If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Perioperative handoffs are a particularly high-risk period given patients' postprocedural physiology, their physical transport through the hospital, and the triad transfer of personnel, information, and technology. The authors piloted a new perioperative handoff process to guide patient transfers from the cardiac operating room (OR) to the cardiac surgical intensive care unit (CSICU). The aim of the study was to evaluate the impact of a standardized handoff process on patient care and provider satisfaction.
A prospective, unblinded intervention study.
A CSICU in a teaching hospital.
Two hundred thirty-eight health care practitioners during the transfer of care of 60 patients.
The implementation of a standardized handoff protocol and checklist.
Measurements and Main Results
After the protocol's implementation, the presence of all handoff core team members at the bedside increased from 0% at baseline to 68% after intervention. The percentage of missed information in the surgery report decreased from 26% to 16% (p = 0.03), but the percentage of missed information in the anesthesia report showed no significant change (19% to 17%, p > 0.05). Handoff satisfaction scores among intensive care unit (ICU) nurses increased from 61% to 81%. On average, the duration of handoff increased by 1 minute.
A standardized handoff protocol that guides the transfer of care from the OR team to the CSICU team can reduce the risk of missed information and improve satisfaction among perioperative providers.
Much of the handoff improvement work that ensued has focused on internal medicine and pediatric house staff sign-outs that occur at the end of each shift. Although one study does create a methodology for conducting safe and effective sign-outs on a surgical service, it is limited to floor patients.
Even less has been published regarding patient handoffs that occur in the perioperative setting. House staff shift work sign-outs are typically unidisciplinary (eg, internal medicine, pediatrics, surgery) and nonhierarchal (eg, resident to resident, not resident to attending), and involve patients already situated on a hospital unit. By contrast, perioperative handoffs are multidisciplinary, often consist of providers at different levels of training, and involve patients in transit who are by definition at high risk of clinical instability during this acute phase of care. Perioperative handoffs are thus, by nature, more multidimensional and complex events than within-unit sign-outs. Such handoffs not only involve the exchange of information but also include the transfer of technology, such as monitors, transducers, ventilators, and catheters to name but a few.
Perioperative handoff is described eloquently by Catchpole et al
in a 2007 study of pediatric patients who enter the pediatric intensive care unit (ICU) after congenital heart surgery. Additional perioperative research of adult patients has been confined to observational studies of postanesthesia care units, but no data exist for adult cardiac surgical care.
Because minimal data are available that evaluate best practices in perioperative handoffs in adult cardiac surgery and this is a group of high-risk patients with complex perioperative management, the authors conducted a study to systematically evaluate the handoff of patients from the operating room (OR) to the cardiac surgical ICU (CSICU). The hypothesis was that the implementation of a new OR-to-ICU protocol would improve provider satisfaction, increase information sharing, and decrease the number of technical defects. Technical defects were related to anything that was not information-sharing based.
This was a prospective, unmasked study that used a pre-post intervention design to evaluate a novel OR-to-ICU handoff protocol. The study was performed between December 2008 and June 2009 in a 15-bed adult CSICU of a large tertiary care center in which more than 1,200 procedures with postoperative ICU admissions are performed annually. All providers typically involved in the handoff of patients from the OR to CSICU were included. The following types of providers were represented: CSICU nurses, intensivists, midlevel practitioners (eg, nurse practitioners and physician assistants), surgeons, anesthesiologists, and house staff from the OR and the CSICU. After institutional review board approval, study participants were provided a letter explaining the purpose of the study, that their involvement was voluntary and anonymous, and that they could refuse to participate.
The protocol was developed through the input of frontline providers from multiple clinical areas through the use of focus groups and surveys. Frontline providers included “senders” and “receivers” of the report. Senders of reports typically include anesthesia and surgical providers, whereas receivers of reports include ICU practitioners (intensivists and midlevel practitioners) and ICU nurses. The handoff protocol includes clinical and technical components and directs elements, such as who should be at the bedside; team members responsible for the patient before, during, and after the handoff; the order in which information and technology are to be transferred; a distinct opportunity for questions and clarification of events; and a defined way to denote completion of the report. A visual aid was created that depicts the procedure of conducting handoffs and checklists to guide information sharing used during handoffs.
