Cost Effectiveness Analysis of a Pharmacist-Led Clinical Decision Support System Powered by Machine Learning to Reduce Drug-Associated Acute Kidney Injury in the Adult Non-ICU Setting
Author(s)
Britney A. Stottlemyer, PharmD1, Kenneth J. Smith, MD, MS2, Caiden J. Lukan, PharmD1, Tezcan Ozrazgat-Baslanti, PhD3, Azra Bihorac, MD, MS FCCM, FASN3, Sandra L. Kane-Gill, PharmD, MS, FCCM, FCCP1;
1University of Pittsburgh School of Pharmacy, PIttsburgh, PA, USA, 2University of Pittsburgh School of Medicine, Pittsburgh, PA, USA, 3University of Florida, Department of Medicine, Division of Nephrology, Hypertension, and Renal Transplantation, Gainesville, FL, USA
1University of Pittsburgh School of Pharmacy, PIttsburgh, PA, USA, 2University of Pittsburgh School of Medicine, Pittsburgh, PA, USA, 3University of Florida, Department of Medicine, Division of Nephrology, Hypertension, and Renal Transplantation, Gainesville, FL, USA
Presentation Documents
OBJECTIVES: Medications may contribute to the development and progression of acute kidney injury (AKI) and increase adverse drug event risk. Nephrotoxin stewardship strategies can promote kidney health by preventing drug-associated AKI (D-AKI) and reducing its severity, however the potential cost benefits associated with targeted programs are unclear. This study evaluates the cost-effectiveness of the “Multi-hospital Electronic Decision Support System for Drug-associated AKI” (MEnD-AKI) clinical trial intervention to reduce the risk and progression of D-AKI.
METHODS: We estimated the cost-effectiveness of MEnD-AKI vs standard-of-care for hospitalized non-intensive care patients using a decision-tree model from the health-system perspective. The model time horizon was 30 days; discounting was not performed. Clinical data were derived from the MEnD-AKI clinical trial conducted at eight hospitals; economic data were derived from relevant published literature. The effectiveness term was AKI progression risk. The robustness of results was tested using one-way and probabilistic sensitivity analyses.
RESULTS: In base-case analyses, the MEnD-AKI intervention was less expensive and more effective than standard-of-care. Compared to standard-of-care, the MEnD-AKI strategy cost $57 per patient less while decreasing absolute risk of AKI progression by 9.3% (29.9% vs. 39.2%). In one-way sensitivity analyses, MEnD-AKI remained cost saving with individual variation of all model parameters. In probabilistic sensitivity analyses, varying all parameters simultaneously over distributions 10,000 times, MEnD-AKI was cost saving in >98% of model iterations.
CONCLUSIONS: MEnD-AKI was less expensive and more effective than standard of care management in hospitalized non-intensive care patients.
METHODS: We estimated the cost-effectiveness of MEnD-AKI vs standard-of-care for hospitalized non-intensive care patients using a decision-tree model from the health-system perspective. The model time horizon was 30 days; discounting was not performed. Clinical data were derived from the MEnD-AKI clinical trial conducted at eight hospitals; economic data were derived from relevant published literature. The effectiveness term was AKI progression risk. The robustness of results was tested using one-way and probabilistic sensitivity analyses.
RESULTS: In base-case analyses, the MEnD-AKI intervention was less expensive and more effective than standard-of-care. Compared to standard-of-care, the MEnD-AKI strategy cost $57 per patient less while decreasing absolute risk of AKI progression by 9.3% (29.9% vs. 39.2%). In one-way sensitivity analyses, MEnD-AKI remained cost saving with individual variation of all model parameters. In probabilistic sensitivity analyses, varying all parameters simultaneously over distributions 10,000 times, MEnD-AKI was cost saving in >98% of model iterations.
CONCLUSIONS: MEnD-AKI was less expensive and more effective than standard of care management in hospitalized non-intensive care patients.
Conference/Value in Health Info
2025-05, ISPOR 2025, Montréal, Quebec, CA
Value in Health, Volume 28, Issue S1
Code
EE225
Topic
Economic Evaluation
Disease
SDC: Urinary/Kidney Disorders