In healthcare, communication errors are common1 and may cause clinical incidents and adverse events in relation to surgery.2–4 Attempts have been made to improve communication, one such method being the communication tool ‘Situation-Background-Assessment-Recommendation’ (SBAR). SBAR is recommended for the transfer of critical information in complex environments.5,6 It is thought to make communication consistent, by providing a predictable structure and bridging differences in communication styles,7 and it is recommended by WHO for use in handover situations.5 SBAR imposes the following structure: Situation, what is going on with the patient? Background, what is the clinical background, or context? Assessment, what do I think the problem is? Recommendation, what would I do to correct it?7 The present study reports on the effect of using SBAR during the handover of patients between the operating theatre and the postanaesthesia care unit (PACU) and examines the recall of information given during the process. In the postoperative period, the patient is in a vulnerable position because serious physiological changes can occur rapidly. For this reason, it is essential that the healthcare personnel remember as much as possible of the verbal information conveyed during handover from the operating theatre to the PACU.
Handover is defined as ‘the transfer of information and professional responsibility and accountability between individuals and teams, within the overall system of care’.8 To guarantee safe care, handovers require consistent and clear communication,9 but handovers from the operating theatre to the PACU have been characterised as involving incomplete transfer of information,10,11 being unstructured12 and having poor standardisation.12,13 In addition, perioperative personnel often work under great stress14 in an environment that is event-driven and marked by distractions.15,16 Postoperative handovers have been described as a complex work process characterised by lack of a supporting framework, time pressure and interruptions.17
Although some attention has been paid to the crucial handover situation, there is not enough research to inform evidence-based handover strategies.18 There is a lack of both interventional studies to investigate improvements in handover effectiveness,18,19 as well as studies that include a control comparison group.20–24 Interventional studies in PACU have primarily focused on the handover of patients undergoing surgery for congenital cardiac defects. There is thus a need for studies involving patients from other surgical specialities.17 We hypothesised that implementation of the communication tool SBAR would improve the receivers’ recall of information given during postoperative handover and that these receivers would show better recall than receivers in a nonintervention comparison group. The aim of the study was to investigate whether implementation of the communication tool SBAR affects retention of information by receivers of postoperative handovers.
Materials and methods
The study was approved by the Regional Ethical Review Board in Uppsala, Sweden (reg. no. 2011/061) on 9 March 2011 and was registered at http://www.controlled-trials.com (Identifier: ISRCTN37251313) on 8 November 2012. The head of the clinic made the decision to implement the communication tool SBAR. The participants and the involved patients received written information about the study and were told that participation was strictly voluntary and could be discontinued at any time without explanation. The design was a prospective intervention study, with an intervention and a comparison group as well as preassessments [Time 1 (T1)] and postassessments [Time 2 (T2)]. A consecutive recruitment of handovers during patient transfer of care from the operating theatre to the PACU was used. All personnel working daytime shift and taking part in the handovers between the operating theatre and the PACU at the anaesthetic clinic at the two hospitals (one hospital as intervention group, the other hospital as comparison group) were invited to participate. Participants serving as ‘senders’ were nurse anaesthetists, anaesthesiologists, specialist trainee physicians (physicians in specialist training to be anaesthesiologists) and those serving as ‘receivers’ were Specialist Nurses in Intensive Care (registered nurses with 1-year training and graduation in intensive care) and registered nurses working at the PACU. Four personnel declined to participate. During the observation period, all adult patients planned for surgery received written information about the study along with the letter of notice for the operation. If the patient had agreed that the handover could be included in the study, he/she brought a written consent form at admission to the hospital. The handover was included if the written consent was attached to the patient's anaesthetic record upon arrival at the PACU and if the sender and receiver of the handover had agreed to participate.
The two hospitals are located in central Sweden. At the intervention clinic, 9428 surgical procedures (whereof 33% were unplanned) were performed in 2011, and 9327 surgical procedures (whereof 34% were unplanned) were performed in 2012. The corresponding figures for the comparison group were 6031 (whereof 31% were unplanned) and 6138 (whereof 32% were unplanned), respectively.
