Implementing and Maintaining an Increased Frequency of Nighttime Education Sessions

Our 271-bed, freestanding children’s hospital is in the Southeastern United States. The residency program includes 68 categorical pediatric residents, 6 pediatric neurology residents, and 24 medicine-pediatric residents. Every pediatric resident rotates on the general wards service and works 5 to 7 nights during their rotation. The night team includes 1 upper level resident and 2 interns. The upper level is expected to supervise each admission with the intern. An attending physician from the hospital medicine division is present from 7 pm to midnight and staffs admissions with the team. The number of admits per night ranges from 2 to 22.

A team composed of 3 hospital medicine physicians (2 in residency leadership positions) and a second-year pediatrics resident met throughout the project and used the model for improvement.⁹  We developed our aim and 7 key drivers, which included remembering to teach, data gathering tool, resident understanding of survey results and importance of learning opportunities, flexibility on when teaching can occur, perception of adequate time to teach, flexibility on who is present for teaching, and comfort with teaching skills and content (Fig 1).

Data were initially obtained by verbally asking upper level residents how frequently they taught after completing the rotation. Residents often had difficulty recalling this information, so a calendar was placed in the resident library for real-time data acquisition. Residents were asked to state whether teaching occurred and explain why if it did not. A REDCap survey was also sent to residents and interns after their ward rotation to collect number of teaching occurrences and content; however, many did not respond, so this survey was no longer sent.¹⁰ 

Our primary measure was the weekly number of reported occurrences of formal (15-minute case presentation, chalk talk, or discussion of board preparation questions) nighttime teaching led by upper level residents during wards night rotations.

Nighttime teaching occurrences were counted based on those marked on the calendar in the resident library. QI Charts software (Performance Improvement Products; Austin, TX) was used for run chart creation. Signal of change was identified by using 6 consecutive points above or below the median line.⁹ 

Our institutional review board approved our study.

As seen in the run chart, the initial introduction of the nighttime education sessions was associated with a signal of change in weekly teaching to a median of 1 but still less than our goal (Fig 2). This remained true despite several initial interventions that are annotated in the chart. These interventions included placing a calendar in the location teaching occurred and sharing resident survey results, in which residents overwhelmingly reported that education and conferences during night shifts was important to them. Additional interventions were tested using plan, do, study, act (PDSA) cycles (Table 1). Most notably, our PDSA cycles taught us that a rigid teaching day and time did not fit within the resident workflow or in night-to-night workload variation. Also, we learned that discomfort with teaching or perception of knowledge base prevented some teaching, so providing a binder of topics helped mitigate these fears.

These insights led us to encourage teaching to occur any night and at any time, provide a binder containing teaching resources, and reiterate that 15 minutes was an adequate time frame for teaching. These changes, along with the start of a new academic year, were associated with an increase in the median to our goal of 3 times weekly, which has been sustained for 32 weeks (Fig 2).

By using quality improvement (QI) methods and working collaboratively with residents, occurrence of formal nighttime teaching increased from a baseline of 0 to 3 times weekly. Simple introduction of the education sessions was not sufficient. Deliberate PDSA cycles incorporating resident feedback created a system that allowed us to reach our goal.

Our results indicate that several interventions had little impact on increasing the nighttime teaching frequency. These included reminding people of the educational importance of conference based on the survey results from our own program, as well as a teaching calendar that served as a reminder. However, the flexibility in the timing of teaching based on resident feedback, providing teaching resources, and the start of a new academic year were associated with achieving our goal. Although the start of a year with excited new interns, energetic second years, and a lower summer census could have impacted the increased teaching, this could not be the sole contributor because those factors did not have that same effect the year before. The main shift likely stemmed from incorporating feedback from those responsible for teaching and who understood the barriers. Providing flexibility and more ownership, allowing teaching to occur any time of night and any night of the week, proved to be successful. Encouraging a variety of teaching methods (ie, morning report style cases or clinical practice guideline review) and providing residents with additional materials in the binder (online resources and ready-made outlines) led to an increased teaching frequency. These modifications took stress off the upper level, allowing them to adjust for a high volume of admissions or unstable patients on the floor.

To track teaching, we discovered that a physical calendar in the workroom was more successful in real-time data acquisition. Verbally asking residents or sending out a survey after the rotation was not successful because many had a difficult time recalling their teaching sessions or did not complete the survey.

Although there is no signal of change, the last several weeks of data reveal multiple points below our median line. It remains unclear whether this is secondary to fatigue with reporting teaching sessions or an actual decrease in teaching. Our hypothesis is that this is due more to reporting fatigue, because we continue to witness nighttime teaching to this day. Additionally, this timing occurred over the holiday block, where the schedule is altered for Christmas and New Year’s, so less teaching could have occurred because of changing schedules. Because this was also December and January, it is possible less teaching occurred given a busier wards service during winter months.

Despite descriptions of night curricula in the literature that have demonstrated positive feedback from residents after implementation,^1,7  few have illustrated how to implement and sustain night curricula and educational opportunities outside of experiential learning. Although positive feedback is important, problem solving with residents to incorporate education sessions is also paramount if the purpose is to replace missed conference opportunities. By using QI methodology, one can continue to evaluate, intervene, and improve night education.

Several limitations exist regarding the results of our study. We relied on self-report to capture teaching occurrences, which could have lead to inaccuracies. If recording was not done in real time, teaching sessions could have been forgotten or incorrectly completed in the calendar, leading to a falsely lower or higher number of teaching occurrences, respectively. Additionally, the data collection was manual and ceased after 84 weeks. Finally, the quality of teaching is variable, and we have not evaluated whether the teaching is effective, leading to changed behavior in clinical practice or knowledge acquisition. Next steps include evaluating the effectiveness of the teaching and expanding the teaching sessions to include subspecialty topics.

As with many attempts to change behavior, simply informing residents that nighttime teaching is an “expectation” does not often achieve intended results. Providing flexibility and teaching resources based on residents’ feedback was vital to the increase in observed teaching. These specific interventions and the methods used to incorporate them can inform other programs’ desires to successfully weave teaching expectations into their night shifts, providing more learning opportunities even with Accreditation Council for Graduate Medical Education duty hour requirements.