You are here

Article Text

Article menu
General practice / Family practice
Original research
Impact of influenza vaccination on GP-diagnosed COVID-19 and all-cause mortality: a Dutch cohort study
  1. Arjan van Laak1,
  2. Ruud Verhees1,
  3. J André Knottnerus1,
  4. Mariëtte Hooiveld2,
  5. Bjorn Winkens3,
  6. Geert-Jan Dinant1
  1. 1Department of General Practice, CAPHRI, Maastricht UMC+, Maastricht, The Netherlands
  2. 2General Practice Care, Otterstraat 118, Nivel, Utrecht, The Netherlands
  3. 3Department of Methodology and Statistics, CAPHRI, Maastricht UMC+, Maastricht, The Netherlands
  1. Correspondence to MD, MSc Arjan van Laak; a.vanlaak{at}alumni.maastrichtuniversity.nl

Abstract

Objectives As clinical presentation and complications of both viruses overlap, it was hypothesised that influenza vaccination was associated with lower general practitioner (GP)-diagnosed COVID-19 rates and lower all-cause mortality rates.

Study design From a primary care population-based cohort in the Netherlands, GP-diagnosed COVID-19 (between 10 March and 22 November 2020) and all-cause mortality events (between 30 December 2019 and 22 November 2020) were recorded. 223 580 persons were included, representing the influenza vaccination 2019 target group (all aged ≥60 years, and those <60 years with a medical indication). Proportional hazards regression analyses evaluated associations between influenza vaccination in 2019 and two outcomes: GP-diagnosed COVID-19 and all-cause mortality. Covariables were sex, age, comorbidities and number of acute respiratory infection primary care consultations in 2019.

Results A slightly positive association (HR 1.15; 95% CI 1.08 to 1.22) was found between influenza vaccination in 2019 and GP-diagnosed COVID-19, after adjusting for covariables. A slightly protective effect for all-cause mortality rates (HR 0.90; 95% CI 0.83 to 0.97) was found for influenza vaccination, after adjusting for covariables. A subgroup analysis among GP-diagnosed COVID-19 cases showed no significant association between influenza vaccination in 2019 and all-cause mortality.

Conclusions Our hypothesis of a possibly negative association between influenza vaccination in 2019 and GP-diagnosed COVID-19 was not confirmed as we found a slightly positive association. A slightly protective effect on all-cause mortality was found after influenza vaccination, possibly by a wider, overall protective effect on health. Future research designs should include test-confirmed COVID-19 cases and controls, adjustments for behavioural, socioeconomic and ethnic factors and validated cause-specific mortality cases.

  • COVID-19
  • primary care
  • immunology
  • epidemiology
  • public health
  • preventive medicine

Data availability statement

Data are available upon reasonable request. The corresponding author can be contacted for details, such as a technical appendix, statistical code and data set.

http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

https://doi.org/10.1136/bmjopen-2022-061727

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Strengths and limitations of this study

    • We used a large and representative sample using routinely recorded data from general practices in the Netherlands.

    • Sensitivity analyses were performed to test the robustness of models.

    • General practitioner-diagnosed COVID-19 recordings were used as SARS-CoV-2 testing capacity was limited until June 2020.

    • We could not use validated and cause-specific mortality data.

    Introduction

    In December 2019, the first COVID-19 cases were reported in Wuhan (China), causing a pandemic with disastrous effects in many ways.1 Although treatments are improving and COVID-19 vaccines are being used worldwide, there is unclarity about the mechanism of action of SARS-CoV-2 on cellular level and about which immunological agents could interact with the disease course of COVID-19. Over the last few decades, it has been reported that vaccines can trigger trained immunity and stimulate defence mechanisms against different pathogens.2 Recently, it was found that influenza vaccines protected against respiratory syncytial virus, parainfluenza and other non-influenza viruses.3 Nevertheless, a suggested positive association with the incidence of COVID-19 caused negative media attention for taking the influenza vaccination during the pandemic,3 after which the author rectified that older coronaviruses were studied, not the 2019 strain.4 Subsequently, more studies were done. A prospective cohort study among 11 201 health workers found that influenza vaccination in the 2019–2020 season was not associated with the risk of SARS-CoV-2 infection.5 Case–control, observational and retrospective pilot studies showed that SARS-CoV-2 infection rates were reduced for healthcare workers who were vaccinated against influenza, compared with the unvaccinated, and protection was also found in an ecological population-based study.6–9 In an observational study that was performed during the first COVID-19 wave among 10 631 hospital workers, a COVID-19 risk reduction of 39% was found for those receiving an influenza vaccination in the 2019/2020 winter season, compared with the unvaccinated.10 However, this only concerned a relatively young and healthy population. For high-risk groups, such as the elderly, protective effects of influenza vaccination on the occurrence of influenza have been demonstrated,11 but there is no evidence that influenza vaccines protect against COVID-19 rates and mortality rates in times of COVID-19 pandemic. More insights into either protective or non-protective effects of influenza vaccination on COVID-19 are relevant for future vaccination campaigns, which may focus on combined vaccinations against COVID-19, influenza and pneumococcal pneumonia.

    Against this background, it was therefore hypothesised that—in the target group for influenza vaccination—influenza vaccines reduce COVID-19 incidence rates and mortality rates. To test this hypothesis, in a primary care population-based cohort it was investigated whether COVID-19 incidence and all-cause mortality rates are lower among influenza vaccinees compared with non-vaccinees, taking sex, age, number of previous acute respiratory infection episodes and medical indications for influenza vaccination into account.

    Methods

    Study design

    A cohort study was conducted in a primary care population that represented the target group for influenza vaccination in 2019 in the Netherlands.

    Setting and participants

    The research focused on a large cohort of persons enlisted in general practices participating in Nivel Primary Care Database. This database is the largest one available in the Netherlands and it is used for national health and disease monitoring. To be included in the Nivel database, general practices needed to meet completeness and quality criteria. Therefore, sufficient data had to be provided on medical procedures, morbidity figures and prescriptions in 2018 and 2019. To ensure high recording quality, at least 70% of all morbidity data had to be provided with meaningful International Classification of Primary Care (ICPC) codes in correspondence with Dutch College for General Practitioners Clinical Standards (Nederlands Huisartsen Genootschap Standaarden) and which had been measured in ≥42 reporting weeks/year.12 Only practices with complete data on influenza vaccinations were included, based on <10% difference between the number of recorded influenza vaccinations and number of claimed vaccines.13

    Persons were included in a Nivel general practice throughout 2019 and belonged to the target group for the 2019/2020 influenza vaccination campaign in the Netherlands. This group included persons under 60 years of age with a medical indication for influenza vaccination and all persons aged 60 years and over.

