Summary
Background
While cardiovascular disease (CVD) is a risk factor for severe COVID-19, the association between predicted cardiovascular risk and severe COVID-19 among people without diagnosed CVD is unclear.
Methods
We carried out historical, population-based cohort studies among adults aged 40–84 years in England using linked data from the Clinical Practice Research Datalink. Individuals were categorized into: existing CVD, raised cardiovascular risk (defined using QRISK3 score ≥10%) and low risk (QRISK3 score <10%) at 12/03/2020. We described incidence and severe outcomes of COVID-19 (deaths, intensive care unit [ICU] admissions, hospitalisations, major adverse cardiovascular events [MACE]) for each group. Among those with a COVID-19 record to 31/12/2020, we re-classified cardiovascular risk at infection and assessed the risk of severe outcomes using multivariable Cox regression with complete case analysis. We repeated analyses using hypertension to define raised cardiovascular risk.
Findings
Among 6,059,055 individuals, 741,913 (12.2%) had established CVD, 1,929,627 (31.8%) had a QRISK3 score ≥10% and 3,387,515 (55.9%) had a QRISK3 score <10%. Marked gradients were seen in the incidence of all severe COVID-19 outcomes by cardiovascular risk profile. Among those with COVID-19 (N = 146,760), there was a strong association between raised QRISK3 score and death: adjusted hazard ratio [aHR] 8.77 (7.62–10.10), N = 97,725, which remained present, though attenuated in age-stratified results. Risks of other outcomes were also higher among those with raised QRISK3 score: aHR 3.66 (3.18–4.21) for ICU admissions, 3.38 (3.22–3.56) for hospitalisations, 5.43 (4.44–6.64) for MACE. When raised cardiovascular risk was redefined by hypertension status, only the association with MACE remained: aHR 1.49 (1.20–1.85), N = 57,264.
Interpretation
Individuals without pre-existing CVD but with raised cardiovascular risk (by QRISK3 score) were more likely to experience severe COVID-19 outcomes and should be prioritised for prevention and treatment. Addressing cardiovascular risk factors could improve COVID-19 outcomes.
Funding
BMA Foundation for Medical Research/Rosetrees Trust, Wellcome, BHF.
Keywords: Cardiovascular risk, QRISK3, Hypertension, Severe COVID-19 outcomes, Electronic health records, Cohort study
Research in context.
Evidence before this study
While severe outcomes of COVID-19 occur more frequently among individuals with pre-existing health conditions, the role of underlying cardiovascular risk is incompletely understood. We searched PubMed from inception to 11 April 2022 using the terms ((“cardiovascular risk” OR “hypertension”) AND (“COVID-19” OR “SARS-CoV-2”) AND (“severe outcomes” OR “mortality”)). Due to a large number of results, we limited to studies of adults in non-specialist populations and filtered by systematic reviews and meta-analyses. Results were obtained from 28 relevant systematic reviews, many covering overlapping studies. Most included studies were small (100s–1000s of patients) and conducted among hospitalised COVID-19 patients. Cardiovascular disease, and to a lesser extent hypertension, were typically associated with raised risks of severe outcomes and death from COVID-19 in these studies. Later population-based studies show that existing cardiovascular disease and some individual cardiovascular risk factors (diabetes, hypertension) were associated with COVID-19-related deaths. No studies assessed cardiovascular risk using risk prediction tools such as QRISK3 which combine different elements of risk.
Added value of this study
This is, to our knowledge, the only study of COVID-19 outcomes to date to characterize underlying cardiovascular risk profile comprehensively using a validated risk prediction score (QRISK3) in a large population-based cohort. We demonstrate a gradient in the risks of hospitalisation, intensive care unit (ICU) admission, major adverse cardiovascular events (MACE) and mortality by cardiovascular risk level, with the highest incidence of severe outcomes occurring among individuals with existing cardiovascular disease, followed by those at raised cardiovascular risk then those at low risk. In cohorts with confirmed and suspected COVID-19, we show that having an elevated QRISK3 score was associated with a higher risk of all categories of severe outcomes after accounting for sociodemographic, lifestyle and clinical confounders, while hypertension status was associated only with a higher risk of MACE.
Implications of all the available evidence
Being at raised cardiovascular risk, defined by having an elevated QRISK3 score, is associated with severe outcomes after COVID-19. Individuals at raised cardiovascular risk represent an important target for COVID-19 prevention and management, as an addition to the current focus on those with diagnosed cardiovascular disease. Strategies to improve cardiovascular health could also improve outcomes following COVID-19.
Introduction
By the end of 2020, the COVID-19 pandemic had led to an estimated 3 million deaths worldwide.1 Large, population-based studies show that existing cardiovascular disease (CVD) and some individual cardiovascular risk factors (such as diabetes and hypertension) are associated with COVID-19-related deaths.2, 3, 4 Other studies support associations between CVD or individual risk factors and severe COVID-19 outcomes among hospitalised patients.5, 6, 7 CVD is a component of the QCOVID risk prediction tool which predicts risks of hospitalisation and mortality from COVID-19.8 However, it is unclear how being at raised cardiovascular risk, defined by commonly-used clinical risk prediction tools such as QRISK3, affects severe COVID-19 outcomes among individuals without existing CVD. Such individuals were not considered ‘clinically vulnerable’ in England during the COVID-19 pandemic.9
Cardiovascular complications of COVID-19 have increasingly been recognized: population-based self-controlled case series studies from Scotland,10 Sweden11 and Denmark12 show an early elevation in acute cardiovascular events such as myocardial infarction (MI) and stroke following COVID-19. Similar transient elevations in the risks of MI and stroke occur following other laboratory-confirmed respiratory infections including influenza and Streptococcus pneumoniae.13 Although evidence from before the COVID-19 pandemic showed that such complications are more frequent after respiratory infections among individuals at raised cardiovascular risk,14 this has not been comprehensively investigated for COVID-19.
Population-based studies with detailed cardiovascular risk assessments are needed to assess the burden of acute severe outcomes of COVID-19, including cardiovascular complications, among individuals with differing levels of underlying cardiovascular risk to guide accurate stratified prevention and management. Here we aimed to quantify the incidence and severe outcomes of SARS-CoV-2 infections and to assess the risk of severe COVID-19 outcomes following infection by underlying cardiovascular risk profile among adults in England.
Methods
Data sources
We used the Clinical Practice Research Datalink (CPRD) Aurum15 January 2022 dataset, with individual level linked data from Hospital Episode Statistics Admitted Patient Care (HES APC), Office for National Statistics (ONS) deaths, Second Generation Surveillance System (SGSS) SARS-CoV-2, and COVID-19 Hospitalisations in England Surveillance System (CHESS).16
The CPRD Independent Scientific Advisory Committee (application 20_000135) and the London School of Hygiene and Tropical Medicine (LSHTM) Ethics Committee (application 22717) approved the study. CPRD provided relevant HES, ONS, SGSS and CHESS data for the study population. All code lists are published on LSHTM Data Compass.17
Incidence study population and follow-up
All individuals aged 40–84 years with at least one year of post-registration time at their primary care practice who are eligible for linkage to HES were eligible for inclusion in our incidence study. Follow-up of individuals started at the latest of; age 40 years, 12 months post-registration, or 12 March 2020, and ended at the earliest of; date of death or outcome of interest, administrative censor (date of leaving the practice or date of last data collection from the practice), or 31 December 2020 (Fig. 1). We started follow-up from 12 March 2020 when daily reporting to CHESS was initiated.18
Fig. 1.
