Explore the vast range of titles available.

In this Perspective, Aziz Sheikh discusses the importance of research to understand the impact of air pollution on human health, commenting on a study by Yaohua Tian and colleagues that examined associations between ambient air quality and risk of hospitalization for pneumonia in adults in China.

Emerging evidence has shown that COVID-19 survivors could suffer from persistent symptoms. However, it remains unclear whether these symptoms persist over the longer term. This study aimed to systematically synthesise evidence on post-COVID symptoms persisting for at least 12 months. We searched PubMed and Embase for papers reporting at least one-year follow-up results of COVID-19 survivors published by 6 November 2021. Random-effects meta-analyses were conducted to estimate pooled prevalence of specific post-COVID symptoms. Eighteen papers that reported one-year follow-up data from 8591 COVID-19 survivors were included. Fatigue/weakness (28%, 95% CI: 18–39), dyspnoea (18%, 95% CI: 13–24), arthromyalgia (26%, 95% CI: 8–44), depression (23%, 95% CI: 12–34), anxiety (22%, 95% CI: 15–29), memory loss (19%, 95% CI: 7–31), concentration difficulties (18%, 95% CI: 2–35), and insomnia (12%, 95% CI: 7–17) were the most prevalent symptoms at one-year follow-up. Existing evidence suggested that female patients and those with more severe initial illness were more likely to suffer from the sequelae after one year. This study demonstrated that a sizeable proportion of COVID-19 survivors still experience residual symptoms involving various body systems one year later. There is an urgent need for elucidating the pathophysiologic mechanisms and developing and testing targeted interventions for long-COVID patients.

An analysis of mortality among more than 114,000 SARS-CoV-2–infected people in Scotland revealed that vaccine effectiveness against death caused by the delta variant 14 days or more after the second dose was 90% for the BNT162b2 vaccine and 91% for the ChAdOx1 nCoV-19 vaccine.

We reflect on our experiences of using Generative Pre-trained Transformer ChatGPT, a chatbot launched by OpenAI in November 2022, to draft a research article. We aim to demonstrate how ChatGPT could help researchers to accelerate drafting their papers. We created a simulated data set of 100 000 health care workers with varying ages, Body Mass Index (BMI), and risk profiles. Simulation data allow analysts to test statistical analysis techniques, such as machine-learning based approaches, without compromising patient privacy. Infections were simulated with a randomized probability of hospitalisation. A subset of these fictitious people was vaccinated with a fictional vaccine that reduced this probability of hospitalisation after infection. We then used ChatGPT to help us decide how to handle the simulated data in order to determine vaccine effectiveness and draft a related research paper. AI-based language models in data analysis and scientific writing are an area of growing interest, and this exemplar analysis aims to contribute to the understanding of how ChatGPT can be used to facilitate these tasks.

The emergence of robotics is transforming industries around the world. Robot technologies are evolving exponentially, particularly as they converge with other functionalities such as artificial intelligence to learn from their environment, from each other, and from humans. The goal of the research was to understand the emerging role of robotics in health care and identify existing and likely future challenges to maximize the benefits associated with robotics and related convergent technologies. We conducted qualitative semistructured one-to-one interviews exploring the role of robotic applications in health care contexts. Using purposive sampling, we identified a diverse range of stakeholders involved in conceiving, procuring, developing, and using robotics in a range of national and international health care settings. Interviews were digitally recorded, transcribed verbatim, and analyzed thematically, supported by NVivo 10 (QSR International) software. Theoretically, this work was informed by the sociotechnical perspective, where social and technical systems are understood as being interdependent. We conducted 21 interviews and these accounts suggested that there are significant opportunities for improving the safety, quality, and efficiency of health care through robotics, but our analysis identified 4 major barriers that need to be effectively negotiated to realize these: (1) no clear pull from professionals and patients, (2) appearance of robots and associated expectations and concerns, (3) disruption of the way work is organized and distributed, and (4) new ethical and legal challenges requiring flexible liability and ethical frameworks. Sociotechnical challenges associated with the effective integration of robotic applications in health care settings are likely to be significant, particularly for patient-facing functions. These need to be identified and addressed for effective innovation and adoption.