    Variables and data sources

    Primary outcomes were time to general practitioner (GP)-diagnosed COVID-19 and time to all-cause mortality. Next to the main independent variable, influenza vaccination in 2019, the following covariables were available: age, sex (male/female), comorbidities (cardiovascular diseases, pulmonary diseases, diabetes mellitus, impaired resistance to infections, chronic renal insufficiency, respiratory disorders due to neurological conditions and HIV) and the number of acute respiratory infection consultations in primary care (in 2019).

    GP-diagnosed COVID-19 cases were reported from 10 March 2020, which corresponds to the time of notification of the first COVID-19 cases in the Netherlands, until 22 November 2020, when the statistical analyses started. GP-diagnosed COVID-19 was recorded for medically attended persons when (1) GPs diagnosed an acute infection of upper airways, other respiratory infections, influenza, pneumonia, other viral disease, other infectious disease, shortness of breath or coughing; and (2) recorded this as indicating a COVID-19 infection (whether or not test confirmed, also given circumstances such as test-confirmed COVID-19-positive housemates). Most cases had not been tested, as population-wide PCR testing was introduced in June 2020.

    Mortality events were recorded throughout 2020 until 22 November 2020. Mortality labels were assigned to persons in our database when GPs recorded the ICPC code (A96) that corresponded with death, which was the best alternative for validated CBS (Statistics Netherlands) data that we were unable to incorporate in our database. Covariable information, such as presence of comorbidities, was obtained from patients’ medical records at their GPs.

    Bias

    Confounding was reduced by adjusting for covariables with respect to demographic characteristics and specific (medical) indications for influenza vaccination. A randomised controlled trial would offer the highest methodological value. However, as it would be unethical to not vaccinate persons with the influenza vaccine proven to be effective in preventing influenza,11 the best alternative was performed: a primary care population-based cohort study. Information bias was limited due to blinding of the researchers in terms of pseudonymisation of patients and general practices. Coding of variables was executed by Nivel investigators already before this study was set up.

    Study size

    Prior to the start of this study, it was calculated for different practical scenarios that the study size of our database would meet the study power (80%) requirements to detect a relative risk (RR) reduction of 25% for GP-diagnosed COVID-19 between influenza vaccinees and non-vaccinees (online supplemental parts I–III). A significance level of 0.05 was used in this study.

    Statistical analysis and quantitative variables

    Differences in numerical and categorical variables between the unvaccinated and vaccinated groups were assessed using independent samples t-test and χ2 test, respectively. Cox regression analyses were used to assess the effect of vaccination in 2019 on the time to GP-diagnosed COVID-19 and all-cause mortality. The following covariables were included in the models: age, sex (male/female), comorbidities and the number of acute respiratory infection consultations in primary care in 2019. For the all-cause mortality models, GP-diagnosed COVID-19 status (yes vs no) was also incorporated as a covariable. As all-cause mortality data collection started earlier than the recording of the first COVID-19 cases, models were adjusted for these time-to-event differences and the starting point was the same for GP-diagnosed COVID-19 cases and controls. In addition, an all-cause mortality analysis was performed for GP-diagnosed COVID-19 cases.

    The linearity assumption was checked for numerical covariables. Due to a non-linear association with the outcomes and simplicity reasons, the numerical variable ‘age’ was divided into three age groups (0–59, 60–74, 75+ years) and included in the model. The same applied for ‘acute respiratory infection consultations in 2019’, where three groups were analysed (0, 1 and 2 or more consultations in 2019). In addition, proportional hazards assumptions were checked for all variables in the model via plotting of log-log (survival) curves and Schoenfeld residual testing. In case the proportional hazard assumption was violated, time-dependent covariables (interaction of a covariable with time) were computed and added to the proportional hazards (Cox regression) models. Covariables were removed via a top-down procedure if they did not contribute significantly to the model. Two-sided p values ≤0.05 were considered statistically significant. Multicollinearity was checked using variance inflation factors (VIF), where VIF≥10 indicates a (multi)collinearity problem.

    Sensitivity analyses were performed to test the robustness of GP-diagnosed COVID-19 models and to exclude a potential effect by PCR testing which only became possible for the Dutch population after 1 June 2020. Therefore, two time periods were analysed separately, one representing the first COVID-19 wave (10 March to 1 June 2020) and the other representing the second wave (1 June to 22 November 2020) (online supplemental part IV). Since we were unable to use validated mortality data, mortality rates of our study were compared with those of the entire Dutch population (online supplemental part V). All analyses were performed using IBM SPSS Statistics for Windows (V.26.0, IBM). By reasons of large study groups which resulted in very small and non-informative p values, HRs with 95% CIs were assumed to be most useful and p values were omitted.

    Patient and public involvement

    Patients or the public were not involved in the design, or conduct, or reporting or dissemination plans of our research.

    Privacy and ethical considerations

    The use of electronic health records for research purposes was allowed under certain conditions. When these conditions were fulfilled, neither obtaining informed consent from patients nor approval by a medical ethics committee was obligatory for this type of observational studies containing no directly identifiable data (Art 24 Dutch General Data Protection Regulation (GDPR) Implementation Act jo, Art 9.2 sub jGDPR).

    Nivel received pseudonymised data from the participating general practices. After a second pseudonymisation for patients and general practices, data were transferred to this study’s main researcher.14

    Results

    For this study, data were available from 145 general practices, representing 223 580 enlisted persons with an indication for influenza vaccination. Numbers of outcome events and proportions to respective groups are presented in figure 1.

    Figure 1
    Figure 1

    Total number of persons, by GP-diagnosed COVID-19 status and by all-cause mortality. GP, general practitioner.

    Descriptive data

    Compared with the unvaccinated group, the vaccinated group was older (mean age±SD 67.0±15.2 vs 60.6±16.1), represented relatively more females (53.7% vs 50.1%), had more often comorbidities and had relatively more acute respiratory infection consultations in 2019 (table 1).

