We use the restricted mean time to recovery (RMTR) as a global and robust measure for efficacy. Results: For the lopinavir-ritonavir trial, with the increase of sample size from 160 to 199, the type I error rate was inflated from 0.05 to 0.071. The difference of RMTIs between the two groups evaluated at day 28 was –1.67 days (95% CI –3.62 to 0.28; P=.09) in favor of lopinavir-ritonavir but not statistically significant. For the remdesivir trial of Wang et al, the difference of RMTIs at day 28 was –0.89 days (95% CI –2.84 to 1.06; P=.37). The planned sample size was 453, yet only 236 patients were enrolled. The conditional prediction shows that the hazard ratio estimates would reach statistical significance if the target sample size had been maintained. For the remdesivir trial of Beigel et al, the difference of RMTRs between the remdesivir and placebo groups at day 30 was –2.7 days (95% CI –4.0 to –1.2; P<.001), confirming the superiority of remdesivir. The difference in the recovery time at the 25th percentile (95% CI –3 to 0; P=.65) was insignificant, while the differences became more statistically significant at larger percentiles. Conclusions: Based on the statistical issues and lessons learned from the recent three clinical trials on COVID-19 treatments, we suggest more appropriate approaches for the design and analysis of ongoing and future COVID-19 trials.
- Open Access
- 22 April 2020
- Published by JMIR Publications Inc. in JMIR Preprints
Statistical Issues and Lessons Learned From COVID-19 Clinical Trials With Lopinavir-Ritonavir and Remdesivir (Preprint)
BACKGROUND Recently, three randomized clinical trials on coronavirus disease (COVID-19) treatments were completed: one for lopinavir-ritonavir and two for remdesivir. One trial reported that remdesivir was superior to placebo in shortening the time to recovery, while the other two showed no benefit of the treatment under investigation. OBJECTIVE The aim of this paper is to, from a statistical perspective, identify several key issues in the design and analysis of three COVID-19 trials and reanalyze the data from the cumulative incidence curves in the three trials using more appropriate statistical methods. METHODS The lopinavir-ritonavir trial enrolled 39 additional patients due to insignificant results after the sample size reached the planned number, which led to inflation of the type I error rate. The remdesivir trial of Wang et al failed to reach the planned sample size due to a lack of eligible patients, and the bootstrap method was used to predict the quantity of clinical interest conditionally and unconditionally if the trial had continued to reach the originally planned sample size. Moreover, we used a terminal (or cure) rate model and a model-free metric known as the restricted mean survival time or the restricted mean time to improvement (RMTI) to analyze the reconstructed data. The remdesivir trial of Beigel et al reported the median recovery time of the remdesivir and placebo groups, and the rate ratio for recovery, while both quantities depend on a particular time point representing local information. We use the restricted mean time to recovery (RMTR) as a global and robust measure for efficacy. RESULTS For the lopinavir-ritonavir trial, with the increase of sample size from 160 to 199, the type I error rate was inflated from 0.05 to 0.071. The difference of RMTIs between the two groups evaluated at day 28 was –1.67 days (95% CI –3.62 to 0.28; P=.09) in favor of lopinavir-ritonavir but not statistically significant. For the remdesivir trial of Wang et al, the difference of RMTIs at day 28 was –0.89 days (95% CI –2.84 to 1.06; P=.37). The planned sample size was 453, yet only 236 patients were enrolled. The conditional prediction shows that the hazard ratio estimates would reach statistical significance if the target sample size had been maintained. For the remdesivir trial of Beigel et al, the difference of RMTRs between the remdesivir and placebo groups at day 30 was –2.7 days (95% CI –4.0 to –1.2; P<.001), confirming the superiority of remdesivir. The difference in the recovery time at the 25th percentile (95% CI –3 to 0; P=.65) was insignificant, while the differences became more statistically significant at larger percentiles. CONCLUSIONS Based on the statistical issues and lessons learned from the recent three clinical trials on COVID-19 treatments, we suggest more appropriate approaches for the design and analysis of ongoing and future COVID-19 trials.
- 22 April 2020
- Published by JMIR Publications Inc. in JMIR Preprints
Open AccessStatistical Issues and Lessons Learned From COVID-19 Clinical Trials With Lopinavir-Ritonavir and Remdesivir (Preprint)
BACKGROUND Recently, three randomized clinical trials on coronavirus disease (COVID-19) treatments were completed: one for lopinavir-ritonavir and two for remdesivir. One trial reported that remdesivir was superior to placebo in shortening the time to recovery, while the other two showed no benefit of the treatment under investigation. OBJECTIVE The aim of this paper is to, from a statistical perspective, identify several key issues in the design and analysis of three COVID-19 trials and reanalyze the data from the cumulative incidence curves in the three trials using more appropriate statistical methods. METHODS The lopinavir-ritonavir trial enrolled 39 additional patients due to insignificant results after the sample size reached the planned number, which led to inflation of the type I error rate. The remdesivir trial of Wang et al failed to reach the planned sample size due to a lack of eligible patients, and the bootstrap method was used to predict the quantity of clinical interest conditionally and unconditionally if the trial had continued to reach the originally planned sample size. Moreover, we used a terminal (or cure) rate model and a model-free metric known as the restricted mean survival time or the restricted mean time to improvement (RMTI) to analyze the reconstructed data. The remdesivir trial of Beigel et al reported the median recovery time of the remdesivir and placebo groups, and the rate ratio for recovery, while both quantities depend on a particular time point representing local information. We use the restricted mean time to recovery (RMTR) as a global and robust measure for efficacy. RESULTS For the lopinavir-ritonavir trial, with the increase of sample size from 160 to 199, the type I error rate was inflated from 0.05 to 0.071. The difference of RMTIs between the two groups evaluated at day 28 was –1.67 days (95% CI –3.62 to 0.28; P=.09) in favor of lopinavir-ritonavir but not statistically significant. For the remdesivir trial of Wang et al, the difference of RMTIs at day 28 was –0.89 days (95% CI –2.84 to 1.06; P=.37). The planned sample size was 453, yet only 236 patients were enrolled. The conditional prediction shows that the hazard ratio estimates would reach statistical significance if the target sample size had been maintained. For the remdesivir trial of Beigel et al, the difference of RMTRs between the remdesivir and placebo groups at day 30 was –2.7 days (95% CI –4.0 to –1.2; P<.001), confirming the superiority of remdesivir. The difference in the recovery time at the 25th percentile (95% CI –3 to 0; P=.65) was insignificant, while the differences became more statistically significant at larger percentiles. CONCLUSIONS Based on the statistical issues and lessons learned from the recent three clinical trials on COVID-19 treatments, we suggest more appropriate approaches for the design and analysis of ongoing and future COVID-19 trials.