SA could run out of ICU beds by early June


SA is going to run out of ICU beds, and it could be as early as June.

And the situation would be have been a lot bleaker without the lockdown, which reduced the transmission of Covid-19 in the country by between 40% and 60%.

Critical patients on average require six days in a hospital ward plus another 10 in ICU if recovery happens. Should they die, it usually happens after six days in ICU.

By July 16.4-million N95 face masks, and 1.3-million testing swabs will be needed.

These figures are according to different modellers who took part in a presentation by the department of health on Thursday.

Health minister Zweli Mkhize said he wanted to share with the public how modelling works, and the large amount of data that had to be factored in.

He also called on the public to acknowledge that this crisis is unfolding before our eyes, and there is thus no way of knowing if a model is perfectly accurate until after the fact.

“We must appreciate that this has been an unfolding pandemic and we have all been learning over the past few months what this means for our country. The whole world is struggling with the same pandemic and nobody has all the answers,” he said, “but there are many lessons to learn that can guide our response as South Africans.”

University of Stellenbosch professor, Juliet Pulliam, speaking on behalf of the SA Covid-19 Modelling consortium, said the lockdown has worked and yet we will still face a shortage of ICU beds, and probably very soon.

The consortium concluded that “the initial social distancing and lockdown measures have worked” by delaying the peak and allowing time for “mitigation measures for the upcoming wave”.

It said “the extension of the lockdown to five weeks bought us critical additional time to ramp up community testing and to prepare”.

Optimistically, lockdown reduced transmissibility by as much as 60%, or at the lowest, by 40%.

During level 4 lockdown alone, we have reduced transmissibility by between 25% and 35%, and if we stick to social distancing from the end of May, we can reduce it by between 10% and 20%.

It said the pessimistic estimate for a peak in active cases would be from early July, while a more optimistic view would peg it as coming in early August.

While there are many variables, there is one certainty, they say: “under almost all scenarios, hospital and ICU capacity will be exceeded”.

Timing and extent are uncertain, but we will definitely run out.

“The longer into the future you project, the more uncertain it becomes,” she said, adding that either way, a shortage of ICU beds “is of particular concern”, especially since that could happen as early as the first week of June.

Pulliam says that the modelling doesn’t capture superspreading events like funerals (on the downside) or the positive effect like behavioural change, which could reduce transmissions substantially.

“We know from situations like ebola and Covid-19 in other settings that people adapt their behaviours in response to seeing a lot of mortality resulting from an event. We don’t factor that in so we may overpredict deaths but other factors might mean we are underpredicting.”

Barry Childs, speaking on behalf of the Actuarial Society of SA (ASSA), also emphasised the fact that with so many variables at play, it was difficult to have a definitive answer.

He said the organisation had experienced wide-ranging and harsh feedback, but that the overriding difficulty is “having to make such significant decisions amid such uncertainty”.

An example, some critics said the ASSA was “four times too high on the death estimates for the country, while others said it was four times too low”.

One variable changes another.

For example, as we discovered how many asymptomatic cases were in the mix, we had to reduce the estimate of the fatality rate.

Despite these uncertainties, it is crucial to “produce a model for use by the broader profession that is sound in its basis and provides guidance in its application”.

Within that model, using base parameters, useful data for planning emerges.

For example, he said that averages for hospital stays showed “10 days if not critical, six days if critical plus another 10 in ICU if recovery happens, or six days before dying”.

Also, about a fifth of all Covid-19 hospital admissions would end in ICU.

Deloitte has been focused on identifying the level of PPE, ventilators and hospitalisation that will be required as we move towards the peak.

Their modelling shows how the demand for such resources grows exponentially with each new week.

For example, according to their modelling, says health care actuary Ashleigh Theophanides, in June, almost 3.5-million N95 masks would be needed, and about 280,000 testing swabs. But, in a single month, those numbers would have increased to 16.4-million N95 masks, and 1.3-million testing swabs.

In June, we would need about 10,971 ICU beds, but by August that figure would approximately double to 20,482.

Like the other modellers, Theophanides says their models “are recalibrated every few weeks” and that their current estimates of deaths sit at about 40,000, but “the work the department of health has done to set up field hospitals could have a significant impact on the expected increase in fatalities”.

One dissenting voice in the modelling is that of Nick Hudson, an actuary who co-ordinates Panda (Pandemic Data and Analytics), who argued that an empirical approach shows that “there is no correlation between lockdown stringency and how many days to reach the peak of infections”.

He said it was possible that parts of the worldwide population “are simply not susceptible to the infection”, and that “lockdowns do not result in any flattening of the curve”.

Hudson cautioned against “repeating mistakes in other countries where huge overestimations in deaths” had occurred, and said that “the consequences of lockdown present a vast humanitarian disaster”.

He said: “We mustn’t use models that cause us to stay in lockdown — poverty kills as much as viruses.”


Please enter your comment!
Please enter your name here

CAPTCHA ImageChange Image