Mapping the outcomes of covid-19 testing reveals the best opportunities for system improvement

To be sure that testing programmes are providing public benefit, we need empirical data on their real world consequences, say Angela Raffle, Sian Taylor-Phillips, and Alice Sitch

Efficient and effective nationwide testing systems for the current and any future pandemics are now an essential prerequisite for health and prosperity. 

The priority now is to look where risk from covid-19 is highest, identifying and isolating/quarantining symptomatic cases, their contacts, and people travelling from overseas, and managing outbreaks. Yet recent investigations show low effectiveness of the NHS Test and Trace service and quality failings in “lighthouse labs.” Rather than solving these problems, the UK government is focusing instead on rolling out non-systematic, poor quality, society wide repeat testing for symptomless people. This goes against the advice, experience, and evidence provided by scientific experts.

When national independent scientific advisory groups, like the UK National Screening Committee and international equivalents, assemble evidence for mass testing programmes to inform policy makers they compare, to a baseline of “usual care,” a range of options for who to test, how to invite participants, what test(s) to use, how to deliver them, to what quality, accompanied by what information and with what interventions. Flowcharts for the testing/intervention pathway are often used to map potential outcomes. 

A common barrier to shared understanding of asymptomatic testing is the stumbling block of case definition. In symptomless people, a “case” can only be defined by what is measurable at the moment of testing. What an individual result signifies, and what will happen in the future, can never be certain, which is why the real life measurement of outcomes for a testing programme is so important. With asymptomatic testing for SARS-CoV-2, only some of the people who turn out to be “true positives” (verified by expert sampling and a standardised assay) will have gone on to transmit infection. 

We have developed an interactive flowchart showing in simplified form two current components of SARS-CoV-2 testing: symptomatic testing and mass asymptomatic testing. Other elements exist too, but these are not shown. We wanted to describe the key dependencies of SARS-CoV-2 testing and highlight opportunities for improvement, rather than to model the complexities of infection transmission. Readers can input their own evidence, projecting the impact it will have on outcomes. Populating the flowchart requires evidence from diverse, and often uncertain, sources. 

If public benefit from a testing programme is to be assured, then empirical data on its real life consequences are essential. Yet no data currently exist for the government’s home self-testing for asymptomatic people, although some evaluation in schoolchildren may be forthcoming. Random allocation by area, or by timing (stepped wedge trial), could have provided evidence as to whether mass home testing brings more benefit than harm. Yet the government seems to have no plans for evaluating this multi-billion pound venture. 

So far asymptomatic testing has used rapid lateral flow tests that can give results in 30 minutes. In a major community testing pilot in Liverpool, 25% of people participated, and this was skewed towards those at lowest risk, as generally occurs with screening. An evaluation found no direct evidence of the pilot having had an impact on hospital admissions and deaths. Important conclusions were: that clear information for participants is needed; that non-targeted community testing is unlikely to be cost effective; and that uniform high quality delivery on such a scale may not be feasible. The yield of cases in Liverpool was higher than expected, suggesting possible “leakage” of symptomatic people who opted for the more convenient (and less sensitive) local rapid test. Going forward, if poorly performed home self-tests substitute for quality assured sampling/testing in people with symptoms then the outcome could be an increase in transmission. 

The UK government’s testing service for symptomatic cases is improving but there is still much to do. Test and trace aims to detect symptomatic SARS-CoV-2, which, according to estimates, probably accounts for around 80% of transmission. Non-classic symptoms are common yet inexplicably only three symptoms currently qualify for eligibility. Transmission can occur before or without symptoms manifesting, which is why the immediate tracing, testing, and isolation of contacts is so critical. Yet only around 50% of people can correctly name the symptoms, and up to 80% of people with symptoms report not getting tested. The estimated SARS-CoV-2 prevalence from REACT data for England has been consistently three to five times higher than cases detected by Test and Trace, suggesting that attendance by those with symptoms is indeed low. 

Even among the cases that are detected by Test and Trace, less than half are able to adhere to advised isolation, largely due to economic pressures and carer duties. In an average week, around 85% of people who test positive give details of at least one contact, of whom 90% are traced and of those as few as 11% may isolate. Undoubtedly, other countries are doing better and we need to learn from their approaches. Relatively modest improvements in adherence with symptomatic testing and with successful isolation may, because they are focused at the point of most impact, have equivalent impact to an intervention that tests an entire community. 

It is essential to transform existing piecemeal testing into a coherent, trusted, easy to navigate, and evidence based programme aimed at protecting the entire population through high quality pathways and supported interventions. Unless people with confirmed infection can isolate, even perfect testing will have no impact. The focus must be where yield of infectious cases will be highest. Properly resourced public health and primary care teams are best placed to ensure this. 

Angela Raffle, consultant in public health and honorary senior lecturer, Bristol Medical School Population Health Sciences, University of Bristol, UK. Twitter @angelaraffle

Sian Taylor-Phillips, professor of population health, Department of Population Evidence and Technologies, Warwick Medical School .

Alice Sitch, senior lecturer in biostatistics, Institute of Applied Health Research, University of Birmingham and NIHR Birmingham Biomedical Research Centre, University Hospitals Birmingham NHS Foundation Trust, and University of Birmingham UK. Twitter @AliceSitch

Competing interests:

Angela Raffle: none to declare.

Sian Taylor-Phillips: none to declare.

Alice Sitch: Alice Sitch’s work is supported by the Birmingham Biomedical Research Centre. This paper and the interactive flowchart it links to presents independent research supported by the NIHR Birmingham Biomedical Research Centre at the University Hospitals Birmingham NHS Foundation Trust and the University of Birmingham.

The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR, or the Department of Health and Social Care.