Search

Rare events meta-analyses of randomized controlled trials (RCTs) are often underpowered because the outcomes are infrequent. Real-world evidence (RWE) from non-randomized studies may provide valuable complementary evidence about the effects of rare events, and there is growing interest in including such evidence in the decision-making process. Several methods for combining RCTs and RWE studies have been proposed, but the comparative performance of these methods is not well understood. We describe a simulation study that aims to evaluate an array of alternative Bayesian methods for including RWE in rare events meta-analysis of RCTs: the naïve data synthesis, the design-adjusted synthesis, the use of RWE as prior information, the three-level hierarchical models, and the bias-corrected meta-analysis model. The percentage bias, root-mean-square-error, mean 95% credible interval width, coverage probability, and power are used to measure performance. The various methods are illustrated using a systematic review to evaluate the risk of diabetic ketoacidosis among patients using sodium/glucose co-transporter 2 inhibitors as compared with active-comparators. Our simulations show that the bias-corrected meta-analysis model is comparable to or better than the other methods in terms of all evaluated performance measures and simulation scenarios. Our results also demonstrate that data solely from RCTs may not be sufficiently reliable for assessing the effects of rare events. In summary, the inclusion of RWE could increase the certainty and comprehensiveness of the body of evidence of rare events from RCTs, and the bias-corrected meta-analysis model may be preferable.

There is a growing demand for the use of high‐quality real‐world evidence (RWE) to support regulatory decision‐making worldwide and in China, which highlights the need for conducting literature reviews to evaluate the available data and evidence. This study aims to review the use of RWE in Chinese regulatory decisions and to summarize relevant regulatory and methodological considerations to inform the future use of RWE in China. We identified policy documents, technical guidance documents, and cases on official Chinese government websites and extracted their contents separately. We consulted experts from the National Medical Products Administration (NMPA) and academic institutes and searched case‐related articles for enrichment. We also searched and included articles related to the use of RWE/Real‐world data in Chinese regulatory decisions. Six trial versions of technical guidance documents, 7 case studies, and 40 articles related to the Chinese regulatory decisions were included in this study. Based on the technical guidance, data quality, and appropriate study design and statistical analysis are the main concerns for RWE generation. The cases and articles related to regulatory decisions revealed 9 main concerns, including data sources and applicability, data quality, strength of existing evidence, appropriate study design and statistical analysis, regulated and transparent process for analysis and evidence generation, product safety and efficacy, product characteristics and clinical needs, ethical considerations and data security, and communicate adequately with regulatory authorities. Among these concerns, data issues are central. Preliminary attempts have been made by the NMPA to promote the use of RWE, but substantial challenges still remain. [ABSTRACT FROM AUTHOR]

The meta‐analysis of rare events presents unique methodological challenges owing to the small number of events. Bayesian methods are often used to combine rare events data to inform decision‐making, as they can incorporate prior information and handle studies with zero events without the need for continuity corrections. However, the comparative performances of different Bayesian models in pooling rare events data are not well understood. We conducted a simulation to compare the statistical properties of four parameterizations based on the binomial‐normal hierarchical model, using two different priors for the treatment effect: weakly informative prior (WIP) and non‐informative prior (NIP), pooling randomized controlled trials with rare events using the odds ratio metric. We also considered the beta‐binomial model proposed by Kuss and the random intercept and slope generalized linear mixed models. The simulation scenarios varied based on the treatment effect, sample size ratio between the treatment and control arms, and level of heterogeneity. Performance was evaluated using median bias, root mean square error, median width of 95% credible or confidence intervals, coverage, Type I error, and empirical power. Two reviews are used to illustrate these methods. The results demonstrate that the WIP outperforms the NIP within the same model structure. Among the compared models, the model that included the treatment effect parameter in the risk model for the control arm did not perform well. Our findings confirm that rare events meta‐analysis faces the challenge of being underpowered, highlighting the importance of reporting the power of results in empirical studies. [ABSTRACT FROM AUTHOR]

