Related Experiment Video
Updated: Aug 13, 2026
Introduction of an Integrated Pathology Image Management, Artificial Intelligence, and Reporting System
Published on: July 11, 2025
Postimplementation evaluation of cytopathology amendment reports after adoption of a new laboratory information
Nathalie Andon1, Ronald N Arpin1, Vanda F Torous1
1Department of Pathology and Laboratory Medicine, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts.
Introduction:
Amendment reports are a recognized but underutilized cytopathology quality assurance metric that can reveal diagnostic, reporting, and workflow vulnerabilities other indicators miss. Implementation of a new laboratory information system (LIS) is an underappreciated driver of amendments, as transitions introduce altered workflows, unfamiliar templates, and reconfigured terminology. The impact of LIS go-live events on amendment patterns has not been systematically studied. We characterized amendment frequency, categorization, and workflow implications during the early postimplementation period following an enterprise LIS migration.
Materials And Methods:
All gynecologic and nongynecologic cytopathology amendments issued during the first 6 months after new LIS implementation were reviewed. Amendments were reviewed and reclassified into a standardized taxonomy distinguishing diagnostic from nondiagnostic changes, misassignment errors, and system-level causes.
Results:
Fifty-five amendments were issued across 18,952 cases (0.29%), with half (50.9%) reclassified upon structured review. Rates peaked in the first full operational month (0.47%) and declined to 0.14% by month 6. Nondiagnostic amendments predominated (50.9%), most commonly corrected nondiagnostic information (27.3%), followed by diagnostic amendments (36.4%), of which corrected diagnostic information was most frequent (25.5%). One wrong-patient misassignment (1.8%) occurred in the go-live month, and 6 amendments (10.9%) reflected system-level errors.
Conclusions:
LIS go-live is associated with a characterizable, time-limited pattern of amendment activity concentrated in the earliest months. Structured, prospective amendment monitoring offers a low-cost, high-yield quality assurance tool for uncovering system-level errors requiring informatics or vendor intervention and for protecting patient safety during LIS transitions, supporting intensified surveillance during the earliest implementation window.