We read with great interest the retrospective cohort study by Sarma et al (Am Heart J 2025;286:88-96) examining the association between vasopressin administration and in-hospital mortality in patients with cardiogenic shock (CS). This study addresses a critical and timely issue in critical care medicine. As anesthesiologists and intensivists at Helios Klinikum Schleswig, where we frequently manage patients with various shock states, including CS, we recognize the importance of exploring alternative vasoactive strategies like vasopressin. We commend the authors for their efforts and offer the following constructive commentary on the study’s strengths, limitations, and clinical implications.
The study’s use of a large cohort (n = 721) and robust statistical methods, including multivariable adjustment and propensity score matching, enhances its internal validity. The subgroup analyses of patients receiving high-dose vasopressors (HDV) and those classified as SCAI Stage C are noteworthy for identifying potential populations that may benefit from vasopressin, addressing a significant therapeutic challenge in clinical practice.
The authors appropriately acknowledged several key limitations, including the retrospective design, lack of standardized vasopressor protocols, and absence of invasive hemodynamic data. We agree that these factors constrain the interpretation of the observed associations. However, we believe their implications—particularly regarding the potential for indication bias, evolving clinical practices, and challenges in phenotyping mixed shock—warrant further emphasis. In our view, these limitations highlight the urgent need for prospective studies with standardized protocols and precise hemodynamic profiling to truly clarify vasopressin’s role in cardiogenic shock management.
Specifically, the absence of a standardized vasopressor protocol introduces potential indication bias, especially given the higher sepsis prevalence in the vasopressin group (33.8% vs 17.5%), suggesting that vasopressin may have been preferentially used in patients with suspected distributive shock features. Despite statistical adjustments, unmeasured confounders may persist.
The study period (2007-2015) reflects outdated practices, notably the prevalent use of dopamine (57.6% of patients), which has since been largely replaced by norepinephrine based on evidence of harm in CS (eg, SOAP-II trial). Notably, lower dopamine use in the vasopressin group (44% vs 63%) may partly explain improved outcomes, potentially reflecting reduced catecholamine toxicity rather than a direct vasopressin effect.
Furthermore, the lack of invasive hemodynamic data, such as pulmonary artery catheter measurements, limits the ability to confirm the hypothesis that vasopressin benefits patients with mixed cardiogenic-vasodilatory shock. Noninvasive SVR estimates via echocardiography were inconclusive, leaving patient phenotyping speculative. In clinical practice, precise hemodynamic profiling is essential prior to initiating vasopressin, which may significantly increase SVR and afterload.
The study also omits important data on vasopressin dosing, timing beyond the first 24 hours, and adverse effects such as coronary vasoconstriction or myocardial ischemia—considerations that are particularly relevant in CS patients. The borderline statistical significance in the adjusted analysis (OR 0.59, 95% CI 0.35-0.99; P =.05) and lack of correction for multiple comparisons, especially in the SCAI Stage C subgroup (adjusted OR 0.408, P =.05), suggest caution due to the risk of type I error.
In our ICU, we cautiously use vasopressin in patients on high-dose catecholamines with suspected vasodilatory features, but limited evidence in CS mandates prudence. While Sarma et al’s findings are encouraging and hypothesis-generating, they do not yet establish vasopressin’s role in routine CS management.
We strongly support the authors’ call for prospective randomized controlled trials (RCTs) to clarify vasopressin’s utility. Such trials should incorporate modern vasopressor protocols, invasive hemodynamic monitoring, detailed CS phenotyping, and evaluation of adverse effects. We also recommend the use of statistical corrections such as Bonferroni to mitigate type I error in subgroup analyses, and advocate for further research into optimal dosing and timing of vasopressin administration.
In conclusion, Sarma et al provide valuable preliminary evidence suggesting a potential mortality benefit of vasopressin in CS patients requiring HDV. However, the retrospective design, outdated practices, and physiologic uncertainties limit applicability. This study is a significant contribution and should serve as a foundation for future research rather than a basis for immediate clinical change.
Sincerely,
Niloufar Dadashpour
Majid Golestanieraghi
Authorship statement
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
