Effect of Case-Based Learning Combined with a Workshop Teaching Model on Medical Record Writing Quality Among Residency Trainees

Summary

This study evaluates whether case-based learning, combined with a workshop-based teaching model, improves the quality of medical record writing among standardized residency trainees.

Abstract

This study evaluated the effect of a case-based learning (CBL) combined with a workshop teaching model on medical record writing quality among standardized residency trainees. This single-center, non-randomized controlled educational study included 200 standardized residency trainees, with 100 in the intervention group and 100 in the control group. The intervention group received the CBL combined with the workshop teaching model, whereas the control group received traditional lecture-based learning. Medical record writing scores were compared before and after the intervention. A linear mixed-effects model was used for the primary analysis, and analysis of covariance was performed as a sensitivity analysis. Subgroup and item-level analyses were also conducted. Baseline age, sex, specialty, training year, and pre-intervention total score were comparable between the two groups (all p > 0.05). After the intervention, total scores increased significantly in both groups, from 69.26 ± 9.66 to 85.02 ± 6.86 in the control group and from 68.92 ± 9.65 to 89.87 ± 6.44 in the intervention group (all p < 0.001). Score improvement was greater in the intervention group than in the control group (20.95 ± 6.96 vs 15.76 ± 6.19, p < 0.001). The linear mixed-effects model showed a significant time effect (β = 15.76, 95% CI 14.47–17.06, p < 0.001) and a significant group × time interaction (β = 5.19, 95% CI 3.35–7.03, p < 0.001). Sensitivity analysis yielded similar results. Greater improvements were also observed in several key items, including diagnostic and treatment planning, treatment adjustment, complication management, surgical/procedural records, informed consent forms, and discharge instructions. The CBL combined with the workshop teaching model was associated with greater improvement in medical record writing quality than traditional lecture-based learning and may represent an effective educational strategy for standardized residency training.

Introduction

Standardized residency training is an essential component of postgraduate medical education¹. The quality of such training is directly related to the development of clinical competence among residents and is also closely associated with healthcare quality and patient safety². During residency training, medical record writing serves not only as a legal documentation of clinical practice but also as an important reflection of residents’ clinical reasoning, information integration, and standardized diagnostic and therapeutic abilities³. High-quality medical records require not only completeness of content and standardized expression, but also the ability to accurately summarize key clinical information, formulate appropriate diagnostic and treatment plans, and demonstrate continuous clinical judgment in the course records⁴. Therefore, medical record writing competence has become one of the key indicators for evaluating the overall clinical competence of standardized residency trainees. In China, postgraduate medical education includes a nationally unified Standardized Residency Training (SRT) program, which provides competency-based clinical training for medical graduates before independent clinical practice. Residents complete structured clinical rotations and standardized assessments over approximately three years, depending on specialty.

In the course of teaching, however, several issues related to residents' medical record writing emerge, including incomplete history-taking procedures, inadequate diagnostic and therapeutic approaches, weak logic in analyzing disease development, and non-standardized medical record writing. Lecture-based learning (LBL), which focuses largely on transferring knowledge to trainees unidirectionally, proves effective at systematically explaining basic knowledge to trainees, but ineffective in encouraging active involvement, case analyses, problem integration, and standardized medical writing⁵. In particular, for cultivating medical record-writing skills, which involve both clinical reasoning and documentation, reliance on knowledge transmission alone is often insufficient to achieve optimal teaching outcomes.

Case-based learning (CBL) emphasizes the use of real clinical cases as learning vehicles. It promotes active thinking through problem-oriented and discussion-based learning, thereby helping improve clinical reasoning and comprehensive analytical ability⁶. Workshop-based teaching, by contrast, emphasizes standardized demonstration, group-based practice, immediate feedback, and error correction, making it more suitable for training standardized skills and behaviors^7,8. From the perspective of teaching mechanisms, CBL emphasizes whether trainees can analyze clinical problems, whereas workshop-based learning emphasizes whether they can correctly implement standardized practices. The combination of these two approaches may therefore provide complementary advantages in the teaching of medical record writing. Previous studies have shown that the combination of CBL and PBL can significantly improve clinical reasoning, critical thinking, and teamwork skills, and can also markedly enhance the clinical thinking ability of assistant general practitioners⁹. Nevertheless, the effectiveness of combining CBL with workshops remains insufficiently explored.

