About

COVID-19 infection fatality rate (IFR) varies by orders of magnitude across demographics. A 10-year-old faces an IFR near 0.002%, while an 85-year-old confronts roughly 15%. Ignoring comorbidity adjustments produces dangerously misleading estimates. This calculator applies the Levin et al. (2020) exponential meta-regression model for age-stratified IFR, then multiplicatively adjusts using hazard ratios from the OpenSAFELY study (Williamson et al., Nature 2020, n = 17.3 million). Vaccination effectiveness reductions are derived from CDC MMWR surveillance reports.

Limitations: this tool approximates individual risk under population-level statistical models. It does not account for viral variant-specific virulence shifts, individual immune response heterogeneity, or access-to-care disparities. The output is a statistical estimate, not a clinical diagnosis. For patients with multiple severe comorbidities, multiplicative hazard ratios may overestimate compound risk due to overlapping pathophysiology. Consult a physician for personalized medical assessment.

Formulas

The base infection fatality rate is computed using the Levin et al. (2020) exponential meta-regression model:

IFR_base = 10^{(−3.27 + 0.0524 × age)}

A sex adjustment factor S is applied. Males carry approximately 1.4× higher risk than females (Peckham et al., Nature Communications 2020):

S =

{

1.4 if male1.0 if female

Comorbidity hazard ratios are applied multiplicatively:

HR_total = n∏i=1 HR_i

Vaccination effectiveness (VE) reduces the final risk:

IFR_final = IFR_base × S × HR_total × (1 − VE)

Where VE values are: 0 (unvaccinated), 0.60 (partial / 1 dose), 0.90 (fully vaccinated 2 doses), 0.95 (boosted 3+ doses). The result is clamped to [0, 100]%.

Where: age = chronological age in years, HR_i = hazard ratio for the i-th comorbidity, S = sex adjustment factor, VE = vaccine effectiveness against death.

Reference Data

Age Group	Base IFR (Unvaccinated)	Approx. Hospitalization Rate	Approx. ICU Admission Rate
0 - 9	0.002%	0.1%	< 0.01%
10 - 19	0.003%	0.3%	0.02%
20 - 29	0.01%	1.0%	0.1%
30 - 39	0.03%	2.0%	0.3%
40 - 49	0.08%	3.5%	0.7%
50 - 59	0.23%	5.5%	1.5%
60 - 69	0.75%	10%	3.5%
70 - 79	2.5%	18%	7%
80 - 89	8.3%	28%	12%
90+	16.0%	35%	15%

Comorbidity	Hazard Ratio (HR)	Source
Diabetes (Type 2)	1.95	OpenSAFELY
Obesity (BMI ≥ 40)	1.92	OpenSAFELY
Chronic Kidney Disease	2.52	OpenSAFELY
Chronic Heart Disease	1.57	OpenSAFELY
Chronic Liver Disease	1.75	OpenSAFELY
Chronic Respiratory Disease (excl. Asthma)	1.63	OpenSAFELY
Asthma (severe, OCS use)	1.25	OpenSAFELY
Cancer (non-haematological, recent)	1.72	OpenSAFELY
Haematological Cancer (recent)	2.80	OpenSAFELY
Stroke / Dementia	2.16	OpenSAFELY
Organ Transplant (Immunosuppressed)	3.53	OpenSAFELY
Autoimmune / Immunosuppressive Therapy	1.70	OpenSAFELY
Hypertension	1.09	Meta-analysis (Lippi 2020)
Smoking (current)	1.25	WHO meta-analysis

Frequently Asked Questions

The Levin et al. exponential model produces IFR values below 0.01% for ages under 20. These low values are consistent with population-level surveillance data from the CDC and ECDC. The model output is floored at 0.0001% (1 in 1,000,000) to avoid displaying misleadingly precise infinitesimal numbers. For pediatric populations, COVID-19 mortality remains exceptionally rare, but Multisystem Inflammatory Syndrome in Children (MIS-C) represents a separate risk not captured by IFR.

The Cox proportional hazards framework underlying the OpenSAFELY study models risk factors as multiplicative on the hazard scale. A patient with diabetes (HR = 1.95) and chronic kidney disease (HR = 2.52) carries a combined HR of 1.95 × 2.52 = 4.91. This multiplicative assumption holds when comorbidities act on independent biological pathways. However, for highly correlated conditions (e.g., obesity and diabetes), the true combined HR may be lower than the product due to shared pathophysiology. The calculator notes this limitation.

The VE values used (60% for partial, 90% for full, 95% for boosted) represent averages from CDC MMWR studies conducted during Delta-era surveillance. VE against death from Omicron subvariants may differ. VE also wanes over time: effectiveness at 6+ months post-vaccination is lower than at 2 weeks. This calculator uses peak VE estimates. For time-since-vaccination adjustments, consult your local public health authority's updated booster guidance.

No. The base IFR model derives from data collected primarily during wild-type and Alpha-wave periods (2020-2021). Omicron subvariants show lower intrinsic virulence (estimated 50-70% reduction in IFR compared to Delta). The calculator provides a conservative (higher) estimate. To approximate Omicron-era risk, the output could be mentally discounted by roughly 50%, but this tool does not automatically apply variant-specific adjustments due to the rapidly evolving evidence base.

Yes. Selecting many comorbidities simultaneously can push the computed IFR above biologically plausible levels. For example, a 90-year-old male with 5 severe comorbidities might compute an IFR exceeding 100% before clamping. The calculator caps output at 100%. In practice, patients with 4+ severe conditions face IFR in the range of 20-40%, as competing risks and correlated pathways prevent simple multiplication from reflecting true compound risk. The result should be interpreted as a relative risk indicator, not an absolute probability.

The OpenSAFELY study reported hazard ratios in categorical BMI bands: BMI 30-35 (HR 1.27), BMI 35-40 (HR 1.56), and BMI ≥ 40 (HR 1.92). This calculator uses the BMI ≥ 40 (Class III obesity) category as a single checkbox for simplicity. Users with BMI 30-39 face lower but still elevated risk. A continuous BMI input would require interpolation between these categorical estimates, introducing additional modeling assumptions.