Due to the current political climate, the author remains anonymous, but you can get in touch via:



Reported deaths are taken from CDC's weekly dataset. The baseline is calculated as average of 2015-2019, adjusted population changes and smoothed using a Loess & Gauss window function. The Normal range is +/- two standard deviations and the substantial increase threshold is four standard deviations above the baseline.

Up to eight weeks might be additionally corrected for reporting delay based on last eight weeks of reporting. The blue dashed line shows the prediction and the light yellow area shows the 95% prediction interval. Any counts fewer than 10 are suppressed by CDC, therefore an imputation algorith is used to calculate an estimation for those deaths.

Non COVID-19 Deaths

Non COVID-19 Deaths are calculated by substracing "COVID-19 (U071, Underlying Cause of Death)" from "All Cause" deaths.


Z-scores are used to standardize series and enable comparison mortality pattern between different populations or between different time periods. The standard deviation is the unit of measurement of the z-score. It allows comparison of observations from different normal distributions. Z-Scores are computed by first calculating the baseline. For each state and week, the baseline is the age- & population-adjusted 5y average of mortality for the years 2015-2019. In the same manner, the standard deviation is calculated. Z-score is then calculated by: z-score = (currentMortality - baseline) / stddv.

Excess Deaths

The baseline is calculated as average of 2015-2019, adjusted for age and population changes and smoothed using a Gauss window function. The excess deaths are calculated as observed deaths minus baseline. For the most recent weeks, excess deaths are calculated using the midpoint of the prediction.


Mortality is a basic indicator of health. Therefore, understanding its epidemiology is fundamental for effective public health planning and action. Vital statistics are accessible for all US states, but in most instances, these data are not readily available during crises or for imminent health threats. With the emergence of new diseases or threats of epidemics (e.g. pandemic influenza, SARS), decision makers will need such data to estimate the severity of the problem and inform any initiatives to be put in place as part of an effective public health response. As many public health threats are not restricted by borders, international unified approaches are critical to detect and estimate the magnitude of excess deaths, as pooling of data increases power to detect changes quickly. Mortality monitoring should be ongoing to detect when and where excess mortality occurs. Mortality monitoring becomes pivotal during influenza or other pandemics for several reasons. In a severe pandemic, mortality monitoring can be a robust way to monitor the pandemics progression and its public health impact when other systems are failing, due to an overburdened health care sector. Decision makers will require data on the pandemics impact and on deaths by age and geographical area in various stages of the pandemic.

European Data: EuroMOMO Worldwide Data: No medical advice & guarantee for correctness. Please always refer to: CDC & WHO