Explore the vast range of titles available.

Meta-analysis is the quantitative, scientific synthesis of research results. Since the term and modern approaches to research synthesis were first introduced in the 1970s, meta-analysis has had a revolutionary effect in many scientific fields, helping to establish evidence-based practice and to resolve seemingly contradictory research outcomes. At the same time, its implementation has engendered criticism and controversy, in some cases general and others specific to particular disciplines. Here we take the opportunity provided by the recent fortieth anniversary of meta-analysis to reflect on the accomplishments, limitations, recent advances and directions for future developments in the field of research synthesis. Meta-analysis—the quantitative, scientific synthesis of research results—has been both revolutionary and controversial, with rapid advances and broad implementation resulting in substantial scientific advances, but not without pitfalls. The rise of research synthesis Four decades after its introduction, meta-analysis has become a widely accepted research synthesis tool. In this Review, Jessica Gurevitch and colleagues explore the history, development and current state of meta-analytic practice in the biological sciences. They outline the contributions that it has made to several disciplines, particularly ecology, evolutionary biology and conservation, where the number of meta-analyses has increased exponentially over time. They discuss some of the pitfalls of these types of analyses and summarize recent developments such as the use of machine learning and artificial intelligence. They suggest that evidence synthesis should become a regular companion to primary scientific research to maximize the effectiveness of scientific inquiry, but call for the rigorous application of stricter quality criteria for the publication of meta-analyses.

Poor medication adherence is a pervasive issue with considerable health and socioeconomic consequences. Although the underlying reasons are generally understood, traditional intervention strategies rooted in patient-centric education and empowerment have proved to be prohibitively complex and/or ineffective. Formulating a pharmaceutical in a drug delivery system (DDS) is a promising alternative that can directly mitigate many common impediments to adherence, including frequent dosing, adverse effects and a delayed onset of action. Existing DDSs have already positively influenced patient acceptability and improved rates of adherence across various disease and intervention types. The next generation of systems have the potential to instate an even more radical paradigm shift by, for example, permitting oral delivery of biomacromolecules, allowing for autonomous dose regulation and enabling several doses to be mimicked with a single administration. Their success, however, is contingent on their ability to address the problems that have made DDSs unsuccessful in the past.Improving medication adherence is recognized as one of the most impactful and cost-effective strategies for improving the health of the general population. Here, Baryakova and colleagues assess the potential of next-generation drug delivery systems to mitigate many common impediments to adherence and discuss the impact that drug delivery systems have had across different disease types.

Non-invasive assessment of the risk of lymph node metastasis (LNM) in patients with papillary thyroid carcinoma (PTC) is of great value for the treatment option selection. The purpose of this paper is to develop a transfer learning radiomics (TLR) model for preoperative prediction of LNM in PTC patients in a multicenter, cross-machine, multi-operator scenario. Here we report the TLR model produces a stable LNM prediction. In the experiments of cross-validation and independent testing of the main cohort according to diagnostic time, machine, and operator, the TLR achieves an average area under the curve (AUC) of 0.90. In the other two independent cohorts, TLR also achieves 0.93 AUC, and this performance is statistically better than the other three methods according to Delong test. Decision curve analysis also proves that the TLR model brings more benefit to PTC patients than other methods. A non-destructive and efficient method for predicting the risk of lymph node metastasis (LNM) in papillary thyroid carcinoma (PTC) patients is highly needed. Here, the authors develop a transfer learning radiomics model for preoperative prediction of LNM in patients with PTC in a multicenter scenario.

The influence of comorbidities on COVID-19 outcomes has been recognized since the earliest days of the pandemic. But establishing causality and determining underlying mechanisms and clinical implications has been challenging—owing to the multitude of confounding factors and patient variability. Several distinct pathological mechanisms, not active in every patient, determine health outcomes in the three different phases of COVID-19—from the initial viral replication phase to inflammatory lung injury and post-acute sequelae. Specific comorbidities (and overall multimorbidity) can either exacerbate these pathological mechanisms or reduce the patient’s tolerance to organ injury. In this Review, we consider the impact of specific comorbidities, and overall multimorbidity, on the three mechanistically distinct phases of COVID-19, and we discuss the utility of host genetics as a route to causal inference by eliminating many sources of confounding. Continued research into the mechanisms of disease-state interactions will be crucial to inform stratification of therapeutic approaches and improve outcomes for patients. This Review discusses the effect of comorbidities and multimorbidity on the three mechanistically distinct phases of COVID-19, evaluating the evidence in the context of confounding factors and our evolving understanding of the disease.

