Alex Nowbar reviews the latest research from the top medical journals.
Annals of Internal Medicine
Have you heard the one about covid-19 and the Canadian model?
Not the fashion variety, I’m afraid. This research letter describes one of the computer variety. Giannakeas et al made an interactive web app where you input the number of acute care beds, critical care beds, and mechanical ventilators available for patients with covid-19 in your local health care system and it outputs the maximum number of covid-19 cases that could be cared for in that system. The tool uses data on the local distribution of covid-19 cases by age and the distribution of severity to estimate lengths of stay. This output can highlight the need for more of a particular resource, e.g. ventilators, in that healthcare system. Neat. It can only be as good as the data that goes in, which at the moment is based on figures from Canada, US, Italy and China, but the authors have committed to updating the tool. For example, the number of cases of covid-19 will increase as testing increases but many of these will be people who will not require hospitalisation.
How do you know if you have covid-19?
Option A: You have a fever and/or cough
Option B: You got a positive PCR result (i.e. there was viral RNA on a swab of your upper airways)
Option C: Someone you’re in close contact with has had A or B and you feel a bit peaky
Option D: You sneezed once yesterday and once again today and the person behind you in the queue at Lidl last week was only 1.2m away
Option E: Your flatmate threw you in the river and you floated
Unfortunately all of the above are compatible with having COVID-19, and yet are also all compatible with not having COVID-19 (although it’s very unlikely with option B, positive PCR). Go figure. This question is probably one of the most important of our time, especially if you deal with vulnerable people at work or in your family. Chow et al present the findings of interviews with 48 healthcare professionals in King County in the US state of Washington who tested positive for SARS-CoV-2 following symptoms. Half of them presented with cough. Fever was the next most common initial symptom occurring in 42% of them. 35% had myalgia. One individual’s only symptoms were corzya and headache. But we have to bear in mind that this data is likely to be affected by recall bias and that this sample wasn’t rigorously selected to be representative of any particular population. However, it is clear that this is a heterogenous condition and screening by symptoms seems like haphazard approach.
Is pregnancy a risk factor for severe covid-19?
8% of the 118 pregnant women in this study had severe covid-19 whereas 15.7% of the general population in China got severe disease. The authors conclude that pregnant women do not have an increased risk of severe disease. I’m not convinced. This comparison is very crude. I would say we just don’t know. And next time, include a comparator group, please. Why weren’t they compared with matched non-pregnant women for example? Sure, there would plenty of confounders but better than nothing! Rates of miscarriage (3%), ectopic pregnancy (2%), premature delivery (21%) and Caesarean sections (93%) in pregnant women with COVID-19 are interesting but the impact of COVID-19 on these rates is unknown.
News from New York
Two hospitals in Manhattan describe the characteristics of their first 393 consecutive patients. A third had “respiratory failure leading to invasive mechanical ventilation”—10 times higher than reports from China. These hospitals had what they call an “early intubation strategy with limited use of high-flow nasal cannulae during this period.” This has a responsible, scientific-sounding ring to it, as if anyone knows whether this is a good idea. This report indicates that early intubation leads to, well, lots of people being intubated and also not a lot of people being able to be extubated. And lo and behold, demand for mechanical ventilation is set to outweigh the supply. Does anyone remember this thing for comparing two interventions of unknown benefit called a randomised controlled trial?
Iceland on the ball
Not football. Football’s off, remember? Iceland has done a good job of trying to understand how covid-19 spreads in a population. People were tested if they were “high-risk” which usually meant they were symptomatic, had contact with an infected person or had recently travelled to a high-risk country. They also screened the population using 2 strategies, an online invitation open to any resident of Iceland who was symptom-free or who had mild symptoms of the common cold and direct invitations to a random sample of the population. 13.9% of the high-risk group tested positive for the virus compared to 0.8% in the open-invite group and 0.6% in the random sample group. The latter numbers are probably indicative of the true prevalence in Iceland at the time. The incidence of infection was lower in women and children. The virus’s age and gender predilection may be the key to a profound pathological insight—can we get that worked out in time for next week’s column please? Many thanks.
Alex Nowbar is a clinical research fellow at Imperial College London
Competing interests: None declared