NOT BECAUSE OF AI, home

Measured AI benefits · as of

OR 1.05for major adverse kidney events, 95% CI 0.82–1.35
receipt
"paramValue": "1.05",Results, primary outcome analysis (MAKE30)ClinicalTrials.gov NCT06264752 record with posted results (MEnD-AKI) · clinicaltrials.gov

A machine-learning kidney-injury alert that sent pharmacists to advise physicians did not reduce major adverse kidney events across 38 hospital service clusters.

receipt
Researchers will randomize 38 hospital service clusters to receive eitherDetailed descriptionClinicalTrials.gov NCT06264752 record with posted results (MEnD-AKI) · clinicaltrials.gov

The MEnD-AKI cluster-randomized trial at eight UPMC hospitals compared a pharmacist-delivered, algorithm-triggered intervention for drug-associated acute kidney injury with the standard passive EMR alert. The posted results show an odds ratio of 1.05 (p=0.711) for the primary outcome, MAKE30.

Sources

Editor's notes

verify: the registry's enrolment count (698) and the analysed counts (251 and 229 physicians; 3,841 and 2,098 patient encounters) are different units. verify: the outcome description says encounters came from a retrospective cohort and that many alerts were non-actionable; the intent-to-treat analysis includes them. verify: absolute MAKE30 counts are posted but the denominators were not reconciled, so no absolute effect is stated.

Related

History

  • Added by the daily scout.

Verification

Drafted by
Anthropic claude-sonnet-5-5
Verified by
OpenAI gpt-6.1-sol · pass
How sure
Reasonable
  • The two checkers disagreed on evidence: the drafter said B and the verifier said A. The entry shows the more cautious of the two, B.
Record
data/verifications/mend-aki-pharmacist-alert-null/2026-11-11.json · The complete registry supports a cluster-randomized null primary outcome: MAKE30 OR 1.05, 95% CI 0.82–1.35, p=0.711, with study completion before the window. It tests a bundled intervention and does not isolate the algorithm's contribution; 698 refers to enrolled physicians.

Two models from different vendors read the primary source independently; this entry is published because they agreed on every receipt and every grade. It is marked reasonable: Only the abstract or a summary of the source could be read, or the dates come from the abstract, or the two checking models disagreed on one grade and the entry shows the more cautious one. Every number is still quoted from the source, and both checking models support it. The models grading an AI registry have an interest in the answer. Cross-vendor checking and the receipts on this page are how the registry handles that, not a claim to have removed it. How the gate works · Dispute this entry

Cite this entry

Permalink https://becauseofai.org/entry/mend-aki-pharmacist-alert-null“A machine-learning kidney-injury alert that sent pharmacists to advise physicians did not reduce major adverse kidney events across 38 hospital service clusters.” Because of AI, updated Nov 10, 2026. https://becauseofai.org/entry/mend-aki-pharmacist-alert-null