Implementation of SBAR in the intervention group meant preparation of a modified SBAR pocket card, an in-house training course, information material and 168 structured observations. The SBAR template prepared by the Swedish Association of Local Authorities and Regions25 was slightly modified by a local inter-professional group (nurses and the anaesthesiologist responsible for the PACU) to adapt to needs at the clinic. The in-house training course was conducted at a Clinical Training Centre and involved 2.5 h of lectures, instruction and role-playing. During the role-play sessions, personnel used the SBAR structure in simulated handovers in groups of three, wherein they took turns being senders, receivers and observers, all under the supervision of instructors from the Clinical Training Centre. During the 4-month-long introduction period (May to September 2011), 80% (155 of 194) of the personnel participated in the in-house training course and the rest were thereafter offered training. The pocket card describing the SBAR structure to be used was distributed to all staff and attached to patients’ tables at PACU, wherein most handovers were conducted. In other places at the anaesthetic clinic wherein handovers were conducted, information about SBAR was made available. In the comparison group, no structured communication system was used.
The 4-month long introduction period was followed by a 7-month long monitoring period during which adherence to SBAR at handovers was measured. It consisted of 168 structured observations of handovers at the anaesthetic clinic, performed by four members of the local inter-professional workgroup. The observations were conducted to determine whether SBAR was used as intended, and the results of these observations were used by the head of department, the managers and local inter-professional group to provide feedback to the intervention group and to monitor the intervention process.
For interpretation of the findings, careful control of the implementation is required.26 During the 7-month long monitoring period, one author (M.R.) contacted staff to collect data on their perceptions of the intervention. Using a random sample of 10 personnel each month, except for 1 month when only six personnel were reached, structured telephonic interviews were conducted. In total, 66 personnel from the whole clinic were interviewed, of whom 27 worked as potential participants in handovers at the PACU. Results showed that of these 27, 23 had taken the in-house training course, 21 had used the SBAR tool and 25 felt confident using the SBAR.
When the patient was transferred from the operating theatre to the PACU, the verbal report made by the sender was audio-recorded. After completion of the handover, the receiver was, within 10 min, asked to reproduce as much as he/she could remember without looking at the patient's records. Data on interruptions and breaks-in-task were also collected, as a covariate, through both audio-recording and noted in a study-specific protocol based on previous research. Interruptions are frequent during postoperative handovers;17 they interfere with memory27 and are therefore necessary to take into account, as a covariate. Similar to the study by Chisholm et al.,28 we defined an interruption as ‘any event that briefly required the attention of the subject but did not result in switching to a new task’ and break-in-task as an event that required attention for more than 10 s and subsequently resulted in changing tasks. In the present study, these two variables are presented as interruptions. All data were collected by the first author (M.R.), who is an intensive care nurse/nurse anaesthetist, not employed by the anaesthetic clinics but known by some personnel.
Data were analysed using descriptive statistics, frequencies and percentages, means and standard deviations (SD) when the data were approximately normally distributed, and medians and ranges when the data were not. The data were tested for normality using the Shapiro–Wilk test. Because each combinations of sender, receiver and patient only occurred once, they were regarded as independent observations and the data were analysed using an independent samples t-test or Mann–Whitney U-test as appropriate for the distribution of the data. However, because some receivers participated in several handovers and this could have resulted in unknown correlations among measurements, a linear generalised estimating equation (GEE) model was also used. We assumed that the correlation structure was compound symmetry. P values, two-tailed, below 0.05 were considered statistically significant. Sizes of estimated mean differences with 95% confidence intervals are also shown. All analyses were performed using IBM SPSS 20.0 (IBM Svenska AB, Stockholm, Sweden).
Data from the audio-recordings of the verbal reports made by the senders during handover and the reproduced reports made by the receivers were transcribed verbatim. Thereafter, information sequences were identified in the report and the reproduced report. For example, ‘has drainage’ and ‘bleeding 500 ml’ were each assessed as separate information sequences.
