Application of problem-based learning in EBERs-ISH theoretical and practical teaching

Abstract

Background

In situ hybridization for Epstein-Barr virus-encoded small RNAs (EBERs-ISH) is a cornerstone of clinical pathology, essential for the diagnosis of EBV-associated diseases. However, traditional lecture-based learning (LBL) in EBERs-ISH instruction often emphasizes unidirectional knowledge transfer, which frequently results in diminished student engagement and inadequate practical troubleshooting skills. While problem-based learning (PBL) has demonstrated efficacy in enhancing clinical competencies within medical education, its application in EBERs-ISH training remains underexplored. This study aimed to evaluate the impact of PBL on the practical proficiency of pathology students regarding EBERs-ISH techniques.

Methods

This study involved postgraduate students majoring in Pathophysiology at The First Affiliated Hospital, College of Medicine, Zhejiang University. Participants were allocated into two cohorts: a control group (n = 32) receiving LBL and an experimental group (n = 29) engaging in a PBL curriculum. Both cohorts underwent a foundational training session on EBERs-ISH techniques. Subsequently, the experimental group participated in PBL modules structured around three distinct clinical scenarios. Upon completion of the instructional period, student proficiency was evaluated through theoretical and practical EBERs-ISH examinations. Statistical analysis was performed using non-parametric and chi-square tests, with statistical significance set at p < 0.05.

Results

Analysis of performance outcomes revealed that the PBL group achieved a significantly higher proportion of “excellent” scores compared to the LBL group (p < 0.013). However, no significant differences were observed in the proportions of “good” or “fail” scores between the two cohorts. In the practical EBERs-ISH assessment, the PBL group’s mean total score (81.9 ± 6.8) was substantially higher than that of the LBL group (65.9 ± 7.9; p < 0.001). Specifically, significant inter-group differences were identified in scores for sectioning quality (p < 0.001), artifact presence (p = 0.003), and staining quality (p < 0.001). Furthermore, Receiver Operating Characteristic (ROC) curve analysis confirmed that these assessment criteria effectively discriminated between the skill levels of the two groups.

Conclusions

PBL significantly enhances student proficiency in both the theoretical and practical aspects of EBERs-ISH when compared to traditional LBL. These findings indicate that PBL is a highly effective pedagogical strategy with considerable potential for broader application in clinical pathology technology education.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12909-026-08864-x.

Introduction

Epstein-Barr virus-encoded small RNAs in situ hybridization (EBERs-ISH) is an essential technique in clinical pathology, critical for the detection of Epstein-Barr virus within tissue specimens [1]. The accurate detection of Epstein-Barr virus (EBV) in pathological specimens is critical for the diagnosis, prognostic evaluation, and therapeutic planning of EBV-associated diseases [2]. As the gold-standard method, EBERs-ISH offers high sensitivity and specificity for visualizing viral presence within tissue sections [3]. However, proficiency in EBERs-ISH requires an integration of theoretical knowledge, technical skill, and critical thinking to interpret results and troubleshoot artifacts [4]. Traditional lecture-based learning (LBL) models, often centered on passive knowledge transmission, may not adequately equip trainees for the complex, problem-solving demands of real-world clinical practice [5, 6].

In medical education, Problem-Based Learning (PBL) has been established as an effective student-centered pedagogy [7]. This approach utilizes authentic clinical problems as the stimulus for learning, wherein students collaboratively identify knowledge gaps, conduct self-directed research, and apply their findings to resolve the initial problem [8]. The PBL framework fosters active learning and is proven to enhance critical thinking, problem-solving capabilities, and lifelong learning skills—competencies that are indispensable for medical professionals [9]. Consequently, applying a PBL model to EBERs-ISH instruction holds significant potential to create a more engaging, relevant, and effective learning experience [10].

While PBL has been successfully implemented in related disciplines such as immunohistochemistry and histopathology [11, 12], its application specifically to EBERs-ISH training remains largely unexplored. For a technique like EBERs-ISH, a PBL approach could provide trainees with structured opportunities to diagnose and resolve common technical errors (e.g., poor staining, sectioning artifacts), thereby deepening their understanding of underlying principles and improving final slide quality. Despite this potential, the existing literature focuses on the general application of PBL in clinical practical teaching [13–15], with a notable lack of studies specifically evaluating its impact on proficiency in a specialized molecular pathology technique such as EBERs-ISH [16].

This study aims to address this research gap by comparing the efficacy of a hybrid PBL-LBL model against a traditional LBL model for teaching EBERs-ISH practical skills to students of clinical pathology technology. By evaluating practical skill performance, we seek to determine the effectiveness of PBL in this context and provide evidence to support a new pedagogical approach for optimizing training in advanced clinical techniques.

Materials and methods

Study design

This investigation was structured as a randomized controlled trial (RCT) to assess the pedagogical efficacy of PBL versus traditional LBL for training in EBERs-ISH. The study protocol received ethical approval from the institutional review board, and written informed consent was secured from all participants prior to enrollment.

Participants and enrollment

Between June 2022 and June 2025, an initial cohort of 68 residents enrolled in a standardized pathology training program at our institution was screened for eligibility. Inclusion criteria stipulated that participants must be recent graduates (≤ 3 years) of a medical or biotechnology program, have completed all prerequisite pathology coursework, and be naive to EBERs-ISH theory and practice. Individuals were excluded for prior EBERs-ISH training, a documented history of poor compliance, or pre-existing psychological conditions that could interfere with the learning process. Of the initial cohort, 61 residents who met these criteria and provided written informed consent were enrolled. These participants were subsequently allocated via a computerized randomization algorithm to either the Problem-Based Learning (PBL) group (n = 29) or the traditional Lecture-Based Learning (LBL) group (n = 32). The core EBERs-ISH educational content was identical for both cohorts; the sole variable was the pedagogical delivery method.

