STRONG-HF and the Four Pillars: What the Trial Really Tested About Heart-Failure Therapy

The short version#

STRONG-HF was mostly a trial about pace and follow-up, not about the order of the drugs. Published in The Lancet in 2022, it enrolled 1,085 patients recently hospitalized for acute heart failure and randomized them to one of two strategies. One group received a fast, protocol-driven push to full recommended doses of several foundational medicines, backed by frequent office visits and lab checks. The other received usual post-discharge care. The high-intensity group had fewer deaths or heart-failure readmissions by 180 days, and the trial was halted early once that gap appeared. Read carefully, its lesson is about how quickly and how safely you can reach target doses after a hospitalization, not about which pillar to start first.

Key points#

• STRONG-HF tested a strategy (fast up-titration plus close monitoring), not a specific drug order.
• It up-titrated three classes toward target: a renin-angiotensin inhibitor, a beta-blocker, and a mineralocorticoid receptor antagonist. SGLT2 inhibitors were not part of the mandated push.
• The primary endpoint (all-cause death or heart-failure readmission at 180 days) favored the intensive arm, roughly 15 percent versus 23 percent.
• The trial was open-label and stopped early, two features that argue for reading the effect size as a plausible best case.
• The scheduled visits and lab checks were part of the intervention. Speed without the monitoring is not what was studied.

The four pillars, and which ones the trial pushed#

Current therapy for heart failure with reduced ejection fraction is built on four drug classes, often called the four pillars: renin-angiotensin system inhibitors (an ACE inhibitor, an ARB, or an ARNI), beta-blockers, mineralocorticoid receptor antagonists, and SGLT2 inhibitors. A frequent misreading is that STRONG-HF tested all four. It did not. The high-intensity protocol drove three of them, the renin-angiotensin inhibitor, the beta-blocker, and the MRA, toward full recommended doses. SGLT2 inhibitors sat outside the mandated up-titration, a reflection of when the trial was designed and run.

That gap matters if you are weighing a claim about "sequencing the pillars." STRONG-HF cannot answer when to add the fourth pillar, because the fourth pillar was never the thing it manipulated. Its evidence is about bringing the older three classes to target quickly and completely after a hospital stay, and about the monitoring that kept doing so tolerable. That is a narrower claim than the slogan.

What "high-intensity care" actually was#

Think of the intervention as two levers pulled at once. The first was pace: aim to reach 100 percent of recommended doses within about two weeks of discharge, rather than nudging doses up over months. The second, and the one that tends to get lost in the headline, was structure. Patients in the intensive arm attended several scheduled visits across the two months after discharge. Before each dose increase, a clinician checked clinical status, blood pressure, kidney function, potassium, and NT-proBNP. The comparison arm received usual care, which in many places meant slower, less standardized follow-up.

It helps to picture the difference less as a stronger dose of medicine and more as a tighter safety net, and the medicines were the same ones a patient might have received anyway. What changed was how fast they were escalated and how closely the escalation was watched.

Why it stopped early, and why that calls for a caveat#

The primary endpoint was all-cause death or heart-failure readmission at 180 days, and the high-intensity group hit that endpoint less often, roughly 15 percent versus 23 percent, a relative risk near 0.66, with the difference driven mainly by fewer readmissions. A data safety monitoring board recommended stopping once the prespecified boundary was crossed.

Early stopping deserves a note of caution rather than a round of applause. Trials halted at an interim look for benefit tend, on average, to overstate the size of the effect, because they are stopped at a favorable moment in the accumulating data. The direction of the STRONG-HF result is credible, and a mortality component supports the combined signal, but the exact magnitude is better read as a plausible best case than as a fixed figure.

The safety and blinding fine print#

Faster titration came at a cost; low blood pressure, high potassium, and worsening kidney function were more common in the intensive arm, with adverse events in roughly 41 percent versus 29 percent. The reassuring counterweight is that serious adverse events and treatment-related serious events were not meaningfully higher. That pattern supports a specific interpretation: the monitoring, as much as the medicines, is what made rapid up-titration workable. Speed without the visits is not the intervention that was tested.

Two design features limit how far the result travels. STRONG-HF was open-label, so patients and treating clinicians knew the assignment; a softer endpoint like readmission is more open to ascertainment bias under those conditions than a hard endpoint like death, and readmission carried most of the benefit. The enrolled population was also selected: people with very low blood pressure, marked kidney impairment, or high potassium at screening were less likely to qualify. The tolerability seen in the trial may therefore look more favorable than it will on your ward.

A short checklist for any speed-or-sequence claim#

When you meet a confident statement about how to sequence or how fast to push the pillars, a few questions separate evidence from extrapolation.

• Name the trial. Is the claim grounded in a randomized comparison, and is it STRONG-HF or something else?
• Ask what was randomized. STRONG-HF randomized an intensity-and-monitoring strategy across three classes, not one drug order against another. It does not establish that any particular starting sequence beats another.
• Check the endpoint and the blinding. An open-label trial resting largely on readmission warrants more caution than a blinded trial resting on mortality.
• Check the population. Excluding low-pressure and low-kidney-function patients narrows how widely the tolerability applies.

Hold the two ideas apart. STRONG-HF gives reasonable support for reaching target doses quickly with close follow-up after an admission. It says little about the ordering that the phrase "sequencing the pillars" implies, and nothing about the timing of the SGLT2 inhibitor it never tested. Keeping those claims separate is most of the work of reading this literature well.