Related Experiment Video
Updated: Aug 14, 2026
Establishing a Competing Risk Regression Nomogram Model for Survival Data
Published on: October 23, 2020
Discrepancy of death diagnosis in gynecology oncology
Joseph T Santoso1, Christine M Lee, Judith Aronson
1University of Tennessee-West Clinic, Memphis, 1588 Union Avenue, Memphis, TN 38104, USA. jsantoso@westclinic.com
Objective:
The medical records, death certificates, and autopsy reports of patients have important medical, legal, and economic implications. This study evaluated the discrepancy between premortem diagnoses (as documented in the medical record and on death certificates) and postmortem diagnoses in a cohort of gynecological cancer patients.
Methods:
The records of all patients with a diagnosis of gynecological cancer who were followed by the Department of Obstetrics and Gynecology at the University of Texas Medical Branch and who died between January 1, 1988 and January 1, 2000 were evaluated. All patients who were admitted initially to Obstetrics and Gynecology service were included, even if they were transferred to a different service after admission. Patients with incomplete charts or in which the death certificate could not be obtained were excluded. The cause of death as recorded in the clinical progress note, death certificate, and autopsy was examined. Each record was evaluated independently by two of the authors to document the cause of death for each patient. The immediate and underlying causes of death were recorded.
Results:
One hundred and seventy eight OB/GYN patients died in the Hospital within the study period. Twelve charts were incomplete or the death certificates could not be identified. The majority of the deaths (91%) occurred on the Gynecologic Oncology service of which 29% (44/151) underwent postmortem examination. All patients in this subgroup had diagnosis of gynecologic malignancy. Twelve patients died while admitted to the Obstetrics service of which 7 underwent autopsy, and 3 patients died while admitted to the Gynecology service of which one patient underwent postmortem examination. The immediate cause of death was most commonly due to sepsis. We found a 66% discrepancy rate between the cause of death recorded in the medical record and/or death certificate and the postmortem diagnosis.
Conclusions:
Multiple studies have documented significant discordance between the medical record, the death certificate, and the postmortem diagnosis in reporting the cause of death. Our findings further confirm the importance of the postmortem examination in determining an accurate cause of death.
