In medical emergencies, time is of the essence. So is the quality of communication. A team of researchers from York University, University Health Network (UHN) and Trillium Health Partners studied text messages sent between nurses and physicians in deteriorating internal medicine patients requiring escalation to intensive care unit (ICU) to identify issues in failures to rescue. Looking at records from 2012 to 2014 at the Toronto General Hospital, the team found that message quality was positively linked to survival.
This new research, published this spring in the International Journal of Medical Informatics, provides policy-makers and hospital administrators with much-needed evidence to directly affect patient outcomes. It highlights the need for a standardized and responsive text-based communication system.
Researchers wanted to learn about the quality of the text messages between doctors and nurse, and if this were related to patient death
This research will help inform future policy initiatives and new approaches for hospital administrators.
Hannah Wong
“Communication about critically ill patients could be improved. There appear to be significant gaps in the quality of messages, their responses and delays in rapid response team (RRT) activation,” wrote lead author Hannah Wong (Assistant Professor in York’s Faculty of Health) and colleagues.
Wong, a PhD in Industrial Engineering, utilizes statistical regression modeling techniques, in conjunction with system dynamics computer simulation, to study problems important to clinicians, health care managers and policy-makers. The goal is to guide the better design of policies to improve appropriateness and quality of care.
Text-based communications more effective than traditional pagers
Text-based communication is gaining popularity in the hospital sector and has been shown to be more efficient than traditional pagers. Since 2011, the Toronto General Hospital, part of UHN, has been using secure electronic text messaging.
Additionally, frameworks such as SBAR (Situation, Background, Assessment, Recommendation) are highly recommended tools to improve communication between nurses and physicians and have been found to reduce unexpected deaths.
“Communication about critically ill patients could be improved. There appear to be significant gaps in the quality of messages, their responses and delays.” – Hannah Wong
Study findings fill major void
Within this text-based context, the research team found that little is known about the communication between nurses and physicians about critically ill, or deteriorating, patients.
More specifically, the researchers wanted to know the following:
- How many patients had a text message communicating deterioration?
- What was the quality of the message?
- Was message quality related to death?
Researchers studied 236 ICU admissions from internal medicine
The team reviewed all 236 admissions from General Internal Medicine (GIM) to ICU between January 1, 2012 and August 31, 2014. The majority of patients (160 patients representing 68 per cent) either did not have a critical text message that met calling criteria for RRT activation in the 48 hours prior to ICU transfer or already had RRT activation prior to a text message.
A critical message is defined as one that involves respiratory distress, oxygen saturation or an abnormal level of oxygen in the blood, an elevated heart rate or blood pressure, or a severe change in mental status.
The remaining 76 patients had a critical message. All critical messages in the 48-hour period prior to ICU transfer were analyzed for RRT calling criteria, time to RRT activation, message quality, presence of vitals, and the quality and timeliness of physician response.
Table: Examples of critical messages and responses |
|
| Legend: RR = respiratory rate; NP = nasal prongs; bpm = beats per minute; BP = blood pressure; pt = patient; NS = normal saline; TM = tracheostomy mask; A&O = alert and oriented; ABG = arterial blood gas; CCRT = critical care response team (rapid response team) | |
| Critical message from nurse | Response from doctor |
| pt has increased RR. desaturating to 88% will put him on 2L NP. he is also tachy after transferring from chair to bed at 125bpm. his BP at 1700 was 83/49. Pls come and assess pt thnx. | Will assess shortly Please give 500cc NS bolus over 1 h |
| Hi, pt noted to have fever 39.4, tachy at 108, 02 sat 88% on 32%TM, changed to 50% TM o2 sat improved 93%. Please assess. Thanks | Thanks |
| Hi, Pt more confused compared to yesterday, A&Ox1, not responding to questions appropriately or not responding at all. please assess, thanks | Thanks |
| Hi, critical lab value returned for ABG, PCO2 = 99, pH = 7.22, Bicarb = 40, thanks | Will call CCRT now |
Researchers found that message quality was positively correlated with in-hospital survival
Study findings point to areas for improvement
The key findings reveal some shortcomings in the system:
- In only 40 per cent of cases was text messaging used to relay critical information about patient deterioration.
- Message quality, specifically, the number of vital signs and SBAR elements, was poor: Only 45 per cent of messages contained two or more vitals, and only 3 per cent of messages contained SBAR (Situation, Background, Assessment, Recommendations).
- There was a median delay of nine hours from critical message to RRT activation in patients where RRT was activated after a critical message was sent.
- Physicians responded to critical messages quickly, but response quality was poor with nearly a quarter of responses only acknowledging receipt.
Study findings showed that message quality was poor: Only 45 per cent of messages contained two or more vitals.
This research will help inform future policy initiatives and new approaches for hospital administrators. “Educational and continuous improvement efforts in the adoption and optimization of text messaging capabilities within clinical messaging systems could improve rapid response and the quality of communication from nurses and physicians,” says Wong.
The article, “An analysis of messages sent between nurses and physicians in deteriorating internal medicine patients to help identify issues in failures to rescue,” was published in the International Journal of Medical Informatics (April 2017). To learn about this research, visit the website.
By Megan Mueller, manager, research communications, Office of the Vice-President Research & Innovation, York University, muellerm@yorku.ca