Long-term outcomes of critically ill patients with hematological malignancies: what is the impact of the coronavirus disease 2019 pandemic? Author's reply

Laveena Munshi, Guillaume Dumas, Sangeeta MehtaIntensive 2024 · DOI · OpenAlex

Introduction

The COHO study was designed to provide a comprehensive understanding of the long-term survival and functional outcomes for patients with hematological malignancies following admission to the intensive care unit (ICU). Given the high-risk nature of this population, establishing a baseline for prognosis is vital for clinical decision-making. However, the emergence of the COVID-19 pandemic during the study's final recruitment phase introduced a potential confounding variable. In this formal reply, the authors address concerns regarding whether the pandemic biased their findings, specifically examining the timing of recruitment and the outcomes of the pandemic-era subgroup.

The Pandemic as a Potential Confounder

In clinical research, a sudden global event like the COVID-19 pandemic can skew results in two primary ways. First, there is the direct impact, where patients might succumb to the virus itself rather than their underlying malignancy. Second, there is the systemic impact, where overwhelmed ICUs might provide a different standard of care due to resource strain, potentially leading to higher mortality for all patients regardless of their diagnosis. If either of these factors were significant during the COHO study, the reported long-term outcomes would not accurately reflect the baseline prognosis for this patient population.

Methodology of the Subgroup Analysis

To address these concerns, the authors performed a retrospective subgroup analysis of the COHO study data. They focused on the temporal relationship between their recruitment window, which concluded in June 2020, and the progression of the pandemic in Ontario, Canada.

The authors specifically isolated a "pandemic-era" cohort, defined as patients admitted after the official onset of the pandemic in the region (March 20, 2020). By comparing the mortality rates and clinical characteristics of this subgroup against the total study population, the authors sought to determine if a statistically significant "pandemic effect" existed.

Key Findings and Data Insights

The analysis revealed several critical points that defend the internal validity of the COHO study:

  • Minimal Temporal Overlap: The majority of data collection occurred in a "pre-COVID" environment. Only 44 patients, representing 11% of the total cohort of 414, were admitted during the early stages of the pandemic.
  • Low Viral Prevalence: Only two patients (0.48% of the total study population) were admitted to the ICU specifically due to COVID-19. This indicates that the direct impact of the virus on the study's mortality data was negligible.
  • The Paradox of Lower Mortality: Surprisingly, the ICU mortality rate for the pandemic-era subgroup was 43% (19/44), which was lower than the 51% mortality rate observed for the entire cohort. This suggests that the quality of care and survival outcomes for these patients were not compromised by the early stages of the pandemic.
  • Systemic Resilience: In Ontario, the ICU system did not face significant, overwhelming strain until the second and third waves in late 2020 and beyond. Since COHO recruitment ended in June 2020, the study avoided the period of greatest systemic resource strain.

Impact on Study Completion

While the pandemic did not bias the patient data, it did have a significant impact on the study's execution. The authors noted that the pandemic caused a massive shift in research resources toward COVID-19-specific studies. This "research displacement" forced the early termination of the COHO study before it could reach its original target sample size. While this reduces the overall statistical power of the study, it does not introduce the type of directional bias that would invalidate the reported outcomes.

Conclusion

The authors conclude that the COVID-19 pandemic was not a major driver of the long-term mortality or functional outcomes reported in the COHO study. The minimal overlap with the pandemic's peak in Ontario and the favorable mortality rates in the pandemic-era subgroup confirm that the study provides a robust look at the prognosis for critically ill patients with blood cancers under standard care conditions. This exchange highlights the importance of temporal context in longitudinal research and the need for future studies to clearly demarcate pandemic-related cohorts when analyzing survival trends.

Summary updated

BibTeX
@inproceedings{munshi2024long,
  author = {Laveena Munshi and Guillaume Dumas and Sangeeta Mehta},
  title = {{Long-term outcomes of critically ill patients with hematological malignancies: what is the impact of the coronavirus disease 2019 pandemic? Author's reply}},
  booktitle = {Intensive},
  year = {2024},
  doi = {10.1007/s00134-024-07614-1},
}
Figure