    View this table:
    Table 1

    Study population: baseline characteristics

    Main results

    GP-diagnosed COVID-19

    It was found that there was a slightly positive association between influenza vaccination in 2019 and COVID-19 rates (HR 1.15; 95% CI 1.08 to 1.22) (table 2). GP-diagnosed COVID-19 rates were higher for females, for persons with cardiovascular disease, pulmonary disease, diabetes mellitus or chronic renal insufficiency and for persons with one or more GP visits for acute respiratory infections in 2019, and lower for older age groups (≥60 years compared with <60 years) (table 2).

    View this table:
    Table 2

    HRs for GP-diagnosed COVID-19, adjusted for covariables, total group (n=223 580)

    All-cause mortality

    Regarding all-cause mortality rates, a slightly protective effect was found for influenza vaccination (HR 0.90; 95% CI 0.83 to 0.97) (table 3). Higher all-cause mortality rates were found for older age groups, males and for persons with GP-diagnosed COVID-19, cardiovascular disease, pulmonary disease, diabetes mellitus, impaired resistance to infections, chronic renal insufficiency, respiratory disorder by neurological conditions or those with one or more GP visits for acute respiratory infections in 2019 (table 3).

    View this table:
    Table 3

    Adjusted HRs for all-cause mortality and its covariables, total group (n=223 580)

    All-cause mortality, subanalysis for GP-diagnosed COVID-19 persons

    Specifically, for GP-diagnosed COVID-19 persons, no significant association was found between influenza vaccination and all-cause mortality rates, and outcomes for other covariables were comparable to those in the total group of persons in this study (tables 3 and 4).

    View this table:
    Table 4

    Adjusted HRs for all-cause mortality and its covariables, subanalysis COVID-19 group (n=4061)

    Sensitivity analyses

    Sensitivity analyses showed that associations between influenza vaccination and GP-diagnosed COVID-19 rates were consistent over time, as shown by the models for the first and second COVID-19 waves, respectively (online supplemental part IV). All-cause mortality numbers from our study population were similar to those of the total Dutch population and also when different age groups and sexes were compared (online supplemental part V).15

    Discussion

    Key findings

    In this primary care cohort study among 223 580 persons from the target group for influenza vaccination in 2019 in the Netherlands, a slightly positive association was found between influenza vaccination in 2019 and GP-diagnosed COVID-19 rates in 2020 (HR 1.15; 95% CI 1.08 to 1.22). We also observed a slightly negative association between influenza vaccination and all-cause mortality rates (HR 0.90; 95% CI 0.83 to 0.97). For the group GP-diagnosed COVID-19 cases, no significant association was found between influenza vaccination in 2019 and all-cause mortality rates.

    Strengths and limitations

    Important strengths of this study were using a representative sample of the Dutch target group for influenza vaccination, the primary care population-based cohort design, adjustments for relevant covariables and sensitivity analyses. But there were also several limitations. First, there is a possibility of residual confounding by risk factors that could not be evaluated in this study. Literature suggests that such COVID-19 risk factors could be high population density,16 low socioeconomic status accompanied by a possible lack of health-seeking behaviour and insufficient health resources,17 18 belonging to ethnic minorities,16 18–21 being a healthcare worker16 19 22 or low compliance to COVID-19 regulations (eg, difficulties keeping 1.5 m distance because of mental disorders23 or scepticism on COVID-19 existence), high Clinical Frailty Scale scores,24 unhealthy behaviour such as smoking25 26 and high body mass index, especially in combination with low physical activity.27 28 These factors could have been differently distributed between vaccinated and unvaccinated groups.

    Additionally, misclassification could induce information bias because PCR testing options were limited during the first wave. However, our sensitivity analyses found consistent associations between influenza vaccination and GP-diagnosed COVID-19 for the first wave and the second wave (when nationwide PCR testing was available free of charge) (online supplemental part IV) and COVID-19 incidences were similar between our population and total population data.29 All-cause mortality rates were similar in our study population and the general population (online supplemental table SV-1), even though we used ICPC recordings instead of mortality data validated by Statistics Netherlands (CBS). Interestingly, the General Practice Research Consortium Netherlands described under-reporting of COVID-19-related deaths because of a low SARS-CoV-2 testing capacity in the first wave of the pandemic, based on a national registration study.30 However, there is no reason to assume that either misclassification or under-reporting would lead to more missed cases for either unvaccinated or vaccinated groups, and sensitivity analyses indicated the robustness of models as well.

    Interpretation

    Influenza vaccination and GP-diagnosed COVID-19 rates

    The slightly positive association between influenza vaccination and GP-diagnosed COVID-19 was not found in other observational studies. In a cohort of 27 201 patients with a laboratory SARS-CoV-2 test within the Michigan healthcare system, influenza vaccination was protective against COVID-19 (0.76; 95% CI 0.68 to 0.86).31 A similar association was also found in an Italian case-negative study (OR 0.89; 95% CI 0.80 to 0.99)32 and a Spanish cohort with patients from both nursing homes and primary care centres (HR 0.63; 95% CI 0.44 to 0.91).33 A large multicentre primary care study in five European countries found no significant association (OR 0.93; 95% CI 0.66 to 1.32) in a case–control design.34 As we previously discussed, the slight effects that we found may have been caused by residual confounding factors that we could not adjust for, such as socioeconomic and behavioural circumstances, and the use of clinical diagnoses instead of test-confirmed cases. However, also other observational studies are confronted with residual confounding. Moreover, differences between study designs and study populations may be an explanation for the differences in study results. As said before, in contrast to other studies, our study is a primary care population-based cohort study regarding the influenza vaccination target group.

    Similar to our study, other research also found positive associations between comorbidities and COVID-19 incidences.25 31 33 35–37 Comorbidities were more present among the vaccinated group in our study, who therefore may require more assistance in everyday tasks and may live together with more individuals per household, indirectly making them more at risk for COVID-19 transmission. Studies showed that nursing home residence was a predisposing factor for COVID-19,33 and comorbidities were associated with poorer clinical outcomes among COVID-19 cases.38 Health-seeking behaviour could have played a role and is known to be more frequent among women with influenza-like illness,39 which was more found among the vaccinated group. For these reasons, models were adjusted by different covariables including age, sex and comorbidities.

