COVID-19: The Missing Middle

On Wednesday, I Tweeted, “I’ve just been alerted to this statement by @WGHealthandCare @HywelDdaHB @PublicHealthW My head is metaphorically in my hands! This priority list is not designed to find the true positives. I’ll try and blog later . . . .” and linked to: https://gov.wales/written-statement-prioritisation-covid-19-testing

There has been considerable media attention on the pressure being experienced by our testing services, both the demands for appointments for sampling and the laboratory capacity. Working in the three laboratories in the Hywel Dda University Health Board, I can confirm that the local demands have increased week by week, since the laboratories acquired testing technologies beginning in April/May but equally for the larger laboratories in Public Health Wales and the Lighthouse Laboratories.

The priority list suggested by the Welsh Government statement above misunderstands the role of the tests we have at our disposal and will contribute to the problems we are experiencing. Let me explain by looking at the priorities. Let me however be clear from the outset, I do understand why what is listed are felt to be important but hopefully I can explain why I would advocate a different approach.

“Priority one will be to support NHS clinical care and focus on hospital patients, including all admissions, so that clinical judgements can be made to ensure the best care for these individuals.”

Within hospital, we always prioritise symptomatic patients. However, the recurring theme in these responses will be the use of COVID testing for asymptomatic individuals. This is screening and the tests we use for diagnostic purposes are not suitable for screening.

The laboratory tests we are using are designed to confirm or reject a clinical diagnosis. When a clinician arrives at a potential diagnosis, s/he has excluded many who would test negative, thereby narrowing those in the population to test, increasing the prevalence in a group that is to be tested. If however we test the general population with no symptoms, then we are testing at the much lower prevalence, essentially that of the geographical population where the patient lives.

Currently as I write, Llanelli is in local lockdown with a prevalence of 151positive results/100,000 population or 0.151%. Laboratory tests can never be 100% sensitive or specific. If the tests have a sensitivity and specificity of 98% then for every 1,000 admissions from Llanelli with a prevalence of 0.151%, we will get 21 positive results and will therefore respond by putting 21 people in isolation or in COVID wards. However, only 1 of those 21 will be a true positive, with 20 false positives. For a fuller review of this, a copy of the paper I wrote for colleagues locally is available to download. This is a Word document with embedded Excel spreadsheet. While the text can be read via a browser, to enable the spreadsheet, it will be necessary to download.

The figures will vary for each region, depending on the local prevalence. However, currently even the higher rates in some parts of Wales are still at a low prevalence such that there will be more false positives than real positives when the test is used as a screen, potentially meaning that inappropriate clinical decisions may be made for these false positives, which may influence the appropriateness of the care provided.

“Priority two will be to protect those in care homes. We know that people living in our care homes are particularly vulnerable to COVID. Based on SAGE and TAG advice, we will continue to test staff, with or without symptoms, fortnightly or, where local prevalence is significant, weekly.  We will also test all new admissions and, in the event of an outbreak with a home, all residents and staff will be tested.”

We need to move away from this now that we are at a low prevalence. This is a huge drain on the testing resources and also falls into the issue of asymptomatic testing as discussed above. At this stage, it would be better to primarily ask nursing homes to be on the lookout for any sniff of a respiratory outbreak and go in quickly to test any symptomatic patients for COVID and if negative, test a small sample against other viruses. While in HDUHB we are still doing the fortnightly testing, we have equally been in rapidly when a home has identified symptomatic patients. In two such recent incidents, we found that the rhinovirus was responsible.

“Priority three will be the testing of NHS staff, including GPs and pharmacists where possible, as protecting our NHS has been at the heart of our COVID response and is a clear priority in our Testing Strategy. We will continue to test NHS staff with symptoms but will move test asymptomatic staff in outbreaks and in areas of higher prevalence.”

I entirely concur with testing symptomatic staff. However, here again there is the suggestion of moving to asymptomatic testing, with the risk of detecting more false positives than true positives and thereby inappropriately inviting staff and close contacts to go into self-isolation. This will have an adverse effect on staffing numbers and make responding to increases in real cases increasingly difficult.

