Clusters of post-acute COVID-19 symptoms: a latent class analysis across 9 databases and 7 countries (preprint)

Prior evidence has suggested the multisystem symptomatic manifestations of post-acute COVID-19 condition (PCC). Here we conducted a network cluster analysis of 24 WHO proposed symptoms to identify potential latent subclasses of PCC. Individuals with a positive test of or diagnosed with SARS-CoV-2 after 09/2020 and with at least one symptom within ≥ 90 to 365 days following infection were included. Sub-analyses were conducted among people with ≥ 3 different symptoms. Summary characteristics were provided for each cluster. All analyses were conducted separately in 9 databases from 7 countries, including data from primary care, hospitals, national health claims and national health registries, allowing to validate clusters across the different healthcare settings. 787,078 persons with PCC were included. Single-symptom clusters were common across all databases, particularly for joint pain, anxiety, depression and allergy. Complex clusters included anxiety-depression and abdominal-gastrointestinal symptoms. Substantial heterogeneity within and between PCC clusters was seen across healthcare settings. Current definitions of PCC should be critically reviewed to reflect this variety in clinical presentation.

Healthy lifestyle for the prevention of post-COVID-19 multisystem sequelae, hospitalization, and death: a prospective cohort study (preprint)

Abstract Background Post-COVID complications are emerging as a global public health crisis. Effective prevention strategies are needed to inform patients, clinicians and policy makers, and to reduce their cumulative burden. We aimed to investigate whether a habitual healthy lifestyle predated pandemic is associated with lower risks of multisystem sequelae and other adverse outcomes of COVID-19, and whether the potential protective effects are independent of pre-existing comorbidities. Methods The prospective population-based cohort study enrolled participants with SARS-CoV-2 infection confirmed by a positive polymerase chain reaction test result between March 1, 2020, and March 1, 2022. Participants with no history of the related outcome one year before infection were included and followed up for 210 days. Exposures included ten modifiable healthy lifestyle factors including past or never smoking, moderate alcohol intake ([≤]4 times week), body mass index <30 kg/m2, at least 150 minutes of moderate or 75 minutes of vigorous physical activity per week, less sedentary time (<4 hours per day), healthy sleep duration (7-9 hours per day), adequate intake of fruit and vegetables ([≥]400 g/day), adequate oily fish intake ([≥]1 portion/week), moderate intake of red meat ([≤]4 portions week) and processed meat ([≤]4 portions week). Outcomes included multisystem COVID-19 sequelae (consisting of 75 diseases/symptoms in 10 organ systems), death, and hospital admission following SARS-CoV-2 infection, confirmed by hospital inpatient and death records. Risk was reported in relative scale (hazard ratio [HR]) and absolute scale (absolute risk reduction [ARR]) during both the acute (the first 30 days) and post-acute (30-210 days) phases of infection using Cox models. Findings A total of 68,896 participants (mean [SD] age, 66.6 [8.4]; 32,098 women [46.6%]) with COVID-19 were included. A favorable lifestyle (6-10 healthy lifestyle factors; 46.4%) was associated with a 36% lower risk of multisystem sequelae of COVID-19 (HR, 0.64; 95% CI, 0.58-0.69; ARR, 7.08%; 95% CI, 5.98-8.09), compared with unfavorable lifestyle (0-4 factors; 12.3%). Risk reductions were observed across all 10 prespecified organ systems including cardiovascular, coagulation, metabolic and endocrine, gastrointestinal, kidney, mental health, musculoskeletal, neurologic, and respiratory disorders, and general symptoms of fatigue and malaise. This beneficial effect was largely attributable to direct effects of healthy lifestyle, with mediation proportion ranging from 44% to 93% across organ systems. A favorable lifestyle was also associated with lower risk of post-COVID death (HR, 0.59; 95% CI, 0.52-0.66; ARR, 1.99%; 95% CI, 1.61-2.32) and hospitalization (HR, 0.78; 95% CI, 0.73-0.84; ARR, 6.14%; 95% CI, 4.48-7.68). These associations were observed after accounting for potential misclassification of lifestyle factors, and during acute and post-acute infection, in those tested positive in the hospital and community setting, and independent of vaccination status or SARS-CoV-2 variant. Interpretation Adherence to a healthy lifestyle predated pandemic was associated with substantially lower risk of complications across organ systems, death, and hospitalization following COVID-19, regardless of phases of infection, vaccination status, test setting, and SARS-CoV-2 variants, and independent of comorbidities. These findings illustrate the benefits of adhering to a healthy lifestyle to reduce the long-term adverse health consequences following SARS-CoV-2 infection.

