Highlights
In a multinational CCCTN cohort of 3282 CICU admissions for cardiogenic shock from 2019 to 2023, use of vasoactive agents and inodilators varied substantially across institutions.
Patient characteristics explained much of inodilator selection, but center-level practice patterns also accounted for a notable proportion of variation, suggesting that clinician and institutional preferences remain influential.
History of heart failure, biventricular failure, valvular disease, and Society of Cardiovascular Angiography and Interventions (SCAI) stage D were associated with greater inodilator use, whereas cardiac arrest, right ventricular failure, SCAI stage E, acute myocardial infarction-related shock, older age, and better kidney function were associated with lower use.
The choice between dobutamine and milrinone also differed across centers, reinforcing the need for higher-quality comparative effectiveness and randomized trial data.
Clinical Context
Cardiogenic shock remains one of the most dangerous syndromes encountered in contemporary cardiac intensive care. It is defined by inadequate tissue perfusion due to primary cardiac dysfunction and carries high short-term mortality despite advances in reperfusion, mechanical circulatory support, and critical care. In practice, clinicians often must make rapid decisions about vasopressors, inotropes, and inodilators with incomplete evidence, substantial physiologic heterogeneity, and limited consensus regarding the optimal drug for a given phenotype.
This problem is especially relevant because cardiogenic shock is not a single disease. Patients may present with acute myocardial infarction-cardiogenic shock (AMI-CS), acute decompensated heart failure, valvular catastrophe, fulminant myocarditis, or post-cardiac arrest shock. Hemodynamic profiles differ widely: some patients are vasoconstricted and hypoperfused, others are vasodilated, and many have mixed left, right, or biventricular failure. In such settings, the appeal of tailoring vasoactive therapy to individual physiology is obvious, but the scientific foundation for specific agent selection remains thin.
Against this background, the study by Hamilton and colleagues provides an important descriptive view of how clinicians actually treat cardiogenic shock across modern cardiac ICUs and how much of treatment selection appears related to patient factors versus institutional practice style.
Study Design and Methods
This analysis used data from the Critical Care Cardiology Trials Network (CCCTN), an international, multicenter collaborative network of cardiac intensive care units coordinated by the TIMI Study Group. The investigators evaluated CICU admissions for cardiogenic shock between 2019 and 2023 across 37 centers, yielding a final cohort of 3282 admissions.
The principal question was not whether one drug improved survival over another, but rather how variable vasoactive treatment patterns were in routine care and what factors were associated with use of inodilators, defined here as dobutamine or milrinone. The investigators also examined variation in the selection of dobutamine versus milrinone among patients receiving inodilator therapy.
To address these questions, the study used multivariable mixed-effects logistic regression models. This approach is appropriate for practice-variation research because it allows simultaneous assessment of patient-level covariates and random effects attributable to individual hospitals or CICUs. In practical terms, it helps estimate how much observed treatment variation can be explained by measurable clinical features and how much likely reflects local practice culture, clinician preference, or unmeasured institutional factors.
Key Results
Substantial variation in vasoactive and inodilator use across centers
The headline finding is straightforward and important: vasoactive treatment selection in cardiogenic shock varied markedly across institutions. This was true not only for overall vasoactive use but also specifically for inodilator therapy and the choice of dobutamine versus milrinone.
Such variability is clinically meaningful because these agents are not interchangeable in all scenarios. Dobutamine is often favored when rapid beta-adrenergic inotropic support is desired and blood pressure can tolerate it. Milrinone, a phosphodiesterase-3 inhibitor, may be preferred in some patients with chronic beta-blocker exposure or pulmonary vascular dysfunction, but its longer half-life and renal clearance complicate use in unstable or renally impaired patients. The fact that major differences in use persist across centers strongly suggests unresolved uncertainty rather than settled evidence-based consensus.
Patient factors associated with greater inodilator use
Several clinical features were independently associated with increased odds of receiving an inodilator. The strongest association was a history of heart failure, with an odds ratio of 1.98 (95% CI, 1.61-2.44). This is intuitive: clinicians may be more likely to view inodilators as appropriate in chronic or acute-on-chronic pump failure phenotypes, where low output rather than profound vasoplegia predominates.
Biventricular failure was also associated with greater inodilator use (odds ratio, 1.59; 95% CI, 1.27-2.00), suggesting that clinicians may use these agents when global myocardial dysfunction is prominent. Other factors associated with higher inodilator use included SCAI stage D shock (odds ratio, 1.34; 95% CI, 1.07-1.68), valvular disease (odds ratio, 1.34; 95% CI, 1.03-1.74), and male sex (odds ratio, 1.23; 95% CI, 1.02-1.49).
The sex association is noteworthy but should be interpreted cautiously. It may reflect residual confounding, differences in phenotype presentation, referral patterns, or treatment bias rather than a biologically meaningful preference for men. The study was not designed to explain this finding mechanistically.
