Article review: Internal Medicine team redesign

The Institute of Medicine has re-sparked discussions about limiting and further reducing resident duty hours in the United States (IOM's Duty Hours Report from Dec 2008). In response to this, the Brigham and Women's Internal Medicine residency program created an innovative inpatient team model, which was published in this month's New England Journal of Medicine.


Study Design

Setting:

• Faulkner Hospital, an affiliated community hospital with 72 inpatient medicine beds

Control team = General Medical Service (GMS) team

• A single attending drawn from a "faculty pool"
• 1 resident, 2 interns
• A different attending drawn from a "faculty pool" to teach 3 times weekly
• Intern call was every 4th night until 10 pm
  
• Resident stays until 7 pm and then night float resident
  

Experimental team = Integrated Teaching Unit (ITU)

• 2 attendings with each covering half of patients on team
• One attending is hospitalist and the other is an internist or specialist
• Attendings were selected based on excellent teaching evaluations.
• Addition of multidisciplinary team members
  
• 2 residents, 3 interns
  
• Both attendings participated in daily morning rounds for 2 hours
• One attending was available all day for teaching and clinical care
  
• Intern call was every 6th night and leaving by noon next day
• Resident supervised all interns every 4th night until 10 pm, and then night float resident
• Team census cap = 15 patients

Methodology
Patients were randomized onto one of 4 teams over a 12-month period. There were 2 GMS and 2 ITU teams each month. An independent observer recorded intern activity on these 4 teams.

Results
Results for the GMS team are shown in blue and ITU team in red.

• Average patient census per intern = 6.6 patients vs 3.5 patients
• Overall trainee satisfaction (83% response rate) = 55% vs 78%
  
• Intern time spent with learning activities = 10% of total time vs 20% (p=0.01)
  
• Intern time spent with teaching activities = 2% of total time vs 8% (p=0.006)
• Duration of patient hospital stay = 4.61 days vs 4.10 days (p=0.002)
• There was no difference in time spent with patient care, patient outcome, and patient satisfaction.
  

The authors noted through qualitative responses, that having a dual-attending team was an important factor in greater trainee and attending satisfaction with the ITU team model. Trainees and attendings on the ITU teams enjoyed the increased exchange of ideas and debate, feedback, and different teach styles.

Bottom Line
The ITU team is a novel pro-education model which promotes greater attending supervision, a lower patient census, and more time for both self-reflective and structure learning on inpatient Medicine services. This study demonstrates that changes which prioritize education can be implemented without negatively impacting on the quality of clinical care.

How is this relevant for Emergency Medicine?
There are some fascinating findings from this study which are directly relevant to the field of EM. To me, this study seems to suggest:

• There should be greater resident and attending staffing in the ED so that the ratio of providers-to-patients is lower. This is especially relevant given that our ED's are so often crowded and overwhelmed with clinical demands. This may help move education up on the priority list for residents and attendings on shift.
  
• For residency programs with ED teaching attending shifts, where an attending physician's sole responsibility is to teach without clinical responsibilities, preference should be given to attendings who are stellar educators. This was the model used in this study.
• Future studies should look at the ideal patient load in balanced learning for EM residents.
  

Reference McMahon, G., Katz, J., Thorndike, M., Levy, B., & Loscalzo, J. (2010). Evaluation of a Redesign Initiative in an Internal-Medicine Residency New England Journal of Medicine, 362 (14), 1304-1311 DOI: 10.1056/NEJMsa0908136