For most of my career, the patient with an ejection fraction of 44% was a clinical shrug.
Not HFrEF — those patients had the big landmark trials behind them, clear guideline-directed therapy, a treatment algorithm anyone could follow. Not HFpEF — that group was at least getting studied and discussed, even if treatment options were frustratingly limited for a long time. The 44% patient was somewhere in the middle. Symptomatic, structurally impaired, but sitting in a category the field had never quite decided what to do with.
That’s HFmrEF. Heart failure with mildly reduced ejection fraction, now formally defined as LVEF 41–49%. The middle range. And for too long, the default approach was watchful waiting — monitor the EF, see if it recovers, hold on initiating therapy until things declare themselves in one direction or the other.
That posture made sense when we had no data. We’re past that now.
Why the Middle Got Left Behind
The short answer is that it wasn’t included.
The landmark trials that built our guideline foundation for heart failure enrolled patients at the extremes. RALES, EMPHASIS-HF, PARADIGM-HF — these were HFrEF trials, EF typically 35% and below. When HFpEF started getting serious research attention, the focus moved to the preserved end of the spectrum. The mid-range didn’t fit neatly into either category, and trial populations reflected that.
The result was guidelines that were detailed and directive at both ends and vague in the middle. The 2022 ACC/AHA/HFSA Heart Failure Guideline formalized the HFmrEF category and acknowledged the evidentiary gap directly. Subgroup analyses from a handful of trials suggested potential benefit from renin-angiotensin-aldosterone system blockade and MRAs, particularly in patients at the lower end of the mid-range — closer to 41% than to 49%. That evidence is real, but it is weaker and more exploratory than what we have for HFrEF. Honest clinical communication requires saying that plainly.
What was not weak was the signal from the SGLT2 inhibitor trials.
Where the Evidence Is Now
SGLT2 inhibitors changed the conversation across the entire ejection fraction spectrum.
The evidence base for dapagliflozin and empagliflozin now extends well beyond HFrEF. The 2022 guideline — and updates that followed — gives SGLT2 inhibitors a Class 2a recommendation in HFmrEF. That is not fringe or experimental territory. It is a guideline recommendation backed by trial data showing consistent benefit in hospitalization reduction and meaningful signal on cardiovascular outcomes in patients with mid-range EF.
Mechanistically, it makes sense. SGLT2 inhibitors reduce volume load, lower filling pressures, and appear to have direct myocardial effects that are not fully explained by their diuretic properties. Those mechanisms do not suddenly stop working at an EF of 41%.
For patients who are symptomatic and meeting criteria — which includes the large proportion of HFmrEF patients who already have NYHA class II symptoms or greater and an elevated natriuretic peptide — there is now a clear evidence-based option. The question is not whether to wait for more data. The question is why we are not starting a therapy with this profile in patients who qualify.
For ARNI therapy and MRAs, the picture is more nuanced. There is signal for benefit in HFmrEF, particularly in patients whose EF sits closer to the reduced end of the range. Some of this comes from subgroup analyses rather than primary HFmrEF trials, and the guideline language reflects that: these therapies may be reasonable. Not the same level of directive confidence as the SGLT2 recommendation. Clinically, I apply them with attention to where in the mid-range the patient’s EF falls and whether the patient’s profile matches those from the subgroups where benefit was observed.
The Clinical-Practice Problem This Creates
When a category is under-studied, clinical culture fills the gap with inertia.
I’ve seen HFmrEF managed for years on diuretics alone, with repeat echos every six months and no substantive change to the therapeutic plan. Sometimes the EF recovers — cardiac sarcoidosis, tachycardia-mediated cardiomyopathy, peripartum cardiomyopathy — and watchful waiting was exactly right. But that’s a specific subset of patients with a specific reason to expect recovery. Applying that posture to the whole HFmrEF population means a lot of patients sitting in a holding pattern that no longer has guideline support.
The other pattern I see is EF-threshold thinking. “She’s not below 40, so I’ll hold on the sacubitril/valsartan.” That’s a reasonable read of the HFrEF indication, but it shouldn’t translate into no therapy at all. An EF of 43% in a symptomatic patient with a dilated left ventricle and an elevated BNP is not a patient to watch. It’s a patient to treat.
NPs and PAs practicing in cardiology carry significant responsibility for this patient population. In many practice settings, we’re the ones managing these patients between attending visits, adjusting therapy, placing the follow-up calls. If the default framing in our training was that HFmrEF is a diagnostic middle ground rather than a treatable condition with accumulating evidence, that framing needs to be revised.
What This Patient Deserves
The shift is not complicated to summarize.
HFmrEF patients with symptoms and appropriate clinical criteria should be evaluated for SGLT2 inhibitor therapy — that is now a guideline recommendation with meaningful evidential backing, not a speculative choice. The question of whether to pursue ARNI or MRA therapy requires more individualized judgment, with honest acknowledgment that the evidence is thinner and concentrated at the lower end of the EF range.
What HFmrEF patients do not deserve is deferral because their EF doesn’t fall into one of the two better-studied categories. The middle was under-studied for years. The field has started to correct that. Clinical practice should follow.
A patient with an EF of 44% and stage B or C heart failure is not a patient to put on hold. They are a patient to treat — with the therapies we now have evidence for, with appropriate titration and monitoring, and with the same rigor we bring to the ends of the spectrum.
The diagnostic shrug is no longer defensible.