Background: Bedside rounds (BR) are an ideal method for teaching in the hospital, with many proposed benefits. However, the effect of BR on educational outcomes is incompletely understood. Our objective was to determine whether BR, compared to other forms of hospital ward rounds, improves learning outcomes in medical education. To accomplish this aim, we conducted a systematic review of controlled studies measuring the effect of BR on learning outcomes.

Methods: Data sources included Ovid MEDLINE, Ovid Embase, Scopus, and Ovid Cochrane Central Registry of Clinical Trials from database inception through February 20, 2020. We included randomized and non-randomized experimental studies comparing BR to any other form of rounds in a hospital-based setting. Studies reporting a quantitative comparison of a learning outcome among physicians in training were included. Eligibility was assessed independently and in duplicate, with disagreements resolved by consensus. Learning outcomes were categorized into one of four domains: (1) learner satisfaction, (2) learner knowledge/skills, (3) learner behavior, and, (4) healthcare systems/processes (e.g., duration of rounds, interprofessional teamwork). Patient outcomes were not included, as they have been the subject of previous systematic reviews. Given the heterogeneity of measures and outcomes we did not pursue quantitative pooling. Instead, we summarized extraction elements using descriptive statistics, along with a narrative synthesis of design, implementation, and outcomes.

Results: Twenty studies met inclusion criteria, including seven randomized trials. General medical wards were the most frequent study setting (n=9, 45%), and all studies involved resident physicians, with the majority involving medical students (n=11, 55%). The design and implementation of BR varied widely, with most studies (n=13, 65%) involving co-interventions such as staff education, communication scripts, pre-rounds huddles, and real-time order entry.The fifteen studies reporting learner satisfaction yielded mixed results, with seven (46.7%) favoring BR, while four studies each favored the control or were equivocal. Of the five studies reporting an outcome of learners’ knowledge/skills, only one favored BR. The effect of BR on learner behavior was more consistently positive, with five of eight studies (62.5%) favoring the intervention. Likewise, eight of the fourteen studies (55%) reporting an outcome related to change in the healthcare system favored BR. Most studies (17, 85%) were deemed to be at high risk of bias, most often related to deviation from the intended intervention and selective reporting of results. The overall strength of evidence was low because of the high risk of bias and unexplained heterogeneity across studies.

Conclusions: In hospital-based settings, BR appear to have neutral-to-mixed effects on learners’ satisfaction, knowledge, and skills, but a positive effect on learner behavior and healthcare processes. Additional research is needed to clarify how BR’s design and implementation relate to its effect on learning outcomes.