Post-trial monitoring of a randomised controlled trial of intensive glycaemic control in type 2 diabetes extended from 10 years to 24 years (UKPDS 91).

BACKGROUND: The 20-year UK Prospective Diabetes Study showed major clinical benefits for people with newly diagnosed type 2 diabetes randomly allocated to intensive glycaemic control with sulfonylurea or insulin therapy or metformin therapy, compared with conventional glycaemic control. 10-year post-trial follow-up identified enduring and emerging glycaemic and metformin legacy treatment effects. We aimed to determine whether these effects would wane by extending follow-up for another 14 years. METHODS: 5102 patients enrolled between 1977 and 1991, of whom 4209 (82·5%) participants were originally randomly allocated to receive either intensive glycaemic control (sulfonylurea or insulin, or if overweight, metformin) or conventional glycaemic control (primarily diet). At the end of the 20-year interventional trial, 3277 surviving participants entered a 10-year post-trial monitoring period, which ran until Sept 30, 2007. Eligible participants for this study were all surviving participants at the end of the 10-year post-trial monitoring period. An extended follow-up of these participants was done by linking them to their routinely collected National Health Service (NHS) data for another 14 years. Clinical outcomes were derived from records of deaths, hospital admissions, outpatient visits, and accident and emergency unit attendances. We examined seven prespecified aggregate clinical outcomes (ie, any diabetes-related endpoint, diabetes-related death, death from any cause, myocardial infarction, stroke, peripheral vascular disease, and microvascular disease) by the randomised glycaemic control strategy on an intention-to-treat basis using Kaplan-Meier time-to-event and log-rank analyses. This study is registered with the ISRCTN registry, number ISRCTN75451837. FINDINGS: Between Oct 1, 2007, and Sept 30, 2021, 1489 (97·6%) of 1525 participants could be linked to routinely collected NHS administrative data. Their mean age at baseline was 50·2 years (SD 8·0), and 41·3% were female. The mean age of those still alive as of Sept 30, 2021, was 79·9 years (SD 8·0). Individual follow-up from baseline ranged from 0 to 42 years, median 17·5 years (IQR 12·3-26·8). Overall follow-up increased by 21%, from 66 972 to 80 724 person-years. For up to 24 years after trial end, the glycaemic and metformin legacy effects showed no sign of waning. Early intensive glycaemic control with sulfonylurea or insulin therapy, compared with conventional glycaemic control, showed overall relative risk reductions of 10% (95% CI 2-17; p=0·015) for death from any cause, 17% (6-26; p=0·002) for myocardial infarction, and 26% (14-36; p<0·0001) for microvascular disease. Corresponding absolute risk reductions were 2·7%, 3·3%, and 3·5%, respectively. Early intensive glycaemic control with metformin therapy, compared with conventional glycaemic control, showed overall relative risk reductions of 20% (95% CI 5-32; p=0·010) for death from any cause and 31% (12-46; p=0·003) for myocardial infarction. Corresponding absolute risk reductions were 4·9% and 6·2%, respectively. No significant risk reductions during or after the trial for stroke or peripheral vascular disease were observed for both intensive glycaemic control groups, and no significant risk reduction for microvascular disease was observed for metformin therapy. INTERPRETATION: Early intensive glycaemic control with sulfonylurea or insulin, or with metformin, compared with conventional glycaemic control, appears to confer a near-lifelong reduced risk of death and myocardial infarction. Achieving near normoglycaemia immediately following diagnosis might be essential to minimise the lifetime risk of diabetes-related complications to the greatest extent possible. FUNDING: University of Oxford Nuffield Department of Population Health Pump Priming.

Biomarkers to improve functional outcome prediction after ischemic stroke: Results from the SICFAIL, STRAWINSKI, and PREDICT studies.

BACKGROUND AND AIMS: Acute ischemic stroke (AIS) outcome prognostication remains challenging despite available prognostic models. We investigated whether a biomarker panel improves the predictive performance of established prognostic scores. METHODS: We investigated the improvement in discrimination, calibration, and overall performance by adding five biomarkers (procalcitonin, copeptin, cortisol, mid-regional pro-atrial natriuretic peptide (MR-proANP), and N-terminal pro-B-type natriuretic peptide (NT-proBNP)) to the Acute Stroke Registry and Analysis of Lausanne (ASTRAL) and age/NIHSS scores using data from two prospective cohort studies (SICFAIL, PREDICT) and one clinical trial (STRAWINSKI). Poor outcome was defined as mRS > 2 at 12 (SICFAIL, derivation dataset) or 3 months (PREDICT/STRAWINSKI, pooled external validation dataset). RESULTS: Among 412 SICFAIL participants (median age 70 years, quartiles 59-78; 63% male; median NIHSS score 3, quartiles 1-5), 29% had a poor outcome. Area under the curve of the ASTRAL and age/NIHSS were 0.76 (95% CI 0.71-0.81) and 0.77 (95% CI 0.73-0.82), respectively. Copeptin (0.79, 95% CI 0.74-0.84), NT-proBNP (0.80, 95% CI 0.76-0.84), and MR-proANP (0.79, 95% CI 0.75-0.84) significantly improved ASTRAL score's discrimination, calibration, and overall performance. Copeptin improved age/NIHSS model's discrimination, copeptin, MR-proANP, and NT-proBNP improved its calibration and overall performance. In the validation dataset (450 patients, median age 73 years, quartiles 66-81; 54% men; median NIHSS score 8, quartiles 3-14), copeptin was independently associated with various definitions of poor outcome and also mortality. Copeptin did not increase model's discrimination but it did improve calibration and overall model performance. DISCUSSION: Copeptin, NT-proBNP, and MR-proANP improved modest but consistently the predictive performance of established prognostic scores in patients with mild AIS. Copeptin was most consistently associated with poor outcome in patients with moderate to severe AIS, although its added prognostic value was less obvious.