In order to measure the structure, the information sequences were categorised using the structure of SBAR. Then, the sequences reported in an irregular part of the handover were counted (e.g. reporting background data during the report of assessments). Thus, a good structure was achieved when few information sequences were transferred during such an irregular part of the handover.
To compare how much the receiver remembered after handover at T1 and T2, the reported and reproduced information sequences were identified and counted in each handover and the percentage of recalled information sequences was calculated. Two authors (M.R. and M.E.) also independently identified information sequences in 40 handovers to test the inter-rater reliability. An intraclass correlation coefficient (ICC) was calculated on percentage of remembered information sequences; the ICC was 0.96.
At T1, 73 observations (intervention group, n = 40; comparison group, n = 33) were included in the study and 72 personnel participated. At T2, 91 observations (intervention group, n = 44; comparison group, n = 47) and 57 personnel participated (Table 1). Of the included personnel, 15 participated both at T1 and T2 (intervention group, n = 7; comparison group, n = 8).
The observed handovers at T1 concerned patients who had undergone general surgery [n = 35 (48%)], gynaecological surgery [n = 9 (12%)] or orthopaedic surgery [n = 29 (40%)]. At T2, the observed handovers concerned patients who had undergone general surgery [n = 42 (46%)], gynaecological surgery [n = 4 (4%)] or orthopaedic surgery [n = 45 (50%)]. The median anaesthetic duration was 105 min (range 15 to 395 min). Because neither the type of surgery nor the anaesthetic duration showed any statistical differences between or within the intervention and comparison groups at T1 or T2, the data have been summarised for the combined groups.
Percentage of information sequences recalled
In the intervention group, the percentage of information sequences remembered by the receivers [presented as mean and (SD)] increased from 43.4% (13.9%) at T1 to 52.6% (14.6%) at T2 (P = 0.004). In the comparison group, the corresponding figures were 51.3% (17.8%) at T1 and 52.6% (14.5%) at T2 (P = 0.725). The results of the Mann–Whitney U-test (presented as median (range)) showed that the structure of the verbal reports improved significantly (P = 0.028) in the intervention group [T1: 3 (0 to 17); T2: 2 (0 to 12)] but not in the comparison group [T1: 3 (0 to 26); T2: 3 (0 to 10); P = 0.889]. In the intervention group, there was no significant change over time for interruptions [T1: 4 (0 to 66); T2: 7.5 (0 to 25); P = 0.115], but in the comparison group, the number of interruptions had increased significantly [P = 0.024, T1: 2 (0 to 64); T2: 4 (0 to 28)]. The duration for handovers ranged between 20 and 524 s and was not significantly different for the intervention group and the comparison group. We proceeded by using GEE to test whether the intervention and time had an effect on the information sequences remembered and whether this change over time differed between the two groups, controlling for variations in interruptions. In the first GEE model, we used recalled information sequences as the dependent variable and the independent variables were the two groups (intervention and comparison) and time (T1 and T2), as well as an interaction effect (time*group) with interruptions as a covariate. The results showed a nonsignificant interaction term, indicating that change over time was not significantly different in the two groups (P = 0.219) (Model 1, Table 2). We continued with a model to test the main effect only. The results were significant for the variable time (P = 0.041) and interruptions (P = 0.002) (Model 2, Table 2). We also ran a model that explicitly tested the differences between groups at T1 and between T1 and T2 (Model 3, Table 2). The results of this model showed that compared with the intervention group at T1, both the comparison group and intervention group had a much better performance at T2 (P = 0.020 and P = 0.025, respectively). Furthermore, the covariate interruptions were significant (B = -0.29; P = 0.004). From the results, we can also see that there was no significant difference between the intervention and comparison groups at T1 (P = 0.130). In Table 3, the sizes of estimated mean differences with 95% confidence intervals for all pairwise comparisons are shown. There is an improvement in recalled information between T1 and T2 in the intervention group after controlling for interruptions, but as no significant time*group interaction effect was found, we cannot claim that this improvement over time is different from the change over time in the comparison group.
