California's coronavirus test result data may be flawed - Los Angeles Times
Advertisement

California’s coronavirus test result data may be flawed, top health official says

A man wears a mask in Huntington Beach
A man wears a mask in Huntington Beach.
(Allen J. Schaben / Los Angeles Times)
Share via

A steep decline in California’s coronavirus infection rate announced this week by Gov. Gavin Newson may not be accurate, according to the state’s top public health official who said Tuesday that the state’s data system used to process COVID-19 test results is marred with technical issues.

The problems have caused delays in analyzing test results and cast doubt on Newsom’s announcement Monday of a 21.2% decline in the seven-day average rate for positive infections compared with the average from the week before.

California Health and Human Services Secretary Dr. Mark Ghaly said that “the seven-day positivity rate is absolutely affected” by the issue. It’s unclear to what extent and for how long cases have been undercounted, and how this situation differs from the more routine delays when test reporting lags over weekends.

Advertisement

The news comes as officials have expressed guarded optimism that the surge of COVID-19 cases in California might be peaking.

A Times analysis found that California has now experienced its first weekly reduction in new confirmed coronavirus cases for the first time in 12 weeks. For the seven-day period that ended Sunday, California reported 59,697 new cases, a drop of 9% from the previous week of 65,634 cases, which was a pandemic record.

On Tuesday, Los Angeles County public health officials reported an additional 1,901 coronavirus cases and 57 related deaths. In a statement, the health department said it learned of new technical issues related to the collection of test results during an emergency meeting Monday night with state officials.

Advertisement

“This issue has undercounted the county’s positive cases and affects the number of COVID-19 cases reported each day and our contact tracing efforts,” the county said.

The health department is working to contact at least 81 laboratories to obtain test results since July 26 in order to determine the accurate count.

The latest maps and charts on the spread of COVID-19 in California.

The update follows the county’s warning last week that it was expecting a backlog in cases “due to previous reporting delays in the state electronic lab system.” The following day, the county reported highs of 4,825 new coronavirus cases and 91 deaths. Other counties, including Sacramento, Placer and Orange — which reported 263 additional cases Tuesday and two deaths — have recently included a warning on their dashboards that case counts may not be accurate.

Advertisement

“The state’s electronic disease reporting system has been experiencing issues processing incoming reports. Therefore, recent data published on the Sacramento County Public Health COVID-19 dashboards are likely to be an underestimate of true cases in the County,” Sacramento Public Health Department said.

The undercount issue does not affect hospitalization and intensive care data, Ghaly said. Those numbers recently plateaued across California after the state saw two days of record-setting fatalities last week and surpassed 9,000 coronavirus- related deaths.

Some counties continue to be hit hard as others see a decline in severity. On Tuesday, San Bernardino County reported 59 additional deaths — its highest single-day count to date.

Hospitalization data are collected differently than the state’s test result numbers, which are gathered through CalREDIE — an electronic system that feeds information from laboratories to the state and local health systems.

California Department of Public Health officials are trying to determine “where data is getting stuck,” Ghaly said.

“We’re not sure when we’ll have a definitive fix to the problem.”

In the meantime, the health department is implementing manual processes to retrieve the information. During a news briefing Tuesday afternoon, Ghaly stressed the importance of focusing on long-term trends over snapshot data.

Advertisement