How CDC Estimates the Burden of Flu Prevented by Vaccination

At a glance

CDC estimates the burden of influenza (flu) prevented by vaccination with a model that uses flu burden data, vaccination coverage data, and vaccine effectiveness data during a specific flu season.

person using tablet to see how CDC classifies flu severity

Overview

CDC estimates the total number of flu illnesses, medical visits, hospitalizations, and deaths prevented by vaccination using a model that incorporates season-specific data on burden of disease, vaccine coverage and vaccine effectiveness (VE) for five age groups. A more detailed description of the method can be found in an article by Tokars, et al1.

CDC uses the annual estimates of influenza (flu) vaccination coverage by month during each season and the final end-of-season vaccine effectiveness measurements, adjusting for how many people have already been infected by flu, to estimate how many people did not get a flu vaccine or infection and were therefore not protected from an influenza virus infection (or were “unprotected” from flu) during the season. The rate of each outcome (illnesses, medical visits, hospitalizations, or deaths) among unprotected people is calculated and then used to estimate the number of flu-related outcomes that would have been expected in the population if no one had been protected by flu vaccination. Finally, the outcomes prevented by flu vaccination are calculated as the difference between outcomes in the hypothetical fully-unvaccinated population and the observed vaccinated population.

Estimates of flu vaccination coverage by month were based on self-report or parental report of vaccination status using data from the National Immunization Survey for children aged 6 months-17 years and Behavioral Risk Factor Surveillance Survey data for adults 18 years and older.2

Vaccine effectiveness estimates for each season are derived from existing CDC flu vaccine effectiveness networks.3 These networks estimate the effectiveness of vaccination in preventing flu diagnosed illnesses by real-time reverse transcription polymerase chain reaction-positive testing among persons with acute respiratory illness of ≤7 days duration seen in hospitals, emergency departments, or outpatient clinics in the United States. Since vaccine effectiveness estimates are available by clinical setting (i.e., outpatient and inpatient), the outpatient VE estimates were used to calculate prevented burden for symptomatic and medically attended illnesses and inpatient VE estimates were used to calculate prevented burden for hospitalizations and deaths. In addition, because estimates by age group and clinical setting were available from multiple data sources, the estimates were combined into an overall estimate of VE for each age group and clinical setting.

Prevented burden calculations were stratified by month of the year to account for annual variations in the timing of disease and vaccination and then summed across the whole season.

Limitations

These estimates are subject to several limitations. First, flu vaccination coverage estimates were derived from reports by survey respondents, not vaccination records, and are subject to recall bias. These coverage estimates are based on telephone surveys with relatively low response rates; non-response bias may remain after weighting for the survey design. Estimates of the number of people vaccinated based on these survey data have often exceeded the actual number of flu vaccine doses distributed, indicating that coverage estimates used in this report may overestimate the numbers of flu illnesses and hospitalizations prevented by flu vaccination. The model of prevented flu illnesses calculates outcomes directly prevented among people who were vaccinated. If indirect protection from decreased exposure to infectious people in a partially flu-vaccinated population (i.e., herd immunity) also occurred, the model would underestimate the number of flu illnesses and hospitalizations prevented by flu vaccination. Finally, because the data and methods used to make these calculations are continually updated, current estimates are preliminary and will be updated when final data are available.

How flu burden is estimated

Rates of laboratory-confirmed flu-related hospitalizations by age group were obtained from FluSurv-NET, a collaboration between CDC, the Emerging Infections Program Network, and selected state and local health departments in 14 geographically distributed areas in the United States that conduct population-based flu surveillance. Reported hospitalization rates were adjusted using a multiplier to correct for underreporting, which is calculated from the percent of people hospitalized with respiratory illness who were tested for flu and the average sensitivity of flu tests used in the surveillance hospitals. The percentage of people hospitalized and tested for flu was measured from data collected each season since the 2010-2011 flu season. When adjusting the hospitalization rates, testing data from prior seasons is used until testing data from that specific season is available, usually 2 years after the season ends.

Adjusted rates were applied to the U.S. population by age group to calculate the numbers of flu-related hospitalizations. The numbers of flu illnesses were then estimated from hospitalizations based on previously measured ratios that reflect the estimated number of ill persons per hospitalization in each age group.4

The number of people seeking medical care for flu were then calculated using age group-specific data on the percentages of people with a respiratory illness who sought medical care, which was estimated from results of the 2010 Behavioral Risk Factor Surveillance Survey.5

All estimates were rounded to two significant digits.

  1. Tokars JI, Rolfes MA, Foppa IM, Reed C. An evaluation and update of methods for estimating the number of influenza cases averted by vaccination in the United States. Vaccine 2018; 36(48): 7331-7.
  2. Centers for Disease Control and Prevention. Influenza Vaccination Coverage.
  3. Centers for Disease Control and Prevention. CDC Seasonal Flu Vaccine Effectiveness Studies.
  4. Reed C, Kim IK, Singleton JA, Chaves SS, Flannery B, Finelli L, et al. Estimated influenza illnesses and hospitalizations averted by vaccination–United States, 2013-14 influenza season. MMWR Morb Mortal Wkly Rep. 2014 Dec 12;63(49):1151-4.
  5. Biggerstaff M, Jhung M, Kamimoto L, Balluz L, Finelli L. Self-reported influenza-like illness and receipt of influenza antiviral drugs during the 2009 pandemic, United States, 2009-2010. Am J Public Health. 2012 Oct;102(10):e21-6.