Over an 8-week preintervention period, a convenience sample of 30 handoffs were observed between 8:00 am and 7:00 pm. Observers were trained and calibrated by one of the investigators (MP). The observer was not a member of the clinical team nor did he/she serve in a supervisory role or participate in the handoff itself. Handoff participants then were given an anonymous posthandoff survey. The survey asked whether handoff team members experienced any technical problems during the handoff and, if encountered, the number and nature of these problems. Two hours after the handoff, observers collected data about events that occurred after the handoff.
After preintervention data collection, educational sessions introducing the OR-to-ICU handoff protocol were held. Education emphasized key points of the protocol, including defining the handoff team members and mandating their presence at the patient bedside throughout the handoff and the format and contents of the handoff process. The format was chronologic nonsimultaneous transfer of equipment (monitors/catheters) followed by the transfer of information (verbal report), which was to include pertinent clinical information. This was to be followed by a distinct question-and-answer period and the formal verbal conclusion of the handoff by the receiving team, marking the earliest time at which handoff team members could leave the patient bedside. This educational process required 2.5 months and was limited by scheduling logistics.
After the completion of educational sessions, a letter signed by administrative and clinical leaders of the CSICU nursing, ICU, anesthesia, and surgery departments announcing the start date for implementation of the novel OR-to-ICU handoff protocol was distributed to all frontline practitioners. The CSICU hosted a kickoff event to introduce the new protocol. In addition, large laminated handoff protocol schematics were posted over each bed, and handoff champions distributed laminated checklist tool pocket cards. Handoff champions were the investigators and those staff members from the CSICU who were part of a working group interested in the handoff process. One of the investigators was available for each handoff to guide the use of the new protocol for the first week. Orientation handoffs were guided in the following manner: as the patient and OR team entered the ICU room, the investigator brought a stand to the foot of the patient bed on which laminated posters of the novel OR-to-ICU handoff protocol and checklist tools were displayed and served as a visual guide. This guided orientation process continued for 4 weeks to capture and orient providers from all shifts. After the 4-week period of guided intervention, a convenience sample of 30 handoffs was observed using the same methodology as in the preintervention period. Approximately 1 month after the completion of postintervention data collection, study participants were provided with a satisfaction survey to assess opinions about use of the novel OR-to-ICU handoff process and tools.
Data elements collected included a description of the handoff procedure, environment, duration, type of providers at bedside, and the information shared during the handoff. Observers counted the number of parallel conversations when more than one person could be heard simultaneously. Handoff duration was defined as the time between patient arrival in the ICU and the departure of all OR team members from the bedside.
To quantify information sharing, specific data were assessed. Elements specified for anesthesia providers included past medical and surgical histories, allergies, baseline vital signs, weight/height, baseline laboratory values, intraoperative procedures (eg, epidural/spinal/block), invasive monitoring (eg, arterial catheter/central venous pressure), venous access (central/peripheral IV size and location), fluid totals, paralytic status, narcotic totals, and antibiotics. Surgical providers reported information regarding the chief complaint, the surgery performed, surgical findings, surgical complications, drains/tubes, postsurgical diagnosis, special instructions, and guidance on when to call for help or with questions. An information-sharing score (ISS) for each discipline (anesthesia and surgery) was calculated using the following equation: (number of categories verbally reported by discipline provider/total number of categories for the respective discipline) × 100. The ISS for the handoff as a whole was calculated by the following equation: (total number of categories verbally reported by both anesthesia and surgery/18) × 100, where 18 is the total number of categories.
Provider satisfaction was assessed using an anonymous 9-item survey based on a 5-point Likert scale (ie, strongly disagree, disagree, neutral, agree, and strongly agree). Satisfaction survey responses from each of the 4 disciplines (ie, OR surgery practitioners, OR anesthesia practitioners, ICU practitioners, and ICU nurses) were analyzed, and a Cronbach alpha coefficient was calculated. The same satisfaction survey was administered to handoff senders and receivers. Of the 9 questions used, 4 were applicable to handoff senders (Q1, Q4, Q7, and Q8), but all were applicable to receivers (Table 1). To measure the frequency of technical problems, a question on whether any technical defects during handoff were encountered and the number and nature of these problems was included.
