BACK TO ALL
CTN Article Call Centre
10-02-2021
How to design a Covid-19 testing strategy
ARTICLES

We invite Dr David Shand, Director of Working Health Solutions to share his expertise in designing a Covid-19 testing strategy.  


Occupational Health specialists have been working hand in hand with Covid Testing Network to design and deliver testing plans to meet the needs of businesses small and large with simple to complex needs.  A well planned and executed programme can mean the difference between a business operating safely and productively within Covid-19 restrictions or having to close and send employees home.


Dr David Shand is a Consultant in Occupational Medicine with 30 experience and has worked across a broad variety of business sectors.  He offers us his views on how he, and Working Health Solutions, design a successful Covid-19 testing strategy for the workplace.


It is the purpose of this article to provide some basic guidance, and hopefully create a structured approach, to answering some of the key questions that are raised when planning a testing strategy, and implementing a workplace testing programme.


Dr. David Shand MB.ChB. MRCGP MFOM

Consultant in Occupational Medicine

Director - Working Health Solutions and Hampshire Health Ltd


Working


In current times, by far the most common enquiry we receive from employers, specifically in relation to the current pandemic, is with regard to testing.  We have been involved in Covid-19 testing since summer 2020, when access to testing, outside of the hospital diagnostic and public health arena, became available to occupational health (OH).


Initially, the sole mode of testing available, for workplace testing initiatives, were laboratory processed amplification tests, principally the RT-PCR test. This has been followed by the emergence of Point of Care Testing (POCT), particularly Lateral Flow Devices (LFD). LFD's have been heralded as an alternative, more affordable and accessible means of testing, with the benefit of rapid turnaround of results. Although these have some advantages over the more expensive laboratory-based tests, with their slower turnaround times, it is important to recognise that these testing methods are quite different. The principal difference is their relative sensitivities, which in turn affects how the results should be interpreted. Due to the lateral flows significantly lower sensitivity, in detecting the virus, their role, and their value has been repeatedly questioned[1].


As I will discuss in this article, I believe that both test methods have a valuable part to play in the fight to control the spread of SARS-COV-2. Understanding these differences is essential, and key to ensuring each is deployed effectively, and for the appropriate application. The debate I believe, should not be around whether one test is overall preferable to the other, because of its higher sensitivity in detecting the virus, but it should be around which type of test to use for any given scenario, who to test, and when, and how frequently to test.


Referring back to the enquires we typically receive; these can broadly be divided into three groups:


  1. A request for help and advice in response to an incident, for example an employee, or immediate family member, testing positive. I will refer to this scenario as Outbreak Management.
  2. Enquiries about how to approach implementing a screening programme, in the workplace, for asymptomatic employees. Asymptomatic Workforce Screening.
  3. A third distinct group is what I will refer to as boundary or entry testing. This is where there is a system in place to try and minimise the risk of importing infection into a controlled, or safety critical area, a workplace 'bubble'. This may be a national requirement, for the purposes of crossing international boundaries, or at the discretion of the employer, to enter a controlled work environment, for example boarding a ship, or oil platform. The vocational aspect of this I will refer to as The Workplace Bubble.

Each of the 3 scenarios described, presents a different problem, with different goals, which necessitate a different approach. For the time being, there is little published evidence, examining existing testing strategies, which can guide us in choosing an appropriate strategy, particularly with regard to asymptomatic screening[2]. It is simply too early in the pandemic for such data to have been collected and analysed.


There have been some studies reported, which have strongly supported the value of asymptomatic screening. For example, weekly screening of asymptomatic health-care workers, and other high-risk groups utilising the RT-PCR test method, was estimated to reduce their contribution to SARS-CoV-2 transmission by 23%[3]. Until more studies are published, which allows some comparison of the relative value of different approaches to screening, we are left to design our strategy based on first principles, utilising our knowledge of the methodology of testing, rather than having the benefit of comparisons of outcome. 


Which test method is most appropriate?