Study design overview with 2020 England COVID-19 timeline.
Cohort study population and follow up
Our cohort study included individuals with COVID-19. In our main analysis we defined this as laboratory-confirmed SARS-CoV-2, identified using SGSS and CHESS data. All individuals in either of the two datasets were considered to have SARS-CoV-2, with the date of infection taken as the earliest specimen date. In a secondary analysis, we defined COVID as clinically reported COVID-19 (CPRD or HES APC [any diagnostic position] recorded) without laboratory-confirmed SARS-CoV-2. We only considered one infection, the earliest recorded, per individual. Follow-up in the cohort study started at this date and ended at the earliest of the dates set out in our incidence study (Fig. 1). We stratified the study population further in time based on the UK COVID-19 waves (one; 12 March to 16 August and two; 17 August to 31 December), during which different testing practices were in operation.
Outcomes
Our primary outcome of interest was death attributable to COVID-19. We defined COVID-19 attributable deaths as those coded as U07.1 or U07.2 in ONS data. In a sensitivity analysis we explored broadening our primary outcome of death attributable to COVID-19 to all-cause death which occurred within 28 days of the individual's diagnosis (based on test result among those with laboratory-confirmed SARS-CoV-2 or consultation date for those with clinically reported COVID-19). Our secondary outcomes were hospitalisation due to COVID-19 (defined by COVID-19 in the primary diagnosis field of any episode recorded in HES APC or presence in CHESS dataset), ICU admission due to COVID-19 (defined by ICU admission recorded in CHESS), need for respiratory support due to COVID-19 (defined by mechanical ventilation recorded in CHESS), or major adverse cardiovascular event (MACE [composite of acute coronary syndrome which included myocardial infarction and unstable angina, ischaemic stroke, acute left ventricular failure, or major ventricular arrhythmia recorded in CPRD or HES APC]). These definitions were informed by a systematic review of the validity of cardiovascular event recording in electronic health records.19
Exposure
Our exposure of interest was cardiovascular risk. First, we identified individuals with established CVD (CPRD Aurum or HES APC recorded) diagnosed before baseline. Among individuals without CVD, we then used QRISK3 score to identify individuals with and without raised cardiovascular risk. Individuals with established CVD were included in our incidence study but excluded from our cohort study.
QRISK3 is a validated UK ten-year risk prediction score for myocardial infarction or stroke based on a combination of known risk factors,20 such as age, sex, ethnicity, socio-economic status, family history of coronary heart disease in a first degree relative aged <60 years, and comorbid health conditions (further outlined in Supplementary methods). We classified individuals as having raised cardiovascular risk (QRISK3 ≥10%) or low cardiovascular risk (QRISK3 <10%) at baseline, based on NICE thresholds for recommending statins for primary prevention.21 In a secondary analysis, we further stratified QRISK3 scores into <10%, 10–20%, or ≥20%.
In separate analyses, we redefined raised cardiovascular risk based on hypertension status within the five years before baseline as a pragmatic method to identify individuals at raised cardiovascular risk in settings where QRISK3 is not widely used. We classified hypertension using coded CPRD diagnoses or the most recent to baseline blood pressure (BP) reading with systolic BP of ≥140mmHg or diastolic BP of ≥90mmHg.
Covariates
Covariates differed depending on the method used to define cardiovascular risk. In analyses where raised cardiovascular risk was defined by QRISK3, we included covariates which were not part of determining the QRISK3 score as detailed in the Supplementary methods. In analyses where hypertension was used to define raised cardiovascular risk, we included covariates accounted for in the QRISK3 algorithm as well as those not included in the algorithm and adjusted for in QRISK3 analysis, as outlined in the Supplementary methods.
Statistical analysis
We described the baseline characteristics, for both the incidence and cohort study populations, using numbers and percentage for categorical variables and mean with standard deviation or median with interquartile range for continuous variables.
For our incidence study population, stratified by cardiovascular risk, we calculated incidence of the primary outcome of COVID-19 death and secondary outcomes of ICU admission, respiratory support, hospitalisation, and MACE, among the whole population, regardless of COVID-19 status. We then calculated the incidence of SARS-CoV-2 infection and clinically reported COVID-19, as well as our primary and secondary outcomes following laboratory-confirmed SARS-CoV-2 or clinically reported COVID-19. We further stratified results by time according to COVID-19 wave. Additionally, we generated age standardised incidence rates, stratified by sex, using one-year age bands from the ONS mid-year population estimates for 2020.22
Among our cohort study population (those with COVID-19), we used Cox proportional hazards regression finely adjusted for calendar time to generate hazard ratios for the association between cardiovascular risk and each outcome, initially adjusting models in hypertension analysis for age and sex, and then in a full model adjusted for all potential confounders. A complete case-analysis approach was used for multivariable analyses. We reported numbers in unadjusted and full models, compared characteristics of those included and excluded from the complete case analysis and also re-ran unadjusted models in the complete case analysis population. We did not conduct multiple imputation because data in CPRD are unlikely to be missing at random. We examined non-proportionality using Schoenfeld's residuals. In a post-hoc analysis, we stratified QRISK3 results by age group to evaluate the effect of age. There were no individuals aged 75–84 years with a QRISK3 score <10%, so age-stratified results were only generated for age groups of 40–54, 55–64, and 65–74 years. Age, like all risk factors included in the calculation of QRISK3 score, had not been adjusted for in the main QRISK3 analysis as the variable is also considered in the assignment of individual scores. However, in a further post-hoc analysis we additionally adjusted for age given the strong association between age and risk of severe COVID-19 outcomes.2 We conducted all analyses in Stata, version 16.
Role of the funding source
The funders of the study had no role in study design, data collection, data analysis, data interpretation or writing of the report.
Results
Description of incidence study population
The incidence study population included 6,059,055 individuals aged 40–84 years of age (Fig. 2), 12.2% (741,913) had established CVD and among those without established CVD, 31.9% (1,929,627) had a QRISK3 score ≥10% and 55.9% (3,387,515) had a QRISK3 score <10%, and 31.1% (1,881,654) had hypertension and 56.7% (3,435,488) had no hypertension. The baseline characteristics of the study population are described in Supplementary Table S1.
Fig. 2.
Study population flow chart.
Incidence of COVID-19 and severe outcomes
Among all individuals the incidence of COVID-19 death was 1.7 (95% CI 1.7–1.8) per 1000 with the highest incidence among those with established CVD (7.4 [7.2–7.7] per 1000), followed by those with raised cardiovascular risk (QRISK3 ≥10%; 2.2 [2.1–2.2] and hypertension; 1.4 [1.3–1.5] per 1000), and was lowest among those at low cardiovascular risk (QRISK3 <10%; 0.2 [0.2–0.2] and no hypertension; 0.7 [0.6–0.7] per 1000). The same gradient by cardiovascular risk level was observed for hospitalisations and MACE, and for outcomes among individuals with laboratory-confirmed SARS-CoV-2 and clinically reported COVID-19 (Table 1). Results by COVID-19 wave showed a higher incidence of outcomes of interest in wave 1 compared to the beginning of wave 2 (Supplementary Tables S2 and S3). Employing the sensitivity analysis definition of death (all cause within 28 days of diagnosis), resulted in similar incidence as COVID-19 death among individuals with laboratory-confirmed SARS-CoV-2 and a higher incidence than COVID-19 death among those with clinically reported COVID-19 (Table 1). Age standardized rates, stratified by sex, are shown in Supplementary Table S4. While age standardized rates showed similar trends across the cardiovascular risk levels, these rates were diminished compared to crude estimates.
