HIT — Targeted Hyponatremia Correction in Hospitalized Patients

1. Publication

• Title: A Randomized Trial of Targeted Hyponatremia Correction in Hospitalized Patients
• Acronym: HIT (Hyponatremia Intervention Trial)
• Year & Journal: NEJM Evidence, 2026;5(3):EVIDoa2500086 (epublished February 24, 2026)
• Citation: Refardt J, Potasso L, Pelouto A, et al. A Randomized Trial of Targeted Hyponatremia Correction in Hospitalized Patients. NEJM Evid. 2026;5(3):EVIDoa2500086. doi:10.1056/EVIDoa2500086

2. Context & Rationale

Background: Chronic hyponatremia is associated with falls, neurocognitive dysfunction, fractures, rehospitalization, and mortality, but whether correction improves outcomes or hyponatremia merely marks underlying disease severity was unresolved. Prior interventional evidence (e.g., the SALT program with tolvaptan) showed sodium improvement but no established 30-day mortality benefit, in narrow populations.

Research Question/Hypothesis: In hospitalized adults with chronic hypotonic hyponatremia (<130 mmol/L), does targeted, etiology-driven sodium correction reduce 30-day death or rehospitalization compared with routine care?

Why This Matters: HIT was the first large randomized trial designed to move beyond associative epidemiology and test whether actively improving sodium changes hard clinical outcomes — a genuinely novel causal test in a extremely common inpatient electrolyte disorder.

3. Design & Methods

• Study Type: Pragmatic, randomized, controlled, parallel-group, international, multicenter superiority trial with blinded 30-day outcome assessment
• Setting & Centers: 9 European centers (university and regional hospitals); recruited August 2018 – April 2024
• Population:
• Inclusion: Adults ≥18y, hospitalized, hypotonic hyponatremia <130 mmol/L with plasma osmolality ≤300 mOsm/kg, judged chronic
• Exclusions: Severe symptomatic hyponatremia requiring immediate 3% saline (unless persisting >48h); end-of-life care; RRT-dependent kidney disease; acute liver failure; Wernicke encephalopathy; hepatic encephalopathy within 2 months; hepatorenal syndrome; pregnancy/lactation; strict hematology isolation
• Intervention: Trial-team-led, etiology-based diagnostic/treatment algorithm with daily review; stepwise escalation (e.g., fluid restriction → oral urea → tolvaptan for SIAD); intensified/maintained/stopped based on correction rate (target 2-12 mmol/L/24h, avoiding >12/24h or >18/48h overcorrection)
• Comparator: Routine care per attending physician discretion, no trial-team input
• Randomization: 1:1, site-stratified, permuted blocks; 2173 patients (1079 intervention, 1094 control)
• Blinding: Open-label treatment; blinded 30-day outcome assessors (not involved in active care)
• Statistical Power & Follow-Up: 2050 evaluable patients needed to detect 5-percentage-point absolute reduction (23%→18%), 80% power. Primary outcome at 30 days.

4. Key Results

Key dissociation: Biochemical success (60.4% vs 46.2% normonatremia) did NOT translate to clinical benefit (20.5% vs 21.8% primary outcome). Exploratory subgroup: age ≥70y showed lower primary-outcome risk (OR 0.76, 95% CI 0.57-1.00).

5. Internal Validity Assessment

• Randomization & Allocation: 1:1, site-stratified, block-randomized — selection bias unlikely.
• Protocol Adherence & Separation: Process separation was substantial (no treatment 2.5% vs 35%; combined therapies 56% vs 31%), but biochemical separation more modest (max sodium change 10.0±5.6 vs 8.7±5.6 mmol/L). By 30 days, hyponatremia persistence/recurrence was similar (41.7% vs 40.9%) — separation eroded post-discharge.
• Missing Data: High follow-up completeness (98.7% and 98.1%).
• Blinding & Detection Bias: Open-label treatment but blinded 30-day outcome assessment — limits detection bias for the objective primary outcome.
• Overall Internal Validity Conclusion: Moderate — strong randomization, high follow-up, and blinded outcome assessment support the null finding; open-label design, modified ITT population, clinical heterogeneity, and modest (not just process, but biochemical) separation limit confidence that a true small effect would have been detected.

