Prof Adam Hill: Recognition and Management of the Deteriorating Patient with COVID-19

So onto our second talk. And again, a colleague and friend. This time, a local colleague and friend. And that’s professor Adam Hill, who’s a consultant respiratory physician in the Royal Infirmary here in Edinburgh, where he is clinically and academically working on translational research on respiratory infection. So, Adam, over to you. Many thanks. Thank you. I’d like to thank the college for asking me to speak. I’m going to talk about the care pre-critical care. And a lot of what I’ve got to say today is learning from international colleagues throughout the world and also experience from previous epidemics. I’m going to discuss the recognition of a deteriorating patient with pneumonia, it’s management, and ask any questions at the end.

The largest study was from our Chinese colleagues. And they split the number of cases into nonsevere pneumonia, severe pneumonia, defined by the American Thoracic Society, and very severe cases, where they needed intensive care, mechanical ventilation, or they passed away. Fever or admission varied between 36% and 43%, but was much more common throughout the admission, between 88% and 92%. Myalgia, arthralgia occurred between 9% and 17%. Headache, 12% to 15%. Nasal congestion, 3% to 5%. Sore throat, 9% to 14%. Breathlessness was much more common in very severe disease. 15% in nonsevere. 38% in severe and 54% in very severe. Cough occurred between 67% and 71% of cases. Sputum production, 30% to 35% cases. Hemoptysis was rare, 1% to 3%.

Sickness and vomiting, 5% to 7%. And diarrhoea, 4% to 6%. In that big very large case serious, 5% were admitted to intensive care unit, of which 2.3% had mechanical ventilation, and 1.4% percent passed away. So looking at the laboratory data was interesting. The white cell count didn’t really significantly change with severity of illness, ranging between 3.7% and 6.1%. Quite characteristic of viral pneumonia. The lymphocytes were reduced, and it seemed to get worse with more severe disease. 1% for nonsevere, 0.8% for severe, and 0.7% for very severe. We then looked at the percentage of patients that had a c-reactive protein, 10 milligrammes a litre or more. And again, this varied with illness severity.

56% in nonsevere, 82% in severe, and 91% in very severe. Interestingly, in a nonsevere group, septic shock, ARDS, acute kidney injury, and DIC were all very rare. These things were much more common with very severe disease. In very severe disease, 13.4% had septic shock, 40% had ARDS, 6% acute injury, 1.5% percent DRC, but no cases of rhabdomyolysis. And this is what you’d really expect. One of the interesting things is that there’s been a lot of data now on the use of CT scanning in viral pneumonia. And I think this has been really helpful. And this is really a move of previous literature, where we’ve depended on chest X-ray. As you’d expect, CT abnormalities was very frequent, from 84% to 95%.

And interestingly, we saw a lot of ground-glass change, between 53% and 61% of cases. Local patchy shadowing, 39% to 55%. Partial patchy shadowing was more common with very severe disease, 46% in nonsevere, 82% in severe, and 70% in very severe. Again, interstitial abnormalities worsened with worsening severity of disease. 12% in nonsevere disease and 26% in severe and very severe disease. There’s another lovely study done by the Chinese colleagues that looked at CT scanning in China published in The Lancet. And they commented the predominant pattern of abnormality was bilateral changes in 79%. Peripheral, 54%; ill-defined, 81%; ground-glass, 65%; and telling us there’s a predilection to the right lower lobe in 27%. We then looked at individuals who were asymptomatic at presentation.

And the predominant pattern at that stage was unilateral in 60%; multifocal, 32%; with ground-glass opacities in 93%. If they presented one to seven days after developed symptoms, lesions quickly evolved to come bilateral, 90%; diffused, 52%; and ground-glass opacity predominance, 81%. Thereafter, the prevalence of ground-glass opacities continued to decrease. 57% eight to 14 days, and 33% at 15 days after the developed symptoms. And interestingly, consolidation and mixed patterns became more frequent. 40% at eight to 14 days, and 53% at 15 days onwards. We then looked at data looking at the white cell count, lymphocytes count, c-reactive protein, and the number of segments affected with pneumonia.

In actual fact, they couldn’t find any difference when looking at white cell count or lymphocyte count. But interestingly, in the asymptomatic group, the c-reactive protein was very low, and the number of segments affected by pneumonia was very low. And this increased when the patient developed symptoms. There’s been other work with c-reactive protein in COVID-19 pneumonia, and it is recognised that the mean CRP is higher if you’re hypoxic compared to if you’re not hypoxic. The mean levels being 66 milligrammes per litre compared to 11 milligrammes per litre. Ruan and colleagues, in an intensive care study, compared patients that had died to those that survived.

