Effects of Patient Load and Other Monitoring System Design Choices on Inpatient Monitoring Quality

Effects of Patient Load and Other Monitoring System Design Choices on Inpatient Monitoring Quality August 10, 2015 8/10/2015 1

AAMI Foundation Vision: To drive the safe adoption and use of healthcare technology Visit our website to learn more about our alarm initiative National Coalition for Alarm Management Safety: Get involved and consider making a donation to this important national effort! http://www.aami.org/thefoundation/content.aspx?item Number=1494&navItemNumber=672 Contact Sarah Lombardi at slombardi@aami.org

Thank You to Our Industry Partners This Patient Safety Seminar is offered at no charge thanks to funding from our National Coalition for Alarm Management Safety industry partners. The AAMI Foundation and its co-convening organizations appreciate their generosity. The AAMI Foundation is managing all costs for the series. The seminar does not contain commercial content. Platinum Gold Silver 9/25/2013 3

LinkedIn Questions Please post questions about alarms on the AAMI Foundation s LinkedIn page: http://www.linkedin.com/groups/healthcare- Technology-Safety-Institute-HTSI-4284508 9/25/2013 4

Speaker Introductions Melanie Wright, PhD Program Director, Patient Safety Research, Trinity Health and Saint Alphonsus Health System Noa Segall, PhD Assistant Professor, Duke University Medical Center

Disclosures Melanie Wright, Research funding from the Agency for Healthcare Research and Quality, National Library of Medicine, and Trinity Health Noa Segall, Research funding from the Agency for Healthcare Research and Quality

Effects of Patient Load and Other Monitoring System Design Choices on Inpatient Monitoring Quality Recognized problems Delayed response to life-threatening emergencies Delayed response to deterioration Research challenges Outcome metrics not well-defined Complexity and variety of implementations Low frequency of events Difficult environment to observe or control

Failure Modes in Delayed Response to Emergencies Patient not monitored when event occurs, System fails to deliver alert to bedside RN, Alert delivered to bedside RN but not seen or heard, Alert delivered to bedside RN and seen/heard but not heeded. Pennsylvania Patient Safety Authority. Alarm Interventions During Medical Telemetry Monitoring: A Failure Modes and Effects Analysis. Harrisburg, PA: Pennsylvania Patient Safety Authority;2008. Andersen PO, Maalã e R, Andersen HB. Critical incidents related to cardiac arrests reported to the Danish Patient Safety Database. Resuscitation. Vol 81 2010:312-316. 8

Factors Associated with Failures Alarm management (false alarms) Monitoring practice model/technologies Continuity and load of monitor watch stations Practices related to maintaining continuity of monitoring Training and assessment of watcher, bedside RN Patient acuity and bedside RN load Communication technologies, practices, culture Ordering and escalation practices More 9

Defining Best Practices Requires metrics and experimental comparisons, more than SRE review. 10

Defining and Measuring Monitoring Quality Mortality, Failure-to-Rescue Codes outside of ICU Response time to emergencies/other events Time not monitored Defining and counting late recognition Defining and counting poor patient states 9/25/2013 11

Patient Load Effects on Response Time to Critical Arrhythmias Introduction

Patient Load Effects on Response Time to Critical Arrhythmias Introduction

Patient Load Effects on Response Methods Time to Critical Arrhythmias Study design Randomized trial Independent variable Patient load: 16, 24, 32, 40, and 48 patients Dependent variables Response time to a simulated ventricular fibrillation Participants requiring 20 sec or longer to respond Task performance Rhythm interpretation Participants 15 remote telemetry technicians 27 nurses from cardiac units

Patient Load Effects on Response Time to Critical Arrhythmias Methods Simulation design

Patient Load Effects on Response Time to Critical Arrhythmias Methods Simulation design Patient Event Required Tasks Weight Call HUC 5 Tachycardia (a 30-bpm Make phone call within 1 minute 4 increase from baseline) Ask to speak to the patient s nurse 3 Bradycardia (< 45 bpm) State correct problem 4 Print rhythm strips 1 Converting to a different Send 1 strip to the patient s nurse 1 rhythm, e.g., Afib Document the patient s number, the current time, the nurse s name, and the rhythm 1

