Table 1.

Conceptual framework: Indicators and conclusions for presence of confounding by indication and healthy vaccinee bias in influenza vaccine effectiveness studies

Indicator	Conclusion	References
Vaccinated study participants have a higher proportion of comorbidities than unvaccinated study participants, as indicated by baseline characteristics	High risk of confounding by indication in the unadjusted data set	[6, 38]
Vaccinated study participants have a lower proportion of comorbidities than unvaccinated study participants, as indicated by baseline characteristics	High risk of healthy vaccinee bias in the unadjusted data set	[35, 36]
Inclusion of comorbidities in the regression model increases vaccine effectiveness	Confounding by indication has led to underestimation of vaccine effectiveness in the unadjusted data set	[7]
Inclusion of comorbidities in the regression model decreases vaccine effectiveness	Healthy vaccinee bias has led to overestimation of vaccine effectiveness in the unadjusted data set	[7]
Significant effects of influenza vaccination appear outside the influenza season (“off-season estimates”), despite adjustment for comorbidities	Residual confounding by healthy vaccinee bias	[3, 11, 36]