The results showed that, compared with T1, the percentage of recalled information sequences among receivers in the intervention group increased significantly at T2 with no significant change in the comparison group. These results were also confirmed using a GEE model focusing on the intervention group and controlling for interruptions. However, when using a GEE model to investigate the time*group interaction term, there was no statistically significant increase in recalled information sequences by receivers in the intervention group compared with the increase in recalled information sequences by receivers in the comparison group at T2. The results of these models should not be regarded as contradictory but complement one another as the models test different hypothesis. In the model looking at the estimation of the interaction term, we could not show a significant difference between the increase in recalled information sequences between the two groups. However, the results of the model focusing on the intervention group only and controlling for interruptions are promising and may indicate that receivers remember more when the information is reported in a predictable structure known by both the sender and receiver. It has been suggested that formalizing the handover improves communication as well as memory,13 but, to date, few studies have tested approaches to formalising communication using a given structure in order to improve handovers. There are, however, some results indicating that structuring the handover process may improve the quality of information transfer.29 Surprisingly, we found only five studies that have measured retention of information after handover. These were conducted in a simulated setting, in an emergency department, in an ICU and in a PACU.20–24 The results, reported as a percentage of the correct retention of information, varied between 27 and 90% in these five studies. The data collection methods differed, though observation protocols were mostly used. None of the five studies used audio-recorded fully transcribed records of the information given by senders and reproduced by receivers. In our study, about 50% of the reported information was recalled.
In earlier studies of postoperative handovers, the most commonly used outcome measures have been information omission (information not transferred to the receiver), quality of teamwork, technical errors and handover duration.17 In the present study, we focused on the structure of information transfer and the receivers retention of that information. The information was given in a methodical manner and both the sender and receiver knew the structure in advance, a factor that is thought to improve the receiver's memory. Checklists that focus on the content of information that must be reported during handover29 may also serve as a memory aid. However, when checklist information is used, then omission is the focus, and omissions are supposed to decrease if the sender complies with the checklist. Both kinds of communication tools, those promoting a predictable structure and checklists focused on the exact content that must be reported, are of value in improving the handover process.
Our results also showed that interruptions had a deleterious effect on recalled information with an increase in the number of interruptions during handover resulting in a decrease in the amount of information recalled by receivers. It is well known that in the PACU, the handover procedure is marked by interruptions.17 Research within psychology has shown the disruptive effects of interruption on human cognition,30,31 and it has been recommended that interruptions should be minimised in workplaces with high demands on memory.27 There are different kinds of memory. Working memory is described as being responsible for operating and selecting strategies and is assumed to be associated with attention.32 Prospective memory is memory for things we should remember to do in the future without having to be reminded of them,33 and studies have shown that interruptions interfere with prospective memory.27 In the area of healthcare, two investigations have shown causality between interruptions and errors. The results from these two investigations showed that interruptions caused up to 11% of medical-dispensing errors34 and led to a 13% increase in clinical errors during administration of medication.35 As both the structure of the communication process and interruptions during the process were associated with recall of information after postoperative handover, one can assume that these factors may be associated with errors that impact on patient safety. A decrease in the proportion of incident reports due to communication errors has been reported after implementation of SBAR in an anaesthetic clinic.36 The personnel's experience of communication accuracy between different professional groups also improved, as did the safety climate.36 However, any possible links to patient outcome have to be examined.
The present study has several strengths: First, implementing SBAR significantly improved the structure of postoperative handover, showing that the SBAR was being used as intended.
