Modelling during the Pandemic

Sharing a well thought out LinkedIn post from a colleague, Jawahar Bhalla, who is a leading Systems Engineer and Systems Thinker in Australia, regarding modelling efforts during the pandemic.

Many who have a systems background have bemoaned the global approach of narrow, single point solutions to address the pandemic.

JB’s post references an episode of the podcast "UnHerd", hosted by Freddie Sayers, with guest Dr Camilla Holten Møller (chair of the Expert Group for Mathematical Modelling at Denmark’s public health agency, Statens Serum Institut).

The Danish modelling effort, from all reports, has led to reasonable predictions of cases and hospitalisations in that country. This is in contrast to the results of efforts in other countries with more renowned institutions. The group doing the work in Denmark is presumably smaller than in the elite institutions. My impression is that there was probably less interference, political and otherwise, in their work. Data collection in Denmark appears to be amongst the best. What comes out in the podcast is the calm, clear approach.

The feeling many have is that all modelling presented from elite institutions, often named after Nobel Laureates, is that they have been inaccurate. Either too pessimistic, leading to unnecessary Draconian measures applied to the public or too optimistic and leading to an out-of-control situation. My own feeling is that there is no appetite to accept “uncertainty”. In other words, being able to say we are not sure exactly how things will go, but it could range from here to here.

Taking the worst-case scenario appears to be considered the safest solution to offer. But what happens when the measures applied to deal with the worst-case scenario (independently to the modelling exercise) lead to harms that are worse than the original problem?

We are dealing with what is called a “complex system”.

Problems can be classed as complicated or complex. It took me a while, in my systems engineering education, to really understand the difference. Understandable, given the terms are synonyms in the thesaurus. Problems/Systems can be complicated, meaning there are lots of underlying details, there can be many components interacting in known ways, but if one applies the required effort to the problem it can be solved. Complex problems on the other hand have behaviours that can be “emergent”, ie that we do not anticipate, due to interactions we don’t understand. The term “wicked problem” is used for social or cultural problems that appear to have no solution.

The pandemic is clearly a complex problem and Systems Thinking should have relevance here.

To finish this article with an example of a component of the system that has been poorly dealt with, particularly in Australia, is Aged Care. Under-funding, poor conditions, staff shortages, overwork have been going on for many years. In Australia we have “Royal Commissions” to address big national problems. There was one recently for Aged Care with little done as a result.

An easy solution is to lock everyone up, not allowing families to visit. But what is the human cost? Isolation is not a viable part of the solution. COVID is currently having a big impact on the Aged Care sector and the lack of consideration for the sector has come into focus.

At our 2021 Systems Engineering Society of Australia Healthcare Working Group session we brainstormed issues in Australian Healthcare, not necessarily the pandemic. My takeaway was that most people identified Aged Care as an issue, often through people’s personal experience.

Better “thinkers” and less interference are needed.