Frequently Asked Questions about Estimated Flu Burden

Illnesses

What is the main lesson learned from annual estimates of annual influenza (flu) illness?

Seasonal flu is associated with large numbers of illnesses, which can impact school attendance, worker absenteeism, and daily productivity. The results of CDC’s flu burden estimates demonstrate the substantial health impact of flu and underscore the importance of yearly flu vaccination for everyone   6 months and older.

How many people get sick with flu every year?

CDC conducts surveillance for people who see their health care provider for flu-like illness through the Outpatient Influenza-like Illness Surveillance Network (ILINet); a network of thousands of health care providers who report the proportion of patients seeking care for flu-like illness weekly to CDC. This system allows CDC to track levels of medically attended flu-like illness over the course of the flu season.

CDC does not know exactly how many people get sick with seasonal flu each year. There are several reasons for this, including that ILINet does not include every health care provider in the United States and that it monitors flu-like illness, not laboratory-confirmed flu cases. Also, seasonal flu illness is not a reportable disease, and not everyone who gets sick with flu seeks medical care or gets tested for flu.

To estimate the number of flu illnesses that occur in the United States each year, CDC uses mathematical modeling in combination with data from traditional flu surveillance systems. CDC estimates that from 2010 to 2020, flu has resulted in between 9 million and 41 million illnesses annually in the United States.

More information on these estimates is available on CDC’s Disease Burden of Influenza page. More information on CDC surveillance systems is also available at CDC’s Overview of Influenza Surveillance in the United States.

Hospitalization

What are the main lessons learned from the updated estimates of flu-related hospitalization?

Seasonal flu is associated with large numbers of hospitalizations. These estimates of flu-related hospitalizations highlight flu’s potential severity, and that being sick with flu can also make some health conditions worse (such as lung disease) or lead to other complications that require hospitalization. Flu vaccination is the first and best way to prevent flu and its serious complications. Everyone 6 months and older should get vaccinated against flu yearly, particularly people at increased risk for serious complications including young children, adults 65 years and older, and people with certain chronic medical conditions.

CDC’s current estimates of hospitalization only go back to 2010. What were the previous estimates of the number of people in the United States hospitalized as a result of seasonal influenza?

Before CDC had FluSurv-NET to gather data on laboratory-confirmed flu hospitalizations, CDC periodically made estimates of flu hospitalizations using a statistical model of data on hospitalizations with flu-like illnesses.  A study conducted by CDC and published in the Journal of American Medical Association (JAMA) in September 2004 provided information on the number of people in the United States that were hospitalized from flu-related complications each year. The study was based on records from 1979 to 2001 from about 500 hospitals across the United States. The study concluded that, on average, more than 200,000 people in the United States are hospitalized each year for respiratory and heart conditions illnesses associated with influenza virus infections. Prior to that, in a paper published in 2000, looking at records from 1970 – 1995, CDC estimated that an average of 114,000 people were hospitalized as a result of flu-related infections each year (7).

The current methodology used to estimate flu-related hospitalizations is straightforward, uses directly observed hospitalization rates from FluSurv-NET surveillance data, and can be updated annually and, starting in 2019, even during a flu season Previously it took longer to arrive at estimates because of delays in the availability of large administrative databases that were used in statistical models.

It’s important to note that the results of both estimation methods have similarly shown that hundreds of thousands of people are hospitalized with flu each year.

Why are the current estimates of flu-related hospitalizations different from previous estimates?

The estimated number of flu hospitalizations has always varied from season to season. In the past, CDC had referred to an average number of hospitalizations that was estimated based on statistical models using hospital discharge records from 1979 to 2001 from selected hospitals across the United States. In addition to year-to-year variation, changes in health care practices and the increasing age of the US population, with a much greater number of adults over 65, may result in different numbers of flu hospitalizations now than in previous decades. Because of these differences, it is challenging to directly compare current estimates of flu disease burden to those from many years ago.

Why does CDC use a range to describe illnesses, hospitalizations and deaths?

CDC believes that using a range is a better way to represent the variability of flu. The more recent estimates use surveillance data on patients hospitalized with laboratory-confirmed flu in geographically distributed areas in the United States.

Why do flu burden estimates that use the same methodology change?

CDC releases preliminary flu burden estimates because some of the data used in the burden estimation model lag by 2 years. Providing preliminary estimates allows for the release of timely information about US flu burden, but also means that these preliminary estimates may change as more data become available. Flu burden estimates are considered preliminary and subject to change until all data are complete. Once all data are complete for a particular year, that flu season’s estimates are considered final.

How many people are hospitalized from flu every year?

