Use case | What we have | Estimation based on cases with known outcomes | Mathematical explanation for bias | Estimation based on expected number of known death outcomes | Simulated comparison of above methods | Deaths reported but not recoveries | Only total deaths reported | Only total cases and deaths reported

A brief primer on distributions in R | Using delay distribution densities in cfr | Preparing delay distribution density for cfr | Passing delay distribution density to cfr functions | Using other distribution representations | Using distributional | Using distcrete | Using continuous and discrete distributions | Links to epiparameter | References

Use case | What we have | Severity of the 1976 Ebola Outbreak | Onset-to-death delay distribution | Intermediate step: Estimating cases with known outcomes | Estimating the naive and corrected CFR | Severity estimation methods | Severity of the COVID-19 pandemic in the U.K. | Onset-to-death distribution for Covid-19 | Details: Adjusting for delays between two time series | References

Use case | What we have | What we assume | Case and death data | Obtaining data on reporting delays | Estimate disease severity | Estimate ascertainment ratio | Concept: How reporting delays bias CFR estimates | References

Use case | What we have | Potential reasons for changing disease severity | Changing severity of the Covid-19 pandemic in the U.K. | Preparing the raw data | Onset-to-death distribution for Covid-19 | Estimating the naive and corrected CFR | Severity of Covid-19 in multiple countries | Details: Adjusting for delays between two time series | References

Use case | What we have | Ascertainment for the Covid-19 pandemic in the U.K. | Estimating the proportion of cases that have been ascertained | Ascertainment in countries with large early Covid-19 pandemics | References

Scope of regulations | Data privacy and integrity | Internet access | Registration and third parties | Updates