Date of Award

May 2018

Degree Type

Dissertation

Degree Name

Doctor of Philosophy

Department

Biomedical and Health Informatics

First Advisor

Derek Nazareth

Committee Members

Hemant Jain, Jasmin Flores, Christine Kovach, Mark D Srite, Hyunkyoung Oh

Keywords

Alarms, Cognitive systems, Comparative, Nurse Call, Patient Events, Perceptions

Abstract

Medical device alarm systems are expected to improve patient care by alerting clinicians about conditions that require attention. However, due to a variety of circumstances, including inadequate training, muting alarms, alarm fatigue, and staffing shortages, the effectiveness of alarm systems may be questionable. This research looked at the appropriateness of time- to-respond (TTR) to alarms, the alarm system configuration, policies and procedures regarding alarms, and the extent of alarm-specific training and education alarms. Using concepts from cognitive systems engineering, organization policy, and organizational learning, a research model was assembled to investigate these relationships.

Quantitative data analysis included an online survey conducted in four hospitals, retrospective review of alarm data related to patient harms, review of Nurse Call download data used to compare self-report of alarms to actual numbers of alarms as well as to assist in answering exploratory questions. Qualitative data analysis included the clinician survey comments, review of alarm-related policy and procedure, and staff interviews.

Alarm survey data were collected from a total of 107 respondents over a three-month timeframe. Data download of alarms totaled 88,307. Using a logistic regression approach, partial support for the hypotheses was found across contexts of high, medium, and low priority alarms. The overall prediction of appropriateness of alarm response was good, except in the case of medium priority alarms. Examination of the alarm data revealed that clinician response to medium priority alarms was considerably slower than anticipated.

The results indicated that alarm configuration, policy, education, and training provided some explanation about alarm response. However, resulting data also indicated that the relationship between the alarm priorities and response times are not fully understood. While high priority and low priority alarms were approached appropriately, medium priority alarms did not elicit the same response. This is of some concern given that they form the bulk of the alarms in some hospitals. While alarm configuration, policy and procedures, education and training provided some explanation about alarm response, other factors may contribute to the disparity in response which were not clarified in this research. As more devices with alarm capabilities are introduced into patient care, it is imperative that the appropriate response is elicited in clinicians.

Download

Share

COinS