Collateral effects of the COVID-19 pandemic on endocrine treatments for breast and prostate cancer in the UK: implications for bone health (preprint)

BackgroundThe COVID-19 pandemic affected cancer screening, diagnosis and treatment pathways. This study examined the impact of the pandemic on incidence and trends of endocrine treatments in patients with breast or prostate cancer; and endocrine treatment-related side-effects. MethodsPopulation-based cohort study using UK primary care Clinical Practice Research Datalink (CPRD) GOLD database. There were 13,701 newly diagnosed breast cancer patients and 12,221 prostate cancer patients with [≥]1-year data availability since diagnosis between January 2017-June 2022. Incidence rates (IR) and incidence rate ratios (IRR) were calculated across multiple time periods before and after lockdown to examine the impact of changing social restrictions on endocrine treatments and treatment-related outcomes, including osteopenia, osteoporosis and bisphosphonate prescriptions. ResultsIn patients with breast cancer, aromatase inhibitor prescriptions increased during lockdown compared to pre-pandemic (IRR: 1.22 [95% Confidence Interval: 1.11-1.34]), followed by a decrease post-first lockdown (IRR: 0.79 [95%CI: 0.69-0.89]). In patients with prostate cancer, first-generation antiandrogen prescriptions increased compared to pre-pandemic (IRR: 1.23 [95% CI: 1.08-1.4]). For breast cancer patients on aromatase inhibitors, diagnoses of osteopenia, osteoporosis and bisphosphonate prescriptions were reduced across all lockdown periods compared to pre-pandemic (IRR range: 0.31-0.62). ConclusionDuring the first two years of the pandemic, newly diagnosed breast and prostate cancer patients were prescribed more endocrine treatments compared to pre-pandemic, due to restrictions on hospital procedures replacing surgeries with bridging therapies. But breast cancer patients had fewer diagnoses of osteopenia and osteoporosis, and bisphosphonate prescriptions. These patients should be followed up in the coming years for signs of bone thinning. Evidence of poorer management of treatment-related side-effects will allow us to determine whether there is a need to better allocate resources to patients at high risk for bone-related complications.

The impact of the COVID-19 pandemic on short-term cancer survival in the United Kingdom: a cohort analysis (preprint)

Objectives: The COVID-19 pandemic profoundly affected healthcare systems and patients. There is a pressing need to comprehend the collateral effects of the pandemic on non-communicable diseases. Here we examined the impact of the COVID-19 pandemic on short-term cancer survival in the United Kingdom (UK). We hypothesised that short-term survival from nine cancers would be reduced during the pandemic, particularly cancers that benefit from screening and early detection (e.g., breast and colorectal cancer). Design: Population-based cohort study. Setting: Electronic health records from UK primary care Clinical Practice Research Datalink (CPRD) GOLD database. Participants: There were 12,259,744 eligible patients aged [≥]18 years with [≥]one year of prior history identified from January 2000 to December 2021. Main outcome measures: We estimated age-standardised incidence rates (IR) and short-term (one- and two-year) survival of several common cancers (breast, colorectal, head and neck, liver, lung, oesophagus, pancreatic, prostate, and stomach cancer) from 2000 to 2019 (in five-year strata) compared to 2020 to 2021 using the Kaplan-Meier method. Results: Apart from pancreatic cancer, IRs decreased for all cancers in 2020 and recovered to different extents in 2021. Short-term survival improved for most cancers between 2000 to 2019, but then declined for those diagnosed in 2020 to 2021.This was most pronounced for colorectal cancer, with one-year survival falling from 79.3% [95% confidence interval: 78.5%-80.1%] in 2015 to 2019 to 76.3% [74.6%-78.1%] for those diagnosed in 2020 to 2021. Conclusion: Short-term survival for many cancers was impacted by the management of the COVID-19 pandemic in the UK. This decline was most prominent for colorectal cancer, with reductions in survivorship equivalent to returning to mortality seen in the first decade of the 2000s. These results illustrate the need for an immediate and well-funded investment in resolving the current backlog in cancer screening and diagnostic procedures in the UK National Health Service to improve patient outcomes.

The Impact of the UK COVID-19 Lockdown on the Screening, Diagnostics and Incidence of Breast, Colorectal, Lung and Prostate Cancer in the UK: a Population-Based Cohort Study (preprint)

