Min­is­ter of­fers sight of Covid-19 mod­els

• While health min­is­ter has lifted lid on pro­jec­tions, num­bers can­not be prop­erly eval­u­ated with­out peer re­views or re­lease of codes

Business Day - - FRONT PAGE - Ta­mar Kahn Science & Health Writer kahnt@busi­nesslive.co.za

In the face of grow­ing pres­sure from busi­ness and citizens, health min­is­ter Zweli Mkhize this week lifted the lid on the mod­el­ling work guid­ing the gov­ern­ment’s re­sponse to the Covid-19 cri­sis.

In the face of grow­ing pres­sure from busi­ness and citizens, health min­is­ter Zweli Mkhize this week lifted the lid on the mod­el­ling work guid­ing the gov­ern­ment’s re­sponse to the Covid-19 cri­sis.

For the first time on Tues­day evening the public was given sight of the pro­jec­tions from a con­sor­tium of ex­perts con­vened by the Na­tional In­sti­tute for Com­mu­ni­ca­ble Dis­eases (NICD), which in­di­cate there could be as many as 3.7-mil­lion de­tected cases and be­tween 40,000 and 48,000 deaths by No­vem­ber.

Mil­lions more cases will go un­de­tected, and 475,000680,000 pa­tients are likely to re­quire hos­pi­tal­i­sa­tion, said the con­sor­tium, which in­cludes ex­perts from the SA Cen­tre for Epi­demi­o­log­i­cal Mod­el­ling and Analysis at Stel­len­bosch Univer­sity, the Mod­el­ling and Sim­u­la­tion Hub, Africa at the Univer­sity of Cape Town (UCT), and the Health Economics and Epi­demi­ol­ogy Re­search Of­fice at Wits. The con­sor­tium’s mod­els cap­ture the public health im­pact of Covid-19, but do not con­sider the so­cial and eco­nomic bur­den of the dis­ease or the lock­down.

The scale of the pro­jected num­bers ex­plains why the gov­ern­ment has sought to de­lay the in­evitable avalanche of in­fec­tions with re­stric­tions on trade, travel and so­cial in­ter­ac­tion, and buy pre­cious time to ramp up the health sys­tem’s ca­pac­ity.

While the an­tic­i­pated num­ber of deaths is dwarfed by those as­so­ci­ated with HIV, TB and other in­fec­tious dis­eases, the ex­tent of Covid-19 in­fec­tion and the sheer vol­ume of pa­tients re­quir­ing ad­mis­sion to hospi­tal is set to stretch the health sys­tem to break­ing point.

In parts of the coun­try, that day may ar­rive soon: the Western Cape, which has more than 60% of SA’s cases, is see­ing a rapid uptick in hospi­tal ad­mis­sions, and al­ready the 23 ICU beds in Tyger­berg Hospi­tal are re­port­edly all oc­cu­pied. At the last count, there were over 19,137 cases of Covid-19 in SA, more than 12,000 of them in the Western Cape. The province has recorded a sim­i­lar pro­por­tion of the 369 recorded deaths.

To put this in con­text, in 2017 there were more than 205,000 deaths from in­fec­tious dis­eases in SA, and 115,000 from non-com­mu­ni­ca­ble dis­eases, ac­cord­ing to the Ac­tu­ar­ial So­ci­ety

of SA. And in 2018 there were 61,000 deaths from tu­ber­cu­lo­sis alone, ac­cord­ing to the World Health Or­gan­i­sa­tion.

How­ever, most Covid-19 deaths are ex­pected to oc­cur in no more than two to three months, rather than the course of a year, pos­ing an ex­treme chal­lenge to the health sys­tem.

The min­is­ter has also sought to show that his of­fi­cials are pay­ing at­ten­tion to other groups work­ing on the po­ten­tial tra­jec­tory and ef­fect of the dis­ease, and on Thurs­day show­cased a raft of mod­els pro­duced by Deloitte, the Ac­tu­ar­ial So­ci­ety of SA (Assa), var­i­ous in­de­pen­dent aca­demic groups at Wits and UCT, and the Pan­demic Data and An­a­lyt­ics (Panda) con­sor­tium.

Deloitte’s mod­els have in­formed the work of Busi­ness for SA (B4SA), which is sup­port­ing sev­eral gov­ern­ment depart­ments and helping with the pro­cure­ment of es­sen­tial med­i­cal sup­plies. It es­ti­mates there will be about 2.9-mil­lion cases and 40,000 deaths, and of­fered a glimpse of its pro­jec­tions for the de­mand for personal pro­tec­tive equip­ment, ven­ti­la­tors and oxy­gen, which have not pre­vi­ously been pub­lished.

Most of the other work high­lighted by the min­is­ter is al­ready in the public do­main. Assa, which has elicited a del­uge of com­ments from the pro­fes­sion in re­sponse to its pre­lim­i­nary model, has pro­jected cu­mu­la­tive deaths for var­i­ous sce­nar­ios, which hover around the 48,000 mark.

Wits aca­demic Alex van den Heever has been vo­cal in his crit­i­cism of the eco­nomic and so­cial costs of the lock­down, as has the Panda con­sor­tium’s Nick Hud­son. Van den Heever’s work has fo­cused on the ef­fect of var­i­ous in­ter­ven­tions and con­cluded it would be far more cost ef­fec­tive to rad­i­cally scale up test­ing and con­tact trac­ing than con­tinue with the lock­down.

Panda’s analysis con­cluded that the lock­down would trig­ger such eco­nomic dam­age that it could cause 29 times more deaths in the long run than Covid-19.

But pre­sent­ing re­sults is not enough. It is im­pos­si­ble for ex­perts to eval­u­ate the ma­jor­ity of these plan­ning tools: few groups have sub­jected their work to peer re­view or re­leased their model code so other sci­en­tists can repli­cate their work.

The stakes are so high, in terms of public health and the eco­nomic and hu­man­i­tar­ian con­se­quences of the lock­down, that the gov­ern­ment, B4SA and in­de­pen­dent groups alike need to pro­vide far greater dis­clo­sure of their work.

Zweli Mkhize

Newspapers in English

Newspapers from South Africa

© PressReader. All rights reserved.