CDC conducts surveillance for flu-related hospitalizations through the Influenza Hospitalization Surveillance Network (FluSurv-NET), a collaboration between CDC, the Emerging Infections Program, and additional Influenza Hospitalization Surveillance Project (IHSP) states in 13 geographically distributed areas in the United States. The network includes hospitals that serve roughly 9 percent of the U.S. population. The data collected through FluSurv-NET allows CDC to calculate an overall hospitalization rate, as well as by age group, but this system does not capture every flu hospitalization that occurs annually in the United States. To estimate the actual number of flu hospitalizations, reported FluSurv-NET hospitalization rates are adjusted to correct for under-detection and under-reporting.,. Using this estimation model, CDC estimates that, from 2010 to 2020, flu has resulted in between 140,000 and 710,000 hospitalizations each year.

More information on these estimates is available on CDC’s Disease Burden of Influenza page.

More information on CDC surveillance systems is also available at CDC’s Overview of Influenza Surveillance in the United States.

Deaths

What are seasonal flu-related deaths?

Seasonal flu-related deaths are deaths that occur in people for whom flu was likely a contributor to the cause of death, but not necessarily the primary cause of death.

Does CDC know the exact number of people who die from seasonal flu each year?

CDC does not know exactly how many people die from seasonal flu each year. There are several reasons for this. First, states are not required to report individual flu illnesses or deaths among people older than 18 years old to CDC. Second, flu is infrequently listed on death certificates of people who die from flu-related complications. Third, many flu-related deaths occur one or two weeks after a person’s initial infection, either because the person may develop a secondary bacterial co-infection (such as bacterial pneumonia) or because flu can aggravate an existing chronic illness (such as congestive heart failure or chronic obstructive pulmonary disease). Also, most people who die from flu-related complications are not tested for flu, or they seek medical care later in their illness when flu can no longer be detected from respiratory samples (sensitive flu tests are only likely to detect flu if performed within a week after onset of illness). In addition, some commonly used tests to diagnose flu in clinical settings are not highly sensitive and can provide false negative results (i.e., they misdiagnose flu illness as not being flu.) For these reasons, many flu-related deaths may not be recorded on death certificates. These are some of the reasons that CDC and other public health agencies (in the United States and globally) use statistical and mathematical models to estimate the annual number of flu-related deaths.

Flu deaths in children are slightly different though because these are nationally notifiable, which means that individual flu deaths must be reported to CDC. States report flu-related child deaths in the United States through the Influenza Associated Pediatric Mortality Surveillance System. However, even deaths in children may be under-reported, for many of the same reasons listed above.

Why are the current estimates of flu-related deaths different from previous estimates?

Previously, CDC periodically made estimates of flu deaths using a statistical model of data on deaths with respiratory and circulatory causes.  An August 27, 2010, MMWR report entitled “Updated Estimates of Mortality Associated with Seasonal Influenza through the 1976-2007 Influenza Season” (MMWR 2010; 59(33): 1057-1062.), provided estimates of the range of flu-associated deaths that occurred in the United States during the three decades from the 1976-1977 season to the 2006-2007 flu season. In that era, CDC estimated that flu-associated deaths ranged from a low of about 3,000 to a high of about 49,000 deaths.

The previous range used to describe influenza-related deaths, from 3,000 to 49,000, was based on data from 30 influenza seasons from 1976 through 2007 used in a statistical model (1). The range described in the tables above, 12,000 to 79,000, is based on data from the 2010-2011 through 2017-2018 influenza seasons using a different mathematical model.

While there are differences in methods used to estimate the deaths in the two time periods, other factors may also contribute to why some seasons have different numbers of flu deaths than seen in the past, including changes in the way that death certificates are filed, changes in the age structure of the population, or changes in the prevalence of chronic medical conditions that put people at higher risk of flu complications.

Why did the death burden estimation method change?

CDC has made periodic estimates of deaths associated with flu for five decades, initially relying on direct counts of deaths with flu listed as a cause of death and then focusing on deaths with pneumonia or flu listed as a cause. Recognizing that these direct counts of deaths were underestimating the true burden of flu, CDC turned to using statistical models.

Initially, these methods relied on designating specific weeks during the winter and spring when influenza viruses circulated as ‘at risk’ weeks. Increases above periodic, regular variations in deaths during these weeks were attributed to the circulation of influenza viruses, and often termed excess deaths. These statistical methods were used to estimate excess deaths that occurred from 1976-2007 (8) and more recently from 2005-2016 (9). Current estimates of flu-related deaths are now based on the mathematical model described above.