Those that died had a median CRP around 125 milligrammes per litre compared to survivors with a median c-reactive proteins around 40 milligrammes per litre. So it’s quite nice showing that c-reactive protein does relate to severity of disease. This is what we’ve done previously with community acquired pneumonia, where you check the c-reactive protein at baseline and check the c-reactive protein at day three and four. We have shown that if the c-reactive protein at day three or four is less than 50% of baseline, you have a low incidence of complicated pneumonia, need for invasive ventilation or inotropic support, and a low 30 day mortality.

Whereas if you’re CRP is above 50% of baseline, you’ve got a higher incidence of complicated pneumonia, invasive ventilation or inotropic support, and a higher 30 day mortality. It is very interesting that you can actually use straightforward, clinical signs to assess prognosis. There was a lovely study published many years ago using the modified halms technique.

They said if your temperature is 37.2 degrees centigrade or less, your heart rate 100 beats per minute or less, your respiratory rate 24 breaths per minute or less, your systolic blood pressure 90 millimetres of mercury or more, and your oxygen saturation 90% or more– if you had all these features and you had a low CRP less than 30 milligrammes per litre, or a little procalcitonin less than 0.25 micrograms per litre, all patients survived and had a good clinical outcome. There’s also been international guidelines that tell us you should not routinely discharge patients with community acquired pneumonia if in the last 24 hours, they’ve had two or more of the following findings.

A temperature higher than 37.5 degrees centigrade, a respiratory rate 24 breaths per minute or more, a heart rate over 100 beats per minute, a systolic blood pressure 90 millimetres of mercury or less, oxygen saturation under 90% on room air, abnormal mental status, or inability to eat without assistance. So let us talk about the deteriorating patient. In my opinion, the best way to detect this is to look at the worsening NEWS score. And there’s an excellent explanation of NEWS score on the Royal College of Physicians website. I put it into three boxes on the causes of deterioration, and I take this from experience when we had our H1N1 outbreak previously.

So it could be due to worsening COVID-19 pneumonia, which is box one. And we would detect that not only from the worsening NEWS, but when you look at the full blood count, we may see worsening lymphopenia. The c-reactive protein may remain elevated or get higher. There may be renal dysfunction. There may be liver dysfunction. And it’s recognised in severe cases there could be a tropinin rise. In these cases, I’d recommend an up-to-date chest X-ray and consideration of a CT scan of a chest. Other than that, box two his pulmonary changes. And actually, there was a lot of patients from the previous epidemic that actually came in with an exacerbation of the airways disease.

They had an exacerbation of asthma, COPD, or bronchiectasis. And indeed, a small number even had a pneumothorax. And then there’s the other box. We have to think that it might not be a pulmonary complication. They may have a cardiac complication, such as arrhythmia, myocardial infarction, heart failure, or other vascular event. It’s been well recognised that actually, pneumonia and other lower respiratory tract infections increases your risk of a future vascular event. You’ve already heard earlier, you could have a pulmonary thromboembolism. And another thing we need to remember is it could just be an exacerbation of an underlying comorbid condition.

So I’m talking about the management outside the critical care unit, so I think it is an absolutely essential you assess the cause of decline. You look at the comorbidities. I would do an electrocardiogram and troponin level. As stated earlier, I would repeat the full blood count, urine electrolytes, liver function test, and c-reactive protein. And I would say under sputum sample for a routine culture, and blood cultures of they are pyrexial. And as stated in a previous slide, I would do an up-to-date X-ray and consider a CT scan of the chest.

If their airways disease are actually worse, which occurred in the last H1N1 epidemic, you would use nebulized bronchodilators, oral steroids, antibiotics if there’s increasing sputum volume and purulence, and we’d consider NIV if there is type 2 failure in COPD patients. So we have just completed a metanalysis recently looking at four randomised controlled trials involving 218 patients, where it compared NIV versus standard medical care of pneumonia. And NIV significantly reduced the rate of endotracheal intubation, reduced the ICU mortality rate, but had no significant effect on hospital mortality rate.

