Read on Twitter
1/ #UncleBob presents his take on: @Broy3445 , @acastigl1 , @ryankraemr , Salanitro AH, @LisaWillett13 , Shewchuk RM, Qu H, Heudebert G, @medrants. Using cognitive mapping to define key domains for successful attending rounds. J Gen Intern Med. 2012 Nov;27(11):1492-8.
2/ I must start with a history lesson. @acastigl1 and I started this project early in the century, asking the important question - what do learners want from attending rounds.
3/ We started with Castiglioni A, Shewchuk RM, Willett LL, Heudebert GR, Centor RM. A pilot study using nominal group technique to assess residents' perceptions of successful attending rounds. J Gen Intern Med. 2008 Jul;23(7):1060-5
4/ We then met with the great Richard Shewchuk, Ph.D. He suggested that we use a consumer preference model to define domains and rank the many results from the NGT study. These techniques include aggregating the 70 suggested attributes down to 30.
5/ So the study took the aggregated attributes and had learners and attendings both do a card sorting process plus a Likert scaling of the 30 attributes. (For some thoughts on card sorting - https://www.usability.gov/how-to-and-tools/methods/card-sorting.html
6/ We analyzed data from the unforced card-sorting exercise using multi-dimensional scaling (MDS), then hierarchical cluster analysis (HCA), to create a visual interpretation of the underlying psychological dimensions of the attributes by placing them onto a “cognitive map.”
7/ Enough of methods! The participants in the study: Participant DemographicsN = 125
Attendings 36
Internal Medicine Residents36
Family Medicine Residents12
Third-year Medical Students40
Attendings 36
Internal Medicine Residents36
Family Medicine Residents12
Third-year Medical Students40
8/ Here is the big result: we found 5 domains. The axes as we labeled them (arbitrarily) show the multidimensional nature of rounds. Knowledge <-> Interpersonal skills. AND Managerial <-> Clinician. We feel that those rough dichotomies describe the complexity of our job.
9/ The domains (again the labels are somewhat arbitrary) further have face validity: clinical teaching, teaching style, learning atmosphere, communicating expectations, and team management. The best attendings have this complex collection of strengths.
10/ The Likert scaling revealed more specifics. Learners rated "Sharing of attending’s thought processes" as the most important attribute. In conversations with learners since we learned the results of the analysis, I am happy to report that almost all learners endorse this.
11/ Other top 6 attributes: “Teaching by example (e.g. good bedside manner)”, "Be approachable—not intimidating”, “Insist on respect for all team members”, “Conduct rounds in an organized, efficient & timely fashion”, and “State expectations for residents/students”
12/ What have I learned from this study 8 years later. Certainly, I am even more explicit about labeling my through processes. I have embraced @CPSolvers and helped start @UnremarkableLab that both focus on making clinical reasoning explicit.
13/ I have always worked on being approachable. I focus even more on actively showing respect for our leaners and our patients. I explicitly work hard to finish rounds on time every day. Every new team gets a very clear chat on expectations. Role modeling is very important.
14/ Several other things were made clear in this study. There is no preference for the style of rounds, as long as part of rounds occurs at the bedside. At the bedside we are role models. No matter what we tell the learners what to do, they will more likely copy our behaviors.
15/ I often debrief the learners after I interact with patients. We try to make explicit the skills used in talking with patients. When I do a mediocre job, I criticize myself explicitly so that the learners understand how to evaluate themselves.
16/ This study owes much to Dr. Kelley Skeff. His great paper: Skeff KM. The chromosomal analysis of teaching: the search for promoter genes. Trans Am Clin Climatol Assoc. 2007;118:123-32. is always worth reading
17/. I hope that this tweetorial is helpful. Please ask questions for clarification. If you want to read the article: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3475821/
18/ Thanks again to @gradydoctor who encouraged me to do this tweetorial. Obviously my colleagues who participated in the study and the physicians who did the card sorting and Likert scaling made this work possible.
