Paper
Friday, 21 July 2006
This presentation is part of : Critical Care Initiatives
Evaluating the Statistical Impact of Using Different Approaches to Operationalize Treatment Seeking Delay for Acute Myocardial Infarction
Maher M. El-Masri, RN, PhD and Susan M. Fox-Wasylyshyn, RN, PhD. Faculty of Nursing, University of Windsor, Windsor, ON, Canada
Learning Objective #1: Develop an awareness of the impact of selecting various operational definition on outcomes of statistical models
Learning Objective #2: Increase awareness of the statistical and clinical issues arising from lack of consistency in operationalizing delay in seeking treatment for acute myocardial infarction

Problem: Knowledge of the factors that contribute to delay in seeking medical treatment for acute myocardial infarction (AMI) provides the basis for interventions that are intended to facilitate prompt care seeking behavior. However, operationalization of delay time varies across different research studies. The use of inconsistent operational definitions of delay time is likely to compromise comparability among the findings of these studies and ultimately limit their generalizability.
Purpose: The purpose of this paper is to examine the impact of inconsistent approaches to measuring delay on the validity of research findings pertaining to identifying its predictors.
Method: A retrospective, descriptive cross-sectional survey was conducted on a convenience sample of 135 in-patients who had recently experienced out-of-hospital AMI. Several regression models were used to examine the influence of using mean delay time as compared to using different cut-off times (1 hour, 2 hours, 3 hours, 6 hours, and 12 hours, median delay) on the number and nature of predictors of AMI delay. Data analysis also included examination of the explained variance, sensitivity, specificity, positive predictive value, and negative predictive value for each regression model.
Results: The number of independent predictors varied from 1 - 5 based on the selection of different definitions of delay. The variance explained by the different regression models ranged from 5.8% to 48%. The sensitivity and specificity of the models ranged from 25% to 79% and 43% to 91%, respectively.  
Conclusion: Use of different criteria for the measurement of delay time resulted in inconsistent results with regard to predictors of delay, produced regression models that explained varied percentages of variance, and had different classification indices. Thus, it is recommended that criteria be established among clinicians and researchers with regard to how care seeking delay is operationalized.

See more of Critical Care Initiatives
See more of The 17th International Nursing Research Congress Focusing on Evidence-Based Practice (19-22 July 2006)