Infectious disease - opinion piece

By Dr Nigel Raymond

19 March 2020

At the moment NZ seems to have a window of opportunity to contain the spread of COVID-19. With the border and other control measures so far, we have few reported cases so far.

What we’ve seen with other countries is that there have been more cases and transmission than realised, with reported cases then appearing some 2-3 weeks or more later. No individual containment measure will be perfect, but the more areas of most risk that are identified and responded to in a multipronged way, the experience in China and some other countries has shown that containment can work.

It does seem from reports that fever, cough, and perhaps sputum are the most common symptoms in the clinical presentation, more than for colds and influenza where nasal symptoms are more typically present, albeit with considerable overlap in symptoms.

Requiring both fever AND respiratory symptoms will be more specific but will be less sensitive i.e. missing a moderate number of cases. The experience of countries that have successfully contained COVID-19 and modelling indicates that finding mild cases and getting them isolated early is crucial, and pays off because it can reduce the efforts required to reign in bigger clusters if cases are found even a few days later. Public personal protective measures will all work together to contribute. It seems that in the 80% or so with milder illness (out of hospital), that there is perhaps most infectiousness during the first week or so.

Some asymptomatic people have infection but seem to contribute less, they were reported in China to be the minority of ‘transmission chain.’ So, identifying people with mild infection early seems to be important, at least at this phase in NZ where there is a chance of containment.

The challenge then is to increase the sensitivity of finding mild cases early, but still focusing resources on where they will make the most benefit. Some people arriving from overseas before the travel restrictions were put in place may develop symptoms, and there could have been some secondary spread we don’t yet know about. Testing people with no symptoms and not contacts of confirmed cases would generally seem to be a low yield undertaking, even though there will be some, which is why that is outside what’s outside testing criteria. If someone has symptoms more typical of a cold, the specificity of testing will be lower, but the benefit of finding COVID-19 considerable. ‘Symptomatic’ is also a clearer message, as deciding exactly which symptoms or combinations is likely to become confusing. ‘All overseas countries except the Pacific’ seems to have made that message clearer, than when new countries were being added stepwise.

The criteria used for close and casual exposure was made more by our colleagues including those with infection control and public health expertise. Both less specific mild ‘cold’ symptoms and casual contact are each lower risk, so the product is very low risk. While NZ is trying to ramp up testing (as advised today by the WHO DG), constrained GP and test capacity should be used to best effect as able.  The point made about ‘physical distancing’ is an interesting one. With respect to GP waiting rooms, I note how a proximity within of a couple of seats has been used by PH for airline passengers.

I think there could be a discussion about these criteria particularly applying to healthcare workers (HCW) in the primary and secondary sectors. If a GP or hospital doc has symptoms, which while non-specific sometimes turn out to be COVID, they are likely to expose more other HCWs and people with health conditions.

It is also worth noting that obviously the PCR tests are not 100% sensitive, maybe 70-80% (our micro colleagues can advise further). Strategies to sample more than one site (deep nasal, NP, throat, sputum) could improve the sensitivity. Repeat testing after a day or two could improve the sensitivity when there is a higher consequence or pre-test clinical likelihood. It’s all about risk and probability, so again no perfect measure.