Objective: Time‐varying treatments are common in observational studies. However, when assessing treatment effects, the methodological framework has not been systematically established for handling time‐varying treatments. This study aimed to examine the current methods for dealing with time‐varying treatments in observational studies and developed practical recommendations. Methods: We searched PubMed from 2000 to 2021 for methodological articles about time‐varying treatments, and qualitatively summarized the current methods for handling time‐varying treatments. Subsequently, we developed practical recommendations through interactive internal group discussions and consensus by a panel of external experts. Results: Of the 36 eligible reports (22 methodological reviews, 10 original studies, 2 tutorials and 2 commentaries), most examined statistical methods for time‐varying treatments, and only a few discussed the overarching methodological process. Generally, there were three methodological components to handle time‐varying treatments. These included the specification of treatment which may be categorized as three scenarios (i.e., time‐independent treatment, static treatment regime, or dynamic treatment regime); definition of treatment status which could involve three approaches (i.e., intention‐to‐treat, per‐protocol, or as‐treated approach); and selection of analytic methods. Based on the review results, a methodological workflow and a set of practical recommendations were proposed through two consensus meetings. Conclusions: There is no consensus process for assessing treatment effects in observational studies with time‐varying treatments. Previous efforts were dedicated to developing statistical methods. Our study proposed a stepwise workflow with practical recommendations to assist the practice. [ABSTRACT FROM AUTHOR]

Rare events meta‐analyses of randomized controlled trials (RCTs) are often underpowered because the outcomes are infrequent. Real‐world evidence (RWE) from non‐randomized studies may provide valuable complementary evidence about the effects of rare events, and there is growing interest in including such evidence in the decision‐making process. Several methods for combining RCTs and RWE studies have been proposed, but the comparative performance of these methods is not well understood. We describe a simulation study that aims to evaluate an array of alternative Bayesian methods for including RWE in rare events meta‐analysis of RCTs: the naïve data synthesis, the design‐adjusted synthesis, the use of RWE as prior information, the three‐level hierarchical models, and the bias‐corrected meta‐analysis model. The percentage bias, root‐mean‐square‐error, mean 95% credible interval width, coverage probability, and power are used to measure performance. The various methods are illustrated using a systematic review to evaluate the risk of diabetic ketoacidosis among patients using sodium/glucose co‐transporter 2 inhibitors as compared with active‐comparators. Our simulations show that the bias‐corrected meta‐analysis model is comparable to or better than the other methods in terms of all evaluated performance measures and simulation scenarios. Our results also demonstrate that data solely from RCTs may not be sufficiently reliable for assessing the effects of rare events. In summary, the inclusion of RWE could increase the certainty and comprehensiveness of the body of evidence of rare events from RCTs, and the bias‐corrected meta‐analysis model may be preferable. [ABSTRACT FROM AUTHOR]

Background: Many rare events meta-analyses of randomized controlled trials (RCTs) have lower statistical power, and real-world evidence (RWE) is becoming widely recognized as a valuable source of evidence. The purpose of this study is to investigate methods for including RWE in a rare events meta-analysis of RCTs and the impact on the level of uncertainty around the estimates. Methods: Four methods for the inclusion of RWE in evidence synthesis were investigated by applying them to two previously published rare events meta-analyses: the naïve data synthesis (NDS), the design-adjusted synthesis (DAS), the use of RWE as prior information (RPI), and the three-level hierarchical models (THMs). We gauged the effect of the inclusion of RWE by varying the degree of confidence placed in RWE. Results: This study showed that the inclusion of RWE in a rare events meta-analysis of RCTs could increase the precision of the estimates, but this depended on the method of inclusion and the level of confidence placed in RWE. NDS cannot consider the bias of RWE, and its results may be misleading. DAS resulted in stable estimates for the two examples, regardless of whether we placed high- or low-level confidence in RWE. The results of the RPI approach were sensitive to the confidence level placed in RWE. The THM was effective in allowing for accommodating differences between study types, while it had a conservative result compared with other methods. Conclusion: The inclusion of RWE in a rare events meta-analysis of RCTs could increase the level of certainty of the estimates and enhance the decision-making process. DAS might be appropriate for inclusion of RWE in a rare event meta-analysis of RCTs, but further evaluation in different scenarios of empirical or simulation studies is still warranted. [ABSTRACT FROM AUTHOR]

