 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
a Department of Emergency Medicine, Royal Children's Hospital, Melbourne, Australia
b Children's University Hospital, Zurich, Switzerland
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
METHOD. The resuscitation of children who presented to the emergency department of the Royal Children's Hospital between February 19, 2001, and August 18, 2002, for whom the trauma team was activated was video recorded. The tapes were analyzed, and management was compared with Advanced Trauma Life Support guidelines. Deviations from these guidelines were recorded as errors. Fifty video recordings were analyzed independently by 2 reviewers. Medical record review was undertaken for a cohort of the most seriously injured patients, and errors were identified. The errors detected with the 2 methods were compared.
RESULTS. Ninety resuscitations were video recorded and analyzed. An average of 5.9 errors per resuscitation was identified with this method (range: 1–12 errors). Twenty-five children (28%) had an injury severity score of >11; there was an average of 2.16 errors per patient in this group. Only 10 (20%) of these errors were detected in the medical record review. Medical record review detected an additional 8 errors that were not evident on the video recordings. Concordance between independent reviewers was high, with 93% agreement.
CONCLUSIONS. Video recording is more effective than medical record review in detecting management errors in pediatric trauma resuscitation. Management errors in pediatric trauma resuscitation are common and often involve basic resuscitation principles. Resuscitation of the most seriously injured children was associated with fewer errors. Video recording is a useful adjunct to trauma resuscitation auditing.
Key Words: video recording • trauma • resuscitation
Abbreviations: CCRTF—Consultative Committee on Road Traffic Fatalities • ISS—injury severity score • ED—emergency department • RCH—Royal Children's Hospital • ATLS—Advanced Trauma Life Support
ASSESSING TRAUMA RESUSCITATION management has always been an imprecise science. Most published works are retrospective reviews of the medical records of adult patients, with only a few studies using real-time video recording and fewer looking at the pediatric population. Studies from Australia1,2 and the United States3 identified errors in trauma management, including errors that contributed to death. All of these studies used retrospective medical record review and likely underestimated the size of the problem. The Consultative Committee on Road Traffic Fatalities (CCRTF) identified a minimum of 3 errors per resuscitation in its review.2 The CCRTF also noted that the majority of resuscitation errors occurred in the emergency department (ED).1
For some time, video recording has been used to evaluate patient management. Video recording has been shown to be a useful practical tool.4–8 It has the advantages of providing accurate times to interventions that are not hindered by poor documentation or the memory of those involved.
The Royal Children's Hospital (RCH) (Melbourne Australia) is a 240-bed tertiary pediatric hospital and the state's only pediatric trauma service. There are 
1500 trauma admissions per year, of which 150 are defined as major trauma, ie, meeting 
1 of the following criteria: injury severity score (ISS)9 of >15; abbreviated injury score9 of 2 in 2 ISS body regions (excluding external genitalia); admission to an ICU for >24 hours; requiring mechanical ventilation; urgent surgery for treatment of intracranial, intrathoracic, or intraabdominal injury or for fixation of pelvic or spinal fractures; or death after injury; 210 are classified as serious injuries, ie, major trauma or ISS of >11. Because this is the only pediatric trauma center for the state of Victoria, 50% of the major trauma cases are transferred from other institutions. In our institution, there are 2 areas that receive trauma patients for resuscitation, the ED and the ICU. Children who arrive by helicopter and need urgent resuscitation or are intubated are examined and resuscitated in the ICU. Approximately 40% of all major trauma admissions are treated directly in the ICU.
There are specific guidelines controlling the activation of the trauma team, and a response by all members is expected to be immediate. A decision was made to video record the trauma team resuscitating children in the ED, with the aims of (1) auditing the trauma team response, (2) investigating the ability of video recording to identify management errors made by the trauma team, and (3) determining whether video recording provided information that was not available in the medical record.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
The camera used is a ceiling-mounted, Panasonic WV-651, analog, combination video/audio camera. Each camera can be activated through a wall-mounted switch in the resuscitation room and has a remote control in an adjacent room for panning and zooming. Each camera is linked to a Panasonic AG-TL700 SVHS video recorder, which is also located in the adjacent room. Remote manipulation of the camera was used rarely in the study.
Before commencement of the study, the errors to be assessed were defined and acceptable deviations from management were agreed on. Advanced Trauma Life Support (ATLS) guidelines for the resuscitation of seriously injured patients were used as the standard (eg, a delay to oxygen therapy of >5 minutes was agreed as an error, and 1 intravenous cannula for a child who required fluid resuscitation or intubation or who had obvious injuries to 2 body regions was considered inadequate). Initially, the investigators reviewing the video recording did so together and agreement on interpretation of events was reached. All subsequent videotapes were reviewed by Dr Stocker or Dr Staubli, with 50 tapes also being reviewed by Dr Oakley. No tape was reviewed by a person involved in the resuscitation. Comments were added to explain omissions of care that were deemed reasonable for the individual patient. All videotapes were viewed within 2 weeks after the resuscitation and then erased.