    Influenza vaccination and all-cause mortality rates

    GP-diagnosed COVID-19 was strongly associated with all-cause mortality rates (HR 10.90; 95% CI 8.51 to 13.96), which was also found in another population-based cohort study.40 Although we did not have access to COVID-19-associated mortality data in our study, other research showed that COVID-19 mortality risks were lower after influenza vaccination.9 41 42 Possibly, influenza vaccinations have a wider beneficial impact on health, such as on cardiometabolic diseases, which are associated with COVID-19 hospitalisation or death.43–47 Furthermore, it may protect against other viral respiratory infections2 3 10 which are risk factors48 for the leading global cause of death: cardiovascular diseases.49

    An overall health benefit is plausible as other death causes most likely did not differ between influenza vaccinees and non-vaccinees throughout 2020, such as traffic incidents and natural disasters. The quality of the influenza vaccine was considered good, as circulating influenza subtypes were similar to the strains in the quadrivalent vaccine,50–52 resulting in a mild and short 2019 influenza epidemic in the Netherlands.51 52 Next to the associations we found between influenza vaccination, COVID-19 and all-cause mortality rates and considering that we did not have access to data on different causes of mortality, it is also interesting to elaborate on severe COVID-19 disease in other literature. In a population of patients positive for COVID-19, it was found that severe outcomes were more likely for those who were not vaccinated against influenza, specifically sepsis (RR 1.36–1.45, where RRs were measured 1–4 months after COVID-19 diagnosis), stroke (RR 1.45–1.58), intensive care unit (ICU) admissions (RR 1.17–1.20), deep venous thrombosis events (RR 1.41–1.53) and emergency department visits (RR 1.20–1.58).53 In American studies, influenza vaccination protected against hospitalisation (OR 0.58; 95% CI 0.46 to 0.73),31 resulted in shorter hospital stays (OR 0.76; 95% CI 0.65 to 0.89)31 and reduced the number of ICU admissions.54 Protective effects were also found in a study with test-positive patients with COVID-19 only whereas patients who were not vaccinated against influenza had higher odds to be hospitalised (OR 2.44; 95% CI 1.68 to 3.61) or to be admitted to ICUs (OR 3.29; 95% CI 1.18 to 13.77).55 In a case–control study, however, no significant association was found between influenza vaccination and hospitalisation (OR 1.00; 95% CI 0.84 to 1.29) or mortality (HR 1.14; 95% CI 0.95 to 1.37) in patients with COVID-19.32

    Lastly, in our study, being female was associated with lower all-cause mortality rates (OR 0.82; 95% CI 0.76 to 0.89). Mortality data validated by CBS (Statistics Netherlands) also showed that all-cause mortality rates were higher for males (online supplemental part V).15 Considering international data from 1 million excess deaths in 29 high-income countries, age-standardised excess death rates in 2020 were higher in men compared with women.56 International studies on severe course of COVID-19, including systematic reviews and meta-analyses, also found higher mortality risks for older age as well as for males.26 40 57–66

    Generalisability

    In a broader context than our study setting, these recent findings are representative for the target group influenza vaccination in the Netherlands. We made use of the largest national health monitoring source for infectious disease data, which met high completeness and quality criteria. This resulted in a considerable number of patients (n=223 580) and associations remained consistent in sensitivity analyses.

    Since there are differences in the findings of international studies on the impact of influenza vaccination on COVID-19 incidences, also given the limitations of the various studies, further studies should evaluate these findings. On an immunological level, it is interesting to further investigate to what extent influenza vaccinations trigger trained immunity against other pathogens, and in particular if immunological memory is stimulated, which is known to persist for at least 6–8 months among around 90% from those with a previous COVID-19 infection.67

    Conclusion

    Our study findings contribute to the knowledge about the impact of influenza vaccination on other respiratory illnesses. Our hypothesis of a possible negative association between influenza vaccination in 2019 and GP-diagnosed COVID-19 rates was not confirmed, as we found a slightly positive association. A slightly protective effect on all-cause mortality was found after influenza vaccination, possibly by a wider, overall protective effect on health. Future research designs should include test-confirmed COVID-19 cases and controls, adjustments for behavioural, socioeconomic and ethnic factors and validated disease-specific mortality causes.

    Data availability statement

    Data are available upon reasonable request. The corresponding author can be contacted for details, such as a technical appendix, statistical code and data set.

    Ethics statements

    Patient consent for publication

    Ethics approval

    This study involves human participants and was approved according to the governance code of Nivel Primary Care Database (NZR-00320.070). In addition, this study was approved by the Medical Ethical Committee of Maastricht UMC+ (METC 2020-2377) which assessed this study as not being subject to the Medical Research Involving Human Subjects Act.

