A key part of a viable trauma system is the Trauma Registry (TR), used for research, education, and performance improvement.
This study sought to assess the consistency of data abstraction, interpretation, and entry by two hospitals with an identical TR database program.
In phase I, trauma service personnel were queried as to how data were abstracted and entered into the TR.
In phase II, a 1-year retrospective review was conducted of TR data for two trauma centers in San Antonio, Texas.
The phase I review revealed substantial variances in the coding and abstracting of TR data in 30 of the 500 elements (6%). Phase II demonstrated that, because of these variances, considerable differences resulted in coded types and causes of injury.
This study illustrates that these variances can impact attempts to combine databases, establish norms, or assess institutional outcomes.
To ensure the standardization and accuracy of this valuable information, changes may be required.
Recommendations include standardization and education.
A uniform trauma registry or national certification may be necessary.
Mots-clés Pascal : Polytraumatisme, Unité soin intensif, Centre santé, Programme, Base donnée, Registre, Performance, Amélioration, Education, Recherche scientifique, Interprétation, Homme, Traumatisme
Mots-clés Pascal anglais : Multiple injury, Intensive care unit, Health center, Program, Database, Register, Performance, Improvement, Education, Scientific research, Interpretation, Human, Trauma
Notice produite par :
Inist-CNRS - Institut de l'Information Scientifique et Technique
Cote : 99-0370767
Code Inist : 002B16K. Création : 14/12/1999.