Data collected included personnel present and timing of their arrival; trauma team leadership and communication; management of the airway, breathing, and circulation; management of the spine; resuscitation undertaken and interventions performed at these stages; completeness of the secondary survey; and time to definitive care. The trauma service database was used to identify patients with an ISS of >11. Medical records of these patients were reviewed to determine whether the history documented the resuscitation errors seen on the videotape and whether management was documented in the history that was not seen on the video recording.
Patient outcomes were assessed by 2 surgeons who were blinded to the management errors identified. They gave a consensus opinion on whether the outcome was poor and whether this was unexpected on the basis of the injuries sustained. All results are presented in a descriptive format. Interrater reliability was determined with the 50 videotapes that were reviewed by 2 reviewers.
The project was reviewed by the RCH Ethics in Human Research Committee and was approved as an auditing tool. The RCH medicolegal office was consulted regarding privacy issues for staff members and patients and advised that, as an auditing tool, the project was exempt from existing privacy legislation.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
|
Errors Identified in Patient Assessment and Management
A total of 532 errors of management were identified, with an average of 5.9 errors per resuscitation; 219 errors involved management of the airway and breathing, 124 errors involved management of the circulation, 32 errors were failures of spinal immobilization, and 157 errors occurred with the secondary survey. There was no resuscitation that was free of errors, with a range of errors of 1 to 12. Management errors are described in detail in Table 2.
|
The only detected error in cervical spine management was not providing head stabilization while transferring the patient. No patients had the cervical spine cleared before arrival or before the video was started. The Victorian ambulance service treats all potential major trauma with a cervical collar.
In management of the circulation, 36 (40%) children had 1 intravenous cannula, 32 (36%) had 2 cannulas, and 22 (24%) had no intravenous access. Not all of the children were seriously injured, and the reviewing researcher recorded that the level of venous access was inappropriate in 18 cases (20%). To be regarded as an error, a child with 1 or no intravenous cannula had to need an intravenous fluid bolus, to have expected injury to 2 body regions on the visual examination, or to need intubation. The size of the cannula could rarely be determined on the videotape.
Thirty-seven patients (40%) received a fluid bolus. Volumes of fluid delivered were difficult to assess, and vital signs were often not visible on the videotape; therefore, the need for and adequacy of fluid resuscitation were difficult to determine (Fig 1). Circulation assessment was adequate for 31 patients (35%).
The secondary survey was incomplete for 71 patients (79%) before the patient left the resuscitation room. The abdomen and chest were not examined for only 2 and 3 patients, respectively.
Errors Among Seriously Injured Children (ISS of >11)
Children with an ISS of >11 represented 28% of our patients (25 patients). There were a total of 54 errors made in this group, an average of 2.16 errors per patient. Fifteen errors for 5 children involved management of the airway and breathing, 23 errors for 12 children involved management of the circulation, and 12 children received an inadequate secondary survey. These errors are described in detail in Table 3.
|
|
Patient Outcomes
There were 10 patients with poor outcomes, ie, 4 deaths and 6 children who remained with significant disability. None of these outcomes was judged to be unexpected from the severity of the injury (Table 5). Five patients were documented to have unexpected complications. Two of these were related to surgery (1 wound infection and 1 complication of external fixation), 1 was related to drug error (a patient who was allergic to penicillin received flucloxacillin after admission and developed a rash), and 2 were infections related to intravenous access.
|
score for the interrater agreement was 0.83 (95% confidence interval: 0.7–0.96). |
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
We have demonstrated that video recording, if used alone, has definite limitations. It was difficult to assess accurately the clinical state of the patient on the video recording without contemporaneous vital signs. This restricted the ability of the reviewers to judge the timeliness and effectiveness of interventions. Combining video recording with a direct stream of vital signs from a central monitoring system that allowed continuous recording would overcome this problem.
Video recording gave only limited information on drug doses and fluid volumes. These were recorded more accurately in the medical record. Medical record review still needs to be performed to check these aspects of resuscitation.
The high rate of errors detected in our study is partly attributable to the effectiveness of video recording and partly attributable to the strict definition of what constituted an error. Deviations from ATLS guidelines were recorded as errors and reasonable timelines were agreed on to ensure timely care irrespective of the severity of the child's injuries. It is possible that experienced clinicians would not adhere to all of the ATLS guidelines for all patients that we video recorded, especially those with more minor trauma. It could be questioned whether ATLS guidelines were appropriate for the 37 patients with an ISS of <5. However, because they had already met the criteria for trauma team activation, we thought this was reasonable, given the difficulties in ascertaining the degree of injury initially. Indeed, far fewer errors were detected in the more seriously injured group, which reassured us that, on the whole, this group received appropriate care.