This may have resulted from the rigour of the implementation, which involved the use of different methods to inform and train the personnel. All of these components have also been reported to increase the sustainability of interventions. Second, interruptions were taken into account when analysing information retention. Third, the data were audio-recorded and were fully transcribed. Fourth, to strengthen the results, inter-rater reliability was high and we took clustering effects into account using GEE analysis and corrected for differences in the covariate interruptions. The study also has limitations: Possible selection bias needs to be considered, as no randomisation was performed and a non-equivalent comparison group was used. The comparison group was chosen because it had not used a structural communication tool previously or been exposed to implementation of any communication tool during the study period. The total annual amount of surgery differed between the two hospitals, but not the proportion of planned surgery. Furthermore, at baseline, the intervention and the comparison group did not show any statistically significant differences concerning anaesthetic duration or type of surgery performed. There were also no statistically significant differences over time within groups, regarding these parameters. The selection of handovers was based on consecutive sampling of handovers following planned surgery, which limits generalisability. Thus, the results may be generalised to handovers from operating theatre to PACU in relation to planned surgery that takes place in small to medium-sized hospitals with clinics that perform different kinds of surgery, if the receiver's memory is measured within 10 min after the handover report. However, a multicentre study would have increased the generalisability of results. We audio-recorded the reproduced report within 10 min. It would have been interesting to also measure retention after a longer interval, but as the observations were conducted in an authentic milieu, repeatedly disturbing the personnel may have entailed a safety risk. Another possible limitation of the present study is that the participants may have felt more pressure or modified their handover behaviour because they knew they were being observed; the ‘Hawthorne effect’. However, this possible influence would have been present in both groups. Furthermore, the results showed no significant interaction effect between the groups, which might have been seen with a larger sample size. Still another limitation is that neither the data collector nor the outcome assessors were blind to which group the senders/receivers belonged to. The potential investigator bias would have been minimised if they had been blinded. However, although the examination was conducted in a planned and highly structured manner, there were differences between the two hospitals that had the potential to confound the results. In the comparison hospital, the personnel rated their teamwork climate, safety climate, job satisfaction, working condition and within-group communication accuracy on a higher level than the intervention hospital at baseline,36 which could indicate that there were differences in the working environment between the two groups.
As there are limitations to human cognition, we consider it essential to increase our knowledge about different factors within healthcare that may interfere with the transfer and recall of information. The present study is an initial study of the subject area, but further studies need to be conducted to confirm our results.
Compared with the comparison group with no intervention, we were unable to show a statistically significant improvement in recalled information among receivers following postoperative handover when SBAR was implemented in an anaesthetic clinic.
Acknowledgements relating to this article
Assistance with the study: we would like to thank the local inter-professional group from the Department of Anaesthesia for their collaboration in the present study. We also wish to thank all participants from the Department of Anaesthesia who participated in the study.
Financial support and sponsorship: this work was supported by the Faculty of Health and Occupational Studies, University of Gävle and by the County Council Gävleborg. It was also supported by the Patient Insurance LÖF and the Swedish Society of Nursing, but these organisations had no role in the design and running of the study.
Conflicts of interest: none.
1. Greenberg CC, Regenbogen SE, Studdert DM, et al. Patterns of communication breakdowns resulting in injury to surgical patients. J Am Coll Surg
2. Nagpal K, Vats A, Ahmed K, et al. An evaluation of information transfer through the continuum of surgical care: a feasibility study. Ann Surg
3. Kluger MT, Bullock MF. Recovery room incidents: a review of 419 reports from the Anaesthetic Incident Monitoring Study (AIMS). Anaesthesia
4. White AA, Pichert JW, Bledsoe SH, et al. Cause and effect analysis of closed claims in obstetrics and gynecology. Obstet Gynecol
6. Flemming D, Hubner U. How to improve change of shift handovers and collaborative grounding and what role does the electronic patient record system play? Results of a systematic literature review. Int J Med Inform
7. Leonard M, Graham S, Bonacum D. The human factor: the critical importance of effective teamwork and communication in providing safe care. Qual Saf Healthcare
2004; 13 (suppl 1):i85–i90.
