Thread by @NicolaMedical

THREAD
The latest @ICNARC report is now out with, for the first time, multivariate analysis. So, stepping into @ActuaryByDay shoes, here’s the summary, and a link to the full report. icnarc.org/Our-Audit/Audi… @COVID19actuary @ICS_updates
1/

The report captures all patients in E & W with confirmed COVID-19 admitted to ICU up until 4pm on 7 May 2020 where outcome data is available. So far, of the 10,758 notified admissions, 6143 have outcomes reported and 2107 are still receiving critical care
2/

The downward trend for admissions to ICU continues; the effects of social distancing evident
3/

There’s very little change in patient characteristics compared to last week, the majority of patients are still considered to be physically independent and less likely to have severe comorbidities compared to the comparator cohort
4/

As predicted by @ActuaryByDay last week, London no longer dominates the top 3 as far as geographical concentration is concerned though overall, London continues to have the greatest number of cases. Otherwise there has been very little change in the order
5/

The age and sex distribution analysis again showing that males are disproportionately vulnerable with over half of those admitted being males over the age of 50
6/

Multivariate analysis was conducted on a cohort of ~ 6000 patients and included those discharged alive from hospital within 30 days and those ending critical care within 30 days. Prognostic factors were selected based on what is already understood about critically ill patients
7/

The next graph demonstrates that those in ICU with COVID-19 aged 70 have twice the risk of death compared to those aged 60, and those aged 40 have half the risk of death compared to those aged 60
8/

The next chart may come as a surprise given that it suggests that the mortality HR for males is broadly similar to females once known variables have been accounted for. However, we can also see that ethnic disparities persist, this must be addressed urgently
9/

Hazrard ratios for deprivation are displayed next; health disparities for the most deprived are evident
10/

Risk of death was associated with increasing BMI compared against a reference of a BMI of 30. This is in line with what is understood about the health impacts of obesity, though would be interesting to see if there is any data on those considered underweight
11/

The impact of comorbidities is next demonstrating the impact of increasing dependency . The immunocompromised group were also at increased risk, (inc metastatic & blood cancers). The other severe comorbs were too small in number to include in the analysis
12/

Additional analysis on physiological factors are captured in the full document. Please note that in multivariate analysis, association DOES NOT in any way confirm causation. Thank you to @ICNARC for compiling this analysis in such a timely manner and for the earlier presentation