Background Clinical information may be lost during the transfer of critically injured trauma patients from the emergency department (ED) to the intensive care unit (ICU). The aim of this study was to investigate the causes and frequency of information discrepancies with handover and to explore solutions to improving information transfer.
Methods A mixed-methods research approach was used at our level I trauma centre. Information discrepancies between the ED and the ICU were measured using chart audits. Descriptive, parametric and non-parametric statistics were applied, as appropriate. Six focus groups of 46 ED and ICU nurses and nine individual interviews of trauma team leaders were conducted to explore solutions to improve information transfer using thematic analysis.
Results Chart audits demonstrated that injuries were missed in 24% of patients. Clinical information discrepancies occurred in 48% of patients. Patients with these discrepancies were more likely to have unknown medical histories (p<0.001) requiring information rescue (p<0.005). Close to one in three patients with information rescue had a change in clinical management (p<0.01). Participants identified challenges according to their disciplines, with some overlap. Physicians, in contrast to nurses, were perceived as less aware of interdisciplinary stress and their role regarding variability in handover. Standardising handover, increasing non-technical physician training and understanding unit cultures were proposed as solutions, with nurses as drivers of a culture of safety.
Conclusion Trauma patient information was lost during handover from the ED to the ICU for multiple reasons. An interprofessional approach was proposed to improve handover through cross-unit familiarisation and use of communication tools is proposed. Going beyond traditional geographical and temporal boundaries was deemed important for improving patient safety during the ED to ICU handover.
- Critical care
- Emergency department
- Team training
Statistics from Altmetric.com
Trauma patients represent some of the most critically ill patients admitted to the hospital. In addition to the severity of their injuries, they are at high risk of harm, not infrequently due to communication errors. While the causes of preventable and potentially preventable trauma deaths are multifactorial, improvements in communication remain a common strategy to mitigate avoidable errors.1 According to the Joint Commission, two out of every three in-hospital deaths are due to communication errors as they cause the majority of sentinel events.2 ,3 This preventable mortality occurs in up to 10% of critically injured trauma patients, with otherwise survivable injuries, admitted to level I trauma centres in the USA.4 The majority of these deaths occur in the intensive care unit (ICU) and are related to haemorrhage and a delay in appropriate intervention.5 Similar to the standardisation of trauma resuscitation with the promulgation of the Advanced Trauma Life Support course,6 ,7 the standardisation of trauma communication has been proposed as a method to reduce such errors.8 ,9
Patient handover from one unit to another represents a vulnerable time for communication errors that result in the loss of clinical information.10 The transfer of critically ill patients between units has been implicated in fragmentation of patient care, potentially resulting in critical incidents or in omissions in the care being delivered.11 Structure and process consistency are thus paramount to reduce errors and improve outcomes, and standardisation tools such as checklists, within the context of a culture of communication have been proposed for improved outcomes.12 ,13 This decrease in variability of the information transmitted reduces the amount of information loss,10 facilitates communication within the constraints of an 80 h work week14 and may lead to improved patient outcomes, including reduced mortality.15 Trauma patients, in particular, would benefit from a cohesive handover of information, given their complex presentations.16 Missed injuries in trauma patients remain associated with significant morbidity and mortality.17 ,18 Trauma patients are also at higher risk of information disruption during the transfer of care, especially when patients are more severely injured and resuscitation is more time-dependent.19 Despite the significant risk of information loss, no standardised system for handover for trauma patients exists.
While handover, in general, has been identified as a vulnerable time for information loss, the causes and frequency of such information discrepancies remain unknown, particularly in trauma centres. Specifically, the discrepancies that occur during the process of handover of time-sensitive, critically injured trauma patients from the emergency department (ED) to the ICU require elucidation. Our primary aim was to determine the frequency and causes of lost information for trauma patients transferred from the ED to the ICU in our level I trauma centre and importantly, why. Our secondary aim was to explore solutions proposed by trauma team leaders and nurses of the ED and ICU. We hypothesised that important clinical information was being lost and that the reasons for these losses were complex and unit-specific.
We used a mixed-methods research design with a combination of chart audits, focus group interviews with nurses from both units and individual interviews of physician trauma team leaders. This allowed us to quantify information lost due to variation and to gain insight into the processes and events that led up to the observed variation.
This study was conducted from November 2011 to June 2012 and involved participants from the ED, ICU and trauma division of our level I trauma centre in Toronto, Canada.
We used an embedded mixed-methods design.20 Chart audits were followed with focus groups and individual interviews to provide explanations for the quantitative findings.