The circulating proteome and brain health: Mendelian randomisation and cross-sectional analyses.

Decline in cognitive function is the most feared aspect of ageing. Poorer midlife cognitive function is associated with increased dementia and stroke risk. The mechanisms underlying variation in cognitive function are uncertain. Here, we assessed associations between 1160 proteins' plasma levels and two measures of cognitive function, the digit symbol substitution test (DSST) and the Montreal Cognitive Assessment in 1198 PURE-MIND participants. We identified five DSST performance-associated proteins (NCAN, BCAN, CA14, MOG, CDCP1), with NCAN and CDCP1 showing replicated association in an independent cohort, GS (N = 1053). MRI-assessed structural brain phenotypes partially mediated (8-19%) associations between NCAN, BCAN, and MOG, and DSST performance. Mendelian randomisation analyses suggested higher CA14 levels might cause larger hippocampal volume and increased stroke risk, whilst higher CDCP1 levels might increase intracranial aneurysm risk. Our findings highlight candidates for further study and the potential for drug repurposing to reduce the risk of stroke and cognitive decline.

Plasma fibrinogen and risk of vascular recurrence after ischaemic stroke: An individual participant and summary-level data meta-analysis of 11 prospective studies.

INTRODUCTION: Inflammation is an emerging target for secondary prevention after stroke and randomised trials of anti-inflammatory therapies are ongoing. Fibrinogen, a putative pro-inflammatory marker, is associated with first stroke, but its association with major adverse cardiovascular events (MACE) after stroke is unclear. MATERIALS AND METHODS: We did a systematic review investigating the association between fibrinogen and post-stroke vascular recurrence. Authors were invited to provide individual-participant data (IPD) and where available we did within-study multivariable analyses with adjustment for cardiovascular risk factors and medications. Adjusted summary-level data was extracted from published reports from studies that did not provide IPD. We pooled risk ratios (RR) by random-effects meta-analysis by comparing supra-median with sub-median fibrinogen levels and performed pre-specified subgroup analysis according to timing of phlebotomy after the index event. RESULTS: Eleven studies were included (14,002 patients, 42,800 follow-up years), of which seven provided IPD. Fibrinogen was associated with recurrent MACE on unadjusted (RR 1.35, 95% CI 1.17-1.57, supra-median vs sub-median) and adjusted models (RR 1.21, 95% CI 1.06-1.38). Fibrinogen was associated with recurrent stroke on univariate analysis (RR 1.19, 95% CI 1.03-1.39), but not after adjustment (RR 1.11, 95% CI 0.94-1.31). The association with recurrent MACE was consistently observed in patients with post-acute (⩾14 days) fibrinogen measures (RR 1.29, 95% CI 1.16-1.45), but not in those with early phlebotomy (<14 days) (RR 0.98, 95% CI 0.82-1.18) (Pinteraction = 0.01). Similar associations were observed for recurrent stroke. DISCUSSION AND CONCLUSION: Fibrinogen was independently associated with recurrence after stroke, but the association was modified by timing of phlebotomy. Fibrinogen measurements might be useful to identify patients who are more likely to derive benefit from anti-inflammatory therapies after stroke.

Impact of vaccination on the association of COVID-19 with cardiovascular diseases: An OpenSAFELY cohort study.

Infection with SARS-CoV-2 is associated with an increased risk of arterial and venous thrombotic events, but the implications of vaccination for this increased risk are uncertain. With the approval of NHS England, we quantified associations between COVID-19 diagnosis and cardiovascular diseases in different vaccination and variant eras using linked electronic health records for ~40% of the English population. We defined a 'pre-vaccination' cohort (18,210,937 people) in the wild-type/Alpha variant eras (January 2020-June 2021), and 'vaccinated' and 'unvaccinated' cohorts (13,572,399 and 3,161,485 people respectively) in the Delta variant era (June-December 2021). We showed that the incidence of each arterial thrombotic, venous thrombotic and other cardiovascular outcomes was substantially elevated during weeks 1-4 after COVID-19, compared with before or without COVID-19, but less markedly elevated in time periods beyond week 4. Hazard ratios were higher after hospitalised than non-hospitalised COVID-19 and higher in the pre-vaccination and unvaccinated cohorts than the vaccinated cohort. COVID-19 vaccination reduces the risk of cardiovascular events after COVID-19 infection. People who had COVID-19 before or without being vaccinated are at higher risk of cardiovascular events for at least two years.