Teaching methods

LBL Group (Lecture-Based Learning)

The LBL group received traditional instructor-led teaching. Following the institutional teaching syllabus, instructors delivered theoretical instruction via PowerPoint presentations. The curriculum encompassed ten core topics: (1) historical evolution of EBERs-ISH technology; (2) basic biological characteristics of Epstein-Barr virus (EBV); (3) composition and storage requirements of EBERs-ISH reagent kits; (4) operational procedures for common laboratory instruments employed in EBERs-ISH; (5) impact of tissue fixation quality on staining outcomes; (6) common misconceptions in clinical interpretation of EBERs-ISH-negative results; (7) tissue-specific variations in pretreatment protocols; (8) standardized procedures for reagent preparation; (9) biosafety protocols for EBERs-ISH procedures; and (10) standardized documentation formats for pathological reports. Following each lecture, in-class quizzes focused on core theoretical knowledge were administered, and key learning points were summarized to reinforce understanding.

PBL Group (Problem-Based Learning with flipped classroom)

The PBL group engaged in student-led inquiry-based learning driven by real-world clinical and experimental problems. The pedagogical approach combined problem-based learning with a flipped classroom model and proceeded through the following phases:

• Phase 1: Pre-class Preparation and Self-directed Learning. Instructors distributed comprehensive learning materials, including standardized EBERs-ISH operational videos, PowerPoint presentations on probe design principles, and clinical case analysis guidelines. Students were provided guidance on accessing additional scholarly resources, such as PubMed literature databases and the Guidelines for EBER In Situ Hybridization Detection in Pathological Diagnosis published by the Chinese Society of Pathology. Specific problem scenarios tailored to EBERs-ISH are detailed in Supplementary Table S1. Students were organized into six collaborative learning groups, with each group assigned one to two problem scenarios. Group members self-directed task distribution (e.g., literature retrieval, experimental principle analysis, case discussion) and synthesized prior theoretical knowledge with newly acquired information through collaborative inquiry and discussion, aiming to achieve competency in EBERs-ISH-related knowledge domains and develop solutions to assigned problems.

• Phase 2: In-class Presentation, Discussion, and Instructor Feedback. Each group synthesized their learning outcomes into a structured presentation, which included: (1) analysis of the clinical significance of EBERs-ISH detection in their assigned case; (2) step-by-step explanation of the EBERs-ISH operational process as designed for the case; (3) interpretation of hypothetical or actual staining results; and (4) proposed solutions to the problem scenario. Peer students supplemented presentations with additional insights. Instructors provided targeted feedback and formative assessment, including correcting non-standard descriptions of procedural steps, identifying deficiencies in result interpretation logic, and guiding students toward deeper understanding of technically challenging concepts. Upon completion of all group presentations, the lead instructor provided a comprehensive summary of key and difficult concepts in EBERs-ISH instruction, addressed student-generated questions, and shared contemporary advances in EBERs-ISH technology along with practical expertise in troubleshooting complex clinical cases.

• Phase 3: Structured Laboratory Practice and Technical Validation. Following in-class sessions, students completed hands-on EBERs-ISH preparation on standardized formalin-fixed, paraffin-embedded (FFPE) nasopharyngeal carcinoma tissue samples. This phase allowed students to apply problem-solving insights from earlier phases and validate their hypothesized technical modifications.

PBL implementation framework

Detailed structure for EBERs-ISH

To ensure systematic implementation of PBL principles specifically adapted to the technical demands of EBERs-ISH, the pedagogical framework was structured into five complementary phases:

• Phase 1: Problem Identification and Clinical Contextualization. Students were presented with authentic clinical cases featuring suboptimal EBERs-ISH results (e.g., weak signal intensity, high background noise, or false-negative outcomes in positive controls). These scenarios were drawn from actual diagnostic challenges encountered in clinical practice.

• Phase 2: Learning Objective Definition and Knowledge Gap Analysis. Through facilitated group discussion, students identified core knowledge gaps relevant to the technical failures presented. Representative learning objectives included: understanding the mechanistic role of Proteinase K digestion in tissue permeabilization; elucidating the thermodynamic basis for probe hybridization temperature optimization; and characterizing the signal amplification mechanism of the anti-FITC antibody detection system. Supplementary Table S1 shows ten real-world EBERs-ISH-focused clinical and experimental problems for PBL discussion (Supplementary Table S1).

• Phase 3: Independent Research and Evidence-Based Problem Solving. Students conducted systematic literature reviews and consulted instructional materials to identify evidence-based solutions addressing the identified technical failures. This phase emphasized integration of foundational knowledge (e.g., nucleic acid chemistry, protein enzymatic kinetics) with practical troubleshooting strategies.

• Phase 4: Collaborative Synthesis and Protocol Optimization. Group teams integrated their research findings through structured discussion, synthesizing disparate information into a coherent, revised experimental protocol designed to remediate the identified technical failures and optimize staining outcomes.

• Phase 5: Practical Validation, Laboratory Execution, and Expert Feedback. Students executed their optimized protocols in the laboratory setting using standardized tissue samples. Following completion, a debriefing session included expert evaluation of both the logical reasoning underlying protocol modifications and the objective quality of the resulting stained slides.

Outcome measures

Post-class theoretical assessment

All students completed a standardized post-class assessment immediately following the teaching intervention. The assessment comprised 10 semi-objective EBERs-ISH-related questions (total score: 100 points, 10 points per item). Items required limited-scope, objectively scorable answers that evaluated both theoretical knowledge and application of operational principles. Detailed assessment content is provided in Supplementary Table S2.

Practical slide preparation quality evaluation

Both groups prepared EBERs-ISH-stained slides using standardized formalin-fixed, paraffin-embedded nasopharyngeal carcinoma tissue specimens. Slide quality was evaluated using a structured rubric comprising 20 sub-items organized into four practical operational indicator categories (total score: 100 points, 5 points per sub-item). The evaluation rubric addressed critical technical dimensions including tissue morphology preservation, signal intensity and specificity, background control, and appropriate counterstaining. Detailed evaluation criteria are presented in Supplementary Table S3.

