During 2010–2019, seasonal influenza caused an estimated 9.4–41 million symptomatic illnesses, 100,000–710,000 hospitalizations, and 4,900–52,000 deaths annually in the United States [1]. Influenza vaccination provides protection against influenza-associated illness, hospitalization, and death, with estimated effectiveness of 40 to 60 percent during most seasons when circulating influenza viruses are similar to vaccine components [2]. The U.S. Centers for Disease Control and Prevention (CDC) provides both interim and final estimates of influenza vaccine effectiveness (VE) to assess effects of vaccination each season [2–5]. Evaluation of vaccine effectiveness informs immunization recommendations, policies surrounding vaccination timing, and communications regarding prevention of influenza, antiviral treatment and nonpharmaceutical interventions. VE platforms also document effects of intra-season influenza viral genetic changes and waning immunity [6–8]. Surveillance systems that link individual patient influenza test results and immunization records can provide early season indication of vaccine protection.

Multiple information systems support influenza surveillance in California, including laboratory reports, ambulatory consultations, hospitalizations, and influenza-associated deaths [9]. Influenza-positive laboratory test results became reportable October 1, 2019, including all influenza test results since June 15, 2023 [10]. Mandatory influenza vaccine reporting to the state immunization information system began January 1, 2023. These influenza reporting improvements were made for improved geographic coverage for respiratory virus surveillance and pandemic preparedness. To assess use of linked laboratory reporting and immunization records, the California Department of Public Health (CDPH) conducted an analysis of VE against laboratory-confirmed influenza during October 2023–June 2024, including VE against influenza A and B by age group and time since vaccination.

Our findings demonstrated feasibility of monitoring effectiveness of seasonal influenza vaccination using public health information systems that linked influenza vaccination records to influenza laboratory test results (including negative results) [11]. Influenza vaccination reduced the odds of laboratory-confirmed influenza by 41%. Although we were unable to assess effectiveness by illness severity, estimates from these information systems were consistent with those from U.S. and Canadian estimates for the 2023–2024 season, especially those from ambulatory settings [3, 14]. Previous studies have documented feasibility of assessing vaccine effectiveness using routinely reported public health data and compared results to traditional test-negative design studies [15]. However, consistency of VE estimated from surveillance data depends on data completeness and linkages between information systems [16]. Influenza test positivity and distribution of circulating influenza virus types in laboratory testing data were similar to overall influenza activity in California (Supplemental Figure 1), but lower compared with other multi-center U.S. studies [17, 18]. California’s laboratory reporting system includes all influenza test results rather than only tests among patients with influenza-like illness. While most influenza testing occurred in outpatient settings, laboratory reporting did not indicate whether tests were ordered for ambulatory or hospitalized patients.

Mandatory reporting requirements in California contributed to data completeness for both influenza testing and immunization data, with a high degree of matching between information systems. Although not all influenza immunizations were reported to CAIR, coverage estimates for children aged <18 years and adults aged ≥65 years in CAIR were similar to estimates from the National Immunization Survey and the Behavioral Risk Factor Surveillance System [19, 20]. Linkage between California’s state immunization information system and laboratory reporting can help assess vaccine effectiveness during future seasons and has the potential to provide early season estimates.

Lower estimates of vaccine effectiveness with increasing patient age are consistent with findings of a meta-analysis of test-negative studies conducted during 2004–2015 [21]. Age-related factors that contribute to lower VE might include diminished adaptive immune response, increased prevalence of comorbidities, and frailty [22]. Influenza VE estimates may inform strategies to communicate benefits of vaccination to persons at higher risk, in addition to importance of early influenza antiviral treatment to reduce the risk for severe influenza-related complications [23]. Demonstrated effectiveness among children inform providers and parents to increase influenza vaccination rates in children. Our findings should be examined alongside VE studies that assess clinical outcomes such as hospitalization and death; vaccination is likely to reduce the risk for severe influenza outcomes, even when not preventing infection [24].

