Development and Implementation of an AI System for Generating Clinical Urine Drug Test Sign-Outs - Summary - MDSpire

Development and Implementation of an AI System for Generating Clinical Urine Drug Test Sign-Outs

  • By

  • Nathan Laha

  • Michael Keebaugh

  • Hsuan-Chieh Liao

  • Bright Amankwaa

  • Olumuyiwa Adesoye

  • Abed Pablo

  • William S. Phipps

  • Andrew N. Hoofnagle

  • Geoffrey S. Baird

  • Patrick C. Mathias

  • Brody H. Foy

  • June 23, 2026

  • 0 min

Share

Objective:

To develop and validate an AI tool for aiding interpretation of urine drug tests (UDTs).

Approach:
    Key Findings:
    • The AI tool achieved expert-level accuracy in generating preliminary textual interpretations of UDTs, with a specific accuracy rate of X%.
    • The integration of AI significantly reduced sign-out time by Y% in the clinical workflow.
    Interpretation:

    The study provides a framework for integrating AI and natural language processing tools into clinical workflows for urine drug testing.

    Limitations:
    • The study did not involve patient or public participation in the design.
    • The AI tool was designed solely for use by trained clinical chemistry clinicians.
    Conclusion:

    The AI tool supports operational workflows in UDT interpretation, enhancing efficiency and accuracy.

Original Source(s)

Development and Implementation of an AI System for Generating Clinical Urine Drug Test Sign-Outs

  • by Nathan Laha, Michael Keebaugh, Hsuan-Chieh Liao, Bright Amankwaa, Olumuyiwa Adesoye, Abed Pablo, William S. Phipps, Andrew N. Hoofnagle, Geoffrey S. Baird, Patrick C. Mathias, Brody H. Foy

  • June 23, 2026

Related Content

Why Real-Time Data Matters in Infection Control

AI, genomics, and integrated data systems could reshape infection prevention workflows

Why Refrigerants Are the Next Critical Sustainability Frontier

With stricter regulations on the horizon, biopharma labs are being forced to rethink how they cool, store, and protect critical samples

Performance of deepseek-R1 and ChatGPT-5.4 thinking in the medical laboratory professional title examination: accuracy, stability, and comparison with interns