In addition, improvements in medical record writing quality may not be reflected merely by an increase in total scores. Still, they may be more specifically manifested in key domains such as diagnostic and treatment planning, treatment adjustment, prevention of complications, surgical or procedural documentation, informed consent, and discharge instructions. These aspects are closely related to the quality of clinical decision-making, continuity of care, physician–patient communication, and the legal validity of medical documentation, and therefore may better reflect the practical value of educational interventions. Meanwhile, trainees from different specialties and at different stages of training may differ in knowledge structure, clinical exposure, and learning needs^10,11. Whether different teaching models produce consistent effects across these subgroups also warrants further investigation.

Therefore, the study used a controlled design to analyze the impact of a teaching model combining case-based learning and workshops with a conventional lecture-based model on the quality of medical record writing by residents during standardized residency training. In particular, the current study examined differences between pre- and post-intervention total scores, improvements in specific score items, and patterns of which certain subgroups benefited from the intervention, to generate empirical evidence on optimal teaching models for medical record writing.

Protocol

This study was exempt from formal ethics review because it was an educational research study conducted among standardized residency trainees and did not involve patients, patient clinical data, biological samples, or any intervention affecting patient diagnosis or treatment. Before the study began, all participating trainees were fully informed of the study purpose, procedures, and intended use of the data, and written informed consent was obtained from all participants. Participation was voluntary, and trainees could withdraw at any time without affecting their training rights or benefits.

Study design and ethical considerations
This single-center, non-randomized controlled educational study was conducted among standardized residency trainees at a tertiary Grade A hospital in Hebei Province, China, between January 2024 and June 2024. Participants were allocated to either the intervention group, which received the CBL combined with the workshop teaching model, or the control group, which received traditional lecture-based learning, according to the routine teaching arrangements implemented during the study period rather than by random assignment. Therefore, no randomization or individual matching procedure was performed.

Participant enrollment and group allocation
A total of 200 trainees who participated in standardized residency training at a tertiary Grade A hospital in Hebei Province and completed the required teaching evaluation were included, with 100 trainees in the intervention group and 100 in the control group. Because this study was based on a routine educational program, no formal a priori sample size calculation was performed. Consecutive sampling was used, and all eligible trainees who completed the teaching program and met the inclusion criteria during the study period were enrolled. The two groups were not individually matched; instead, baseline comparability was assessed by comparing demographic characteristics and pre-intervention medical record writing scores. The study included standardized residency trainees from 2022, 2023, and 2024 cohorts, representing different stages of residency training rather than only first-year residents. The inclusion criteria were as follows: (1) complete baseline data; (2) available medical record writing scores both before and after the intervention; and (3) completion of the full teaching program during the study period. The exclusion criteria were as follows: (1) missing key baseline data; (2) absence of either pre-intervention or post-intervention score records; and (3) trainees who did not complete the required training.

According to the routine teaching arrangements, trainees who received traditional lecture-based learning were assigned to the control group, whereas those who received the CBL combined with the workshop teaching model were assigned to the intervention group. Baseline characteristics, including age, sex, specialty, training year, and pre-intervention medical record writing score, were compared between groups to evaluate baseline comparability.

Teaching intervention protocol
The teaching intervention lasted 12 consecutive weeks. Teaching sessions were held once weekly, lasting half a day. Both the intervention and control groups followed the same teaching schedule, teaching objectives, assessment requirements, and total contact hours. The only difference between the two groups was the teaching strategy. The teaching was delivered by instructors experienced in standardized residency training and medical record quality management who had received unified training before the study to ensure consistency in teaching objectives, content, and assessment standards.

The intervention and control groups received training during the same study period, but their teaching sessions were organized separately according to the routine residency training schedule. The CBL + workshop case materials, guiding questions, workshop exercises, and feedback documents were used only during the intervention sessions and were not formally distributed to the control group prior to study completion.

To ensure comparability between groups, the total instructional time was identical. In the intervention group, part of the time allocated to conventional lectures was replaced with structured case discussion, workshop-based medical record-writing practice, and instructor feedback, without increasing overall teaching time.

Control group (LBL Group)
The control group received conventional lecture-based instruction in accordance with the standardized residency training syllabus. The teaching intervention lasted 12 consecutive weeks, with one half-day teaching session each week. The teaching content focused on standardized inpatient medical record writing, including admission records, first-course records, daily progress notes, diagnostic and treatment planning, treatment adjustment, complication management, surgical/procedural records, informed consent forms, and discharge instructions. Teaching materials included PowerPoint slides, teaching manuals, instructional videos, and representative examples of standardized inpatient medical records.

