An Analysis of Resident Generated On-Shift Evidence Based Medicine Questions
- Cervantes, Estelle;
- Shobba, Phillip;
- Kudrimoti, Shreyas;
- Albers, Jacob;
- Brown, Jeffrey;
- Kaul, Kashyap;
- Spinosi, William;
- Varadhan, Ajay;
- Zackary, Joseph;
- Kane, Bryan
- et al.
Learning Objective: Using previously validated methodology, to analyze the EBM content of clinical questions generated by EM residents while on shift.
Background: Evidence Based Medicine (EBM) skills allow EM physicians to obtain new information while on shift. There is little documentation EM resident clinical questioning skills.
Objective: Using previously validated methodology, to analyze the EBM content of clinical questions generated by EM residents while on shift.
Methods: With IRB approval, residents (PGY 1-4) were required to submit logs of on-shift EBM activity in the program’s procedure software system New InnovationsTM. The logs are a convenience sample, with an N of 3-5 per 28-day EM rotation. The logs include a patient description, clinical question, search strategy, information found, and subsequent application. The questions were analyzed through the lens of Patient-Intervention-Comparison-Outcome (PICO) using a rubric previously described by Ramos et al, BMJ, 2003. Anchor words/phrases were established for each of the PICO elements, with exemplars in Table 1. Data was analyzed descriptively.
Results: From 6/2013 until 5/2020, 10,450 discrete completed logs were identified for inclusion. A total of 49 were excluded (45 logs because they were intentionally left blank or only contained a punctuation mark and 4 were exact copies of the previous log) leaving 10,401 for analysis. These were submitted by 143 residents, of which 51 were female (35.7%). Table 2 demonstrates analysis of the questions via the PICO framework using the Fresno rubric. The average score each of the 4 PICO categories for all logs was 1.20. When excluding zero scores the average was 2.33.
Conclusions: In this single site cohort, resident description of P was most detailed, followed by I. C was the most excluded clinical question element. Having more patient oriented O would strengthen that category. When residents include a PICO category in clinical questions, their ability to do so appears strong, so educational interventions to encourage the use of all 4 PICO elements may yield the most improvement.