Objectives: This study aimed to assess the impact of the COVID-19 lockdown on the screening and diagnosis of breast, colorectal, lung, and prostate cancer. The study also investigated whether the rates returned to pre-pandemic levels by December 2021. Design: Cohort study. Setting: Electronic health records from UK primary care Clinical Practice Research Datalink (CPRD) GOLD database. Participants: The study included individuals registered with CPRD GOLD between January 2017 and December 2021, with at least 365 days of prior observation. Main outcome measures: The study focused on screening, diagnostic tests, referrals and diagnoses of first-ever breast, colorectal, lung, and prostate cancer. Incidence rates (IR) were stratified by age, sex and region, and incidence rate ratios (IRR) were calculated to compare rates during and after lockdown with the reference period before lockdown. Forecasted rates were estimated using negative binomial regression models. Results: Among 5,191,650 eligible participants, the initial lockdown resulted in reduced screening and diagnostic tests for all cancers, which remained dramatically reduced across the whole observation period for almost all tests investigated. For cancer incidence rates, there were significant IRR reductions in breast (0.69), colorectal (0.74), and prostate (0.71) cancers. However, the reduction in lung cancer incidence (0.92) was non-significant. Extrapolating to the entire UK population, an estimated 18,000 breast, 13,000 colorectal, 10,000 lung, and 21,000 prostate cancer diagnoses were missed from March 2020 to December 2021. Conclusion: The national COVID-19 lockdown in the UK had a substantial impact on cancer screening, diagnostic tests, referrals and diagnoses. Although incidence rates started to recover after the lockdown, they remained significantly lower than pre-pandemic levels for breast and prostate cancers and associated tests. Delays in diagnosis are likely to have adverse consequences on cancer stage, treatment initiation, mortality rates, and years of life lost. Urgent strategies are needed to identify undiagnosed cases and address the long-term implications of delayed diagnoses.

HLA Alleles, COVID-19 Vaccine Antibody Response and Real-World Breakthrough Outcomes (preprint)

The rapid development, approval, and global distribution of COVID-19 vaccines represent an unprecedented intervention in public health history, with over 13 billion doses administered worldwide in two years. However, our understanding of the HLA genetic underpinnings of COVID-19 vaccine-induced antibody responses and their clinical implications for breakthrough outcomes remain limited. To bridge this knowledge gap, we designed and performed a series of genetic and epidemiological analyses among 368,098 vaccinated individuals, and a subset of 194,371 participants who had antibody serology tests. Firstly, we corroborated earlier findings that SNPs associated with antibody response were predominantly located in Major Histocompatibility Complex region, and that the expansive HLA-DQB1*06 allele family was linked to better antibody responses. However, our findings contest the claim that DQB1*06 alleles alone significantly impact breakthrough risks. Additionally, our results suggest that the specific DQB1*06:04 subtype could be the true causal allele, as opposed to the previously reported DQB1*06:02. Secondly, we identified and validated six new functional HLA alleles that independently contribute to vaccine-induced antibody responses. Moreover, we unravelled additive effects of variations across multiple HLA genes that, concurrently, change the risk of clinically relevant breakthrough COVID-19 outcomes. Finally, we detangled the overall vaccine effectiveness and showed that antibody positivity accounts for approximately 20% protection against breakthrough infection and 50% against severe outcomes. These novel findings provide robust population evidence demonstrating how variations within HLA genes strongly, collectively, and causally influence vaccine-induced antibody responses, and the risk of COVID-19 breakthrough infection and related outcomes, with implications for subsequent functional research and personalised vaccination.

The role of COVID-19 vaccines in preventing post COVID-19 thromboembolic and cardiovascular complications: a multinational cohort study (preprint)

ImportanceThe overall effects of vaccination on the risk of cardiac, and venous and arterial thromboembolic complications following COVID-19 remain unclear. ObjectiveWe studied the association between COVID-19 vaccination and the risk of acute and subacute COVID-19 cardiac and thromboembolic complications. DesignMultinational staggered cohort study, based on national vaccination campaign rollouts. SettingNetwork study using electronic health records from primary care records from the UK, primary care data linked to hospital data from Spain, and national insurance claims from Estonia. ParticipantsAll adults with a prior medical history of [≥]180 days, with no history of COVID-19 or previous COVID-19 vaccination at the beginning of vaccine rollout were eligible. ExposureVaccination status was used as a time-varying exposure. Vaccinated individuals were classified by vaccine brand according to the first dose received. Main OutcomesPost COVID-19 complications including myocarditis, pericarditis, arrhythmia, heart failure (HF), venous (VTE) and arterial thromboembolism (ATE) up to 1 year after SARS-CoV-2 infection. MeasuresPropensity Score overlap weighting and empirical calibration based on negative control outcomes were used to minimise bias due to observed and unobserved confounding, respectively. Fine-Gray models were fitted to estimate sub-distribution Hazard Ratios (sHR) for each outcome according to vaccination status. Random effect meta-analyses were conducted across staggered cohorts and databases. ResultsOverall, 10.17 million vaccinated and 10.39 million unvaccinated people were included. Vaccination was consistently associated with reduced risks of acute (30-day) and subacute post COVID-19 VTE and HF: e.g., meta-analytic sHR 0.34 (95%CI, 0.27-0.44) and 0.59 (0.50-0.70) respectively for 0-30 days, sHR 0.58 (0.48 - 0.69) and 0.71 (0.59 - 0.85) respectively for 90-180 days post COVID-19. Additionally, reduced risks of ATE, myocarditis/pericarditis and arrhythmia were seen, but mostly in the acute phase (0-30 days post COVID-19). ConclusionsCOVID-19 vaccination reduced the risk of post COVID-19 complications, including cardiac and thromboembolic outcomes. These effects were more pronounced for acute (1-month) post COVID-19 outcomes, consistent with known reductions in disease severity following breakthrough vs unvaccinated SARS-CoV-2 infection. RelevanceThese findings highlight the importance of COVID-19 vaccination to prevent cardiovascular outcomes after COVID-19, beyond respiratory disease. Key PointsO_ST_ABSQuestionC_ST_ABSWhat is the impact of COVID-19 vaccination to prevent cardiac complications and thromboembolic events following a SARS-CoV-2 infection? FindingsResults from this multinational cohort study showed that COVID-19 vaccination reduced risk for acute and subacute COVID-19 heart failure, as well as venous and arterial thromboembolic events following SARS-CoV-2 infection. MeaningThese findings highlight yet another benefit of vaccination against COVID-19, and support the recommendations for COVID-19 vaccination even in people at high cardiovascular risk.