Quantitative strand: chart audits
Every third chart of trauma patients admitted from the ED to the ICU over a 6-month period was sampled. Injuries, interventions and significant past medical histories were compared between the trauma team leader ED discharge and the ICU admission note. Discrepancies were categorised by organ system. Parametric and non-parametric analyses were applied as appropriate.
Qualitative strand: focus groups (ED and ICU nurses) and individual interviews (trauma team leaders)
Participants were recruited from both the ED and the ICU.21 Nurse participants included experienced registered nurses working in the ED or ICU. Trauma team leaders included emergency medicine physicians, orthopaedic and trauma surgeons. Potential participants were excluded if they felt uncomfortable expressing opinions in a group, or disagreed with having their opinions tape-recorded for later transcription.
We conducted six focus groups with day and night ED nurses (n=3) and ICU nurses (n=3). Participants were given the option to leave at any time or to decline to answer any questions. No gifts were given to participants. Individual trauma team leader interviews occurred in private offices.
Interview questions were semi-structured and open-ended. Each focus group had similar questions (box 1). Trauma team leaders were asked the same questions.
Open-ended questions posed to the intradisciplinary staff about handover
What are the perceptions of the challenges of handover, if any?
How would you describe the challenges with handover?
What constitutes a poor handover?
What are the consequences of a poor handover?
How is lost information retrieved?
How would the ideal handover look and why?
What can be done to ensure that the information handed over is correct and full?
Notes were taken to include observations made on verbal and non-verbal language. Sessions were audio recorded. A debriefing session between the interviewers was held immediately following the focus groups and interviews. Audio-recorded focus groups and interviews were transcribed and managed using HyperRESEARCH (V.2.8.3).22 Thematic analysis was applied to focus on examining themes within the data.23
Fifty charts were audited. None were excluded due to missing ED or ICU notes. Demographics are displayed in table 1.
About 48% of these patients required some form of operative intervention. The mortality rate was 26%.
Clinical information in the trauma team leader note was missing from the ICU admission record for 12 patients (24%). Missing information was categorised by injuries according to organ system, past medical history and change in interventions (table 2). Recovered clinical information denotes injuries or past medical histories previously missed that were discovered at a later time. We considered recovered clinical information discovered in the ICU as ‘information rescue’.
About 24% of patients had injuries that were not transmitted in handoff. Missed injuries averaged 1.4 per patient. Injuries were missed even though the intensivist responsible for the ICU admission note had access to the trauma team leader note, where the injuries were documented.
Most missed injuries were neurosurgical, including intracerebral haemorrhages, calvarial or cervical spine fractures. Information about the patient's past medical history was also missed, however, at a lesser rate than injuries. Interestingly, neurological disorders, including a previous diagnosis of traumatic brain injury or epilepsy, were also the most commonly missed past medical histories. Patients with missed injuries on the ED note were significantly more likely to also have missed past medical histories when admitted to the ICU (Fisher's exact test statistic=0.0013, p<0.05).
Nineteen patients (38%) had new clinical data on ICU admission notes that were not identified or included in the original ED note. Patients with recovered clinical data in the ICU note were more likely to have had missed injuries on the trauma team leader note prior to admission to the ICU (Fisher's exact test statistic=0.005, p<0.05).
Overall, 48% had information discrepancies recorded in their ED or ICU notes. The clinical management of 32% of patients changed due to recovered information discovered in the ICU (information rescue). More patients with information discrepancies experienced a change in their clinical management, compared with those without (Fisher's exact test statistic=0.0085, p<0.05).
Focus groups and individual interviews
Six focus groups with 46 participants in total were convened, consisting of 22 ED nurses and 24 ICU nurses. One ICU nurse declined to participate on learning that the interviews were tape-recorded. Nine trauma team leaders agreed to participate in individual interviews.
Both nurses and trauma team leaders described the ideal handover as quiet, organised and thorough with one designated leader handing over all clinical information to an attentive ICU team. All injuries would be described as well as all interventions done for the patient, in a standardised fashion and quickly. CT findings would be shared with nurses for a complete handover.
The ideal handover? The trauma ABCs and a complete head to toe, starting with the laceration on the back of the head. Um, for example, we had a patient here. Nobody could understand why they were… their hemoglobins were so low. They had a gash on the back of the head like this, about that deep. Patient turned; there were at least four units of blood on the bed that had been missed. We're now two hours into resuscitation of the patient that should've not have gone that long. Um, so they start at the top of the head. They should know all neuro. Chest. Any chest injuries. Abdomen. Bones. Pulses. Everything. (ICU-2-8)
Conversely, a difficult handover was described as chaotic, noisy with no clear leadership or a leader with poor communication skills. A difficult handover had multiple, parallel handovers segregated by disciplines. Handover chaos increased in parallel with the haemodynamic instability of trauma patients.
