Pittsburgh professors see flaws in coronavirus modeling, predict more grim outlook

Two Pittsburgh academics argue that much of the modeling regarding the trajectory of the covid-19 pandemic has been fundamentally flawed — and the real outlook for the disease is much more dismal.

Wesley Pegden and Maria Chikina have joined an international conversation about projections of how dire the covid-19 pandemic may become.

In a scathing analysis published March 29 on Medium, “A Call to Honesty in Pandemic Modeling,” Pegden and Chikina assert that many of the popular models being cited by national media and President Trump are misleading. Pegden is an associate professor of mathematics at Carnegie Mellon University and Chikina is an assistant professor of molecular biology at the University of Pittsburgh; they are a married couple.

“The idea that normal life can resume in two to three months without having a huge wave in infections — there is just no science behind that,” Pegden said in a Tribune-Review interview.

Their research concludes that models claiming social distancing will flatten the curve in as little as two months are flawed.

They predict that without additional interventions, including additional ventilators — and a vaccine — the pandemic will return with a vengeance the moment society resumes normal activity — regardless of how long a lockdown lasts.

“The duration of containment efforts does not matter, if transmission rates return to normal when they end, and mortality rates have not improved,” they write. “This is simply because as long as a large majority of the population remains uninfected, lifting containment measures will lead to an epidemic almost as large as would happen without having mitigations in place at all.”

In the article, the professors re-examine several models, pointing out that each one maps the pandemic only for a limited time frame, often neglecting what happens after mitigations end.

The idea for their Medium article came after reading a March 25 opinion piece in The New York Times, which makes the case that social distancing for two months will save drastically more lives than the same mitigation for two weeks. The models illustrated in the column, by Nicholas Kristof and Stuart Thompson of the Times and based on the research of mathematicians and epidemiologists, do not show projected covid-19 projections past October. Pegden described the column as “particularly egregious.”

“If the future is the same as the present, you’re still going to have an epidemic that is just as bad,” Pegden said. “I was very upset that somebody was misleading people in this way about what the point of these mitigations are, and I think it’s dangerous to do that.”

But the Times’ piece isn’t the only article pulling information from scientific models to draw conclusions about the pandemic’s scope. At this point, still on the upward slope of “the curve,” the U.S. is obsessed with them.

A FiveThirtyEight article summarizes the complex nature of building a pandemic model: There are a variety of factors to consider and very little reliable data at this point. An April 2 piece in The Atlantic warns that models are not meant to be used for sweeping predictions anyway — they are meant to lay out a range of possibilities, all uniquely dependent upon what actions are taken.

That hasn’t stopped commentators and lawmakers from using the models as definitive tools for policymaking. In Tuesday’s Coronavirus Task Force briefing, Dr. Deborah Birx quoted from a study by Christopher J.L. Murray from the University of Washington. She said the model was being used by the task force to predict the possible “time course” of the disease and the number of people who could potentially die.

Pegden and Chikina mention Murray’s study in their Medium post as well, pointing out that his model only covers four months, stopping in July. “What happens in August?” they ask.

The lockdown is less important than what is actually achieved in the meantime, the authors write. The mitigations being put in place by state governments, including Pennsylvania, should only be viewed as a “stall tactic” while effective strategies are formulated and carried out.

An “honest model,” the authors write, is one only used for short-term planning.

Additionally, “no model whose purpose is to study the overall benefits of mitigations should end at a time-point before a steady state is reached,” they write.

There are more optimistic takes on the future of the pandemic than the one offered by Pegden and Chikina. Current data suggests that only a small portion of those who test positive for covid-19 ever display symptoms, and of those who do, the majority are able to pull through the disease.

But even that data is incomplete, Pegden said. As testing in the U.S. continues to stall, there is still uncertainty in the actual mortality rate for covid-19, as well as the true number of positive cases.

Uncertainty is the theme within Pegden and Chikina’s analysis.

“It’s a very serious problem,” Pegden said. “Right now, people don’t have a full plan worked out for how this will happen or how this will end.”