Comprehensive strategies are needed to improve awareness, recognition, and treatment for acute coronary syndromes (ACS). Because rapid therapy can improve the unfavorable prognosis of individuals with ACS, it is critical that nurses accurately associate the cues of ACS and quickly and aggressively initiate interventional strategies that reduce mortality. Few studies have examined the decision making process of emergency nurses, even though nurses triage decisions often determine initial patient priority and urgency status for emergency evaluation and treatment. Purpose: To develop a genetic algorithm (GA) to examine gender bias in emergency department nurses triage decisions for ACS. Theoretical/Conceptual Framework: The theoretical framework is a synthesis of Hammond’s lens model (1964) and Evans’ (1984) two-stage reasoning process. Subjects: Emergency department registered nurses (N=840) who belong to the Emergency Nurses Association and triage patients for complaints suggestive of ACS. Method: Subjects completed a computer generated vignette with randomized symptoms suggestive or not suggestive of ACS and randomly selected patient characteristics of sex, age, and race. Two GA were then constructed. The first, constructed on the female vignette patients, was tested for sensitivity and specificity separately for both the female and male vignette patients. The second, constructed using only male vignette patients, was similarly tested. A total of 500 chromosomes with a maximum of 50 genes each were randomly constructed and tested. Bootstrapping was used to prevent overfitting. A total of 500 generations were used. After each generation, the survivors underwent selection, mutation, and crossover to produce each subsequent generation. Results/Conclusions: Nurses used different symptoms to evaluate male and female vignette patients for ACS. Despite this, the two GA had similar sensitivities and specificities. Results indicate that nurses do not agree on the symptoms that constitute ACS and that ED nurses triage decisions do not exhibit gender bias in ACS evaluation.
Session #1198 - Decision Making
The 29th Annual MNRS Research Conference (April 1-4, 2005)