The acute clinical manifestations of COVID-19 have been well characterized 1 , 2 , but the post-acute sequelae of this disease have not been comprehensively described. Here we use the national healthcare databases of the US Department of Veterans Affairs to systematically and comprehensively identify 6-month incident sequelae—including diagnoses, medication use and laboratory abnormalities—in patients with COVID-19 who survived for at least 30 days after diagnosis. We show that beyond the first 30 days of illness, people with COVID-19 exhibit a higher risk of death and use of health resources. Our high-dimensional approach identifies incident sequelae in the respiratory system, as well as several other sequelae that include nervous system and neurocognitive disorders, mental health disorders, metabolic disorders, cardiovascular disorders, gastrointestinal disorders, malaise, fatigue, musculoskeletal pain and anaemia. We show increased incident use of several therapeutic agents—including pain medications (opioids and non-opioids) as well as antidepressant, anxiolytic, antihypertensive and oral hypoglycaemic agents—as well as evidence of laboratory abnormalities in several organ systems. Our analysis of an array of prespecified outcomes reveals a risk gradient that increases according to the severity of the acute COVID-19 infection (that is, whether patients were not hospitalized, hospitalized or admitted to intensive care). Our findings show that a substantial burden of health loss that spans pulmonary and several extrapulmonary organ systems is experienced by patients who survive after the acute phase of COVID-19. These results will help to inform health system planning and the development of multidisciplinary care strategies to reduce chronic health loss among individuals with COVID-19. Healthcare data from the US Department of Veterans Affairs are used to characterize the six-month incident sequelae of individuals who survive for at least thirty days after developing COVID-19.

Despite the importance of chemotherapy-associated adverse events in oncology practice and the broad range of interventions available to mitigate them, limited systematic efforts have been made to identify, critically appraise and summarize the totality of evidence on the effectiveness of these interventions. Herein, we review the most common long-term (continued beyond treatment) and late or delayed (following treatment) adverse events associated with chemotherapy and other anticancer treatments that pose major threats in terms of survival, quality of life and continuation of optimal therapy. These adverse effects often emerge during and continue beyond the course of therapy or arise among survivors in the months and years following treatment. For each of these adverse effects, we discuss and critically evaluate their underlying biological mechanisms, the most commonly used pharmacological and non-pharmacological treatment strategies, and evidence-based clinical practice guidelines for their appropriate management. Furthermore, we discuss risk factors and validated risk-assessment tools for identifying patients most likely to be harmed by chemotherapy and potentially benefit from effective interventions. Finally, we highlight promising emerging supportive-care opportunities for the ever-increasing number of cancer survivors at continuing risk of adverse treatment effects.The effective management of treatment-related events remains an unmet need in oncology. The authors of this Review discuss the underlying biological mechanisms, risk factors, most commonly used pharmacological and non-pharmacological management strategies, and clinical practice guidelines for the most common long-term (continuing beyond treatment) and late or delayed (following treatment) adverse events associated with chemotherapy and other anticancer treatments.

Pulsed field ablation (PFA) is an emerging technology for the treatment of atrial fibrillation (AF), for which pre-clinical and early-stage clinical data are suggestive of some degree of preferentiality to myocardial tissue ablation without damage to adjacent structures. Here in the MANIFEST-17K study we assessed the safety of PFA by studying the post-approval use of this treatment modality. Of the 116 centers performing post-approval PFA with a pentaspline catheter, data were received from 106 centers (91.4% participation) regarding 17,642 patients undergoing PFA (mean age 64, 34.7% female, 57.8% paroxysmal AF and 35.2% persistent AF). No esophageal complications, pulmonary vein stenosis or persistent phrenic palsy was reported (transient palsy was reported in 0.06% of patients; 11 of 17,642). Major complications, reported for ~1% of patients (173 of 17,642), were pericardial tamponade (0.36%; 63 of 17,642) and vascular events (0.30%; 53 of 17,642). Stroke was rare (0.12%; 22 of 17,642) and death was even rarer (0.03%; 5 of 17,642). Unexpected complications of PFA were coronary arterial spasm in 0.14% of patients (25 of 17,642) and hemolysis-related acute renal failure necessitating hemodialysis in 0.03% of patients (5 of 17,642). Taken together, these data indicate that PFA demonstrates a favorable safety profile by avoiding much of the collateral damage seen with conventional thermal ablation. PFA has the potential to be transformative for the management of patients with AF. In a post-approval study including more than 17,000 patients on the safety of pulsed field ablation, a new method for treatment of atrial fibrillation, the procedure was found to have a low rate of adverse events but was associated with some unexpected rare complications that will need further study.

