Although CLIA '88 mandates the performance of cytologic-histologic correlation, practice experience reveals that these data are simply documented and rarely used for quality improvement efforts. Precluding meaningful use of these data for practice and patient safety improvement is a lack of knowledge regarding the nature and extent of the impact of these diagnostic discrepancies on patient management and outcomes.
To obtain information regarding the clinical impact of gynecologic cytologic-histologic discrepancies on patient management and outcomes.
Design: Retrospective record review of patients with both a Pap test and a gynecologic surgical specimen obtained within a 6-month period and with at least a “two-step” diagnostic discrepancy during the year 2002 at 4 different laboratories.
Setting: Three academic centers and one community hospital.
Participants: All patients with original or review diagnoses of HSIL/CIN II or III or higher and a random sample of 10% of all female patients with diagnoses of lower than HSIL/CIN II or III.
Main outcome measures: 1) original correlation assessment of discrepancy as due to sampling or interpretation, 2) specific clinical management procedure undertaken, and 3) morbidity associated with clinical management procedures. The clinical impact of each discrepancy event was categorically summarized either as no harm, near miss, or harm (minimal, moderate, or severe).
A total of 475 cases were reviewed. 30% of discrepancies were due to an original interpretation error, the majority cytology undercalls. 10% of patients were lost to follow-up, 40% had repeat Pap tests alone, 23% had colposcopic examinations with further cervical sampling, and 27% had other procedures. 25%, 10%, and 65% of cases were no harm, near miss, and harm events, respectively, with 23% involving unnecessary further diagnostic procedures, and 33% involving delays in diagnosis.
A large opportunity exists for laboratories to improve patient safety by using cytologic-histologic correlation data to guide quality improvement process changes.
See more of Poster Session
See more of The 2005 Institute for Quality in Laboratory Medicine Conference