The emergence of variants of concern (VOCs), each with potential differences in their virulence and ability to evade the immunity system, challenged healthcare providers and systems during the COVID-19 pandemic (1). It is not certain that viral evolution leads to lower severity; therefore, assessing the impact of new and emerging variants on clinical outcomes may be important for proper clinical care optimization and resource allocation (1–3). The risk of death observed with different severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) VOCs has varied, reflecting the net result of a complex interplay between intrinsic VOC characteristics, patient’s individual characteristics, healthcare resources, therapeutics, and infection- and vaccine-related immunity (4–6).

In this issue of Critical Care Medicine, Santosa et al (7) present survival trends of adult critically ill COVID-19 patients in Sweden during the first two and a half years of the pandemic, from March 2020 to September 2022. The study, utilizing a population-based cohort of Swedish ICU patients with COVID-19 listed as a primary or secondary diagnosis, investigated the impact of VOCs, patient characteristics, comorbidities, and treatment approaches on mortality.

Of the 8975 patients included in the study by Santosa et al (7), 32.6% died. Prognostic predictors of death remained constant throughout different VOC periods: older age, impaired immune system-related disease, greater clinical severity at presentation, and restricted treatment strategy (which meant limiting therapeutic interventions given the low probability of hospital discharge). Conversely, the use of steroids during the ICU stay and booster vaccination (third dose) was associated with reduced mortality risk (7).

Interestingly, the ICU mortality rate was highest during the Omicron period (65.9 deaths per 1000 admitted patients) while the ICU mortality rate during the Delta period was 26.4 deaths per 1000, and 18.8 per 1000 during the Alpha period (7). This contradicts previous findings that suggest Omicron is more transmissible but associated with a lower risk of adverse health outcomes (4,8,9).

As also described by the authors, this contradiction may be understood in light of potential confounding whereby there were differences in patient characteristics or factors related to care between VOC periods that also may have influenced the risk of death (10). For example, patients admitted to the ICU during the Omicron period had worse prognostic factors compared with other VOC periods (older age, higher prevalence of comorbidities, and lower income). Additionally, more than 80% of patients were treated with steroids during the Delta period but less than 50% of patients during the Omicron period. Likewise, more than half of patients were admitted to the ICU from hospital wards during Delta but less than half during Omicron. These patterns, overall, suggest that there are differences in the types of patients compared between the Delta and Omicron periods. While the authors attempted to adjust for these factors, unknown, unmeasured, or poorly measured confounding factors may still influence results—a phenomenon known as residual confounding.

Other findings in the study also suggest confounding. For example, the study reports that booster vaccination was associated with reduced risk of death, particularly for those over 70 years old (as compared with the unvaccinated), while one or two doses of vaccine increased the risk of death, which is inconsistent with other data supporting the efficacy of vaccines (11–13). We speculate whether this apparent contradictory association between one or two doses of vaccines and greater risk of death is, in fact, related to baseline patients’ characteristics: patients under greater risk of death were those prioritized initially for vaccination with the scheme of one and two doses (14). Finally, the study did not genetically confirm SARS-CoV-2 VOCs but rather assumed it according to calendar period. Potential misclassification of VOCs may also contribute for the paradoxical findings, especially since the Delta VOC, which is known for being associated with worse health outcomes, immediately precedes Omicron (4,15).

These findings, however, also do not rule out other explanations for increased mortality associated with Omicron VOC: while the Omicron VOC is associated with lower risk of severe illness, patients who do experience severe illness may be at higher risk of death (16).

Despite limitations, the study by Santosa et al (7) has many strengths. It was a large-scale, nationwide, multilinkage registered-based study involving all critically ill COVID-19 patients admitted to the ICU during the first two and a half years of the COVID pandemic (17). The large number of patients included allowed authors to investigate a broader number of prognostic factors associated with ICU mortality confirming some of the previous prognostic findings in COVID, such as older age, impaired immune system-related disease, greater clinical severity at presentation, and restricted treatment strategy.