Because vaccine effectiveness varies by influenza virus type and antigenic difference between vaccine components and circulating viruses [25, 26], vaccine effectiveness requires monitoring every season. Based on observed VE for influenza A and B, vaccines were protective during the 2023–24 respiratory season when vaccine components were antigenically similar to circulating viruses [27]. In California, approximately 85% of positive influenza specimens tested at Respiratory Laboratory Network facilities were influenza A viruses, of which 73% were A(H1N1)pdm09 viruses [28]. Because most clinical laboratory tests identify influenza A and B, and influenza A subtyping is not routine, virus subtype was only available for 5% of results. VE against influenza A subtypes, including A(H1N1)pdm09- and A(H3N2)-specific estimates should be interpreted with caution. Influenza A positive specimens from more severe or hospitalized cases may have been more likely to be submitted to public health laboratories for subtyping and be less representative of tests performed. We also observed higher VE against influenza B late in the influenza season in contrast to lower estimates observed against influenza A. Differences in viral evolution may contribute to waning vaccine effectiveness or immune escape [29].

Documentation of vaccine product in California’s immunization information system allowed estimation of VE by vaccine type, which may inform specific vaccine recommendations. CDC’s Advisory Committee for Immunization Practices preferentially recommends that adults aged ≥65 years receive a high-dose, recombinant, or adjuvanted influenza vaccine [13]. While all products were associated with protection, confidence intervals were wide and differences between product-specific estimates were not statistically significant among adults aged ≥65 or children aged 6 months to 17 years.

The 2023–2024 season was the first since mandatory submission of influenza immunizations to California’s state information system and reporting of laboratory results, including negative results of influenza testing. Substantial increases in influenza test positivity during December 2023 and January 2024 coincided with indicators of influenza activity from other surveillance systems in California [28]. During January–June 2024, cumulative estimates of VE for the season were unchanged, whereas influenza activity decreased. However, we observed lower estimated VE against both influenza A and B beginning 60 to 89 days after vaccination, with the lowest estimates (including null estimates) occurring ≥180 days after vaccination. With most vaccinations occurring during September or October, protection was highest during peak influenza activity. After peak influenza activity, VE estimates might be lower late in the season due to higher attack rates among unvaccinated persons, referred to as differential depletion of susceptible persons [30, 31]. Cumulative VE estimates provide a better indication of protection during a season than when stratified by period [31].

The main strengths of this analysis were data timeliness and the substantial sample size to calculate VE for specific populations (i.e., those receiving a specific vaccine). We reported interim estimates of 2023–2024 influenza vaccine effectiveness in California in February 2024, when interim U.S. estimates were reported [3,11]. Depending on timing of influenza activity in California, surveillance data may provide early estimates due to the timeliness of reporting and large numbers of influenza tests performed. During 2023–2024, the number of influenza tests reported during October–January represented 53% of the total number for the season. The majority (>95%) of influenza laboratory results and influenza vaccination records are available in CalREDIE and CAIR, respectively, within one week, allowing for near real time VE calculations. The captured population is also diverse, allowing for representativeness of California’s population. Mandatory reporting of laboratory and vaccine data in real time allows public health agencies to evaluate vaccine protection early during the influenza season, when developing communication strategies is crucial. The large population of California and these information systems might provide information about effectiveness of specific or new products recommended for children or older adults aged ≥65 years.

This analysis is subject to multiple limitations. First, incomplete influenza test results and influenza vaccinations during the first year of mandatory reporting may have resulted in misclassification of unvaccinated persons, under ascertainment of cases (i.e. unreported positive tests) and bias due to under reporting of negative test results. Second, laboratory data did not include influenza-associated symptoms, care setting, or illness outcomes. VE against medically attended acute respiratory infection, hospitalization, or death could not be calculated. Estimates of VE might have been biased if case-patients and control subjects were tested in different clinical settings, such that persons without influenza had less severe illness than persons who tested positive. Third, differences among preexisting conditions associated with likelihood of severe influenza, healthcare seeking, testing behavior, vaccination history, and prior influenza infection could not be assessed. Fourth, this analysis did not exclude persons who tested positive for SARS-CoV-2 infection. While associations between influenza and COVID-19 vaccination can lead to underestimation of VE estimates [32], updated COVID-19 vaccination coverage was lower than 2023–2024 influenza vaccine coverage in California. Finally, this approach might be limited if patterns of clinical influenza testing, indications for testing, or use of rapid antigen tests versus molecular tests change over time.

During the 2023–2024 influenza season in California, data from influenza laboratory test results and vaccination information systems allowed health officials to assess VE. Influenza vaccination was protective against laboratory-confirmed influenza, though with heterogeneity across population subgroups and influenza type. Interim and final season estimates were consistent. Public health information systems can provide timely assessment of influenza vaccine effectiveness during the influenza season.