Teaching was instructor-centered. Instructors explained the principles of standardized documentation, demonstrated common documentation errors, and summarized key writing requirements. Trainees mainly learned through listening, note-taking, and reviewing representative examples. No structured case discussion, workshop-based practice, or immediate individualized feedback was provided.

Intervention group (CBL + workshop group)
The intervention group received the CBL combined with the workshop teaching model over the same 12-week period, with one half-day teaching session per week. The total instructional time was identical to that of the control group; however, part of the conventional lecture time was replaced by structured case discussion, workshop-based practice, and instructor feedback.

The teaching process consisted of three stages: (1) Pre-class preparation: instructors selected representative clinical cases according to the teaching objectives and prepared guiding questions related to clinical reasoning and medical record writing. Trainees reviewed the relevant literature, medical record writing standards, and case materials before each session. (2) In-class implementation: CBL sessions were first conducted. Trainees discussed the cases in small groups, analyzed clinical problems, developed diagnostic and treatment plans, and summarized key documentation points under instructor guidance. This was followed by workshop training, during which instructors demonstrated standardized medical record writing using the hospital's inpatient medical record templates, along with the Inpatient Medical Record Content Quality and Clinical Thinking Ability Evaluation Form. Trainees then completed structured medical record writing exercises based on the same clinical cases, while instructors provided on-site guidance and corrected documentation errors. (3) Post-class feedback: After each teaching session, instructors reviewed the completed medical records and provided immediate verbal feedback during the workshop, followed by written comments identifying common deficiencies and suggestions for improvement. Trainees revised their medical records according to the feedback before the subsequent teaching session.

A representative CBL case, including the case summary, discussion questions, medical record writing task, and instructor feedback framework, is provided in Supplementary File 1 to facilitate replication of the educational protocol.

Outcome assessment
The medical records included in the evaluation were inpatient records completed by standardized residents during routine clinical practice. Medical records generated from actual hospitalized patients were used rather than simulated or standardized written cases. To minimize the influence of differences in case complexity, representative inpatient medical records with comparable documentation requirements were selected for both the pre-intervention and post-intervention assessments. All records were evaluated using the same standardized assessment form and predefined scoring criteria. The assessment focused on documentation quality rather than disease-specific characteristics. Medical records were evaluated by a multidisciplinary assessment team consisting of one senior clinician with experience in standardized residency training and one senior medical record quality-control specialist. The senior clinician evaluated clinically oriented domains, including diagnostic accuracy, diagnostic and treatment planning, clinical reasoning as reflected in the record, treatment adjustment, and continuity of care. The medical record quality-control specialist evaluated documentation completeness, standardization, legal compliance, informed consent documentation, and discharge documentation. Both evaluators were blinded to group allocation and scoring time points, and the final score was calculated as the average of the two ratings.

Primary outcome
The primary outcome was the quality of medical record writing. In this study, medical record writing quality was defined as the completeness, accuracy, structure, standardization, legal compliance, and appropriate documentation of clinical reasoning within inpatient medical records. This form was developed by the institutional medical record quality management team based on standardized residency training requirements and hospital medical record quality-control standards. It was routinely used for residency teaching evaluation and medical record quality management. The total score of the form was 100 points, with higher scores indicating better medical record writing quality.

The assessment focuses on multiple domains of medical record quality, including completeness of documentation; history-taking and physical examination; diagnostic accuracy documented in the record; diagnostic and treatment planning; continuity of care; treatment adjustment; standardized documentation; physician–patient communication; legal compliance; and discharge documentation.

Medical record writing quality was assessed using the institution's Inpatient Medical Record Content Quality and Clinical Thinking Ability Evaluation Form. Despite its historical name, this institutional assessment form was developed primarily for evaluating the quality of inpatient medical records during standardized residency training. The clinically oriented items (e.g., diagnostic accuracy, diagnostic and treatment planning, and continuity of care) were used to assess whether clinical reasoning was appropriately reflected in the written medical record, rather than to directly measure clinical thinking ability as an independent educational outcome.

Item-level evaluation indicators
In addition to the total score analysis, improvements in individual scoring items were further compared. These items included diagnostic and treatment planning, timeliness of treatment regimen adjustment, prevention and management of complications, surgical/procedural records, standardization of informed consent forms, and discharge instructions, to explore differences between teaching models in specific aspects of medical record writing.

Time points of measurement
Two measurement time points were set in this study: T0, the pre-intervention assessment, and T1, the post-intervention assessment at the end of the intervention. The difference between T1 and T0 was defined as the improvement score and was used to evaluate the intervention effects of different teaching models.