A comprehensive analysis of APOE genotype effects on human brain structure in the UK Biobank.

Alzheimer's disease (AD) risk is increased in carriers of the apolipoprotein E (APOE) ε4 allele and decreased in ε2 allele carriers compared with the ε3ε3 genotype. The aim of this study was to determine whether: the APOE genotype affects brain grey (GM) or white matter (WM) structure; and if differences exist, the age when they become apparent and whether there are differential effects by sex. We used cross-sectional magnetic resonance imaging data from ~43,000 (28,494 after pre-processing) white British cognitively healthy participants (7,446 APOE ε4 carriers) aged 45-80 years from the UK Biobank cohort and investigated image-derived phenotypes (IDPs). We observed no statistically significant effects of APOE genotype on GM structure volumes or median T2* in subcortical structures, a measure related to iron content. The volume of white matter hyperintensities differed significantly between APOE genotype groups with higher volumes in APOE ε4ε4 (effect size 0.14 standard deviations [SD]) and ε3ε4 carriers (effect size 0.04 SD) but no differences in ε2 carriers compared with ε3ε3 carriers. WM integrity measures in the dorsal (mean diffusivity [MD]) and ventral cingulum (MD and intracellular volume fraction), posterior thalamic radiation (MD and isotropic volume fraction) and sagittal stratum (MD) indicated lower integrity in APOE ε4ε4 carriers (effect sizes around 0.2-0.3 SD) and ε3ε4 (effect sizes around 0.05 SD) carriers but no differences in ε2 carriers compared with the APOE ε3ε3 genotype. Effects did not differ between men and women. APOE ε4 homozygotes had lower WM integrity specifically at older ages with a steeper decline of WM integrity from the age of 60 that corresponds to around 5 years greater "brain age". APOE genotype affects various white matters measures, which might be indicative of preclinical AD processes. This hypothesis can be assessed in future when clinical outcomes become available.

Mortality Outcomes in a Large Population with and without Covert Cerebrovascular Disease.

Covert cerebrovascular disease (CCD) is frequently reported on neuroimaging and associates with increased dementia and stroke risk. We aimed to determine how incidentally-discovered CCD during clinical neuroimaging in a large population associates with mortality. We screened CT and MRI reports of adults aged ≥50 in the Kaiser Permanente Southern California health system who underwent neuroimaging for a non-stroke clinical indication from 2009-2019. Natural language processing identified incidental covert brain infarcts (CBI) and/or white matter hyperintensities (WMH), grading WMH as mild/moderate/severe. Models adjusted for age, sex, ethnicity, multimorbidity, vascular risks, depression, exercise, and imaging modality. Of n=241,028, the mean age was 64.9 (SD=10.4); mean follow-up 4.46 years; 178,554 (74.1%) had CT; 62,474 (25.9%) had MRI; 11,328 (4.7%) had CBI; and 69,927 (29.0%) had WMH. The mortality rate per 1,000 person-years with CBI was 59.0 (95%CI 57.0-61.1); with WMH=46.5 (45.7-47.2); with neither=17.4 (17.1-17.7). In adjusted models, mortality risk associated with CBI was modified by age, e.g. HR 1.34 [1.21-1.48] at age 56.1 years vs HR 1.22 [1.17-1.28] at age 72 years. Mortality associated with WMH was modified by both age and imaging modality e.g., WMH on MRI at age 56.1 HR = 1.26 [1.18-1.35]; WMH on MRI at age 72 HR 1.15 [1.09-1.21]; WMH on CT at age 56.1 HR 1.41 [1.33-1.50]; WMH on CT at age 72 HR 1.28 [1.24-1.32], vs. patients without CBI or without WMH, respectively. Increasing WMH severity associated with higher mortality, e.g. mild WMH on MRI had adjusted HR=1.13 [1.06-1.20] while severe WMH on CT had HR=1.45 [1.33-1.59]. Incidentally-detected CBI and WMH on population-based clinical neuroimaging can predict higher mortality rates. We need treatments and healthcare planning for individuals with CCD.

Legacy benefits of blood pressure treatment on cardiovascular events are primarily mediated by improved blood pressure variability: the ASCOT trial.

