Milrinone or Dobutamine After Bypass: The Same Choice, Pointed by Two Different Risks
Two patients come off bypass in low cardiac output syndrome needing inotropic support. One's borderline kidneys argue against a renally cleared drug; the other's recent ventricular arrhythmia argues against a catecholamine — and between them, neither inotrope is the safe default.
Rosa D., a 74-year-old woman, has watched her grandchildren three afternoons a week since she retired from three decades of hospital-lab work, and came in for isolated CABG after an abnormal stress test found severe three-vessel disease she had no symptoms from until a routine cardiology follow-up flagged it. Her chronic kidney disease — stage 3b, creatinine 1.6, eGFR 34, attributed to longstanding hypertension — was known and stable going into surgery, not itself a reason to avoid the operation, but directly relevant to what happens now that she is struggling to come off bypass. Her preload has already been optimized and her rhythm is sinus, with no arrhythmia documented anywhere in her chart — two explanations the team has excluded before reaching for any drug at all.
Her cardiac index sits at 1.9 L/min/m² despite an already-optimized preload and rhythm — a genuine low cardiac output state rather than a volume or rhythm problem the team can correct without pharmacologic support. The two standard inotropes carry different liabilities in her specific case: milrinone is predominantly renally cleared, which runs directly against an eGFR of 34 and raises the prospect of accumulation and a hypotension slow to titrate back once the drug builds up in a kidney that cannot eliminate it at the expected rate. The literature the team will reach for does not describe her cleanly. DOREMI, the largest head-to-head comparison of these two drugs, enrolled patients in cardiogenic shock of predominantly ischemic origin — not patients in post-cardiotomy low output syndrome an hour off cardiopulmonary bypass, which is a different physiology arriving by a different route. Its finding of no significant difference is a reason not to treat either drug as generally superior; it is not a finding about her.
Post-bypass low cardiac output with borderline renal function
Dobutamine for Rosa, not milrinone. Her eGFR is 34 and milrinone is predominantly renally cleared — giving it here risks real accumulation and a prolonged hypotension we'd struggle to titrate back once it builds up in a kidney that can't eliminate it at the expected rate.
I don't think the renal argument carries as much as you're making it carry. Milrinone's clearance being renal doesn't mean it can't be given to a patient with an eGFR of 34 — it means the dose comes down and the infusion doesn't get loaded. We titrate it. What you're describing as an untitratable accumulation is a dosing problem with a known answer, not a contraindication.
And I'd push back on treating dobutamine as the safe default just because it isn't renally cleared. She's 74 with three-vessel disease and myocardium that has just come off a cross-clamp. Direct beta-1 agonism raises myocardial oxygen demand in exactly the territory we've just revascularized, and it drives heart rate in a patient whose diastolic filling time is doing real work at a cardiac index of 1.9. Dobutamine has liabilities here too; they're just not renal ones, so they're easier to overlook.
That's fair on the titration point, and I'll take it — I was treating renal clearance as disqualifying when it's really a reason to dose differently. Where I still land on dobutamine is narrower than where I started: it isn't that milrinone can't be given safely, it's that if she does become hypotensive on it, her kidneys are the reason we'll be waiting longer than usual for that to resolve, and we're weaning her off bypass tonight rather than managing her over days. I'd rather hold the drug with the shorter offset in a patient whose offset I can't shorten.
On the oxygen-demand point, I don't have a good answer, and I don't want to pretend I do. We'll keep her rate under 100 and accept that I'm trading one real risk for another rather than avoiding risk.
Marcus T., a 55-year-old man, sustained a witnessed cardiac arrest four weeks ago while coaching his son's Little League practice — ventricular tachycardia degenerating into ventricular fibrillation, resuscitated on scene by a parent who happened to be an ICU nurse, and found on workup to reflect a large area of scarred myocardium from a silent prior infarct nobody had known about. An ICD was placed at that admission and interrogated again this morning, confirming appropriate sensing and no recorded arrhythmia in the four weeks since — a quiet month, but a short one. Today he is undergoing surgical revascularization of the vessel responsible for that scar, on otherwise normal renal function with no prior kidney disease of any kind — a clean clearance profile that will turn out to matter as much to his plan as the arrest itself.
