Consistency between coded poison center data and fatality abstract narratives for therapeutic error deaths in older adults

Abstract

Introduction: The national poison center database provides unique opportunities for research on medication safety. Much of this research utilizes only the coded portions of the case record. The primary goal of this study was to evaluate the consistency of the case scenario details on the coded poison center record compared to written fatality abstracts for fatal therapeutic error cases in older adults. A secondary goal was to elucidate causes and coding consistency for fatal iatrogenic errors.

Methods: A retrospective analysis of fatalities because of unintentional therapeutic errors reported to the National Poison Data System from January 1, 2002 to December 31, 2006, for adults > or =65 years old was performed. Coded case scenarios from National Poison Data System were compared with those described in the fatality abstract narrative.

Results: The primary outcome measure revealed 110 deaths in older adults because of therapeutic errors. In 21 of the 110 cases (19.1%), additional or alternative case scenario details were found by evaluating the written fatality abstracts. Secondary outcome analysis found that in the 27 iatrogenic error cases, almost half had either inconsistent or incomplete scenario detail coding compared to the fatality abstract narrative. Eighteen of these cases had a second case scenario coded, which helped elucidate the reason for the error.

Conclusion: This study found consistency between coded scenario detail and the fatality abstract narrative in approximately 80% of cases. The fact that the exact reason could not be ascertained from either the coded record or the fatality abstract in one-third of cases is troublesome. Poison center data are a potentially valuable source of toxicology and poisoning information that necessitates accuracy and consistency to provide optimal utility.