Anesthesiologists have long prided themselves as being patient safety advocates. From the introduction of the pulse oximeter, adopting technologies to detect physiologic changes, and preventing incorrect equipment configurations, our specialty has a tradition of evolving to address challenges in safe care.1 This attention to safety is credited with a decrease in anesthesia-related mortality to approximately 8 per million hospital surgical discharges.2 Now, because anesthesia-related outcomes have become more expansively defined to include major postoperative complications and death days to weeks after surgery, anesthesiologists must also expand their definition of patient safety beyond the care rendered in the operating room (OR).
One particular care process undergoing more scrutiny is the handoff, the process of transferring patient care, and key medical information, from one provider to another. Anesthesia practice is unique, in that we have at least 3 kinds of handoffs across our many practice areas: shift-to-shift relief, duty breaks, and transitions to other sites of care.3 Investigators have shown that the OR to postanesthesia care unit4 and OR to intensive care unit handoffs5 (types of care transitions) are fraught with miscommunication. Until recently, however, there has been little rigorous study of intraoperative anesthesia handoffs (a type of shift-to-shift relief). In just the past year though, 3 groups have used administrative and clinical data to address the question of whether anesthesia handoffs are associated with harm. In the most recent of these studies, conducted by Hyder et al.6 of the Mayo Clinic and reported in this issue of Anesthesia & Analgesia, data from colorectal surgery patients were analyzed to tease out associations between intraoperative handoffs and patient adverse outcomes.
Although linking handoffs to outcomes is conceptually straightforward, detecting the impact of intraoperative handoffs on outcomes is methodologically challenging for at least 4 reasons. First, and perhaps most easily solved, is the need for very granular intraoperative staffing data. Hyder et al.6 and others7,8 accomplished this task by using robust institutional databases that contained information about the specific timing of staff changes. These databases also contained detailed clinical information that allowed these teams to control for a variety of patient- and procedure-level factors that influence outcomes. There are few multi-institutional databases with this level of granularity with the notable exception of the Multicenter Perioperative Outcomes Group,9 which has not yet been used to study intraoperative handoffs.
A second methodological concern with studying intraoperative handoffs is that handoffs are just one of many health care processes that affect patient outcomes after surgery. As such, it is difficult to detect the isolated effect of handoffs on a single outcome such as mortality. Although 1 of the 3 recent database studies examined death, all 3 studies have also used a composite outcome of adverse events to increase the ability to confidently associate handoffs with outcomes. Although this may seem like manipulation to achieve statistical significance, it seems reasonable, given the multiple mechanisms through which handoffs may contribute to harm, which include medication administration errors, lack of adherence to care plans, and delays in care.
A third problem with intraoperative handoff analyses using databases is the time problem. There is a necessary association between number of handoffs and case duration. That is, the longer an operative case, the more likely that there will be handoffs. Certainly, longer cases also tend to be more technically complicated or involve operative complications. A related consideration is operative start time. With most staffing schemes, cases that start later in the day are more likely to experience handoffs. The increased risk conferred by these transfers may be compounded by fatigue, waning attention, or limited availability of support staff. Increased case duration or late start time may themselves be associated with untoward outcomes, so complex statistical procedures are necessary to show any additional effect of handoff on outcomes. Two of the 3 recent intraoperative handoff database studies used propensity score matching to control for relevant confounders.7,8 This procedure models the likelihood of an event (in this case, a handoff), which is then used to match cases (patients who received a handoff) to controls (patients who did not receive a handoff).
In contradistinction to the propensity score matching approach, Hyder et al. used a different valid tactic: they limited the patient population to a specific surgery type to ensure homogeneity of the analytic sample. To quote the authors, “…any potential association between care transitions and complications in an observational study would be meaningful only in a specific context…”6 In their case, the context was elective colorectal surgery. Using the Mayo Clinic database, the team of Hyder et al. identified 927 patients with ASA physical status I to IV who underwent elective colorectal surgery between 2006 and 2010. The authors then used multivariable logistic regression to test the association between the number of attending anesthesiologists caring for a patient (primary exposure) and whether the patient experienced a composite outcome of death or major complication in the 30 days after surgery. The authors controlled for a number of confounding variables, including operative duration, case complexity, surgeon and assistant experience, and patient preoperative health states. A major consequence of the Hyder et al. approach is that the analytic sample was much smaller than if a more heterogeneous patient population had been studied. For example, Cleveland Clinic study by Saager et al.7 and University of Ottawa cardiac surgery study by Hudson et al.8 included 138,932 and 14,421 patients, respectively. However, restricting the study sample to a homogeneous population helps limit the effects of unmeasured confounders that may persist even after propensity score matching.
A fourth problem in handoff research is that handoff quality differs from provider to provider. This issue is not addressed by any of the intraoperative handoff studies because handoff quality is not captured as a variable in the study databases. Indeed, this is currently an impossible characteristic to capture because there are no accepted metrics of intraoperative handoff quality. There is a validated metric of handoff quality used in other settings,10 but this would need to be adapted and validated for use with intraoperative handoffs. Also, given the likelihood that handoff quality varies both within and between providers over time, this attribute may be best captured by direct observation or peer rating.
Interestingly, despite the challenges discussed earlier, all 3 recent intraoperative handoff studies showed that intraoperative handoffs are associated with adverse patient outcomes, including mortality and major morbidity (e.g., acute renal failure, cardiac arrest, and stroke). Although the effect sizes are variable (7%–44% increased odds of a composite outcome per handoff across 3 studies, 43% increased odds of death in the cardiac surgery cohort), the associations are clear and statistically significant, irrespective of the population or approach. In the study by Hyder et al., the association between handoffs and adverse outcomes persisted across adjustment strategies and sensitivity analyses. Although the number of attending anesthesiologists was the primary exposure, the authors also examined whether an increased number of in-room providers (including residents and nurse anesthetists) was associated with an adverse outcome; it was.
