Infectious events in patients with severe COVID-19: results of a cohort of patients with high prevalence of underlying immune defect

Anastasia Saade, Giulia Moratelli, Guillaume Dumas, Asma Mabrouki, Jean‐Jacques Tudesq, Lara Zafrani, Élie Azoulay, Michaël DarmonAnnals 2021 · DOI

Abstract

In this population of COVID-19 patients with high prevalence of underlying immune defect, a high risk of infections was noted. Mechanical ventilation (MV) and use of steroids were independently associated with infection rate.

Introduction

Early in the COVID-19 pandemic, clinicians frequently prescribed empirical antibiotics due to concerns regarding co-infections. However, data on the actual frequency and risk factors for these infections remained sparse, particularly for "high-risk" patients—those with underlying malignancies or organ transplants. In severe COVID-19 patients, especially those with pre-existing immune defects, secondary infections represent a major complication, often affecting over a quarter of patients within the first 10 days of ICU admission. While steroids like dexamethasone have become standard for treating COVID-19-related inflammation, emerging evidence suggests they may significantly increase the risk of hospital-acquired infections (HAIs) in vulnerable populations without a clear mortality benefit.

Methods

This retrospective single-center study analyzed 100 consecutive adult patients admitted to the ICU at Saint-Louis University Hospital in Paris between March and May 2020. A defining characteristic of this cohort was its unique composition: 34% of patients had underlying malignancies or were solid organ transplant recipients, a much higher prevalence of immune defects than the typical 4% seen in other COVID-19 studies.

To address the complexities of ICU data, the researchers employed several robust statistical techniques, including competing-risk analysis to avoid overestimating infection risk and propensity score matching to create comparable groups for evaluating the impact of dexamethasone.

Fig. 1: Study flow-chart detailing the breakdown of the 100-patient cohort by mechanical ventilation status and infection outcomes.

Results

The study revealed a stark contrast between community-acquired and hospital-acquired infections. Only 7% of patients presented with a bacterial infection upon admission. However, the incidence density of HAIs in the ICU was 125 events per 1,000 ICU-days. By day 10, 27% of the total population had developed an infection.

Risk Factors for Infection

Mechanical ventilation (MV) emerged as the primary gateway for secondary hits, with a sub-hazard ratio of 11.0. For patients on MV, the infection risk jumped to 48% at day 10, compared to just 2.2% for non-ventilated patients. The physical breach of the airway combined with the patient's weakened state makes intubation a massive risk factor for pneumonia, predominantly caused by Gram-negative bacilli such as Pseudomonas aeruginosa and Klebsiella species.

Fig. 2: Cumulative incidence of infectious events, ICU discharge, and ICU mortality for the overall population (a) and stratified by invasive mechanical ventilation (b).

The Impact of Steroids

Dexamethasone use was independently associated with a nearly threefold increase in infections (35% vs. 13% in matched controls; $p=0.01$). Crucially, in this specific cohort, the use of steroids did not show a significant influence on mortality, highlighting a potential "double-edged sword" effect in immunocompromised patients.

Discussion and Conclusion

While secondary infections are frequent, overall ICU mortality (28%) was independently associated with patient severity (SAPS II score) and underlying malignancy rather than the infections themselves.

The study is limited by its single-center, retrospective design and a relatively small sample size of 100 patients. Despite propensity matching, confounding by disease severity remains a challenge. These findings suggest a need for refined antibiotic stewardship to avoid unnecessary usage at admission and call for larger studies to explore the influence of steroid dosing on infection rates in immunocompromised COVID-19 populations.

Summary updated

Abstract

In this population of COVID-19 patients with high prevalence of underlying immune defect, a high risk of infections was noted. MV and use of steroids were independently associated with infection rate.
BibTeX
@article{saade2021infectious,
  author = {Anastasia Saade and Giulia Moratelli and Guillaume Dumas and Asma Mabrouki and Jean‐Jacques Tudesq and Lara Zafrani and Élie Azoulay and Michaël Darmon},
  title = {{Infectious events in patients with severe COVID-19: results of a cohort of patients with high prevalence of underlying immune defect}},
  journal = {Annals},
  year = {2021},
  doi = {10.1186/s13613-021-00873-x},
}
Figure