Table 1.
Number and incidence rates of laboratory-confirmed SARS-CoV-2 and clinically reported COVID-19 and outcomes of interest.
| All | Established CVD | QRISK3 score |
Hypertension |
|||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Raised risk |
Low risk |
Raised risk |
Low risk |
|||||||||
| N | Rate (95% CI) per 1000 | N | Rate (95% CI) per 1000 | N | Rate (95% CI) per 1000 | N | Rate (95% CI) per 1000 | N | Rate (95% CI) per 1000 | N | Rate (95% CI) per 1000 | |
| All individuals | 6,059,055 | 741,913 | 1,929,627 | 3,387,515 | 1,881,654 | 3,435,488 | ||||||
| COVID-19 deatha | 7866 | 1.7 (1.7–1.8) | 4164 | 7.4 (7.2–7.7) | 3203 | 2.2 (2.1–2.2) | 499 | 0.2 (0.2–0.2) | 2014 | 1.4 (1.3–1.5) | 1688 | 0.7 (0.6–0.7) |
| Hospitalisationb | 28,013 | 6.1 (6.0–6.2) | 10,880 | 19.4 (19.0–19.8) | 10,794 | 7.3 (7.2–7.4) | 6339 | 2.5 (2.4–2.6) | 8481 | 5.9 (5.8–6.0) | 8652 | 3.4 (3.3–3.4) |
| Major adverse cardiovascular event | 71,035 | 15.5 (15.4–15.6) | 49,318 | 88.0 (87.2–88.8) | 16,604 | 11.2 (11.0–11.4) | 5113 | 2.0 (2.0–2.1) | 12,633 | 8.8 (8.6–8.9) | 9084 | 3.5 (3.4–3.6) |
| Laboratory-confirmed SARS-CoV-2 | 174,129 | 38.0 (37.8–38.2) | 24,779 | 44.2 (43.7–44.8) | 41,416 | 28.0 (27.7–28.2) | 107,934 | 42.5 (42.2–42.7) | 50,854 | 35.3 (35.0–35.6) | 98,496 | 38.2 (37.9–38.4) |
| COVID-19 deatha | 6475 | 48.9 (47.7–50.1) | 3493 | 199.9 (193.4–206.6) | 2597 | 82.9 (79.8–86.2) | 385 | 4.6 (4.2–5.1) | 1664 | 42.5 (40.5–44.6) | 1318 | 17.4 (16.5–18.4) |
| All cause death within 28 days of diagnosis | 6649 | 50.2 (49.0–51.4) | 3592 | 205.6 (198.9–212.4) | 2661 | 85.0 (81.8–88.3) | 396 | 4.7 (4.3–5.2) | 1716 | 43.9 (41.8–46.0) | 1341 | 17.7 (16.8–18.7) |
| ICU admissionc | 2024 | 15.3 (14.6–16.0) | 499 | 28.6 (26.2–31.2) | 930 | 29.7 (27.8–31.7) | 595 | 7.1 (6.6–7.7) | 811 | 20.7 (19.4–22.2) | 714 | 9.4 (8.8–10.1) |
| Respiratory supportd | 1084 | 8.2 (7.7–8.7) | 230 | 13.2 (11.6–15.0) | 526 | 16.8 (15.4–18.3) | 328 | 3.9 (3.5–4.4) | 475 | 12.1 (11.1–13.3) | 379 | 5.0 (4.5–5.5) |
| Hospitalisationb | 17,893 | 135.2 (133.2–137.2) | 6555 | 375.1 (366.1–384.3) | 7047 | 225.0 (219.8–230.3) | 4291 | 51.3 (49.8–52.9) | 5628 | 143.9 (140.1–147.7) | 5710 | 75.3 (73.4–77.3) |
| Major adverse cardiovascular event | 2251 | 17.0 (16.3–17.7) | 1422 | 81.4 (77.3–85.7) | 616 | 19.7 (18.2–21.3) | 213 | 2.5 (2.2–2.9) | 486 | 12.4 (11.4–13.6) | 343 | 4.5 (4.1–5.0) |
| Clinically reported COVID-19 | 70,700 | 15.4 (15.3–15.5) | 13,668 | 24.4 (24.0–24.8) | 20,240 | 13.7 (13.5–13.9) | 36,792 | 14.5 (14.3–14.6) | 21,374 | 14.8 (14.6–15.0) | 35,658 | 13.8 (13.7–14.0) |
| COVID-19 deatha | 723 | 13.4 (12.5–14.4) | 365 | 36.2 (32.6–40.1) | 318 | 20.7 (18.5–23.1) | 40 | 1.4 (1.0–1.9) | 178 | 10.8 (9.3–12.5) | 180 | 6.6 (5.7–7.6) |
| All cause death within 28 days of diagnosis | 1599 | 29.6 (28.2–31.1) | 823 | 81.5 (76.2–87.3) | 657 | 42.7 (39.5–46.1) | 119 | 4.2 (3.5–5.0) | 375 | 22.8 (20.6–25.2) | 401 | 14.6 (13.3–16.1) |
| Hospitalisationb | 3692 | 68.4 (66.2–70.6) | 1336 | 132.4 (125.5–139.7) | 1385 | 89.9 (85.3–94.8) | 971 | 34.1 (32.0–36.3) | 1124 | 68.2 (64.3–72.3) | 1232 | 44.9 (42.5–47.5) |
| Major adverse cardiovascular event | 2002 | 37.1 (35.5–38.7) | 1282 | 127.0 (120.3–134.2) | 569 | 37.0 (34.0–40.1) | 151 | 5.3 (4.5–6.2) | 418 | 25.4 (23.0–27.9) | 302 | 11.0 (9.8–12.3) |
Ascertained from ONS death certificate data in which the COVID related ICD-10 codes U07.1 or U07.2 were present in the record.
Ascertained from presence in CHESS dataset or HES APC record coded with primary diagnosis of U07.1 or U07.2.
Ascertained from CHESS records coded with ICU/HDU admission, only available for those with laboratory confirmed SARS-CoV-2.
Ascertained from CHESS record coded with use of respiratory support via invasive mechanical ventilation, only available for those with laboratory confirmed SARS-CoV-2.
Description of cohort study population
After excluding those with established CVD, 146,760 people had laboratory-confirmed SARS-CoV-2 and 56,197 had clinically-reported COVID-19 during our study period (Fig. 2). Among those with laboratory-confirmed SARS-CoV-2, when cardiovascular risk was classified by QRISK3 score, 26.8% (39,295) had raised risk (a score ≥10%) and 73.2% (107,465) had low risk (a score <10%). When hypertension was used to classify cardiovascular risk, 34.0% (49,955) had raised risk (hypertension) and 66.0% (96,805) had low risk (no hypertension). Individuals with laboratory-confirmed SARS-CoV-2 and raised cardiovascular risk (QRISK3 ≥10% or hypertension) were older and a higher proportion were men. Baseline characteristics of the laboratory-confirmed SARS-CoV-2 study population are shown in Table 2 and of the clinically reported COVID-19 study population in Supplementary Table S5. When compared to individuals with laboratory-confirmed SARS-CoV-2, a higher proportion of those with clinically reported COVID-19 were older, women, less affluent, and lived in London.
