Timothy M. Cook, MBBS, on COVID-19 ICU Mortality Rates

Although the COVID-19 pandemic has been associated with increased rates of intensive care unit (ICU) admittance, mortality rates among these patients remain unclear.

In order to estimate these rates, as well as to examine how factors such as geography and phases of the pandemic affected them, researchers performed a systematic review and meta-analysis¹ using data from 24 observational studies (N = 10,150 patients) conducted from December 19, 2019, to May 28, 2020.

Overall, pooled ICU mortality was 41.6%, which represents a decrease of roughly one-third since March 2020 (59.5%) and is significantly lower than some estimates reported at the beginning of the pandemic.

Consultant360 reached out to study author Timothy M. Cook, MBBS, who is a consultant in anaesthesia and intensive care medicine at Royal United Hospitals NHS Trust in Bath, England, and is an honorary professor of anaesthesia at Bristol University, to further discuss the results of the study.

Consultant360: Your results showed that the overall mortality rate did not differ by geographic location. Was this an unexpected result?

Timothy Cook: It was always a possibility, which is why we set out to explore it. “Intensive care” is a rather undefined entity and it is quite plausible that what is described as an ICU in one region may not be the same in all countries or geographies. For this reason, it is also plausible that patients may have different characteristics and illness severity when they are admitted in different geographic locations.

There was some evidence of variations in patient characteristics, admission criteria, and treatments delivered by geography but as reporting in studies is not standardized, we were not able to explore these differences in detail. Overall mortality did not differ across the continents.

Consultant360: In your study, you noted that there was a drop in in-ICU mortality from above 50% in March 2020 to approximately 40% in May 2020. What factors do you think played a role in this decrease?

TC: We looked at completed episodes of ICU care—so only included patients who were either discharged alive from ICU or who had died while on ICU. These results sometimes differ a lot from the “crude mortality” in reports, because in many studies the majority of patients were still in ICU when the study was reported. This means the real mortality may be quite different depending on the outcomes of those still receiving ICU care.

Mortality in papers published up to the end of March mortality was 59.5%. By the end of May, the mortality in all studies was down to below 42%.

We suggest there may be 4 explanations. 

• During the early surges, many healthcare systems were very stretched and “stressed” and it is possible that as the surges have passed these healthcare systems may have recovered and been able to deliver better care. This may also have led to altered admission criteria as the pandemic has resolved.
• It is possible that we have simply improved the quality of care delivered as we have learned more about the disease. This has been supported by prompt publication of clinical reports early in the pandemic. There is no doubt that ICU treatment has changed (for example, improvements in how patients’ lungs are ventilated, fluid administration strategies and anticoagulation regimens) and patients may well be benefiting from these changes.
• Many patients admitted to ICU with COVID-19 spend several weeks in ICU. In the UK approximately 20% of patients stay in ICU at least 4 weeks and 9% more than 6 weeks. For some patients it takes a long time to wean from a ventilator, and good outcomes from these “long-stayers” may have contributed to improved outcomes.
• Finally, research may have helped. The RECOVERY study's dexamethasone data was published in June so our results preceded its publication, but the large number of patients who were included in research studies throughout April and May, including those randomized to receive dexamethasone, may have contributed a small element of improvement in outcome.

There has been some speculation elsewhere that mortality may have fallen because the virus has become less potent—but evidence from outside our study does not support this.

Consultant360: This is the first systematic review and meta‐analysis of outcomes of patients admitted to ICU with COVID‐19. What should future studies focus on to help improve our understanding of this subject?

TC: Our study has only looked at 3 geographical regions (Asia—mostly China, Europe and North America) and it will be interesting to see whether geographic variations become evident when studies from Southern Europe, South America, Africa, and Australasia are included. These studies were not included in our study as the pandemic has hit these regions rather later, and such studies are awaited.

If possible, future reviews should not only look at outcomes but may also explore whether there are differences in patient populations, admission criteria, treatments delivered, and how these link to outcome. However, we were not able to explore this because of the ways that studies are reported. Improved standardization of reporting of such studies would improve our understanding of how patient, disease, and treatment factors are correlated with good or poor outcomes.

Consultant360: You noted that you “chose in‐ICU mortality as our primary outcome measure as a useful metric of the efficacy of ICU care.” What do your findings show about the efficacy of ICU care across the world?

TC: This is a really interesting question. The results show that overall a patient admitted to ICU on one side of the world has broadly the same chance of surviving as a patient admitted to ICU on the other side of the globe. However, it remains uncertain whether the ICUs (in terms of patients, disease severity, and treatments) that we compared with each other were equivalent. We noted a lot of variation (heterogeneity) in our results and this merits further exploration.

Consultant360: How might your findings, overall, affect practitioners across the world?

TC: Firstly, I think the study can be considered good news. In March, the figures coming out of ICU were pretty depressing with well over half of all ICU patients dying and amongst those needing mechanical ventilation, perhaps three quarters were dying. Just a few months later, the mortality rate is dramatically lower, and this appears to be the case in all the geographical regions we explored.

This is good news for patients and for all those treating them. As we now also know that dexamethasone can significantly reduce mortality in patients with COVID-19 we have 2 pieces of good news. Ongoing research provides the possibility of new therapeutics and refining current treatment further. So, it is plausible that ICU mortality may fall even lower and we need to be sure that ICU treatment is optimized globally. In the next months and possibly years, we can anticipate surges in SARS-CoV-2 activity and we also need to make efforts to ensure we have enough ICU provision to manage this. 

Reference:

1. Armstrong RA, Kane AD, Cook TM. Outcomes from intensive care in patients with COVID‐19: a systematic review and meta‐analysis of observational studies. Anaesthesia. Published online June 30, 2020. https://doi.org/10.1111/anae.15201