6. External Validity Assessment

• Population Representativeness: Older (median age 73y), multimorbid (median Charlson 4), broad etiology mix (euvolemic, hypovolemic, hypervolemic; community- and hospital-acquired) — closely reflects real-world general inpatient hyponatremia practice.
• Important Exclusions: Severe symptomatic hyponatremia requiring immediate hypertonic saline excluded; only 6.2% had sodium <120 mmol/L — results should not extrapolate to severe/profound or ICU-dominant hyponatremia.
• Overall External Validity Conclusion: Moderate-to-good for older, general-hospital patients with predominantly moderate chronic hyponatremia; limited for severe symptomatic cases, profound hyponatremia, or ICU populations.

7. Strengths & Limitations

Strengths:

• Largest randomized trial to date testing whether inpatient hyponatremia correction improves hard outcomes
• Pragmatic, multicenter, broad etiology inclusion
• High follow-up completeness, blinded outcome assessment
• Clear process separation between arms

Limitations:

• Open-label intervention, performance bias/contamination risk
• Mostly moderate hyponatremia (only 6.2% <120 mmol/L)
• Intervention duration tied to hospital stay — discharge timing outside trial control
• Biochemical separation modest despite substantial process separation
• Modified (not strict) intention-to-treat primary analysis

8. Interpretation & Practice Impact

• Clinical Implications: Does not support expending additional inpatient effort purely to normalize sodium in mostly moderate chronic hyponatremia when routine care is already delivered — but does not argue against treating severe symptoms or selected high-risk patients.
• Mechanistic Coherence: The dissociation between improved sodium correction and unchanged hard outcomes weakens the simple causal view that hyponatremia itself substantially drives short-term prognosis in this population — low sodium may be more a disease-severity marker, or its causal contribution too small/delayed for a 30-day composite.
• Systems-Level Takeaway: Future trials should target more persistent SIAD, profound hyponatremia, patients where sodium blocks discharge, or use longer follow-up with symptom/mobility/cognition/discharge-readiness endpoints rather than short-term mortality alone.

9. Controversies & Subsequent Evidence

• Editorial Commentary: Accompanying NEJM Evidence editorial (Kumachev, MacMillan) emphasizes the achieved biochemical gap was modest (max sodium change differed by only 1.3 mmol/L) and that the trial team had no control over discharge timing — concluding the result is narrower than "don't treat hyponatremia": routine inpatient care simply wasn't improved by additional in-hospital intensification for this population.
• Guideline Integration: Published into an active debate on correction speed/overcorrection limits; cannot resolve that dispute since HIT enrolled mostly moderate chronic cases, excluding severe symptomatic hyponatremia.

10. Summary & Executive Takeaway

Summary: HIT randomized 2173 hospitalized adults with chronic hypotonic hyponatremia to targeted, etiology-driven correction or routine care. The intervention clearly improved biochemical correction (normonatremia 60.4% vs 46.2%), but the 30-day composite of death or rehospitalization was unchanged (20.5% vs 21.8%, P=0.45).

Overall Takeaway: HIT separates biochemical efficacy from clinical efficacy — more sodium correction did not produce fewer 30-day deaths or rehospitalizations in a broadly representative, mostly moderate chronic hyponatremia population, arguing against routine escalation purely to normalize sodium during admission (though not against treating severe symptoms or selected high-risk patients).

11. Bibliography

• Schrier RW, Gross P, Gheorghiade M, et al; SALT Investigators. Tolvaptan for hyponatremia. N Engl J Med. 2006;355:2099-2112.
• Kumachev A, MacMillan TE. Targeted Treatment of Moderate Hyponatremia in Hospitalized Patients — Helpful or Unnecessary? [editorial]. NEJM Evid. 2026;5(3):EVIDe2600002.