Texas Children's Hospital (Houston, TX) was selected as the winner of ECRI Institute's 10th Health Devices Achievement Award for its development of a project that demonstrated how to better integrate alarm decisions into patient care discussions, and the value of doing so.
The Health Devices Achievement Award recognizes outstanding initiatives undertaken by member healthcare institutions to improve patient safety, reduce costs, or otherwise facilitate better strategic management of health technology. ECRI Institute announced Texas Children's as the winner of the 10th award in February 2016. For details about the other submissions that achieved recognition, see The Health Devices Achievement Award: Recognizing Exceptional Health Technology Management.
ECRI Institute congratulates the project team members: Dr. Fernando Stein, Dr. Eric Williams, Dr. Kevin Roy, Gail Parazynski, Jennifer Sanders, Kimi Raffie, Mellissa Rauch, Darlene Acorda, Samantha Jacques, PhD, Melita Howell, Andrew Peterman, Emma Fauss, PhD, Vincent Gange, and John Weimert.
David Jamison, Executive Director, Health Devices Group (third in from right), presents the 10th Health Devices Achievement Award to the team from Texas Children's Hospital.
To improve patient safety by effectively incorporating alarm management into the fabric of bedside care
1. The need to improve the management of clinical alarms is recognized as a patient safety priority:
a) In 2014, the Joint Commission enacted a National Patient Safety Goal (NPSG) requiring improved management of clinical alarm systems.
b) Since 2007, the topic of alarm hazards has been ranked at or near the top of ECRI Institute's annual list of the Top 10 Health Technology Hazards.
2. Over the past decade, alarm-related projects initiated at Texas Children's led to incremental improvements in reducing the number of unnecessary alarms.
3. However, the projects did little to address the underlying need for a clinical alarm strategy that would consistently yield alarms that were both meaningful and actionable.
4. Thus, Texas Children's decided that a "reboot" of its alarm management program was in order.
1. Like many healthcare facilities, Texas Children's translated the Joint Commission's NPSG Elements of Performance into corresponding action plans. This included activities such as:
a) Forming a Clinical Alarm Management Steering Team to provide strategic direction, governance oversight, and communication channels.
b) Reviewing the equipment inventory, examining alarm data, and obtaining feedback from work groups to identify the most important alarm signals to manage.
c) Establishing policies and procedures.
d) Communicating the plan in appropriate forums and scheduling formal training.
2. For the steering team, an important component of the plan was to find ways to better integrate alarm decisions into patient care discussions.
3. A key preliminary step was to change the organizational mindset to put the patient back into the center of the conversation:
a) Alarm management discussions often revolve around the concept of alarm fatigue, in which staff become overwhelmed by, distracted by, or desensitized to the number of alarms that activate.
b) While acknowledging that alarm fatigue is an important consideration, the team understood it to be a symptom of an underlying problem, not the problem itself.
c) The risk of focusing too narrowly on that one symptom, observed Jennifer Sanders, the hospital's director of clinical informatics, nursing, is that doing so can lead to the issue being perceived as simply a technology management problem or a nursing challenge.
d) The team believed that refocusing attention toward the patient would help clarify that the issue is a patient care consideration that requires input from the whole care team.
e) Sanders notes that the change in mindset helped increase the engagement of stakeholders from many different disciplines, which in turn played a large role in the program's success.
4. With the patient as the focal point, it became clear that substantive changes needed to begin at the bedside. Accordingly, the teams' search for a better approach was led by two guiding principles:
a) Alarms that activate should be actionable and should reflect a patient's specific need.
b) As the patient's condition changes, alarm settings should be routinely evaluated by the entire care team.
5. To achieve this in practice, Texas Children's needed a better system for collecting, analyzing, and viewing alarm data. Ultimately, the organization decided to develop its own system.
a) After assessing commercially available solutions, Texas Children's opted to partner with an outside vendor to create a system that would meet its needs.
b) The result is an alarm dashboard that provides an analytics platform that the care team can use to make decisions around alarm settings.
c) The dashboard can display alarm data by care area, by assigned nurse, and by patient.
(1) The care-area-level data shows the alarms by bed to help identify the patients who are frequently in alarm status.
(2) The nurse-level information depicts the alarm load by nurse, allowing nursing leaders to review patient assignments for appropriateness, considering the total alarm load for each staff nurse.
