Explore the vast range of titles available.

With increasing numbers infected by SARS-CoV-2, understanding long-COVID is essential to inform health and social care support. A Scottish population cohort of 33,281 laboratory-confirmed SARS-CoV-2 infections and 62,957 never-infected individuals were followed-up via 6, 12 and 18-month questionnaires and linkage to hospitalization and death records. Of the 31,486 symptomatic infections,1,856 (6%) had not recovered and 13,350 (42%) only partially. No recovery was associated with hospitalized infection, age, female sex, deprivation, respiratory disease, depression and multimorbidity. Previous symptomatic infection was associated with poorer quality of life, impairment across all daily activities and 24 persistent symptoms including breathlessness (OR 3.43, 95% CI 3.29–3.58), palpitations (OR 2.51, OR 2.36–2.66), chest pain (OR 2.09, 95% CI 1.96–2.23), and confusion (OR 2.92, 95% CI 2.78–3.07). Asymptomatic infection was not associated with adverse outcomes. Vaccination was associated with reduced risk of seven symptoms. Here we describe the nature of long-COVID and the factors associated with it. In this population-based cohort study from Scotland, the authors investigate the prevalence of symptoms in the post-acute phase of COVID-19 infection compared to matched uninfected controls. They identify persistent symptoms associated with infection and identify factors associated with failure to recover.

Long-COVID prevalence estimates vary widely and should take account of symptoms that would have occurred anyway. Here we determine the prevalence of symptoms attributable to SARS-CoV-2 infection, taking account of background rates and confounding, in a nationwide population cohort study of 198,096 Scottish adults. 98,666 (49.8%) had symptomatic laboratory-confirmed SARS-CoV-2 infections and 99,430 (50.2%) were age-, sex-, and socioeconomically-matched and never-infected. While 41,775 (64.5%) reported at least one symptom 6 months following SARS-CoV-2 infection, this was also true of 34,600 (50.8%) of those never-infected. The crude prevalence of one or more symptom attributable to SARS-CoV-2 infection was 13.8% (13.2%,14.3%), 12.8% (11.9%,13.6%), and 16.3% (14.4%,18.2%) at 6, 12, and 18 months respectively. Following adjustment for potential confounders, these figures were 6.6% (6.3%, 6.9%), 6.5% (6.0%, 6.9%) and 10.4% (9.1%, 11.6%) respectively. Long-COVID is characterised by a wide range of symptoms that, apart from altered taste and smell, are non-specific. Care should be taken in attributing symptoms to previous SARS-CoV-2 infection. Determining the prevalence of Long COVID is challenging because many symptoms attributed to the syndrome could have other causes. Here, the authors estimate the prevalence of Long COVID in Scotland by comparing rates of symptoms reported by people with and without history of SARS-CoV-2 infection.

Heart Failure (HF) is common, with worldwide prevalence of 1%-3% and a lifetime risk of 20% for individuals 40 years or older. Despite its considerable health economic burden, techniques for early detection of HF in the general population are sparse. In this work we tested the hypothesis that a simple Transformer neural network, trained on comprehensive collection of secondary care data across the general population, can be used to prospectively (three-year predictive window) identify patients at an increased risk of first hospitalisation due to HF (HHF). The model was trained using routinely-collected, secondary care health data, including patient demographics, A&E attendances, hospitalisations, outpatient data, medications, blood tests, and vital sign measurements obtained across five years of longitudinal electronic health records (EHRs). The training cohort consisted of n = 183,894 individuals (n = 161,658 age/sex-matched controls and n = 22,236 of first hospitalisation due to HF after a three-year predictive window). Model performance was validated in an independent testing set of n = 8,977 patients (n = 945 HHF patients). Testing set probabilities were well-calibrated and achieved good discriminatory power with Area Under Receiver Operating Characteristic Curve (AUROC]) of 0.86, sensitivity of 36.4% (95% CI: 33.33%-39.56%), specificity of 98.26% (95% CI: 97.95%-98.53%), and PPV of 69.88% (95% CI: 65.86%-73.62%). At Probability of HHF ≥ 90% the model achieved 100% PPV (95% CI: 96.73%-100%) and sensitivity of 11.7% (95% CI: 9.72%-13.91%). Performance was not affected by patient sex or socioeconomic deprivation deciles. Performance was significantly better in Asian, Black, and Mixed ethnicities (AUROC 0.932–0.945) and in the 79–86 age group (AUROC 0.889). We present the first evidence that routinely collected secondary care health record data can be used in the general population to stratify patients at risk of first HHF.

