TL;DR: ⚪ Null — a dedicated kidney action team with EHR-based recommendations improved implementation but did not reduce worsening AKI, dialysis, or death (19.8% vs 18.4%).
1. Publication
- Title: Early, Individualized Recommendations for Hospitalized Patients With Acute Kidney Injury: A Randomized Clinical Trial
- Acronym: KAT-AKI (Kidney Action Team)
- Year & Journal: JAMA, published October 25, 2024 (332(24):2081-2090)
- Citation: Aklilu AM, Menez S, Baker ML, et al; KAT-AKI Team. JAMA. 2024;332(24):2081-2090. doi:10.1001/jama.2024.22718
2. Context & Rationale
Background: AKI affects up to 20% of hospitalized patients and often goes undiagnosed/undertreated, since AKI management is heterogeneous and a "one-size-fits-all" alert-based approach has previously shown limited benefit (and possible unintended harm from alert fatigue).
Research Question/Hypothesis: In hospitalized patients with AKI, do personalized diagnostic/therapeutic recommendations from a dedicated "kidney action team" (physician + pharmacist) delivered via the EHR reduce worsening AKI stage, dialysis, or mortality compared with usual care?
3. Design & Methods
- Study Type: Randomized, investigator-blinded clinical trial
- Setting & Centers: 7 hospitals, 2 health systems (Yale New Haven Health, Connecticut/Rhode Island; Johns Hopkins, Baltimore); Oct 2021–Feb 2024
- Population: 4003 hospitalized patients with AKI
- Intervention: Kidney action team (physician + pharmacist) reviewing charts and sending personalized EHR recommendations across 5 categories (diagnostic testing, volume, potassium, acid-base, medications)
- Comparator: Usual care
- Statistical Power & Follow-Up: Primary: composite of worsening AKI stage, dialysis, or mortality.
4. Key Results
4003 patients randomized.
Outcome | Intervention | Usual Care | Notes | |
Composite: worsening AKI, dialysis, or death (primary) | 19.8% | 18.4% | Not significantly different (numerically higher with intervention) | |
Recommendation implementation within 24h (secondary) | 33.8% (2459/7270) | 24.3% (1766/7269 undelivered recs) | Diff 9.5pp, 95% CI 8.1–11.0 | Significantly higher implementation, but did not translate to outcome benefit |
5. Internal Validity Assessment
Large (4003-patient), investigator-blinded RCT across two academic health systems. Overall: Strong — clean null primary result despite confirmed process improvement (higher recommendation implementation), demonstrating a genuine dissociation between process metrics and patient-centered outcomes.
6. External Validity Assessment
7-hospital, 2-health-system population — broadly representative of general hospitalized AKI populations in academic health systems with EHR-integrated alert capability.
7. Strengths & Limitations
Strengths: Large, well-conducted trial with confirmed intervention delivery (higher implementation rate); addresses a genuine care-heterogeneity problem in AKI management.
Limitations: Author explicitly noted potential "unintended harm from alerts and limited experience/confidence in managing the diverse nature of AKI, where a one-size-fits-all approach will not work" — even with individualized recommendations, this did not translate to outcome benefit.
8. Interpretation & Practice Impact
Does not support a dedicated kidney action team model for preventing AKI progression, dialysis, or death — despite successfully improving process measures (recommendation implementation), this did not translate to better patient outcomes.
9. Controversies & Subsequent Evidence
Lead author Abinet Aklilu explicitly framed this alongside prior AKI-alert intervention studies as part of a consistent pattern where process improvements do not reliably translate to outcome benefits in AKI management, suggesting future interventions may need to address deeper barriers (clinical confidence, AKI heterogeneity) rather than just improving information delivery.
10. Summary & Executive Takeaway
Summary: KAT-AKI randomized 4003 hospitalized AKI patients across 7 hospitals to receive personalized kidney-action-team recommendations via EHR or usual care. The composite outcome (worsening AKI, dialysis, or death) was not significantly reduced (19.8% vs 18.4%), despite significantly higher recommendation implementation in the intervention group.
Overall Takeaway: A dedicated kidney action team providing individualized EHR-based recommendations improves process measures but does not improve hard AKI outcomes — reinforcing that better information delivery alone is insufficient to address the heterogeneous, complex nature of AKI management.
11. Bibliography
- Aklilu AM, O'Connor KD, Martin M, et al. KAT-AKI protocol and early data. BMJ Open. 2023.