The burden of long COVID: a multinational cohort analysis of Spanish and UK data including SARS-CoV-2 infections, reinfections, and matched contemporaneous test negative controls (preprint)

As limited data was available on the effect of persisting COVID-19 symptoms, we characterised long COVID and identified key symptoms associated with persistent disease. Using primary care data from Spain and UK, we estimated incidence rates of long COVID in the population and among COVID-19 patients over time. Subsequently, we investigated which WHO-listed symptoms were particularly differential for long COVID by comparing their frequency in COVID-19 patients vs matched test-negative controls. Lastly, we compared persistent symptoms after first infections vs. reinfections. Fortunately, the proportion of COVID-19 cases resulting in long COVID declined over the study period.  Risk for altered smell/taste, dyspnoea, and fatigue were consistently higher in long COVID patient vs controls [RR between 5.97-1.09]. All persistent symptoms were less common after reinfection than first infection. More research is needed into the definition of long COVID, and the effect of interventions to minimise the risk and impact of persistent symptoms.

Observational methods for COVID-19 vaccine effectiveness research: a trial emulation and empirical evaluation (preprint)

Despite much research on the topic, little work has been done comparing the use of methods to control for confounding in the estimation of COVID-19 vaccine effectiveness in routinely collected medical record data. We conducted a trial emulation study to replicate the ChAdOx1 (Oxford/AstraZeneca) and BNT162b2 (BioNTech/Pfizer) COVID-19 phase 3 efficacy studies. We conducted a cohort study including individuals aged 75+ from UK CPRD AURUM (N = 916,128) in early 2021. Three different methods were assessed: Overlap weighting, inverse probability treatment weighting, and propensity score matching. All three methods successfully replicated the findings from both phase 3 trials, and overlap weighting performed best in terms of confounding, systematic error, and precision. Despite lack of trial data beyond 3 weeks, we found that even 1 dose of BNT162b2 was effective against SARS-CoV-2 infection for up to 12 weeks before a second dose was administered. These results support the UK Joint Committee on Vaccination and Immunisation modelling and related UK vaccination strategies implemented in early 2021.

Digital ethnicity data in population-wide electronic health records in England: a description of completeness, coverage, and granularity of diversity (preprint)

