Characteristics differentiating problem representation synthesis between novices and experts

BACKGROUND

Formulating a thoughtful problem representation (PR) is fundamental to sound clinical reasoning and an essential component of medical education. Aside from basic structural recommendations, little consensus exists on what characterizes high-quality PRs.

OBJECTIVES

To elucidate characteristics that distinguish PRs created by experts and novices.

METHODS

Early internal medicine residents (novices) and inpatient teaching faculty (experts) from two academic medical centers were given two written clinical vignettes and were instructed to write a PR and three-item differential diagnosis for each. Deductive content analysis described the characteristics comprising PRs. An initial codebook of characteristics was refined iteratively. The primary outcome was differences in characteristic frequencies between groups. The secondary outcome was characteristics correlating with diagnostic accuracy. Mixed-effects regression with random effects modeling compared case-level outcomes by group.

RESULTS

Overall, 167 PRs were analyzed from 30 novices and 54 experts. Experts included 0.8 fewer comorbidities (p < .01) and 0.6 more examination findings (p = .01) than novices on average. Experts were less likely to include irrelevant comorbidities (odds ratio [OR] = 0.4, 95% confidence interval [CI] = 0.2-0.8) or a diagnosis (OR = 0.3, 95% CI = 0.1-0.8) compared with novices. Experts encapsulated clinical data into higher-order terms (e.g., sepsis) than novices (p < .01) while including similar numbers of semantic qualifiers (SQs). Regardless of expertise level, PRs following a three-part structure (e.g., demographics, temporal course, and clinical syndrome) and including temporal SQs were associated with diagnostic accuracy (p < .01).

CONCLUSIONS

Compared with novices, expert PRs include less irrelevant data and synthesize information into higher-order concepts. Future studies should determine whether targeted educational interventions for PRs improve diagnostic accuracy.

Developing a Web-Based Asynchronous Case Discussion Format on Social Media to Teach Clinical Reasoning: Mixed Methods Study

BACKGROUND

Case-based learning conferences are valuable to trainees, but growing clinical demands hinder consistent attendance. Social media increasingly acts as a venue for trainees to supplement their education asynchronously. We designed and implemented a web-based asynchronous clinical case discussion series on the Twitter social media platform to fill this educational gap.

OBJECTIVE

The aim of this mixed methods study is to examine the nature of interactions among web-based case discussion participants and assess local attitudes regarding the educational intervention.

METHODS

Starting in February 2018, we posted clinical vignettes to a dedicated Twitter account with the prompt "What else do you want to know?" to stimulate discussion. The authors replied in real time when case discussion participants requested additional details. Additional data about the case were posted at regular intervals to the discussion thread to advance the overall case discussion. Participants were asked to explain their reasoning and support their conclusions when appropriate. Web-based engagement was assessed using Twitter Analytics. Participants' posts were qualitatively analyzed for themes, with special attention to examples of using clinical reasoning skills. A codebook of types of participant posts and interactions was refined iteratively. Local engagement and attitudes at our institution were assessed by surveying internal medicine trainees (n=182) and faculty (n=165) after 6 months.

RESULTS

Over a 6-month period, 11 live case discussions were engaged with by users 1773 times. A total of 86 Twitter profiles spanning 22 US states and 6 countries contributed to discussions among participants and the authors. Participants from all training levels were present, ranging from students to faculty. Interactions among participants and the case moderators were most commonly driven by clinical reasoning, including hypothesis-driven information gathering, discussing the differential diagnosis, and data interpretation or organization. Of 71 respondents to the local survey, 29 (41%) reported having a Twitter account. Of the 29 respondents with Twitter accounts, 17 (59%) reported participating in the case discussions. Respondents agreed that case participation increased both their clinical reasoning skills (15/17, 88%) and clinical knowledge (13/17, 76%).

CONCLUSIONS

A social media-based serialized case discussion was a feasible asynchronous teaching method for engaging web-based learners of all levels in a clinical reasoning discussion. Further study should examine what factors drive trainee participation in web-based case discussions and under what circumstances asynchronous discussion might be preferred over in-person teaching activities.

Brain-Type Natriuretic Peptide and Amino-Terminal Pro-Brain-Type Natriuretic Peptide Discharge Thresholds for Acute Decompensated Heart Failure: A Systematic Review

BACKGROUND

Acute decompensated heart failure (ADHF) requiring hospitalization is associated with high postdischarge mortality and readmission rates.

PURPOSE

To examine the association between achieving predischarge natriuretic peptide (NP) thresholds and mortality and readmission rates in adults hospitalized for ADHF.

DATA SOURCES

Multiple databases from 1947 to October 2016 (English-language studies only).

STUDY SELECTION

Trials and observational studies that compared mortality and readmission outcomes between patients with ADHF achieving a specific predischarge NP goal and those not achieving the goal.

DATA EXTRACTION

Two investigators independently extracted study characteristics and assessed study risk of bias. One author graded the overall strength of evidence, with review by a second author.

DATA SYNTHESIS

One randomized trial, 3 quasi-experimental studies, and 40 observational studies were identified. The most commonly used thresholds were a brain-type NP (BNP) level of 250 pg/mL or less or an amino-terminal pro-brain-type NP (NT-proBNP) decrease of at least 30%. Achievement of absolute BNP thresholds reduced postdischarge all-cause mortality (7 of 8 studies) and the composite outcome of mortality and readmission (12 of 14 studies). Achievement of percentage-change BNP thresholds reduced the composite outcome (5 of 6 studies), and achievement of percentage-change NT-proBNP thresholds reduced all-cause and cardiovascular mortality (2 of 4 studies) and the composite outcome (9 of 9 studies). All findings were low-strength. The randomized trial, assessed as having high risk of bias, suggested that a predischarge decrease in NT-proBNP level was associated with lower risk for the composite outcome. Two quasi-experimental studies and 5 observational studies had low risk of bias. Low-risk-of-bias studies had outcome estimates similar in magnitude and direction to estimates from high-risk-of-bias studies.

LIMITATION

Most studies failed to adjust for critical confounders and had inadequate definition or assessment of exposures and outcomes.

CONCLUSION

Low-strength evidence suggests an association between achieving NP predischarge thresholds and reduced ADHF mortality and readmission.

PRIMARY FUNDING SOURCE

None.

Streamlined protocol for mRNA display

mRNA display is a powerful method for in vitro directed evolution of polypeptides, but its time-consuming, technically demanding nature has hindered its widespread use. We present a streamlined protocol in which lengthy mRNA purification steps are replaced with faster precipitation and ultrafiltration alternatives; additionally, other purification steps are entirely eliminated by using a reconstituted translation system and by performing reverse transcription after selection, which also protects input polypeptides from thermal denaturation. We tested this procedure by performing affinity selection against Her2 using binary libraries containing a nonspecific designed ankyrin repeat protein (DARPin) doped with a Her2-binding DARPin (dopant fraction ranging from 1:10 to 1:10 000). The Her2-binding DARPin was recovered in all cases, with an enrichment factor of up to 2 orders of magnitude per selection round. The time required for 1 round is reduced from ∼4-7 days to 2 days with our protocol, thus simplifying and accelerating mRNA display experiments.