The first question, for each of the above scenarios, is which test method is appropriate for the purpose? This choice is in turn dependent, on the performance characteristics of the different modes of testing available. There remains ongoing debate around the value of the highly sensitive amplification tests, versus the more affordable, rapid tests which deliver a result within 20-30 mins, but which are less sensitive in terms of the virus level they can detect, the limit of detection (LOD).


The RT-PCR test method, the gold standard amplification test, which involves replication of viral RNA, prior to the detection component of the process, will detect as few as 100-200 viral copies per ml. Conversely, the most sensitive of the LFD's, has a lower limit of detection in the region of 2-300,000 viral copies per ml, many requiring viral loads of greater than 106 viral copies/ml, and are therefore 1000x less sensitive than the amplification tests.[4].


It is this issue of sensitivity, alongside questions around usage by untrained persons, which has led to questions over the value of LFD's, and whether they have a place in mass screening - as in the Liverpool initiative towards the end of 2020. This raises the question - Do LFD's have a role to play in a workplace testing strategy? Or, should we be focussing our attention on the more sensitive amplification tests.


To explore this dilemma further, it is important to understand more about the virus dynamics following infection, and how this is reflected in the results of the two different test types. The figure below, presents a graphical illustration of viral load over time changes, following infection, and how this relates to the different LOD's for the two test methods we are considering. The graph is intended as illustrative, and the time scale vs. viral load figures displayed are approximate.


Graph


From the graph, it can be seen that the high sensitivity test, will likely detect the presence of the virus somewhere around 24-36 hours before the low sensitivity test, and will continue to detect the virus for perhaps up to 10 days, after the low sensitivity test is no longer able to detect the virus.


The purpose of this illustration is not to demonstrate one test as performing better than the other. The important point is the different characteristics relevant to the tests' interpretation, and to inform which test to choose, in each of the different testing scenarios described above.


The principal difference is that the RT-PCR is a much more sensitive tool to detect an individual who has been infected by the virus, showing a positive result around 24 hours sooner that the POCT/LFD, and continuing to do so for around 10 days after the viral level has fallen below POCT devices LOD. There are therefore pros and cons, with the LFD being more likely to return a false negative in the early period of infection, but the PCR potentially returning what is effectively a "false positive", in terms of the intended purpose of the test, when there is a long post-infectivity period, when the PCR test continues to return a positive result. In the case of an asymptomatic screening programme, this demonstrates the importance of implementing frequent testing. A single "snap-shot" test using the PCR will perhaps pick up more individuals who are in the post-infectivity stage of infection, as this is a longer period, due to the long tail of declining viral load, after the infectious stage has passed. Therefore, where the intention is to undertake regular testing, regardless of which test mode is chosen, it is important to understand the need for reasonably frequent testing, to create a time-line, and give context to any positive test results.


Although the LFD may be slightly behind in returning a positive result, in an infected individual following infection, the actual period of infectivity, where the PCR test is positive, and the LFD remains negative, is relatively short, perhaps 24 hours. Conversely, the LFD will only remain positive for a shorter period, more aligned with the period of infectivity, and the post-infectious period, where the test continues to return a "false-positive" with regard to the purpose of testing, is relatively short, around 5 days, as opposed to more than 10 days with the PCR test. In practice we also see a high number of inconclusive PCR tests in individuals who have had a positive test up to 4 weeks previously. An inconclusive test is encountered sometimes in the very early stage of infection, where the test is detecting some viral RNA, but more often in the long post-infectious period of the virus infection.


The main point is that where the aim is to identify infective individuals in the workforce, to reduce spread within the workforce, the more data that is available, the more accurate the interpretation of the test result is going to be. This will reduce the unnecessary need to isolate for those that fall into the "false positive" post-infective window, and reduce the impact of a "false negative" where a test is undertaken in the early stage of infection below the tests LOD. As explained above, a "false positive", in this context is an individual who tests positive, but who no longer presents a significant risk of infection to colleagues, but is then excluded from the workforce unnecessarily.


Both types of test therefore, have a role to play in a workplace Covid-19 test strategy, but their performance characteristics are different, and it is essential to understand this difference, before deciding where and when to use the different test types available.