Background The link between ethnicity and healthcare inequity, and the urgency for better data is well-recognised. This study describes ethnicity data in nation-wide electronic health records in England, UK. Methods We conducted a retrospective cohort study using de-identified person-level records for the England population available in the National Health Service (NHS) Digital trusted research environment. Primary care records (GDPPR) were linked to hospital and national mortality records. We assessed completeness, consistency, and granularity of ethnicity records using all available SNOMED-CT concepts for ethnicity and NHS ethnicity categories. Findings From 61.8 million individuals registered with a primary care practice in England, 51.5 (83.3%) had at least one ethnicity record in GDPPR, increasing to 93·9% when linked with hospital records. Approximately 12·0% had at least two conflicting ethnicity codes in primary care records. Women were more likely to have ethnicity recorded than men. Ethnicity was missing most frequently in individuals from 18 to 39 years old and in the southern regions of England. Individuals with an ethnicity record had more comorbidities recorded than those without. Of 489 SNOMED-CT ethnicity concepts available, 255 were used in primary care records. Discrepancies between SNOMED-CT and NHS ethnicity categories were observed, specifically within “Other-” ethnicity groups. Interpretation More than 250 ethnicity sub-groups may be found in health records for the English population, although commonly categorised into “White”, “Black”, “Asian”, “Mixed”, and “Other”. One in ten individuals do not have ethnicity information recorded in primary care or hospital records. SNOMED-CT codes represent more diversity in ethnicity groups than the NHS ethnicity classification. Improved recording of self-reported ethnicity at first point-of-care and consistency in ethnicity classification across healthcare settings can potentially improve the accuracy of ethnicity in research and ultimately care for all ethnicities. Funding British Heart Foundation Data Science Centre led by Health Data Research UK. Research in context Evidence before this study Ethnicity has been highlighted as a significant factor in the disproportionate impact of SARS-CoV-2 infection and mortality. Better knowledge of ethnicity data recorded in real clinical practice is required to improve health research and ultimately healthcare. We searched PubMed from database inception to 14 th July 2022 for publications using the search terms “ethnicity” and “electronic health records” or “EHR,” without language restrictions. 228 publications in 2019, before the COVID-19 pandemic, and 304 publications between 2020 and 2022 were identified. However, none of these publications used or reported any of over 400 available SNOMED-CT concepts for ethnicity to account for more granularity and diversity than captured by traditional high-level classification limited to 5 to 9 ethnicity groups. Added value of this study We provide a comprehensive study of the largest collection of ethnicity records from a national-level electronic health records trusted research environment, exploring completeness, consistency, and granularity. This work can serve as a data resource profile of ethnicity from routinely-collected EHR in England. Implications of all the available evidence To achieve equity in healthcare, we need to understand the differences between individuals, as well as the influence of ethnicity both on health status and on health interventions, including variation in the behaviour of tests and therapies. Thus, there is a need for measurements, thresholds, and risk estimates to be tailored to different ethnic groups. This study presents the different medical concepts describing ethnicity in routinely collected data that are readily available to researchers and highlights key elements for improving their accuracy in research. We aim to encourage researchers to use more granular ethnicity than the than typical approaches which aggregate ethnicity into a limited number of categories, failing to reflect the diversity of underlying populations. Accurate ethnicity data will lead to a better understanding of individual diversity, which will help to address disparities and influence policy recommendations that can translate into better, fairer health for all.

Contribution of genetics and lifestyle to the risk of major cardiovascular and thromboembolic complications following COVID-19 (preprint)

Clinical determinants for cardiovascular and thromboembolic (CVE) complications of COVID-19 are well-understood, but the roles of genetics and lifestyle remain unknown. We performed a prospective cohort study using UK Biobank, including 25,335 participants with confirmed SARS-CoV-2 infection between March 1, 2020, and September 3, 2021. Outcomes were hospital-diagnosed atrial fibrillation (AF), coronary artery disease (CAD), ischemic stroke (ISS), and venous thromboembolism (VTE) within 90 days post-infection. Heritable risk was represented by validated polygenic risk scores (PRSs). Lifestyle was defined by a composite of nine variables. We estimated adjusted hazard ratios (aHR) and confidence intervals (CI) using Cox proportional hazards models. In the COVID-19 acute phase, PRSs linearly predicted a higher risk of AF (aHR 1.52 per standard deviation increase, 95% CI 1.39 to 1.67), CAD (1.59, 1.40 to 1.81), and VTE (1.30, 1.11 to 1.53), but not ISS (0.92, 0.64 to 1.33). A healthy lifestyle was associated with a substantially lower risk of post-COVID-19 AF (0.70, 0.53 to 0.92), CAD (0.64, 0.44 to 0.91), and ISS (0.28, 0.12 to0.64), but not VTE (0.82, 0.48 to 1.39), compared with an unhealthy lifestyle. No evidence for interactions between genetics and lifestyle was found. Our results demonstrated that population genetics and lifestyle considerably influence cardiovascular complications following COVID-19, with implications for future personalised thromboprophylaxis and healthy lifestyle campaigns to offset the elevated cardiovascular disease burden imposed by the ongoing pandemic.

Contribution of genetics and lifestyle to the risk of major cardiovascular and thromboembolic complications following COVID-19 (preprint)

Clinical determinants for cardiovascular and thromboembolic (CVE) complications of COVID-19 are well-understood, but the roles of genetics and lifestyle remain unknown. We performed a prospective cohort study using UK Biobank, including 25,335 participants with confirmed SARS-CoV-2 infection between March 1, 2020, and September 3, 2021. Outcomes were hospital-diagnosed atrial fibrillation (AF), coronary artery disease (CAD), ischemic stroke (ISS), and venous thromboembolism (VTE) within 90 days post-infection. Heritable risk was represented by validated polygenic risk scores (PRSs). Lifestyle was defined by a composite of nine variables. We estimated adjusted hazard ratios (aHR) and confidence intervals (CI) using Cox proportional hazards models. In the COVID-19 acute phase, PRSs linearly predicted a higher risk of AF (aHR 1.52 per standard deviation increase, 95% CI 1.39 to 1.67), CAD (1.59, 1.40 to 1.81), and VTE (1.30, 1.11 to 1.53), but not ISS (0.92, 0.64 to 1.33). A healthy lifestyle was associated with a substantially lower risk of post-COVID-19 AF (0.70, 0.53 to 0.92), CAD (0.64, 0.44 to 0.91), and ISS (0.28, 0.12 to0.64), but not VTE (0.82, 0.48 to 1.39), compared with an unhealthy lifestyle. No evidence for interactions between genetics and lifestyle was found. Our results demonstrated that population genetics and lifestyle considerably influence cardiovascular complications following COVID-19, with implications for future personalised thromboprophylaxis and healthy lifestyle campaigns to offset the elevated cardiovascular disease burden imposed by the ongoing pandemic.

