Evaluating the SPIKES Model for Improving Peer-to-Peer Feedback Among Internal Medicine Residents: a Randomized...

Feedback improves trainee clinical performance, but the optimal way to provide it remains unclear. Peer feedback offers unique advantages but comes with significant challenges including a lack of rigorously studied methods. The SPIKES framework is a communication tool adapted from the oncology and palliative care literature for teaching trainees how to lead difficult conversations.

To determine if a brief educational intervention focused on the SPIKES framework improves peer feedback between internal medicine trainees on inpatient medicine services as compared to usual practice.

Randomized, controlled trial at an academic medical center during academic year 2017-2018.

Seventy-five PGY1 and 49 PGY2 internal medicine trainees were enrolled. PGY2s were randomized 1:1 to the intervention or control group.

The intervention entailed a 30-min, case-based didactic on the SPIKES framework followed by a refresher email on SPIKES sent to PGY2s before each inpatient medicine rotation. PGY1s were blinded as to which PGY2s underwent the training.

The primary outcome was PGY1 evaluation of the extent of feedback provided by PGY2s. Secondary outcomes included PGY1 report of feedback quality and PGY2 self-report of feedback quantity and quality. Outcomes were obtained via anonymous online survey and reported using a Likert scale with a range of one to four.

PGY1s completed 207 surveys (51% response rate) and PGY2s completed 61 surveys (42% response rate). PGY1s reported a higher extent of feedback (2.5 vs 2.2; p = 0.02; Cohen's d = 0.31), more specific feedback (2.3 vs 2.0; p < 0.01; d = 0.33), and higher satisfaction with feedback (2.6 vs 2.2; p < 0.01; d = 0.47) from intervention PGY2s. There were no significant differences in PGY2 self-reported outcomes.

With modest implementation requirements and notable limitations, a brief educational intervention focused on SPIKES increased PGY1 perception of the extent, specificity, and satisfaction with feedback from PGY2s.