Abstract
This paper assesses the age specificity of the infection fatality rate (IFR) for COVID-19. Our benchmark meta-regression synthesizes the age-specific IFRs from six recent large-scale seroprevalence studies conducted in Belgium, Geneva, Indiana, New York, Spain, and Sweden. The estimated IFR is close to zero for children and younger adults but rises exponentially with age, reaching about 0.3 percent for ages 50-59, 1.3 percent for ages 60-69, 4.6 percent for ages 70-79, and 25 percent for ages 80 and above. We compare those predictions to the age-specific IFRs implied by recent seroprevalence estimates for nine other U.S. locations, three smale-scale studies, and three countries (Iceland, New Zealand, and Republic of Korea) that have engaged in comprehensive tracking and tracing of COVID-19 infections. We also review seroprevalence studies of 32 other locations whose design was not well-suited for estimating age-specific IFRs. Our findings indicate that COVID-19 is not just dangerous for the elderly and infirm but also for healthy middle-aged adults, for whom the fatality rate is more than 50 times greater than the risk of dying in an automobile accident. Consequently, the overall IFR for a given location is intrinsically linked to the age-specific pattern of infections. In a scenario where the U.S. infection rate reaches 20 percent, our analysis indicates that protecting vulnerable age groups could prevent more than 200,000 deaths.
Objective Determine age-specific infection fatality rates for COVID-19 to inform public health policies and communications that help protect vulnerable age groups.
Methods Studies of COVID-19 prevalence were collected by conducting an online search of published articles, preprints, and government reports identified by online searches. Studies were identified covering a total of 52 locations in advanced economies. Studies were screened using three specific criteria: (i) representative sample of the general population; (ii) effective pandemic containment by the time of the study; and (iii) reporting of age-specific prevalence estimates and confidence intervals. Age-specific IFRs were computed using reported fatalities four weeks after the midpoint date of each study, reflecting the typical pattern of lags in fatalities and reporting. Six studies were identified as benchmarks and used in meta-regression of the infection fatality rate (IFR) as a function of age, using the STATA metareg procedure. The meta-regression results were then compared with age-specific IFRs for 15 other locations --an “out-of-sample” exercise that statisticians commonly use in assessing the validity of forecasting models.
Results Our analysis finds a highly significant exponential relationship between age and IFR for COVID-19. The estimated age-specific IFRs are close to zero for children and younger adults but rise to about 0.3 percent for ages 50-59, 1.3 percent for ages 60-69, 4.6 percent for ages 70-79, and 25 percent for ages 80 and above. Nearly all of the age-specific IFRs included in our meta-analysis fall within the 95% prediction interval of the meta-regression.
Discussion Our results indicate that COVID-19 is hazardous not only for the elderly but for middle-aged adults, for whom the infection fatality rate is more than 50 times greater than the annualized risk of a fatal automobile accident. Moreover, the overall IFR for COVID-19 should not be viewed as an exogenously fixed parameter but as intrinsically linked to the age-specific pattern of infections. Consequently, individual and collective efforts that minimize infections in older adults could substantially decrease total deaths. In a scenario where the infection rate of the U.S. population reaches 20%, our analysis indicates that protecting vulnerable age groups could prevent over 200,000 deaths.
Competing Interest Statement
The authors have declared no competing interest.
Funding Statement
No external funding was received for this research.
Author Declarations
I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained.
Yes
The details of the IRB/oversight body that provided approval or exemption for the research described are given below:
This study is a meta-analysis using information from publicly available studies (published articles, Medrxiv preprints, and government reports).
All necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived.
Yes
I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance).
Yes
I have followed all appropriate research reporting guidelines and uploaded the relevant EQUATOR Network research reporting checklist(s) and other pertinent material as supplementary files, if applicable.
Yes
Footnotes
Declaration The authors have no financial interests nor any other conflicts of interest related to this study. No funding was received for conducting this study. The views expressed here are solely those of the authors and do not represent the views of any other person or institution.
14 Brown and Walensky (2020), p.82.
17 Valenti et al. (2020), p.12.
52 See Molenberghs et al. (2020), Table 6. This study used the seroprevalence findings of Herzog et al. (2020).
53 See Perez-Saez et al. (2020), Table S2. This study used the seroprevalence findings of Stringhini et al. (2020).
54 See Menachemi et al. (2020), Table 1, total population prevalence based on seroprevalence and active infections. Population data by single year of age as of July 1, 2019 was obtained from U.S. Vital Statistics System (2020). Cumulative fatalities by age as of May 25 were obtained from Indiana State Department of Health (2020).
