Prediction of in-hospital mortality in patients aged 75 years and older with acute ST-segment elevation myocardial infarction using logistic regression and classification tree

• Published in: Kardiovaskuli͡a︡rnai͡a︡ terapii͡a︡ i profilaktika Vol. 23; no. 3; p. 3896
• Main Authors: Pereverzeva, K. G., Yakushin, S. S.
• Format: Journal Article
• Language: English; Russian
• Published: SILICEA-POLIGRAF» LLC 01.04.2024
• Subjects: cardiogenic shock; chronic heart failure; diabetes mellitus; leukocytes; prognosis; st-segment elevation myocardial infarction; ventricular tachycardia
• ISSN: 1728-8800; 2619-0125
• DOI: 10.15829/1728-8800-2024-3896

Aim. To study risk factors for in-hospital mortality in patients with acute myocardial infarction (MI) with ST-segment elevation (STEMI) aged ≥75 years. Material and methods. This retrospective analysis analyzed medical records of 174 patients aged ≥75 years hospitalized at the regional cardiology clinic due to STEMI in 2020-2021. There were 52 (29,9%) men. The median age for all patients was 81 [79; 85] year. Results. Inhospital mortality was 24,1%. Binary logistic regression revealed the following factors increasing the risk of inhospital mortality: cardiogenic shock (CS) — odds ratio (OR) 39,04; 95% confidence interval (CI): 8,87-171,74 (p<0,001); ventricular tachycardia (VT) — OR=60,52; 95% CI: 5,21-703,45 (p=0,001); white blood cell (WBC) count — for each increase by 1×10 9 /l — OR=1,15; 95% CI: 1,05-1,26 (p=0,002). At the same time, history of heart failure (CHF) reduced the death risk (OR=0,04; 95% CI: 0,009-0,21 (p<0,001)). Classification tree found that 100% mortality was predicted in patients with CS and without diabetes and a history of MI and in patients with pulmonary edema and without CS, VT and HF. Mortality of 83,3% was predicted in patients without CS and without VT with a history of HF, WBC count ≥14,5×10 9 /L and body mass index ≤23,7 kg/m 2 . In addition, mortality of 75,0% was predicted in patients without CS, but with VT, as well as in patients with CS without diabetes and with a history of MI. In patients with CABG and diabetes, the predicted mortality rate was 50,0%. Conclusion. The strongest factors associated with the risk of inhospital death were CS, VT, and high WBC count. The positive effect on the prognosis of pre-existing diseases, in particular HF, is apparently associated with the protective effect of previously taken therapy. This assumption needs to be confirmed based on prospective studies.