The federal requirement that nursing homes report COVID-19 cases and deaths did not kick in until May 24, 2020 — three months after the first known nursing home outbreak at Life Care Center of Kirkland, Washington.

Facilities could voluntarily provide such data to the CDC National Healthcare Safety Network, but just how many did so remains unclear. But factors ranging from data availability to reputation concerns may have motivated many, like Life Care Center of Kirkland, not to do so.

A new study in the journal JAMA Network Open attempts to fill in that blank.

By comparing federal tallies to nursing home data in 20 states that required reporting from the pandemic's onset, the authors found a shortfall of 44% of cases and 40% of deaths.

Extrapolating to include all U.S. states, that translates to more than 68,000 cases and 16,000 deaths going unreported.

Understanding those gaps and where they occurred is essential for tracking the pandemic's early impact. For example, unreported cases and deaths accounted for a larger portion of year-end totals in the Northeast than in regions where coronavirus impacts were felt more strongly later, such as the South and West.