As the pandemic races towards an unenviable two-year milestone, the Otago Global Health Institute’s Covid-19 Masterclass Series is bringing together a network of experts to discuss key Covid-19 topics. We’ll be running a piece daily until December 5.
James Ussher and Philip Hill assess our current testing options and what’s available worldwide – and make some suggestions on what avenues we should be taking
Diagnostic testing is crucial in the response to Covid-19, but adequate testing capacity and fast enough turnaround times for results are major challenges globally.
People who have symptoms of Covid-19 need to be tested as soon as possible. If they seek a test and the result is available quickly, public health teams have a chance to find their contacts and stop onward transmission of SARS-CoV-2.
Covid-19 is highly infectious even for up to two days before an infected person becomes symptomatic, requiring surveillance testing of asymptomatic people in particular situations.
As the pandemic continues, many businesses and organisations will be developing testing strategies. These should be developed with health experts, taking into account the following issues:
How do we estimate the performance of tests for Covid-19?
The properties of the tests inform how we use them. We estimate their sensitivity (percentage of those who have the virus that test positive) and the specificity (percentage of those without the virus that test negative). When a test has low specificity, it will give too many false positive results.
We also estimate predictive values – positive predictive value tells you what percentage of those with a positive test have the virus; negative predictive value tells you what percentage of those with a negative test do not have the virus.
Predictive values change with the prevalence of infection in a population. For example, if a test is 80 percent sensitive and 98 percent specific (such as the current rapid antigen tests described below), at 1 percent community prevalence of Covid-19, the positive predictive value is less than 30 percent.
With SARS-CoV-2, tests often pick up both viable and non-viable virus, complicating the interpretation of results.
What tests have been developed?
Tests for Covid-19 that require a laboratory tend to use high-throughput polymerase chain reaction (PCR) machines and have high sensitivity (over 90 percent) and extremely high specificity (almost 100 percent). Two key sampling methods for these tests are the nasopharyngeal swab and the saliva sample.
The sensitivity of the best saliva tests has reached similar levels to nasopharyngeal testing. However, transport to the laboratory, processing of multiple samples, batch testing, and reporting back to the patient, can often combine to a wait of over 24 hours for a result.
There are machines that perform PCR testing after a sample is placed in a cartridge, with a result available in less than an hour, but they can only handle a few cartridges at a time and cartridge availability is severely restricted globally.
A further simplification of these methods to detect viral RNA is loop-mediated isothermal amplification (LAMP), which does not require a laboratory and provides a result in less than half an hour. Few studies have assessed the performance of LAMP tests so far and they appear to be less sensitive than PCR tests.
Rapid turnaround antigen tests (RATs) for Covid-19 have become increasingly available. RATs detect the presence of the ‘spike protein’ of the virus. They tend to use nasal swab samples, with few large studies so far assessing saliva as a sample.
RATs vary considerably with respect to their performance properties and turnaround time and can generally only process one sample at a time. There are differences in sensitivity in symptomatic (averaging around 65 per cent) versus asymptomatic (averaging around 55 percent) individuals.
Reported sensitivity estimates are often inflated through calculating percentage positive agreement with PCR tests. Sensitivity also varies at different stages of illness, averaging around 70 percent in the first week versus around 45 percent in the second week. Specificity is often very high (approaching 100 percent) with many of the better-quality antigen tests.
Some of those with higher sensitivity (approaching 90 percent) may have a small drop in specificity down to as low as 97 percent. At low infection prevalence, this can lead to the majority of positive tests being false positive.
Other diagnostic tests under development use technologies such as microfluidics, holographic microscopes to look at nano-scale structures, and gas chromatography. Breath tests and trained sniffer dogs hold some promise.
Waste-water sampling PCR-based testing is hampered by the variation in the number of cases that shed virus in their stool, dilution of virus, and the inability to distinguish viable and non-viable virus.
What testing strategies should be used?
While many countries have struggled to establish large enough testing capacities to respond effectively to outbreaks, lack of strategic use of large testing capacities has led to serious inefficiencies in other countries and subsequent failure to contain outbreaks.
Symptomatic cases are much more infectious than those who never become symptomatic and testing of asymptomatic people can lead to the diversion of testing away from symptomatic individuals.
Furthermore, since never-symptomatic people probably play a minor role in overall transmission, large numbers of positive results from such individuals may place unnecessary work on the health response.
We can think of at least five different groups of people in relation to testing:
- Symptomatic people in the community should be able to get a test. Both nasopharyngeal swab and saliva PCR testing should be available to maximise uptake.
- All those entering the country should be subject to a testing plan, with a mix of pre-departure, arrival and post-arrival testing adjusted with expert advice over time.
- Symptomatic and asymptomatic individuals being admitted to at-risk situations (e.g. hospitals) should be subject to a testing plan to protect vulnerable people. It may not be feasible to screen visitors or even all emergency department attendees.
- Symptomatic and asymptomatic individuals attending large events that pose super-spreading risk are candidates for self-testing using RATs prior to attendance and requirements may be influenced by vaccination status.
- Outbreaks will vary in relation to the strategic use of tests. Testing of asymptomatic casual contacts may be avoided to maximise the value of limited laboratory resources.
*Conflict of Interest statement: Associate Professor James Ussher is employed by the Southern Community Laboratories in Dunedin, which conducts Covid-19 testing. Professor Philip Hill has been on four ministerial advisory committees for COVID-19.