Graph Neural Networks (GNNs) are powerful tools in graph application areas. However, recent studies indicate that GNNs are vulnerable to adversarial attacks, which can lead GNNs to easily make wrong predictions for downstream tasks. A number of works aim to solve this problem but what criteria we should follow to clean the perturbed graph is still a challenge. In this paper, we propose GSP-GNN, a general framework to defend against massive poisoning attacks that can perturb graphs. The vital principle of GSP-GNN is to explore the similarity property to mitigate negative effects on graphs. Specifically, this method prunes adversarial edges by the similarity of node feature and graph structure to eliminate adversarial perturbations. In order to stabilize and enhance GNNs training process, previous layer information is adopted in case a large number of edges are pruned in one layer. Extensive experiments on three real-world graphs demonstrate that GSP-GNN achieves significantly better performance compared with the representative baselines and has favorable generalization ability simultaneously. [ABSTRACT FROM AUTHOR]

Efficacy of acupuncture-related therapy in the treatment of knee osteoarthritis: a network meta-analysis of randomized controlled trials. Knee osteoarthritis (KOA) is a degenerative disease in which articular cartilage and synovial fluid in the knee are lost, leading to lower limb disability, particularly in elderly individuals.[1] While nonsteroidal anti-inflammatory drugs (NSAIDs) can improve the physical function of KOA patients, they pose a significant risk of gastrointestinal adverse events that hinder disease management.[2] Acupuncture (AT) is one of the most common non-pharmacological therapies for the treatment of musculoskeletal disease-related symptoms.[3] Although evidence supports the potential effectiveness of acupuncture in treating KOA-related symptoms, the findings are controversial.[4] Therefore, we conducted a systematic review, quantitative analysis, and summary of existing evidence to investigate the safety and effectiveness of acupuncture for KOA. Acupuncture Trialists C. The persistence of the effects of acupuncture after a course of treatment: a meta-analysis of patients with chronic pain. [Extracted from the article]

Objective: To date, evidence regarding the effectiveness and safety of two consecutive cycles of single embryo transfer (2SETs) compared with one cycle of double embryo transfer (DET) has been inadequate, particularly considering infertile women with different prognostic factors. This study aimed to comprehensively summarize the evidence by comparing 2SETs with DET. Methods: PubMed, Embase, Cochrane Library databases, ClinicalTrails.gov, and the WHO International Clinical Trials Registry Platform were searched up to March 22, 2022. Peer-reviewed, English-language randomized controlled trials (RCTs) and observational studies (OS) comparing the outcomes of 2SETs with DET in infertile women with their own oocytes and embryos were included. Two authors independently conducted study selection, data extraction, and bias assessment. The Mantel–Haenszel random-effects model was used for pooling RCTs, and a Bayesian design-adjusted model was conducted to synthesize the results from both RCTs and OS. Main Results: Twelve studies were finally included. Compared with the DET, 2SETs were associated with a similar cumulative live birth rate (LBR; 48.24% vs. 48.91%; OR, 0.97; 95% credible interval (CrI), 0.89–1.13, τ2 = 0.1796; four RCTs and six observational studies; 197,968 women) and a notable lower cumulative multiple birth rate (MBR; 0.87% vs. 17.72%; OR, 0.05; 95% CrI, 0.02–0.10, τ2 = 0.1036; four RCTs and five observational studies; 197,804 women). Subgroup analyses revealed a significant increase in cumulative LBR (OR, 1.33; 95% CrI, 1.29–1.38, τ2 = 0) after two consecutive cycles of single blastocyst transfer compared with one cycle of double blastocyst transfer. Moreover, a lower risk of cesarean section, antepartum hemorrhage, preterm birth, low birth weight, and neonatal intensive care unit admission but a higher gestational age at birth and birth weight were found in the 2SETs group. Conclusion: Compared to the DET strategy, 2SETs result in a similar LBR while simultaneously reducing the MBR and improving maternal and neonatal adverse outcomes. The 2SETs strategy appears to be especially beneficial for women aged ≤35 years and for blastocyst transfers. [ABSTRACT FROM AUTHOR]

