Author

Tara N. Cohen

Date of Award

Spring 2017

Document Type

Dissertation - Open Access

Degree Name

Doctor of Philosophy in Human Factors

Department

Human Factors and Systems

Committee Chair

Dr. Scott A. Shappell, Ph.D.

First Committee Member

Dr. Albert J. Boquet, Ph.D.

Second Committee Member

Dr. Douglas A. Wiegmann, Ph.D.

Third Committee Member

Dr. Elizabeth H. Lazzara, Ph.D.

Fourth Committee Member

Dr. Scott R. Reeves, M.D.

Abstract

INTRODUCTION: The purpose of the current research was to assess the utility of the Human Factors Analysis and Classification System (HFACS), a tool that has historically been used reactively to look at accidents and incidents, for classifying observational data from various healthcare venues.

METHOD: Three studies are presented to investigate the reliability of HFACS for classifying observational data. In Study I, HFACS was applied to observational human factors data collected from the cardiovascular operating room (CVOR) at an academic medical university. Three trained analysts categorized the data using HFACS and several approaches were used to evaluate its reliability during the categorization task. The same method was repeated for Study II, which utilized CVOR data collected from a non-academic hospital. To investigate the ability of HFACS for differentiating between hospitals, the data from the academic and non-academic hospitals were compared. Finally, to explore the utility of HFACS in another venue, Study III employed the same approach as Study I and II however, observational data from a trauma center was utilized.

RESULTS: Results of the three studies revealed that the framework was substantially reliable (k=0.635 (95% CI, .611-.659), p = 0.000; k =0.642 (95% CI, .633-.652), p = 0.000; k=0.680 (95% CI, .662 to .698), p = 0.000) for classifying observational healthcare data. In all three data sets, preconditions for unsafe acts were the most common area of systemic weakness. However, differences in the distributions of these categories did exist when data-sets were compared.

CONCLUSION: This study is a first step in establishing the reliability of the HFACS framework as a tool for classifying observational human factors data. As HFACS appears to be a reliable observation tool, findings associated with its use could help to identify where errors and adverse events are likely to occur. Therefore, the proactive identification of human factors issues associated with patient harm represents the next step in the evolution of patient safety. Predictably, hospital administrators could put in place targeted interventions to help mitigate human factors issues before they manifest and become harmful events in the future.

Share

COinS