Correlation of Influenza Vaccination and Influenza Incidence on COVID-19 Severity and Other Perspectives
80 Pages Posted: 10 Apr 2020 Last revised: 29 Jul 2021
Date Written: April 10, 2020
The pandemic of Covid-19 is evolving worldwide, and it is associated with high mortality and morbidity. There is a growing need to discuss the elements of a coordinated strategy to control the spread and mitigate the severity and mortality of Covid-19. H1N1 vaccine and streptococcus pneumonia vaccines are available. The current analysis was performed to correlate the severity of Covid-19 and influenza (H1N1) vaccination statistics and also influenza lower respiratory tract incidence. There is a correlation between Covid-19 related mortality and morbidity and influenza vaccination status, which appears protective. The tendency of correlation is more visualized as the pandemic is evolving. The case incidence and recovery parameters also showed a beneficial trend. Since evolutionarily, influenza is close to SARS-CoV-2 viruses and shares some common epitopes and mechanisms, partial protection is possible to reduce the Covid-19 related severity using the influenza vaccination. In countries where influenza immunization is less, there is a correlation between lower respiratory tract infections (LRI) and influenza attributable to lower respiratory tract infections incidence and Covid-19 severity, which is beneficial. WHO regions with an increased incidence of LRI associated mortality was associated with a reduction in Covid19 mortality by about 16%. Independently low values of the influenza LRI parameter are significantly associated with all levels of Covid-19 mortality(>200 to >1000/million, P<0.05 for all, December 12, 2020). The influenza LRI*population density parameter was significantly associated with Covid19 mortality >200 to >800/million (P<0.05, for each). In the logistic regression model inclusive of parameters - influenza LRI incidence, population density, influenza LRI*population density and total population in numbers, low influenza LRI incidence to indicate Covid19 mortality >400/million had beta coefficient of -0.501, CI -0.971 to -0.031 p=0.037, >500/million the beta coefficient was -0.59, CI -1.116 to -0.065, p=0.028, >600/million the beta coefficient was -0.541, CI -1.068 to -0.013, P=0.045, and for Covid19 mortality >700/million the beta coefficient was -0.694, CI -1.414 to +0.026 P= 0.059 (December 12, 2020). The calculated odds ratio from beta coefficient, i.e., odds ratio=exp(beta coefficient) shows a significant odds ratios for influenza LRI parameter which is about 0.4 and 1-odds ratio of about 0.6 or 60% protection if the parameter is corrected. For influenza LRI incidence*population density the calculated odds ratio for an average beta coefficient [-2.5 and e^(-2.5)=0.08] across various cut-offs was 0.08 and 1-odds ratio of 0.92 which is nearing 1, which indicates the event (=1 or Covid 19 mortality) would happen in these areas with low influenza LRI and low population density which is discussed in the paper. ROC curve analysis (December 11 data) shows a cut off of 397/100 000 has a sensitivity of 88.9% (54 to 99.8) and specificity of 76.3% (69.4 to 82) and PPV (positive predictive value) 16.3 and NPV (negative predictive value) 99.2 and (likelihood ratio, LR) LR+ 3.751 and LR- 0.146 with an AUC of 0.837 to indicate Covid19 mortality of >900/million. The low influenza LRI parameter is associated with Covid19 case fatality rates of >2 percent.
Latest results: In the logistic regression model inclusive of the parameters-influenza LRI incidence, Influenza LRI*Population density, population density and population in numbers, influenza LRI parameter had beta or standardised coefficients of -1.192 (-2.168 to -0.216, P=0.017, data Jan 9, 2021) to indicate Covid19 mortality >1000/million (corresponding odds ratios 0.303 CI 0.11 to 0.805. By partial least square regression (PLS-R regression) the influenza LRI had standardised coefficient of -0.21(-0.34 to -0.08), and for influenza LRI*population density the standardised coefficient was -0.092 (-0.142 to 0.041). For Covid19 mortality >400/million the influenza LRI*population density had a beta regression value of -2.1(CI -3.63 to -0.6, P=0.006, Jan 9, 2021) and the corresponding odds ratio was 0.122, CI 0.026 to 0.548. By partial least square (PLS) regression the influenza LRI had a standardised coefficient of -0.133 (CI-0.286 to 0.02), and influenza LRI*population density had a standardised coefficient of -0.129(CI-0.18 to 0.07). ROC curve shows cut-off of influenza LRI parameter 477/100 000 has a sensitivity of 93.8%, specificity of 75.9%, PPV 27.3% and NPV 99.2%, LR+ 3.891 and LR- 0.082 to indicate a Covid19 mortality of >1000/million. Independent, influenza LRI and Influenza LRI*population density is associated with Covid19 mortality of various cut-offs from 300 to 1000/million (P<0.05, for each, Jan 9, 2021).
Conclusion: In countries or regions with low influenza LRI incidence, influenza vaccine (single dose) will be protective against Covid19 severity. This strategy will be more beneficial in countries/areas with lesser population density. Also, the addition of pneumococcal vaccines (single dose) would further lessen Covid19 severity in areas with a low incidence of influenza LRI and lower respiratory tract infections. These are useful interim measures before Covid19 vaccines are delivered in the routine practise worldwide.
Note: Peer reviewed article: Considering Interim Interventions to Control COVID-19 Associated Morbidity and Mortality—Perspectives. Front Public Health 8:444. https://doi.org/10.3389/fpubh.2020.00444 (Sept 22, 2020)
Conflict of Interest: None.
Keywords: COVID-19, Influenza vaccination, Influenza lower respiratory tract infections, Mortality, Morbidity, Immunity
JEL Classification: I, E
Suggested Citation: Suggested Citation