This study provides evidence supporting the use of NIV as a potential means of avoiding ET intubation and ICU mortality in patients with acute respiratory failure due to pneumonia in a critical care setting. But clearly, large international studies are needed before it can become rigidly recommended. So what can we do? In my opinion, we’ve got very severely ill group. I would use targeted antibiotics based on their microbiology. If they don’t oppose the microbiology, I would treat with co-amoxiclav or if they’re penicillin allergic, doxycycline. My view is that we should be entering as many of these into the recovery trial, which is comparing standard care versus the combination of lopinavir and ritonavir versus dexamethasone versus hydroxychloroquine.

There is a query for another arm, which is inhaled interferon beta 1 alpha, but it’s not currently available. There’s been recent British Thoracic Society guidelines in 2020 showing that they would prefer early use of invasive mechanical ventilation, and that’s preferred over NIV. NIV is, however, preferred over high flow nasal oxygen because high flow nasal oxygen needs high volumes of oxygen. And there’s also concerns regarding transmission risk rather than NIV. That’s my last slide and thank you for the talk. So Adam, thank you very much indeed for that. There’s a number– again, there’s more questions than I can certainly manage here.

But one that’s kind of recurring quite a lot that did come up on your slides was in what specific situations do you feel that a CT is kind of absolutely necessary for care? Some people were asking about whether it should be done routinely at the front door. Others were asking about at what time would you say there’s a definite indication that’s going to influence clinical management? Well, so the times I’ve used it in the last epidemic was when you’re really concerned about thromboembolic disease in particular. So if you’re concerned about pulmonary emboli, a CT pulmonary angiogram is helpful.

I think the first concern on the plain X-ray that you’re worried about some other pathology, like lung abscess or pneumothorax, you may want to consider a CT scan if there is an empyema. However, the case series have actually mainly shown for these viral ones, it’s more like an ARDS type picture. So I think the main indication, in my own head, would be worry about thromboembolic disease. And I think the rest we could probably managed with a plain chest X-ray. OK, now I’m going to take advantage of one of your other roles, and that’s as a dean here, and ask you a question that I know lots of our trainees are asking. And the question is this.

I’m an FY2 rotating into A and E in one of the very few trusts who are going to be rotating on the 1st of April. Do you have any advice on how I can best prepare for A and E?

That’s a very good question. There’s going to be– the COVID-19 epidemic has caused a major staff crisis because we have rates of up to 25% absenteeism. And so we’re having to redeploy a lot of people. There’s talk about getting medical students upgraded quicker. There’s a whole lot of strategies to try and increase the workforce. We in the Deanery have created a website that is designed to try and bring people up to speed with common respiratory conditions, but also helps– there’s a lot of information on it for nurses and allied health care professionals. There’s also a bit on pharmacy and psychology. In addition to that, there’ll be an extensive induction process on all the local health boards.

So I think whenever a trainee is going into an area, they should have had extensive induction, not only related to the COVID-19, but also locally as well. OK, and then you mentioned some criteria that really would suggest a patient shouldn’t be discharged from the front door. And clearly, we can quite quickly get to a situation when it will be difficult to go on admitting. And indeed, patients may not be sent up to the hospital at all. So a specific question that relates to that is any advice as to which patients can be safely discharged rapidly from the emergency room and which require further observation before a decision regarding admission?

Well, my view is that the international guidelines suggest that you shouldn’t be discharged if you’ve got any of the risk factors I identified in the previous slide. And to me, if you don’t have any of these risk factors, you may be able to safely manage at home. But I think if you do have these risk factors, then we’ve got a duty to admit these patients because they’re at real risk of further deterioration. OK, and then a question that’s been coming up a lot actually– it’s a very general one. It’s about the fact that some apparently young and apparently healthy, definitely young and apparently healthy, individuals are getting quite severe disease.

And it’s whether you’ve got any views on what’s going on there. Do you think they’re, for example, harbouring an unknown lung or other disease? Oh, I’d love to be able to answer that question if I’m honest. We’ve been seeing that with pneumonia for years. But actually, the reality is is that it is much more infrequent in the younger, but sometimes the younger have much more severe disease. And sometimes, they end up with acute lung injury. And it’s thought to be due to a high immune response. The reality is it is quite rare in reality compared to– it’s much more common to cause trouble in people that are older with comorbid disease.

I think we’d really love to look at innate immunity and see why these young people develop acute lung injury. I think there’s a lot of work needing done in that. And I know there are several groups that are looking at that. But I don’t really have an answer today. But there must be something in their innate immune system that’s leading to this. OK, Adam. Well it just remains for me to thank you again. I know that you’re a busy clinically, and you’re busy with your educational role as well. So thanks very much indeed on behalf of everybody for taking the time tonight to speak to us. Thank you very much. Pleasure.