Patient factors associated with lower inodilator use
In contrast, several variables were associated with lower odds of inodilator use. Cardiac arrest showed the strongest inverse association (odds ratio, 0.33; 95% CI, 0.27-0.42). This likely reflects the complex physiology of postarrest shock, where vasopressor support often takes priority because of severe vasoplegia, neurologic uncertainty, or concern about arrhythmias and hypotension.
Right ventricular failure was associated with less inodilator use (odds ratio, 0.50; 95% CI, 0.33-0.73), a finding that may seem counterintuitive because inodilators can reduce pulmonary vascular resistance and support the right ventricle. However, this result may indicate clinician concern about systemic hypotension, a preference for alternative support strategies, or the difficulty of phenotype classification in real-world care.
Patients in SCAI stage E had lower inodilator use than others (odds ratio, 0.57; 95% CI, 0.41-0.81), consistent with a shift toward vasopressors, mechanical circulatory support, or both in the most extreme shock states. AMI-CS was also associated with less inodilator use (odds ratio, 0.71; 95% CI, 0.56-0.90), which may reflect more common use of vasopressors in infarct-related hypotension and concern about tachyarrhythmia or ischemia with some inotropes.
Older age (odds ratio, 0.77 per 10-year increase; 95% CI, 0.72-0.83), peripheral arterial disease (odds ratio, 0.73; 95% CI, 0.54-0.99), and estimated glomerular filtration rate (odds ratio, 0.96 per 10 mL/min per 1.73 m2 increase; 95% CI, 0.93-0.99) were also associated with lower inodilator use. The eGFR association is interesting because milrinone is renally cleared, yet the reported direction suggests more complex prescribing behavior than a simple pharmacokinetic rule.
How much variation came from patients versus centers?
The mixed-effects modeling provides one of the study’s most useful insights. For overall inodilator use, 45.7% of variation was attributed to patient-level factors, while 22.7% was attributed to the random effect of individual CICU centers. For the choice between dobutamine and milrinone, 35.3% of variation was attributed to patient-level factors and 32.6% to individual centers.
These figures indicate that patient phenotype matters, as expected, but they also show that center-specific practice style is nearly as important when selecting between dobutamine and milrinone. Put differently, two otherwise similar patients might receive different inodilator strategies depending on where they are admitted.
Importantly, no measured institution-level characteristic, such as transplant center status, explained the variability. That implies the driver is less likely to be an obvious structural feature and more likely to be embedded local culture, attending preference, training background, hemodynamic philosophy, or unmeasured care processes.
Clinical Interpretation
This study does not tell clinicians which vasoactive agent is best. Instead, it clarifies just how much uncertainty currently shapes practice. The findings fit with long-standing frustrations in cardiogenic shock care: despite high disease severity and major physiologic consequences of drug choice, comparative evidence remains sparse, and recommendations often rely on limited trials, observational data, and expert opinion.
From a bedside perspective, the observed patterns make clinical sense in several areas. Inodilators were more commonly used in chronic heart failure-like shock phenotypes and less commonly in postarrest, AMI-CS, and extremis presentations. That suggests clinicians are using broad physiologic heuristics. Yet the magnitude of center effect reveals that these heuristics are not standardized.
The variation between dobutamine and milrinone is especially relevant. Contemporary practice often treats these agents as reasonable alternatives, but each carries distinct hemodynamic tradeoffs. Dobutamine has a shorter half-life and may be easier to titrate quickly; milrinone may be attractive in patients on chronic beta-blockade or with pulmonary hypertension, but it can accumulate in renal dysfunction and may worsen hypotension. In the absence of definitive outcome data, many decisions understandably remain preference-sensitive.
Strengths and Limitations
The study has several strengths. It draws from a large, contemporary, international multicenter CICU network and reflects real-world cardiogenic shock practice rather than narrowly selected trial populations. The use of mixed-effects modeling is methodologically appropriate for disentangling patient and center contributions to variation.
At the same time, the study is observational and primarily descriptive. It cannot establish that one treatment pattern is superior or inferior. Residual confounding is inevitable, especially in shock, where unmeasured bedside factors such as invasive hemodynamics, lactate trajectory, vasoactive dose intensity, mechanical support timing, clinician gestalt, and perceived reversibility often drive treatment choices. The analysis also cannot fully capture the dynamic nature of shock management, in which multiple vasoactive agents may be started, stopped, and sequenced over hours.
Another limitation is that center-level random effects show variability but do not reveal the specific local mechanisms behind that variability. Future work should examine whether practice differences relate to standardized protocols, pulmonary artery catheter use, mechanical circulatory support availability, or interdisciplinary shock team activation.
How This Fits With Existing Evidence and Guidelines
Current cardiogenic shock guidance generally supports norepinephrine as a first-line vasopressor when hypotension is prominent, while inotropes are considered when there is persistent hypoperfusion from low cardiac output. However, major guidelines acknowledge that the evidence base for selecting individual inotropes and inodilators is limited.