Our results indicate that management errors in pediatric trauma resuscitation are frequent (an average of 5.9 errors per patient resuscitation). The errors made involved basic resuscitation principles. There were no errors of intubation or other invasive interventions. The error rate was greater than that found by the CCRTF, which documented 3 errors per resuscitation in the ED.2 Like most other reviews of resuscitation,1–3 that review was a retrospective review of hospital and ambulance records and would have missed some errors because of lack of documentation. Indeed, our history review showed that it would have missed up to 80% of simple management errors.
The review of timing of the arrival of team members provided useful information. Historically, the trauma team leader in our institution was the surgical registrar. After we identified that they were not present at the beginning of resuscitation in 50% of cases and that this created confusion among the team, the team leader role was moved to the emergency physician. To address the pattern and frequency of errors, we instituted 12 hours of teaching, focusing on basic principles of trauma management, which all members of the trauma team receive when beginning at the hospital.
There is a growing body of evidence that video recording can be a useful tool in resuscitation evaluation.4–8 Those studies compared trauma resuscitation to ATLS guidelines and neonatal resuscitation to existing guidelines. They documented errors that occurred, but only the neonatal review examined the frequency of errors. Unlike Carbine et al,6 who demonstrated a 48% perfect neonatal resuscitation rate, we were unable to demonstrate a single resuscitation that was devoid of error. Some feedback on the management and communication errors was given to the trauma team. However, there was no difference in behavior or errors at the beginning versus the end of the study.
The patients in this study who had poor outcomes experienced a number of management errors. Some of these errors might have contributed to slow or delayed resuscitation, but none was seen to be the sole cause of the poor outcome This differs from the findings of the CCRTF, which identified an average of 1 error per resuscitation that contributed to death.2 The CCRTF studies only road traffic fatalities, and our population was not as seriously injured.
Patient and staff privacy is an issue that is often put forward as an impediment to video recording.7,10 This study was reviewed by our institutional ethics committee and deemed to be a standard auditing tool. No patient or parental consent was deemed necessary; indeed, it was thought to be impractical. The staff members were educated about the role of the video recording and the auditing process. All staff members agreed to the video recording, and no staff members expressed complaints. Patient privacy issues were investigated with the hospital's medicolegal office, and it was thought that using videotaping as a standard auditing tool exempted the procedure from privacy legislation and it could be used in an ongoing process to improve patient management.
The cost of video recorder installation and the time taken to review the videotapes (on average, 40 minutes) means that not all departments would have the resources necessary to use this tool. The time taken to review the histories (on average, 1 hour) had a significant impact on our ability to continue to audit all trauma resuscitations. We plan to develop an electronic resuscitation record and video recording that can be viewed concurrently, to limit the time taken to audit trauma care.
The number of children in this study limits the ability to identify clearly the relationship between adverse events and management. We documented a number of significant omissions in practice that potentially could contribute to poor outcomes and might have contributed to the outcomes of some of the patients in this study, but the number of poor outcomes was too small to determine this. Although management errors have been shown to be relatively frequent events,2,11,12 the proportion leading to adverse events has been shown to be 1%.11,12 Therefore, many hundreds of patients would need to be studied to identify the errors that are associated with adverse events. Furthermore, the identified difficulties in viewing the videotapes mean that some management errors might have been missed and some management appropriate for the individual child might have been considered an error. One half of the videotapes were reviewed by 1 reviewer. Although the inter-rater agreement was high, there is a potential for bias. We were unable to videotape all trauma resuscitations at our hospital, and excluding patients who were resuscitated in the ICU potentially excluded a more seriously injured cohort.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Address correspondence to Ed Oakley, MBBS, FACEM, Department of Emergency Medicine, Royal Children's Hospital, Flemington Road, Parkville 3052, Australia. E-mail: ed.oakley{at}rch.org.au
The authors have indicated they have no financial relationships relevant to this article to disclose.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  C. P F O'Donnell, C O. F Kamlin, P. G Davis, and C. J Morley Ethical and legal aspects of video recording neonatal resuscitation Arch. Dis. Child. Fetal Neonatal Ed., March 1, 2008; 93(2): F82 - F84. [Full Text] [PDF]
|
|
|
|
 |
|
 T. A. Leone, W. Rich, and N. N. Finer Educational Perspectives: Addressing the Complexities of Medical Education With Video NeoReviews, December 1, 2006; 7(12): e597 - e601. [Full Text] [PDF]
|
|
|
|
 |
|
 J. S. Upperman, D. Balise, Y. Morad, H. Ford, and G. H. Mahour Trauma Video Review: A Valuable Resource Pediatrics, August 1, 2006; 118(2): 839 - 839. [Full Text] [PDF]
|
|
|
|
|
|
E. Oakley, S. Young, S. Stocker, and G. Staubli Trauma Video Review: A Valuable Resource: In Reply Pediatrics, August 1, 2006; 118(2): 839a - 840. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
||
|
||
Home | My Pediatrics | Journal Information | Current Issue | Past Issues | Subscriptions & Services | Contact Us Click here for faster international access © 2006 American Academy of Pediatrics. All rights reserved. |
||
|
||
