Another week, another @ICNARC report.
When I dug through this report two weeks ago to examine whether COVID-19 patients were really “at death’s door” (see article), I didn’t expect it to become a regular weekend fixture.
They are compelling though, so let’s take a look. /1
When I dug through this report two weeks ago to examine whether COVID-19 patients were really “at death’s door” (see article), I didn’t expect it to become a regular weekend fixture.
They are compelling though, so let’s take a look. /1
Full report can be found here: icnarc.org/DataServices/A…
I’m doing this plain-English summary on behalf of @ICS_updates and @COVID19actuary. In truth though the report is well written and very accessible.
I’ll keep the tone light, but I know what these numbers represent. /2
I’m doing this plain-English summary on behalf of @ICS_updates and @COVID19actuary. In truth though the report is well written and very accessible.
I’ll keep the tone light, but I know what these numbers represent. /2
The latest report goes to 4pm on 16 April and covers 5,578 patients. We have far more outcome data now with 1,437 patients discharged from ICU and 1,499 deaths. 2,642 patients are still receiving critical care. /3
I’ll start with some hopeful news. This chart shows us that the number of COVID-19 patients in ICU appears to have stabilised (before the lag period, which we should ignore).
It’s a different chart from the cumulative one I showed last week. Plateau is clearer in this one. /4
It’s a different chart from the cumulative one I showed last week. Plateau is clearer in this one. /4
I can hear the shouts of “capacity constraints” already. I don’t *think* that’s what’s driving this. I’m not certain what proportion of our ICU capacity is included in this report but I believe it is high. Others may be able to correct/support me on this point. /5
Fig 4 shows the outbreak is still concentrated in certain places. Top 3 networks are all London! East & South East, Birmingham, Manchester all have high admissions too.
The variation in the lag between notification and receipt of patient data is curious. Why are some slower? /6
The variation in the lag between notification and receipt of patient data is curious. Why are some slower? /6
Here is our old friend Table 1 (with headings!)
As before, 93% could live without assistance in daily activities and 93% were free of severe comorbidities. Much higher than typical viral pneumonia patients.
These ICU patients were not dying before they caught COVID-19. /7
As before, 93% could live without assistance in daily activities and 93% were free of severe comorbidities. Much higher than typical viral pneumonia patients.
These ICU patients were not dying before they caught COVID-19. /7
The age and sex distribution is little changed. Much more concentration among males aged 50-80 than typical viral pneumonia. Three quarters of ICU patients are males.
Other groups underrepresented. /8
Other groups underrepresented. /8
A few people have commented on both ethnicity and BMI. Both results look interesting at first, but in both cases the distribution is largely, but not entirely, explained by the age, sex and location of the patients.
For ethnicity in particular I’d still want more data. /9
For ethnicity in particular I’d still want more data. /9
Turning to outcomes, with nearly 3,000 now, it’s still very much 50/50 overall whether someone dies or is discharged.
Age is the biggest factor, with chances of survival falling significantly as age increases.
Below age 40, 1-in-4 dies.
Above age 80, 1-in-4 survives.
/10
Age is the biggest factor, with chances of survival falling significantly as age increases.
Below age 40, 1-in-4 dies.
Above age 80, 1-in-4 survives.
/10
Chances of survival fall as BMI increases. 👆
Females are slightly more like to survive.
Those with severe comorbidities or needing assistance with daily activities also do less well (but numbers are very small). /11
Females are slightly more like to survive.
Those with severe comorbidities or needing assistance with daily activities also do less well (but numbers are very small). /11
I discussed the applicability of these results to the broader hospital population last week so I won’t go back over it here.
Clearly, not all people who get sick make it to hospital. And not all hospital patients are admitted to ICU. /12
Clearly, not all people who get sick make it to hospital. And not all hospital patients are admitted to ICU. /12
We have here a large and growing data set of people who it is clear mostly had many years ahead of them.
I’ll quote one of the intensivists who replied to me last week. /13
I’ll quote one of the intensivists who replied to me last week. /13
I’ll wrap things up there.
Thank you to the incredible @aroradrn for the suggestion provide these regular updates.
If you wish, you can support the intensive care community here. /14
Thank you to the incredible @aroradrn for the suggestion provide these regular updates.
If you wish, you can support the intensive care community here. /14
Two typos in tweet 10 (sorry!)
Below age [50], 1-in-4 dies.
Above age 80, 1-in-[3] survives.
(No-one has called me out on them, thanks, but they were bothering me).
Below age [50], 1-in-4 dies.
Above age 80, 1-in-[3] survives.
(No-one has called me out on them, thanks, but they were bothering me).
