Technology Hazards and Alarm Management
by John Zaleski, Ph.D., CAP, CPHIMS
Chief Analytics Officer
ECRI recently published its annual Top 10 Technology Hazards report for 2019. Items 4 and 7 on the list are, respectively:
- Improperly set ventilator alarms put patients at risk for hypoxic brain injury or death; and,
- Improper customization of physiologic monitor alarm settings may result in missed alarms.
Both of these hazards pertain to alarm settings on mechanical ventilators and multi-parameter physiologic monitors, respectively, that can and have led to patient harm, principally through the actions of missing clinically-actionable alarms due to inaccurate alarm sensitivity and specificity.
Hazard number four specifically argues: “Mechanically ventilated patients are at risk if user-adjustable ventilator alarms are not tailored to the patient’s respiratory parameters. Leaks, disconnections, and other failures associated with a ventilator’s consumable components are a fairly common occurrence and can quickly lead to harm if the condition is not identified and rectified promptly… Loose connections, manufacturing defects, or other problems with these components can prevent adequate ventilation. Within minutes, inadequate ventilation can result in hypoxic brain injury or death… ECRI institute continues to investigate deaths resulting from breathing circuit disconnections during which no alarm activated. In two cases from early 2018, alarms to detect inadequate ventilation, such as the minute-volume and low-pressure alarms, were not set appropriately.”
Loose connections, for example, can be represented by the dislodgement of the endotracheal tube in the trachea or disconnection of mechanical ventilator tubing. The effects of these events can become perilous for the patient. When endotracheal tube dislodgement does occur, the immediate impact can be a rapid drop in positive end-expiratory pressure (PEEP) and, concomitantly, expiratory or exhaled minute volume (MVe). The end result is failure to ventilate the patient. When such an event goes undetected, the patient can lapse into respiratory failure and arrest.
In terms of detection, the PEEP and MVe are recognizable in the output data from the mechanical ventilator. If not watched closely, then a patient may not receive adequate—or any—ventilation. For patients not breathing spontaneously, this can result in imminent death if not quickly rectified, or can mean harm if the patient is denied oxygen for an extended period of time. Hence, the ability to detect disconnects and dislodgements via measurement of the clinical parameters of PEEP and MVe can mean more rapid intervention.
Hazard number seven, addressing improper customization of physiologic monitor alarm settings, details that: “Improper customization of the alarms on a physiologic monitoring system could prevent staff from learning about significant changes in the patient’s physiologic status or about problems with the medical device or system. Failure to recognize and respond to such conditions in a timely manner can result in serious patient injury or death… Physiologic monitoring systems must be designed and configured to strike the proper balance between activating too many alarms (specifically nuisance alarms that can lead to alarm fatigue) and activating too few alarms (which can lead to hazardous conditions being missed). Alarm customization is one practice that can help achieve this balance. Alarm customization involves selecting alarm values or settings based on the particular needs of a care area and the condition of the patient. When customization is done properly, alarms are less likely to activate for nonactionable conditions, thereby reducing the number of nuisance alarms that activate. But if done improperly, alarm customization can create opportunities for missed alarms, and thus patient harm…”
The ability to adjust multi-parameter physiologic alarms to meet the needs of a patient is only one mechanism upon which to focus improvement. The settings of individual parameters, such as tachycardia and bradycardia thresholds tailored to a particular patient’s needs, still does not take into account information to ensure high enough alarm specificity. Alarm signal threshold adjustments alone do not account for or provide enough specificity into linked cardiopulmonary events, such as pulmonary embolism, pulmonary edema or congestive heart failure or shock. For example, changes in blood pressure related to early onset of septic shock, or changes in pulse related to increasing respiratory distress and early onset respiratory failure and arrest, require monitoring not simply single parameters but multiple parameters from the several devices so that better context regarding patient state can be achieved.
Bernoulli commends ECRI in its continuing efforts to highlight areas of patient safety concern. The two examples identified herein can impact patient safety and are, therefore, of utmost importance. Customization of alarm signal settings and accommodating individual patient needs are definitely a step in the right direction. Bernoulli further suggests that multi-parameter customization, including combinatorial parameter assessments, would provide even more specificity and sensitivity in identifying patients who are trending towards adverse events.
About the Author:
John Zaleski,
Ph.D., CAP, CPHIMS
Chief Analytics Officer
Dr. Zaleski brings more than 25 years of experience in researching and ushering to market devices and products to improve health care. He received his PhD from the University of Pennsylvania, with a dissertation that describes a novel approach for modeling and prediction of post-operative respiratory behavior in post-surgical cardiac patients. Dr. Zaleski has a particular expertise in designing, developing and implementing clinical and non-clinical point of care applications for hospital enterprises. Dr. Zaleski is the named inventor or co-inventor on seven issued patents related to medical device interoperability and has authored three seminal texts on integrating medical device data with electronic health record systems and using medical device data for real-time clinical decision making. View John’s profile on Linkedin.