The brain requires a continuous supply of energy in the form of ATP, most of which is produced from glucose by oxidative phosphorylation in mitochondria, complemented by aerobic glycolysis in the cytoplasm. When glucose levels are limited, ketone bodies generated in the liver and lactate derived from exercising skeletal muscle can also become important energy substrates for the brain. In neurodegenerative disorders of ageing, brain glucose metabolism deteriorates in a progressive, region-specific and disease-specific manner — a problem that is best characterized in Alzheimer disease, where it begins presymptomatically. This Review discusses the status and prospects of therapeutic strategies for countering neurodegenerative disorders of ageing by improving, preserving or rescuing brain energetics. The approaches described include restoring oxidative phosphorylation and glycolysis, increasing insulin sensitivity, correcting mitochondrial dysfunction, ketone-based interventions, acting via hormones that modulate cerebral energetics, RNA therapeutics and complementary multimodal lifestyle changes.Accumulating evidence indicates that impaired glucose metabolism in the brain is involved in the cause and progression of neurodegenerative disorders of ageing such as Alzheimer disease. This Review discusses the status and prospects of therapeutic strategies for countering neurodegenerative disorders of ageing by rescuing, protecting or normalizing brain energetics.

The recent release of large-scale healthcare datasets has greatly propelled the research of data-driven deep learning models for healthcare applications. However, due to the nature of such deep black-boxed models, concerns about interpretability, fairness, and biases in healthcare scenarios where human lives are at stake call for a careful and thorough examination of both datasets and models. In this work, we focus on MIMIC-IV (Medical Information Mart for Intensive Care, version IV), the largest publicly available healthcare dataset, and conduct comprehensive analyses of interpretability as well as dataset representation bias and prediction fairness of deep learning models for in-hospital mortality prediction. First, we analyze the interpretability of deep learning mortality prediction models and observe that (1) the best-performing interpretability method successfully identifies critical features for mortality prediction on various prediction models as well as recognizing new important features that domain knowledge does not consider; (2) prediction models rely on demographic features, raising concerns in fairness. Therefore, we then evaluate the fairness of models and do observe the unfairness: (1) there exists disparate treatment in prescribing mechanical ventilation among patient groups across ethnicity, gender and age; (2) models often rely on racial attributes unequally across subgroups to generate their predictions. We further draw concrete connections between interpretability methods and fairness metrics by showing how feature importance from interpretability methods can be beneficial in quantifying potential disparities in mortality predictors. Our analysis demonstrates that the prediction performance is not the only factor to consider when evaluating models for healthcare applications, since high prediction performance might be the result of unfair utilization of demographic features. Our findings suggest that future research in AI models for healthcare applications can benefit from utilizing the analysis workflow of interpretability and fairness as well as verifying if models achieve superior performance at the cost of introducing bias.

Secondary use of electronic health records (EHRs) promises to advance clinical research and better inform clinical decision making. Challenges in summarizing and representing patient data prevent widespread practice of predictive modeling using EHRs. Here we present a novel unsupervised deep feature learning method to derive a general-purpose patient representation from EHR data that facilitates clinical predictive modeling. In particular, a three-layer stack of denoising autoencoders was used to capture hierarchical regularities and dependencies in the aggregated EHRs of about 700,000 patients from the Mount Sinai data warehouse. The result is a representation we name “deep patient”. We evaluated this representation as broadly predictive of health states by assessing the probability of patients to develop various diseases. We performed evaluation using 76,214 test patients comprising 78 diseases from diverse clinical domains and temporal windows. Our results significantly outperformed those achieved using representations based on raw EHR data and alternative feature learning strategies. Prediction performance for severe diabetes, schizophrenia, and various cancers were among the top performing. These findings indicate that deep learning applied to EHRs can derive patient representations that offer improved clinical predictions, and could provide a machine learning framework for augmenting clinical decision systems.