Table 2.
Baseline characteristics of the laboratory-confirmed SARS-CoV-2 study population by cardiovascular risk.
| All | QRISK3 score |
Hypertension |
|||
|---|---|---|---|---|---|
| Raised risk |
Low risk |
Raised risk |
Low risk |
||
| N = 146,760 | N = 39,295 | N = 107,465 | N = 49,955 | N = 96,805 | |
| Age (years), Mean (SD)a | 54.0 (10.1) | 65.3 (9.5) | 49.9 (6.6) | 57.7 (10.6) | 52.2 (9.3) |
| Age group (years)a | |||||
| 40-54 | 84,928 (57.9%) | 5150 (13.1%) | 79,778 (74.2%) | 21,382 (42.8%) | 63,546 (65.6%) |
| 55-64 | 39,757 (27.1%) | 13,967 (35.5%) | 25,790 (24.0%) | 16,435 (32.9%) | 23,322 (24.1%) |
| 65-74 | 14,782 (10.1%) | 12,885 (32.8%) | 1897 (1.8%) | 7819 (15.7%) | 6963 (7.2%) |
| 75-84 | 7293 (5.0%) | 7293 (18.6%) | 0 (0.0%) | 4319 (8.6%) | 2974 (3.1%) |
| Sexa | |||||
| Women | 80,805 (55.1%) | 14,316 (36.4%) | 66,489 (61.9%) | 24,608 (49.3%) | 56,197 (58.1%) |
| Men | 65,955 (44.9%) | 24,979 (63.6%) | 40,976 (38.1%) | 25,347 (50.7%) | 40,608 (41.9%) |
| Ethnicitya | |||||
| White or not stated | 104,902 (71.5%) | 28,657 (72.9%) | 76,245 (70.9%) | 36,280 (72.6%) | 68,622 (70.9%) |
| South Asian | 12,340 (8.4%) | 4588 (11.7%) | 7752 (7.2%) | 3893 (7.8%) | 8447 (8.7%) |
| Black | 3408 (2.3%) | 519 (1.3%) | 2889 (2.7%) | 1378 (2.8%) | 2030 (2.1%) |
| Mixed/Other | 11,793 (8.0%) | 2624 (6.7%) | 9169 (8.5%) | 4114 (8.2%) | 7679 (7.9%) |
| Unknown | 14,317 (9.8%) | 2907 (7.4%) | 11,410 (10.6%) | 4290 (8.6%) | 10,027 (10.4%) |
| Townsend quintilea | |||||
| 1 (most affluent) | 28,068 (19.1%) | 6224 (15.8%) | 21,844 (20.3%) | 9175 (18.4%) | 18,893 (19.5%) |
| 2 | 28,488 (19.4%) | 6968 (17.7%) | 21,520 (20.0%) | 9619 (19.3%) | 18,869 (19.5%) |
| 3 | 28,259 (19.3%) | 7281 (18.5%) | 20,978 (19.5%) | 9612 (19.2%) | 18,647 (19.3%) |
| 4 | 28,947 (19.7%) | 8099 (20.6%) | 20,848 (19.4%) | 10,027 (20.1%) | 18,920 (19.5%) |
| 5 (least affluent) | 32,940 (22.4%) | 10,712 (27.3%) | 22,228 (20.7%) | 11,505 (23.0%) | 21,435 (22.1%) |
| Unknown | 58 (0.0%) | 11 (0.0%) | 47 (0.0%) | 17 (0.0%) | 41 (0.0%) |
| Region of residence | |||||
| North East | 6207 (4.2%) | 1746 (4.4%) | 4461 (4.2%) | 2276 (4.6%) | 3931 (4.1%) |
| North West | 34,059 (23.2%) | 9696 (24.7%) | 24,363 (22.7%) | 12,364 (24.8%) | 21,695 (22.4%) |
| Yorkshire and the Humber | 4908 (3.3%) | 1328 (3.4%) | 3580 (3.3%) | 1718 (3.4%) | 3190 (3.3%) |
| East Midlands | 2508 (1.7%) | 690 (1.8%) | 1818 (1.7%) | 897 (1.8%) | 1611 (1.7%) |
| West Midlands | 24,071 (16.4%) | 6916 (17.6%) | 17,155 (16.0%) | 8972 (18.0%) | 15,099 (15.6%) |
| East of England | 5347 (3.6%) | 1207 (3.1%) | 4140 (3.9%) | 1662 (3.3%) | 3685 (3.8%) |
| South West | 32,542 (22.2%) | 8765 (22.3%) | 23,777 (22.1%) | 10,067 (20.2%) | 22,475 (23.2%) |
| South Central | 26,156 (17.8%) | 6050 (15.4%) | 20,106 (18.7%) | 8170 (16.4%) | 17,986 (18.6%) |
| London | 10,734 (7.3%) | 2812 (7.2%) | 7922 (7.4%) | 3737 (7.5%) | 6997 (7.2%) |
| Unknown | 228 (0.2%) | 85 (0.2%) | 143 (0.1%) | 92 (0.2%) | 136 (0.1%) |
| BMI categorya,b | |||||
| Underweight (<18.5 kg/m2) | 865 (0.6%) | 336 (0.9%) | 529 (0.5%) | 182 (0.4%) | 683 (0.7%) |
| Normal (18.5–24.9 kg/m2) | 25,789 (17.6%) | 5980 (15.2%) | 19,809 (18.4%) | 5727 (11.5%) | 20,062 (20.7%) |
| Overweight (25.0–29.9 kg/m2) | 39,501 (26.9%) | 12,415 (31.6%) | 27,086 (25.2%) | 13,618 (27.3%) | 25,883 (26.7%) |
| Obese (30.0–39.9 kg/m2) | 34,394 (23.4%) | 12,652 (32.2%) | 21,742 (20.2%) | 16,361 (32.8%) | 18,033 (18.6%) |
| Severely obese (≥40.0 kg/m2) | 5558 (3.8%) | 1985 (5.1%) | 3573 (3.3%) | 3153 (6.3%) | 2405 (2.5%) |
| Unknown | 40,653 (27.7%) | 5927 (15.1%) | 34,726 (32.3%) | 10,914 (21.8%) | 29,739 (30.7%) |
| Cholesterol:HDL, Mean (SD)a,b | 3.8 (1.2) | 4.0 (1.3) | 3.7 (1.1) | 3.8 (1.2) | 3.7 (1.2) |
| Systolic blood pressure, Mean (SD)a,b,c | 128.1 (14.6) | 134.3 (14.4) | 125.5 (13.8) | 138.7 (13.5) | 122.1 (11.3) |
| Smoking statusa,b | |||||
| Non-smoker | 73,789 (50.3%) | 18,643 (47.4%) | 55,146 (51.3%) | 26,210 (52.5%) | 47,579 (49.1%) |
| Ex-smoker | 34,028 (23.2%) | 12,543 (31.9%) | 21,485 (20.0%) | 13,287 (26.6%) | 20,741 (21.4%) |
| Current smoker | 12,400 (8.4%) | 4630 (11.8%) | 7770 (7.2%) | 3872 (7.8%) | 8528 (8.8%) |
| Unknown | 26,543 (18.1%) | 3479 (8.9%) | 23,064 (21.5%) | 6586 (13.2%) | 19,957 (20.6%) |
| Alcohol consumptionb | |||||
| No heavy drinking | 85,406 (58.2%) | 26,746 (68.1%) | 58,660 (54.6%) | 32,332 (64.7%) | 53,074 (54.8%) |
| Heavy drinking | 12,319 (8.4%) | 3760 (9.6%) | 8559 (8.0%) | 4512 (9.0%) | 7807 (8.1%) |
| Unknown | 49,035 (33.4%) | 8789 (22.4%) | 40,246 (37.5%) | 13,111 (26.2%) | 35,924 (37.1%) |
| Family history of CHDa | 13,116 (8.9%) | 4579 (11.7%) | 8537 (7.9%) | 4371 (8.7%) | 8745 (9.0%) |
| Consultation frequency in prior 12 months, Median (IQR) | 3 (1–7) | 6 (2–10) | 3 (1–6) | 4 (1–9) | 3 (1–6) |
| Medication used | |||||
