End of shift handoffs for hospitalized patients in the VA system remain a significant patient safety risk and opportunity for improvement. This is especially true in graduate medical education where recent changes in duty hours have dramatically increased the number of handoffs that occur daily. Few studies have evaluated interventions to improve handoffs and many that have been tried demonstrate only short term gains. As well, many interventions are poorly conceived and do not take full advantage of principles of instructional technology that might improve long-term success. One particular challenge has been that methods of study have often overlooked critical contextual dimensions of handoffs, such as cognitive strategies used by resident in preparing for and enacting handoffs and the local microsystem culture in which handoffs take place.
This multi-method study had three aims:
Aim 1: What are expert residents thinking as they approach, enact, and reflect on the handoff process?
Aim 2: How do residents actually behave during the handoff process?
Aim 3: How do residents learn effective handoff practices?
AIM 1: Cognitive Task Analysis (CTA)
CTA is a methodology designed to tap into experts' cognitive schemas or framework for making sense of situations and tasks. Originally developed in the 1950's it was known as the critical incident technique and was used in the military and elsewhere to understand the complexity of thought and tasks in such activities as flying a plane or managing a nuclear power plant. In medicine, CTA has been used to understand the relationship between the complexity of tasks and its effects on performance. We used CTA in this project to better understand how expert medical residents approach, enact, and reflect on end of shift handoffs. This was done through skillful interviews that highlighted where critical decisions were, or needed to be made. The results were then subjected to qualitative thematic analysis.
AIM 2: Direct Observation and videotaping of end of shift handoffs.
Complementing the focus on internal thought processes, we made and used audiovisual recordings of end of shift handoffs to provide an external, or what anthropologists term an 'etic" point of view to understand the stream and structure of behavior that constituted the end of shift handoff. Such data was useful in comparing and contrasting what residents said they thought about when approaching, enacting, and reflecting on handoffs with what they actually did Microinteractional Analysis (MA) of linguistic and non-verbal aspects of end of shift handoffs.
MA is an approach to understanding how individuals co-create meaning in and through face-to-face interaction. MA focuses on sequence of spoken turns in a conversation to analyze the underlying rules in use for "sense making". Non-verbal behavior is an important dimension of face-to-face interaction as it carries information about the speaker's intention as well as the content of what they are saying. Turn-by-turn sequential analysis was used to identify patterned behaviors during handoffs associated with such things as errors in hearing, understanding, and interpreting information being provided during the handoff.
AIM 3: How do residents learn effective handoff practices.
The goal of Aim 3 was to integrate the findings from Aims 1 & 2 into a handoff curriculum that is sensitive to the local culture in which handoffs take place and uses evidence based educational methods that will optimize for success.
Aim 1 Preliminary findings from CTA interviews demonstrated the importance for researchers, educators and policy makers, of understanding local culture and implicit conversational rules used to enact handoffs.
-One important finding revealed by the CTA interviews was that outgoing residents implicitly took into account the "reputation" of the incoming resident. A resident with a strong reputation required fewer details and instructions on what to do than "weak" residents who required a great deal of detailed instruction on patients' status and what to do in the event that it changed during the shift. From these data, we identified a phenomenon we called "recipient design" in which outgoing residents made informal appraisals of the personality style and preferences for information of the incoming resident.
-CTA analysis also revealed that residents were finely attuned to different shift types (morning or evening) and that they differed in style, content, and objectives. Residents preparing to hand off in the evening thought about what an incoming resident, who might know nothing about the patients s/he is covering, needed to know and anticipate over the next 8-12 hours. By contrast, handing off to the morning shift was more of an abbreviated status report in which overnight changes were reported to residents who knew and cared for the patients in question, plus any new patients who had been admitted overnight.