In summary, we commend the authors’ careful and thorough work. The study by Santosa et al (7) illustrates well the challenges of evaluating the virulence and sequelae of VOCs within the backdrop of evolving patient characteristics and care patterns. The possibility that future VOCs may exhibit varying levels of virulence underscores the need to invest in VOC surveillance systems that can facilitate timely research and enhance preparedness for future outbreaks.

1. Sigal A, Milo R, Jassat W: Estimating disease severity of Omicron and Delta SARS-CoV-2 infections. Nat Rev Immunol. 2022; 22:267–269 2. Nyberg T, Ferguson NM, Nash SG, et al.; COVID-19 Genomics UK (COG-UK) consortium: Comparative analysis of the risks of hospitalisation and death associated with SARS-CoV-2 omicron (B.1.1.529) and delta (B.1.617.2) variants in England: A cohort study. Lancet. 2022; 399:1303–1312 3. Nyberg T, Bager P, Svalgaard IB, et al.: A standardised protocol for relative SARS-CoV-2 variant severity assessment, applied to Omicron BA.1 and Delta in six European countries, October 2021 to February 2022. Euro Surveill. 2023; 28:2300048 4. Relan P, Motaze NV, Kothari K, et al.: Severity and outcomes of Omicron variant of SARS-CoV-2 compared to Delta variant and severity of Omicron sublineages: A systematic review and metanalysis. BMJ Glob Health. 2023; 8:e012328 5. Janke AT, Mei H, Rothenberg C, et al.: Analysis of hospital resource availability and COVID-19 mortality across the United States. J Hosp Med. 2021; 16:211–214 6. Karagiannidis C, Mostert C, Hentschker C, et al.: Case characteristics, resource use, and outcomes of 10 021 patients with COVID-19 admitted to 920 German hospitals: An observational study. Lancet Respir Med. 2020; 8:853–862 7. Santosa A, Oras J, Li H, et al.: Survival of critically ill COVID-19 patients in Sweden during the first two and a half years of the pandemic. Crit Care Med. 2024; 52:1194–1205 8. Ulloa AC, Buchan SA, Daneman N, et al.: Estimates of SARS-CoV-2 omicron variant severity in Ontario, Canada. JAMA. 2022; 327:1286–1288 9. Ward IL, Bermingham C, Ayoubkhani D, et al.: Risk of covid-19 related deaths for SARS-CoV-2 omicron (B.1.1.529) compared with delta (B.1.617.2): Retrospective cohort study. BMJ. 2022; 378:e070695 10. Becher H: The concept of residual confounding in regression models and some applications. Stat Med. 1992; 11:1747–1758 11. Yang ZR, Jiang YW, Li FX, et al.: Efficacy of SARS-CoV-2 vaccines and the dose-response relationship with three major antibodies: A systematic review and meta-analysis of randomised controlled trials. Lancet Microbe. 2023; 4:e236–e246 12. Graña C, Ghosn L, Evrenoglou T, et al.: Efficacy and safety of COVID-19 vaccines. Cochrane Database Syst Rev. 2022; 12:CD015477 13. Xu Y, Li H, Kirui B, et al.: Effectiveness of COVID-19 vaccines over 13 months covering the period of the emergence of the omicron variant in the Swedish population. Vaccines (Basel). 2022; 10:2074 14. Baral S, Chandler R, Prieto RG, et al.: Leveraging epidemiological principles to evaluate Sweden’s COVID-19 response. Ann Epidemiol. 2021; 54:21–26 15. Champredon D, Becker D, Peterson SW, et al.: Emergence and spread of SARS-CoV-2 variants of concern in Canada: A retrospective analysis from clinical and wastewater data. BMC Infect Dis. 2024; 24:139 16. Becher T, Frerichs I: Mortality in COVID-19 is not merely a question of resource availability. Lancet Respir Med. 2020; 8:832–833 17. Nyberg F, Franzén S, Lindh M, et al.: Swedish Covid-19 investigation for future insights—a population epidemiology approach using register linkage (SCIFI-PEARL). Clin Epidemiol. 2021; 13:649–659