STRONG-HF and the Four Pillars: What the Trial Actually Reframed About Sequencing Heart-Failure Therapy

What STRONG-HF changed about the question

For years the debate about heart-failure medicines centered on order: which foundational drug class to start first, and how long to wait before adding the next. The STRONG-HF trial, published in The Lancet in 2022, quietly moved the argument somewhere else. In 1,085 patients recently hospitalized for acute heart failure, an open-label strategy of rapid up-titration paired with close outpatient monitoring lowered the combined risk of 180-day death or heart-failure readmission compared with usual care, and it was stopped early once that difference emerged. The reframing is best stated plainly: STRONG-HF is mostly a trial about speed and follow-up intensity, rather than about the sequence in which the pillars are introduced.

This piece is educational and not medical advice; treatment decisions belong to a person and their own clinician.

The four pillars, and what the trial actually up-titrated

Contemporary therapy for heart failure with reduced ejection fraction rests 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 common point of confusion is that STRONG-HF tested all four. It did not. The high-intensity protocol up-titrated three classes, the renin-angiotensin inhibitor, the beta-blocker, and the MRA, toward full recommended doses. SGLT2 inhibitors were not part of the mandated up-titration, reflecting the period in which the trial was designed and run.

That detail matters for anyone weighing a "sequence the pillars" claim. STRONG-HF cannot tell you when to add the fourth pillar, because the fourth pillar was not what it manipulated. Its evidence speaks to how quickly and completely to bring the older three classes to target after a hospitalization, and to the monitoring that made doing so tolerable.

What "high-intensity care" meant, and why it stopped early

The intervention had two moving parts that are easy to conflate. The first was speed and completeness: aim to reach 100 percent of recommended doses within about two weeks of discharge. The second, and arguably the underappreciated one, was structure. Patients received several scheduled outpatient visits over the two months after discharge, with checks of clinical status, blood pressure, kidney function, potassium, and NT-proBNP before each dose increase. The comparison arm received usual post-discharge care, which in many settings meant slower and less protocolized follow-up.

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

Early stopping deserves a caveat rather than applause. Trials halted at an interim analysis for benefit tend, on average, to overstate the size of the effect, because they are stopped at a random high point in the accumulating data. The direction of the STRONG-HF result is credible and the death-or-readmission signal is supported by a mortality component, but the precise magnitude should be read as a plausible best case rather than a fixed number.

Reading the safety and blinding fine print

Faster titration was not free of adverse events. Hypotension, hyperkalemia, and worsening kidney function were more common in the high-intensity arm, with adverse events reported in roughly 41 percent versus 29 percent. The counterweight is that serious adverse events and treatment-related serious events were not meaningfully higher, which supports the interpretation that 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 temper 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 susceptible to ascertainment bias under those conditions than a hard endpoint like death, and the readmission component carried much 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, so the tolerability seen in the trial may be more favorable than in an unselected ward.

How to appraise a sequence-and-speed claim

When you encounter a confident statement about how to sequence 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 actually randomized. STRONG-HF randomized an intensity-and-monitoring strategy across three classes, not a head-to-head of one drug order against another, so 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. Look at the population and whether it resembles the person in question, since exclusion of low-pressure or low-kidney-function patients limits generalizability. Finally, separate the two claims hiding inside "sequence and speed." STRONG-HF offers reasonable support for reaching target doses quickly with close follow-up after an admission. It offers little on the ordering question that the phrase "sequencing the pillars" implies, and nothing on the timing of the SGLT2 inhibitor it never tested. Holding those apart is most of the work of reading this literature well.