Predictors of Death in Comorbid Patients with Thrombotic Complications: a Single-Center Retrospective Cross- Sectional Study

Abstract

   INTRODUCTION. Despite the improvements and success in the management of thrombosis, the prognosis of thrombotic complications in comorbid patients remains extremely unfavorable. Understanding this problem suggests that it is rational to pay more attention to the prevention of thrombosis in order to avoid thrombotic complications in the first place, and there is an urgent need to improve the ability to predict the development of complications and, most importantly, death. Therefore, some inexpensive, but effective methods of risk assessment need to be developed and integrated in clinical practice.   AIM OF STUDY. To assess which laboratory markers can be associated with higher probability of fatal outcome in patients with thrombotic complications.   MATERIAL AND METHODS. The retrospective cross-sectional study included 283 patients who were admitted to the N.V. Sklifosovsky Research Institute for Emergency Medicine in 2021. Minimum age was 21 years, maximum age was 96 years, median age – 68 (59,0; 76,5) years. 161 (56,9 %) patients were males, 122 (43,1 %) – females. Group I included 226 patients with thrombotic complications; Group II included 57 (20,1 %) patients who died during hospitalization. The patients were diagnosed with the following thrombotic complications: pulmonary embolism (PE) was diagnosed in 28 (9.9 %), acute coronary syndrome with ST segment elevation in 38 (13.4 %), arterial thrombosis in 84 (29.7 %), venous thrombosis in 54 (19.1 %), thrombophlebitis in 22 (7.8 %), and systemic thromboembolism in 7 (2.5 %) patients.   RESULTS. It was established that the decrease in the estimated glomerular filtration to the levels of stage 3 (and lower) chronic kidney disease (AUC — 0,881, sensitivity — 86.92 %, specificity — 85.71 %), hemoglobin level — to less than 120 g/L (AUC — 0,690, sensitivity — 77,21 %, specificity — 55,1 %), as well as hypoalbuminemia (AUC — 0,905, sensitivity — 78,89 %, specificity — 91,18 %) statistically significantly increased the likelihood of death by 19 (OR 19.276, 95 % CI [7,792–47.687], p < 0.001), 4 (OR 4.158, 95 % CI [2.177–7.939], p < 0.001), and 30 (OR 30,000, 95 % CI [9.93–90,610], p < 0.001) times, respectively. The statistical analysis using the univariate logistic regression model revealed that such diseases as coronary artery disease (OR 8,6, 95 % CI [2,6–28,466], p < 0.001), chronic heart failure (CHF) (OR 13,714, 95 % CI [4,784–39,313], p < 0.001), atrial fibrillation (OR 3,455, 95 % CI [1,830–6,525], p < 0.001), type 2 diabetes (OR 2.5, 95 % CI [1,286–4,858[, p = 0.007), postinfarction cardiosclerosis (OR 3,734, 95 % CI [1,953–7,142], p < 0.001), and previous stroke (OR 3,319, 95 % CI [1,519–6,490], p = 0.002) made an independent contribution to death prediction. During the study, we calculated the patients’ serum albumin-to-creatinine ratio (sACR). ROC analysis revealed a cut-off point for sACR to be 0.33 g/mmol as having the best predictive ability of death (AUC — 0.920, 84.3 % sensitivity, 85.29 % specificity). It was established that sACR less than 0.33 g/mmol increased the probability of death by 26 times (OR 26.3806, 95 % CI [9.4573.57], p < 0.001).   CONCLUSION. Serum albumin-to-creatinine content ratio can be used as a predictor of fatal outcome in comorbid patients with thrombotic complications.