55 See Rosenberg et al. (2020). Population data by single year of age as of July 1, 2019 was obtained from U.S. Vital Statistics System (2020). Some seroprevalence age brackets were adjusted (+/- 5 years) to match the age structure of the New York Department of Health (2020) COVID-19 fatality report; see the technical appendix for further detail.
56 Age-specific IFRs were constructed using the seroprevalence findings of Pollán et al. (2020), Table S7 (both tests positive) and excess mortality data for Week 25 reported by Spain National Institute of Statistics (2020).
57 See Sweden Public Health Authority (2020a,b,c,d,e) for information about the seroprevalence program design, antibody test standards, results for weeks 18 to 21, and COVID-19 fatalities as of week 24, respectively.
58 See Connecticut Department of Health & Human Services (2020).
60 The Minneapolis-St. Paul-St. Cloud combined statistical area covers 20 counties in Minnesota; this CSA also covers two counties in Wisconsin that were not included in the seroprevalence study or the IFR tabulations. Cumulative fatalities in the 20-county MSA accounted for 89% of Minnesota’s cumulative COVID-19 fatalities as of June 4 and hence are assumed to have the same relative age distribution as the statewide fatality count.
61 See Utah Department of Health (2020).
62 The Philadelphia metropolitan statistical area covers 5 counties in Pennsylvania; this MSA also includes counties in Delaware, Maryland, and New Jersey that were not included in the seroprevalance study or the IFR tabulations. Cumulative fatalities in the 5-county MSA accounted for 61% of Pennsylvania’s cumulative COVID-19 fatalities as of May 23 and hence are assumed to have the same relative age distribution as the statewide fatality count.
63 The San Francisco Bay Area covers nine counties that are members of the Association of Bay Area Governments. Cumulative fatalities in the 9-county San Francisco Bay Area are not published by age groups and hence are assumed to have the same relative age distribution as the statewide fatalities as of May 25.
64 South Florida includes four counties that span Miami, Ft. Lauderdale, and Palm Beach. Cumulative fatalities by age for each of those four counties are tabulated by Florida Department of Health (2020).
65 See Missouri Department of Health & Senior Services (2020).
66 See Western Washington spans Seattle, Tacoma, and nearby counties. Cumulative fatalities are reported by Washington Department of Health (2020).
67 See Iceland Directorate of Health (2020) for finalized data thru June 14, when Iceland had 1,796 recovered cases, 10 fatalities, and 4 individuals in isolation (none hospitalized).
70 See Pagani et al. (2020).
71 See Italy National Institute of Statistics (2020a,b) for Castiglione d’Adda population by age and excess mortality by age in 2020 compared to the average mortality during the same calendar dates in 2015 to 2019, respectively.
72 See Streeck et al. (2020).
73 See Kreis Heinsberg District Administration (2020) and Stat Germania (2020) for Gangelt COVID-19 fatalities and population by age, respectively.
74 See Mizumoto et al. (2020), Russell et al. (2020), Leffler and Hogan (2020), and Salje et al. (2020a,b).
75 See Japan National Institute for Infectious Diseases (2020), Mizumoto et al. (2020), and Salje et al. (2020a,b).
76 See Reifer et al. (2020).
77 See Doi et al. (2020).
78 See Takita et al. (2020a,b).
79 See Emmenegger et al. (2020).
87 See Snoeck et al. (2020).
88 See Bryan et al. (2020).
90 See Fraehling et al. (2020).
91 See Fontanet et al. (2020).
92 See Armann et al. (2020).
98 See Skowronski et al. (2020).
99 See Czech Ministry of Health (2020).
100 See Verity et al. (2020).
101 See Nishiura et al. (2020).
102 See Jersey (U.K.) Health & Community Services (2020a,b).
103 See Feehan et al. (2020).
104 See Stadlbauer et al. (2020).
105 See Weis et al. (2020).
106 See Chamie et al. (2020).
107 See San Miguel County Department of Health & Environment (2020).
108 See Vodičar et al. (2020a, b).
110 See Lavezzo et al. (2020).
Data Availability
This study is a meta-analysis using information from published articles, preprints, and government reports; all sources are listed in the bibliography with active URLs. The data and Stata code used in performing the meta-regression analysis are provided as Supplementary Materials.