Subgroup classification
Participants were divided by specialty into the internal medicine and surgery groups, and by training year into the 2022, 2023, and 2024 cohorts. Subgroup analyses were performed to compare changes in scores across specialties and training years under different teaching models. No randomization procedure was performed because group allocation followed the department's routine educational arrangement.

Statistical analysis
All statistical analyses were performed using R software, version 4.3.2. After testing for normality, continuous variables were expressed as mean ± standard deviation (x̄ ± s), and categorical variables were expressed as frequencies and percentages. For comparisons of baseline characteristics between the two groups, continuous variables were analyzed using the independent-samples t-test, and categorical variables were analyzed using the chi-square test or Fisher’s exact test.

For changes in total scores, descriptive results before and after the intervention were first compared between the two groups. Paired t tests were used to compare within-group differences before and after the intervention, and independent-samples t tests were used to compare differences in improvement scores between the two groups. Furthermore, a linear mixed-effects model was constructed with the medical record writing score as the dependent variable; group, time, and the group × time interaction as fixed effects; and age, sex, specialty, and training year as covariates, with participant ID included as a random effect, to evaluate the effects of different teaching models on score changes. Among these, the group × time interaction term was considered the main effect indicator. To verify the robustness of the results, a sensitivity analysis using a covariance analysis (ANCOVA) was conducted, with the post-intervention total score as the dependent variable and the pre-intervention total score as the covariate.

For subgroup analyses, participants were classified by specialty and training year, and a linear mixed-effects model was used to assess differences in improvement in scores between the intervention and control groups. For item-level analyses, the improvement score for each item was used to compare improvements between the two groups. Two-tailed tests were used, with p < 0.05 considered significant.

Results

Baseline characteristics
A total of 200 standardized residency trainees were included in this study, with 100 in the control group and 100 in the intervention group. The mean age was (26.72 ± 3.44) years in the control group and (25.95 ± 3.23) years in the intervention group. The total medical record writing scores before the intervention were (69.26 ± 9.66) and (68.92 ± 9.65), respectively. There were no statistically significant differences between the two groups in sex, specialty, or training year distribution (all p > 0.05), indicating good comparability of baseline characteristics between the two groups. In the baseline comparisons, the difference in age was not statistically significant (p = 0.104), and the difference in pre-intervention total score was also not statistically significant (p = 0.804). Likewise, there were no statistically significant differences in the distributions of sex, specialty, or training year (p = 0.109, 0.466, and 1.000, respectively) (Table 1).

Changes in total scores before and after the intervention in the two groups
After the intervention, the total medical record writing scores in both groups were markedly higher than before. In the control group, the total score increased from (69.26 ± 9.66) to (85.02 ± 6.86), whereas in the intervention group, it increased from (68.92 ± 9.65) to (89.87 ± 6.44). Paired t-tests showed that the differences before and after the intervention were statistically significant in both groups (all p < 0.001). Specifically, the mean increase was 15.76 ± 6.19 points in the control group and 20.95 ± 6.96 points in the intervention group. Comparison of the improvement scores between the two groups showed that the intervention group had a significantly greater increase than the control group (t = -5.57, p < 0.001) (Table 2, Figure 1).

Results of the linear mixed-effects model
The linear mixed-effects model showed a significant main effect of time (β = 15.76, 95% CI 14.47–17.06, p < 0.001), indicating that scores after the intervention were higher than those before. More importantly, the group × time interaction was significant (β = 5.19, 95% CI 3.35–7.03, p < 0.001), suggesting that, after adjustment for age, sex, specialty, and training year, the intervention group achieved an additional improvement of approximately 5.19 points compared with the control group. The ANCOVA sensitivity analysis yielded results consistent with the primary analysis, showing that the post-intervention total score remained significantly higher in the intervention group than in the control group (β = 5.15, 95% CI 3.86–6.44, p < 0.001), indicating good robustness of the findings (Table 3).

Subgroup analysis
Subgroup analysis revealed that, within the two specialty subgroups of internal medicine and surgery, the intervention group achieved a greater change in score than the control group. The interaction term between the group and time within the internal medicine subgroup was β = 4.34 (95% CI 1.98 to 6.70, p < 0.001), while the interaction term for the surgery subgroup was β = 6.51 (95% CI 3.51 to 9.51, p < 0.001). However, the interaction term among specialties was insignificant (p = 0.260).

In the stratified analysis by training year, it can be observed that all trainees in the cohort years of 2022, 2023, and 2024 demonstrated significantly greater performance increases when exposed to the intervention than when in the control group. Yet, none of the three-way interactions involving training year were significant (all p > 0.05) (Table 4).