Chest radiography (CXR) is the most commonly used imaging modality and deep neural network (DNN) algorithms have shown promise in effective triage of normal and abnormal radiograms. Typically, DNNs require large quantities of expertly labelled training exemplars, which in clinical contexts is a major bottleneck to effective modelling, as both considerable clinical skill and time is required to produce high-quality ground truths. In this work we evaluate thirteen supervised classifiers using two large free-text corpora and demonstrate that bi-directional long short-term memory (BiLSTM) networks with attention mechanism effectively identify Normal, Abnormal, and Unclear CXR reports in internal (n = 965 manually-labelled reports, f1-score = 0.94) and external (n = 465 manually-labelled reports, f1-score = 0.90) testing sets using a relatively small number of expert-labelled training observations (n = 3,856 annotated reports). Furthermore, we introduce a general unsupervised approach that accurately distinguishes Normal and Abnormal CXR reports in a large unlabelled corpus. We anticipate that the results presented in this work can be used to automatically extract standardized clinical information from free-text CXR radiological reports, facilitating the training of clinical decision support systems for CXR triage.

Previous studies on the natural history of long-COVID have been few and selective. Without comparison groups, disease progression cannot be differentiated from symptoms originating from other causes. The Long-COVID in Scotland Study (Long-CISS) is a Scotland-wide, general population cohort of adults who had laboratory-confirmed SARS-CoV-2 infection matched to PCR-negative adults. Serial, self-completed, online questionnaires collected information on pre-existing health conditions and current health six, 12 and 18 months after index test. Of those with previous symptomatic infection, 35% reported persistent incomplete/no recovery, 12% improvement and 12% deterioration. At six and 12 months, one or more symptom was reported by 71.5% and 70.7% respectively of those previously infected, compared with 53.5% and 56.5% of those never infected. Altered taste, smell and confusion improved over time compared to the never infected group and adjusted for confounders. Conversely, late onset dry and productive cough, and hearing problems were more likely following SARS-CoV-2 infection. The long-term natural history of long-COVID is not well understood. In this population-based cohort study from Scotland, the authors describe symptom prevalence and health-related quality of life up to 18 months after a positive SARS-CoV-2 test and compare with matched test-negative controls.

IntroductionVery little is known about possible clinical sequelae that may persist after resolution of acute COVID-19. A recent longitudinal cohort from Italy including 143 patients followed up after hospitalisation with COVID-19 reported that 87% had at least one ongoing symptom at 60-day follow-up. Early indications suggest that patients with COVID-19 may need even more psychological support than typical intensive care unit patients. The assessment of risk factors for longer term consequences requires a longitudinal study linked to data on pre-existing conditions and care received during the acute phase of illness. The primary aim of this study is to characterise physical and psychosocial sequelae in patients post-COVID-19 hospital discharge.Methods and analysisThis is an international open-access prospective, observational multisite study. This protocol is linked with the International Severe Acute Respiratory and emerging Infection Consortium (ISARIC) and the WHO’s Clinical Characterisation Protocol, which includes patients with suspected or confirmed COVID-19 during hospitalisation. This protocol will follow-up a subset of patients with confirmed COVID-19 using standardised surveys to measure longer term physical and psychosocial sequelae. The data will be linked with the acute phase data. Statistical analyses will be undertaken to characterise groups most likely to be affected by sequelae of COVID-19. The open-access follow-up survey can be used as a data collection tool by other follow-up studies, to facilitate data harmonisation and to identify subsets of patients for further in-depth follow-up. The outcomes of this study will inform strategies to prevent long-term consequences; inform clinical management, interventional studies, rehabilitation and public health management to reduce overall morbidity; and improve long-term outcomes of COVID-19.Ethics and disseminationThe protocol and survey are open access to enable low-resourced sites to join the study to facilitate global standardised, longitudinal data collection. Ethical approval has been given by sites in Colombia, Ghana, Italy, Norway, Russia, the UK and South Africa. New sites are welcome to join this collaborative study at any time. Sites interested in adopting the protocol as it is or in an adapted version are responsible for ensuring that local sponsorship and ethical approvals in place as appropriate. The tools are available on the ISARIC website (www.isaric.org).Protocol registration numberosf.io/c5rw3/Protocol version3 August 2020EuroQol ID37035.