| Regular corticosteroidsa | 1545 (1.1%) | 1070 (2.7%) | 475 (0.4%) | 790 (1.6%) | 755 (0.8%) |
| Antihypertensivesa | 35,232 (24.0%) | 15,938 (40.6%) | 19,294 (18.0%) | 21,361 (42.8%) | 13,871 (14.3%) |
| Statins | 19,723 (13.4%) | 13,931 (35.5%) | 5792 (5.4%) | 11,230 (22.5%) | 8493 (8.8%) |
| Antiplatelets | 7415 (5.1%) | 4360 (11.1%) | 3055 (2.8%) | 3830 (7.7%) | 3585 (3.7%) |
| Anticoagulants | 2720 (1.9%) | 1761 (4.5%) | 959 (0.9%) | 1359 (2.7%) | 1361 (1.4%) |
| Comorbid condition | |||||
| Atrial fibrillationa | 1420 (1.0%) | 1303 (3.3%) | 117 (0.1%) | 745 (1.5%) | 675 (0.7%) |
| Migrainesa | 5396 (3.7%) | 907 (2.3%) | 4489 (4.2%) | 1559 (3.1%) | 3837 (4.0%) |
| Diabetesa | 12,238 (8.3%) | 9811 (25.0%) | 2427 (2.3%) | 6603 (13.2%) | 5635 (5.8%) |
| CKD stage 3–5a | 9294 (6.3%) | 7013 (17.9%) | 2281 (2.1%) | 5482 (11.0%) | 3812 (3.9%) |
| Chronic liver disease | 1563 (1.1%) | 775 (2.0%) | 788 (0.7%) | 641 (1.3%) | 922 (1.0%) |
| Chronic respiratory disease (not asthma) | 4880 (3.3%) | 3303 (8.4%) | 1577 (1.5%) | 2267 (4.5%) | 2613 (2.7%) |
| Asthma with recent OCS used | 7558 (5.1%) | 2597 (6.6%) | 4961 (4.6%) | 3120 (6.2%) | 4438 (4.6%) |
| Asthma with no recent OCS use | 14,861 (10.1%) | 3582 (9.1%) | 11,279 (10.5%) | 5056 (10.1%) | 9805 (10.1%) |
| Severe mental illness/antipsychotic usea | 1700 (1.2%) | 956 (2.4%) | 744 (0.7%) | 595 (1.2%) | 1105 (1.1%) |
| Dementia | 2407 (1.6%) | 2046 (5.2%) | 361 (0.3%) | 1060 (2.1%) | 1347 (1.4%) |
| Chronic neurological disease | 1932 (1.3%) | 1034 (2.6%) | 898 (0.8%) | 746 (1.5%) | 1186 (1.2%) |
| Learning/intellectual disability | 1014 (0.7%) | 361 (0.9%) | 653 (0.6%) | 292 (0.6%) | 722 (0.7%) |
| Non-haematological cancer | |||||
| Diagnosed <1 year ago | 3839 (2.6%) | 2275 (5.8%) | 1564 (1.5%) | 1807 (3.6%) | 2032 (2.1%) |
| Diagnosed 1–4.9 years ago | 4554 (3.1%) | 2085 (5.3%) | 2469 (2.3%) | 1878 (3.8%) | 2676 (2.8%) |
| Diagnosed ≥5 years ago | 8436 (5.7%) | 2764 (7.0%) | 5672 (5.3%) | 2998 (6.0%) | 5438 (5.6%) |
| Haematological malignancy | |||||
| Diagnosed <1 year ago | 558 (0.4%) | 373 (0.9%) | 185 (0.2%) | 246 (0.5%) | 312 (0.3%) |
| Diagnosed 1–4.9 years ago | 273 (0.2%) | 162 (0.4%) | 111 (0.1%) | 124 (0.2%) | 149 (0.2%) |
| Diagnosed ≥5 years ago | 278 (0.2%) | 114 (0.3%) | 164 (0.2%) | 113 (0.2%) | 165 (0.2%) |
| Rheumatoid arthritisa | 1276 (0.9%) | 648 (1.6%) | 628 (0.6%) | 560 (1.1%) | 716 (0.7%) |
| Systemic lupus erythematosusa | 164 (0.1%) | 57 (0.1%) | 107 (0.1%) | 46 (0.1%) | 118 (0.1%) |
| HIVa | 234 (0.2%) | 56 (0.1%) | 178 (0.2%) | 100 (0.2%) | 134 (0.1%) |
| Immunosuppressione | 1404 (1.0%) | 643 (1.6%) | 761 (0.7%) | 597 (1.2%) | 807 (0.8%) |
| Erectile dysfunctiona | 7183 (10.9%) | 5293 (21.2%) | 1890 (4.6%) | 3556 (14.0%) | 3627 (8.9%) |
In QRISK3 algorithm, but non-imputed version included here (for smoking status, cholesterol:HDL ratio, systolic BP and BMI).
Most recent measure before baseline. N with missing cholesterol:HDL measurement 55,392 (37.7%).
Used on hypertension definition. N with missing systolic BP measurement 16,713 (11.4%).
At least 1 prescription in the 12 months before baseline. Other than corticosteroids which was defined as at least 2 prescriptions prior to baseline with the most recent ≤28 days before baseline.
Ever history of solid organ transplant or permanent cellular immune deficiency; history in the 24 months before baseline for aplastic anaemia, bone marrow or stem cell transplant; history in the 12 months before baseline for biologics or other immunosuppressant therapy (excluding corticosteroids), other or unspecified cellular immune deficiency.
Risk of death after COVID-19
In unadjusted analysis, raised QRISK3 score was associated with a substantial increase in COVID-19 death overall (HR 16.33 [14.61–18.24] N = 146,760) and in the study population with complete data available (HR 14.95 [13.07–17.10], N = 97,725) (Supplementary Table S6). After adjustment for non-QRISK3 confounders, the association between QRISK3 score and COVID-19 death attenuated but remained substantial (aHR 8.77 [7.62–10.10]). Characteristics of those included and excluded from the fully-adjusted model are shown in Supplementary Table S7. Among all patients aged 40–54 years, 6.1% had a QRISK3 score ≥10% but among those who died from COVID-19, 25.4% had a QRISK3 score ≥10% (Supplementary Table S8). In age-stratified and further age-adjusted analysis, the association between QRISK3 score and COVID-19 death was diminished compared to the main effect estimate but remained statistically significant in all age-group strata (Supplementary Table S9) with an age-adjusted result of 2.91 (2.45–3.45) In comparison, there was no association between hypertension and COVID-19 death (aHR 1.05 [0.94–1.18], N = 57,264) (Fig. 3). Results for all adjustment factors are shown in Supplementary Fig. S1.