Randomized data comparing specific inotropes in cardiogenic shock are sparse and often underpowered for hard outcomes. The DOREMI trial, which compared milrinone with dobutamine in cardiogenic shock, found no significant difference in a composite primary outcome, but it was not definitive for all relevant subgroups and has not eliminated practice variation. The current CCCTN analysis suggests that uncertainty seen in trials and guidelines translates directly into heterogeneous bedside care.
Implications for Practice and Research
For clinicians, the immediate implication is not that variation itself is wrong, but that variation should ideally reflect patient physiology rather than local habit. Centers may benefit from reviewing their own vasoactive prescribing patterns, especially if decisions between dobutamine and milrinone are inconsistent or poorly protocolized. Shock teams, structured hemodynamic assessment, and explicit phenotype-based treatment pathways may help reduce unwarranted variability.
For researchers, the message is clear: comparative effectiveness work in cardiogenic shock remains urgently needed. Future studies should move beyond broad syndrome labels and test therapies within clinically relevant phenotypes such as AMI-CS, acute decompensated heart failure shock, right ventricular failure, biventricular failure, and postarrest shock. Adaptive platform trials, embedded pragmatic trials, and analyses linked to invasive hemodynamic data may be especially valuable.
Outcomes of interest should include not only mortality, but also time to shock resolution, need for mechanical circulatory support escalation, arrhythmias, renal injury, and organ recovery. The goal is not simply to identify a universally “best” inodilator, but to determine which agent is best for which patient and at what stage of shock.
Conclusion
The CCCTN analysis by Hamilton and colleagues shows that vasoactive treatment selection in cardiogenic shock remains highly variable across contemporary CICUs. Patient characteristics explain a substantial portion of this variation, but center-level practice style also plays a major role, particularly in the choice between dobutamine and milrinone. These findings highlight a persistent evidence gap in one of the most consequential decisions in cardiac critical care. Until stronger comparative data are available, clinicians should anchor vasoactive selection to careful hemodynamic phenotyping, structured reassessment, and transparent local protocols designed to minimize unwarranted practice variation.
Funding and ClinicalTrials.gov
The abstract does not report a funding source for this specific analysis. No ClinicalTrials.gov registration number is provided, which is reasonable given the observational registry-based design.
References
1. Hamilton DE, Shriver JL, Patel SM, Park JG, Michos ZE, Mathis MR, Adie SK, Alviar CL, Barnett CF, Berg DD, Bennett CE, Bohula EA, Carnicelli AP, Daniels LB, Dodson MW, Gage A, Gidwani U, Goldfarb M, Katz JN, Ketcham SW, Kwon Y, Leibner ES, Loriaux DB, Luk A, Marano P, Miller PE, Mukundan SV, Papolos AI, Pisani BA, Proudfoot AG, Roswell RO, Shah KS, Solomon MA, Tomey MI, van Diepen S, Zakaria S, Morrow DA, Thompson AD, CCCTN Investigators. Variation in Vasoactive Treatment Selection for Cardiogenic Shock: Insights From the Critical Care Cardiology Trials Network (CCCTN). Circulation. Heart failure. 2026-05-07:e013778. PMID: 42093634.
2. Mathew R, Di Santo P, Jung RG, Marbach JA, Hutson J, Simard T, Ramirez FD, Harnett DT, Merdad A, Almufleh A, Weng W, Abdel-Razek O, Di Ciano P, Perry Arnesen H, Sellers SL, Appadurai V, Bainey KR, Goodman SG, Lala A, Van Diepen S, Overgaard CB, Hibbert B, So DYF, Froeschl M, Chan V, Labinaz M, Glover C, Dick A, Marquis JF, O’Brien ER, Gallo R, Coppens M, Whitlock R, Devereaux PJ, Lawler PR, Teoh K, Rao V, Velianou JL, Stadnick E, Abdel-Qadir H, Mazer CD, Takeda K, Rao-Melacini P, Luk A, Fan E, Natarajan MK, Goodman SG, Ko DT, Austin PC, Udell JA. Milrinone as Compared with Dobutamine in the Treatment of Cardiogenic Shock. N Engl J Med. 2021;385:516-525.
3. van Diepen S, Katz JN, Albert NM, Henry TD, Jacobs AK, Kapur NK, Kilic A, Menon V, Ohman EM, Sweitzer NK, Thiele H, Washam JB, Cohen MG. Contemporary Management of Cardiogenic Shock: A Scientific Statement From the American Heart Association. Circulation. 2017;136:e232-e268.
4. Thiele H, Ohman EM, de Waha-Thiele S, Zeymer U, Desch S. Management of cardiogenic shock complicating myocardial infarction: an update 2019. Eur Heart J. 2019;40:2671-2683.
5. Garan AR, Kanwar M, Thayer KL, Whitehead E, Zweck E, Hernandez-Montfort J, Sinha SS, Stelling K, DeFilippis EM, Mahr C, Khandelwal A, Badiye A, Kapur NK. Complete hemodynamic profiling with pulmonary artery catheters in cardiogenic shock is associated with lower in-hospital mortality. JACC Heart Fail. 2020;8:903-913.