“Priority four is targeted testing to support management of outbreaks and surveillance studies which will help to manage outbreaks in high-risk environments such as closed residential settings or higher-risk workplaces, where the risks of the virus spreading and chances of finding more positive cases is high. Surveillance testing refers to the population-wide Office for National Statistics study, essential trials for new potential vaccines, and studies of particular at-risk populations.”

Prioritising outbreaks makes sense as these will be symptomatic individuals. Surveillance testing is important but it should not be prioritised over symptomatic testing. Thus this priority mixes two issues that should be separated.

“Priority five is to prioritise testing for symptomatic staff working in education or childcare settings where is it needed to keep the settings open. We are continuing to improve the testing system to ensure staff can get priority access when they have symptoms. Those who test negative can return to work, ensuring our education and childcare settings can remain open.”

This is again symptomatic testing and therefore should have a much higher priority. This statement is I assume just talking about COVID testing and the result. The statement, “Those who test negative can return to work,” is questionable and is not consistent with good infection prevention and control practice. I suspect this is trying to tease out the debate of COVID versus a simple head cold. However, children or staff with ongoing respiratory symptoms, while COVID negative, are potentially at risk of spreading and causing an outbreak associated with other respiratory viruses. As we move into the winter season, we anticipate we will have the tools available to test for a wider range of viruses, including COVID. We must be ready to respond appropriately, depending on the nature of the viruses involved.

“Priority six is testing all symptomatic individuals irrespective of local prevalence although it is important that individuals only book a test if they have key symptoms (a high temperature, a new continuous cough and a loss of taste or smell).   People should self-isolate if they have any of these symptoms, or if you are asked to by a contact tracer.  By following this advice, we will be able ensure that the resources that we have available are being used as effectively as possible.”

And I guess in my view, this is one of the most important group of individuals to test – new symptomatic individuals in the community. This is where the emphasis should be placed, if we are to keep on top of new community outbreaks. It also plays into the title of this post, “The Missing Middle,” because of the persistent advice around the coronavirus from all parts of the UK and devolved governments. Typical is the advice being offered by the Welsh Government on their Facebook page at: https://www.facebook.com/welshgovernment/photos/a.287087867998975/4344276955613359/ where the advice is:

Only book a coronavirus test if you have these symptoms.

🔺 New continuous cough
🔺 High temperature
🔺 Loss of taste and/or smell.

I am now tending to refer to these as “The COVID 3,” because they are all the public seem to be bombarded with and as a result, I fear they may be part of the problem in relation the the current rise in cases, which in turn is being reinforced by the difficulties we now face with increased demand for testing. We are testing more people who are symptomatic with these three symptoms but also a lot of people who are asymptomatic: testing for negativity pre-procedure as our health service tries to get back to delivering services that were put on hold during lockdown, testing for negativity before a care home resident is discharged back to their care home after an admission to hospital, testing for negativity of care home staff weekly and residents every two weeks and testing asymptomatic people who are part of a COVID-19 incident.

This testing can have a place but we would be better placed if we were to review quite what we are seeking to achieve, review the processes and perhaps release testing resource for “The Missing Middle.” The COVID 3 are not the only symptoms associated with this new virus.

Take a look at the CDC website at https://www.cdc.gov/coronavirus/2019-ncov/symptoms-testing/symptoms.html the list I copied below was as of
11th October 2020 when I last looked and says:
“People with COVID-19 have had a wide range of symptoms reported – ranging from mild symptoms to severe illness. Symptoms may appear 2-14 days after exposure to the virus. People with these symptoms may have COVID-19:

• Fever or chills
• Cough
• Shortness of breath or difficulty breathing
• Fatigue
• Muscle or body aches
• Headache
• New loss of taste or smell
• Sore throat
• Congestion or runny nose
• Nausea or vomiting
• Diarrhea

This list does not include all possible symptoms. CDC will continue to update this list as we learn more about COVID-19”

