In response to an inquiry from TSL, The Claremont Colleges Services confirmed this week that Student Health Services over-reported the apparent number of COVID-19 cases at the 5Cs in recent weeks, an error that translated to some of the colleges’ individual testing dashboards. More than one hundred positive tests were mistakenly reported since March 27, TSL found.
In January, SHS began using antigen tests to determine whether those who had already tested positive for COVID-19 could be released from isolation after at least five days, the result of new guidance from the Los Angeles County Department of Public Health. If and when those people tested positive on the additional tests, the results were inadvertently added to case counts for the week, artificially inflating the totals, The Claremont Colleges Services spokesperson Laura Muna-Landa said in an email.
“Tests completed for isolation release are follow-up tests of previous positive test results and should have been excluded from the count totals,” Muna-Landa said.
TSL first discovered the error through an automatic screenshotting tool it developed to track changes in college data reporting for its COVID Dashboard project, the only publicly available tracker for cases across the 5Cs.
On April 19, Pomona’s dashboard indicated 205 student positive tests had occurred between March 27 and April 16. But two days later, the screenshots show, the data changed to reflect 159 positives during the same time period.
“TCCS has corrected the data, and the updated information is now reflected in the dashboard. The revisions do not alter the overall upward trend in cases after spring break, though they do reduce the scale,” Dean of Students Avis Hinkson said in an email to TSL. “Numbers provided in messages I sent to all students that referenced counts on specific days were accurate.”
Pitzer College’s dashboard, meanwhile, was revised to 101 positive student tests for the same three-week period, down from a previous total of 136. Harvey Mudd College’s dashboard decreased from 100 cases to 78.
“The majority of the duplicate reported positives were contained to a small timeframe in the second half of the post-spring break surge, reflective of more individuals having follow-up antigen tests to determine if they were still positive or not,” Muna-Landa said.
TCCS, the organization that administers SHS, first discovered the issue on April 18 and informed the Consortial Response Group on April 20, Muna-Landa said. The data was corrected April 22 and the entire data posting procedure was reviewed “to ensure no other undetected data reporting issues.”
Claremont McKenna College tests through a separate organization, as do faculty at Pomona, Pitzer and Scripps College, so those case counts were not affected.
“PCR tests and antigen tests are tracked separately,” Muna-Landa said. “SHS also separately tracks the antigen tests used for early release from isolation and the antigen tests used for symptomatic individuals requiring an immediate test result. The latter antigen tests have always been included in the testing totals on our dashboard as they are initial tests, just as the Shield T3 saliva PCR tests are initial tests.”