Reviews/Heart Failure Management/SGLT2 inhibitors improve outcomes in acute decompensated heart failure
Adhf · HFM-ADHF-01 · v2
Borderline: within 4 points of the Likely/Uncertain boundary, so a re-grade could land either side.
This claim has two parts: that starting an SGLT2 inhibitor in or just after a hospital stay for acute heart failure both relieves fluid overload and improves overall clinical outcomes. The overall-outcome part is carried by a purpose-built trial in hospitalised patients [10] and a randomised pre-discharge analysis [1], with several real-world cohorts [7][9][11] pointing the same way. The fluid-relief part, by contrast, stands on just one small 102-patient pilot trial [2], which found more weight loss and less diuretic use but no survival or readmission benefit. Many of the supporting studies do not cleanly compare the drug against usual care: some compare starting it earlier versus later [7], and one pairs it with another drug class [11]. The large real-world benefits may partly reflect that steadier patients are the ones chosen to start the drug in hospital. The panel judged the outcome evidence strong but the fluid-relief and comparator sides thin, which is why it did not place the combined claim in a higher tier.
A dedicated in-hospital randomised trial [10] found a clinical-benefit win ratio of 1.36 (95% CI 1.09-1.68) for empagliflozin started when patients were stable. A pre-discharge post hoc of randomised data [1] reduced the cardiovascular-death or heart-failure-event composite (HR 0.54 at 90 days, 0.49 at 30 days) and all-cause mortality (HR 0.39). Observational cohorts corroborate the direction: inpatient versus post-discharge initiation [7] (post-discharge start an independent predictor, HR 2.40), the MALEOS cohort [8][9] (composite HR 0.37), and dual SGLT2i-MRA registries [11] (90-day odds 0.320 and 0.504). The decongestion signal is confined to the pilot [2], which reported greater weight loss and lower loop-diuretic doses but no in-hospital or 30-day prognostic gain. The contested point: The division is the conjunction: strong randomised support for the composite-outcome limb [10][1] versus a decongestion limb resting on one small pilot [2] that itself found no prognostic benefit. A second fault line is whether 'versus standard care' is cleanly tested, since [7] contrasts timing and [11] tests a combination, not SGLT2i against usual care.
All 11 papers behind this claim with their DOIs and what each does to it, the panel’s arbitration, the evidence gap map and the open research questions. Free, and sent to your inbox.
We’ll also send you other claims from Heart Failure Management, one at a time, and tell you when a release or a pivotal trial changes a grade. Unsubscribe by replying.
One reviewer's only job is to attack the finding. The challenges below were weighed before the grade was set.
The claim conjoins decongestion AND composite outcomes, but decongestion is supported only by [2], a single-centre 102-patient pilot whose authors explicitly state dapagliflozin 'did not improve the in-hospital and 30-day prognosis after discharge'. The decongestion limb is far weaker than the composite limb.
Comparator drift from 'standard care': [7] compares inpatient vs post-discharge SGLT2i (both arms receive the drug) and [11] compares SGLT2i+MRA vs neither (a combination), so neither cleanly isolates SGLT2i versus standard care.
Where it remains uncertain. Decongestion is supported only by the 102-patient pilot [2], which found more weight loss and less diuretic use but stated dapagliflozin did not improve in-hospital or 30-day prognosis.
Aleatory
The only decongestion evidence [2] enrolled 102 patients, so its weight-loss and diuretic differences carry wide random variation and its null 30-day outcomes (9 vs 12 deaths) are underpowered.
Epistemic
No adequately powered trial in the corpus measures decongestion as a primary outcome against standard care; [10] measures a composite including a symptom score, not decongestion.
Inferential
Retrospective cohorts [8]/[9] (HR 0.37) and [11] (odds 0.320) report effects large enough to suggest stabler patients were selected to start the drug, which multivariable adjustment cannot fully remove.
The load-bearing papers from a corpus of 11, each classified by what it does to the claim. Every DOI is taken from the corpus the claim was graded on.