Long-term mental health outcomes after SARS-CoV-2 infection: prospective cohort study (preprint)

Background Despite previous evidence from retrospective cohorts suggest that survivors of COVID-19 may be at increased risk of psychiatric sequelae, questions remain on the incidence and absolute risk of psychiatric outcomes, and on the potential protective effect of vaccination. Addressing these knowledge gaps will help public health and clinical service planning during the ongoing pandemic. Methods Based on UK Biobank prospective data, we constructed a SARS-CoV-2 infection cohort including participants with a positive PCR test for SARS-CoV-2 between March 1, 2020 and September 30, 2021; a contemporary control cohort with no evidence of SARS-CoV-2, and a historical control cohort predating the COVID-19 pandemic. Additional control cohorts were constructed for benchmarking, including participants diagnosed with other respiratory tract infection, or with a negative SARS-CoV-2 test. We used propensity score weighting using predefined (clinically informed) and data-driven covariates to minimize confounding. We then estimated incidence rates and risk of first psychiatric disorders diagnosed by ICD-10 codes and psychotropic prescriptions after SARS-CoV-2 infection using cause-specific Cox models. Results In this prospective cohort including 406,579 adults (224,681 women, 181,898 men; mean [SD] age 66.1 [8.4] years), 26,181 had a SARS-CoV-2 infection. Compared with contemporary controls (n=380,398), COVID-19 survivors had increased risks of subsequent psychiatric diagnoses (HR: 2.02, 95% CI 1.85-2.21; difference in incidence rate: 24.85, 95 CI 20.69-29.39 per 1000 person-years) and psychotropic prescriptions (HR: 1.61, 95% CI 1.48-1.75; difference in incidence rate: 21.77, 95% CI 16.59-27.54 per 1000 person-years). Regarding individual mental health related outcomes, the SARS-CoV-2 infection cohort showed an increased risk of psychotic disorders (2.26, 1.28-3.98), mood disorders (2.19, 1.92-2.50), anxiety disorders (2.08, 1.82-2.38), substance use disorders (1.59, 1.34-1.90), sleep disorders (1.95, 1.60-2.39); and prescriptions for antipsychotics (3.78, 2.74-5.21), antidepressants (1.55, 1.29-1.87), benzodiazepines (1.82, 1.58-2.11), and opioids (1.40, 1.26-1.55). Overall, the risk of any mental health outcome was increased with a HR of 1.58, 95% CI 1.47-1.70; and difference in incidence rate of 32.04, 25.76-38.81 per 1000 person-years. These results were consistent when comparing to a historical control cohort. Additionally, mental health risks were increased even further in participants who tested positive in hospital settings. Finally, participants who were fully vaccinated had a lower risk of mental health outcomes compared to those infected when unvaccinated or partially vaccinated. All observed risks of mental health outcomes were attenuated or even lower after SARS-CoV-2 infection compared with those with other respiratory infections, or with participants in the test-negative control cohort. Conclusions In this prospective cohort study, people who survived COVID-19 were at increased risk of psychiatric outcomes and related psychotropic medications. These risks were higher in those with more severe disease, treated in hospital settings, and were significantly reduced in fully vaccinated people. Of note, compared to participants with other respiratory infections or with only negative testing results, those infected with SARS-CoV-2 had an even lower risk of mental health outcomes, warranting further research into causation. The early identification and treatment of psychiatric disorders among survivors of COVID-19 should be a priority in the long-term management of COVID-19. Particular attention might be needed for those with severe (hospitalized) disease and those who were not fully vaccinated at the time of infection.

Genetic risk and incident venous thromboembolism in middle-aged and older adults following Covid-19 vaccination (preprint)

