OBJECTIVES: To develop a simulated “control” group to compare Phase 1 trial data from acute kidney failure (AKF) patients receiving novel renal-replacement therapy (RRT) with a Selective Cytopheretic Device (SCD), and to compare clinical outcomes and economic impact of SCD treatment to hemodialysis. METHODS: Clinicians collected data on the first 35 patients treated with the SCD.  To add a second interventional trial arm, the developer needed a simulated control group (control) for comparative purposes.  Data from the literature and 2009 MEDPAR files were utilized to develop a control with clinical characteristics matching those of the trial population but undergoing hemodialysis.  Critically, cohorts were matched on diagnosis, SOFA score, mechanical ventilation, sepsis, and RRT treatment.  Secondary sources contained mortality data, survivor RRT dependency, and resource consumption units for the control -- variables captured for trial patients.  Unit resource costs for trial patients were based on MedPAR data and the literature.  SCD costs were manufacturer-supplied. We computed relative resource cost impact for both cohorts from imputed units and unit costs.  RESULTS: Our methods yielded sufficient information to compare clinical outcomes and budget impact of SCD treatment relative to hemodialysis. SCD-treated patients showed 60-day mortality rates of 31% versus control rates >50%. Savings from reduced ALOS in trial patients exceeded incremental costs of SCD treatment.  We estimated future payer budget impact using USRDS (2010) treatment costs for survivors in both groups with continuing RRT needs progressing to ESRD care.  Results showed cost avoidance of ~$2.0M for 100 AKF patients. Confidence in the model was supported by sensitivities to ICU ALOS and survival rates, non-estimated variables for both cohorts. The model withstood key variable manipulation of 20%+ before conclusions changed directionally. CONCLUSIONS: A simulated control approach is a cost-effective means for small manufacturers to conduct early hypothesis testing and build a foundation for later-stage development.