It is tempting to think of a patient care handoff as a simple information processing task. With this perspective, handoff improvement efforts naturally focus on those members of the healthcare team to whom tasks are delegated, often staff nurses, advanced practice providers, and, in academic settings, resident physicians. Indeed, many of the published studies on handoffs within and outside of the ICU focus on these team members (1–3). Handoffs between attending physicians—often considered to be the leader of the multidisciplinary ICU team (4)—are less studied and less well understood. The evidence we do have suggests that attending physician handoffs are multifunctional (4), highly variable (5), and sensitive to human factors-related concerns such as ordering effects (6).
In this issue of Critical Care Medicine, Dutra et al (7) shed light on the cognitive processes at play during attending handoffs. In this single-center observational study, daytime intensivists are present on consecutive weekdays for 12 hours, and they hand off each evening to another intensivist. Working with the assumption that it is important for the nighttime relieving clinician to understand the daytime clinician’s understanding of patient diagnoses and clinical goals, the authors set out to determine the extent to which daytime and nighttime clinicians developed a shared mental model (8) of each patient, measured by nighttime clinicians’ “correct” identification of diagnoses and clinical goals. The primary study outcome was the percentage of correct diagnoses, adjusted with the use of generalized estimating equations to account for clustering within the data. As might be expected, the authors did not find perfect concordance between the daytime and nighttime intensivists’ understanding of patient diagnoses and goals. The degree of the imperfection found, however, should give us pause.
In handoffs of 352 patients (distributed over 44 handoff sessions with 30 intensivists), nighttime clinicians were interviewed immediately after the handoff, before they cared for patients. Nighttime clinicians were only able to identify 53% of the daytime clinicians’ diagnoses (95% CI, 50–56%), even though each patient had a median of just two relevant diagnoses. Fascinatingly, nighttime intensivists identified 306 diagnoses that were not reported by the daytime intensivists at all. The degree of concordance varied as a function of the broad category of diagnosis; agreement was much higher for shock (about 90%) as compared to diminished level of consciousness (about 55%). The nighttime intensivists also did not identify 27 of the 31 patients flagged by the daytime intensivist as having unclear diagnoses. The nighttime intensivists were no better at identifying the daytime intensivists’ clinical goals, such as weaning vasopressors. Just 40% of the daytime clinicians’ patient care goals were correctly recalled (95% CI, 35–46%).
Although this work (7) is provocative, it has notable limitations. One key limitation is that data were collected by structured interview (i.e., self-report) rather than by direct or surreptitious observation. This approach may minimize the Hawthorne effect in which observation influences the behavior of the observed (9), but it introduces recall bias. The authors attempted to limit recall bias by conducting interviews immediately following the handoff session, but human memory is notoriously fallible. Audio recordings and/or chart reviews could have been used to more precisely document the diagnoses and clinical goals that were discussed during handoff sessions. A second important limitation is that the authors focused entirely on the clinical patient care aspects of handoffs, offering no insight into the other “big picture” issues that intensivists handle, including patient and family dynamics and ICU team functioning (4). Third, the work was conducted at a single center, limiting the transferability and generalizability of the results.
Despite the inherent limitations of single-center observational research, this work (7) does offer us an opportunity to ask difficult questions about intensivists’ patient care delivery. Clearly, we believe that our role in patient care is important, as evidenced by the large and growing body of work examining associations between intensivist staffing models and patient outcomes (10). It stands to reason that the reliability with which we accomplish patient care transitions is worthy of scrutiny. Yet, despite near universal agreement that intensivist handoffs are important, at least in a United States sample (5), I have found among my fellow intensivists a notable lack of will to engage with this topic. We seem to believe that regulatory imperatives to standardize handoffs (11) do not apply to us, perhaps because we are smart enough to “figure it out.” (4) Or maybe we work too much (12) and are too burned out to turn our attention to a nonpatient care activity that encroaches on our unpaid time (4 , 5).
At the same time, we recognize the positive and negative impacts that intensivist handoffs may have on patient care (5). Having witnessed the look of relief after reassuring a family member that I understood their loved one’s complex care, or arriving on Monday morning being ready to spar with a reluctant consultant service, I can attest to the importance of hearing the perspectives of a trusted colleague prior to starting an ICU rotation. I have also been “burned,” having to apologize to a family member for suggesting an already-tried-and-failed course of action or revealing a fundamental misunderstanding of a patient’s course, errors that could have been prevented with effective handoffs.