Teaching satisfaction assessment

Student satisfaction with the teaching intervention was assessed using a standardized university teaching evaluation form. This instrument comprises 10 items evaluating key pedagogical dimensions, including instructor effectiveness, classroom interaction quality, content depth, teaching methodology alignment with technical instruction, and classroom management, with each item rated on a four-point Likert scale: “Very Satisfied” (4 points), “Satisfied” (3 points), “General” (2 points), or “Dissatisfied” (1 point). Student responses were dichotomized: selections of “Very Satisfied” or “Satisfied” were classified as “satisfied”, while selections of “General” or “Dissatisfied” were classified as “dissatisfied”.

Student interest assessment and classification

A shorter version, the study interest questionnaire-short form (SIQ-SF, Supplementary Table S4), was developed in 2013, containing 9 items while maintaining strong reliability and validity, thus saving time for respondents [17]. Student learning interest was assessed using a structured Likert scale designed to evaluate three key dimensions: situational interest (immediate engagement during the EBERs-ISH procedure), individual interest (long-term predisposition toward molecular pathology), and perceived utility (relevance to future clinical practice). The instrument comprised 9 items (Supplementary Table S4), each rated on a four-point Likert scale: “Strongly Agree” (4 points), “Agree” (3 points), “Uncertain” (2 points), and “Disagree” (1 point). Based on the mean questionnaire scores, students were classified into three interest levels: High Interest (mean score > 3.0), Medium Interest (mean score 2.0–3.0), and Low Interest (mean score < 2.0).

Reliability and validity of assessment instruments

To ensure the scientific rigor and generalizability of findings, comprehensive reliability and validity analyses were conducted on all assessment instruments.

For Internal Consistency, Cronbach’s alpha coefficient (α\alphaα) was calculated separately for the theoretical and practical assessments using SPSS version 22.0. Interpretation criteria were: α ≥ 0.8 (excellent), 0.7 ≤ α < 0.8 (good), and α < 0.7 (unacceptable). Additionally, split-half reliability with Spearman-Brown correction was calculated for the theoretical assessment, and item-total correlations (ITC ≥ 0.3) were examined to identify items with weak discriminative power.

For the practical assessment, inter-rater reliability among the evaluating experts was quantified using the Intraclass Correlation Coefficient (ICC) with a two-way random effects model and absolute agreement specification, calculated using R version 4.4.2. Interpretation criteria were ICC ≥ 0.75 (excellent), 0.5 ≤ ICC < 0.75 (moderate), and ICC < 0.5 (poor). Practical slides were independently evaluated by five expert pathologists, each with ≥ 5 years of experience in EBERs-ISH and clinical pathology instruction.

For Content Validity, five expert raters (≥ 5 years of EBERs-ISH and clinical pathology teaching experience) independently evaluated each assessment item across three dimensions—relevance to core EBERs-ISH competencies, importance to clinical and educational needs, and clarity of wording—using a four-point Likert scale. Item-level content validity indices (I-CVI) were calculated as 1 if ≥ 80% of experts scored ≥ 3 on all three dimensions, and 0 otherwise. Scale-level content validity index (S-CVI) was computed as the average of all I-CVIs, with S-CVI ≥ 0.8 considered acceptable. Kendall’s coefficient of concordance (WWW) was used to assess inter-expert agreement (W ≥ 0.5 W indicating good consensus), and consensus meetings were convened to address items with low agreement.

For the theoretical assessment, construct validity was evaluated using exploratory factor analysis (EFA) with principal component analysis and varimax rotation. The instrument was hypothesized to measure two latent constructs: “basic EBERs-ISH principles” (Items 1–4) and “practical application and troubleshooting” (Items 5–10). Factors with eigenvalues ≥ 1 were retained. Acceptable construct validity was indicated by: (1) factor loadings ≥ 0.5 on hypothesized constructs and (2) cumulative variance explained ≥ 50%.

Statistical analysis

All statistical analyses were conducted using R software (version 4.4.2, https://www.r-project.org/). Categorical variables were expressed as counts and percentages, and analyzed using the chi-square test, Yates’ corrected chi-square test, or Fisher’s exact test as appropriate. Given that variance homogeneity testing indicated unequal variances between groups for theoretical and practical examination scores, and that ordinal data (e.g., excellent [≥ 80 points], good [≥ 70 points], failed [< 60 points] categories) did not conform to normal distribution assumptions, non-parametric statistical tests were employed. The Kruskal-Wallis H test was used to compare average scores between the PBL and LBL groups. The Mann-Whitney U test with Bonferroni correction was applied for pairwise group comparisons. Chi-square tests were used to compare categorical outcomes (excellent rate, qualified rate) between groups. For multiple comparisons, the Bonferroni-corrected significance level was α = 0.0125, with P < 0.0125 considered statistically significant. For primary comparisons between the two groups, the significance level was α = 0.05, with P < 0.05 considered statistically significant (two-tailed tests). Effect sizes were calculated using appropriate methods (e.g., rank-biserial correlation for non-parametric comparisons) to quantify the magnitude of differences between groups.

Results

Baseline characteristics

The flow of participants through the trial, from initial screening to final allocation, is depicted in Fig. 1. A baseline comparison confirmed that the PBL and LBL groups were homogenous, with no statistically significant differences observed in key demographic or academic characteristics prior to the intervention (Table 1).

Fig. 1.

Flow diagram of participant enrollment and allocation. The diagram illustrates the progression of participants from initial eligibility screening through to random assignment into either the Problem-Based Learning (PBL) or the traditional Lecture-Based Learning (LBL) intervention groups

Table 1.