Each approach, the statistical and the mathematical model, has its own strengths as well as limitations. The statistical methods require final US mortality records, which are not routinely available until 2-3 years after the date of an individual death. Thus, national vital statistics data cannot be used currently to estimate flu-related deaths in an expedient manner.

Because of these reasons, CDC has chosen to use a mathematical model that is based on FluSurv-NET, a strong surveillance platform of laboratory-confirmed flu hospitalizations. This mathematical model is straightforward and can be used to compare one season to another in a way that is simple, yet comprehensive.

Why doesn’t CDC base its seasonal flu mortality estimates only on death certificates that specifically list influenza?

Seasonal flu may lead to death from other causes, such as pneumonia, congestive heart failure, or chronic obstructive pulmonary disease. It has been recognized for many years that flu is underreported on death certificates and patients aren’t always tested for seasonal flu virus infection, particularly the elderly who are at increased risk of seasonal flu complications and death. Some deaths – particularly among the elderly – are associated with secondary complications of flu (including bacterial pneumonias). Influenza virus infection may not be identified in many instances because influenza virus is only detectable for a short period of time and/or many people don’t seek medical care until after the first few days of acute illness. For these and other reasons, statistical modeling strategies have been used to estimate seasonal flu-related deaths for many decades. Only counting deaths where flu was included on a death certificate would be a gross underestimation of seasonal flu’s true impact.

How many flu-associated deaths occur in people who were not vaccinated?

Flu-related deaths in adults are not a nationally notifiable condition, and so states are not required to report flu-related deaths in adults to CDC. In contrast, flu-related deaths in children are a nationally notifiable condition, and so jurisdictions (inclusive of state, city or local public health departments) do provide data to CDC on flu-related deaths in children. These data generally include demographic information, flu laboratory test results, clinical information, and information on the child’s vaccination history, when it is available. From 2010-2020, approximately 80% of flu-related deaths in children have occurred in children who were not vaccinated.

The latest surveillance data on flu deaths in children is available.

Why are estimated pediatric deaths in this report different from the number of pediatric deaths reported through the Influenza-Associated Pediatric Mortality Surveillance System?

Deaths associated with laboratory-confirmed flu in children less than 18 years old became nationally notifiable in 2004 and are reported to CDC through the Influenza-Associated Pediatric Mortality Surveillance System. The number of reported deaths is published each week in FluView. However, the number of reported deaths is likely an underestimate of the total number of flu-related pediatric deaths because not all children may be tested for flu or children may be tested later in their illness when seasonal flu can no longer be detected from respiratory samples.

CDC estimates the numbers of flu-related deaths using mathematical models to account for likely under-reporting. The estimates of deaths associated with flu that we report are from one such model. Previously published reports have found that the estimated numbers of flu-related deaths in children from statistical models may be two to three times higher than the number of reported deaths. (10)

How many adults die from flu each year?

Flu deaths in adults are not nationally notifiable. In order to monitor flu-related deaths in all age groups, CDC tracks pneumonia and influenza (P&I)–attributed deaths through the National Center for Health Statistics (NCHS) Mortality Reporting System. This system tracks the proportion of death certificates processed that list pneumonia or flu as the underlying or contributing cause of death. This system provides an overall indication of whether flu-associated deaths are elevated but does not provide an exact number of how many people died from flu. As it does for the numbers of flu cases, doctor’s visits and hospitalizations, CDC also estimates deaths in the United States using mathematical modeling. The model used to estimate flu-related deaths uses a ratio of deaths-to-hospitalizations in order to estimate the total flu-related deaths during a season.

CDC estimates that, from 2010- to 2020, flu-related deaths in the United States ranged from a low of 12,000 (during 2011-2012) to a high of 52,000 (during 2017-2018).

For more information visit How CDC Estimates Burden and Overview of Influenza Surveillance in the United States, “Mortality Surveillance.”

How many children die from flu each year?

Flu-related deaths in children (people younger than 18) became nationally reportable in 2004. Since that time the number of pediatric flu deaths reported to CDC each year has ranged from 37 (2011-2012 season) to 199 deaths (2019-2020 season). It’s important to note that the actual number of flu deaths in children is thought to be higher than what is reported by states to CDC because not all flu deaths in children are detected/reported. CDC also estimates the numbers of flu-related deaths using statistical models. Estimates of deaths in children since 2010 have ranged from 37 (2011-2012) to about 1,200 (2012-2013).

For more information on these estimates see CDC’s Disease Burden of Influenza page.

Why is it difficult to know exactly how many people die from flu?