BACKGROUND AND AIMS: Visit-to-visit systolic blood pressure variability (BPV) is an important predictor of cardiovascular (CV) outcomes. The long-term effect of a period of blood pressure (BP) control, but with differential BPV, is uncertain. Morbidity and mortality follow-up of UK participants in the Anglo-Scandinavian Cardiac Outcomes Trial-Blood Pressure-Lowering Arm has been extended for up to 21 years to determine the CV impact of mean systolic blood pressure (SBP) control and BPV during the trial, and amongst those allocated to amlodipine- and atenolol-based treatment. METHODS: Eight thousand five hundred and eighty hypertensive participants (4305 assigned to amlodipine ± perindopril-based and 4275 to atenolol ± diuretic-based treatment during the in-trial period (median 5.5 years) were followed for up to 21 years (median 17.4 years), using linked hospital and mortality records. A subgroup of participants (n = 2156) was followed up 6 years after the trial closure with a self-administered questionnaire and a clinic visit. In-trial mean SBP and standard deviation of visit-to-visit SBP as a measure of BPV, were measured using >100 000 BP measurements. Cox proportional hazard models were used to estimate the risk [hazard ratios (HRs)], associated with (i) mean with SBP and BPV during the in-trial period, for the CV endpoints occurring after the end of the trial and (ii) randomly assigned treatment to events following randomization, for the first occurrence of pre-specified CV outcomes. RESULTS: Using BP data from the in-trial period, in the post-trial period, although mean SBP was a predictor of CV outcomes {HR per 10 mmHg, 1.14 [95% confidence interval (CI) 1.10-1.17], P < .001}, systolic BPV independent of mean SBP was a strong predictor of CV events [HR per 5 mmHg 1.22 (95% CI 1.18-1.26), P < .001] and predicted events even in participants with well-controlled BP. During 21-year follow-up, those on amlodipine-based compared with atenolol-based in-trial treatment had significantly reduced risk of stroke [HR 0.82 (95% CI 0.72-0.93), P = .003], total CV events [HR 0.93 (95% CI 0.88-0.98), P = .008], total coronary events [HR 0.92 (95% CI 0.86-0.99), P = .024], and atrial fibrillation [HR 0.91 (95% CI 0.83-0.99), P = .030], with weaker evidence of a difference in CV mortality [HR 0.91 (95% CI 0.82-1.01), P = .073]. There was no significant difference in the incidence of non-fatal myocardial infarction and fatal coronary heart disease, heart failure, and all-cause mortality. CONCLUSIONS: Systolic BPV is a strong predictor of CV outcome, even in those with controlled SBP. The long-term benefits of amlodipine-based treatment compared with atenolol-based treatment in reducing CV events appear to be primarily mediated by an effect on systolic BPV during the trial period.

Cognitive performance following stroke, transient ischaemic attack, myocardial infarction, and hospitalisation: an individual participant data meta-analysis of six randomised controlled trials.

BACKGROUND: Survivors of stroke are often concerned about cognitive problems, and information on the risk of cognitive problems often comes from small studies. We aimed to estimate years of cognitive ageing associated with stroke compared with transient ischaemic attack, myocardial infarction, and other hospitalisations in a large population. METHODS: Using data from six randomised controlled trials (ORIGIN, ONTARGET, TRANSCEND, COMPASS, HOPE-3, and NAVIGATE ESUS), we completed an individual participant data meta-analysis using data requested from the Public Health Research Institute to estimate the association of stroke (by type and severity), transient ischaemic attack, myocardial infarction, and other hospitalisations with cognitive performance measured at the end of each trial. We included participants in any of these randomised controlled trials with a cognitive assessment at baseline and at least one other timepoint. Cognitive performance was measured with the Mini-Mental State Examination or the Montreal Cognitive Assessment, transformed into Z scores. We estimated Z score differences in end of trial cognitive performance between people with and without events and calculated corresponding years of cognitive ageing in these trials, and additionally calculated using a population representative cohort-the Cognitive Function and Ageing Study. FINDINGS: In 64 106 participants from 55 countries, compared with no event, stroke was associated with 18 years of cognitive ageing (1487 strokes included in the model, 95% CI 10 to 28; p<0·0001) and transient ischaemic attack with 3 years (660 transient ischaemic attacks included in the model, 0 to 6; p=0·021). Myocardial infarction (p=0·60) and other hospitalisations (p=0·26) were not associated with cognitive ageing. The mean difference in SD compared with people without an event was -0·84 (95% CI -0·91 to -0·76; p<0·0001) for disabling stroke, and -0·12 (-0·19 to -0·05; p=0·0012) for non-disabling stroke. Haemorrhagic stroke was associated with worse cognition (-0·75, -0·95 to -0·55; p<0·0001) than ischaemic stroke (-0·42, -0·48 to -0·36; p <0·0001). INTERPRETATION: Stroke has a substantial effect on cognition. The effects of transient ischaemic attack were small, whereas myocardial infarction and hospitalisation had a neutral effect. Prevention of stroke could lead to a reduction in cognitive ageing in those at greatest risk. FUNDING: Population Health Research Institute and Chief Scientist Office of Scotland.

Understanding the performance and reliability of NLP tools: a comparison of four NLP tools predicting stroke phenotypes in radiology reports.