Item-level analysis
The item-level analysis showed that the intervention group achieved greater improvement than the control group in several key documentation domains, including diagnostic and treatment planning, timeliness of treatment regimen adjustment, prevention and management of complications, surgical/procedural records, standardization of informed consent forms, and discharge instructions. For each item, pre-intervention score, post-intervention score, improvement score, between-group mean difference in improvement, 95% CI, and p-value are presented in Table 5.

Data Availability:
The anonymized datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request and with permission from the institution. Public deposition of the raw score data was not performed because the dataset was derived from institutional educational evaluation records of residency trainees and is subject to institutional data management restrictions.

Figure 1. Changes in total medical record writing scores before and after the intervention. The figure shows the pre-intervention and post-intervention total medical record writing scores for the control and intervention groups. Error bars represent standard deviations. Both groups showed increased scores after the intervention, with a greater improvement observed in the intervention group. Please click here to view a larger version of this figure.

Figure 2. Schematic diagram of the study design and intervention workflow.
The schematic diagram shows trainee recruitment, group allocation, teaching intervention, pre-intervention and post-intervention assessments, blinded scoring, and statistical analysis. Please click here to view a larger version of this figure.

Variable	Control group (n = 100)	Intervention group (n = 100)	Statistic	p-value
Age (years, x̄ ± s)	26.72 ± 3.44	25.95 ± 3.23	t = 1.634	0.104
Pre-intervention total score (points, x̄ ± s)	69.26 ± 9.66	68.92 ± 9.65	t = 0.249	0.804
Sex [n (%)]			χ² = 2.568	0.109
Male	44	32
Female	56	68
Specialty [n (%)]			χ² = 0.531	0.466
Internal medicine	65	59
Surgery	35	41
Training year [n (%)]			χ² = 0.000	1
2022 cohort	34	34
2023 cohort	33	33
2024 cohort	33	33

Table 1: Comparison of baseline characteristics between the two groups of standardized residency trainees. This table presents the baseline demographic and training-related characteristics of the control group and intervention group, including age, pre-intervention total score, sex, specialty, and training year.

Group	n	Pre-intervention (x̄ ± s)	Post-intervention (x̄ ± s)	Improvement (x̄ ± s)	Within-group p-value
Control group	100	69.26 ± 9.66	85.02 ± 6.86	15.76 ± 6.19	<0.001
Intervention group	100	68.92 ± 9.65	89.87 ± 6.44	20.95 ± 6.96	<0.001
NOTE: The between-group comparison of improvement scores showed a statistically significant difference (t = -5.57, p < 0.001).

Table 2: Comparison of total medical record writing scores before and after the intervention between the two groups. This table presents the pre-intervention, post-intervention, and improvement scores, along with within-group p-values, for the control and intervention groups. The between-group comparison of improvement scores showed a statistically significant difference.

Model	Variable	β	95% CI	p-value
Linear mixed-effects model	Group (intervention group vs control group)	-0.29	-2.02–1.45	0.746
Linear mixed-effects model	Time (post-intervention vs pre-intervention)	15.76	14.47–17.06	<0.001
Linear mixed-effects model	Group × time	5.19	3.35–7.03	<0.001
ANCOVA	Group (intervention group vs control group)	5.15	3.86–6.44	<0.001
NOTE: The linear mixed-effects model was adjusted for age, sex, specialty, and training year. In the ANCOVA model, the post-intervention total score was used as the dependent variable, with adjustment for the pre-intervention total score and baseline covariates.

Table 3: Results of the primary analysis and sensitivity analysis for total medical record writing scores. This table presents the results of the linear mixed-effects model and ANCOVA sensitivity analysis. The linear mixed-effects model was adjusted for age, sex, specialty, and training year. In the ANCOVA model, the post-intervention total score was used as the dependent variable, with adjustment for the pre-intervention total score and baseline covariates.

Analysis	β	95% CI	p-value
Subgroup analysis by specialty
Three-way interaction for specialty: group × time × specialty	—	—	0.26
Internal medicine subgroup: group × time	4.3	1.98–6.70	<0.001
Surgery subgroup: group × time	6.5	3.51–9.51	<0.001
Subgroup analysis by training year
Three-way interaction for training year: group × time × 2023 cohort	—	—	0.319
Three-way interaction for training year: group × time × 2024 cohort	—	—	0.753
2022 cohort subgroup: group × time	4.4	1.67–7.09	0.002
2023 cohort subgroup: group × time	6.2	4.06–8.43	<0.001
2024 cohort subgroup: group × time	5	2.02–7.92	0.001

Table 4: Results of the subgroup analyses. This table presents subgroup analyses by specialty and training year, including the group-by-time effects within each subgroup and the three-way interaction terms used to assess whether intervention effects differed across subgroups.