Patient Load Effects on Response Time to Critical Arrhythmias Methods Simulation design

Patient Load Effects on Response Time to Critical Arrhythmias Results Response time 25 20 Individual Response Times Mean Response Time Response Time (seconds) 15 10 5 0 16 24 32 40 48 Number of Patients Monitored

Patient Load Effects on Response Time to Critical Arrhythmias Results Task performance 90% 80% 70% Task Performance Score (%) 60% 50% 40% 30% 20% 10% 0% 16 24 32 40 48 Number of Patients Monitored

Patient Load Effects on Response Results Survey Time to Critical Arrhythmias Strongly Disagree/ Disagree Neutral Agree/ Strongly Agree The experiment was long enough to accurately assess my workload 1 0 39 The lethal rhythm (VF) was realistic (similar to a real VF) 1 0 41 The waveforms were clear enough to interpret 3 5 34 The pace of patient events was realistic (technician responses only) 0 1 14 My documentation tasks were realistic (technician responses only) 0 2 13 The phone conversations were realistic (technician responses only) 1 0 14 Overall, the experiment was realistic (similar to real cardiac monitoring) (technician responses only) 1 2 12

Patient Load Effects on Response Time to Critical Arrhythmias Discussion Response times to the simulated arrhythmia increased as patient load increased More outliers in 48-patients group Task performance decreased as patient load increased

Trinity Health Research to Identify Best Practices Develop and validate a metric of patient monitoring quality that can be used in comparisons across hospitals and care settings. Define characteristics of monitoring system design expected to influence monitoring effectiveness. Conduct case studies to identify best practices in in-hospital monitoring.

Factors that impact monitoring quality From 11 interviews, 7 site visits (preliminary data) 9/25/2013 23

Interviewee initiated responses: Alarm management Communication directness Architecture/environment Communication quality/culture Training of monitor watchers Training of bedside nurses Involvement/use of rapid response teams Technology coverage Technology usability Technology familiarity Interviewer raised and confirmed: Human watching and data filtering Performance monitoring and feedback Ordering/discontinuation policies and enforcement Monitor watcher patient load Bedside nurse patient load Accessibility of expert resources Patient acuity unit type Interviewer raised, not confirmed Backup or escalation alerting Secondary alerting practices Multi-parameter early warning signs Red variability within Trinity Health Green variability outside Trinity Health 2014 Trinity Health - Lvonia, MI 24

Next Steps Observational, case-based studies: Validate and select relevant outcome metrics. Make comparisons across units and hospitals to identify best practices. Lab-based and simulation studies: Computer simulation to evaluate the impact of design choices on response time to emergencies. Expand monitor watch lab-based patient load research to include patient load and task assignment.

Thank you for attending! Slides & Recording Available Here: http://www.aami.org/thefoundation/content.aspx?itemnumber=1498&n avitemnumber=671 9/25/2013 26

Free Alarm Resources Safety Innovations Series White Papers Patient Safety Seminar Recordings Alarms Management Patient Safety Seminars Webinar Recordings Webinar Slides Key Points Checklists

Questions? Please visit the AAMI Foundation s LinkedIn page to post a question: http://www.linkedin.com/groups/health care-technology-safety-institute- HTSI-4284508 Or you can email your question to: slombardi@aami.org.

Thank You to Our Industry Partners This Patient Safety Seminar is offered at no charge thanks to funding from our National Coalition for Alarm Management Safety industry partners. The AAMI Foundation and its co-convening organizations appreciate their generosity. The AAMI Foundation is managing all costs for the series. The seminar does not contain commercial content. Platinum Gold Silver 9/25/2013 29

Consider Making a Donation to the AAMI Foundation Today! Making Healthcare Technology Safer, Together Click here to donate online: http://my.aami.org/store/donation.aspx Thank you for your support! 9/25/2013 30