Randomised evidence for pre-discharge initiation improving the composite, though limited to a T2D post hoc subgroup.
“Sotagliflozin reduced the main endpoint at 90 days after discharge (HR: 0.54 [95% CI: 0.35-0.82]; P = 0.004) and at 30 days (HR: 0.49 [95% CI: 0.27-0.91]; P = 0.023) and all-cause mortality at 90 days (HR: 0.39 [95% CI: 0.17-0.88]; P = 0.024).”
Panel auditInternal validity: medium · scope of inference: Patients with type 2 diabetes hospitalised for worsening HF who began sotagliflozin on or before discharge; composite CV death/HF events at 30 and 90 days.Key limitations: post hoc subgroup, not randomised to timing · diabetes-only population · decongestion not reported
Sole direct decongestion evidence (weight loss, diuretic sparing); same trial reports no prognostic benefit, capping the decongestion limb.
“In the dapagliflozin group, there was less tendency to increase the dose of loop diuretics (14% vs 30%; p=0.048), lower average doses of loop diuretics (78.46±38.95 mg/day vs 102.82±31.26 mg/day; p=0.001) and more significant weight loss (4100 (2950; 5750) g vs 3000 (1380; 4650) g; p=0.02).”
Panel auditInternal validity: low · scope of inference: Acute HF inpatients (mean LVEF 44.9%); decongestion/diuretic use and 30-day prognosis.Key limitations: small pilot · prognosis (mortality/rehospitalisation) showed no difference · single centre
Observational support for earlier (inpatient) initiation on rehospitalisation; comparator is post-discharge SGLT2i, not standard care.
“SGLT2i inpatients had significantly reduced cardiovascular rehospitalisations (22.3%) versus post-discharge patients (44.4%) (p=0.005), and significantly reduced heart failure specific rehospitalisations (10.1%) versus post-discharge patients (27.8%) (p=0.018).”
Panel auditInternal validity: low · scope of inference: HFrEF exacerbation inpatients across four English hospitals; rehospitalisation and death by timing of SGLT2i initiation.Key limitations: comparator is later initiation, not standard care · retrospective and small · no mortality difference · no decongestion endpoint
In-hospital SGLT2i initiation reduced the composite of death, prolonged stay, or inotrope need in an adjusted model.
Panel auditInternal validity: low · scope of inference: Acute decompensated HF with LVEF ≤40% and no prior SGLT2i; in-hospital composite outcome (death, stay >21 days, inotropes).Key limitations: retrospective, non-randomised · duplicate of study [9] · small sample · confounding by indication
4 of 11 papers shown. The full review lists all of them, with the panel’s arbitration →
Claims are graded at each quarterly release, and re-graded between releases only when a pivotal trial lands. Every change is a new version of the claim.
Strong randomised evidence carries the composite-outcome limb, but a thin single-pilot decongestion limb, comparator drift, and confounded observational data hold the conjunctive claim below the higher tiers.
Release 2026-q4 re-graded every claim from its corpus under protocol v0.19. An earlier release's grade is shown for the record; it came from an earlier protocol and is not a movement.
Assumptions that carry weight but are not themselves established.
The panel records the gap that would most move the claim, and the study that would close it.
Powered decongestion trial. The decongestion limb rests on a single 102-patient pilot [2] with no prognostic benefit, leaving the claim's first half thinly supported. The panel rates this the highest-importance open gap for the claim, and names the study design that would close it: A randomised trial powered on objective decongestion endpoints (weight, diuretic requirement, natriuresis) comparing in-hospital SGLT2i against standard care.
Open question: Does in-hospital SGLT2i initiation improve objective decongestion against standard care in an adequately powered trial? Why it is still open: Only the 102-patient pilot [2] measures decongestion directly, and it found no prognostic benefit.
Priors is a sceptical first reader, not peer review, and not medical advice.
Release 2026-q4 · graded 21 Oct 2026 · protocol v0.19