Item	Control pre-intervention	Control post-intervention	Control improvement	Intervention pre-intervention	Intervention post-intervention	Intervention improvement	Mean difference in improvement (95% CI)	p-value
Diagnostic and treatment planning	6.53 ± 2.47	7.81 ± 1.55	1.28 ± 2.27	6.67 ± 2.31	8.64 ± 1.28	1.97 ± 1.85	0.69 (0.03 to 1.36)	0.042
Timeliness of treatment regimen adjustment	3.52 ± 1.07	4.23 ± 0.73	0.71 ± 0.82	3.43 ± 1.14	4.60 ± 0.55	1.17 ± 1.02	0.47 (0.17 to 0.76)	0.002
Prevention and management of complications	3.41 ± 1.20	4.19 ± 0.78	0.77 ± 0.89	3.33 ± 1.27	4.52 ± 0.55	1.19 ± 1.07	0.41 (0.09 to 0.73)	0.011
Surgical/procedural records	6.56 ± 2.17	8.00 ± 1.42	1.44 ± 1.52	6.37 ± 2.32	8.67 ± 1.24	2.29 ± 1.88	0.85 (0.30 to 1.40)	0.003
Standardization of informed consent forms	3.40 ± 1.27	4.20 ± 0.75	0.80 ± 1.09	3.32 ± 1.40	4.56 ± 0.53	1.24 ± 1.25	0.44 (0.06 to 0.82)	0.023
Discharge instructions/orders	3.52 ± 1.21	4.39 ± 0.68	0.87 ± 0.96	3.24 ± 1.29	4.61 ± 0.54	1.37 ± 1.16	0.51 (0.16 to 0.85)	0.004

Table 5: Comparison of significantly improved items at the item level. This table presents item-level pre-intervention scores, post-intervention scores, improvement values, between-group mean differences in improvement, 95% confidence intervals, and p-values for documentation domains that showed greater improvement in the intervention group.

Supplementary File 1. Representative case-based learning (CBL) workshop case used for teaching medical record writing. This supplementary file provides the representative clinical case, learning objectives, discussion questions, workshop activities, instructor feedback framework, and documentation tasks used during the CBL, combined with the workshop teaching intervention, to facilitate replication of the educational protocol.Please click here to download this file.

Discussion

Based on standardized residency trainees' medical record writing score data, this study compared the effects of the CBL combined with the workshop teaching model and the traditional LBL teaching model on the quality of medical record writing. The results showed that the total scores of both groups increased significantly after the intervention compared with before, but the magnitude of improvement was greater in the intervention group. The linear mixed-effects model showed a significant group × time interaction, indicating that, after adjustment for age, sex, specialty, and training year, the intervention group achieved additional score gains compared with the control group. The ANCOVA sensitivity analysis yielded findings consistent with those of the primary analysis, indicating good robustness of the results. These findings suggest that, compared with the traditional lecture-based teaching model, the CBL combined with the workshop teaching model may be associated with greater improvement in medical record writing quality among standardized residency trainees.

These results generally agree with other research on educational interventions in medical record writing and workshop training. Medical record writing workshops for residents have been found to enhance the quality of medical records in prior research^12,13,14. Moreover, it has been noted that there is a slow but steady improvement in documentation when resident physicians receive structured chart documentation training in discharge summaries. In addition, several other researchers have found that education in medical record documentation improves the ability to document for medical students and resident physicians^15,16.

From a teaching perspective, CBL emphasizes real clinical cases as the core of learning and promotes active integration of medical history, physical findings, auxiliary examinations, and diagnostic and therapeutic reasoning through problem-oriented learning and group discussion, thereby strengthening the clinical reasoning process. Workshop-based teaching, in contrast, emphasizes standardized demonstration, group practice, immediate feedback, and error correction, which is more conducive to translating abstract writing standards into executable documentation behaviors. Existing medical education literature suggests that the core value of CBL lies in bridging theory and practice while enhancing learners’ active participation and clinical application ability¹⁷. Meanwhile, medical documentation is not merely a process of recording information, but also an important means of supporting the development of clinical reasoning and the integration of clinical problems. Therefore, the combination of CBL and workshops may have produced a more marked improvement in total scores by simultaneously influencing both the development of clinical thinking and the implementation of standardized documentation.

