Evaluation of the Modified HASBLED Score for Prediction of In-hospital Mortality in Patients with COVID-19

Abstract

Objective: The COVID-19 outbreak continues to be the common cause of deaths worldwide in recent times. Preventing poor outcomes (death, intubation, non-invasive ventilation, need for intensive care) is the first goal for hospitalized patients. Identifying high-risk patients during hospitalization can provide more effective follow-up and treatment. The HASBLED score is highly predictive for bleeding events in atrial fibrillation patients. We used the HASBLED score to identify patients with frailty, vulnerability, and comorbid diseases, not as a bleeding score. So, we used albumin level instead of labile INR in the score because it shows both the nutritional status and inflammation. We aim to evaluate the modified HASBLED score for predicting mortality and poor outcomes in hospitalized COVID-19 patients. Method: In total, 2850 hospitalized COVID-19 patients were screened retrospectively and, after the exclusions, 2041 patients were included in the study. The patients were divided into two groups according to the M-HASBLED score as <3 and ≥3. The demographic, laboratory, and clinical characteristics of the patients were obtained from the hospital registry system and the database of the Turkish Ministry of Health. Patients with two negative PCR tests and patients with missing data were excluded. Results: 582 patients were included in the M-HASBLED RS ≥3 group, and 1459 patients in the M-HASBLED RS<3 group. Hypertension, diabetes mellitus, coronary artery disease, heart failure, cerebrovascular disease, and chronic renal failure were higher in the M-HASBLED RS≥3 group. In-hospital stay (7 (6-11), 9.5 (7-15) IQR), need for ICU (21.9%, 54%), NIMV (14%, 29.2%), and intubation (13%, 36.6%) were higher in the M-HASBLED RS ≥3 group compared to the other group. Death was observed in 208 (14.3%) patients in the M-HASBLED RS <3 group, and in244 (41.9%) patients in the M-HASBLED RS ≥3 group, and it was statistically higher in the M-HASBLED RS≥3 group (p <0.001). The M-HASBLED score was found as an independent factor associated with in-hospital mortality (OR: 1.20, 95% confidence interval CI: 1.03-1.37, p=0.010). Conclusion: M-HASBLED RS showed poor in-hospital outcomes accurately and is an independent factor associated with mortality in hospitalized COVID-19 patients.