Background: Natural language processing (NLP) has the potential to automate the reading of radiology reports, but there is a need to demonstrate that NLP methods are adaptable and reliable for use in real-world clinical applications. Methods: We tested the F1 score, precision, and recall to compare NLP tools on a cohort from a study on delirium using images and radiology reports from NHS Fife and a population-based cohort (Generation Scotland) that spans multiple National Health Service health boards. We compared four off-the-shelf rule-based and neural NLP tools (namely, EdIE-R, ALARM+, ESPRESSO, and Sem-EHR) and reported on their performance for three cerebrovascular phenotypes, namely, ischaemic stroke, small vessel disease (SVD), and atrophy. Clinical experts from the EdIE-R team defined phenotypes using labelling techniques developed in the development of EdIE-R, in conjunction with an expert researcher who read underlying images. Results: EdIE-R obtained the highest F1 score in both cohorts for ischaemic stroke, ≥93%, followed by ALARM+, ≥87%. The F1 score of ESPRESSO was ≥74%, whilst that of Sem-EHR is ≥66%, although ESPRESSO had the highest precision in both cohorts, 90% and 98%. For F1 scores for SVD, EdIE-R scored ≥98% and ALARM+ ≥90%. ESPRESSO scored lowest with ≥77% and Sem-EHR ≥81%. In NHS Fife, F1 scores for atrophy by EdIE-R and ALARM+ were 99%, dropping in Generation Scotland to 96% for EdIE-R and 91% for ALARM+. Sem-EHR performed lowest for atrophy at 89% in NHS Fife and 73% in Generation Scotland. When comparing NLP tool output with brain image reads using F1 scores, ALARM+ scored 80%, outperforming EdIE-R at 66% in ischaemic stroke. For SVD, EdIE-R performed best, scoring 84%, with Sem-EHR 82%. For atrophy, EdIE-R and both ALARM+ versions were comparable at 80%. Conclusions: The four NLP tools show varying F1 (and precision/recall) scores across all three phenotypes, although more apparent for ischaemic stroke. If NLP tools are to be used in clinical settings, this cannot be performed "out of the box." It is essential to understand the context of their development to assess whether they are suitable for the task at hand or whether further training, re-training, or modification is required to adapt tools to the target task.

Antiplatelet therapy and incident cognitive impairment or dementia-a systematic review and meta-analysis of randomised clinical trials.

OBJECTIVE: The benefit of antiplatelet therapy in preventing cognitive impairment or dementia is uncertain. We investigated the association between antiplatelet therapy and incident cognitive impairment or dementia in randomised clinical trials. METHODS: We searched PubMed, EMBASE and CENTRAL for randomised clinical trials published from database inception through 1 February 2023. Trials that evaluated the association of antiplatelet therapy with incident cognitive impairment or dementia were included. For single-agent antiplatelet, the control group was placebo. For dual agent antiplatelet therapy, the control group was single-agent monotherapy. A random-effects meta-analysis model was used to report pooled treatment effects and 95% confidence intervals (CIs). The primary outcome was incident cognitive impairment or dementia. Secondary outcomes included change in cognitive test scores. RESULTS: A total of 11 randomised clinical trials were included (109,860 participants). All reported the incidence of cognitive impairment or dementia on follow-up. The mean (SD) age of trial participants was 66.2 (7.9) years. Antiplatelet therapy was not significantly associated with a reduced risk of cognitive impairment or dementia (11 trials; 109,860 participants) (3.49% versus 4.18% of patients over a mean trial follow-up of 5.8 years; odds ratio [OR], 0.94 [95% CI, 0.88-1.00]; absolute risk reduction, 0.2% [95% CI, -0.4% to 0.009%]; I2 = 0.0%). Antiplatelet therapy was not significantly associated with mean change in cognitive test scores. CONCLUSION: In this meta-analysis, antiplatelet therapy was not significantly associated with a lower risk of incident cognitive impairment or dementia, but the CIs around this outcome do not exclude a modest preventative effect.

World-wide variations in tests of cognition and activities of daily living in participants of six international randomized controlled trials.

Background: Better understanding of worldwide variation in simple tests of cognition and global function in older adults would aid the delivery and interpretation of multi-national studies of the prevention of dementia and functional decline. Method: In six RCTs that measured cognition with the mini-mental state examination (MMSE), Montreal cognitive assessment (MoCA), and activities of daily living (ADL) with the Standardised Assessment of Everyday Global Activities (SAGEA), we estimated average scores by global region with multilevel mixed-effects models. We estimated the proportion of participants with cognitive or functional impairment with previously defined thresholds (MMSE≤24 or MoCA≤25, SAGEA≥7), and with a country-standardised z-score threshold of cognitive or functional score of ≤-1. Results: In 91,396 participants (mean age 66.6 years [SD 7.8], 31% females) from seven world regions, all global regions differed significantly in estimated cognitive function (z-score differences 0.11-0.45, p<0.001) after accounting for individual-level factors, centre and study. In different regions, the proportion of trial participants with MMSE≤24 or MoCA≤25 ranged from 23-36%; the proportion below a country-standardised z-score threshold of ≤1 ranged from 10-14%. The differences in prevalence of impaired IADL (SAGEA≥7) ranged from 2-6% and by country-standardised thresholds from 3-6%. Conclusions: Accounting for country-level factors reduced large differences between world regions in estimates of cognitive impairment. Measures of IADL were less variable across world regions, and could be used to better estimate dementia prevalence in large studies.