One additional interesting result obtained through the present study is that the benefits of using the CBL and workshop teaching model did not seem to be equally beneficial for all evaluation criteria, but rather were mostly observed within certain important areas, such as making diagnostics and developing a treatment plan, timely adjustments in the treatment protocol, preventive measures for possible complications, recording surgical procedures, standardizing the informed consent form, and discharge instructions. In comparison with more fundamental aspects of medical record keeping, such as collecting patient history and general information, these criteria require a higher level of trainee competence in evaluating patients' overall health status, maintaining continuity of diagnosis and treatment, and understanding medical documentation laws and regulations.

This finding is consistent with previous studies suggesting that documentation quality is not merely a matter of format, but is also closely related to clinical reasoning and the quality of feedback^18,19. Clinical reasoning-oriented documentation assessment tools developed by previous researchers have emphasized that high-quality medical records should not only document facts, but also clearly present problem summaries, differential diagnoses, and diagnostic reasoning. Other studies have shown that focused workshops on progress notes can improve residents’ understanding of the structure and evaluation standards of high-quality documentation and enhance their confidence in writing^20,21. These findings suggest that educational interventions are more likely to improve higher-level documentation domains, such as diagnostic and treatment planning, treatment adjustment, and analysis of disease progression, when they provide both clear assessment criteria and targeted feedback.

From a practical perspective, the item-level findings of this study provide meaningful implications for teaching. Diagnostic and treatment planning, as well as the timeliness of treatment regimen adjustment, reflect whether trainees can formulate appropriate management strategies based on available information and revise them promptly according to changes in the patient’s condition. Prevention and management of complications reflect risk awareness and continuity of clinical thinking. Surgical/procedural records, standardization of informed consent forms, and discharge instructions are directly related to patient safety, physician–patient communication, and legal compliance. Although traditional LBL has certain advantages in knowledge delivery, it often fails to adequately address the process of translating clinical reasoning into standardized medical record documentation. In contrast, the CBL combined with the workshop model more closely resembles the learning cycle in real clinical practice, namely, “case analysis–plan formulation–standardized execution–feedback acquisition–revision.” Therefore, the observed item-level improvements may reflect the CBL's structural characteristics and the workshop teaching model, although this interpretation should be confirmed in future multicenter studies.

Within both the internal medicine and surgery subgroups, the intervention group showed significantly greater improvements in scores than the control group. Moreover, the trainees from the 2022, 2023, and 2024 cohorts also demonstrated a trend toward higher scores. This suggests that CBL and the workshop teaching methods are useful not only for specific specialties or groups of students but also for all students. Regarding standardized training programs, it seems reasonable to say that the method will be more valuable.

However, it should be noted that in this research, the interaction terms between specialties and between training years were not statistically significant. Thus, even though it has been shown through the stratified analysis that there is a trend toward a positive effect of the intervention on the intervention group, it is too early to conclude that the effect of the intervention varied in relation to either the specialties or the training years, since no statistical evidence is currently available for that. Therefore, it would be premature to conclude that the effect is greater in internal medicine than in surgery or any other specialty, and likewise, that some particular training years benefit more from the intervention than others. Some previous studies suggest that competence, experience, and feedback may vary across training stages, thereby affecting teaching efficiency. However, this was not verified in the present study for various reasons, including the sample size and the assessment indicators used.

These findings also suggest that, in reforming standardized residency training, it may be more important to first establish a unified, standardized, and reproducible core teaching framework than to overemphasize specialty-specific differences at an early stage. For competencies such as medical record writing, which involve both clinical reasoning and documentation standardization, a foundational high-quality teaching module may be more important than excessive subgroup differentiation. On this basis, future studies with larger sample sizes, multicenter data, and more refined analyses by specific medical record types across different specialties may be more likely to identify true differential effects.

The findings of this study suggest that teaching in medical record writing should not remain at the level of traditional knowledge transmission, but should place greater emphasis on case-driven clinical reasoning training and feedback-based standardized skills training. For residency training administrators, the value of the CBL combined with the workshop teaching model lies not only in improving total scores but also in enhancing key domains such as diagnostic and treatment planning, dynamic management of changing clinical conditions, surgical/procedural records, and discharge instructions, all of which are closely related to healthcare quality, patient safety, and legal risk control. In other words, the advantages of this teaching model are not limited to helping trainees “write more completely,” but are more clearly reflected in helping them “write in a way that is more consistent with clinical decision-making logic and medical documentation standards.”