Background As digital tools are increasingly used to support COVID‐19 contact tracing, the equity implications must be considered. As part of a study to understand the public's views of digital contact tracing tools developed for the national ‘Test and Protect’ programme in Scotland, we aimed to explore the views of groups often excluded from such discussions. This paper reports on their views about the potential for contact tracing to exacerbate inequalities. Methods A qualitative study was carried out; interviews were conducted with key informants from organizations supporting people in marginalized situations, followed by interviews and focus groups with people recruited from these groups. Participants included, or represented, minority ethnic groups, asylum seekers and refugees and those experiencing multiple disadvantage including severe and enduring poverty. Results A total of 42 people participated: 13 key informants and 29 members of the public. While public participants were supportive of contact tracing, key informants raised concerns. Both sets of participants spoke about how contact tracing, and its associated digital tools, might increase inequalities. Barriers included finances (inability to afford smartphones or the data to ensure access to the internet); language (digital tools were available only in English and required a degree of literacy, even for English speakers); and trust (many marginalized groups distrusted statutory organizations and there were concerns that data may be passed to other organizations). One strength was that NHS Scotland, the data guardian, is seen as a generally trustworthy organization. Poverty was recognized as a barrier to people's ability to self‐isolate. Some participants were concerned about giving contact details of individuals who might struggle to self‐isolate for financial reasons. Conclusions The impact of contact tracing and associated digital tools on marginalized populations needs careful monitoring. This should include the contact tracing process and the ability of people to self‐isolate. Regular clear messaging from trusted groups and community members could help maintain trust and participation in the programme. Patient and Public Contribution Our patient and public involvement coapplicant, L. L., was involved in all aspects of the study including coauthorship. Interim results were presented to our local Public and Patient Involvement and Engagement Group, who commented on interpretation and made suggestions about further recruitment.

Many western countries used shielding (extended self-isolation) of people presumed to be at high-risk from COVID-19 to protect them and reduce healthcare demand. To investigate the effectiveness of this strategy, we linked family practitioner, prescribing, laboratory, hospital and death records and compared COVID-19 outcomes among shielded and non-shielded individuals in the West of Scotland. Of the 1.3 million population, 27,747 (2.03%) were advised to shield, and 353,085 (26.85%) were classified a priori as moderate risk. COVID-19 testing was more common in the shielded (7.01%) and moderate risk (2.03%) groups, than low risk (0.73%). Referent to low-risk, the shielded group had higher confirmed infections (RR 8.45, 95% 7.44–9.59), case-fatality (RR 5.62, 95% CI 4.47–7.07) and population mortality (RR 57.56, 95% 44.06–75.19). The moderate-risk had intermediate confirmed infections (RR 4.11, 95% CI 3.82–4.42) and population mortality (RR 25.41, 95% CI 20.36–31.71) but, due to their higher prevalence, made the largest contribution to deaths (PAF 75.30%). Age ≥ 70 years accounted for 49.55% of deaths. In conclusion, in spite of the shielding strategy, high risk individuals were at increased risk of death. Furthermore, to be effective as a population strategy, shielding criteria would have needed to be widely expanded to include other criteria, such as the elderly.

Chest X-rays (CXRs) are the first-line investigation in patients presenting to emergency departments (EDs) with dyspnoea and are a valuable adjunct to clinical management of COVID-19 associated lung disease. Artificial intelligence (AI) has the potential to facilitate rapid triage of CXRs for further patient testing and/or isolation. In this work we develop an AI algorithm, CovIx, to differentiate normal, abnormal, non-COVID-19 pneumonia, and COVID-19 CXRs using a multicentre cohort of 293,143 CXRs. The algorithm is prospectively validated in 3289 CXRs acquired from patients presenting to ED with symptoms of COVID-19 across four sites in NHS Greater Glasgow and Clyde. CovIx achieves area under receiver operating characteristic curve for COVID-19 of 0.86, with sensitivity and F1-score up to 0.83 and 0.71 respectively, and performs on-par with four board-certified radiologists. AI-based algorithms can identify CXRs with COVID-19 associated pneumonia, as well as distinguish non-COVID pneumonias in symptomatic patients presenting to ED. Pre-trained models and inference scripts are freely available at https://github.com/beringresearch/bravecx-covid .