Profiles And Ouctome In Patients With Respiratory Covid-19 Admitted In Wards Of A French Hospital: A Clustering Approach

RATIONALE: Numerous data regarding both clinical presentation and prognosis of COVID-19 have been published. Most studies focused on individual predictors for mortality. Although some prognostic factors were consistently identified across the different studies such as older age or cardiovascular comorbidities, other discrepancies reflect geographical location of the studies, characteristics of study population, admission in wards and/or intensive care units, and variables incorporated in the statistical model. We aimed to a priori identify specific patient profiles, then assessing their association with the outcomes in COVID-19 patients with respiratory symptoms admitted specifically in hospital wards. METHODS: We conducted a retrospective single center study from February, 27, 2020 to April, 27, 2020. A non-supervised cluster analysis was first used to detect patient profiles based on characteristics at admission of 220 consecutive patients admitted at our institution. Then, we assessed its prognostic value, using Cox regression analyses to predict survival. RESULTS: Three clusters were identified, with 47 patients in cluster 1, 87 in cluster 2, and 86 in cluster 3, and whose presentation differed. Cluster 1 mostly included sexagenarian patients with active malignancy who were admitted early after COVID-19 onset. Cluster 2 included the oldest, overweight patients with high blood pressure and renal insufficiency, while cluster 3 included the youngest patients with gastrointestinal symptoms and delayed admission. These subgroups of patients were associated with different outcomes, with 60 days survival of 74.3% (cluster1), 50.6% (cluster2) and 96.5% (cluster3) (figure 1). This was confirmed by the multivariable Cox analyses that exhibited the prognostic value of those patterns. CONCLUSION: The cluster approach seems appropriate and pragmatic to early identify patient profiles that could help physicians to segregate patients according to their prognosis. Figure 1: Survival since hospital admission according to the clusters .