With the above knowledge we can move on planning how best to approach testing for each of our three different scenarios.


1) Outbreak Management

Where there is concern about possible employee contact with a positive case clearly, where "contact" falls within the Gov.uk definition of 'contact', as set out in the guidance on self-isolation, there is a defined protocol which must be followed. Testing does not negate this need to follow the guidance on self-isolation. However there are many scenarios where "contact" with a positive case does not fall within the Gov.uk definition of contact, but there are still grounds for concern and a potential increase in risk. This may perhaps arise from a recognised case of infection in the work group, family, or social circle of an employee, pointing to increased incidence of infection in the immediate social and work groups.


Here the first step is to temporarily isolate the employee for whom there is concern, over increased risk of contact, separating them from any further contact with colleagues or customers, and presuming the individual in question to be a potential infection hazard. The next step is to review the employee's work environment, and map out which work colleagues are in the immediate secondary contact group of the workforce, for possible onward transmission.  This is then the immediate focus of attention for any containment measures. In this scenario, the potential risk of transmission in the workplace, is best controlled utilising the PCR test (or equivalent amplification test), for both the employee in question, and the immediate secondary group.  Here we have a target group in which we wish to diagnose the presence of absence of infection accurately, at a point in time where eradication of the risk is the goal. The timing of testing should be based on a timeline of any potential contacts with case zero, ensuring a test is not conducted too soon, based on our knowledge of the incubation period viral levels, relative to the PCR's established LOD.


As to whether further broader testing is required, and how quickly, very much depends on the working environment and how practical it has been to implement social distancing guidance in the work environment. The need for broader testing will obviously be informed by the results of initial testing, which ideally should be undertaken within a maximum of 24 hours.


One final point is that, where testing is being undertaken as a follow up to a previously positive test, I would propose that the most appropriate test to conduct is an LFD test, after the appropriate 10 days of self-isolation has elapsed. An RT-PCR test conducted in this scenario, may continue to return a positive result, long after the individual is no longer likely to present an infection risk.


2) Asymptomatic Screening

For the effective surveillance of the asymptomatic workforce, the desired outcomes are different. Here the aim is to reduce the risk of transmission in the workforce whilst, not disrupting operations through unnecessarily excluding employees who do not present a risk. It is here that an understanding of the test characteristics illustrated in the above figure, are perhaps particularly important.


For asymptomatic screening this is where the less-sensitive LFD's potentially have a role to play. Here, detection of potential infection in the workforce, to reduce transmission within the workforce, is the priority, while also avoiding unnecessary infectivity "false positives" for those who may test positive, but who are already recovering from the virus and no longer infectious.


In this scenario, I would propose that frequent testing is the key goal, both to establish a time-line of results to aid interpretation, but also to limit the impact of false negatives where an infected individual is pre-infectious but still below the LOD for the test utilised. It is here that the affordability, and availability of LFD's perhaps has a role, as an alternative to the RT-PCR test. As long as the relative test characteristics are taken into account when planning the strategy, ultimately the decision on which mode of testing is perhaps more dependent on the relative cost, and whether there is access to a PCR test provider who can deliver rapid results. With either test, testing needs to be frequent, probably 3 times per week minimum, if utilising LFD's, or twice a week with PCR tests, but results need to be available quickly, ideally within 24 hours. In my opinion the value of testing, undertaken with a frequency of less than twice per week, is questionable.


Therefore, where there is limited access to affordable PCR testing with a quick turnaround of results, this, in my opinion, is where the rapid, LFD devices have a potential role to play. Testing however needs to be frequent, more so than when utilising RT-PCR testing. The relative costs of increased frequency of testing and the additional requirement for in-house clinical support, for POCT testing, makes the basic cost of the test only one factor which needs to be considered when choosing the most appropriate strategy, to meet the aims of the testing programme.


