Figure 1
Overview of Data Analysis Process and Phenotype Formation.
This flowchart depicts the three-step cross-case analytic process used to identify phenotypes of management reasoning struggle. In Step 1 (Within-Case Analysis), four researchers independently coded each transcript and organized codes into four domains: behaviors of struggle, contextual amplifiers, identification methods, and remediation strategies. In Step 2 (Cross-Case Analysis), a matrix of cases and codes was constructed and examined for patterns of co-occurrence. Conceptually coherent clusters advanced to phenotype formation. In Step 3 (Phenotype Formation), behavioral clusters were labeled and examined for alignment with codes from the other three domains to assign final phenotype descriptions.
Figure 2
Cross-Case Behavioral Matrix Organized by Phenotype.
This matrix displays the 15 learner cases (rows) and behavioral codes (columns) identified through cross-case analysis, organized by the four phenotypes of management reasoning struggle: limited management scripts, can’t change course, impaired task triage and execution, and failure to calibrate. Marked cells indicate the presence of a given behavior within a case, illustrating how co-occurring behavioral codes formed conceptually coherent clusters.
Table 1
Cross-cutting Findings: Behaviors, Contextual Amplifiers, Identification Methods, and Remediation Strategies that Appeared Across Phenotypes.
| CATEGORY | CODE | DESCRIPTION |
|---|---|---|
| Behaviors* | Freezes under uncertainty | Trainees stopped acting when the clinical situation lacked a clear answer or established path. |
| Defers to experts | Trainees handed management decisions to attendings, consultants, or peers rather than forming their own plan. | |
| Inability to articulate rationale | Trainees could not explain the reasoning behind their decisions when asked. | |
| Contextual Amplifiers | High-acuity settings | Management reasoning struggles were more visible when patients were sicker or more complex. |
| Low-supervision settings | Struggles became more apparent when attending oversight was reduced. | |
| Frequent rotation changes | Struggles were amplified by transitions between rotations. | |
| Frequent attending turnover | Frequent changes in supervising attendings worsened management reasoning struggles. | |
| Identification Methods | Faculty evaluations | Written evaluations from frontline supervisors describing concerns about management reasoning. |
| CCC or leadership review | Concerns identified or escalated during CCC meetings or aggregate review of performance data. | |
| Peer trainee feedback | Concerns raised by fellow trainees, including through chiefs who noted patterns across rotations. | |
| Direct observation by program leadership | PDs or associate PDs personally observed the trainee in the clinical environment. | |
| Informal faculty reports | Supervisors raised concerns through phone calls, side conversations, or emails to program leadership outside of formal evaluation systems. | |
| Duty hour review | Trainees staying significantly later than peers was used as an early signal prompting closer review. | |
| Nurse identification | Bedside nurses noticed and reported concerns about a trainee’s management. | |
| Targeted solicited feedback | Program leadership proactively sought narrative feedback from specific supervisors after an initial concern was raised. | |
| Remediation Strategies** | Longitudinal one-on-one coaching | A coach, typically a chief trainee or faculty member, worked with the trainee over time. |
| Clinical reasoning worksheets | Structured written exercises such as compare-and-contrast tasks and management script templates. | |
| Case-based learning | Guided review of clinical cases with a coach outside of active clinical duties. | |
| Schedule modification | Temporary reassignment to rotations with different acuity levels, supervision structures, or clinical exposures. | |
| Increased observation and check-ins | Chiefs or coaches shadowed trainees through their clinical workflow with frequent structured feedback. | |
| Chart-stimulated recall | One-on-one review of the trainee’s own clinical documentation to examine reasoning behind prior management decisions. | |
| External scaffolding | Structural supports added to the trainee’s workflow, including order read-back systems, pre-rounding checklists, and note templates. |
[i] *These behaviors appeared in cases spanning all four phenotypes, suggesting they may represent a common response to management reasoning difficulty regardless of its specific nature.
**No site described a formal curriculum for teaching management reasoning.