Optimizing the Four Pillars: How Long to Keep Titrating Before Reassessing a Recovering Ejection Fraction
A newly diagnosed heart failure patient's ejection fraction has climbed from twenty percent to thirty-three over twelve weeks of drug therapy — real improvement, but still below the threshold guidelines use to reconsider a defibrillator. The disagreement isn't whether to keep adjusting his medications. It's whether three separate dose-limiting events already mean he's reached his ceiling, or whether concluding that this early would be premature.
R.C., a 54-year-old man, has directed his high school's marching band for eighteen years, a job that keeps him on his feet on the practice field most fall afternoons. He'd carried a diagnosis of hypertension for about a decade, treated on and off — lisinopril prescribed years earlier, refilled less consistently than his primary care doctor would have liked. Roughly fourteen weeks ago, in the middle of preparing his students for the season's first competition, he noticed he could no longer cross the field without stopping to catch his breath. Over the following two weeks the breathlessness worsened until he couldn't lie flat without waking up gasping, and his ankles had swollen enough that his usual shoes no longer fit. He came to the emergency department on a Tuesday afternoon and by that evening had been diagnosed with new-onset heart failure with severely reduced systolic function — an EF of 20%, NYHA class III — with a cardiac catheterization the same admission showing no obstructive coronary disease. The picture was non-ischemic dilated cardiomyopathy, with longstanding, undertreated hypertension a real likely contributor, though no single cause fully explained a heart this weakened this quickly.
He spent four days on the inpatient service being decongested and started on low-dose carvedilol before discharge. What has happened since is the part that matters now. Two weeks out, his outpatient heart failure clinic switched him from lisinopril to sacubitril/valsartan, started at 49/51mg twice daily, and added spironolactone at 12.5mg daily; carvedilol was advanced to 6.25mg twice daily the same visit. Six weeks out, dapagliflozin 10mg daily was added, and a check-in echocardiogram already showed real movement — EF up to 30%. Encouraged by that jump, the clinic pushed his sacubitril/valsartan toward its target dose of 97/103mg twice daily; four days later he called reporting lightheadedness that turned out to be a systolic pressure in the mid-80s, and the dose came back down to 49/51mg twice daily. An attempt at the same visit to advance his carvedilol past 6.25mg twice daily was pulled back too, after his resting heart rate settled into the low-to-mid 50s with real fatigue alongside it. Three weeks ago, a routine potassium came back at 5.0, and the planned increase in his spironolactone to 25mg daily never happened. Today, twelve weeks from diagnosis — the point guidelines treat as the first reasonable moment to ask whether a primary-prevention defibrillator is warranted — his repeat echocardiogram shows an EF of 33%, and his symptoms have eased to NYHA class II. Real, meaningful improvement. Also, still below the number that would let this question close itself.
In heart failure clinic, three months in
Twenty to thirty-three in twelve weeks is a good number, and I don't think we're done. Non-ischemic myocardium recovers on its own timeline, and that timeline regularly runs past ninety days — sometimes well past it. And DANISH, the largest trial to look specifically at defibrillators in non-ischemic cardiomyopathy, didn't find an overall survival benefit from the device — only in a younger subgroup. That changes how urgently this particular decision needs to be made, even before we get to the pharmacology.
Right now we have a patient whose improvement has slowed, not stopped, with three of his four foundational drugs still sitting well under target dose. I'd want one real attempt to push further before I accept that thirty-three is where this settles.
I'm not arguing we implant anything today — nobody's suggesting that. I'm arguing the reassessment point itself, the one guidelines put at ninety days, shouldn't keep quietly sliding without us naming that's what's happening. He's had three separate dose-limiting events getting here: a pressure in the eighties, a heart rate in the fifties with real fatigue, a potassium at five. At some point, treating every one of those as "not yet a ceiling" starts to look like moving the goalposts rather than following the evidence.
He's thirty-three percent today. If we're having this same conversation at six months, I want us to be honest that we extended a three-month window to six — not that we were still inside the original one.
The disagreement here is smaller than either of you is treating it, because neither of you is actually describing a confirmed ceiling. Look at what happened at each of those three events. The sacubitril/valsartan went up, he got lightheaded four days later, and it came back down — nobody checked whether he was also a little volume-depleted that week, or tried a smaller step instead of reverting all the way. The carvedilol stalled at a heart rate in the fifties with fatigue that was never actually tested against whether it was rate-related or something else entirely. And the potassium of five is three weeks old and has never been repeated.
None of that is a ceiling — it's three data points where a first attempt didn't go perfectly and nobody tried again. Before either of you is right about what thirty-three percent means, I want a real second attempt at all three, done more carefully, with an actual stopping point written down this time — not another open-ended extension.
Agreed within the visit: recheck potassium today rather than treat the three-week-old value as final; if it has normalized, cautiously resume the spironolactone increase to 25mg daily; retry the carvedilol titration in smaller steps with a symptom diary; hold the sacubitril/valsartan at its current dose for now, given how recently it caused symptomatic hypotension. Repeat the echocardiogram in another eight to twelve weeks — extending the total observation window toward roughly five to six months from diagnosis — before the ICD question is revisited.
Not agreed, and the reason a hard branch point was written into the follow-up rather than a single shared expectation:
The Heart Failure Cardiologist's stance: individualize again. If the trend is still moving upward, even slightly, extend once more. Only a genuinely flat or falling number should close the pharmacologic question.
The General Cardiologist's stance: refer for formal ICD reassessment without further extension. The ninety-day window will have already become six months, and guideline criteria were satisfied well before that.
Both cardiologists accepted the Clinical Pharmacologist's read as the reason they're extending the window at all rather than referring today — but neither changed their own answer to what happens if the next number looks the same as this one.