Comparison of basic characteristics between PBL and LBL groups (n = 61)

Characteristic	Subcategory	PBL Group (n = 29)	LBL Group (n = 32)	P-value
Age	18–30 years	18	21	0.9825
	31 years and above	11	11
Gender	Male	8	7	0.8262
	Female	21	25
Year of study	Freshman	17	19	0.2344
	Sophomore	8	7
	Junior	4	6
Major background	Medicine	25	27	1.0000
	Biotechnology	4	5
Internship experience	Yes	19	21	1.0000
	No	10	11
Previous similar course	Yes	15	15	0.9030
	No	14	17
Family support	High	10	11	1.0000
	Medium	19	21
Motivation for learning	Intrinsic	4	5	0.0738
	Extrinsic	3	4
	Mixed	22	23

Teaching effectiveness evaluations

As detailed in Table 2, student evaluations of teaching effectiveness revealed significant pedagogical differences between the two groups. Participants in the PBL group rated their instructors significantly higher on metrics related to fostering an engaging and interactive learning environment. This included perceptions of teaching enthusiasm, motivation, classroom atmosphere, and the ability to inspire critical thinking (P < 0.05 for all). Furthermore, the PBL instructors were perceived as more effective in utilizing diverse teaching media and aligning their methods with the specific technical demands of EBERs-ISH.

Table 2.

Comparison of students' evaluation of course teaching effectiveness between the PBL Group and the LBL Group

Teaching evaluation systems	Group (satisfied/dissatisfied, n)		P-value*
	PBL (n = 29)	LBL (n = 32)
The instructor demonstrates enthusiasm in teaching and effectively motivates students' engagement	25/4	14/18	0.0011
Excellent interaction effectiveness and an active classroom atmosphere	27/2	16/16	0.0002
The instructor proficiently uses Mandarin and professional English in teaching	25/4	24/8	0.3430
The instructor sets clear teaching objectives, highlights key points, and presents content with logical thinking	28/1	29/3	0.6141
The instructor's teaching content is rich and in-depth (covering both theoretical and practical dimensions)	26/3	27/5	0.7100
The instructor exhibits excellent classroom management and control capabilities	27/2	31/1	0.6003
The instructor inspires students’ thinking, association, and innovative awareness	27/2	23/9	0.0455
The instructor rationally utilizes various teaching media, leading to good teaching effectiveness	28/1	24/8	0.0277
The instructor maintains appropriate demeanor and a professional appearance	24/5	28/4	0.7242
The instructor adopts appropriate teaching methods that align with the characteristics of EBERs-ISH technical teaching	27/2	19/13	0.0027

*Bold italic indicates statistically significant differences

Conversely, no significant differences were observed in foundational teaching competencies (p > 0.05). Students in both cohorts held equally positive views of their instructors’ professionalism, including their command of language, clarity of objectives, depth of content, classroom management, and professional demeanor (Table 2).

Theoretical knowledge assessment

An analysis of EBERs-ISH theoretical test scores from 2022 to 2025 is summarized in Table 3. While performance was comparable in the initial year, a clear divergence emerged from 2023 onward. During this period, the PBL group consistently achieved a significantly higher proportion of “excellent” scores (≥ 80 points) compared to the LBL group. The annual excellent rates for the PBL versus LBL groups were 100% vs. 27.3% in 2023 (P < 0.001), 75% vs. 0% in 2024 (P = 0.001), and 83.3% vs. 0% in 2025 (P = 0.002), respectively (Fig. 2A).

Table 3.

The statistical analysis of excellent rate, good rate and failed rate for EBERs-ISH-related theoretical scores in each year

Year	Indicator	Group (meet/not meet, n)		P-value#	Bonferroni-corrected significant differences
		PBL	LBL
2022	Excellent rate	3/2	1/4	0.5186	ns
	Good rate	5/0	2/3	0.0384	ns
	Failed rate	0/5	1/4	0.2918	ns
2023	Excellent rate	10/0	3/8	0.0006	*
	Good rate	10/0	7/4	0.0341	ns
	Failed rate	0/10	3/8	0.0745	ns
2024	Excellent rate	6/2	0/9	0.0012	*
	Good rate	7/1	5/4	0.3630	ns
	Failed rate	0/8	3/6	0.0719	ns
2025	Excellent rate	5/1	0/7	0.0021	*
	Good rate	6/0	3/4	0.0261	ns
	Failed rate	0/6	1/6	0.3352	ns

ns No significant difference

*statistical significance after correction (P <0.0125)

^#Bold italic indicates statistically significant differences. Excellent (≥80 points), Good (≥70 points), and failed (<60 points)

Fig. 2.

Comparison of annual assessment scores between the PBL and LBL groups (2022–2025). Bar graphs depict the mean ± standard deviation for (A) theoretical knowledge test scores and (B) practical skills test scores. Statistical significance between groups for each year is indicated. *, P < 0.05; **, P < 0.01; ***, P < 0.001,****, P < 0.0001; ns, not significant

Practical skills assessment

The PBL group demonstrated significantly superior performance on the EBERs-ISH practical skills assessment (Table 4). The total mean score for the PBL group was substantially higher than that of the LBL group (81.9 ± 6.8 vs. 65.9 ± 7.9; P < 0.001). A breakdown by scoring criteria revealed that the PBL cohort achieved significantly higher scores in minimizing slice defects (21.2 ± 3.9 vs. 12.3 ± 5.7; P < 0.001), reducing artifacts (20.2 ± 4.5 vs. 18.3 ± 4.5; P = 0.003), and executing proper staining protocols (30.2 ± 4.3 vs. 24.1 ± 6.4; P < 0.001).

Table 4.