The findings from Dutra et al (7) support what most of us already know; intensivist handoffs are far from perfect. We are subject to all the cognitive biases that challenge our other colleagues, but we are perhaps less likely to be questioned than other members of the team. This special status elevates the importance of intensivist handoffs above what their paltry footprint in the published literature would suggest. Prospective mixed methods inquiries would help develop an understanding of how intensivist handoffs contribute to patient care and the patient and family experience. Empiric studies of interventions to improve the reliability of intensivist handoffs are also needed.
Where should we go from here? Handoff standardization is a proven strategy for other types of handoffs (3), but its role is unclear for intensivist handoffs (5). Standardization would decrease defects in information transmission but could also contribute to clinician dissatisfaction if it introduces work that is thought to be unnecessary. Also, given the variability in schedules, workflow, and electronic health record documentation, it is unlikely that a single approach to standardization would work across settings (13). A reasonable first step might be the creation of explicit expectations for intensivist handoffs in a given ICU or practice group. In a 2016 study, just 50% of surveyed intensivists reported knowing of expectations for their handoffs (5), indicating an opportunity for ICU leadership to develop local guidelines that might increase handoff reliability and consistency.
Other strategies to support intensivist handoffs include the use of robust electronic health record documentation, handoff training with audit and feedback, use of incentives and disincentives to promote desired behavior, scheduling overlapping shifts to allow time for handoffs, and relying on other members of the ICU team to maintain continuity and fill in gaps in understanding following handoffs. None of these approaches will work in isolation, as handoffs are a complex interaction involving not just information exchange but social interaction, sharing of the cognitive burdens of patient care, and reinforcement of organizational cultural norms (14). Irrespective of the idiosyncrasies of our practice environments, however, it is incumbent upon researchers and clinicians to develop robust strategies to safely transition the care of our most vulnerable patients.
1. Riesenberg LA, Leitzsch J, Cunningham JMNursing handoffs
: A systematic review of the literature. Am J Nurs 2010; 110:24–34; quiz 3536
2. Riesenberg LA, Leitzsch J, Massucci JL, et alResidents’ and attending physicians’ handoffs
: A systematic review of the literature. Acad Med 2009; 84:1775–1787
3. Starmer AJ, Spector ND, Srivastava R, et alI-PASS Study Group: Changes in medical errors after implementation of a handoff program. N Engl J Med 2014; 371:1803–1812
4. Lane-Fall MB, Speck RM, Ibrahim SA, et alAre attendings different? Intensivists explain their handoff ideals, perceptions, and practices. Ann Am Thorac Soc 2014; 11:360–366
5. Lane-Fall MB, Collard ML, Turnbull AE, et alICU attending handoff practices: Results from a national survey of academic intensivists. Crit Care Med 2016; 44:690–698
6. Ilan R, LeBaron CD, Christianson MK, et alHandover patterns: An observational study of critical care physicians. BMC Health Serv Res 2012; 12:11
7. Dutra M, Monteiro MV, Ribeiro KB, et alHandovers Among Staff Intensivists: A Study of Information Loss and Clinical Accuracy to Anticipate Events. Crit Care Med 2018; 46:1717–1721
8. Brindley PG, Reynolds SFImproving verbal communication in critical care medicine. J Crit Care 2011; 26:155–159
9. Yanes AF, McElroy LM, Abecassis ZA, et alObservation for assessment of clinician performance: A narrative review. BMJ Qual Saf 2016; 25:46–55
10. Sakr Y, Moreira CL, Rhodes A, et alExtended Prevalence of Infection in Intensive Care Study Investigators: The impact of hospital and ICU organizational factors on outcome in critically ill patients: Results from the extended prevalence of infection in intensive care study. Crit Care Med 2015; 43:519–526
12. Sevransky JE, Chai ZJ, Cotsonis GA, et alSurvey of annual staffing workloads for adult critical care physicians working in the United States. Ann Am Thorac Soc 2016; 13:751–753
14. Patterson ES, Wears RLPatient handoffs
: Standardized and reliable measurement tools remain elusive. Jt Comm J Qual Patient Saf 2010; 36:52–61