An illustration of a coronavirus, created for Centers for Disease Control and Prevention. (Image by Alissa Eckert, MS; Dan Higgins, MAM)
After months of declining to answer reporters’ questions about the accuracy of its COVID-19 data, the Iowa Department of Public Health has reportedly confirmed that it is improperly backdating thousands of positive COVID-19 test results.
The backdating has resulted in a lower number of new infections reported by the state each day.
Bleeding Heartland reported Monday that Bob Ramaekers, the lead epidemiologist for the department’s Surveillance Unit, acknowledged in an Aug. 14 email that IDPH’s COVID-19 website has recorded recent cases of the virus as occurring weeks or months in the past. According to Ramaekers, state officials are aware that this is a problem and are working on a fix.
As the Iowa Capital Dispatch has previously reported, IDPH appears to routinely under count the number of new infections and deaths that it reports each day. The IDPH daily numbers not only disagree with the daily counts reported by others, such as the New York Times, but conflict with other data published by IDPH.
The Capital Dispatch told IDPH of that problem in May, and received no substantive response to its inquiries about how the numbers were being calculated. The news organization then began relying more heavily on the the New York Times’ COVID-19 tracker for daily counts of infections and deaths in Iowa.
The acknowledgement that IDPH is under-reporting daily infection counts follows the Dubuque Telegraph Herald’s recent report showing that the state is using an unusual methodology to calculate average infection rates in a community. The manner in which those rates are calculated is critical because the averages play a significant role in the state’s determination of whether school districts are granted waivers from requirements for in-person classroom instruction.
The backdating of cases could provide an answer to a question that IDPH has long declined to answer: Why does the agency’s cumulative total for infections remain relatively consistent with the numbers reported by others, even though its daily count of new infections is often dozens, or even hundreds, below what others are reporting?
According to Bleeding Heartland, nurse practitioner Dana Jones recently asked IDPH about the fact that many new cases of COVID-19 appear to be assigned to dates weeks or months in the past. She wrote: “I would like to know why new positive COVID cases were added to dates as far back as March over the last week. This has been a consistent occurrence since I started following the data.”
Ramaekers replied to Jones on Aug. 14, explaining that positive test results were being linked to the date a person was first tested — even if that first test was months before and had come back negative. “We recognize this is a problem and have been working on logic to handle it,” he wrote, according to Bleeding Heartland.
Last month, the Capital Dispatch told IDPH spokeswoman Amy McCoy that because of obvious problems with the agency’s COVID-19 data, the news organization couldn’t rely on IDPH for accurate numbers.
In an email to McCoy, the Capital Dispatch wrote: “Because these issues are so obvious to anyone who has used the IDPH site, and because these issues have been raised so many times with the department, and because these issues haven’t been explained, let alone addressed, I have to assume all of this is deliberate.”
That communication followed several weeks of back-and-forth with the department, beginning on May 11, when the Capital Dispatch wrote to McCoy and stated:
“On the IDPH site, the graph showing new infections indicates 389 newly confirmed cases for today. But the difference between today’s numbers and yesterday’s appears to be 414. So media are reporting both numbers. Can you tell me what’s going on there? Also, will the newly revamped site include the newly confirmed deaths each day, or has that reporting been abandoned?”
After receiving no response, the Capital Dispatch wrote again to McCoy in an email marked “URGENT,” and noted: “We’re getting conflicting information as to the answer to the question I posed …. This is something we need to get settled ASAP because I can’t tell what each day’s ‘new infection’ count is — the number shown in the bar graph on the site, or the number derived by taking the ‘total to date’ for yesterday and subtracting it from the ‘total to date’ for today.”
The Capital Dispatch also asked the governor’s spokesman, Pat Garrett, about the numbers, stating: “I do understand why cases might be added to a day retroactively, but what I don’t understand is why the website simultaneously reports conflicting numbers.”
In response, Garrett provided a one-sentence statement: “If we revise a day like 5/10 upwards 5-6 cases let’s say so the larger to date number will show an increase of 5-6 cases.”
On June 30, the Capital Dispatch asked McCoy about the number of COVID-19 deaths being reported by IDPH: “As I understand it, IDPH is using more restrictive criteria than the CDC is using in determining what constitutes a COVID-19 death, with the CDC reporting the deaths in which the virus is considered the cause as well as those deaths in which the virus is one of several contributing factors. Iowa apparently is only reporting the former, and not the latter. Can you tell me whether that is, in fact, the case — and, if it is, when the state established this specific method of calculating and reporting COVID-19 deaths?”
After receiving no response to that question, the Capital Dispatch wrote again to McCoy: “…The data on your COVID-19 Long-Term Care site doesn’t say whether the numbers are cumulative or current. Judging by the size of the numbers, they would appear to be cumulative … but that can’t be right as the total number of recovered and dead individuals is greater than the total number reported to have been infected. Also, the data you provide for homes with outbreaks seems to be current, not cumulative, as evidenced by homes with known outbreaks, such as Bishop Drumm, not being listed there anymore. Can you shed any light on all of this?”
McCoy responded by saying there needed to be “a system fix” to address the issue with the agency’s reporting of long-term care outbreaks. She didn’t address the question about IDPH’s reporting of deaths.
On July 20, the Capital Dispatch wrote again to McCoy about new infections and noted that “the numbers you report for individual days are sometimes hundreds less than what the (New York Times) reports for those same days, and they are also less than the day-to-day difference in IDPH’s own reporting.”
On July 21, after receiving no response, the Capital Dispatch wrote again to McCoy: “I’ll ask again … Why don’t the number of daily new infections as reported by IDPH agree with the difference between one day’s total and the next day’s total as reported by IDPH?”
McCoy responded, “As we have explained many times including at news conferences, the website is constantly updating. If cases need to be reassigned, for example because we got more information that someone actually lives in a different county, the numbers will be adjusted. Or, if someone has been tested multiple times and we need to remove duplicates, we will adjust. This is a constant process.”
In response, the Capital Dispatch stated the adjustments cited by McCoy couldn’t explain such large discrepancies in the numbers, adding, “Obviously, IDPH has no intention of answering the questions we ask or explaining the inconsistencies in the data that it reports … but it’s happy to continually boast of its commitment to transparency and openness.”
McCoy replied, “My apologies, I didn’t mean to sound like I wasn’t wanting to be responsive. Can you help me understand better which discrepancies you are referring to? … Happy to try and get more information to you soon.”
The Capital Dispatch responded: “Pick any day you choose — the IDPH report of new infections for any given day (shown in the website’s bar graph) rarely comes close to the IDPH-reported increase in the total (cumulative) cases for that day.” The news organization also noted on July 10, IDPH reported 628 new COVID-19 cases in Iowa, but the New York Times reported 916, a record at that time.
McCoy never responded to the email raising those issues. Agency spokesman Matt Highland said he would set up a time for the department’s medical director, Dr. Caitlin Pedati, to discuss the issues. The department has yet to schedule that discussion.
Our stories may be republished online or in print under Creative Commons license CC BY-NC-ND 4.0. We ask that you edit only for style or to shorten, provide proper attribution and link to our web site. Please see our republishing guidelines for use of photos and graphics.