Descriptive statistics of EBERs-ISH-related practical operational scores between PBL and LBL group

Theoretical assessment			PBL Group (Mean ± SD)	LBL Group (Mean ± SD)	P-value*
Four levels of practical operational skills (5 points each term)		Abbreviation
Misidentification	Wrong Wax Block Retrieval	WWBR	10.4±3.3	11.2±3.7	0.1507
	Incorrect Labeling	IL
	Mismatch Between Request Form and Waxblock Information	MFW
Slices defects	Frequent Wrinkles in Slices	FWS	21.2±3.9	12.3±5.7	<0.001
	Scratches, Damage, or Section Loss	SDSL
	Bubbles in Slices	BS
	Tissue Contamination	TC
	Uneven thickness	UT
Artifacts	Missing Controls	MC	20.2±4.5	18.3±4.5	0.0031
	False Positives in Controls	FPC
	False Negatives in Controls	FNC
	Inadequate proteinase digestion	IPD
	Excessively proteinase digestion	EPD
Staining	Overstaining	OS	30.2±4.3	24.1±6.4	0.0001
	Understaining	US
	Non-specific Staining	NSS
	Regional Lack of Staining	RLS
	Slices Not Restained	SNR
	Restaining Too Light	RTL
	Restaining Too Dark	RTD
Total score			81.9±6.8	65.9±7.9	<0.001

*Bold italic indicates statistically significant differences

A year-over-year analysis of practical test “excellent” rates further confirmed these findings (Table 5). The PBL group consistently outperformed the LBL group, with excellent rates of 100% vs. 20% in 2022 (p = 0.013), 50% vs. 0% in 2023 (P = 0.007), 62.5% vs. 0% in 2024 (P = 0.005), and 83.3% vs. 42.9% in 2025 (P = 0.002), respectively (Fig. 2B).

Table 5.

The statistical analysis of excellent rate, good rate and failed rate for EBERs-ISH-related practical operational scores in each year

Year	Indicator	Group (meet/not meet, n)		P-Value#	Bonferroni-corrected significant differences
		PBL	LBL
2022	Excellent rate	5/0	1/4	0.0098	*
	Good rate	5/0	3/2	0.1138	NS
	Failed rate	0/5	1/4	0.2918	NS
2023	Excellent rate	5/5	0/11	0.0072	*
	Good rate	9/1	3/8	0.0037	NS
	Failed rate	1/9	4/7	0.1566	NS
2024	Excellent rate	5/3	0/9	0.0048	*
	Good rate	8/0	5/4	0.0311	NS
	Failed rate	0/8	1/8	0.3311	NS
2025	Excellent rate	5/1	0/7	0.0021	*
	Good rate	6/0	3/4	0.0261	NS
	Failed rate	0/6	3/4	0.0675	NS

NS No significant difference

*statistical significance after correction (P<0.0125)

^#Bold italic indicates statistically significant differences

Inter-year score stability

To assess score consistency over the study period, a rank sum test was performed on the scores from 2022 to 2025 for both groups. Pairwise comparisons between all year combinations yielded no statistically significant differences for either the PBL or LBL group (all P > 0.05), indicating stable scoring standards across the four-year duration of the study (Fig. 3).

Fig. 3.

Inter-year stability of assessment scores for the PBL and LBL groups. Box-and-whisker plots illustrate the distribution of scores for each year from 2022 to 2025, demonstrating the consistency of the evaluation standards over time. Panels show (A) theoretical scores for the LBL group, (B) theoretical scores for the PBL group, (C) practical scores for the LBL group, and (D) practical scores for the PBL group

Performance distribution analysis

A Sankey diagram was generated to visualize the overall performance distribution across years, groups, and test types (Fig. 4A). The diagram clearly illustrates the superior outcomes of the PBL intervention. The PBL group accounted for 90% of all “excellent” scores across both theoretical and practical assessments. In contrast, the LBL group constituted 95% of all failing participants. The majority of LBL participants clustered in the “good” and “other (60–70 points)” categories, representing 70% and 95% of participants in those tiers, respectively. This demonstrates a clear shift in performance distribution, with PBL driving students toward excellence and LBL resulting in a higher incidence of mediocrity and failure.

Fig. 4.

Performance distribution and key skills analysis. (A) A Sankey diagram visualizes the flow of participants from their assigned group (PBL or LBL) through various performance categories (e.g., Excellent, Good, Fail) for both theoretical and practical assessments. (B) Receiver Operating Characteristic (ROC) curve analysis for subgroups of the EBERs-ISH practical skills assessment. The curves evaluate the power of specific technical skills to discriminate between the performance of the PBL and LBL groups, with the Area Under the Curve (AUC) serving as the primary metric

Identification of key differentiating skills via ROC analysis

To identify the specific technical skills that best discriminated between the two groups, a subgroup Receiver Operating Characteristic (ROC) curve analysis was conducted on the practical assessment criteria (Fig. 4B). The analysis revealed that mastery of several key steps was strongly associated with the PBL group’s superior performance. The highest discriminatory power was observed for minimizing “Scratches, Damage, or Section Loss” (AUC = 0.815), preventing “False Negatives in Controls” (AUC = 0.801), ensuring “Adequate Proteinase Digestion” (AUC = 0.786), and avoiding “Non-specific Staining” (AUC = 0.765). These high AUC values indicate that these skills were the primary drivers of the performance gap. Conversely, skills such as maintaining “Uneven Thickness” (AUC = 0.526), avoiding “Bubbles in Slices” (AUC = 0.420), and preventing “False Positives in Controls” (AUC = 0.384) showed poor discriminatory ability, suggesting both groups performed comparably on these metrics.

Psychometric validation of assessment instruments

The assessment instruments used in this study underwent rigorous psychometric validation (Supplementary Tables S4–S7). The Chinese version of the SIQ-SF demonstrated high content validity (I-CVI: 0.9–1.0) and stability (ICC: 0.714–0.848). The EBERs-ISH assessment framework exhibited excellent internal consistency for both theoretical (Cronbach’s α = 0.85) and practical (Cronbach’s α = 0.82) tests. Furthermore, high inter-assessor reliability was confirmed (ICC = 0.88), and content validity was robust (S-CVI ≥ 0.92) with significant inter-expert consensus (Kendall’s W, p < 0.001). Factor analysis confirmed the theoretical test captured its intended constructs, explaining 62.30% of the cumulative variance. Collectively, these data establish the psychometric integrity and reliability of the instruments used to evaluate educational outcomes.

