Optimizing Clinical Reasoning Assessments With Analytic and Holistic Ratings: Examining the Validity, Reliability, and Cost of a Simplified Patient Note Scoring Procedure

Automated Patient Note Grading: Examining Scoring Reliability and Feasibility

PURPOSE

Scoring postencounter patient notes (PNs) yields significant insights into student performance, but the resource intensity of scoring limits its use. Recent advances in natural language processing (NLP) and machine learning allow application of automated short answer grading (ASAG) for this task. This retrospective study evaluated psychometric characteristics and reliability of an ASAG system for PNs and factors contributing to implementation, including feasibility and case-specific phrase annotation required to tune the system for a new case.

METHOD

PNs from standardized patient (SP) cases within a graduation competency exam were used to train the ASAG system, applying a feed-forward neural networks algorithm for scoring. Using faculty phrase-level annotation, 10 PNs per case were required to tune the ASAG system. After tuning, ASAG item-level ratings for 20 notes were compared across ASAG-faculty (4 cases, 80 pairings) and ASAG-nonfaculty (2 cases, 40 pairings). Psychometric characteristics were examined using item analysis and Cronbach's alpha. Inter-rater reliability (IRR) was examined using kappa.

RESULTS

ASAG scores demonstrated sufficient variability in differentiating learner PN performance and high IRR between machine and human ratings. Across all items the ASAG-faculty scoring mean kappa was .83 (SE ± .02). The ASAG-nonfaculty pairings kappa was .83 (SE ± .02). The ASAG scoring demonstrated high item discrimination. Internal consistency reliability values at the case level ranged from a Cronbach's alpha of .65 to .77. Faculty time cost to train and supervise nonfaculty raters for 4 cases was approximately $1,856. Faculty cost to tune the ASAG system was approximately $928.

CONCLUSIONS

NLP-based automated scoring of PNs demonstrated a high degree of reliability and psychometric confidence for use as learner feedback. The small number of phrase-level annotations required to tune the system to a new case enhances feasibility. ASAG-enabled PN scoring has broad implications for improving feedback in case-based learning contexts in medical education.

Applying a validated scoring rubric to pre-clerkship medical students' standardized patient notes: a pilot study

BACKGROUND

With the elimination in 2021 of the United States Medical Licensing Examination Step 2 Clinical Skills test, it is incumbent upon U.S. medical schools to develop local validated assessments of clinical reasoning. While much attention has been paid to summative exams for graduating students, formative exams for pre-clerkship students have not been well studied.

METHODS

We applied the University of Illinois at Chicago College of Medicine (UIC-COM) Patient Note (PN) Scoring Rubric to templated PNs written by 103 pre-clerkship students for two cases in an objective structured clinical examination (OSCE) at the Yale School of Medicine. The rubric consists of four section scores (Documentation, Differential Diagnosis, Justification, and Workup, each scored 1 to 4) and a composite score (scaled 23 to 100). We calculated item discrimination for each section score and Cronbach's alpha for each case. We surveyed students about their experience writing the templated PN.

RESULTS

Mean Documentation, Differential Diagnosis, Justification, Workup, and composite scores for case A were 2.16, 1.80, 1.65, 2.29, and 47.67, respectively. For case B, the scores were 2.13, 1.21, 1.60, 1.67, and 40.54, respectively. Item discrimination ranged from 0.41 to 0.80. Cronbach's alpha for cases A and B was 0.48 and 0.25, respectively. A majority of the students felt that the exercise was useful and appropriate to their level of training.

CONCLUSIONS

Despite performing poorly, pre-clerkship students found the note-writing task beneficial. Reliability of the scoring rubric was suboptimal, and modifications are needed to make this exercise a suitable measure of clinical reasoning.