Jumbled Numbers

by Carol Bova

The Virginia Mercury published an article by Kate Masters April 6th about Virginia’s $27 million order for personal protective equipment (PPE). As of April 20th, the Governor’s Office had not responded to the Mercury’s Freedom of Information Act request for a copy of the contract or the specifics of the deal.

The article did quote Virginia Health Commissioner Dr. Norman Oliver, who described the types of equipment ordered. In discussing how many sets of PPE are needed to care for those who are presumed positive for COVID-19, Oliver also explained why hospitalization numbers from the Virginia Hospital and Healthcare Association (VHHA) differ from those of the Virginia Department of Health (VDH).

“The new numbers vary from the state’s tally of hospitalized cases because VDH only counts the number of lab-confirmed patients who have been discharged from the hospital, Oliver said at the briefing,” he said. “That number lags behind the number you would get if you were to ask that same question of VHHA, which actually tracks the current census in the hospital.”

This may explain why VHHA began posting its COVID-19 dashboard online on April 6 with current hospitalization details, and why on April 13 it added the number of discharged  patients with positive tests. It does not explain why, as of April 20, VDH still posts a footnote to its hospitalization numbers that says, “Hospitalization status at time case was investigated by VDH. This underrepresents the total number of hospitalizations in Virginia.” It seems likely that Oliver misspoke.

At the April 20 press conference, Governor Ralph Northam said that the Department of Health will expand the data it shares with the public. On a daily basis, VDH will share number of cases, hospitalizations and fatalities by locality and at the health-district level with demographic information like age and race.

Health Commissioner Oliver said long-term care facilities account for 77 of 139 outbreaks and 10% of state cases, but 26% of deaths.

“We have been tracking as best we can, the race and ethnic breakdown on both cases and deaths,” he said. “African American cases number about 28 percent of the cases, for those cases for which we have race and ethnicity data, and for the deaths for which we have data, they number about 34 percent.”

He said they have been improving race and ethnicity data and are missing only about one-third for cases and 3% missing in death data. One-third is still a lot of missing information, but improved from the 50% unknown that Richmond/Henrico Health District Director Danny Avulo reported five days ago.

The VDH website on April 20 shows percentages by ethnicity based on all cases. (Deaths are not currently shown by ethnicity.)

African American: 18.5%; 1,659 cases of which 417 are hospitalized.
Other: 13.2%; 1,183 cases with 266 hospitalizations
White: 33.6%; 3,018 cases with 574 hospitalizations
Race/ethnicity not reported: 34.8%, 3,130 with 243 hospitalizations

Excluding the 3,130 cases with unknown ethnicity, the breakdown of the known 5,860 cases is:

African American 28.3%
Other: 20.2%
White: 51.5%

Oliver’s figure of 34% African Americans of all but 3% of deaths is an alarming number. Knowing the age breakdown will show how alarming that is, especially if more deaths are in younger categories or if the number includes many of the high percentage of deaths in long term care facilities.

Carol Bova is a writer who lives in Mathews County.