Unveiling Clinically Meaningful Phenotypes in Japanese COVID-19 Patients through Cluster Analysis

Introduction:

The clinical course of coronavirus disease (COVID-19) is characterized by a diverse range of symptoms and outcomes. Identifying clinically meaningful phenotypes can be crucial for predicting the severity or prognosis of the disease, particularly in high-risk populations such as older adults. In this study, we aimed to investigate the usefulness of cluster analysis in identifying prognostically significant phenotypes among Japanese COVID-19 patients.

Methods:

We conducted a retrospective cohort study of 1322 Japanese patients with COVID-19 who were admitted to hospitals between April 2020 and May 2021. Patients aged 18 years or older who agreed to participate in the study were included. We analyzed demographic, clinical, and laboratory data using hierarchical cluster analysis to identify patterns of co-variates associated with COVID-19 severity or prognosis.

Results:

Our results revealed four distinct clusters, each characterized by unique profiles of clinical and laboratory parameters. Clusters 3 and 4 exhibited higher rates of oxygen supplementation, intensive care unit admission, and mechanical ventilation compared to the other two clusters. However, mortality rates were lower in Cluster 3. Despite having a similar age distribution as Cluster 2, Cluster 3 tended to have poorer outcomes.

Interestingly, patients in Clusters 3 and 4 had elevated levels of serum lactate dehydrogenase, ferritin, KL-6, d-dimer, and C-reactive protein compared to the other two clusters. In contrast, symptoms such as sore throat, dysosmia, and dysgeusia were less frequent in these clusters, while shortness of breath was more common.

Discussion:

Our study demonstrates the usefulness of cluster analysis in identifying clinically meaningful phenotypes among Japanese COVID-19 patients. The results suggest that distinct subgroups exist within the population, each characterized by unique profiles of clinical and laboratory parameters. These findings have important implications for predicting the severity or prognosis of COVID-19 and informing personalized treatment strategies.

In conclusion, our study highlights the importance of considering individual patient characteristics when managing COVID-19 patients. The identification of clinically meaningful phenotypes can help healthcare providers tailor their approach to each patient's needs, potentially improving outcomes and reducing healthcare resource utilization. Future studies should aim to validate these findings in larger populations and explore the potential applications of cluster analysis in predicting COVID-19 severity or prognosis.

'Aya Endo Akane'

Aya endo akane