8. Jeffcott SA, Evans SM, Cameron PA, et al. Improving measurement in clinical handover. Qual Saf Healthcare
9. Manser T. Teamwork and patient safety in dynamic domains of healthcare: a review of the literature. Acta Anaesthesiol Scand
10. Nagpal K, Abboudi M, Fischler L, et al. Evaluation of postoperative handover using a tool to assess information transfer and teamwork. Ann Surg
11. Siddiqui N, Arzola C, Iqbal M, et al. Deficits in information transfer between anaesthesiologist and postanaesthesia care unit staff: an analysis of patient handover. Eur J Anaesthesiol
12. Nagpal K, Arora S, Abboudi M, et al. Postoperative handover: problems, pitfalls, and prevention of error. Ann Surg
13. Nagpal K, Vats A, Ahmed K, et al. A systematic quantitative assessment of risks associated with poor communication in surgical care. Arch Surg
14. Rosenstein AH, O’Daniel M. Impact and implications of disruptive behavior in the perioperative arena. J Am Coll Surg
15. Smith AF, Mishra K. Interaction between anaesthetists, their patients, and the anaesthesia team. Br J Anaesth
16. Chen JG, Wright MC, Smith PB, et al. Adaptation of a postoperative handoff communication process for children with heart disease: a quantitative study. Am J Med Qual
17. Moller TP, Madsen MD, Fuhrmann L, Ostergaard D. Postoperative handover: characteristics and considerations on improvement: a systematic review. Eur J Anaesthesiol
18. Foster S, Manser T. The effects of patient handoff characteristics on subsequent care: a systematic review and areas for future research. Acad Med
19. Gordon M, Findley R. Educational interventions to improve handover in healthcare: a systematic review. Med Educ
20. Dowding D. Examining the effects that manipulating information given in the change of shift report has on nurses’ care planning ability. J Adv Nurs
21. Talbot R, Bleetman A. Retention of information by emergency department staff at ambulance handover: do standardised approaches work? Emerg Med J
22. Scott LA, Brice JH, Baker CC, Shen P. An analysis of paramedic verbal reports to physicians in the emergency department trauma room. Prehosp Emerg Care
23. Pickering BW, Hurley K, Marsh B. Identification of patient information corruption in the intensive care unit: using a scoring tool to direct quality improvements in handover. Crit Care Med
24. Arenas A, Tabaac BJ, Fastovets G, Patil V. Undivided attention improves postoperative anesthesia handover recall. Adv Med Educ Pract
26. Kazdin AE. Research design in clinical psychology. 4th ed.Boston, MA: Allyn and Bacon; 2003.
27. Li SY, Magrabi F, Coiera E. A systematic review of the psychological literature on interruption and its patient safety implications. J Am Med Inform Assoc
28. Chisholm CD, Collison EK, Nelson DR, Cordell WH. Emergency department workplace interruptions: are emergency physicians ‘interrupt-driven’ and ‘multitasking’? Acad Emerg Med
29. Segall N, Bonifacio AS, Schroeder RA, et al. Can we make postoperative patient handovers safer? A systematic review of the literature. Anesth Analg
30. Li SY, Blandford A, Cairns P, Young RM. The effect of interruptions on postcompletion and other procedural errors: an account based on the activation-based goal memory model. J Exp Psychol Appl
31. Altmann EM, Trafton JG. Timecourse of recovery from task interruption: data and a model. Psychon Bull Rev
32. Baddeley A. Baddeley AD, Eysenck MW, Anderson M. What is memory? Memory hove
. East Sussex; New York: Psychology Press; 2009. 1–17.
33. Eysenck M. Baddeley AD, Eysenck MW, Anderson M. Prospective memory. Memory hove
. East Sussex; New York: Psychology Press; 2009. 343–356.
34. Ashcroft DM, Quinlan P, Blenkinsopp A. Prospective study of the incidence, nature and causes of dispensing errors in community pharmacies. Pharmacoepidemiol Drug Saf
35. Westbrook JI, Woods A, Rob MI, et al. Association of interruptions with an increased risk and severity of medication administration errors. Arch Intern Med
© 2016 European Society of Anaesthesiology
36. Randmaa M, Martensson G, Leo Swenne C, Engstrom M. SBAR improves communication and safety climate and decreases incident reports due to communication errors in an anaesthetic clinic: a prospective intervention study. BMJ Open