Abstract
This secondary analysis of the COVIP study shows a higher 30-day mortality in critically ill elderly COVID-19 patients who received steroids as part of their treatment @cjungMD https://bit.ly/3xdyEur
To the Editor:
More than a year after the onset of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic, treating patients with coronavirus disease 2019 (COVID-19) remains a challenge. In contrast to the rapid development of effective vaccines against SARS-CoV-2, the development of specific and effective therapeutics against COVID-19 remains largely unresolved.
In addition to standard intensive care, including oxygen therapy and organ support when required, the use of systemic corticosteroids was found to have a positive effect in randomised trials. However, data regarding treatment of elderly COVID-19 patients are scarce.
Importantly, treatment with corticosteroids has well documented deleterious effects [1]: While the immunosuppressive effect in patients with COVID-19 is presumably responsible for the desired therapeutic effect, it may also render the patients more prone to secondary bacterial infections and potentially decrease viral clearance [2]. Corticosteroid therapy is also associated with hyperglycaemia, has catabolic effects and is associated with neuropathy. This could potentially affect the risk–benefit balance, especially in vulnerable patient groups, such as elderly, frail patients.
The aim of this secondary analysis was to investigate the effects of corticosteroid therapy in an international observational prospective study of critically ill elderly patients with COVID-19.
The COVIP study (“Corona Virus disease (COVID19) in Very Elderly Intensive care Patients (VIPs)”; NCT04321265) included patients aged 70 years or older with proven COVID-19 and admitted to an intensive care unit (ICU) [3]. 30-day mortality was defined as the primary endpoint. The study was conducted by the Very old Intensive care Patient (VIP) network [4] across 207 ICUs in 35 countries. Data were collected through an electronic case report form. A prospective study design was chosen to achieve high-quality data. Informed consent was taken if not waived by the local ethical committee.
Two multi-level logistic regression models were utilised: the first model used the hospital unit as random effect and the steroid use as fixed effect; the second model was a multi-variable model adjusting for “The Sequential Organ Failure Assessment” (SOFA) score and frailty as assessed by the Clinical Frailty Scale. Sensitivity analyses complemented the analysis.
In total, 3082 patients were included in the COVIP study; 2115 patients received corticosteroids, and 967 patients received none. Median age was 75 (interquartile range (IQR) 72–79) years in both groups. With a median SOFA score of 5 (IQR 3–8), there was no difference between the two groups.
30-day mortality was 53% in the group treated with corticosteroids and 42% in the no-corticosteroid group (p<0.001).
The univariate 30-day mortality rates were higher in patients receiving corticosteroids (53 versus 42%; aOR 1.16, 95%CI 1.28–2.02; p<0.001). This association of corticosteroid use was even more pronounced after 3 months (69% versus 49%; p<0.001; figure 1a). In addition, we found that corticosteroids remained associated with increased odds of 30-day mortality after multivariable adjustment (aOR 1.60, 95% CI 1.26–2.04; <0.001). Further sensitivity analyses consistently confirmed the finding in subgroups stratifying for age (<80/≥80 years), frailty (fit/vulnerable/frail), mechanical ventilation (yes/no), noninvasive ventilation (yes/no), sex (female/male), symptom onset (≤7 days/>7 days), and vasopressor therapy (yes/no) (figure 1b). Furthermore, in sensitivity analyses evaluating patients in the first surge (March–May, aOR 1.38, 95% CI 1.05–1.82; p=0.02; n=1448) and the second surge (September–December, aOR 2.09, 95% CI 1.04–4.21; p=0.04; n=1414) the finding was sustained.
In this prospective study of more than 3000 critically ill COVID-19 patients aged 70 years and older, we have found an independent association of steroid use with increased mortality.
These results question the routine use of corticosteroid treatment in elderly COVID-19 patients. While the immunosuppressive effect of steroids is undisputed and desirable in the context of severe COVID-19 treatment, the adverse effects of steroid treatment in elderly patients may outweigh the potential benefits.
This is the largest prospective analysis of critically ill elderly patients in relation to corticosteroid use to treat severe COVID-19 disease. Corticosteroid therapy has been established as standard of care in all ICU patients. However, even landmark randomised controlled trials do not support this with evidence in elderly patients. The RECOVERY study [5] showed no effect of corticosteroids in their subgroup of patients >70 years. Of note, only 169 patients in this group were on mechanical ventilation. The CoDEX study found no effect on mortality after 28 days, but no data was supplied specifically looking at patients above 70 years old [6]. Thus, both studies included far fewer patients than our current analysis. It is important to note that our data does not question the corticosteroid strategy in younger COVID-19 patients. It just emphasises that the decision to use corticosteroids needs to be individually tailored, first and foremost according to age, but also with regards to comorbidities and other factors [7].
Our analysis has limitations: First, this is a secondary analysis of a prospective study. Second, our study is not randomised and, despite multivariable adjustment, it is likely that unknown confounding factors may have contributed to our findings. Third, we have no detailed information about dosage and duration of corticosteroid treatment.
In conclusion, in this prospective observational study we found a higher 30-day mortality in critically ill elderly COVID-19 patients who received steroids as part of their treatment.
Shareable PDF
Supplementary Material
This one-page PDF can be shared freely online.
Shareable PDF ERJ-00979-2021.Shareable
Footnotes
This study is registered at Clinicaltrials.gov as NCT04321265.
Conflict of interest: C. Jung has nothing to disclose.
Conflict of interest: B. Wernly has nothing to disclose.
Conflict of interest: J. Fjølner has nothing to disclose.
Conflict of interest: R.R. Bruno has nothing to disclose.
Conflict of interest: D. Dudzinski has nothing to disclose.
Conflict of interest: A. Artigas reports grants from Grifols and Fisher&Paykel, personal fees for advisory board work from Grifols, Novartis and Lilly Foundation, outside the submitted work.
Conflict of interest: B. Bollen Pinto has nothing to disclose.
Conflict of interest: J.C. Schefold has nothing to disclose.
Conflict of interest: G. Wolff has nothing to disclose.
Conflict of interest: M. Kelm has nothing to disclose.
Conflict of interest: M. Beil has nothing to disclose.
Conflict of interest: S. Sigal has nothing to disclose.
Conflict of interest: P.V. van Heerden has nothing to disclose.
Conflict of interest: W. Szczeklik has nothing to disclose.
Conflict of interest: M. Czuczwar has nothing to disclose.
Conflict of interest: M. Elhadi has nothing to disclose.
Conflict of interest: M. Joannidis has nothing to disclose.
Conflict of interest: S. Oeyen has nothing to disclose.
Conflict of interest: T. Zafeiridis has nothing to disclose.
Conflict of interest: B. Marsh has nothing to disclose.
Conflict of interest: F.H. Andersen has nothing to disclose.
Conflict of interest: R. Moreno has nothing to disclose.
Conflict of interest: M. Cecconi has nothing to disclose.
Conflict of interest: S. Leaver has nothing to disclose.
Conflict of interest: A. Boumendil has nothing to disclose.
Conflict of interest: D.W. De Lange has nothing to disclose.
Conflict of interest: B. Guidet has nothing to disclose.
Conflict of interest: H. Flaatten has nothing to disclose.
Support statement: This work was supported by the European Commission, APHP and Health Region West (Norway). Funding information for this article has been deposited with the Crossref Funder Registry.
- Received April 6, 2021.
- Accepted June 2, 2021.
- Copyright ©The authors 2021.
This version is distributed under the terms of the Creative Commons Attribution Non-Commercial Licence 4.0. For commercial reproduction rights and permissions contact permissions{at}ersnet.org