Understanding the Current Epidemic from Dates of Symptom Onset

5/15/20, 4:35 PM

Understanding the Current Epidemic from Dates
of Symptom Onset
4/27/2020

Brief Report Prepared for the Iowa Department of
Public Health by the University of Iowa College of
Public Health COVID-19 Response Group
Submitted by

Grant Brown, Ph.D., Assistant Professor of Biostatistics
Joseph Cavanaugh, Ph.D., Professor and Head of Biostatistics
Aaron Miller, Ph.D., Assistant Professor of Epidemiology
Jacob Oleson, Ph.D., Professor of Biostatistics and Director of the Center for Public Health
Statistics
Michael Pentella, Ph.D., Clinical Professor of Epidemiology and Director of the State Hygienic Lab
Eli Perencevich, M.D., M.S., Professor of Internal Medicine and Epidemiology, and Director of the
Center for Access and Delivery Research and Evaluation
Daniel Sewell, Ph.D., Assistant Professor of Biostatistics

Highlights
1. Use of symptom onset dates in any estimation of COVID peaks and trajectory is incorrect unless
the most recent 7-10 days of data are excluded
2. Current estimates of the epidemic curve using our statistical modeling with Iowa speci!c data
suggest Iowa has not yet reached a peak
3. Opening before the peak is reached is expected to result in a rapid rise of cases throughout the
state

file:///Users/laurabelin/Downloads/IDPH%20Brief%20Report%20-%20Onset%20of%20Symptoms%20Data%20(1).html

Page 1 of 5

 Understanding the Current Epidemic from Dates of Symptom Onset

5/15/20, 4:35 PM

Report
To understand the progression of COVID-19 in Iowa, it is critically important to understand the
number of individuals whose symptoms begin on each day since the beginning of the outbreak in
Iowa. However, due to a systematic pattern of missing reporting of data, we need to be careful in
looking at only the daily reported counts. The result of doing so leads to a misrepresentation of the
current epidemic trajectory.
For each day starting on March 20, we computed the curve showing the daily number of new
infections as determined by symptom onset using the information that was available on that day
(Figure 1). The most current curve is shown in the darkest green, and each daily curve going back in
time is shown in a lighter shade of green. The pattern of the curve based on showing new cases
always appears to be peaking no matter which daily curve we look at. Using this metric, it has
looked like Iowa is soon peaking every day for the past 6 weeks, even though this has not been the
case. The inherent tendency of these curves to erroneously show that Iowa has passed the peak is
due to the time lag between symptom onset and positive test results. This time lag needs to be
incorporated into these !gures to gain an accurate representation of Iowa’s curve.

file:///Users/laurabelin/Downloads/IDPH%20Brief%20Report%20-%20Onset%20of%20Symptoms%20Data%20(1).html

Page 2 of 5

 Understanding the Current Epidemic from Dates of Symptom Onset

5/15/20, 4:35 PM

To account for the time lag between symptom onset and positive test results, we used an
imputation procedure to estimate the daily number of new infections as measured by symptom
onset. The resulting curve is given in Figures 2 and 3. Figure 2 shows the number of new cases. The
observed counts are shown in the red trajectory, but accounting for the lag e"ect, we estimate the
blue curve. Cumulative cases are shown in Figure 3. Note that these !gures only re#ect the state of
the epidemic up to the last day in the data we have received from

file:///Users/laurabelin/Downloads/IDPH%20Brief%20Report%20-%20Onset%20of%20Symptoms%20Data%20(1).html

Page 3 of 5

 Understanding the Current Epidemic from Dates of Symptom Onset

file:///Users/laurabelin/Downloads/IDPH%20Brief%20Report%20-%20Onset%20of%20Symptoms%20Data%20(1).html

5/15/20, 4:35 PM

Page 4 of 5

 Understanding the Current Epidemic from Dates of Symptom Onset

5/15/20, 4:35 PM

In summary, considering the dates of symptom onset is critical to understanding the current state
of the COVID-19 outbreak in Iowa. However, due to a time lag between symptom onset and
receiving a positive test result, there is a systematic missingness in the data that spuriously
demonstrates Iowa has progressed past the peak of the epidemic. In actuality, accounting for the
time lag paints a di"erent picture, namely we are still on an upward trajectory and have not yet
reached the epidemic’s peak.

file:///Users/laurabelin/Downloads/IDPH%20Brief%20Report%20-%20Onset%20of%20Symptoms%20Data%20(1).html

Page 5 of 5