I mean, I think the biggest problem is the doctor transfer though. They're very poor at that, some of them. Some of them haven't a thorough assessment of the patient in emerg. and, uh, it's up to us to find many more injuries than was diagnosed in emerg. (ICU-1-4)
It's just mad chaos up here trying to stabilize somebody…It's just, you know, chickens with their heads cut off running around and there's no order. (ICU-3-5)
This ‘chaos’ led to increased stress and resentment towards the other unit members and was felt to lead to a loss of transferred clinical information. This information loss was then considered to increase patient harm, including preventable mortality. Trauma team leaders identified similar challenges with handover.
So frankly it's chaos. And um, it's unstructured. And it's unclear precisely who I should be talking to because the nurse most responsible isn't identified. (TTL-9)
The majority of nurses attributed a chaotic handover to be due to physician disorganisation, while, as the quote above suggests, the trauma team leaders attributed this occurrence to nurses and their role.
Table 3 organises themes and subthemes that explore the reasons behind challenges with handover.
The handover of clinical information was perceived as lacking a formalised structure, requiring standardisation. The majority of nursing participants agreed that each physician acting as a trauma team leader had a different approach to delivering patient information. At times, handover was omitted altogether. This variability in handover was compounded with new trauma trainees of different disciplines. This lack of standardisation led to increased time needed for handover, causing delay in patient care. Unlike nurses who are trained to follow well-established communication theory, trauma team leaders were not formally or informally taught how to give handover of patient information in a concise and reproducible fashion.24 The more acutely unwell a patient was, the noisier and more distracting handover became. It also remained unclear who was in charge of leading, organising and delivering the handover of patient information, or who was receiving this information. Leadership during handover was felt to be lacking. There was a generalised sense of physician disorganisation.
“I think a lot of times…there is confusion in the trauma room because there's ten different doctors giving ten different orders and with everything being pretty messy. No clear leadership so handover is confusing or missed… (ED-1-4)
Conversely, tension was generated with the ICU nurses as it was unclear what had been done to the patient and what remained to be done for ongoing resuscitation. Physician–nurse communication was also limited at times.
It's usually pretty stressful… the trauma team leader sometimes doesn't tell us what injuries they see on scan. You need to be alert in there instead of going through the chart and reading up on everything to be able to give report so you don't get in trouble for giving a bad report. Do you know what I mean? (ED-3-1)
The doctors talk amongst each other. There's no real information given to the beside nurse. (ICU-1-1)
Study participants described the ED culture as stressful, fast-paced and unpredictable. There was a perception that other hospital units, such as the ICU, did not fully appreciate these working conditions. We discovered a divergence in opinions between nurses and physicians in terms of who was responsible for such chaotic handover, with each discipline assigning responsibility to the other.
Intradisciplinary ED and ICU nurses both described how their colleagues did not understand the stressors of one another's units. For example, ED nurses described feeling ‘judged’ by the ICU nurses for incomplete resuscitation and disorderly delivery of the patient (‘sheets bloody’, ‘tangled IV lines’), with negative clinical repercussions for the patient.
It's a lot of passive aggressive stuff. It's not…it's not good for your patients. (ED-1-2)
Physicians, in contrast, felt that the communication and leadership they provided was adequate. The trauma team leader participants did not describe or acknowledge intradisciplinary and interdisciplinary tension. ‘We get along well as a team’ was a common perception among physicians. Again, physicians rarely recognised that they contributed to variability in handover.
It's never been that big of a problem dropping people off in the ICU…It generally rolls pretty smoothly. (TTL-2)
There was recognition, however, of some disorganisation on a nursing level with handover in the ICU. It was unclear at times which nurse was the receiving nurse for the injured patient.
It is unclear who the primary recipient of this information is. (TTL-4)
Physicians also felt frustrated at the delay in the process of handover due to unavailability of ICU beds and needing to return the trauma patient back to the ED.
The patient needs to be getting ICU care, not ED care. (TTL-7)
Overall, the perception of how handovers were structured varied between nurses and physician participants. This caused communication challenges that were described as an impediment to handovers from the ED to the ICU.
Shared themes and solutions
Both nurses and physicians considered clinical outcomes to be worse due to information lost on patient handover. Improvement in communication started with clear identification of all team members.
The majority of study participants expressed frustration at not clearly knowing the extent of injuries or what the overall plan of care was for the patient, which in turn, increased their stress.