Noninvasive In Vivo Thrombus Imaging in Patients With Ischemic Stroke or Transient Ischemic Attack-Brief Report.

BACKGROUND: 18F-GP1 is a novel positron-emitting radiotracer that is highly specific for activated platelets and thrombus. In a proof-of-concept study, we aimed to determine its potential clinical application in establishing the role and origin of thrombus in ischemic stroke. METHODS: Eleven patients with recent ischemic stroke (n=9) or transient ischemic attack (n=2) underwent 18F-GP1 positron emission tomography and computed tomography angiography at a median of 11 (range, 2-21) days from symptom onset. 18F-GP1 uptake (maximum target-to-background ratio) was assessed in the carotid arteries and brain. RESULTS: 18F-GP1 uptake was identified in 10 of 11 patients: 4 in the carotid arteries only, 3 in the brain only, and 3 in both the brain and carotid arteries. In those with carotid uptake, 4 participants had >50% stenosis and 3 had nonstenotic disease. One case had bilateral stenotic disease (>70%), but only the culprit carotid artery demonstrated 18F-GP1 uptake. The average uptake was higher in the culprit (median maximum target-to-background ratio, 1.55 [interquartile range, 1.26-1.82]) compared with the contralateral nonculprit carotid artery (maximum target-to-background ratio, 1.22 [1.19-1.6]). In those with brain 18F-GP1 uptake (maximum target-to-background ratio, 6.45 [4.89-7.65]), areas of acute infarction on computed tomography correlated with brain 18F-GP1 uptake in 6 cases. Ex vivo autoradiography of postmortem infarcted brain tissue showed focal uptake corresponding to intraluminal thrombus within the culprit vessel and downstream microvasculature. There was also evidence of diffuse uptake within some of the infarcted brain tissue reflecting parenchymal petechial hemorrhage. CONCLUSIONS: 18F-GP1 positron emission tomography and computed tomography angiography is a novel noninvasive method of identifying in vivo cerebrovascular thrombosis, which holds major promise in understanding the role and origin of thrombosis in stroke. REGISTRATION: URL: https://www. CLINICALTRIALS: gov; Unique identifier: NCT03943966.

Ontology-driven and weakly supervised rare disease identification from clinical notes.

BACKGROUND: Computational text phenotyping is the practice of identifying patients with certain disorders and traits from clinical notes. Rare diseases are challenging to be identified due to few cases available for machine learning and the need for data annotation from domain experts. METHODS: We propose a method using ontologies and weak supervision, with recent pre-trained contextual representations from Bi-directional Transformers (e.g. BERT). The ontology-driven framework includes two steps: (i) Text-to-UMLS, extracting phenotypes by contextually linking mentions to concepts in Unified Medical Language System (UMLS), with a Named Entity Recognition and Linking (NER+L) tool, SemEHR, and weak supervision with customised rules and contextual mention representation; (ii) UMLS-to-ORDO, matching UMLS concepts to rare diseases in Orphanet Rare Disease Ontology (ORDO). The weakly supervised approach is proposed to learn a phenotype confirmation model to improve Text-to-UMLS linking, without annotated data from domain experts. We evaluated the approach on three clinical datasets, MIMIC-III discharge summaries, MIMIC-III radiology reports, and NHS Tayside brain imaging reports from two institutions in the US and the UK, with annotations. RESULTS: The improvements in the precision were pronounced (by over 30% to 50% absolute score for Text-to-UMLS linking), with almost no loss of recall compared to the existing NER+L tool, SemEHR. Results on radiology reports from MIMIC-III and NHS Tayside were consistent with the discharge summaries. The overall pipeline processing clinical notes can extract rare disease cases, mostly uncaptured in structured data (manually assigned ICD codes). CONCLUSION: The study provides empirical evidence for the task by applying a weakly supervised NLP pipeline on clinical notes. The proposed weak supervised deep learning approach requires no human annotation except for validation and testing, by leveraging ontologies, NER+L tools, and contextual representations. The study also demonstrates that Natural Language Processing (NLP) can complement traditional ICD-based approaches to better estimate rare diseases in clinical notes. We discuss the usefulness and limitations of the weak supervision approach and propose directions for future studies.

C-Reactive Protein, Interleukin-6, and Vascular Recurrence After Stroke: An Individual Participant Data Meta-Analysis.

