Background: Early-season (interim) rather than end-season (final) seasonal influenza vaccine (SIV) effectiveness estimates from surveillance studies are utilised for decision-making regarding SIV components. Despite, concordance between these estimates is underexplored. Further, SIV effectiveness suggestively varies across geographical regions, population characteristics, vaccine antigenic similarity with circulating virus strains (VAS), and study methods, but the evidence has largely conflicted. This thesis aimed to address these gaps in knowledge. Methods: Systematic evidence reviews with meta-analyses and a mixed methods study, adhering to standard guidelines. Included published full-text articles of SIV effectiveness against laboratory-confirmed (RT-PCR/culture) influenza from test-negative design (TND) studies in outpatients after the 2009/10 influenza pandemic, with patients vaccinated against influenza ≥14days before symptom onset and symptom onset ≤7days. Multivariable logistic regression was used to assess determinants of a substantial difference (≥10%) between paired adjusted point interim/final estimates. Pooled adjusted final estimates against A(H1N1)pdm09, A(H3N2), and influenza B were calculated using an inverse variance, random-effects model. The chi-square statistic (χ2) was used to assess the statistical significance of the difference between estimates. Results: There were 68 pairs of interim/final estimates, with no statistically significant difference between almost all. Inconsistent statistical model in estimations, and interim estimation before influenza circulation peak increased the odds of having a substantial difference between paired estimates. Point pooled estimates were higher in the Southern compared with Northern hemispheres, with statistically significant difference between almost all pooled estimates (76 articles). Pooled estimates decreased with increasing age in the Northern hemisphere, and were almost entirely statistically significantly higher with vaccine antigenically similar compared with dissimilar to circulating virus strains. Pooled estimates were higher with self-reported vaccination compared with from medical records, almost entirely higher with respiratory specimen collection ≤7days compared with ≤4days of symptom onset, and higher with adjustment for age but not medical conditions compared with adjustment for both (70 articles); however, mostly non-statistically significant. Conclusions: Interim SIV effectiveness estimates seem sufficient for SIV composition decision-making. Variations in final estimates across geographical regions and age groups, and according to VAS and study methods necessitate consideration of these factors when designing, evaluating or comparing TND studies of SIV effectiveness.