Table 1Number of Handoff Satisfaction Survey Respondents by Discipline
Data were analyzed using Minitab 15 Statistical Software (Minitab, Inc, State College, PA). To test for statistical significance, the authors used the 2-sample t test and Mann-Whitney U test, as applicable, for continuous variables and the Fisher exact test for categoric variables.
The study included direct observations of 30 consecutive OR-to-CSICU handoffs in both pre- and postintervention periods. The introduction of the protocol minimally increased the handoff duration from an average of 11 minutes to an average of 12 minutes. This increase was not statistically significant (2-sample t test: p = 0.395). The presence of all team members simultaneously at the bedside increased from 0% preintervention to 68% postintervention (Fisher exact test: p < 0.001). The number of parallel conversations decreased from an average of 11.3 per handoff preintervention (median = 11.5, standard deviation [SD] = 6.95) to 3.5 per handoff postintervention (median = 3.0, SD = 3.18). This reduction was statistically significant (Mann-Whitney U test: p < 0.001). On a 5-point Likert scale, the percentage of nurses who responded that they could hear all of the report increased significantly from 45% before intervention to 78% afterward (Fisher exact test: p = 0.023).
As depicted in Figure 1, the sharing of information during verbal report increased with intervention implementation. The overall handoff ISS significantly increased from 78% to 84% after intervention (Mann-Whitney U test: p = 0.01). The surgery ISS also increased significantly from 74% to 83% (Mann-Whitney U test: p = 0.02). Although the anesthesia ISS increased from 82% to 86%, this increase was not statistically significant (Mann-Whitney U test: p = 0.09). No surgery report was given during one preintervention handoff and 3 postintervention handoffs. Therefore, those handoffs were excluded from ISS analyses because no information was shared during a distinct handoff.
There was a statistically significant reduction in the average number of questions from the ICU team at the end of the handoff from 0.23 questions per handoff in the preintervention period (SD = 0.50) to 0.00 (SD = 0.00) in the postintervention period (Mann-Whitney U test: p = 0.0397).
One hundred seventy satisfaction surveys in the preintervention period and 138 in the postintervention period were completed (response rate of 73% and 55%, respectively). The number of responses ranged from 3 to 8 surveys per handoff. The distribution of survey respondents by discipline is summarized in Table 1 and their pre- and postintervention responses in Table 2. The Cronbach α coefficient was calculated based on the survey responses from each of the 4 disciplines. The Cronbach α coefficient was 0.95 for ICU nurses, 0.93 for ICU practitioners, 0.81 for OR surgery practitioners, and 0.90 for OR anesthesia practitioners.
Table 2Satisfaction Ratings Pre- and Postintervention by Discipline and Survey Item
General satisfaction with the postoperative handoffs varied by discipline. During the preintervention period, the percentage of handoff participants who agreed or strongly agreed with the statement “I was satisfied with the OR to ICU handoff” ranged from a low of 62% among ICU nurses (total number of respondents [n] = 21) to 85% among OR anesthesiologists (n = 65), 92% among ICU practitioners (n = 53), and 98% among OR surgeons (n = 29) (Fisher exact test comparing ICU nurses with OR surgeons: p < 0.001). Table 2 depicts the percentage of handoff team members who strongly agreed with each of the 9 items on the handoff satisfaction survey. In the preintervention period, ICU nurses gave lower satisfaction ratings on all survey items compared with the ICU practitioners (Fisher exact test; p value for survey items 1-9 were p < 0.001, p = 0.016, p = 0.003, p < 0.001, p = 0.043, p = 0.02, p = 0.001, p = 0.002, and p < 0.001, respectively).
Compared with the preintervention period, the percentage of respondents answering “strongly agree” among ICU nurses and practitioners in the postintervention period increased for all survey items. This, however, did not reach statistical significance except for items 4 and 9 (“I could hear all of the report” and “I received anticipatory guidance”) among ICU nurses and for item 6 (“I received information on follow-up”) among ICU practitioners (Table 2).
The percentage of respondents answering “strongly agree” among handoff senders increased significantly in the postintervention period compared with preintervention period for all 4 applicable survey items: “I was satisfied with the OR-to-ICU handoff” (Q1, p = 0.001), “I could hear all of the report” (Q4, p = 0.008), “Physical transfers went smoothly” (Q7, p = 0.001), and “Handoff start and end clear” (Q8, p = 0.021) (Table 2).
