Patients Hospitalized With Coronavirus Disease 2019: A Diverse Population

(See the Major Article by Kojima et al. on pages 547–57.)

Historically, patients infected with Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) required hospitalization if they developed serious Coronavirus Disease 2019 (COVID-19) disease, typically resulting from pneumonia, which may progress to acute respiratory distress syndrome (ARDS), and/or with thromboembolic events or acute kidney injury. Treatments, according to guidelines, have encompassed the use of antiviral [1], immunomodulatory, and anticoagulation therapies in addition to organ support, including supplementary oxygen [2]. Initially in the pandemic, when population immunity against SARS-CoV-2 was negligible, the key risk factors for serious COVID-19 were older age and male sex. In this vulnerable section of the population, obesity and comorbidity were additional risk factors.

More recently, population immunity against SARS-CoV-2 has increased because of natural infection, vaccination, or a combination thereof [3]. Vaccines have been shown to reduce the risk of serious COVID-19; a benefit that has been reasonably maintained during the pandemic [4]. Hence, if a vaccinated person gets admitted with serious COVID-19, this can be considered as a vaccine failure. As expected, protective immunity is less frequently seen in patients with compromised immune function, and these patients will hence tend to be enriched among hospitalized patients with serious COVID-19 during the pandemic.

A distinctly different group of SARS-CoV-2–infected hospitalized patients are those not in need of in-hospital care for their infection, but nevertheless admitted for care of COVID-unrelated health care problems. For such patients, COVID-19 treatment guidelines follow the principles for ambulatory patients. During situations with widespread community transmission in populations with reasonably high population immunity, this group is expected to dominate among infected and hospitalized patients.

During the pandemic, the virus has changed its virulence features, increased its transmissibility, and became more immune evasive [5]. In 2021, first the alpha strain and subsequently the delta strain dominated. Delta caused more serious disease than alpha, was slightly more transmissible, and was associated with higher plasma levels of nucleocapsid protein [6]. However, at the end of 2021, delta was outcompeted by the omicron strain. Omicron's ability to cause serious disease was weaker than that of delta for 2 principal reasons. First, omicron replicates less well in pulmonary tissue than delta and hence is less likely to cause serious COVID-19 [7]. Second, benefit from established immunity protecting against serious disease seen in most countries also protected against serious COVID-19, despite the emergence of omicron.

Kojima and colleagues' [8] paper in this issue of the journal reports these time trends nicely. The study uses data from an impressively large prospective multisite cohort of consecutive patients hospitalized with COVID-19 at several participating US sites from March 2021 to August 2022.

Most notable, Kojima et al [8] report a reduction in the proportion of admitted patients with serious COVID-19 disease during the omicron phase of the pandemic compared with the situation in 2021. Several proxies for severity were used to illustrate this. The need for supplementary oxygen fell from 82% to 83% of hospitalized patients in 2021, to 58% to 60% in to 2022. This was mirrored by a decrease in the use of glucocorticoids—indicated for patients with COVID-19 requiring supplementary oxygen—around the time of admission. This drug was used in 74% to 80% in patients admitted in 2021 and fell to 54% in 2022. Most impressively, the need to administer oxygen to treat patients with COVID-19 ARDS also decreased over time (from 40%–45% of patients hospitalized in 2021 to 15%–16% into 2022). As expected, the resulting in-hospital risk of death decreased from 12% during delta to 4% into 2022.

Two important aspects of the report by Kojima et al [8] are worth considering.

First, death was ascertained while patients were admitted but not after discharge. Studies following patients for 60 to 90 days from time of admission, irrespective of admission status, have reported that between 10% and 15% of discharged patients either get readmitted and/or die [9–11]. The decisions to discharge, especially under an epidemic surge with overwhelmed hospital capacity, is complex. Regardless, attempts—if at all feasible—should be made to ascertain serious outcomes in initially hospitalized cohorts over 60 or 90 days from time of admission, to fully understand the negative health impact from being hospitalized with serious COVID-19.

Another part of the report from Kojima et al [8] relates to the selection of patients that entered the study. Not all infected and patients admitted to the hospitals were enrolled. It would have been interesting to know this number. Rather, potential participants had to have acute COVID-19 with at least 1 of a series of symptoms, but not necessarily disease from the lower respiratory tract. How well patients admitted with serious illness resulting from COVID-19 were delineated from those with serious illness from other causes and nonserious COVID-19 is therefore unclear. The study focuses on percentage of the cohort with various manifestations and outcomes; hence, understanding the denominator is important.

Given the data summarized here, it would be possible to identify subgroups among the enrolled participants that likely had COVID-19 pneumonia or COVID-19 ARDS based on clinical and/or therapeutic proxies. The residual group is those with borderline if at any signs of serious COVID-19 disease.

Classifying the cohort as suggested previously would provide better insight to interpret some of the time trends. For example, is the reason why inflammatory markers and the risk of myocardial infarction and venous thrombotic disease decreased over time is that proportionally more patients with milder COVID-19 disease were admitted? Conversely, have COVID-19 pneumonia and associated ARDS disease entities become more benign over time?

More generally, the development of a harmonized and implementable case definition of those with serious illness resulting from COVID-19 among infected hospitalized cohorts would add scientific strength to this area of research. For example, mixing the 2 distinct groups of admitted patients when assessing vaccine effectiveness is obviously problematic because it dilutes the ability to provide a meaningful estimate of protection against serious COVID-19. Also, surveillance-wise, a lot would be gained from doing so. To understand the severity of an ongoing surge, it is important to be able to distinguish serious disease resulting from COVID-19 from serious disease with incidental COVID-19.

Finally, a case definition could assist to improve insights into best practices for managing related respiratory infections. We have invested enormous resources in SARS-CoV-2 for the past 3 years for very good reasons. An impressive amount of research has been generated in record time, and consequently we now stand on reasonably solid footing to manage the future steps of this pandemic. When reviewing the knowledge base, it is almost a paradox that we presently know better how to manage coronavirus epidemics than influenza epidemics. Influenza, like SARS-CoV-2, has a varied scope of disease. Also, a whole separate commentary could also be written to discuss challenges in our understanding of whether the association between admissions with respiratory-syncytial virus infection and poor outcome is causal or not.

SARS-CoV-2 continues to evolve [12], both clinically and at the molecular level, and it is possible that the virus may regain some of the virulence of the delta strain. If so, carefully performed cohort studies, such as that from Kojima et al [8], will be instrumental to detect this in a timely fashion to enable a timely response.

Notes

Financial support. Funded by Danish National Research Foundation (grant 126) and National Institutes of Allergy and Infectious Diseases (contract 75N91019D00024, task order number 75N91020F00039).

I have received support for related research area from the Danish Government, Danish National Research Foundation, and National Institutes of Allergy and Infectious Diseases to perform the ENFORCE study (https://enforce.dk/), and TICO/ACTIV-3 as well as STRIVE via INSIGHT (http://insight.ccbr.umn.edu/).