BACKGROUND Covid-19 vaccination has been associated with an increased risk of venous thromboembolism (VTE). However, it is unknown whether genetic predisposition to VTE is associated with an increased risk of thrombosis following vaccination. METHODS Using data from the UK Biobank, which contains in-depth genotyping data and linked vaccination and health outcomes information, we generated a polygenic risk score (PRS) using 299 genetic variants identified from a previous large genome-wide association study. We prospectively assessed associations between PRS and incident VTE after first and the second-dose vaccination separately. We conducted sensitivity analyses stratified by vaccine type (adenovirus- and mRNA-based) and using two historical unvaccinated cohorts. We estimated hazard ratios (HR) for PRS-VTE associations using Cox models. RESULTS Of 359,310 individuals receiving one dose of a Covid-19 vaccine, 160,327 (44.6%) were males, and the mean age at the vaccination date was 69.05 (standard deviation [SD] 8.04) years. After 28- and 90-days follow-up, 88 and 299 individuals developed VTE respectively, equivalent to an incidence rate of 0.88 (95% confidence interval [CI] 0.70 to 1.08) and 0.92 (95% CI 0.82 to 1.04) per 100,000 person-days. The PRS was significantly associated with a higher risk of VTE (HR per 1 SD increase in PRS, 1.41 (95% CI 1.15 to 1.73) in 28 days and 1.36 (95% CI 1.22 to 1.52) in 90 days). Similar associations were found after stratification by vaccine type, in the two-dose cohort and across the historical unvaccinated cohorts. CONCLUSIONS The genetic determinants of post-Covid-19-vaccination VTE are similar to those seen in historical data. This suggests that, at the population level, post-vaccine VTE has similar aetiology to conventional VTE. Additionally, the observed PRS-VTE associations were equivalent for adenovirus- and mRNA-based vaccines.

Venous Thromboembolism in Ambulatory Covid-19 patients: Clinical and Genetic Determinants (preprint)

Background Substantial evidence suggests that severe Covid-19 leads to an increased risk of Venous Thromboembolism (VTE). We aimed to quantify the risk of VTE associated with ambulatory Covid-19, study the potential protective role of vaccination, and establish key clinical and genetic determinants of post-Covid VTE. Methods We analyzed a cohort of ambulatory Covid-19 patients from UK Biobank, and compared their 30-day VTE risk with propensity-score-matched non-infected participants. We fitted multivariable models to study the associations between age, sex, ethnicity, socio-economic status, obesity, vaccination status and inherited thrombophilia with post-Covid VTE. Results Overall, VTE risk was nearly 20-fold higher in Covid-19 vs matched non-infected participants (hazard ratio [HR] 19.49, 95% confidence interval [CI] 11.50 to 33.05). However, the risk was substantially attenuated amongst the vaccinated (HR: 2.79, 95% CI 0.82 to 9.54). Older age, male sex, and obesity were independently associated with higher risk, with adjusted HRs of 2.00 (1.61 to 2.47) per 10 years, 1.66 (1.28 to 2.15), and 1.85 (1.29 to 2.64), respectively. Further, inherited thrombophilia led to an HR 2.05, 95% CI 1.15 to 3.66. Conclusions Ambulatory Covid-19 was associated with a striking 20-fold increase in incident VTE, but no elevated risk after breakthrough infection in the fully vaccinated. Older age, male sex, and obesity were clinical determinants of Covid-19-related VTE. Additionally, inherited thrombophilia doubled risk further, comparable to the effect of 10-year ageing. These findings reinforce the need for vaccination, and call for targeted strategies to prevent VTE during outpatient care of Covid-19

The impact of COVID passport mandates on the number of cases of and hospitalizations with COVID-19 in the UK: a difference-in-differences analysis (preprint)

BackgroundMandatory COVID-19 certification was introduced at different times in the four countries of the UK. We aimed to study the effect of this intervention on the incidence of cases and hospital admissions. MethodsThe main outcome was the weekly averaged incidence of COVID-19 confirmed cases and hospital admissions. We performed Negative Binomial Segmented Regression (NBSR) and Autoregressive Integrated Moving Average (ARIMA) analyses for the four countries (England, Northern Ireland, Scotland and Wales), and fitted Difference-in-Differences (DiD) models to compare the latter three to England, where COVID-19 certification was imposed the latest. FindingsNBSR methods suggested COVID-19 certification led to a decrease in the incidence of cases in Northern Ireland, but not in hospitalizations. In Wales, they also caused a decrease in the incidence of cases but not in hospital admissions. In Scotland, we observed a decrease in both cases and admissions. ARIMA models confirmed these results. The DiD model showed that the intervention decreased the incidence of COVID compared to England in all countries except Wales, in October. Then, the incidence rate of cases already had a decreasing tendency, as well as in England, hence a particular impact of Covid Passport was less obvious. In Wales, the model coefficients were 2.2 (95% CI -6.24,10.70) for cases and -0.144 (95% CI -0.248, -0.039) for admissions in October and -7.75 (95% CI -13.1, -2.46) for cases and -0.169 (95% CI-0.308, -0.031) for admissions in November. In Northern Ireland, -10.1 (95% CI -18.4, -1.79) for cases and -0.269 (95% CI -0.385, -0.153) for admissions. In Scotland they were 7.91 (95% CI 4.46,11.4) for cases and -0.097 (95% CI - 0.219,0.024) for admissions. InterpretationThe introduction of mandatory certificates decreased cases in all countries except in England. Differences on concomitant measures, on vaccination uptake or Omicron variant prevalence could explain this discrepancy.