    References

    1. ↵
      1. Cascella M,
      2. Rajnik M,
      3. Aleem A
      . Features, evaluation and treatment coronavirus (COVID-19. Treasure Island, FL: StatPearls Publishing, 2020.
    2. ↵
      1. Netea MG,
      2. Domínguez-Andrés J,
      3. Barreiro LB, et al
      . Defining trained immunity and its role in health and disease. Nat Rev Immunol 2020;20:375–88.doi:10.1038/s41577-020-0285-6pmid:http://www.ncbi.nlm.nih.gov/pubmed/32132681
      OpenUrlCrossRefPubMed
    3. ↵
      1. Wolff GG
      . Influenza vaccination and respiratory virus interference among Department of Defense personnel during the 2017-2018 influenza season. Vaccine 2020;38:350–4.doi:10.1016/j.vaccine.2019.10.005pmid:http://www.ncbi.nlm.nih.gov/pubmed/31607599
      OpenUrlCrossRefPubMed
    4. ↵
      1. Wolff GG
      . Letter to the editor. Vaccine 2020;38:4651.doi:10.1016/j.vaccine.2020.04.016pmid:http://www.ncbi.nlm.nih.gov/pubmed/32471778
      OpenUrlPubMed
    5. ↵
      1. Martínez-Baz I,
      2. Trobajo-Sanmartín C,
      3. Arregui I, et al
      . Influenza vaccination and risk of SARS-CoV-2 infection in a cohort of health workers. Vaccines 2020;8:1–7.doi:10.3390/vaccines8040611pmid:http://www.ncbi.nlm.nih.gov/pubmed/33076405
      OpenUrlPubMed
    6. ↵
      1. Massoudi N,
      2. Mohit B
      . A case-control study of the 2019 influenza vaccine and incidence of COVID-19 among healthcare workers. J Clin Immunol 2021;41:324–34.doi:10.1007/s10875-020-00925-0pmid:http://www.ncbi.nlm.nih.gov/pubmed/33244671
      OpenUrlCrossRefPubMed
    7. ↵
      1. Erismis B,
      2. Karabela SN,
      3. Eksi F, et al
      . Annual influenza vaccination effect on the susceptibility to COVID-19 infection. Cent Eur J Public Health 2021;29:14–17.doi:10.21101/cejph.a6573pmid:http://www.ncbi.nlm.nih.gov/pubmed/33831281
      OpenUrlCrossRefPubMed
    8. ↵
      1. Amato M,
      2. Werba JP,
      3. Frigerio B, et al
      . Relationship between influenza vaccination coverage rate and COVID-19 outbreak: an Italian ecological study. Vaccines 2020;8:535.doi:10.3390/vaccines8030535pmid:http://www.ncbi.nlm.nih.gov/pubmed/32947988
      OpenUrlPubMed
    9. ↵
      1. Fink G,
      2. Orlova-Fink N,
      3. Schindler T, et al
      . Inactivated trivalent influenza vaccination is associated with lower mortality among patients with COVID-19 in Brazil. BMJ Evid Based Med 2021;26:192–3.doi:10.1136/bmjebm-2020-111549
      OpenUrlAbstract/FREE Full Text
    10. ↵
      1. Debisarun PA,
      2. Struycken P,
      3. Domínguez-Andrés J
      . The effect of influenza vaccination on trained immunity: impact on COVID-19. medRxiv2020:20212498.
    11. ↵
      1. Govaert TM,
      2. Thijs CT,
      3. Masurel N, et al
      . The efficacy of influenza vaccination in elderly individuals. A randomized double-blind placebo-controlled trial. JAMA 1994;272:1661–5.pmid:http://www.ncbi.nlm.nih.gov/pubmed/7966893
      OpenUrlCrossRefPubMedWeb of Science
    12. ↵
      1. Nederlands Huisartsen Genootschap
      . Richtlijn adequate dossiervorming Met Het elektronisch patiëntendossier (ADEPD). 3rd edn. Utrecht, The Netherlands: Nederlands Huisartsen Genootschap, 2013.
    13. ↵
      1. Heins M,
      2. Hooiveld M,
      3. Korevaar J
      . Monitor vaccinatiegraad nationaal programma grieppreventie 2018. Utrecht, The Netherlands: Nivel, 2019. Available: https://www.nivel.nl/nl/publicatie/monitor-vaccinatiegraad-nationaal-programma-grieppreventie-2018 [Accessed 20 Jul 2020].
    14. ↵
      1. Kuchinke W,
      2. Ohmann C,
      3. Verheij RA, et al
      . A standardised graphic method for describing data privacy frameworks in primary care research using a flexible zone model. Int J Med Inform 2014;83:941–57.doi:10.1016/j.ijmedinf.2014.08.009pmid:http://www.ncbi.nlm.nih.gov/pubmed/25241154
      OpenUrlCrossRefPubMed
    15. ↵
      1. CBS
      . Overledenen; geslacht en leeftijd, per week 2020. Available: https://opendata.cbs.nl/statline/#/CBS/nl/dataset/70895ned/table [Accessed 1 Apr 2021].
    16. ↵
      1. Whitehead M,
      2. Taylor-Robinson D,
      3. Barr B
      . Poverty, health, and covid-19. BMJ 2021;372:n376.doi:10.1136/bmj.n376pmid:http://www.ncbi.nlm.nih.gov/pubmed/33579719
      OpenUrlFREE Full Text
    17. ↵
      1. Zhu Y,
      2. Duan M-J,
      3. Dijk HH, et al
      . Association between socioeconomic status and self-reported, tested and diagnosed COVID-19 status during the first wave in the Northern Netherlands: a general population-based cohort from 49 474 adults. BMJ Open 2021;11:e048020.doi:10.1136/bmjopen-2020-048020pmid:http://www.ncbi.nlm.nih.gov/pubmed/33753448
      OpenUrlAbstract/FREE Full Text
    18. ↵
      1. Singh BM,
      2. Bateman J,
      3. Viswanath A, et al
      . Risk of COVID-19 hospital admission and COVID-19 mortality during the first COVID-19 wave with a special emphasis on ethnic minorities: an observational study of a single, deprived, multiethnic UK health economy. BMJ Open 2021;11:e046556.doi:10.1136/bmjopen-2020-046556pmid:http://www.ncbi.nlm.nih.gov/pubmed/33597146
      OpenUrlPubMed
    19. ↵
      1. Clough HE,
      2. McIntyre KM,
      3. Patterson GE, et al
      . Use of routine death and illness surveillance data to provide insight for UK pandemic planning: lessons from COVID-19. BMJ Open 2021;11:e044707.doi:10.1136/bmjopen-2020-044707pmid:http://www.ncbi.nlm.nih.gov/pubmed/33558359
      OpenUrlAbstract/FREE Full Text
    20. ↵
      1. Apea VJ,