Technical defects were reported primarily by the CSICU nurses. Some examples of technical defects included a forgotten chart, a patient without an identification band, an arterial catheter that was working in the OR but was not working at the time of transfer and the nurse needed to take time to troubleshoot, and the delayed availability of respiratory therapy. Based on the nurses' survey responses, the average number of technical problems reported per handoff was reduced from 0.8 in the preintervention phase (SD = 1.22, median = 0.5) to 0.62 in the postintervention period (SD = 0.97, median = 0). This reduction did not reach statistical significance (Mann-Whitney U test: p = 0.56). At 2 hours after handoff, the authors collected data from the ICU team on whether any calls or pages had been made to the OR team or if patient care adverse events had occurred. The average number of pages and calls to the OR team posthandoff increased from 0.14 per handoff in the pre- (SD = 0.44) to 0.27 per handoff in the postintervention period (SD = 0.64) but that increase was not statistically significant (Mann-Whitney U test: p = 0.325).
Through a comprehensive approach that incorporated literature review and local input, an innovative handoff protocol for the perioperative period was developed and implemented. Although there is broad literature on handoffs in the hospital setting
this is the first that looks at technical errors and information shared in the perioperative adult cardiac surgical population. With protocol implementation, the handoff environment changed from one that was noisy with multiple parallel conversations to an orderly exchange of information. Protocol implementation was associated with improved information sharing during the handoff with no significant increase in duration. Although the mechanisms that were most responsible for this improved transfer cannot be extrapolated from the current data, it likely represents a decrease in distractions and subsequent improvement in information transfer. In fact, the intervention was meant to decrease local distractions by its design of guiding the team to focus first on the transfer of equipment and monitoring followed by the verbal report, which the authors believe was a significant contributing factor in the reduction of parallel conversations. The protocol and leadership support clearly improved the multidisciplinary nature of the handoff with the consistent presence of all the team members. It is unclear why the protocol did not result in fewer follow-up calls in the 2 hours after the handoff, and this study was not designed to answer this adequately. One consideration is that the bar had been raised, and the receivers (ICU nurses and practitioners) of the patient were empowered to call back to the OR for clarification of any issues that had arisen in the first 2 hours. Furthermore, data on whether the callbacks were specifically for information missed in the handoff or for changes in patient status were not captured. In addition, markers for patient complexity were not collected and could not be accounted for either.
A key element that is unique to this protocol is the incorporation of and emphasis on the explicit transmission of anticipatory guidance, with the inclusion of the patient's greatest safety risk as an element. This element was used as a tool for providers to share concerns about the anticipated clinical course based on their recent experience with the patient. It serves to coordinate multiple team members at the time of transfer so that if a deviation or any divergence from the expected clinical course occurs, the receiving practitioners can more rapidly identify a problem and communicate it to other team members. Although this study was not designed to show the direct impact of this component, the authors maintain that it is critical to improving patient safety. This concept has been used in handoff templates in medicine
The authors were unable to show a significant decrease in the average number of reported technical errors. Many technical problems may not be amenable to resolution with this study's intervention. For example, although absence of an orogastric tube was counted as a technical problem, it was not differentiated whether the reporting team failed to notify the receiving team of their difficulty placing it (addressable by the study's intervention) or whether they forgot to place it (not addressable with the study's intervention). Another possible explanation is that with the improved environment during the transfer, providers may have been more attuned to the technical issues. Lastly, given the low number of technical defects at baseline, a larger sample size would be needed to show a statistically significant reduction.