Coming off bypass, his cardiac index is similarly reduced — 1.8 L/min/m², reflecting real stunning of already-scarred myocardium from the bypass run itself — and the same two inotropes are back on the table, but the deciding factor for him runs in the opposite direction from Rosa's. His creatinine of 0.9 and eGFR of 88 remove the clearance liability entirely, so the renal argument that shaped her plan this morning has no purchase at all on his. What he carries instead is a heart that demonstrated, within the past month, that it can degenerate into a lethal arrhythmia under enough physiologic stress — and DOREMI, whose null result the team has already invoked once today, enrolled shock patients rather than patients selected for arrhythmic vulnerability, so it offers nothing on the one variable that actually distinguishes him. The same two drugs, the same low-output physiology, and a decision that has to rest on his own four-week-old history rather than on a trial never built to answer the question he actually poses.
Post-bypass low cardiac output with a recent ventricular arrhythmia history
Milrinone for Marcus, not dobutamine. He had a sustained VT arrest a month ago from scarred myocardium, and dobutamine's direct beta-1 agonism carries real, well-documented arrhythmogenic risk — exactly the wrong profile for a heart that's already proven it can degenerate under stress. Milrinone works through PDE3 inhibition, bypassing the beta-receptor pathway entirely.
I want to be careful about how cleanly we're separating those two drugs on arrhythmia. Milrinone is not an antiarrhythmic. PDE3 inhibition raises intracellular cyclic AMP and calcium the same way beta-1 stimulation does — it just arrives at that endpoint without going through the receptor. The arrhythmia signal for milrinone in the heart failure literature is real, and bypassing the beta receptor isn't the same as bypassing the mechanism that makes an irritable ventricle fire.
I'd also say the ICD cuts against your argument rather than for it. He has a device in place that was interrogated this morning and will treat a recurrence. That doesn't make an arrhythmia harmless, but it does mean an arrhythmia is a monitored, treatable event for him specifically in a way it wasn't four weeks ago on a Little League field — which is different from how you're weighting it.
You're right on the mechanism and I overstated it — milrinone raises calcium too, and I shouldn't have implied the beta receptor was the only route to an arrhythmia. I'll concede that. Where I don't move is the ICD point. The device treats a sustained arrhythmia; it doesn't prevent the hemodynamic collapse in the seconds before it fires, and shocking a freshly revascularized heart in the ICU is a cost, not a neutral safety net. I'd still rather not hand a demonstrated beta-1 responder direct beta-1 agonism on the day we've stunned his myocardium.
So milrinone, but not because it's arrhythmia-free — because the difference in arrhythmic risk still points the same direction, and his kidneys remove the one liability that made us avoid it for Rosa this morning. Same two drugs, same low-output physiology, opposite conclusion, and I want the record to show it was a closer call than it looked at the start.
Agreed for both patients, with the two drugs assigned in opposite directions: Rosa received dobutamine given her reduced renal clearance; Marcus received milrinone given his recent sustained ventricular arrhythmia. Both plans were framed explicitly as the same underlying principle — matching the inotrope's specific liability against each patient's specific risk — rather than either drug being treated as a default. Neither assignment was arrived at cleanly: the renal argument for Rosa was conceded to be a dosing problem rather than a contraindication, and the arrhythmia argument for Marcus was conceded to be a difference of degree rather than a difference in kind, since milrinone raises intracellular calcium by its own route.
Not addressed as a general protocol: whether this patient-specific matching approach should be formalized into a standing algorithm for LCOS inotrope selection, or continue to be decided case by case as it was today. Both voices treated today's reasoning as sound without committing to writing it into a fixed institutional pathway.