Fig. 3.
Adjusted hazard ratios for raised cardiovascular risk effect on COVID-19 severe outcomes presented separately for a) QRISK3 and b) hypertension from complete case analysis. QRISK3 score hazard ratios are for the effect of a score ≥ 10% with <10% as the reference (N = 97,725 for complete case analysis compared to 146,760 for crude model). Hypertension hazard ratios are for the effect of having hypertension with not having hypertension as the reference (N = 57,264 for complete case analysis). Hypertension models were adjusted for age, sex, ethnicity, socioeconomic status, body-mass index, alcohol consumption, smoking status, total cholesterol: high density lipoprotein cholesterol ratio, family history of coronary heart disease, treatment with corticosteroids, antiplatelets, or anticoagulants, diagnosis of atrial fibrillation, migraine, diabetes, chronic kidney disease stage 3–5, chronic liver disease, chronic lung disease, asthma, severe mental illness, dementia, chronic neurological disease, learning disability, or malignancy, and treatment or diagnosis of a immunosuppressive condition; and QRISK3 models were adjusted for alcohol consumption, treatment with antiplatelets or anticoagulants, diagnosis of chronic liver disease, chronic lung disease, asthma, dementia, chronic neurological disease, learning disability, or malignancy, and treatment or diagnosis of an immunosuppressive condition (which are not included in the QRISK3 algorithm).
Risk of other severe outcomes after COVID-19
Significant associations were also found for QRISK3 score ≥10% and the outcomes of ICU admission (aHR 3.66 [3.18–4.21]), respiratory support (aHR 3.73 [3.10–4.49]), hospitalisation (aHR 3.38 [3.22–3.56]), and MACE (aHR 5.43 [4.44–6.64]). There was only a minor reduction in associations after further adjustment for age (Fig. 3). There was no association between hypertension and ICU admission (aHR 1.15 [0.98–1.36]), respiratory support (aHR 1.20 [0.97–1.48]), or hospitalisation (aHR 1.05 [0.99–1.11]) but there was an association between hypertension and MACE (aHR 1.49 [1.20–1.85]).
Additional analyses
Results between wave 1 and wave 2 were broadly similar for all outcomes (Supplementary Table S10). Similar results were also obtained when COVID-19 was clinically reported rather than laboratory-confirmed (Supplementary Table S11). Further stratification of the QRISK3 score showed a substantially greater risk of COVID-19 death in individuals with a QRISK3 score of ≥20% (aHR 15.15 [13.05–17.59]) than 10-<20% (aHR 5.32 [4.54–6.23]) when both were compared to those with a score <10% (Supplementary Table S12). A similar, though less extreme gradient was observed for the other outcomes.
Discussion
In this large, population-based cohort study using linked data from England in 2020, we found a striking gradient in the occurrence of severe COVID-19 outcomes by underlying cardiovascular risk profile among people without pre-existing CVD. The risks of death, ICU admission, hospital admission and MACE were all greater among individuals at raised cardiovascular risk measured by QRISK3 score, compared to those at low risk, despite no increase in recorded infections in this group. Associations between raised cardiovascular risk and COVID-19 deaths remained present, though attenuated, when results were stratified by 15-year age-group and further adjusted for age. When cardiovascular risk was measured by hypertension alone, differences were only evident for MACE outcomes. Analysis by pandemic waves revealed similar patterns, although the incidence of severe outcomes was greatest during the first wave.
Our study used linked electronic health record data from primary and secondary, including intensive, care, mortality records and national laboratory surveillance to capture detailed clinical and laboratory data on SARS-CoV-2 infections and outcomes. It is, to our knowledge, the first UK population-based study to assess COVID-19 outcomes using a comprehensive, combined measure of cardiovascular risk, QRISK3, rather than focusing on individual vascular risk factors. Findings from this large, representative cohort should be generalizable to adults in England aged 40–84 years (the upper age for which QRISK3 can be used to assess cardiovascular risk). Our dataset spanned the first and major part of the second wave of the COVID-19 pandemic in England, allowing comparisons of outcomes between waves. Limiting follow up to the end of December 2020 prevented contamination from the emergence of coronavirus variants or widespread roll out of the COVID-19 vaccination programme in England.
Nevertheless, differences in the availability of laboratory PCR testing are likely to have led to differences in the reported incidence of infection between waves: a laboratory-confirmed definition of SARS-CoV-2 lacked sensitivity to identify cases occurring during wave one before mass testing became widely available. It is also possible that some outcomes such as hospitalisation or MACE may have led to in-hospital testing, strengthening the observed association between vascular risk status and severe outcomes in the laboratory-confirmed cohort during the first wave. However, individuals who were at raised cardiovascular risk defined by QRISK3 score had a lower incidence of laboratory-confirmed infection than those at low cardiovascular risk, suggesting that differential in-hospital testing is unlikely to have biased our results. Reasons for the lower rates of laboratory-confirmed infections among individuals at high cardiovascular risk are unclear but may reflect reduced access to community testing e.g. due to shielding. Stratifying by pandemic wave to explore the effect of expanded testing and advances in clinical management of COVID-19 in later time periods revealed similar results to the main analysis. When we compared results for individuals with confirmed SARS-CoV-2 infection to those with clinically diagnosed COVID-19, we also saw similar patterns. Our descriptive analysis of COVID-19 outcomes alone regardless of recorded infection status supported findings from the cohort analysis.
The magnitude of association between cardiovascular risk status and severe outcomes varied by the method used to classify cardiovascular risk. In general, classification by QRISK3 produced more exaggerated differences between high and low cardiovascular risk groups than classification by hypertension alone. This is perhaps unsurprising as QRISK3 is a more comprehensive measure of cardiovascular risk, which includes additional comorbidities and socio-demographic components of risk. While age is a major driver of severe COVID-19 outcomes, associations with raised cardiovascular risk remained present in both age-stratified and age-adjusted analyses. Although misclassification of cardiovascular risk status could have occurred due to the documented reductions in GP visits and healthcare-seeking for non-COVID conditions during the pandemic,23 under-recognition of individuals at raised cardiovascular risk would have led to bias towards the null. In addition, our sensitivity analysis in which QRISK3 status was graded more finely into three strata (<10%, 10–19%, 20%+), confirmed a gradient of increasing risk of severe outcomes with increasing vascular risk level, which suggests that the main results are robust to any minor exposure misclassification. As QRISK3 scores were developed for the UK population, levels of cardiovascular risk identified in our study population may differ to those in other countries using different risk scores, although results of our hypertension analysis should generalize to other settings.