Indeed, a recent BMJ article discussed how common gastrointestinal symptoms in children are a presenting feature of this new virus. One of the conclusions suggests adding such symptoms to the list i.e. go from COVID 3 to a new COVID 4. To my mind, this misses the point. The CDC list above demonstrates one of the issues I have been seeking to explain to colleagues (and anyone else who will listen) that all respiratory viruses are capable of a range of symptoms and signs from asymptomatic through to the very severe or even death and to limit how the virus is behaving to 3 or 4 symptoms before access to a test is granted, will mean we miss the start of a local outbreak until such time as somebody with one of the COVID 3 manages to get a test. By then, spread will already be underway. As well as the asymptomatic and those with the COVID 3, there will be a lot of people in a middle group with minor symptoms, “The Missing Middle,” who we should also be testing. This would be achievable by prioritising all symptomatic individuals.

There is most definitely an increase in detected positive cases currently. It is also evident from other countries and the UK graphs that the current trend in positive test results is higher than in March and April but to suggest the virus has changed somehow and is mainly affecting the young misses the point. Our testing strategy has changed. In March and April we were only essentially testing people admitted to hospital with respiratory symptoms. Estimates are that 80% of people only ever get a mild illness not requiring admission and therefore these 80% were not tested.

As the BMJ article highlights, the Covid-19 Symptom Study app, had already identified such gastrointestinal symptoms. The study associated with the App was successful in presenting evidence that contributed to lack of taste or smell being added as a new finding as one of the cardinal symptoms of the COVID 3. A colleague on Twitter suggested this was the only unique symptom associated with COVID-19 when he was trying to make the case for not ignoring other symptoms. I had my doubts about the uniqueness; I think this is a reflection of the power of social media and our connectedness. In an internal health meeting I suggested other respiratory viruses this winter will probably be found to also cause a loss of smell or taste, now we were universally alerted to the possibility. At which point a senior nurse in the meeting declared she lost her sense of smell for a period when she had a bad attack of the flu in the past. Because the numbers are likely to vary from virus to virus, as all other symptoms vary from virus to virus, without the power of the internet in the past, we simply had not made the connection. As we go into the winter season, we will be testing for a wider range of viruses and we will be able to explore just how often this new symptom appears with other viruses.

As a result, we are not comparing like with like – the initial peak was an underestimate of symptomatic people, so take any of those early numbers, divide by 20 and multiply by 100 and you get an approximation of how high the symptomatic cases were. But not only that, as I discussed in my first post on this issue in April, below these visible cases are the invisible asymptomatic cases and in the current testing dispensation, we are also now testing for there. Estimates of asymptomatic cases vary from 20% to 80% depending on the populations studied but to your current estimate, you could add another 50%.

This change in testing numbers is best revealed by the number of deaths. Mercifully, these remain low but are increasing. This is to be expected once we released lock-down and as discussed in “COVID-19: Where next” albeit specifically discussing Ceredigion, our biggest risk will probably be with hospital and nursing/residential home outbreaks and we will need to be on our toes to spot quickly and isolate cases.

My Priority List

Priority 1: Symptomatic cases in any group, with a wider recognition that symptoms can be any associated with the typical responses associated with the respiratory system. I cannot over-emphasise how important it is to include all new cases appearing in the community if we are to get on top of the pandemic.

Priority 2: Outbreak support. When managing an outbreak in normal times, we do not normally test everyone. In a closed community (care home, factory, school, ward, household) once we have had three positive results all associated with the same virus, other individuals with the same symptoms can be defined by a case definition and actions taken to limit spread agreed appropriately. However, where there are staff caring for the outbreak population, asymptomatic testing may be helpful in managing the staff because in the closed community, the prevalence will be higher than the general population and such screening is more likely to be detecting true positive asymptomatic or pre-symptomatic individuals.

Priority 3: Any asymptomatic screening. However, as the various government departments seem to misunderstand the use of the current tests for screening, they should perhaps ask the UK National Screening Committee to review the use of the current tests in asymptomatic populations.