PBL enhances academic performance by boosting student satisfaction and interest

To investigate the mechanisms underlying the PBL model’s effectiveness, we evaluated its impact on student satisfaction and learning interest and correlated these metrics with academic performance. The analysis, presented in Fig. 5, reveals that the PBL intervention created a significantly more engaging learning environment. Students in the PBL group reported substantially higher satisfaction scores (p < 0.001, Fig. 5A) and learning interest scores (p < 0.001, Fig. 5B) compared to their counterparts in the LBL group. Crucially, this heightened engagement was directly associated with improved academic outcomes. Correlation analysis demonstrated a significant positive relationship between students’ final overall scores and both their satisfaction levels (Fig. 5C) and their learning interest (Fig. 5D). These results strongly suggest that the PBL model’s success in improving test scores is, at least in part, attributable to its ability to foster a more satisfying and stimulating educational experience.

Fig. 5.

Problem-Based Learning Enhances Student Engagement and Correlates with Improved Academic Performance. (A) A violin plot comparing student Satisfaction or Interest Scores between the LBL and PBL groups. The plot displays the data distribution density, an embedded boxplot indicating the median and interquartile range, and overlaid individual data points. The PBL group reported significantly higher scores. (B) A scatter plot with marginal density distributions showing a significant positive correlation between Satisfaction Score and the final overall Score. The blue line represents the linear regression fit, and the shaded gray area indicates the 95% confidence interval. (C) A scatter plot with marginal density distributions revealing a significant positive correlation between Interest Score and the final overall Score. Statistical significance in panel A was determined using a two-sided Mann-Whitney U test (*** P < 0.001). Correlation analysis in panels C and D was performed using Pearson’s correlation coefficient, with the R and p-values displayed on the plots

Discussion

This study demonstrates that Problem-Based Learning (PBL) significantly enhances students’ performance in EBERs-ISH theoretical assessments and clinical slide preparation quality compared to traditional Lecture-Based Learning (LBL). These findings align with the core pedagogical principles of PBL, which emphasize active learning, problem-solving, and iterative practice [18]. PBL-trained students demonstrated superior proficiency in preventing slide defects, artifacts, and staining errors, likely attributable to hands-on troubleshooting experience embedded within case-based workshops. This observation is consistent with a meta-analysis by Arjanto et al. [19], which found that PBL improves technical skills through exposure to simulated real-world clinical challenges.

The primary contribution of this work lies in its systematic integration of PBL into specialized clinicopathological technical training—a pedagogical context that remains understudied in the existing literature [13]. Historically, PBL applications in medical education have concentrated on foundational medical sciences and general clinical clerkships, where the primary focus remains on theoretical knowledge synthesis and case-based clinical reasoning [11]. In contrast, our study tailored PBL specifically to the unique technical demands of EBERs-ISH, a procedure requiring precise manual dexterity, real-time troubleshooting, and critical assessment of staining quality [20]. We embedded three scenario-based discussions centered on authentic clinical and technical challenges, directly addressing the fundamental limitation of traditional LBL—namely, its reliance on unidirectional knowledge transmission, which fails to prepare learners for the dynamic and error-prone nature of clinical EBERs-ISH operations [16]. The PBL group’s marked dominance in the proportion of excellent performance and minimal failure rates underscores that PBL does not merely elevate test scores; rather, it cultivates the sophisticated technical reasoning and troubleshooting mindset required for safe, accurate clinical practice—an outcome consistently emphasized in prior investigations of PBL in pathological techniques [21].

To deepen our understanding of the mechanisms underlying PBL’s effectiveness in promoting EBERs-ISH competency, we analyzed the learning process through the lens of constructivist learning theory and Bloom’s taxonomy of cognitive domains [22]. Within the PBL framework, students are confronted with an authentic problem—suboptimal EBERs-ISH staining results (e.g., weak signal intensity, elevated background noise)—which catalyzes engagement in higher-order cognitive activities. Specifically, when students deliberate on questions such as “Why is Proteinase K incubation precisely 15 minutes rather than 10 or 20 minutes?”, they transcend rote memorization of procedural steps to engage with underlying enzymatic principles—the dynamic equilibrium between tissue permeabilization and RNA preservation. This problem-centered inquiry propels learners from Bloom’s lower cognitive levels (comprehension) toward intermediate and advanced levels (application, analysis, synthesis, and evaluation) [22]. Subsequently, when students execute optimized probe hybridization temperatures (43 °C vs. 45 °C vs. 48 °C) in the laboratory and observe the corresponding variations in signal intensity and specificity, they achieve the highest cognitive domains of synthesis and evaluation. In marked contrast, LBL students passively receive this information, remaining at the level of factual and procedural recall without mechanisms to promote higher-order cognitive development [23]. This theoretical framework elucidates why PBL demonstrated particularly pronounced advantages in the proportion of excellent performance: PBL directly scaffolds advanced cognitive abilities, whereas LBL predominantly supports foundational knowledge retention [24]. The integration of constructivist principles—wherein learning is constructed through active problem-solving and social interaction—further explains PBL’s superiority [25]; collaborative group discussions and peer feedback inherent in PBL create opportunities for conceptual refinement and perspective-taking that solitary lecture attendance cannot replicate [26]. A recent randomized controlled trial similarly reported that PBL reduces technical errors and fosters deeper understanding of procedural nuances, with these competencies correlating positively with course performance [27].

Although prior research has validated PBL’s efficacy in enhancing students’ knowledge and clinical decision-making skills [28, 29], this investigation uniquely focuses on the specialized context of EBERs-ISH—a technique demanding precision, meticulous attention to detail, and adaptive problem-solving. Our findings extend previous work [30], which identified that PBL improves diagnostic accuracy in EBV-related malignancies through enhanced slide interpretation. However, unlike that prior study, we quantified PBL’s impact on specific technical error categories, thereby providing actionable insights for evidence-based curriculum refinement.

Our utilization of ROC curve analysis to identify variables most strongly associated with slide quality strengthens the generalizability of our findings by pinpointing modifiable factors amenable to targeted instruction. A distinguishing feature of our analysis is the focus on subgroup differentiation; most prior investigations report aggregate outcomes without delineating which specific competencies drive observed improvements [31]. Our identification of four critical technical key points is clinically meaningful: these subdomains directly impact diagnostic accuracy. By linking PBL’s benefits to these high-stakes skills, we provide actionable insights for curriculum refinement, allowing educators to prioritize training on the most impactful aspects of EBERs-ISH training.

