Understanding COVKID Measures

and Data Sources

  1. COVID-19 Identified Cases are the cumulative number of children and teens who have tested positive for novel coronavirus. These data are collected from state health department websites every 3rd day. Case data are reported for 49 states, the District of Columbia, and New York City. Currently New York State does not report COVID-19 case counts by age.

  2. COVID-19 Reported Deaths are the cumulative number of child deaths reported by state health departments. Some states do not report COVID-19 deaths by age. We also include deaths reported by creditable news agencies.

  3. COVID-19 Intensive Care Hospitalizations are the cumulative number of children ever admitted to a pediatric intensive care unit (PICU) for COVID-19. These data are collected from a voluntary PICU hospital registry (myvps.org). Not every PICU in the nation participates in the registry, so these counts are an underestimate.

  4. Estimated Number Ever Infected with Novel Coronavirus. These estimates are calculated using the PICU hospitalization data, based on a methodology developed by the COVKID Project Team and published in a peer-reviewed scientific article on April 10, 2020.

  5. Racial and Ethnic Disparities in Identified Cases. Cumulative counts of identified cases are reported by race and Hispanic ethnicity for the few states that make these data publicly available.


Technical Notes Log


September 22: We identified 16 new PICU cases today, and they have been added to the cumulative totals, although we do not know the exact date of hospitalization for these cases. There were 11 additional PICU patients reported in the state of Minnesota, by the state health department, above and beyond the cases reported in the MYVPS PICU registry. These cases are aged 0-14 years. We also added 5 additional cases from Massachusetts. These children are currently hospitalized in PICU but they have not been reported to the MYVPS PICU registry.


August 13: We have just retrospectively revised COVID-19 case counts for the state of New York with new, imputed data. Previously these data were missing, because the NYS Department of Health does not report cases by age. NYC does, but NYC accounts for only about half of all cases among kids in the state. Our imputation method consists for calculating the date-specific percent of all NYC cases which were aged 0-17, and then applying that percentage to the total cases for NYS to obtain case counts for 0-17 years for the state as a whole. This method assumes that the age distribution of cases in all of NYS is the same as that reported for NYC. The addition of the imputed case counts for NYS impacts the epi curve charts for the Mideast region and for the whole USA.

Today the cumulative total cases in Alabama decreased by several hundred, because Alabama began reporting cases for aged 0-17. Prior to today, they were reporting for ages 0-4 and ages 5-24, so we used an imputation method to derive estimated case counts for ages 0-17. Our estimates were higher than the actual case counts now reported on the Alabama COVID dashboard. Retrospective data were not provided, so we cannot revise our counts backward in time. This has slightly impacted the shape of the epi curve for the Southeast region.

July 24: We draw our counts of children admitted to PICU from the COVID-19 dashboard administered by myvps.org. Last night, VPS updated their dashboard with current new patient counts AND with substantial upward revisions to the counts for many past days. Consequently, the regional PICU patient curves which we posted earlier this week are now incorrect. We have pulled all of these charts. The current national trend chart is correct, and will continue to update this graphic. The current regional cumulative counts are also correct. However, because the shape of the curve keeps changing every time VPS revises hospital data for past days and weeks, we cannot commit to keeping the regional charts updated and correct.