Residual confounding may also have been present in our study. While we adjusted for a broad range of sociodemographic, lifestyle and clinical confounding factors, some variables are either not measured (such as genetic risk profiles) or are sub-optimally recorded (such as BMI) in EHRs. Nevertheless, population-based self-controlled case series analyses of COVID-19 and thrombotic outcomes, which use within-person comparisons to control implicitly for fixed confounding24 show comparable results to cohort studies,11,25 suggesting that confounding is unlikely to explain our cohort results. Missing data on alcohol consumption reduced numbers for the QRISK3 complete case analysis (as other lifestyle factors were imputed in the QRISK3 algorithm if missing) whereas for hypertension, reduced numbers were driven by missing data on alcohol, smoking, BMI, cholesterol and ethnicity. Nevertheless, complete case analysis gives unbiased results when the chance of being a complete case is independent of outcome after taking covariates into account, even when data are missing not at random.26
Our findings extend those from previous smaller studies of individual cardiovascular risk factors and COVID-19 outcomes,27,28 supporting a strong association between raised cardiovascular risk profile and severe COVID-19 outcomes. While a previous Mendelian randomisation study, which by design avoids reverse causation and most confounding, failed to show an association between some genetically-predicted cardiovascular risk factors (blood pressure, BMI, type 2 diabetes and coronary artery disease) and COVID-19 hospitalisation,29 estimates had wide confidence intervals and did not capture full profiles of either cardiovascular risk or severe COVID-19 outcomes. The bidirectional relationship between cardiovascular risk and COVID-19 shown in our study is consistent with pre-COVID era work on cardiovascular complications of acute respiratory infections, showing a gradient in the risk of complications aligned with underlying cardiovascular risk status.14
Potential mechanisms underlying severe outcomes in COVID-19 include pro-inflammatory, pro-thrombotic and vasoconstrictive effects of SARS-CoV-2-mediated imbalances in ACE-2/RAS signalling.30 It has been suggested that individuals with conditions leading to raised cardiovascular risk are likely to have altered cytokine profiles leading to chronic systemic inflammation, which may have a synergistic effect on disease severity in acute COVID-19.31 A substantial burden of cardiovascular disease has also been demonstrated in survivors of acute COVID-19 at one year.32,33 Understanding the natural history of longer-term cardiovascular and other complications including post-COVID-19 syndrome34 in individuals at raised cardiovascular risk, along with the mechanisms underlying both short and long-term health changes, should be a priority for future research. Future studies could also investigate the role of COVID-19 treatments in modifying or mediating the relationship between raised cardiovascular risk and severe COVID-19. Combining mechanistic research with clinical evidence to improve patient care among those at raised cardiovascular risk is essential to prevent and manage severe outcomes of COVID-19 in this group.35 Our study highlights the need for a continued focus on integrated prevention e.g. combining COVID-19 vaccinations with cardiovascular disease prevention to improve health among those at raised cardiovascular risk.
In conclusion, we showed that individuals at raised cardiovascular risk in England were more likely to die or to experience severe outcomes after COVID-19 than those at low cardiovascular risk, despite not initially being identified as a vulnerable group. Those at raised cardiovascular risk should be considered a priority for targeted prevention and treatment strategies for COVID-19. Addressing cardiovascular risk factors could improve outcomes after COVID-19.
Contributors
CWG conceptualized the study and obtained funding. CWG, JAD, HS, EH, LS, JB and AB contributed to study design. JAD managed and analysed data, supported by EH, HS and CWG. JAD and CWG drafted the manuscript. All authors reviewed the manuscript, interpreted data and approved the final version for publication.
Data sharing statement
Data used for the study were obtained from the UK CPRD database under licence from the UK Medicines and Healthcare Products Regulatory Agency. Access to CPRD data is subject to protocol approval via CPRD's Research Data Governance Process (https://cprd.com/data-access). All codelists used for this study are available on LSHTM Data Compass: https://doi.org/10.17037/DATA.00002762. Analytical code is available via GitHub: https://github.com/jenAdavidson/cvrisk_covid_cohort.
Ethical approval
The CPRD Independent Scientific Advisory Committee (application 20_000135) and the London School of Hygiene and Tropical Medicine (LSHTM) Ethics Committee (application 22717) approved the study.
Declaration of interests
CWG reports participation in the data safety and monitoring board for the IAMI trial of influenza vaccine for cardiovascular disease (NCT02831608) ending April 2020. JB reports consulting fees from ARCbio, HVivo and GSK and participation in a data safety and monitoring board for the COM Cov trial, Oxford, now ended. AB reports grants from NIHR, AstraZeneca and the British Medical Association and leadership roles as Vice-President, Digital, Marketing, Communications for the British Cardiovascular Society and Senior Advisor to the Emerging Leaders Programme of the World Heart Federation. All other authors report no conflicts.
Acknowledgements
This work was supported by the BMA Foundation for Medical Research/Rosetrees Trust COVID-19 grant (M958). CWG is funded by a Wellcome Career Development Award (225868/Z/22/Z). JAD was funded by a British Heart Foundation Non-Clinical PhD Studentship (FS/18/71/33938). EH is funded by a National Institute for Health Research (NIHR) post-doctoral fellowship (PDF-2016-09-029). HS is funded by the NIHR though an Advanced Fellowship (NIHR301730). The views expressed in this publication are those of the author(s) and not necessarily those of the funders, NIHR, NHS or the UK Department of Health and Social Care. This work uses data provided by patients and collected by the NHS as part of their care and support. We thank Dr Harriet Forbes for helpful discussions on study design.
Footnotes
Supplementary data related to this article can be found at https://doi.org/10.1016/j.lanepe.2023.100604.
Appendix A. Supplementary data
References
- 1.World Health Organization . 2021. The true death toll of COVID-19: estimating global excess mortality.https://www.who.int/data/stories/the-true-death-toll-of-covid-19-estimating-global-excess-mortality [cited 2022 18/10/2022]. Available from: [Google Scholar]
- 2.Williamson E.J., Walker A.J., Bhaskaran K., et al. Factors associated with COVID-19-related death using OpenSAFELY. Nature. 2020;584(7821):430–436. doi: 10.1038/s41586-020-2521-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Strongman H., Carreira H., De Stavola B.L., Bhaskaran K., Leon D.A. Factors associated with excess all-cause mortality in the first wave of the COVID-19 pandemic in the UK: a time series analysis using the Clinical Practice Research Datalink. PLoS Med. 2022;19(1) doi: 10.1371/journal.pmed.1003870. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Barron E., Bakhai C., Kar P., et al. Associations of type 1 and type 2 diabetes with COVID-19-related mortality in England: a whole-population study. Lancet Diabetes Endocrinol. 2020;8(10):813–822. doi: 10.1016/S2213-8587(20)30272-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Pepera G., Tribali M.S., Batalik L., Petrov I., Papathanasiou J. Epidemiology, risk factors and prognosis of cardiovascular disease in the Coronavirus Disease 2019 (COVID-19) pandemic era: a systematic review. Rev Cardiovasc Med. 2022;23(1):28. doi: 10.31083/j.rcm2301028. [DOI] [PubMed] [Google Scholar]
- 6.Bae S., Kim S.R., Kim M.N., Shim W.J., Park S.M. Impact of cardiovascular disease and risk factors on fatal outcomes in patients with COVID-19 according to age: a systematic review and meta-analysis. Heart. 2021;107(5):373–380. doi: 10.1136/heartjnl-2020-317901. [DOI] [PubMed] [Google Scholar]