• Estimates of PM 2.5 concentrations in 2030 declined compared with 2015. • If the age structure maintains, PM 2.5 -related premature death in 2030 related to 2015 decreased. • With the population aging, PM 2.5 -related premature death would sharply increase in 2030. Fine particulate matter (PM 2.5) pollution is one of the most critical environmental and public health problems in China and has caused an enormous disease burden, especially long-term PM 2.5 exposure. Global climate change represents another environmental challenge in the coming decades and is also an essential factor affecting PM 2.5 pollution. Moreover, China has an aging population with a changing population size and falling age-standardized mortality rates. However, little evidence exists evaluating the potential impacts from climate change and population aging on the long-term PM 2.5 exposure-related disease burden. This study quantifies the impacts of climate and population changes on changes in the disease burden attributed to long-term PM 2.5 exposure from 2015 to 2030 in mainland China, which could add evidence for the revision of relevant environmental standards and health policies. This modeling study investigated long-term PM 2.5 exposure-related mortality across China based on PM 2.5 projections under Intergovernmental Panel on Climate Change Representative Concentration Pathways (RCPs) and population scenarios from shared socioeconomic pathways (SSPs). PM 2.5 concentrations were simulated by the Weather Research and Forecasting (WRF) and Community Multiscale Air Quality (CMAQ) modeling systems. In addition, three types of population projections in 2030 relative to 2015 were set up as follows: (i) the population remained the same as that in 2015; (ii) the population size changed under SSPs, but the age structure remained the same; (iii) both the population size and age structure changed under SSPs. The global exposure mortality model (GEMM) was adopted to estimate PM 2.5 -related premature deaths. Ambient PM 2.5 concentrations decreased from 2015 to 2030 under the two climate and emission scenarios. Estimates of related premature mortality in 2030 declined compared with that in 2015 due to lower PM 2.5 concentrations (RCP4.5: −16.8%; RCP8.5: −16.4%). If the age structure of the population remained unchanged and the population size changed under SSPs, the nonaccidental premature mortality also showed a decrease ranging from −18.6% to −14.9%. When both population size and age structure changed under SSPs, the population in China would become older. Nonaccidental premature mortality would sharply increase by 35.7–52.3% (with a net increase of 666–977 thousand) in 2030. The PM 2.5 pollution in 2030 under both RCP4.5 and RCP8.5 would slightly improve. The population sizes in 2030 projected by SSPs are relatively stable compared with that in 2015. However, the modest decrease due to air pollution improvement and stable population size would be offset by population aging. [ABSTRACT FROM AUTHOR]