Cause often times too, they'll just say a GCS of a number but then it's not like specifically what is it that they're doing, you know, neurologically…Like what is it that they were doing down in emerg. You know, what were their pupils? ‘Cause then all of a sudden, you open a pupil and it's blown and you're like, why? (ICU-3-2)
There was also significant stress generated when patients were described on handover as ‘hemodynamically normal’ when in reality they required ongoing and aggressive resuscitation. Anecdotes were shared that described preventable mortality related to chaotic, disorganised handover.
The majority of study participants viewed standardising handover as a necessary step to improve quality and patient safety. It was understood by all participants that the trauma team leader should be leading a unified handover with multidisciplinary colleagues contributing missing details (‘filling in the blanks’). Specifically, reducing variability in the structure and process of handover was seen as necessary to improve the clinical outcomes for critically unwell patients. Communication tools used to standardised handover, with or without a formal checklist, were suggested as one method to reduce variability.
With all the chaos with sick trauma patients, we need more than new checklists for nurses to fill out. We need one unified handover where the information given is needed and given the same way each time. That way we don't forget important details. (ICU-1-2)
Both nurses and physicians felt that a checklist would be a feasible and necessary tool to implement but only within a global improvement in the overall culture of handover, communication and safety. Participants felt this improvement could be achieved by physician non-technical training in communication, prior to working as a trauma team leader.
We need more non-technical training to learn how to communicate and lead a team (TTL-1).
Understanding different perspectives was also considered important for improving communication between units. One programme that was considered very successful was the exchange programme with nurses working in one another's units to understand local culture.
“Now we had a really great exchange with the ICU where we got to do a day there and they got to do a day down here.” (ED-1-2) “You go from an environment of fast-paced chaos to an environment of controlled slow pace. So the mindset of the two different departments is very different, right?” (ED-3-7)
Additionally, physicians offered the perspective that having the ICU staff witness the care that occurs in the trauma bay during the early stages of resuscitation would also be very helpful.
Handover would be probably smoother and a little bit more, I guess, inclusive of all the nuances…if one of the ICU people whether it's the charge nurse or the potential receiving nurse or the fellow went down and sort of saw the patient in the ED. (TTL-9)
Nurses felt that they must be among the key players in the development of a culture of safety and communication both in the ED and ICU units. Interdisciplinary communication with follow-up was considered important even after formal handover had been completed.
The follow up that the trauma team leader wanted wasn't being done so she just came herself and made sure it was done which was really nice from a nurse's perspective because you, I don't know, you're not banging your head against a wall…so it was nice. So follow-ups. Especially for our newer docs that are here (in the ICU). I mean, I think they appreciate it too. (ICU-3-3)
Solutions proposed were similar between nurses and physicians, extending beyond the use of a checklist, with an emphasis on a unified handover (with key team members identified), training on how to give handover and early cross-unit communication.
A key finding of this study is that 48% of critically injured trauma patients had some form of information discrepancy that occurred during handover from the ED to the ICU, including information about injuries and past medical histories. Omitting data on injury was significantly associated with omitting a portion of the medical history, which places trauma patients at further risk of information loss and potential harm. However, these patients were also more likely to have their information recovered in the ICU. This ‘information rescue’ was frequently associated with subsequent changes in patient management.
Participants identified that variations in handover led to significant objective and subjective information loss. Stressors were identified on interdisciplinary and intradisciplinary levels. At times, nurses felt that physicians were responsible for challenges with handover and vice versa. Additionally, ED and ICU nurses had their own challenges in communication. Losing valuable clinical information was felt to contribute to worse patient outcomes (potentially including mortality) while increasing stress among staff participating in handover.
We also identified that a checklist alone would be insufficient to address information loss. Participants highlighted the importance of standardising the process of handover. In their view, this process should be multidisciplinary and led by the trauma team leader with prior training in communication. Also the traditional notion of handover occurring in one place, at one time, was challenged with the invitation for ICU representatives to observe early treatment in the ED, and conversely for the trauma team leader to ensure ongoing follow-up in the ICU, long after the completion of the ‘official’ handover.