Most screening tests for T2DM in use today were developed using multivariate regression methods that are often further simplified to allow transformation into a scoring formula. The increasing volume of electronically collected data opened the opportunity to develop more complex, accurate prediction models that can be continuously updated using machine learning approaches. This study compares machine learning-based prediction models (i.e. Glmnet, RF, XGBoost, LightGBM) to commonly used regression models for prediction of undiagnosed T2DM. The performance in prediction of fasting plasma glucose level was measured using 100 bootstrap iterations in different subsets of data simulating new incoming data in 6-month batches. With 6 months of data available, simple regression model performed with the lowest average RMSE of 0.838, followed by RF (0.842), LightGBM (0.846), Glmnet (0.859) and XGBoost (0.881). When more data were added, Glmnet improved with the highest rate (+ 3.4%). The highest level of variable selection stability over time was observed with LightGBM models. Our results show no clinically relevant improvement when more sophisticated prediction models were used. Since higher stability of selected variables over time contributes to simpler interpretation of the models, interpretability and model calibration should also be considered in development of clinical prediction models.

Global efforts toward the development and deployment of a vaccine for COVID-19 are rapidly advancing. To achieve herd immunity, widespread administration of vaccines is required, which necessitates significant cooperation from the general public. As such, it is crucial that governments and public health agencies understand public sentiments toward vaccines, which can help guide educational campaigns and other targeted policy interventions. The aim of this study was to develop and apply an artificial intelligence-based approach to analyze public sentiments on social media in the United Kingdom and the United States toward COVID-19 vaccines to better understand the public attitude and concerns regarding COVID-19 vaccines. Over 300,000 social media posts related to COVID-19 vaccines were extracted, including 23,571 Facebook posts from the United Kingdom and 144,864 from the United States, along with 40,268 tweets from the United Kingdom and 98,385 from the United States from March 1 to November 22, 2020. We used natural language processing and deep learning-based techniques to predict average sentiments, sentiment trends, and topics of discussion. These factors were analyzed longitudinally and geospatially, and manual reading of randomly selected posts on points of interest helped identify underlying themes and validated insights from the analysis. Overall averaged positive, negative, and neutral sentiments were at 58%, 22%, and 17% in the United Kingdom, compared to 56%, 24%, and 18% in the United States, respectively. Public optimism over vaccine development, effectiveness, and trials as well as concerns over their safety, economic viability, and corporation control were identified. We compared our findings to those of nationwide surveys in both countries and found them to correlate broadly. Artificial intelligence-enabled social media analysis should be considered for adoption by institutions and governments alongside surveys and other conventional methods of assessing public attitude. Such analyses could enable real-time assessment, at scale, of public confidence and trust in COVID-19 vaccines, help address the concerns of vaccine sceptics, and help develop more effective policies and communication strategies to maximize uptake.

In this Perspective, Aziz Sheikh discusses research to evaluate health policy changes in the provision of care, commenting on a study by James Lopez Bernal and colleagues that examined specialist-dominated hospital care versus community-based care in the United Kingdom.

Background Smokeless tobacco (ST) is consumed by more than 300 million people worldwide. The distribution, determinants and health risks of ST differ from that of smoking; hence, there is a need to highlight its distinct health impact. We present the latest estimates of the global burden of disease due to ST use. Methods The ST-related disease burden was estimated for all countries reporting its use among adults. Using systematic searches, we first identified country-specific prevalence of ST use in men and women. We then revised our previously published disease risk estimates for oral, pharyngeal and oesophageal cancers and cardiovascular diseases by updating our systematic reviews and meta-analyses of observational studies. The updated country-specific prevalence of ST and disease risk estimates, including data up to 2019, allowed us to revise the population attributable fraction (PAF) for ST for each country. Finally, we estimated the disease burden attributable to ST for each country as a proportion of the DALYs lost and deaths reported in the 2017 Global Burden of Disease study. Results ST use in adults was reported in 127 countries; the highest rates of consumption were in South and Southeast Asia. The risk estimates for cancers were also highest in this region. In 2017, at least 2.5 million DALYs and 90,791 lives were lost across the globe due to oral, pharyngeal and oesophageal cancers that can be attributed to ST. Based on risk estimates obtained from the INTERHEART study, over 6 million DALYs and 258,006 lives were lost from ischaemic heart disease that can be attributed to ST. Three-quarters of the ST-related disease burden was among men. Geographically, > 85% of the ST-related burden was in South and Southeast Asia, India accounting for 70%, Pakistan for 7% and Bangladesh for 5% DALYs lost. Conclusions ST is used across the globe and poses a major public health threat predominantly in South and Southeast Asia. While our disease risk estimates are based on a limited evidence of modest quality, the likely ST-related disease burden is substantial. In high-burden countries, ST use needs to be regulated through comprehensive implementation of the World Health Organization Framework Convention for Tobacco Control.

In this Perspective on the two clinical trials of Terry Haines and colleagues that incrementally removed and reinstated allied healthcare services, Aziz Sheikh discusses the evidence base for the routine provision of such services.