Trial for the Early Identification of Acute Kidney Injury

NCT ID: NCT04200950

Last Updated: 2021-09-24

Basic Information

• Recruitment Status: WITHDRAWN
• Clinical Phase: PHASE2
• Study Classification: INTERVENTIONAL
• Study Start Date: 2020-07-31
• Study Completion Date: 2021-06-30

Brief Summary

Previse is a novel, software-based clinical decision support (CDS) system that predicts acute kidney injury (AKI). Previse uses machine learning methods and information drawn from the electronic health record (EHR) to identify the early signs of acute kidney injury; by doing so before the clinical syndrome of AKI is fully developed, Previse can give clinicians the time to intervene with the goals of preventing further kidney damage, and decreasing the sequelae of AKI. It has been demonstrated in retrospective work that Previse can predict AKI with high accuracy at long prediction horizons, but the tool has yet to be validated in prospective settings; therefore, in this project, the clinical utility of Previse will be assessed through an individually randomized controlled multicenter trial.

Detailed Description

The trial is designed as an individually randomized, controlled, and non-blinded multicenter prevention trial with a baseline period and a primary endpoint of proportion of patients meeting one or more criteria for the Major Adverse Kidney Events within 30 days (MAKE30) composite of death, new renal replacement therapy, or persistent creatinine elevation ≥ 200% of baseline, all censored at the first of hospital discharge or 30 days. The trial will evaluate the efficacy of a machine learning algorithm for AKI prediction, in approximately 8,574 patients aged ≥ 18 years admitted to one of three participating study hospitals. Individual patient randomization will be performed at the time of the alert with a 1:1 allocation ratio. Patients will be evaluated for inclusion in the trial upon admission, and will be automatically enrolled upon meeting inclusion criteria. Because data collection will be conducted through noninvasive procedures that are routinely employed in clinical practice, it will require a waiver of informed consent. Trial efficacy will be assessed at regularly scheduled study visits, and safety will be monitored on an ongoing basis for all patients. Safety will be assessed through the collection of adverse events, laboratory tests, vital signs, and physical examinations throughout the study. An independent Data Monitoring Committee (DMC) will be formed to assist in the periodic monitoring of safety, data quality, and integrity of study conduct. In addition, the DMC will review the interim efficacy analysis performed to determine whether the primary endpoint has been met. Total trial duration is expected to be approximately 12 months.

Conditions

Acute Kidney Injury

Study Design

• Allocation Method: RANDOMIZED
• Intervention Model: PARALLEL
• Primary Study Purpose: PREVENTION
• Blinding Strategy: NONE

Study Groups

Intervention

Previse alert arm

Group Type: EXPERIMENTAL

Previse

Intervention Type: DEVICE

Machine learning algorithm for early acute kidney injury (AKI) prediction.

Control

No alert

Group Type: NO_INTERVENTION

No interventions assigned to this group

Interventions

Previse

Machine learning algorithm for early acute kidney injury (AKI) prediction.

Intervention Type: DEVICE

Eligibility Criteria

Inclusion Criteria

• Adult ≥ 18 years admitted to a participating study hospital

Exclusion Criteria

• ﹤18 years of age
• ESRD diagnosis code
• Stage 4 or Stage 5 CKD diagnosis code
• Initial creatinine ≥4.0mg/dl
• Nephrectomy during admission
• Admission to hospice service
• Admission to observation status
• Any organ transplant (including kidney transplant) within 6 months
• Dialysis order prior to AKI onset
• Dialysis order within 24 hours of admission
• Prior admission in which patient was randomized

• Minimum Eligible Age: 18 Years
• Eligible Sex: ALL
• Accepts Healthy Volunteers: Yes

Sponsors

Dascena

INDUSTRY

Sponsor Role: lead

Responsible Party

Responsibility Role: SPONSOR

References

Mohamadlou H, Lynn-Palevsky A, Barton C, Chettipally U, Shieh L, Calvert J, Saber NR, Das R. Prediction of Acute Kidney Injury With a Machine Learning Algorithm Using Electronic Health Record Data. Can J Kidney Health Dis. 2018 Jun 8;5:2054358118776326. doi: 10.1177/2054358118776326. eCollection 2018.

Reference Type: BACKGROUND

PMID: 30094049 (View on PubMed)

Other Identifiers

07012020

Identifier Type: -

Identifier Source: org_study_id