      2. Wan YI,
      3. Dhairyawan R, et al
      . Ethnicity and outcomes in patients hospitalised with COVID-19 infection in East London: an observational cohort study. BMJ Open 2021;11:e042140.doi:10.1136/bmjopen-2020-042140pmid:http://www.ncbi.nlm.nih.gov/pubmed/33455936
      OpenUrlAbstract/FREE Full Text
    21. ↵
      1. Thomas DR,
      2. Orife O,
      3. Plimmer A, et al
      . Ethnic variation in outcome of people hospitalised during the first COVID-19 epidemic wave in Wales (UK): an analysis of national surveillance data using Onomap, a name-based ethnicity classification tool. BMJ Open 2021;11:e048335.doi:10.1136/bmjopen-2020-048335pmid:http://www.ncbi.nlm.nih.gov/pubmed/34408047
      OpenUrlAbstract/FREE Full Text
    22. ↵
      1. Mortgat L,
      2. Verdonck K,
      3. Hutse V, et al
      . Prevalence and incidence of anti-SARS-CoV-2 antibodies among healthcare workers in Belgian hospitals before vaccination: a prospective cohort study. BMJ Open 2021;11:e050824.doi:10.1136/bmjopen-2021-050824pmid:http://www.ncbi.nlm.nih.gov/pubmed/34187832
      OpenUrlAbstract/FREE Full Text
    23. ↵
      1. Francis NA,
      2. Stuart B,
      3. Knight M, et al
      . Predictors of clinical deterioration in patients with suspected COVID-19 managed in a 'virtual hospital' setting: a cohort study. BMJ Open 2021;11:e045356.doi:10.1136/bmjopen-2020-045356pmid:http://www.ncbi.nlm.nih.gov/pubmed/33757955
      OpenUrlAbstract/FREE Full Text
    24. ↵
      1. Hewitt J,
      2. Carter B,
      3. Vilches-Moraga A, et al
      . The effect of frailty on survival in patients with COVID-19 (COPE): a multicentre, European, observational cohort study. Lancet Public Health 2020;5:e444–51.doi:10.1016/S2468-2667(20)30146-8pmid:http://www.ncbi.nlm.nih.gov/pubmed/32619408
      OpenUrlPubMed
    25. ↵
      1. Hamer M,
      2. Kivimäki M,
      3. Gale CR, et al
      . Lifestyle risk factors, inflammatory mechanisms, and COVID-19 hospitalization: a community-based cohort study of 387,109 adults in UK. Brain Behav Immun 2020;87:184–7.doi:10.1016/j.bbi.2020.05.059pmid:http://www.ncbi.nlm.nih.gov/pubmed/32454138
      OpenUrlCrossRefPubMed
    26. ↵
      1. Ho FK,
      2. Celis-Morales CA,
      3. Gray SR, et al
      . Modifiable and non-modifiable risk factors for COVID-19, and comparison to risk factors for influenza and pneumonia: results from a UK Biobank prospective cohort study. BMJ Open 2020;10:e040402.doi:10.1136/bmjopen-2020-040402pmid:http://www.ncbi.nlm.nih.gov/pubmed/33444201
      OpenUrlAbstract/FREE Full Text
    27. ↵
      1. Mahase E
      . Covid-19: why are age and obesity risk factors for serious disease? BMJ 2020;371:4130.doi:10.1136/bmj.m4130pmid:http://www.ncbi.nlm.nih.gov/pubmed/33106243
      OpenUrlPubMed
    28. ↵
      1. Hamrouni M,
      2. Roberts MJ,
      3. Thackray A, et al
      . Associations of obesity, physical activity level, inflammation and cardiometabolic health with COVID-19 mortality: a prospective analysis of the UK Biobank cohort. BMJ Open 2021;11:e055003.doi:10.1136/bmjopen-2021-055003pmid:http://www.ncbi.nlm.nih.gov/pubmed/34732503
      OpenUrlPubMed
    29. ↵
      1. RIVM
      . Coronadashboard - Besmettingen - Positief geteste mensen, 2020. Available: https://coronadashboard.rijksoverheid.nl/landelijk/positief-geteste-mensen [Accessed 6 Jan 2021].
    30. ↵
      1. Cals JWL,
      2. Derckx R,
      3. Blanker MH
      . Intensieve en palliatieve covid-19-zorg door huisarts. Ned Tijdschr Geneeskd 2020;164:D5616.
    31. ↵
      1. Conlon A,
      2. Ashur C,
      3. Washer L, et al
      . Impact of the influenza vaccine on COVID-19 infection rates and severity. Am J Infect Control 2021;49:694–700.doi:10.1016/j.ajic.2021.02.012pmid:http://www.ncbi.nlm.nih.gov/pubmed/33631305
      OpenUrlCrossRefPubMed
    32. ↵
      1. Ragni P,
      2. Marino M,
      3. Formisano D, et al
      . Association between exposure to influenza vaccination and COVID-19 diagnosis and outcomes. Vaccines 2020;8:675.doi:10.3390/vaccines8040675pmid:http://www.ncbi.nlm.nih.gov/pubmed/33198368
      OpenUrlPubMed
    33. ↵
      1. Vila-Córcoles A,
      2. Ochoa-Gondar O,
      3. Satué-Gracia EM, et al
      . Influence of prior comorbidities and chronic medications use on the risk of COVID-19 in adults: a population-based cohort study in Tarragona, Spain. BMJ Open 2020;10:e041577.doi:10.1136/bmjopen-2020-041577pmid:http://www.ncbi.nlm.nih.gov/pubmed/33303459
      OpenUrlAbstract/FREE Full Text
    34. ↵
      1. Kissling E,
      2. Hooiveld M,
      3. Brytting M, et al
      . Absence of association between 2019-20 influenza vaccination and COVID-19: results of the European I-MOVE-COVID-19 primary care project, March-August 2020. Influenza Other Respir Viruses 2021;15:429–38.doi:10.1111/irv.12839pmid:http://www.ncbi.nlm.nih.gov/pubmed/33481344
      OpenUrlCrossRefPubMed
    35. ↵
      1. Docherty AB,
      2. Harrison EM,
      3. Green CA
      . Features of 20 133 UK patients in hospital with Covid-19 using the ISARIC WHO Clinical Characterisation Protocol: prospective observational cohort study. BMJ 2020;369:1985.
    36. ↵
      1. Piroth L,
      2. Cottenet J,
      3. Mariet A-S, et al
      . Comparison of the characteristics, morbidity, and mortality of COVID-19 and seasonal influenza: a nationwide, population-based retrospective cohort study. Lancet Respir Med 2021;9:251–9.doi:10.1016/S2213-2600(20)30527-0pmid:http://www.ncbi.nlm.nih.gov/pubmed/33341155