Regarding the apparent paradox in our results—namely, the statistically significant difference in excellent rate between PBL and LBL groups contrasting with the absence of significant differences in overall scores—our annual comparisons reveal an instructive pattern. The consistent intergroup difference in excellent rate across years, coupled with the lack of notable variation in good rates or failure rates, suggests a differential efficacy pattern whereby PBL uniquely elevates higher-level mastery of EBERs-ISH competencies. This observation aligns with PBL’s foundational educational objective of fostering deep content understanding and autonomous problem-solving ability. Conversely, the absence of differences in good rates and failure rates indicates that both pedagogies adequately support learners in achieving the baseline competency threshold for EBERs-ISH technical proficiency, suggesting the assessment instrument primarily discriminates at advanced proficiency levels—likely a reflection of its intentional design to evaluate higher-order cognitive and technical abilities.

From the perspective of hierarchical learning outcomes, this finding carries important implications: PBL does not uniformly elevate all students’ aggregate performance; rather, it demonstrates specific capacity to elevate intermediate performers to excellence, thereby exemplifying its particular utility in cultivating deep understanding and independent problem-solving capabilities. This selective efficacy is theoretically consistent with constructivist frameworks, wherein authentic problem-solving environments particularly benefit learners capable of engaging in higher-order reasoning. These findings further provide indirect yet meaningful evidence supporting the psychometric quality (reliability and validity) of our assessment instruments.

First, the assessment tool’s consistent ability to detect statistically significant intergroup differences in excellent rate across multiple years aligns with evidence of “construct validity”—it effectively captures meaningful variations in the targeted educational outcome, a cardinal indicator of whether an assessment measures what it purports to measure. Second, the absence of interannual fluctuations in overall scores for both groups suggest consistent test difficulty and standardized scoring procedures over time, thereby indirectly reinforcing the instrument’s “reliability”. This consistency minimizes the impact of random measurement error that could obscure true performance differences, ensuring reproducibility of assessment findings. Notably, the absence of overall score differences between groups does not compromise the assessment’s psychometric integrity; rather, it reflects that while PBL excels at promoting advanced proficiency, both instructional methods support general content mastery comparably. Collectively, these patterns validate the assessment’s ability to discriminate meaningful performance variations reliably and validly, while simultaneously highlighting PBL’s particular advantage in fostering advanced EBERs-ISH competencies. This research directly addresses a well-documented gap in molecular pathology education: the underutilization of PBL in training for specialized techniques like EBERs-ISH, despite its central role in clinical diagnosis of EBV-associated malignancies [32].

Historically, traditional LBL has been critiqued for creating a persistent “theory-practice divide”—trainees often master theoretical constructs yet struggle to apply them to real-world troubleshooting scenarios [33]. This gap is exemplified in our LBL cohort, which exhibited suboptimal performance in managing slide defects and elevated failure proportions in practical assessments—outcomes consistent with systematic review findings indicating that PBL-trained technicians are more likely to benefit from educational interventions [34]. Our findings demonstrate that PBL effectively bridges theoretical knowledge and practical skill by integrating problem-driven inquiry with real-time experiential feedback. Specifically, when students collaboratively deliberate on “Why is ISH background elevated?”, they enumerate multiple potential mechanisms (excessive antibody concentration, prolonged incubation duration, inadequate washing protocols); subsequent laboratory verification of each hypothesis enacts true theory-practice integration through iterative hypothesis-testing cycles. The significance of this finding is amplified given the global shortage of skilled molecular pathology technicians. Implementation challenges in EBERs-ISH training—including resource constraints, teacher professional development requirements, and assessment complexity—substantially impact pedagogical effectiveness. Recent investigations highlight that insufficient training in techniques such as EBERs-ISH delays diagnosis and treatment of EBV-associated diseases in low- and middle-income countries [35]. Unlike advanced lab equipment—often scarce in resource-limited settings—PBL relies on case-based discussions and collaborative problem-solving, making it a scalable solution. This aligns with Zhang’s finding that PBL improves academic performance [36] even in institutions with limited resources [37], as it reduces reliance on costly hands-on practice alone [38].

To maximize the potential impact of PBL in EBERs-ISH education and ensure its translational applicability across diverse institutional contexts, we propose an integrated teaching model centered on a closed-loop “Problem-Learning Objective-Validation” triad (Supplementary Figure S1). Within this framework, curriculum development should prioritize the design of core problems grounded in authentic clinical failures—such as elevated background, weak signals, or false negatives—rather than the direct prescription of standardized protocols. These problems must be mapped to explicit learning objectives aligned with EBERs-ISH technical critical points, ensuring that students understand the complex interactions between antibody concentration, incubation conditions, and washing protocols. By integrating laboratory practice, students are empowered to validate their problem-solving hypotheses, thereby fostering iterative cycles of hypothesis formation, experimental validation, and reflective refinement.

The successful implementation of this model necessitates a fundamental transformation of the instructor’s role, shifting from a “knowledge dispenser” to a “learning facilitator”. Faculty development should focus on training educators to design clinically authentic yet open-ended problems that stimulate deep reasoning while avoiding over-guidance. Furthermore, instruction in facilitative questioning techniques is essential to enable educators to employ recursive questioning that progressively deepens student understanding—for example, advancing from identifying an outcome to addressing alternative scenarios based on underlying mechanisms. Cultivating feedback expertise is equally critical, ensuring that instructors move beyond identifying errors to guiding students toward a comprehensive understanding of error etiology and prevention strategies.

Optimization of the assessment system represents another critical dimension, requiring the development of multi-level frameworks strictly aligned with PBL objectives. Our design of the “Protocol Comprehension Examination”—which incorporates principle-based items, troubleshooting scenarios, and protocol adaptation questions—has proven effective in discriminating student cognitive levels in the current study. We recommend the standardization and dissemination of this assessment framework to other molecular pathology technique curricula, while maintaining expert blind-scoring mechanisms to ensure objectivity and validity.

