Does the occurrence of cardiovascular complications after acute respiratory infection vary by underlying cardiovascular risk? A cohort study using linked primary and secondary care records from over 4.2 million individuals in England, 2008-2018

Talk Code: 
Jennifer Davidson
Jennifer Davidson, Amitava Banerjee, Liam Smeeth, Harriet Forbes, Daniel Grint, Emily Herrett, Helen McDonald, Richard Pebody, Charlotte Warren-Gash
Author institutions: 
London School of Hygiene and Tropical Medicine, University College London, University of Bristol, Public Health England


The effect of underlying cardiovascular risk profile on complications following acute respiratory infections (ARIs) in individuals without established cardiovascular disease (CVD) is unknown. Whether to consider individuals at raised cardiovascular risk, but without CVD, a priority group for vaccination against ARIs, such as influenza, therefore remains unclear. Our study aimed to investigate how cardiovascular risk modified the occurrence of ARI and major adverse cardiovascular events (MACE) after an ARI.


We conducted a cohort study in individuals aged 40-64 years without established CVD or a chronic health condition eligible for influenza vaccination, using primary care Clinical Practice Research Datalink data from 01/09/2008-31/08/2018 linked to Hospital Episode Statistics Admitted Patient Care and Office for National Statistics mortality data from England. We classified underlying cardiovascular risk based on diagnosed hypertension and overall predicted cardiovascular risk estimated using QRISK2 score. Using multivariable Poisson regression models with time-varying age and cardiovascular risk level, we obtained incidence rate ratios (IRR) for ARI. Among individuals who had an ARI, we then used multivariable Cox regression models to obtain hazard ratios (HR) for the risk of major adverse cardiovascular events (MACE) within one year of infection.


4,212,930 individuals were included; 526,480 (12.5%) had hypertension and 607,087 (14.4%) had a QRISK2 score ≥10%. There were 586,147 ARI episodes among 442,408 individuals, of which 107,639 episodes were influenza/ILI and 31,068 pneumonia. After adjusting for confounders, patients with raised cardiovascular risk had a higher incidence of ARI, whether by hypertension (IRR 1.04, 95% CI 1.03-1.05) or QRISK2 score ≥10% (IRR 1.39, 1.37-1.40). Of the 442,408 individuals with an ARI, 4,196 had a MACE within one year of infection. A large proportion (38.5%) of MACE after ARI were in those with pneumonia. After adjustment, having raised cardiovascular risk was associated with substantial increased risk of a MACE after infection (hypertension: HR 1.98, 1.83-2.15 and QRISK2: HR 3.65, 3.42-3.89).


People with no diagnosed CVD but who have raised underlying cardiovascular risk, measured by hypertension diagnosis and, in particular, overall predicted cardiovascular risk, have increased incidence of both ARI and cardiovascular complications following an ARI. The findings show the importance of managing and reducing cardiovascular risk to lessen ARIs and the cardiovascular consequences of infection. Individuals with raised cardiovascular risk are not typically targeted for seasonal influenza or pneumococcal vaccines. QRISK2 score provides a better marker of risk, compared with hypertension alone, to identify individuals at risk of first cardiovascular event following ARI.

Submitted by: 
Jennifer Davidson
Funding acknowledgement: 
British Heart Foundation Non-Clinical PhD Studentship FS/18/71/33938 and Wellcome Trust Intermediate Clinical Fellowship 201440/Z/16/Z.