The actual number of COVID-19 cases in our nation has been vastly under reported, according to new data.

An analysis of blood samples from 16,025 people who underwent routine testing for the virus at 10 U.S. sites from March 23 to May 12, suggests that at least 10 times more people were infected than previously reported.

According to the Center for Infectious Disease Research and Policy at the University of Minnesota, the study that was published in JAMA Internal Medicine was led by researchers from the Centers for Disease Control and Prevention (CDC) and is the first multistate analysis of its kind. The researchers said that actual infection occurred at least one or two weeks before the samples were collected.

The results differed from state to state. In some states the number of people infected was 24 times the reported cases, the CDC team said, noting that across the board, the average was 10 times greater than previously thought.

For example, the CDC data estimates that in New York 642,000 people were infected by the virus by April 1 but only 53,803 actual cases were officially reported by that date.

According to CNN, Dr. Robert Redfield, director of the CDC, said that testing likely had missed 90% of cases. President Donald Trump said the pandemic will “probably, unfortunately, get worse before it gets better,” after the CDC released its analysis.

On Tuesday during a White House briefing Trump said, “We are asking everybody, when you are not able to socially distance, to wear a mask. Whether you like the mask or not, they have an impact.”