Finally, although PBL is often more economical than traditional methods by emphasizing collaborative discussion over high-end equipment, minimal resource support remains essential for sustainability. This includes the provision of adequate standardized tissue specimens for practical validation, the development of comprehensive case repositories populated with authentic clinical data, and the facilitation of literature access and discussion infrastructure. Systematic implementation of this integrated model across diverse institutional settings has substantial potential to achieve translational benefit and facilitate the broader dissemination of PBL’s educational advantages in specialized molecular diagnostics.

Moreover, the application of PBL in EBERs-ISH instruction carries broader implications for contemporary medical education. It aligns with modern pedagogical trends emphasizing development of practical technical skills, critical thinking capacity, and self-directed learning competencies [39]. By anchoring instruction in authentic problems rather than abstract knowledge, PBL renders EBERs-ISH education more contextually relevant and intrinsically motivating, potentially enhancing student engagement, and learning outcomes [39, 40]. The collaborative group-based nature of PBL additionally prepares students for the teamwork-dependent environment of clinical laboratory practice, wherein effective communication and coordinated problem-solving are essential for accurate diagnosis and optimal patient care [41, 42].

Limitations

While this investigation provides robust evidence supporting PBL’s efficacy in EBERs-ISH education, several limitations warrant acknowledgment to appropriately contextualize the findings. First, the sample size was modest and derived exclusively from a single academic institution. This restriction limits generalizability to alternative settings, such as community-based laboratories with limited molecular pathology resources or training programs serving non-resident personnel. Future investigations should expand recruitment across multiple institutions and encompass diverse learner populations (including nurses, laboratory technicians, and medical students), thereby strengthening external validity and broadening applicability. Second, despite implementation of double-blind scoring procedures and pre-validated assessment instruments, residual subjective bias in practical evaluation cannot be entirely eliminated. Although our scoring rubric provided explicit operational definitions for outcomes such as “non-specific staining”, subtle variations in assessor interpretation of “mild” versus “moderate” defects could introduce minor measurement variability. While attained inter-rater reliability mitigates this concern substantially, integration of automated assessment modalities—such as artificial intelligence-driven image analysis platforms [43]—would afford more objective quantification of slide quality. We recommend that future investigations explore incorporation of digital pathology and computational image analysis techniques into PBL outcome assessment, thereby further enhancing measurement objectivity. Third, this study lacks longitudinal follow-up to evaluate skill retention and clinical transferability. We assessed immediate post-training performance but did not determine whether PBL-trained residents maintained superior EBERs-ISH competency at 6–12-month intervals or whether acquired skills transferred to other molecular pathology techniques (e.g., human papillomavirus in situ hybridization, microbial in situ hybridization). Future investigations should employ longitudinal designs to evaluate both skill retention durability and the breadth of skill transfer capacity—assessments critical for establishing PBL as a sustained, scalable educational intervention with enduring value. Additionally, the present study’s PBL intervention targeted foundational EBERs-ISH competency development. The efficacy of PBL application to advanced clinical applications—such as diagnostic reasoning in complex cases or integrated multi-technique diagnostic approaches—remains unexplored. We recommend that subsequent investigations extend PBL application to advanced EBERs-ISH competencies and explore its integration with complementary diagnostic methodologies, thereby further expanding its pedagogical scope and demonstrating its utility across diverse levels of technical expertise.

Conclusion

This study demonstrates that integrating PBL with LBL significantly improves theoretical and practical proficiency in EBERs-ISH training for clinicopathological residents, with benefits in technical subdomains critical to diagnostic accuracy. Its innovation—tailoring PBL to the hands-on demands of EBERs-ISH—fills a gap in molecular pathology education, where PBL has long been underutilized for specialized technical skills. While limitations constrain generalizability, the findings support PBL’s adoption in EBERs-ISH training and provide a replicable model for other molecular techniques. By enhancing trainees’ ability to perform high-stakes EBERs-ISH operations, this work ultimately contributes to improving the quality of clinical pathology services and advancing competency-based medical education.

Abbreviations

EMA

Epithelial Membrane Antigen

EBV

Epstein-Barr Virus

EBERs-ISH

EBV-Encoded Small RNAs In Situ Hybridization

LBL

Lecture-Based Learning

PBL

Problem-Based Learning

RCT

Randomized Controlled Trial

ICC

Intraclass Correlation Coefficient

ITC

Item-Total Correlation

I-CVI

Item-Level Content Validity Index

S-CVI

Scale-Level Content Validity Index

EFA

Exploratory Factor Analysis

ROC

Receiver Operating Characteristic

AUC

Area Under the ROC Curve

Authors’ contributions

All authors contributed to the study conception and design. Material preparation, data collection, and analysis were performed by Meichun Zeng. The first draft of the manuscript was written by Meichun Zeng and Qingjun Jia, and all authors commented on previous versions of the manuscript. Qingjun Jia provided methods, reviewed drafts of the paper, and approved the final draft. All authors read and approved the final manuscript.

Funding

This work was supported by General Research Project of the Zhejiang Provincial Department of Education (Y202455526), and Zhejiang Science and Technology Plan for Disease Prevention and Control (2025JK025).

Data availability

Research data generated during this investigation are not released publicly to safeguard participant anonymity but may be accessed by contacting the corresponding author with a justified request.

Declarations

Ethics approval and consent to participate

The study was approved by the Hangzhou Center for Disease Control and Prevention (Hangzhou Health Supervision Institution) (Approval No. 2021-0008). Prior to study initiation, investigators provided a detailed briefing on the research objectives to all enrolled trainees. All participants subsequently signed written informed consent forms, explicitly authorizing their participation in the educational intervention and the subsequent publication of study outcomes. The confidentiality of personal data was explicitly guaranteed, and participants were formally informed of their right to withdraw from the study at any stage without incurring penalties. Participant information and all data materials were anonymized and securely stored in a database. This study was conducted in accordance with the Declaration of Helsinki.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.