-As residents attempted to reduce coordination errors and safely transfer responsibility for patients to their colleagues, they used six important cognitive activities that occurred during preparation for handoff. These six activities have largely been unexplored in the burgeoning literature on improving patient safety culture. They are:
1. complete open tasks for each patient to reduce decisions required by the night shift,
2. ensure that documentation is up-to-date,
3. organize information so that it is clearly and succinctly communicated to the
4. specify tasks for the night shift,
5. anticipate questions and problems likely to arise during the night shift,
6. prioritize across the patient census.
We could find none of the qualities or behaviors described by residents in the CTA interviews in the current handoff literature.
The goal of Aim 2 was to compare the CTA interview material with video recordings of actual handoffs.
A total of 87 video observations (9.5 hours) were made; 16 in surgery and the remainder in internal medicine during the project period. In addition, 17 interviews with incoming or outgoing residents were audio recorded and transcribed for use in the analysis.
A coding scheme for looking at various dimensions of verbal and non-verbal communication and relationship during handoffs was also developed by the project team. It included categories of handoff organization and efficiency, quality of communication, and humanism/professionalism. Day to night, and night to day handoffs were compared across these dimensions. Overall night to day shifts scored lower on all dimensions than day to night shifts.
Sites were also compared along the following dimensions: % private setting, %use of supporting documents, attentional focus, % no eye contact and control of handoff pace. The best overall average rating was in Phoenix, with Chicago and Indianapolis having lower overall average ratings. Almost 70% of handoffs in Phoenix took place in private settings but only 8.3% and 5.8% took place in private settings in Chicago and Indianapolis, respectively. There was little variation in the use of supporting documents from 90% in Phoenix to 100% in Indianapolis. In terms of joint attentional focus (a desirable state for handoffs), Indianapolis and Chicago scored high while Phoenix scored relatively low. Finally, almost 30% of handoffs in Chicago occurred without mutual eye contact (also a desirable attribute for quality and accuracy in handoffs). In Phoenix all handoffs were accompanied by eye contact while in Indianapolis 11% were lacking eye contact.
Aim 3: A review of the continuing medical education literature concludes that adult learners learn best in small groups with opportunities to safely practice ongoing or newly acquired skills. It also suggests that direct evidence of performance (such as the use of videotaping) as a means of improvement is both salient and effective. To that end, the project team completed development of a curriculum for teaching and learning handoffs that is based on principles of small group learning, findings from the project and is organized conceptually around the mnemonic POISED (Prepare, Orient, Inform, Educate and Debrief). The curriculum has been pilot tested with a small group of medical residents and is awaiting permission from the internal medicine and surgery program directors at Indiana University to be fully implemented with a larger cohort of residents.
The study will impact the VA by highlighting the need to address the local culture and context in which end of shift handoffs take place. Handoffs are a complex socio-technical task that involved implicit rules of face-to-face interaction that need to be integrated into standardized approaches to information transfer that characterize the current literature. Educational interventions based on the findings from this study will positively impact resident behavior and ultimately improve quality and safety of patient care.
- Rattray NA, Flanagan ME, Militello LG, Barach P, Franks Z, Ebright P, Rehman SU, Gordon HS, Frankel RM. "Do You Know What I Know?": How Communication Norms and Recipient Design Shape the Content and Effectiveness of Patient Handoffs. Journal of general internal medicine. 2019 Feb 1; 34(2):264-271.
- Rattray NA, Ebright P, Flanagan ME, Militello LG, Barach P, Franks Z, Rehman SU, Gordon HS, Frankel RM. Content counts, but context makes the difference in developing expertise: a qualitative study of how residents learn end of shift handoffs. BMC medical education. 2018 Nov 3; 18(1):249.
- Militello LG, Rattray NA, Flanagan ME, Franks Z, Rehman S, Gordon HS, Barach P, Frankel RM. "Workin' on Our Night Moves": How Residents Prepare for Shift Handoffs. Joint Commission Journal on Quality and Patient Safety. 2018 Aug 1; 44(8):485-493.