      OpenUrlPubMed
    37. ↵
      1. Ortega E,
      2. Corcoy R,
      3. Gratacòs M, et al
      . Risk factors for severe outcomes in people with diabetes hospitalised for COVID-19: a cross-sectional database study. BMJ Open 2021;11:e051237.doi:10.1136/bmjopen-2021-051237pmid:http://www.ncbi.nlm.nih.gov/pubmed/34301668
      OpenUrlPubMed
    38. ↵
      1. Guan W-J,
      2. Liang W-H,
      3. Zhao Y, et al
      . Comorbidity and its impact on 1590 patients with COVID-19 in China: a nationwide analysis. Eur Respir J 2020;55:2000547.doi:10.1183/13993003.00547-2020pmid:http://www.ncbi.nlm.nih.gov/pubmed/32217650
      OpenUrlAbstract/FREE Full Text
    39. ↵
      1. Peppa M,
      2. John Edmunds W,
      3. Funk S
      . Disease severity determines health-seeking behaviour amongst individuals with influenza-like illness in an internet-based cohort. BMC Infect Dis 2017;17:238.doi:10.1186/s12879-017-2337-5pmid:http://www.ncbi.nlm.nih.gov/pubmed/28359335
      OpenUrlCrossRefPubMed
    40. ↵
      1. Hothorn T,
      2. Bopp M,
      3. Günthard H, et al
      . Assessing relative COVID-19 mortality: a Swiss population-based study. BMJ Open 2021;11:e042387.doi:10.1136/bmjopen-2020-042387pmid:http://www.ncbi.nlm.nih.gov/pubmed/34006026
      OpenUrlAbstract/FREE Full Text
    41. ↵
      1. Candelli M,
      2. Pignataro G,
      3. Torelli E, et al
      . Effect of influenza vaccine on COVID-19 mortality: a retrospective study. Intern Emerg Med 2021;16:1849–55.doi:10.1007/s11739-021-02702-2pmid:http://www.ncbi.nlm.nih.gov/pubmed/33743150
      OpenUrlCrossRefPubMed
    42. ↵
      1. Zanettini C,
      2. Omar M,
      3. Dinalankara W, et al
      . Influenza vaccination and COVID-19 mortality in the USA: an ecological study. Vaccines 2021;9:427.doi:10.3390/vaccines9050427pmid:http://www.ncbi.nlm.nih.gov/pubmed/33923159
      OpenUrlPubMed
    43. ↵
      1. Svensson P,
      2. Hofmann R,
      3. Häbel H, et al
      . Association between cardiometabolic disease and severe COVID-19: a nationwide case-control study of patients requiring invasive mechanical ventilation. BMJ Open 2021;11:e044486.doi:10.1136/bmjopen-2020-044486pmid:http://www.ncbi.nlm.nih.gov/pubmed/33597145
      OpenUrlAbstract/FREE Full Text
    44. ↵
      1. Tornhammar P,
      2. Jernberg T,
      3. Bergström G, et al
      . Association of cardiometabolic risk factors with hospitalisation or death due to COVID-19: population-based cohort study in Sweden (SCAPIS). BMJ Open 2021;11:e051359.doi:10.1136/bmjopen-2021-051359pmid:http://www.ncbi.nlm.nih.gov/pubmed/34475186
      OpenUrlAbstract/FREE Full Text
    45. ↵
      1. Wollenstein-Betech S,
      2. Cassandras CG,
      3. Paschalidis IC
      . Personalized predictive models for symptomatic COVID-19 patients using basic preconditions: hospitalizations, mortality, and the need for an ICU or ventilator. Int J Med Inform 2020;142:104258.doi:10.1016/j.ijmedinf.2020.104258pmid:http://www.ncbi.nlm.nih.gov/pubmed/32927229
      OpenUrlPubMed
    46. ↵
      1. Gu T,
      2. Chu Q,
      3. Yu Z, et al
      . History of coronary heart disease increased the mortality rate of patients with COVID-19: a nested case-control study. BMJ Open 2020;10:e038976.doi:10.1136/bmjopen-2020-038976pmid:http://www.ncbi.nlm.nih.gov/pubmed/32948572
      OpenUrlAbstract/FREE Full Text
    47. ↵
      1. Basu A,
      2. Agwu JC,
      3. Barlow N, et al
      . Hypertension is the major predictor of poor outcomes among inpatients with COVID-19 infection in the UK: a retrospective cohort study. BMJ Open 2021;11:e047561.doi:10.1136/bmjopen-2020-047561pmid:http://www.ncbi.nlm.nih.gov/pubmed/34172549
      OpenUrlAbstract/FREE Full Text
    48. ↵
      1. Behrouzi B,
      2. Araujo Campoverde MV,
      3. Liang K, et al
      . Influenza vaccination to reduce cardiovascular morbidity and mortality in patients with COVID-19: JACC state-of-the-art review. J Am Coll Cardiol 2020;76:1777–94.doi:10.1016/j.jacc.2020.08.028pmid:http://www.ncbi.nlm.nih.gov/pubmed/33032740
      OpenUrlFREE Full Text
    49. ↵
      1. World Health Organization
      . Cardiovascular diseases (CVDs), 2021. Available: https://www.who.int/news-room/fact-sheets/detail/cardiovascular-diseases-%28cvds%29 [Accessed 20 Jul 2021].
    50. ↵
      1. Schurink-Van 't Klooster TM,
      2. Van Gageldonk-Lafeber A,
      3. Wallinga J
      . Influenza vaccination in the Netherlands. Background information for the Health Council of the Netherlands. Bilthoven, The Netherlands, 2019. Available: https://www.rivm.nl/bibliotheek/rapporten/2019-0002.pdf
    51. ↵
      1. World Health Organization
      . FluNet – influenza laboratory surveillance information by the global influenza surveillance and response system (GISRS) – Netherlands. Available: https://www.who.int/initiatives/global-influenza-surveillance-and-response-system [Accessed 10 Jan 2021].
    52. ↵
      1. RIVM
      . Griepprik – Feiten en cijfers. Available: https://www.rivm.nl/griep-griepprik/feiten-en-cijfers [Accessed 1 Feb 2021].
    53. ↵
      1. Taghioff SM,
      2. Slavin BR,
      3. Holton T, et al
      . Examining the potential benefits of the influenza vaccine against SARS-CoV-2: a retrospective cohort analysis of 74,754 patients. PLoS One 2021;16:e0255541.doi:10.1371/journal.pone.0255541pmid:http://www.ncbi.nlm.nih.gov/pubmed/34343191
      OpenUrlCrossRefPubMed
    54. ↵
      1. Umasabor-Bubu OQ,
      2. Bubu OM,