While communication failures are common in trauma care, there is little evidence on how to improve the process.25 In particular, trauma patients are at particular risk of deviation from the natural progression of a procedure during transport to the ICU or the operating room.19 A recent prospective, observational cohort trial reported flow disruptions, especially errors in communication, that compromised trauma patient care.26 We confirmed that communication failures are common in trauma and increase the risk of error generation.27 We found that trauma patients at risk of missed injuries were also at risk of loss of information about their past medical histories. We also confirmed that 25% of disruptions in information transfer led to challenges with clinical progression or a change in management. The causes of this information loss remain unknown. In our study we identified important local challenges. Handover was often performed differently, depending on the trauma team leader. At times, key team members were not identified, or leadership was lacking to ensure a smooth process for handover. Tension existed both within and across disciplines. While physicians and nurses assigned blame on the other, all acknowledged the trauma team leader as having the ultimate responsibility for handover. Dealing with trainee trauma team leaders was also a source of stress, leading to increased chaos. While recent studies in trauma have also recognised challenges in handover in trauma (specifically from the field to the ED), these studies in general lack a deeper analysis of the role of system and culture in why this occurs, or solutions.16 ,28 ,29
Many proposed solutions focus on improving handover through the use of checklists. While checklists are appealing due to their simplicity, we identified barriers to their use. All of our multidisciplinary participants recognised a checklist as a potential tool to reduce variability in handover, but it was viewed as a new ‘paperwork’ burden for nurses with the potential for ‘checklist burnout’. While some major well-cited studies have highlighted the usefulness of checklists in medicine and surgery,30 ,31 in other large, population-based trials, mandatory use of checklists did not significantly improve patient outcomes.32 Other studies have demonstrated that successful use of checklists may simply reflect effective and complex local culture change in processes of care and a culture of safety.33 ,34 Thought leaders in the patient safety field have also called for caution in viewing checklists as a panacea for rectifying errors in communication and threats to patient safety.35 ,36 In fact, since the completion of this study, a trauma care checklist has been piloted in our institution, with significant barrier to use and implementation, due to challenges similar to those highlighted in this study.37 Solutions proposed by our participants included the need for member and leader identification, with a unified, multidisciplinary handover (possibly involving a checklist) with mutual understanding of cross-unit cultures. These results are consistent with the patient safety literature that recommends identifying and solving local unit safety challenges prior to implementing structure or process change for quality improvement.33 ,38
Our study considered solutions as proposed by participants of all units involved in the handover of critically ill trauma patients, including ED and ICU nurses and physicians. Some of their perspectives overlapped while others diverged. The majority of the tension that existed between the ED and the ICU nurses resulted from a lack of understanding challenges in local culture. This was addressed with an ongoing ‘exchange’ programme, which deployed nurses to a new unit, thus exposing them to a different culture. Previous work has focused on nursing handover from one unit to another,11 but we are unaware of studies that have explored the concept of cross-unit exchanges specifically to improve communication and handover. Similarly, the process of handover has, by definition, been considered to occur in one place at one time.39 In our study, our participants had overlapping perspectives on the importance of starting handover early, with ICU presence in the ED, and longitudinally, with the trauma team leader following up on patients long after ICU admission. This challenges the current concept of handover as a static event occurring in one place at one time.
Future studies will need to investigate the impact of loss of clinical information, and subsequent rescue, on clinical outcomes. Additionally, addressing the safety culture within each unit and the ‘neo-culture’ created where and when both intersect merits further study to address challenges prior to checklist implementation. The ultimate outcomes to measure will be a reduction in information loss and plausibly adverse clinical outcomes, once both unit safety culture and checklist implementation are complete.
There are several limitations to this study. We were limited by the data extracted from our chart audit. It was difficult to ascertain if the preponderance of missed information was due to a worse injury severity score of the patient, or due to systematic errors occurring with the same trauma team leader or to other causes. We were also unable to fully demonstrate if the changes in clinical management of patients with information rescue led to an improvement in clinical outcomes or not. Future studies will need to broaden this data capture to fully explore these details, with a focus on clinical outcomes. Also, we elucidated proposed solutions from both nurses and physicians of one level I trauma centre in Toronto, Canada. It is unclear, however, given the unique cultural milieu of each hospital, if these solutions can be applied to other Canadian centres or beyond.
Our study found that the transfer of critically injured trauma patients from the ED to the ICU was accompanied by a significant loss of important clinical information. Both physicians and nurses had varying perspectives of why this occurred. A change in the culture of safety, with improvements in both interdisciplinary and intradisciplinary communication is needed prior to the implementation of communication tools used in quality improvement and patient safety endeavours.
Twitter Follow Tanya Zakrison at @tzakrison
Contributors TLZ, BR, AM, and SR: planning, conducting research and manuscript writing with quantitative data analysis. AJ and SS: qualitative data analysis and interpretation. CA, CS, SR and NN: analysis and interpretation of overall data with critical revisions to manuscript.
Funding St Michael’s Hospital Innovation Fund.
Competing interests None declared.
Ethics approval The institutional review board from St Michael's Hospital approved this study.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.