3) The Workplace Bubble

Finally, where the aim is to protect a cohort of workers who are known to be virus free, or must remain virus free to maintain operation effectiveness, as per the example of a ships' crew, the demands of the testing programme are again different.  Here, if the workforce becomes compromised, this may present significant challenges. Quarantining and segregation, within a confined area, may not be practical, or where there are limited numbers of key staff, the loss of sections of the workforce may compromise the safe running of the operation, which cannot easily be remedied.


Here, in my opinion, the most appropriate test, is almost certainly the more sensitive amplification test. The primary aim is to avoid infection entering the workforce, and not to miss a single case of infection, rather than trying to lower the prevalence and limit transmission within the workforce. Here the disruption of the false positive, post-infection result may still be disruptive, but less of an issue, particularly where testing is undertaken in conjunction with a period of quarantining. Quarantine, on entry into the workforce/working environment, as an adjunct to testing, and with further testing being conducted before release into the wider working environment, should make elimination of infection from the bubble something which can be maintained from a practical perspective. A reduced rotation of staff will also be an important feature of such a strategy, if it is to remain practical and effective. After initial RT-PCR testing, prior to release from quarantine, there is certainly an argument for follow up to be undertaken with LFD tests, to be sufficient.


Finally, on the need for "boundary" testing, there is the specific requirement in relation to business travel, where there is a need to cross international borders. Here decisions on which test are dictated by the entry requirements for the destination country, or any country to be transited through. The majority of countries stipulate the requirement for an RT-PCR test, most commonly within a 72 hour window prior to departure.


Key Messages:

  • Amplification tests (PCR) are more sensitive than LFD's, but it is understanding how to interpret the result, which is the important point, when choosing which type of test, to use in any given scenario.
  • The PCR test can be of value in any scenario, whereas the LFD has a limited range of applications, principally larger scale asymptomatic screening.
  • When planning a testing programme, first define the goals of testing. The goals will depend on the purpose of testing, as per the three typical scenarios described. Asymptomatic screening presents different demands of the test regimen, compared with targeted outbreak management, and protection of a controlled bubble.
  • For asymptomatic screening, the actual choice of test will probably depend on logistics and cost, and whether there is access to a reliable, fast turnaround PCR service, alongside careful consideration of the relative costs. The true costs of POCT, versus laboratory PCR, may not be immediately apparent as the clinical set up for POCT is more complex, if it is to be conducted properly.
  • Ensure testing is undertaken with sufficient frequency to facilitate accurate interpretation, thus minimising the impact of false negatives, and creating a timeline to further aid the interpretation, and management of positive results.

Ultimately, it is essential to have access to appropriate expertise and advice, for any testing regimen, planned or implemented. It is the overall performance of the testing programme, which is the important measure of success, not the number of positive or negative results obtained. The aim is to keep the organisation running efficiently and smoothly, protecting the workforce, whilst avoiding disruption to work activities. Disruption can both arise from an outbreak in the workforce, or conversely, unnecessary suspension of work activities through misinterpretation of test results.


[1] Covid-19: Safety of lateral flow tests questioned after they are found to miss half of cases - BMJ 2020;371:m4744

[2] Covid-19: Innova lateral flow test is not fit for "test and release" strategy, say experts - BMJ 2020;371:m4469

[3] Schuetz AN, et al Alexander J. McAdam, Editor; When Should Asymptomatic Persons Be Tested for COVID-19? 10.1128/JCM.02563-20

[4] Grassly NC, Pons-Salort M, Parker EPK, White PJ, Ferguson FM, Comparison of molecular testing strategies for COVID-19 control: a mathematical modelling study Lancet Infect Dis 2020; 20: 1381??"89

[5] Nguyen NNT et al. Review Article: Development of Diagnostic Tests for Detection of SARS-CoV-2 Diagnostics Volume 10 Issue 11 10.3390/diagnostics10110905

[6] Corman VM et al Comparison of seven commercial SARS-CoV-2 rapid Point-of-Care Antigen tests doi: https://doi.org/10.1101/2020.11.12.20230292

Want a quote?

210911661 0333 050 9582

 

522322432 contact@covid19-testing.org

 

Vector-2 Vector Vector-1