- 7.Silverio A., Di Maio M., Citro R., et al. Cardiovascular risk factors and mortality in hospitalized patients with COVID-19: systematic review and meta-analysis of 45 studies and 18,300 patients. BMC Cardiovasc Disord. 2021;21(1):23. doi: 10.1186/s12872-020-01816-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Clift A.K., Coupland C.A.C., Keogh R.H., et al. Living risk prediction algorithm (QCOVID) for risk of hospital admission and mortality from coronavirus 19 in adults: national derivation and validation cohort study. BMJ. 2020;371:m3731. doi: 10.1136/bmj.m3731. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.NHS Digital . 2020. Risk criteria (shielded patient list)https://digital.nhs.uk/coronavirus/shielded-patient-list/risk-criteria Available from: [Google Scholar]
- 10.Ho F.K., Man K.K.C., Toshner M., et al. Thromboembolic risk in hospitalized and nonhospitalized COVID-19 patients: a self-controlled case series analysis of a nationwide cohort. Mayo Clin Proc. 2021;96(10):2587–2597. doi: 10.1016/j.mayocp.2021.07.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Katsoularis I., Fonseca-Rodríguez O., Farrington P., Lindmark K., Fors Connolly A.M. Risk of acute myocardial infarction and ischaemic stroke following COVID-19 in Sweden: a self-controlled case series and matched cohort study. Lancet. 2021;398(10300):599–607. doi: 10.1016/S0140-6736(21)00896-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Modin D., Claggett B., Sindet-Pedersen C., et al. Acute COVID-19 and the incidence of ischemic stroke and acute myocardial infarction. Circulation. 2020;142(21):2080–2082. doi: 10.1161/CIRCULATIONAHA.120.050809. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Ohland J., Warren-Gash C., Blackburn R., et al. Acute myocardial infarctions and stroke triggered by laboratory-confirmed respiratory infections in Denmark, 2010 to 2016. Euro Surveill. 2020;25(17) doi: 10.2807/1560-7917.ES.2020.25.17.1900199. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Davidson J.A., Banerjee A., Smeeth L., et al. Risk of acute respiratory infection and acute cardiovascular events following acute respiratory infection among adults with increased cardiovascular risk in England between 2008 and 2018: a retrospective, population-based cohort study. Lancet Digit Health. 2021;3(12):e773–e783. doi: 10.1016/S2589-7500(21)00203-X. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Wolf A., Dedman D., Campbell J., et al. Data resource profile: clinical practice research Datalink (CPRD) Aurum. Int J Epidemiol. 2019;48(6):1740–g. doi: 10.1093/ije/dyz034. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Clinical Practice Research Datalink . 2022. CPRD Aurum CHESS January 2022 (Version 2022.01.001)https://cprd.com/cprd-aurum-chess-january-2022 Available from: [Google Scholar]
- 17.Davidson J.A., McDonald H.I., Strongman H., et al. 2022. Codelists for: "Effect of cardiovascular risk profile on severe outcomes of COVID-19 in England in 2020: a population-based cohort study".https://datacompass.lshtm.ac.uk/id/eprint/2762/ Available from: [Google Scholar]
- 18.Public Health England . 2020. COVID-19 Hospitalisation in England Surveillance System (CHESS) – daily reporting.https://www.england.nhs.uk/coronavirus/wp-content/uploads/sites/52/2020/03/phe-letter-to-trusts-re-daily-covid-19-hospital-surveillance-11-march-2020.pdf Available from: [Google Scholar]
- 19.Davidson J., Banerjee A., Muzambi R., Smeeth L., Warren-Gash C. Validity of acute cardiovascular outcome diagnoses recorded in European electronic health records: a systematic review. Clin Epidemiol. 2020;12:1095–1111. doi: 10.2147/CLEP.S265619. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Hippisley-Cox J., Coupland C., Brindle P. Development and validation of QRISK3 risk prediction algorithms to estimate future risk of cardiovascular disease: prospective cohort study. BMJ. 2017;357:j2099. doi: 10.1136/bmj.j2099. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.NICE . 2014. Cardiovascular disease: risk assessment and reduction, including lipid modification.https://www.nice.org.uk/guidance/cg181 Last updated 27 Sept 2016. Clinical guideline: CG181. Available from: [Google Scholar]
- 22.Office for National Statistics Estimates of the population for the UK, England, Wales, Scotland and Northern Ireland. Mid-2020 edition. 2021. https://www.ons.gov.uk/peoplepopulationandcommunity/populationandmigration/populationestimates/datasets/populationestimatesforukenglandandwalesscotlandandnorthernireland Available from:
- 23.Mansfield K.E., Mathur R., Tazare J., et al. Indirect acute effects of the COVID-19 pandemic on physical and mental health in the UK: a population-based study. Lancet Digit Health. 2021;3(4):e217–e230. doi: 10.1016/S2589-7500(21)00017-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Fonseca-Rodríguez O., Fors Connolly A.M., Katsoularis I., Lindmark K., Farrington P. Avoiding bias in self-controlled case series studies of coronavirus disease 2019. Stat Med. 2021;40(27):6197–6208. doi: 10.1002/sim.9179. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Katsoularis I., Fonseca-Rodríguez O., Farrington P., et al. Risks of deep vein thrombosis, pulmonary embolism, and bleeding after covid-19: nationwide self-controlled cases series and matched cohort study. BMJ. 2022;377 doi: 10.1136/bmj-2021-069590. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Hughes R.A., Heron J., Sterne J.A.C., Tilling K. Accounting for missing data in statistical analyses: multiple imputation is not always the answer. Int J Epidemiol. 2019;48(4):1294–1304. doi: 10.1093/ije/dyz032. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Krittanawong C., Virk H.U.H., Narasimhan B., et al. Coronavirus disease 2019 (COVID-19) and cardiovascular risk: a meta-analysis. Prog Cardiovasc Dis. 2020;63(4):527–528. doi: 10.1016/j.pcad.2020.05.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Vasbinder A., Meloche C., Azam T.U., et al. Relationship between preexisting cardiovascular disease and death and cardiovascular outcomes in critically ill patients with COVID-19. Circ Cardiovasc Qual Outcomes. 2022;15(10) doi: 10.1161/CIRCOUTCOMES.122.008942. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Cecelja M., Lewis C.M., Shah A.M., Chowienczyk P. Cardiovascular health and risk of hospitalization with COVID-19: a Mendelian Randomization study. JRSM Cardiovasc Dis. 2021;10 doi: 10.1177/20480040211059374. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Augustine R., Abhilash S., Nayeem A., et al. Increased complications of COVID-19 in people with cardiovascular disease: role of the renin-angiotensin-aldosterone system (RAAS) dysregulation. Chem Biol Interact. 2022;351 doi: 10.1016/j.cbi.2021.109738. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Srivastava A., Rockman-Greenberg C., Sareen N., Lionetti V., Dhingra S. An insight into the mechanisms of COVID-19, SARS-CoV2 infection severity concerning β-cell survival and cardiovascular conditions in diabetic patients. Mol Cell Biochem. 2022;477(6):1681–1695. doi: 10.1007/s11010-022-04396-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Xie Y., Xu E., Bowe B., Al-Aly Z. Long-term cardiovascular outcomes of COVID-19. Nat Med. 2022;28(3):583–590. doi: 10.1038/s41591-022-01689-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Wang W., Wang C.Y., Wang S.I., Wei J.C. Long-term cardiovascular outcomes in COVID-19 survivors among non-vaccinated population: a retrospective cohort study from the TriNetX US collaborative networks. eClinicalMedicine. 2022;53 doi: 10.1016/j.eclinm.2022.101619. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.NICE. SIGN. RCGP . 2021. COVID-19 rapid guideline: managing the long-term effects of COVID-19. [NICE Guideline NG188] Available from: https://www.nice.org.uk/guidance/ng188/resources/covid19-rapid-guideline-managing-the-longterm-effects-of-covid19-pdf-51035515742. [Google Scholar]
- 35.Mohamed M.O., Banerjee A. Long COVID and cardiovascular disease: a learning health system approach. Nat Rev Cardiol. 2022;19(5):287–288. doi: 10.1038/s41569-022-00697-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.