Factors outside of the actual handoff period can impact protocol implementation success. During 4 of the 60 handoffs, a surgery provider was unavailable. One of the major reasons for the surgeon's absence was a prolonged OR wait time (>1-hour delay) for an available CSICU bed; during the wait, the designated surgeons were already involved in the next surgery in a different OR and were no longer available. It is conceivable that such occurrences may be inevitable, and future studies should assess the impact of the absence of the surgeon and investigate processes that ensure high-quality handoffs under those circumstances. Satisfaction with OR-to-ICU handoffs was significantly lower among the ICU nursing staff than among the ICU practitioners and OR surgeons. Postintervention surveys revealed a positive trend of increased satisfaction ratings for all survey items; however, this increase did not reach statistical significance except for items number 4 and 9. This could be because of the small sample size of surveyed nurses. Similarly, satisfaction ratings of the ICU practitioners did not significantly improve except for one item. However, satisfaction ratings of the ICU practitioners were relatively high at baseline, and this rating scale may not have been sensitive enough to capture differences in practitioner satisfaction because of a ceiling effect. Furthermore, the study was not designed to be able to compare pre- and post-satisfaction differences at the individual provider level, and likely different individual providers were surveyed in the pre- and postintervention periods. The response rates for the satisfaction survey were slightly lower in the post- compared with the preintervention period. This could be because the handoff lasted longer after intervention, and it is conceivable that some providers did not want to take additional time to answer the survey.
Although some of the findings were statistically significant and the authors were approached by many of the respective providers after the implementation phase stating that the handoff process was greatly improved, no conclusion can be made that these findings were clinically meaningful because patient outcomes were not tracked. This study had several relevant limitations, the first two of which are because of its nature as a pilot study. First, the intent was to perform a short-term evaluation on the feasibility of implementing a perioperative handoff protocol and assessing its impact on information transfer and satisfaction. As such, the sample size of this study was small. Second, although the authors were able to show the improved transfer of care achieved by organizing and standardizing the handoff process, they did not evaluate sustainability of this model or the impact on patient outcomes. As interventions are designed to decrease patient risks, it is critical that they be sustainable beyond the observation period. Additional, studies are needed to assess sustainability of this intervention.
Third, the use of observers was necessary to assess the impact of the intervention but that approach has the potential of introducing a Hawthorne effect. The authors believe that this effect was limited, however, for the following reasons: Observers did not interact with the primary team during the actual handoff, observers were present during both preintervention and postintervention periods of the study, and observers were not members of the clinical team and had no authority over the clinical team. Fourth, the assessment tool used to evaluate satisfaction was not formally validated. Fifth, a quasi-experimental design was used, and, therefore, the authors cannot account for potential unknown confounders. Because this was piloted in a single clinical area, it was not feasible to randomize patients or providers to different handoff methodologies. Given that no other initiatives focused on handoffs within the perioperative setting during this time period, the authors do not feel that temporal trends were notable confounders. Sixth, because anonymous data were captured, the impact of the intervention on individual providers could not be assessed, and baseline performance of providers varied. Finally, this study was performed in a CSICU, and the findings may not be generalizable to other ICUs or to all perioperative settings. However, the protocol was designed with input from local providers who participate in OR–to–general ICU and OR-to-PACU handoffs. Future studies are needed to evaluate the impact of this intervention in these various settings and should include a measure of sustainability. Findings from research on handoffs and transitions of care that link to outcomes also will be important.
Although improving handoffs is an important national patient safety goal, many questions remain unanswered regarding how such improvements can be accomplished and measured. For example, if the inclusion of a handoff tool is to be a component of a national measure of reportable quality, the research discussed here has not identified what the key elements of such a metric should be or the format of this tool. The authors have presented their handoff tool in this article, but there are other approaches, such as SBAR (situation, background, assessment, and recommendation), that have been assessed in other patient populations. Questions remain as to whether a checklist is what should be measured or whether the quality of the information shared should be evaluated. Like many aspects of patient safety and quality improvement, these are very difficult questions to answer. There is much to be learned about reporting on the processes of care and demonstrating that protocols and tools actually improve patient safety.
The authors thank Ms Claire Levin for editing the manuscript.
Dropping the baton: A qualitative analysis of failures during the transition from emergency department to inpatient care.
Supported by the Department of Anesthesiology and Critical Care Medicine, the Johns Hopkins University School of Medicine (M.A.P.) and the Center for Innovation in Quality Patient Care, The Johns Hopkins Hospital (H.A.). Elizabeth A. Martinez (E.A.M.) was supported by the Agency for Healthcare Research and Quality K08 grant number HS013904-02.
E.A.M. is currently faculty in the Department of Anesthesia, Critical Care and Pain Medicine, Massachusetts General Hospital, Harvard Medical School, Boston, MA.