(3) The patient view summarizes the alarms by patient for the past 24 hours, showing the type and number of alarms as well as the "time in alarm" (the cumulative duration of those alarms). The dashboard also displays: comparisons with the previous day, alarms aggregated by time of day, and data to illustrate the impact of changing an alarm limit (e.g., changing a limit from x to y for this patient would have changed the number of alarms by z%).
6. Texas Children's first implemented the system in the progressive care unit (PCU), a step-down ICU unit that was selected to be the early adopter care area.
a) The pilot program started with a baseline analysis that included:
(1) An on-site assessment of the care area.
(2) Multiple staff interviews.
(3) A comprehensive analysis of PCU alarm and patient data. The new data acquisition platform had passively collected this data from the electronic medical record, cardiac monitoring system, and nurse staff assignments over a 53-day period.
b) Based on these findings, the team decided to start with simple changes and progress to larger-scale ones using a PDSA (plan-do-study-act) deployment methodology. The process progressed as follows:
(1) The first PDSA cycle addressed environmental factors in the care area—specifically, and quite interestingly, the effect that trash can lids had on the alarm load.
(a) The team's analysis of the PCU alarm data identified spikes in the number of alarms that were activating in care areas arranged as "pods," where four children were grouped in one larger room.
(b) The team investigated those areas and discovered that the young patients were being startled by the noise when the lid was closed on the metal trash/linen bins located in those care areas. The startle response frequently led to momentary changes in the patient's condition that would trigger an alarm.
(c) Remediating the problem in this case did not involve a change in clinical practice or high-tech solution, but rather a simple adjustment: Equip the trash/linen bins with soft-closing lids.
(2) The second PDSA cycle used data collected at the care-area level to drive a decision to adjust the default alarm limits for SpO2 (a parameter that measures oxygenation in the blood).
(a) The data for this care area revealed that "SpO2 Lo" was the most frequently activating alarm.
(b) Following collaboration with medical staff and endorsement from the Critical Care Committee, a decision was made to change the lower threshold for this parameter from 93 to 90 in the PCU.
(3) A third PDSA cycle focused on staff education, raising awareness about alarm management practices and reviewing device functionality (e.g., proper use of alarm silence and suspend functions).
(4) The fourth PDSA cycle unveiled patient-specific alarm dashboards.
(a) These dashboards provided a tool to aid clinicians in the alarm-setting process.
(b) The care team for each patient would review the patient's dashboard during rounding, providing an opportunity to evaluate the settings for high-alarming patients.
1. Texas Children's experienced an ongoing reduction in the number of alarms over the six months following implementation. And in the time since, the organization reports that the lower levels have been sustained, with some additional reductions reflecting a continuous improvement process.
2. Following are some of the specific results Texas Children's reported for the four PDSA cycles outlined above:
a) The effects of repairing or replacing the trash/linen bin closures:
(1) A 2.6% reduction in "time in alarm" (38 minutes per day).
(2) Quieter environment; the noise level of the bins was reduced from 80 dB to below 31 dB (the average ambient noise levels).
b) Reducing the low-SpO2 threshold from 93 to 90: This change resulted in a 10% reduction in SpO2 alarms per bed per day.
c) Nursing education: This effort yielded an additional 11% reduction in low-SpO2 alarms per bed per day.
d) Pilot of patient-specific alarm dashboards:
(1) Physicians became engaged in the alarm management improvement process and discussion.
(2) The process of reviewing the alarm dashboard during rounds took an average of less than 2 min per patient.
(3) Clinicians responded positively to the use of data and patient analytics to help in making patient care decisions.
3. Texas Children's is currently working on adopting this management model more widely across the organization.
Texas Children's observed the following about this project:
1. The data acquisition system, implementation methodology, and introduction of real-time dashboards shifted the conversation in two important ways:
a) Discussions shifted from "alarm management" to "patient care," putting the focus on the patient, not on the technology generating an alarm.
b) Feedback shifted from subjective statements ("too many alarms") to actionable information based on analytic dashboards for the care area, assigned nurses, and each patient.
2. Each successful PDSA cycle:
a) Revealed new insights. For example, the effect that the trash/linen bins were having on the numbers of alarms was an eye-opening discovery. This scenario illustrated that simple changes can sometimes help address complex problems.
b) Deepened the collaboration between the medical and nursing staff, as the team members began to understand and appreciate different points of view.
c) Increased acceptance of the methodology. Texas Children's notes that care team members are now requesting more formal research around decision support related to alarm data.
3. Overall, the project provided the building blocks to scale the solution from a single 36-bed unit to an enterprise-wide platform. Data acquisition efforts are already under way in several care areas.