BACKGROUND: Anti-inflammatory therapies reduce recurrent vascular events in coronary disease. Existing studies have reported highly conflicting findings for the association of blood inflammatory markers with vascular recurrence after stroke leading to uncertainty about the potential of anti-inflammatory therapies after stroke and no consensus about the utility of measurement of inflammatory markers in current guidelines. METHODS: We investigated the association between hsCRP (high-sensitivity C-reactive protein), IL-6 (interluekin-6), and recurrent major adverse cardiovascular events (MACE), and stroke from individual participant data from 8420 patients with ischemic stroke/transient ischemic attack from 10 prospective studies. We did within-study multivariable regression analyses and then combined adjusted risk ratio (RR) by random-effects meta-analysis. RESULTS: During 18 920 person-years of follow-up, 1407 (16.7% [95% CI, 15.9-17.5]) patients had MACE and 1191 (14.1% [95% CI, 13.4-14.9]) patients had recurrent stroke. On bivariate analysis, baseline IL-6 was associated with MACE (RR, 1.26 [95% CI, 1.10-1.43]) and recurrent stroke (RR, 1.18 [95% CI, 1.05-1.32]), per unit increase logeIL-6. Similar associations were observed for hsCRP (MACE RR, 1.19 [95% CI, 1.09-1.29]; recurrent stroke RR, 1.12 [95% CI, 1.04-1.21], per unit increase logehsCRP). After adjustment for vascular risk factors and treatment, independent associations remained with MACE (IL-6, RR, 1.12 [95% CI, 1.04-1.21]; hsCRP, RR, 1.09 [95% CI, 1.04-1.15]) and recurrent stroke (IL-6, RR, 1.09 [95% CI, 1.00-1.19]; hsCRP, RR, 1.05 [95% CI, 1.00-1.11]). Comparing the top with the bottom quarters (Q4 versus Q1), IL-6 (RR, 1.35 [95% CI, 1.09-1.67]) and hsCRP (RR, 1.31 [95% CI, 1.07-1.61]) were associated with MACE after adjustment. Similar results were observed for recurrent stroke for IL-6 (RR, 1.33 [95% CI, 1.08-1.65]) but not hsCRP (RR, 1.16 [95% CI, 0.93-1.43]). CONCLUSIONS: Blood markers of inflammation were independently associated with vascular recurrence after stroke, strengthening the rationale for randomized trials of anti-inflammatory therapies for secondary prevention after ischemic stroke/TIA.

New Insights into Stroke from Continuous Passively Collected Temperature and Sleep Data Using Wrist-Worn Wearables.

Actigraphy may provide new insights into clinical outcomes and symptom management of patients through passive, continuous data collection. We used the GENEActiv smartwatch to passively collect actigraphy, wrist temperature, and ambient light data from 27 participants after stroke or probable brain transient ischemic attack (TIA) over 42 periods of device wear. We computed 323 features using established algorithms and proposed 25 novel features to characterize sleep and temperature. We investigated statistical associations between the extracted features and clinical outcomes evaluated using clinically validated questionnaires to gain insight into post-stroke recovery. We subsequently fitted logistic regression models to replicate clinical diagnosis (stroke or TIA) and disability due to stroke. The model generalization performance was assessed using a leave-one-subject-out cross validation method with the selected feature subsets, reporting the area under the curve (AUC). We found that several novel features were strongly correlated (|r|>0.3) with stroke symptoms and mental health measures. Using selected novel features, we obtained an AUC of 0.766 to estimate diagnosis and an AUC of 0.749 to estimate whether disability due to stroke was present. Collectively, these findings suggest that features extracted from the temperature smartwatch sensor may reveal additional clinically useful information over and above existing actigraphy-based features.

Harmonising electronic health records for reproducible research: challenges, solutions and recommendations from a UK-wide COVID-19 research collaboration.

BACKGROUND: The CVD-COVID-UK consortium was formed to understand the relationship between COVID-19 and cardiovascular diseases through analyses of harmonised electronic health records (EHRs) across the four UK nations. Beyond COVID-19, data harmonisation and common approaches enable analysis within and across independent Trusted Research Environments. Here we describe the reproducible harmonisation method developed using large-scale EHRs in Wales to accommodate the fast and efficient implementation of cross-nation analysis in England and Wales as part of the CVD-COVID-UK programme. We characterise current challenges and share lessons learnt. METHODS: Serving the scope and scalability of multiple study protocols, we used linked, anonymised individual-level EHR, demographic and administrative data held within the SAIL Databank for the population of Wales. The harmonisation method was implemented as a four-layer reproducible process, starting from raw data in the first layer. Then each of the layers two to four is framed by, but not limited to, the characterised challenges and lessons learnt. We achieved curated data as part of our second layer, followed by extracting phenotyped data in the third layer. We captured any project-specific requirements in the fourth layer. RESULTS: Using the implemented four-layer harmonisation method, we retrieved approximately 100 health-related variables for the 3.2 million individuals in Wales, which are harmonised with corresponding variables for > 56 million individuals in England. We processed 13 data sources into the first layer of our harmonisation method: five of these are updated daily or weekly, and the rest at various frequencies providing sufficient data flow updates for frequent capturing of up-to-date demographic, administrative and clinical information. CONCLUSIONS: We implemented an efficient, transparent, scalable, and reproducible harmonisation method that enables multi-nation collaborative research. With a current focus on COVID-19 and its relationship with cardiovascular outcomes, the harmonised data has supported a wide range of research activities across the UK.