      3. Mbah AK, et al
      . Association between influenza vaccination and severe COVID-19 outcomes at a designated COVID-only hospital in Brooklyn. Am J Infect Control 2021;49:1327–30.doi:10.1016/j.ajic.2021.04.006pmid:http://www.ncbi.nlm.nih.gov/pubmed/33891988
      OpenUrlPubMed
    55. ↵
      1. Yang M-J,
      2. Rooks BJ,
      3. Le T-TT, et al
      . Influenza vaccination and hospitalizations among COVID-19 infected adults. J Am Board Fam Med 2021;34:S179–82.doi:10.3122/jabfm.2021.S1.200528pmid:http://www.ncbi.nlm.nih.gov/pubmed/33622834
      OpenUrlAbstract/FREE Full Text
    56. ↵
      1. Islam N,
      2. Shkolnikov VM,
      3. Acosta RJ, et al
      . Excess deaths associated with covid-19 pandemic in 2020: age and sex disaggregated time series analysis in 29 high income countries. BMJ 2021;373:1137.doi:10.1136/bmj.n1137pmid:http://www.ncbi.nlm.nih.gov/pubmed/34011491
      OpenUrlPubMed
    57. ↵
      1. Vrotsou K,
      2. Rotaeche R,
      3. Mateo-Abad M, et al
      . Variables associated with COVID-19 severity: an observational study of non-paediatric confirmed cases from the general population of the Basque Country, Spain. BMJ Open 2021;11:e049066.doi:10.1136/bmjopen-2021-049066pmid:http://www.ncbi.nlm.nih.gov/pubmed/33795313
      OpenUrlAbstract/FREE Full Text
    58. ↵
      1. Jiménez E,
      2. Fontán-Vela M,
      3. Valencia J, et al
      . Characteristics, complications and outcomes among 1549 patients hospitalised with COVID-19 in a secondary hospital in Madrid, Spain: a retrospective case series study. BMJ Open 2020;10:e042398.doi:10.1136/bmjopen-2020-042398pmid:http://www.ncbi.nlm.nih.gov/pubmed/33172949
      OpenUrlAbstract/FREE Full Text
    59. ↵
      1. Nachtigall I,
      2. Lenga P,
      3. Jóźwiak K, et al
      . Clinical course and factors associated with outcomes among 1904 patients hospitalized with COVID-19 in Germany: an observational study. Clin Microbiol Infect 2020;26:1663–9.doi:10.1016/j.cmi.2020.08.011pmid:http://www.ncbi.nlm.nih.gov/pubmed/32822883
      OpenUrlCrossRefPubMed
    60. ↵
      1. Pastor-Barriuso R,
      2. Pérez-Gómez B,
      3. Hernán MA, et al
      . Infection fatality risk for SARS-CoV-2 in community dwelling population of Spain: nationwide seroepidemiological study. BMJ 2020;371:4509.doi:10.1136/bmj.m4509pmid:http://www.ncbi.nlm.nih.gov/pubmed/33246972
      OpenUrlPubMed
    61. ↵
      1. Pijls BG,
      2. Jolani S,
      3. Atherley A, et al
      . Demographic risk factors for COVID-19 infection, severity, ICU admission and death: a meta-analysis of 59 studies. BMJ Open 2021;11:e044640.doi:10.1136/bmjopen-2020-044640pmid:http://www.ncbi.nlm.nih.gov/pubmed/33431495
      OpenUrlAbstract/FREE Full Text
    62. ↵
      1. Pedote PD,
      2. Termite S,
      3. Gigliobianco A, et al
      . Influenza vaccination and health outcomes in COVID-19 patients: a retrospective cohort study. Vaccines 2021;9:358.doi:10.3390/vaccines9040358pmid:http://www.ncbi.nlm.nih.gov/pubmed/33917898
      OpenUrlPubMed
    63. ↵
      1. Wingert A,
      2. Pillay J,
      3. Gates M, et al
      . Risk factors for severity of COVID-19: a rapid review to inform vaccine prioritisation in Canada. BMJ Open 2021;11:e044684.doi:10.1136/bmjopen-2020-044684pmid:http://www.ncbi.nlm.nih.gov/pubmed/33986052
      OpenUrlAbstract/FREE Full Text
    64. ↵
      1. Abate BB,
      2. Kassie AM,
      3. Kassaw MW, et al
      . Sex difference in coronavirus disease (COVID-19): a systematic review and meta-analysis. BMJ Open 2020;10:e040129.doi:10.1136/bmjopen-2020-040129pmid:http://www.ncbi.nlm.nih.gov/pubmed/33028563
      OpenUrlCrossRefPubMed
    65. ↵
      1. Fumagalli C,
      2. Rozzini R,
      3. Vannini M, et al
      . Clinical risk score to predict in-hospital mortality in COVID-19 patients: a retrospective cohort study. BMJ Open 2020;10:e040729.doi:10.1136/bmjopen-2020-040729pmid:http://www.ncbi.nlm.nih.gov/pubmed/32978207
      OpenUrlAbstract/FREE Full Text
    66. ↵
      1. Ou M,
      2. Zhu J,
      3. Ji P, et al
      . Risk factors of severe cases with COVID-19: a meta-analysis. Epidemiol Infect 2020;148:e175.doi:10.1017/S095026882000179Xpmid:http://www.ncbi.nlm.nih.gov/pubmed/32782035
      OpenUrlPubMed
    67. ↵
      1. Chivese T,
      2. Matizanadzo JT,
      3. Musa OAH
      . The prevalence of adaptive immunity to COVID-19 and reinfection after recovery – a comprehensive systematic review and meta-analysis of 12 011 447 individuals. medRxiv 2021:21263103.

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Footnotes

    • Contributors G-JD (principal guarantor) is responsible for the overall content, together with AvL. They contributed to the designing, planning, conducting, interpretation, reporting, revision and submission of the work. RV contributed to the study design and revision of the work. AK was responsible for conceptualising, designing and planning the study, and contributed to the statistical analysis, interpretation and reporting of the work. MH contributed to data collection, interpretation and revision of the work. BW contributed to conducting the work, interpretation (in particular, statistical analyses) and revision of the work.

    • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

    • Competing interests None declared.

    • Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting or dissemination plans of this research.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.