There are several factors that make it difficult to determine accurate numbers of deaths caused by flu regardless of reporting. Some of the challenges in counting flu-related deaths include the following:

  • the sheer volume of deaths to be counted.
  • the lack of testing (not everyone that dies with an influenza-like illness is tested for flu);
  • and the different coding of deaths (flu-related deaths often are a result of complications secondary to underlying medical problems, and this may be difficult to sort out).

For more information: Estimating Seasonal Influenza-Associated Deaths in the United States.

2020-2021 Burden

Why aren’t there flu burden estimates for the 2020-21 flu season?

Every year, CDC usually generates estimates of the number of illnesses, medical visits, hospitalizations and deaths that happen during a flu season—these estimates are used to collectively describe the annual burden of flu.  To produce these estimates, CDC uses a mathematical model that is based in part on the number of people that are hospitalized with flu in our hospitalization surveillance network, FluSurv-NET. During the 2020-2021 season, however; the number of people hospitalized with flu was too low to generate stable burden estimates as is done for a typical flu season.

We do know that flu activity last season was the lowest it has been since current reporting began in 1997. We also know that flu illnesses, hospitalizations and deaths were very low last season, probably well below the estimates for the 2011/2012 season, which was the mildest flu season in the decade between 2010 and 2020.

Are there other metrics that can be used to compare the 2020-2021 flu season with past seasons?

One measure to help understand how low flu activity was during 2020-2021 is the cumulative number (or percent of specimens testing positive) of clinical lab-confirmed flu infections for Week 34 from the past two flu seasons.

Season (up to Week 34)
 Season (up to Week 34)  Specimens tested by clinical labs  Total Positive Specimens
 2020-21 1,480,295 2,265 (0.15%)
 2019-20 1,491,430 250,396 (16.8%)

 

Another measure would be the FluSurv-NET cumulative hospitalization rates from 2020-21 and 2019-20.

  • The overall cumulative hospitalization rate for the 2020-21 flu season was 0.8 per 100,000.
  • For the 2019-20 flu season, the overall cumulative end-of-season hospitalization rate was 66.2 per 100,000.
  • It’s important to note that FluSurv-NET covers approximately 9% of the U.S. population and rates may not necessarily be representative of the entire country.

Another measure would be the total number of hospitalizations reported through FluSurv-NET.

  • During 2020-21, 230 lab-confirmed hospitalizations were reported through FluSurv-NET.
  • During 2019-20, 19,302 lab-confirmed hospitalization were reported through FluSurv-NET.
  • It’s important to note that this is not a comprehensive count of all flu hospitalizations in the United States. This is just a small subset reported to CDC for surveillance purposes.
  • FluSurv-NET covers approximately 9% of the U.S. population and rates may not necessarily be representative of the entire country
References
  1. Reed C, Chaves SS, Daily Kirley P, Emerson R, Aragon D, Hancock EB, et al. Estimating influenza disease burden from population-based surveillance data in the United States. PLoS One. 2015;10(3):e0118369.
  2. Centers for Disease Control and Prevention. Estimated influenza illnesses and hospitalizations averted by influenza vaccination – United States, 2012-13 influenza season. MMWR Morb Mortal Wkly Rep. 2013 Dec 13;62(49):997-1000.
  3. 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.
  4. Centers for Disease Control and Prevention. Estimated Influenza Illnesses and Hospitalizations Averted by Vaccination — United States, 2014–15 Influenza Season. 2015 December 10, 2015 [cited 2016 October 27];
  5. Centers for Disease Control and Prevention. Estimated Influenza Illnesses and Hospitalizations Averted by Vaccination — United States, 2014–15 Influenza Season. 2015 December 10, 2015 [cited 2016 October 27]
  6. 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.
  7. Simonsen L, Fukuda K, Schonberger LB, Cox NJ. The impact of influenza epidemics on hospitalizations. J Infect Dis. 2000;181(3):831-7.
  8. Centers for Disease Control and Prevention. Estimates of deaths associated with seasonal influenza — United States, 1976-2007. MMWR Morb Mortal Wkly Rep. 2010 Aug 27;59(33):1057-62.
  9. Foppa IM, Cheng PY, Reynolds SB, Shay DK, Carias C, Bresee JS, et al. Deaths averted by influenza vaccination in the U.S. during the seasons 2005/06 through 2013/14. Vaccine. 2015 Jun 12;33(26):3003
  10. Wong KK, Cheng P, Foppa I, Jain S, Fry AM, Finelli L. Estimated paediatric mortality associated with influenza virus infections, United States, 2003-2010. Epidemiol Infect. 2014 May 15:1-8