A survey on clinical natural language processing in the United Kingdom from 2007 to 2022.

Much of the knowledge and information needed for enabling high-quality clinical research is stored in free-text format. Natural language processing (NLP) has been used to extract information from these sources at scale for several decades. This paper aims to present a comprehensive review of clinical NLP for the past 15 years in the UK to identify the community, depict its evolution, analyse methodologies and applications, and identify the main barriers. We collect a dataset of clinical NLP projects (n = 94; £ = 41.97 m) funded by UK funders or the European Union's funding programmes. Additionally, we extract details on 9 funders, 137 organisations, 139 persons and 431 research papers. Networks are created from timestamped data interlinking all entities, and network analysis is subsequently applied to generate insights. 431 publications are identified as part of a literature review, of which 107 are eligible for final analysis. Results show, not surprisingly, clinical NLP in the UK has increased substantially in the last 15 years: the total budget in the period of 2019-2022 was 80 times that of 2007-2010. However, the effort is required to deepen areas such as disease (sub-)phenotyping and broaden application domains. There is also a need to improve links between academia and industry and enable deployments in real-world settings for the realisation of clinical NLP's great potential in care delivery. The major barriers include research and development access to hospital data, lack of capable computational resources in the right places, the scarcity of labelled data and barriers to sharing of pretrained models.

Automated clinical coding: what, why, and where we are?

Clinical coding is the task of transforming medical information in a patient's health records into structured codes so that they can be used for statistical analysis. This is a cognitive and time-consuming task that follows a standard process in order to achieve a high level of consistency. Clinical coding could potentially be supported by an automated system to improve the efficiency and accuracy of the process. We introduce the idea of automated clinical coding and summarise its challenges from the perspective of Artificial Intelligence (AI) and Natural Language Processing (NLP), based on the literature, our project experience over the past two and half years (late 2019-early 2022), and discussions with clinical coding experts in Scotland and the UK. Our research reveals the gaps between the current deep learning-based approach applied to clinical coding and the need for explainability and consistency in real-world practice. Knowledge-based methods that represent and reason the standard, explainable process of a task may need to be incorporated into deep learning-based methods for clinical coding. Automated clinical coding is a promising task for AI, despite the technical and organisational challenges. Coders are needed to be involved in the development process. There is much to achieve to develop and deploy an AI-based automated system to support coding in the next five years and beyond.

Association of COVID-19 With Major Arterial and Venous Thrombotic Diseases: A Population-Wide Cohort Study of 48 Million Adults in England and Wales.

BACKGROUND: Infection with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) induces a prothrombotic state, but long-term effects of COVID-19 on incidence of vascular diseases are unclear. METHODS: We studied vascular diseases after COVID-19 diagnosis in population-wide anonymized linked English and Welsh electronic health records from January 1 to December 7, 2020. We estimated adjusted hazard ratios comparing the incidence of arterial thromboses and venous thromboembolic events (VTEs) after diagnosis of COVID-19 with the incidence in people without a COVID-19 diagnosis. We conducted subgroup analyses by COVID-19 severity, demographic characteristics, and previous history. RESULTS: Among 48 million adults, 125 985 were hospitalized and 1 319 789 were not hospitalized within 28 days of COVID-19 diagnosis. In England, there were 260 279 first arterial thromboses and 59 421 first VTEs during 41.6 million person-years of follow-up. Adjusted hazard ratios for first arterial thrombosis after COVID-19 diagnosis compared with no COVID-19 diagnosis declined from 21.7 (95% CI, 21.0-22.4) in week 1 after COVID-19 diagnosis to 1.34 (95% CI, 1.21-1.48) during weeks 27 to 49. Adjusted hazard ratios for first VTE after COVID-19 diagnosis declined from 33.2 (95% CI, 31.3-35.2) in week 1 to 1.80 (95% CI, 1.50-2.17) during weeks 27 to 49. Adjusted hazard ratios were higher, for longer after diagnosis, after hospitalized versus nonhospitalized COVID-19, among Black or Asian versus White people, and among people without versus with a previous event. The estimated whole-population increases in risk of arterial thromboses and VTEs 49 weeks after COVID-19 diagnosis were 0.5% and 0.25%, respectively, corresponding to 7200 and 3500 additional events, respectively, after 1.4 million COVID-19 diagnoses. CONCLUSIONS: High relative incidence of vascular events soon after COVID-19 diagnosis declines more rapidly for arterial thromboses than VTEs. However, incidence remains elevated up to 49 weeks after COVID-19 diagnosis. These results support policies to prevent severe COVID-19 by means of COVID-19 vaccines, early review after discharge, risk factor control, and use of secondary preventive agents in high-risk patients.