When NOLA Public Schools released COVID-19 case data on Jan. 31 for the week prior, not a single KIPP New Orleans school was listed as having a positive case or student or staff member in quarantine — quite a feat for a 6,000-student charter network amid a pandemic.
KIPP, the largest charter group in the city, submitted reports to the district and the Louisiana Department of Health that week, network Director of Risk and Compliance James O’Donnell said in an interview this week. O’Donnell could not confirm exactly what was in those reports — submitted through a web survey — but said the charter school network was aware of diagnosed cases from the period in question.
NOLA Public Schools spokeswoman Taslin Alfonzo confirmed in a statement that there was an issue with KIPP’s reporting. It’s not clear, however, if the problem was on KIPP’s end or the district’s end. Asked what may have caused it, she did not provide answers.
“As it has since the start of the pandemic, NOLA-PS will continue to provide COVID-19 tracking data as provided to the District by our community of schools. Schools are encouraged to send their information through our electronic process,” Alfonzo wrote in response.
She added that the district was not aware of other schools facing similar reporting issues.
“We have no verification of a school that could not access our online reporting form. If they experience issues, we are here at the District to provide assistance,” she said.
However, the school district’s overall case count last week appeared suspiciously low. Its report of 384 cases was less than half of those reported by the Louisiana Department of Health. The state department reported 805 cases in Orleans Parish schools the same week. What’s more, the district’s report covers a longer timeframe. Its tally is supposed to include all “active” cases — those diagnosed over the previous two weeks. The LDH count only includes cases newly diagnosed during the previous week.
It’s not unusual for the case counts in the two reports to be different. Apart from the different timeframes, the state’s data can include any school in the parish, not just charter schools overseen by the district.
The state tally — which is released every Wednesday and covers cases reported over the previous week — can include cases from private schools and state-run charter schools as well as cases connected to NOLA Public Schools. (Unlike the school district, which reports case counts by school, the state reports only a cumulative parishwide number every week and does not make more specific data public.) Even so, the district and state counts are typically much closer than what was reported last week.
There were other issues with the district’s report last week. A press release said that district schools performed 20,773 tests with a 2.8 percent positivity rate. That should yield 581 positive cases, yet the district only reported 384 as mentioned above.
Alfonzo said that discrepancy could be caused by testing delays.
“Our schools continue to test students during and through the reporting period, and tests may take up to 72 hours to yield results, which may lead to schools not being notified of all positive test results before the end of each weekly reporting period.”
Ongoing issues
Numbers released this week by the district and the state are much more closely aligned than in last week’s report. On Monday, NOLA Public Schools reported 525 active cases among students and staff at its charters. And on Wednesday, the state reported 567 newly diagnosed cases connected to schools located in Orleans Parish.
But there were still some outstanding issues in the district’s data.
Though KIPP schools are showing 157 COVID-19 cases in this week’s data, the district’s reporting shows zero quarantines at KIPP campuses. O’Donnell said the network has had fewer quarantines in recent months as student vaccinations have increased, but they still have some quarantines. (The district has previously had problems reporting accurate quarantine numbers.)
Of course, data errors are bound to happen occasionally. The district’s system is dependent on schools self-reporting their numbers on time and accurately and on its COVID-19 web survey working properly. However, KIPP’s reporting issue did not appear on last week’s report as an error or non-report. It simply said the network had zero cases.
Tulane University epidemiologist Dr. Susan Hassig reviewed the district’s databoard. She said she thinks there’s an easy fix to show missing data. Under current guidelines, schools don’t submit anything if they have no cases to report, and the district’s dashboard autofills those schools as having zero cases or quarantines. That system works fine until a school with reportable cases either fails to submit its data or encounters problems with the web survey.
If schools had to submit weekly reports even to say they have zero cases, and if non-reporting schools were labeled as such in the data dashboard, the district would immediately be able to distinguish between schools without cases and schools with missing data.
“I am very much opposed to using zero for a field that doesn’t have a report,” she said in an interview Wednesday. “It needs to be a non-valid response for the field, so if it’s a numeric field it should be a non-numeric response like an NR or an X.”
“If you use zero when there’s a ‘no report,’ that’s messing things up big time. It’s not an uncommon mistake. It happens all the time,” she said.
“Sometimes this happens with my teaching assistants,” the professor said. “Zero may be ultimately what a student gets, but I need to know [the assignment] is missing. It’s a distinction and it can be a really important distinction when you’re thinking about what your data is telling you.”
Hassig noted that schools have done a lot of work to set up entirely new testing and reporting systems over the course of the pandemic, which has been a challenge.
“New data systems can be really challenging for environments that are overtaxed,” Hassig said. “They are doing a lot of testing which is really valuable to keeping schools open in New Orleans.”
But, she said, making sure the data is as accurate as possible is important for tracking the spread and making policy decisions
“The coding of non-report versus no cases is important at many levels — oversight at the district level plus case counts,” she said. “Knowing that x number of schools didn’t send in reports allows the district to immediately do follow up.”
“One thing we have learned over our two years experience with Sars-COV-2 is it’s complicated understanding how to capture and report this information. It’s something that is a different responsibility for a lot of people,” she said. “There’s limited experience within the educational environment with this kind of reporting. I think it’s a reminder that from the public health side that we’re supporting the front line institutions and helping them to do it just a little bit better.”