Inotropes and mortality in patients with cardiogenic shock: An instrumental variable analysis from the SWEDEHEART registry

ABSTRACT Background The use of inotropic agents in treating cardiogenic shock (CS) remains controversial. We aimed to investigate the effect of treatment with inotropes on 30-day mortality in patients with CS from the SWEDEHEART registry (The Swedish Web-system for Enhancement and Development of Evidence-based care in Heart disease Evaluated According to Recommended Therapies). Methods We used data from the national SWEDEHEART registry on all patients diagnosed with CS in Sweden between 2000 and 2022. The primary endpoint was 30-day all-cause mortality. We used multilevel Cox proportional-hazards regression with instrumental variable and inverse probability weighting propensity score to adjust for known and unknown confounders. The treatment-preference instrument was the quintile of preference for using inotropes at the treating hospital. Results In total, 16,214 patients (60.5% men and 39.5% women) were included; 23.5% had diabetes, 10.2% had a previous myocardial infarction (MI), and 13.8% had previous heart failure (HF). The median age was 70 years (interquartile range; 19), and 66.4% were >70. Acute coronary syndrome (ACS) was the cause of CS in 82.9% of patients. Inotropic agents were used in 43.8% of patients, while 56.2% did not receive inotropic agents. There were 7,875 (48.1%) deaths. On average, patients treated with inotropes were two years younger and more likely to have ACS. Patients not treated with inotropes were more likely to have previous MI and previous PCI but less likely to undergo PCI. The number of patients with CS decreased by 12% per year (P trend <0.001). There was a considerable variation between hospitals in the preference for using inotropes ranging from 25 to 78% (P<0.001). Inotropes increased by 5% per year (P trend <0.001). The unadjusted mortality in CS increased by 2% per calendar year (P trend <0.001). The risk of death was higher in patients treated with inotropes [adjusted hazard ratio (HR adj ) 1.72; 95% confidence interval (CI) 1.26-2.35; P=0.001]. There was a quantitative interaction between inotrope treatment and age and diagnosis (P interaction < 0.001 and P interaction = 0.018, respectively). Conclusions In this observational study, using inotropes was associated with a higher mortality risk in patients with CS. The increased risk of death was more pronounced in patients younger than 70. The number of patients with CS is decreased, while the use of inotropes and mortality increased in Sweden.