This growing body of literature suggests that there is something about intraoperative handoffs that may be placing our patients at risk for harm. Although database studies cannot answer the question of how to deal with the handoff problem, they are useful for generating and testing hypotheses about the mechanisms through which handoffs may cause harm. Perhaps handoffs are associated with delayed administration of antibiotics, prolonged vital sign derangements, or fluid overload. Maybe it is not the intraoperative handoff itself but the downstream consequences of inadequate handoff to postoperative providers that place patients at risk for adverse outcomes. Going forward, additional research is needed to elucidate these mechanisms, to understand which specific outcomes are most impacted by handoffs (rather than composite outcomes), and to develop strategies to mitigate the increased risk that handoffs may confer.
One natural reaction to these emerging data is to consider whether we ought to try to minimize or eliminate handoffs altogether. This is not feasible for both regulatory and human factors reasons: Duty hour restrictions for residents11 necessitate shift changes at intervals that may be out of sync with patient care, increasing the number of handoffs. In addition, evidence suggesting that physicians’ functioning is impaired when they are fatigued offers a compelling reason to offer duty relief.12 That being said, we should challenge ourselves to consider which handoffs are truly necessary. Additional research is needed to explore the logistics and consequences of changing schedules to decrease the number of intraoperative handoffs.
As we consider what to do about anesthesia handoffs, it bears mentioning that handoffs may not be all bad. Cooper et al.13 demonstrated more than 30 years ago that intraoperative handoffs may facilitate the detection of errors not noticed by one provider, serving as a valuable safety check. More recent studies affirming the positive aspects of anesthesia handoffs are lacking, but the work of Patterson and Wears14 suggests that handoffs serve important purposes other than information transmission. Handoffs build resiliency into health care by distributing the cognitive work associated with patient care, creating opportunities for social interaction, and creating cultural norms around appropriate patient care.14 For these reasons, handoffs conducted well may actually mitigate, rather than increase, preventable harm.
Given the impracticality of eliminating handoffs from anesthetic practice, we are now challenged with figuring out how to keep the good aspects of handoffs, dispense with the bad, and maintain the anesthesiologist’s position as a leader in patient safety. The past record of success in improving the safety of anesthesia care is not sufficient justification to avoid a re-examination of our practice. The challenges enumerated here can and must be surmounted so that we can continue to advance safe care in and outside the OR.
Name: Meghan B. Lane-Fall, MD, MSHP.
Contribution: This author wrote the manuscript.
Attestation: Meghan B. Lane-Fall approved the final manuscript.
This manuscript was handled by: Sorin J. Brull, MD, FCARCSI (Hon).
1. Gaba DM. Anaesthesiology as a model for patient safety in health care. BMJ. 2000;320:785–8
2. Li G, Warner M, Lang BH, Huang L, Sun LS. Epidemiology of anesthesia-related mortality in the United States, 1999-2005. Anesthesiology. 2009;110:759–65
3. Lane-Fall MB, Brooks AK, Wilkins SA, Davis JJ, Riesenberg LA. Addressing the mandate for hand-off education: a focused review and recommendations for anesthesia resident curriculum development and evaluation. Anesthesiology. 2014;120:218–29
4. Anwari JS. Quality of handover to the postanaesthesia care unit nurse. Anaesthesia. 2002;57:488–93
5. Petrovic MA, Martinez EA, Aboumatar H. Implementing a perioperative handoff tool to improve postprocedural patient transfers. Jt Comm J Qual Patient Saf. 2012;38:135–42
6. Hyder JA, Bohman JK, Kor DJ, Subramanian A, Bittner EA, Narr BJ, Cima RR, Montori VM. Anesthesia care transitions and risk of postoperative complications. Anesth Analg. 2016;122:134–44
7. Saager L, Hesler BD, You J, Turan A, Mascha EJ, Sessler DI, Kurz A. Intraoperative transitions of anesthesia care and postoperative adverse outcomes. Anesthesiology. 2014;121:695–706
8. Hudson CC, McDonald B, Hudson JK, Tran D, Boodhwani M. Impact of anesthetic handover on mortality and morbidity in cardiac surgery: a cohort study. J Cardiothorac Vasc Anesth. 2015;29:11–6
9. Kheterpal S. Clinical research using an information system: the multicenter perioperative outcomes group. Anesthesiol Clin. 2011;29:377–88
10. Horwitz LI, Dombroski J, Murphy TE, Farnan JM, Johnson JK, Arora VM. Validation of a handoff assessment tool: the Handoff CEX. J Clin Nurs. 2013;22:1477–86
11. Nasca TJ, Day SH, Amis ES JrACGME Duty Hour Task Force. . The new recommendations on duty hours from the ACGME Task Force. N Engl J Med. 2010;363:e3
12. Howard SK, Rosekind MR, Katz JD, Berry AJ. Fatigue in anesthesia: implications and strategies for patient and provider safety. Anesthesiology. 2002;97:1281–94
13. Cooper JB, Long CD, Newbower RS, Philip JH. Critical incidents associated with intraoperative exchanges of anesthesia personnel. Anesthesiology. 1982;56:456–61
14. Patterson ES, Wears RL. Patient handoffs: standardized and reliable measurement tools remain elusive. Jt Comm J Qual Patient Saf. 2010;36:52–61