Taken together with previous literature and this study's findings, future teaching practice may be further optimized in three areas. First, standardized case discussion, peer assessment, and structured feedback mechanisms should be incorporated into routine medical record writing training for standardized residency trainees, so that trainees can repeatedly practice problem summarization, diagnostic and treatment planning, and written expression in the context of authentic clinical cases. Second, focused teaching modules may be developed around the domains that showed the greatest improvement in this study, such as diagnostic and treatment planning, the rationale for treatment adjustment, complication risk management, informed consent, and discharge instructions. Third, a continuous feedback and retraining mechanism should be established. Previous literature suggests that the effect of one-time training may be limited, whereas programs involving ongoing chart documentation teaching or repeated reinforcement over short cycles may be more likely to achieve sustained improvement in documentation quality.

Study limitations
This study has several limitations. First, this was a single-center, non-randomized controlled study. Participants were allocated according to the teaching model implemented during the study period rather than through random assignment. Therefore, selection bias and potential period effects cannot be completely excluded, although baseline characteristics were comparable between groups and adjusted analyses yielded consistent results. Future multicenter randomized studies are warranted to further confirm these findings. Second, the primary outcome of this study was the medical record writing score, which can adequately reflect changes in documentation quality, but broader clinical competency outcomes, such as clinical reasoning scale scores, actual medical error rates, or patient outcomes, were not included. Therefore, the evaluation of teaching effectiveness remains somewhat limited. Third, although subgroup analyses by specialty and training year were performed, none of the three-way interaction terms reached statistical significance, suggesting that the current sample size may still be insufficient to detect true subgroup differences. Finally, this study adopted total-score and item-level evaluations, which have practical value for teaching management; however, future studies may incorporate assessment tools that place greater emphasis on the quality of clinical reasoning expressed in documentation, in order to further improve the precision of identifying and providing feedback on high-quality medical record writing ability among standardized residency trainees. Because only the final averaged score was retained in the teaching evaluation database, the separate ratings of the two evaluators were unavailable; therefore, inter-rater reliability indicators such as the intraclass correlation coefficient could not be calculated retrospectively. Future studies should retain independent evaluator scores to allow formal assessment of inter-rater reliability. Because the primary outcome was medical record writing quality rather than clinical thinking ability itself, the clinically oriented items included in the assessment form should be interpreted as indicators of how well clinical reasoning was documented in the medical record, rather than as direct measures of clinical reasoning competency.

In addition, although the two groups attended separate teaching sessions and the CBL + Workshop materials were not formally distributed to the control group, all trainees were from the same institution. Therefore, informal communication between trainees cannot be completely prevented, and potential contamination bias cannot be entirely excluded.

Furthermore, although the assessment team included both a senior clinician and a medical record quality-control specialist to improve content validity, the evaluation still relied on an institution-specific assessment form. Future studies should use externally validated assessment instruments and retain independent evaluator scores for formal inter-rater reliability analysis.

Concluding remarks
In conclusion, compared with traditional lecture-based learning, the CBL combined with the workshop teaching model was associated with greater improvement in medical record writing quality among standardized residency trainees. Greater improvements were observed in several key documentation domains, including diagnostic and treatment planning, treatment adjustment, complication management, surgical/procedural records, informed consent documentation, and discharge instructions. These findings suggest that the CBL combined with the workshop teaching model may represent a useful educational strategy for standardized residency training. However, because this was a single-center, non-randomized educational study with short-term follow-up, further multicenter randomized studies are needed to confirm these findings.

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Inpatient Medical Record Connotation Quality and Clinical Thinking Ability Evaluation Form	The Second Hospital of Hebei Medical University	Not applicable	Institutional assessment form used for evaluating inpatient medical record writing quality during standardized residency training.
Microsoft Excel	Microsoft Corporation	https://www.microsoft.com/microsoft-365/excel	Used for data entry, data management, and preparation of table files.
R software	R Foundation for Statistical Computing	RRID: SCR_001905; Version 4.3.2	Used for statistical analysis, including t tests, chi-square tests, linear mixed-effects models, and ANCOVA.
Standardized inpatient medical record template	The Second Hospital of Hebei Medical University	Not applicable	Used during Workshop-based medical record writing practice.
Teaching materials	The Second Hospital of Hebei Medical University	Not applicable	Included PowerPoint slides, teaching manuals, instructional videos, representative medical records, CBL case materials, guiding questions, Workshop exercises, and feedback documents.