When Duty Calls, Resident Physicians Answer: Learning to Take Transfer Calls Free

A pre- and posteducational intervention trial was performed at a large children’s hospital in March 2021. The authors used guided participation methodology¹⁰  to create a workshop addressing skills in taking intra- and interfacility transfer calls (see Supplemental Fig 3). The authors presented a synchronous 1-hour session highlighting the importance of phone communication, outlining an evidence-based handoff tool, and reviewing illustrative cases to 44 resident physicians during scheduled didactics. Participants received a reference card of the SHARING tool to access during the workshop and beyond.

Direct observation is effective in rating behavioral change at Kirkpatrick Level 3¹¹  and was chosen to measure the workshop’s efficacy. Six simulated calls representing transfer requests for common diagnoses were written by pediatric hospital medicine faculty through iterative reviews.

A checklist for scoring phone call behaviors was created from the SHARING tool as it is a well-developed tool created by nominal group technique methods and a person-based approach.⁹  The checklist was constructed to evaluate the residents’ ability to obtain necessary patient details for a thorough handoff, as decided by the multidisciplinary team who developed the SHARING handoff tool. From the handoff tool, 24 unique patient details were converted to dichotomously scored items on a checklist used for directed observation during simulation calls. As the relative value of each item on the SHARING handoff tool has not been determined, the authors gave each checklist item equal weight for a sum of 24 possible points. To decrease interrater variability, the grading checklist and its use were demonstrated to the proctors (pediatric hospital medicine faculty and fellows) in a 30-minute session before the workshop. The proctors were taught about the SHARING tool, graded a simulated case, and asked clarifying questions.

Participants were divided into groups of 3 residents of different training levels. The residents served as their own control and completed a preintervention and a postintervention simulated phone call directly before and after the workshop. We created a blocking design with 20 prespecified orders of the 6 unique simulation cases and assigned 1 to each resident, repeating the block design until every resident received a randomized order of calls. Residents were graded by the same proctor, who was not blinded to the resident, pre- and postintervention. Data were anonymized, scored, and compared by a separate author. In addition to behavioral change in a simulated environment, we measured resident-reported confidence in taking calls, resident-reported knowledge of the call process, and perceived importance of the skill utilizing an internally developed retrospective pre- and postsurvey¹²  completed by 95% of participants. The survey consisted of 6 questions regarding demographics and previous experience (defined as listened to calls, participated in calls with mentor present, or answered calls alone) and rating confidence in taking calls, knowledge of the call process, and importance of skill on a 5-point scale before and after the workshop.

To test for an increase in score comparing preintervention to postintervention, a paired t test was used. Simulation scores were assessed for normality by using a Shapiro-Wilk test and were normally distributed. Comparisons of pre- and postsurvey results, which were not normally distributed, were tested by using Wilcoxon rank tests. To test for a difference in pre/postsimulation score change between postgraduate year (PGY) levels, a Kruskal-Wallis test was used. An α level of P <.05 was used to assess statistical significance for each test.

Before implementation, the hospital institutional review board designated this project exempt from further review.

Of the resident participants, 95% (42/44) completed the pre- and postassessment. Table 1 contains participant demographics. Behaviors in a simulated environment, as measured by an average score on the grading checklist, improved from 9.54 (standard deviation: 3.15) of 24 possible points (37%) to 16.07 (standard deviation: 3.44) of 24 points (71%) after the workshop, with a mean score increase of 6.52 points (P <.0001; 95% confidence interval: 5.53–7.51). Net behavioral change scores are presented in Fig 1. Residency training level did not affect the change in score outcome (P = .3763).

Resident-reported confidence, knowledge of the transfer call process, and importance of transfer call skills significantly increased (P <.0001), with a median score increase of 1.5 (interquartile range [IQR]: 1–2), 1.0 (IQR: 1–2), and 0.5 (IQR: 0–1), accordingly. There was insufficient evidence to conclude that there was an effect of PGY level on confidence (P = .2750) or importance of skill (P = .1734), but PGY1 did have a higher increase in self-reported knowledge rating compared with other PGYs (P = .0053).

There was insufficient evidence to support that previously listening to calls affected the change in scores for confidence, knowledge of the call process, or perceived importance of transfer call skills. Those who had previously participated in calls with mentors had significantly higher increases in confidence scores than those who had not (P = .0017). Conversely, those who had not previously participated in calls with mentors had a significantly higher increase in knowledge score than those who had (P = .0004). Those who had not previously answered calls alone had significantly higher increases in confidence scores than those who previously answered alone (P = .0152). Figure 2 contains survey data based on previous participation.

Given the known risk associated with transitions of care,¹  ensuring residents can obtain necessary information during transfer phone calls to safely transition care is an important educational objective. Because there are lacking curricula teaching resident physicians how to take transfer calls⁸  and standardized handoff tools are effective at improving patient outcomes during transitions of care,⁵  a handoff tool was chosen as the basis for our educational intervention. A workshop that introduces the SHARING handoff tool using guided participation revealed improvement in behavioral scores in a simulated environment, regardless of PGY level. We applied the SHARING handoff tool, previously shown to be effective for interhospital transfers,⁹  to both inter- and intrahopsital simulated transfer calls and found it to be an effective educational and assessment tool. After this educational intervention, residents recognized the applicability of this skill to their futures. PGY1 had a higher increase in knowledge rating compared with other PGY levels, supporting PGY1’s need to learn nonclinical skills.^13,14 

Our findings are likely to be true due to variability in the residents’ experience before educational intervention, randomized order of cases to control for the level of difficulty, and balanced distribution of resident classes within small groups. Limitations to consider include the influence of hearing other residents’ calls; in some groups, more experienced residents went before less experienced residents, which may have inflated scores. Additionally, by having a pre- and posteducation call, residents may have felt more comfortable during their second call, regardless of the workshop. The residents may have also performed better because of Hawthorne effect, which is difficult to control.¹⁵  We attempted to control for interrater variability by utilizing the same proctor pre- and postintervention and comparison of the net change in scores rather than actual scores. Unfortunately, this did not allow proctors to be blinded to residents. Because performance was measured in a structured, simulated environment by familiar proctors, scores may be higher than if measured in real-time communication with unfamiliar physicians. Although we had good participation, results not supported by sufficient evidence may have been due to the study being underpowered. Finally, the survey was unvalidated, and the subjective ratings of pre- and postintervention were analyzed by looking at a change in score, which does not account for a perfect preintervention score that cannot increase postintervention.

Involving residents in transfer calls may encourage resident autonomy, thereby preparing them for independent practice and encouraging greater medical decision-making skills.¹⁶  This teaches systems-based practices¹⁷  and the use of handoffs, which are trackable The Accreditation Council for Graduate Medical Education milestones in residents’ training.³  The long-term goal is for residents to continue improvement in competence and confidence over time using the SHARING tool. Because this was a single-site, single-iteration pilot workshop, next steps may include the use of the grading rubric for standard evaluation of real-time phone calls to show maintenance of learned skills and plans to provide this workshop annually at the end of PGY1. This workshop is generalizable to any residency program looking to encourage resident participation in transfer calls, with only a slight modification of cases.

This guided participation workshop improved resident physician behaviors in a simulated environment, confidence, and knowledge of the inter- and intrafacility transfer call process, thereby demonstrating the utility of providing a standardized tool and education to improve transitions of care.