Death certification errors at an academic institution

Arch Pathol Lab Med. 2005 Nov;129(11):1476-9. doi: 10.5858/2005-129-1476-DCEAAA.

Abstract

Context: The correctly completed death certificate provides invaluable personal, epidemiologic, and legal information and should be thorough and accurate. Death certification errors are common and range from minor to severe.

Objective: To determine the frequency and type of errors by nonpathologist physicians at a university-affiliated medical center.

Design: Fifty random patients were identified who died at this academic medical center between January 2002 and December 2003 and did not undergo an autopsy. From medical chart review, clinical summaries were produced. Two pathologists used these summaries to create mock death certificates. The original and mock death certificates were then compared to identify errors in the original certificate. Errors were graded on a I to IV scale, with grade IV being the most severe.

Results: Of the 50 death certificates reviewed, grade I, II, and III errors were noted in 72%, 32%, and 30%, respectively. Seventeen certificates (34%) had grade IV errors (wrong cause or manner of death). Multiple errors were identified in 82% of the death certificates reviewed.

Conclusions: The rate of major (grade IV) death certification errors at this academic setting is high and is consistent with major error rates reported by other academic institutions. We attribute errors to house staff inexperience, fatigue, time constraints, unfamiliarity with the deceased, and perceived lack of importance of the death certificate. To counter these factors, we recommend a multifaceted approach, including an annual course in death certification and discussion of the death certificate for each deceased patient during physician rounds. These measures should result in increased accuracy of this important document.

MeSH terms

  • Cause of Death*
  • Death Certificates*
  • Diagnostic Errors / statistics & numerical data*
  • Hospitals, University / statistics & numerical data*
  • Humans
  • Medical Records / statistics & numerical data
  • Retrospective Studies
  • Vermont