Existing studies focused on the evaluation of health burden of long-term exposure to air pollutants, whereas limited information is available on short-term exposure, particularly in China. Air pollutant concentrations in 338 Chinese cities in 2017 were used to estimate the air pollutants related health burden which was defined as premature mortalities from all-cause, cardiovascular and respiratory disease as well as hospital admissions for cardiovascular and respiratory disease. Log-linear model was used as the exposure-response function to estimate the health burden attributable to each air pollutant. The value of statistical life and cost of illness methods were used to estimate economic loss of the premature mortalities and hospital admissions, respectively. The national all-cause premature mortalities attributable to all air pollutants was 1.35 million, accounting for 17.2% of reported deaths in China in 2017. Among all-cause premature mortality, contributions of PM 2.5 , PM 2.5-10 , NO 2 , SO 2 , O 3 and CO were11.1%, 5.2%, 28.9%, 9.6%, 23.0%, and 22.2%, respectively. The national cardiovascular and respiratory premature mortalities were 0.77 and 0.21 million, respectively. About 7.8 million cardiovascular and respiratory disease hospital admissions were attributed to short-term exposure to all air pollutants. The economic loss of the overall health burden (premature mortality and hospital admissions) was 2065.54 billion Yuan, which was equivalent to 2.5% of the national GDP in 2017. The health burden and economic loss attributable to short-term exposure to ambient air pollutant are substantial in China. It suggested that the adverse health effects attributable to short-term exposure to air pollutant should not be neglected in China. In order to reduce the health impact of air pollution, each city should develop air pollution prevention and control measures based on existing scientific evidence. • The mortalities and hospital admissions attributable to short-term exposure to six air pollutants were estimated in China. • The national all-cause premature mortality attributable to all air pollutants was 1.35 million. • About 7.8 million hospital admissions were attributed to short-term exposure to air pollutants. • The economic loss of the overall health burden was 2065.54 billion Yuan, accounting for 2.5% of the national GDP. [ABSTRACT FROM AUTHOR]

Text field labelling plays a key role in Key Information Extraction (KIE) from structured document images. However, existing methods ignore the field drift and outlier problems, which limit their performance and make them less robust. This paper casts the text field labelling problem into a partial graph matching problem and proposes an end-to-end trainable framework called Deep Partial Graph Matching (dPGM) for the one-shot KIE task. It represents each document as a graph and estimates the correspondence between text fields from different documents by maximizing the graph similarity of different documents. Our framework obtains a strict one-to-one correspondence by adopting a combinatorial solver module with an extra one-to-(at most)-one mapping constraint to do the exact graph matching, which leads to the robustness of the field drift problem and the outlier problem. Finally, a large one-shot KIE dataset named DKIE is collected and annotated to promote research of the KIE task. This dataset will be released to the research and industry communities. Extensive experiments on both the public and our new DKIE datasets show that our method can achieve state-of-the-art performance and is more robust than existing methods.

Introduction: Although interest in including non-randomised studies of interventions (NRSIs) in meta-analysis of randomised controlled trials (RCTs) is growing, estimates of effectiveness obtained from NRSIs are vulnerable to greater bias than RCTs. The objectives of this study are to: (1) explore how NRSIs can be integrated into a meta-analysis of RCTs; (2) assess concordance of the evidence from non-randomised and randomised trials and explore factors associated with agreement; and (3) investigate the impact on estimates of pooled bodies of evidence when NRSIs are included., Methods and Analysis: We will conduct a systematic survey of 210 systematic reviews that include both RCTs and NRSIs, published from 2017 to 2022. We will randomly select reviews, stratified in a 1:1 ratio by Core vs non-Core clinical journals, as defined by the National Library of Medicine. Teams of paired reviewers will independently determine eligibility and abstract data using standardised, pilot-tested forms. The concordance of the evidence will be assessed by exploring agreement in the relative effect reported by NRSIs and RCT addressing the same clinical question, defined as similarity of the population, intervention/exposure, control and outcomes. We will conduct univariable and multivariable logistic regression analyses to examine the association of prespecified study characteristics with agreement in the estimates between NRSIs and RCTs. We will calculate the ratio of the relative effect estimate from NRSIs over that from RCTs, along with the corresponding 95% CI. We will use a bias-corrected meta-analysis model to investigate the influence on pooled estimates when NRSIs are included in the evidence synthesis., Ethics and Dissemination: Ethics approval is not required. The findings of this study will be disseminated through peer-reviewed publications, conference presentations and condensed summaries for clinicians, health policymakers and guideline developers regarding the design, conduct, analysis, and interpretation of meta-analysis that integrate RCTs and NRSIs., Competing Interests: Competing interests: None declared., (© Author(s) (or their employer(s)) 2023. Re-use permitted under CC BY-NC. No commercial re-use. See rights and permissions. Published by BMJ.)