Use of machine learning for comparing disease risk scores and propensity scores under complex confounding and large sample size scenarios: a simulation study (preprint)

ABSTRACT Background The surge of treatments for COVID-19 in the ongoing pandemic presents an exemplar scenario with low prevalence of a given treatment and high outcome risk. Motivated by that, we conducted a simulation study for treatment effect estimation in such scenarios. We compared the performance of two methods for addressing confounding during the process of estimating treatment effects, namely disease risk scores (DRS) and propensity scores (PS) using different machine learning algorithms. Methods Monte Carlo simulated data with 25 different scenarios of treatment prevalence, outcome risk, data complexity, and sample size were created. PS and DRS matching with 1: 1 ratio were applied with logistic regression with least absolute shrinkage and selection operator (LASSO) regularization, multilayer perceptron (MLP), and eXtreme Gradient Boosting (XgBoost). Estimation performance was evaluated using relative bias and corresponding confidence intervals. Results Bias in treatment effect estimation increased with decreasing treatment prevalence regardless of matching method. DRS resulted in lower bias compared to PS when treatment prevalence was less than 10%, under strong confounding and nonlinear nonadditive data setting. However, DRS did not outperform PS under linear data setting and small sample size, even when the treatment prevalence was less than 10%. PS had a comparable or lower bias to DRS when treatment prevalence was common or high (10% - 50%). All three machine learning methods had similar performance, with LASSO and XgBoost yielding the lowest bias in some scenarios. Decreasing sample size or adding nonlinearity and non-additivity in data improved the performance of both PS and DRS. Conclusions Under strong confounding with large sample size DRS reduced bias compared to PS in scenarios with low treatment prevalence (less than 10%), whilst PS was preferable for the study of treatments with prevalence greater than 10%, regardless of the outcome prevalence. Key Messages When handling nonlinear nonadditive data with strong confounding, DRS estimated by machine learning methods outperforms PS in scenarios with low treatment prevalence (less than 10%). However, if having linear data and small sample size data with strong confounding, we did not observe DRS outperformed PS even when treatment prevalence was less than 10%. Our results suggested that using PS performed better compared to DRS in tackling strong confounding problems with treatment prevalence greater than 10%. Small sample size increased bias for both DRS and PS methods, and it affected DRS more than PS.

Comparative effectiveness of the BNT162b2 vs ChAdOx1 vaccine against Covid-19 (preprint)

Although pivotal trials with varying populations and study methods suggest higher efficacy for mRNA than adenoviral Covid-19 vaccines, no direct evidence is available. Here, we conducted a head-to-head comparison of BNT162b2 versus ChAdOx1 against Covid-19. We analysed 235,181 UK Biobank participants aged 50 years or older and vaccinated with one or two doses of BNT162b2 or ChAdOx1. People were followed from the vaccination date until 18/10/2021. Inverse probability weighting was used to minimise confounding and the Cox models to derive hazard ratio. We found that, compared with two doses of ChAdOx1, vaccination with BNT162b2 was associated with 30% lower risks of both SARS-CoV-2 infection and related hospitalisation during the period dominated by the delta variant. Also, this comparative effectiveness was consistent across several subgroups and persisted for at least six months, suggesting no differential waning between the two vaccines. Our findings can inform evidence-based Covid-19 vaccination campaigns and booster strategies.

Comparative effectiveness and safety of homologous two-dose ChAdOx1 versus heterologous vaccination with ChAdOx1 and BNT162b2: a cohort analysis (preprint)

Small trials have suggested that heterologous vaccination with first-dose ChAdOx1 and second-dose BNT162b2 may generate a better immune response than homologous vaccination with two doses of ChAdOx1. We used linked data from Catalonia (Spain), where those aged <60 who received a first dose of ChAdOx1 could choose between ChAdOx1 and BNT162b2 for their second dose. Comparable cohorts were obtained after exact-matching 14,325/17,849 (80.3%) people receiving heterologous vaccination to 14,325/149,386 (9.6%) receiving homologous vaccination by age, sex, region, and date of second dose. Of these, 238 (1.7%) in the heterologous and 389 (2.7%) in the homologous groups developed COVID-19 between 1st June 2021 and 11th October 2021. The resulting hazard ratio (95% confidence interval) was 0.61 [ 0.52-0.71 ], favouring heterologous vaccination, with a Number Needed to Treat of 94.9 [ 71.8 - 139.8 ]. The two groups had similar testing rates and safety outcomes. Sensitivity and negative control outcome analyses confirmed these findings. In conclusion, we demonstrate that a heterologous vaccination schedule with ChAdOx1 followed by BNT162b2 was more efficacious than and similarly safe to homologous vaccination with two doses of ChAdOx1. Most of the infections in our study occurred when Delta was the predominant SARS-CoV-2 variant in Spain. These data agree with previous phase 2 randomised trials.