Content area
Aim
To summarize existing research reviews on blended learning's effectiveness in nursing and medical education, offering teaching practitioners valuable references for continuous improvement in blended learning.
DesignAn umbrella review of systematic reviews and meta-analyses was undertaken using the PRISMA guidelines.
MethodsThis study retrieved all systematic reviews and meta-analyses related to blended learning, which were included in CNKI, VIP database, PubMed, Web of Science and Cochrane Library as of December 2024. The methodological quality of the included literature was evaluated using the AMSTAR-2 tool, while indications of learning effectiveness at each level were examined through the modified Kirkpatrick Model.
ResultsSeventeen studies were included: four with moderate, four with low and nine with critically low methodological quality. The modified Kirkpatrick Model suggests that blended learning in nursing and medical education may improve student feedback and attitudes and perceptions in the short term. Only nursing education has researched the long-term effects of blended learning, which improves critical thinking, collaborative knowledge and self-directed learning. However, organizational practices, student benefits and patient and community impact research are scarce.
ConclusionBlended learning can offer multiple beneficial impacts on students, but majority of the current systematic reviews examining the effectiveness of blended learning exhibit low methodological quality and lack standardized outcome evaluation systems. Future research and practice will also need to improve the standardization of reporting, focus on the long-term effects of blended learning and understand its impact on patients/communities.
With the advent of the electronic information age, information technology has gradually become the new driving force for the upgrading and development of various industries and has also brought new opportunities to the education field, giving rise to various forms of information-based learning ( Zhan-sheng and Bo-jie, 2014). Among them, blended learning (BL), a learning mode that integrates traditional face-to-face learning and online learning ( Williams, 2008), has gradually come into the public's view. As teachers' one-way transmission of knowledge dominates traditional offline lectures, meeting students' needs for all-round development is difficult. Pure online teaching, despite the application of information technology, leaves too much free space for students and fails to ensure the quality of teaching and learning ( Muller and Mildenberger, 2021),in contrast, the blended learning mode merges the benefits of traditional classroom instruction and digital learning platforms ( Regmi et al., 3 2024), enabling two-dimensional lectures ( Yan and Aola, 2023). Especially after entering the Internet+ era, the Asia-Pacific region has gradually increased its practice of BL in nursing and medical education. The design of several curricula, including microbiology ( Xue et al., 2024), medical immunization ( Bohong et al., 2024), rehabilitation therapy ( Qi et al., 2025) and professionalism education ( Shan et al., 2024), has also incorporated BL.
Despite the rapid development of blended learning, its exact implementation plan and effectiveness remain controversial. BL can potentially improve medical student ( Rowe et al., 2012) and undergraduate nursing education ( McCutcheon et al., 2015) by integrating knowledge outcomes, providing personalized learning paths and increasing student self-confidence and self-esteem compared with traditional learning ( Kononowicz et al., 2019; Fisher et al., 2021; Lee et al., 2023). It has also been argued that BL has limitations, such as the time cost of e-learning ( Rong et al., 2021), teachers not having enough experience in how to design courses in a blended format ( Megahed and Hassan, 2022) and the high demand for student-independent learning, which will significantly affect the effectiveness of teaching and learning if the students lack motivation and initiative. They won't keep up with the class if they don't study independently before it ( Bohong et al., 2024). However, in the post-COVID 19 era, the blended learning method combining online and offline advantages is in high demand ( Imran et al., 2023).
The number of articles on BL research has increased substantially in recent years. Although several systematic reviews have explored the effectiveness of BL, there is still considerable variation in the level of evidence, methods and quality of reporting. Furthermore, there has been no umbrella review of systematic reviews specific to BL. Therefore, this article aims to evaluate the effectiveness of BL in nursing and medical education by thoroughly reviewing existing systematic reviews and meta-analyses. It examines implications and distinctions in its application across the two domains of education, emphasizing the formulation of beneficial recommendations for nursing education. Through this review, we hope to provide valuable references for educators worldwide, particularly in the Asia-Pacific region, to promote the further development and optimization of BL in the field of nursing and medicine.
2 Materials and methods2.1 Study design
The umbrella review method (also known as a review of reviews) was chosen for this study because it highlights different interventions and outcomes, aggregates the results of multiple systematic reviews addressing similar issues ( Grant and Booth, 2009), provides a broader perspective and has a higher level rating for assessing the validity of evidence than other reviews. This review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines ( Page et al., 2021).We systematically retrieved, extracted and organized published systematic reviews and meta-analyses on the application of BL in nursing and medical education. The analysis focused on exploring the effectiveness of BL pedagogy applied to nursing and medical education at all Kirkpatrick Model levels based on results from existing studies.
2.2 Inclusion and exclusion criteria2.2.1 Inclusion criteria
- (a) Study participants included medical students, nursing students and other medical education-related students, with no restrictions on academic qualifications;
- (b) The intervention group was the BL method;
- (c) The control group is the traditional learning methods;
- (d) The outcome indicators include theoretical achievement, practical achievement, knowledge skills, self-directed learning ability, cooperation ability, learning interest, critical thinking, learning satisfaction and other relevant indicators;
- (e) The study design is a systematic review and meta-analysis of BL;
(a) Protocol; Conference summary; Narrative literature review; Network Meta-analysis
(b) Repeated publication or Literature in Non-Medical Education;
(c) Incomplete data or failure to extract valid information;
(d) Non-Chinese or English literature
2.3 Search strategyThe PubMed, Web of Science, Cochrane Library, CNKI and VIP databases were systematically reviewed to evaluate BL in medical education, covering the period from the establishment of each database until October 2024. The search terms were formulated based on the PICOS principle and the search strategy was tailored for different platforms, using a mix of subject terms and free words. The English search terms include: blended learning; blended teaching; hybrid learning; hybrid learning; blend* course; medical education; medical student*; nurs* student*; nurs* education; systematic review; and meta-analysis. The criteria for inclusion and exclusion were closely adhered to ensure the quality of the data. See Supplementary Table 1 for detail search forms.
2.4 Literature screening and data extractionThe study's screening and data extraction process strictly adhered to the PRISMA guidelines ( Liberati et al., 2009). To ensure the quality and relevance of the included studies, two researchers independently screened the literature, extracted the relevant information and cross-checked the results. We began by duplicating the retrieved literature using the EndNote X9 tool. Studies that did not meet the inclusion criteria were excluded based on an initial review of the title and abstract. Following this, the full texts were thoroughly reviewed to identify eligible studies. The following relevant information was extracted from the review using an Excel spreadsheet: (a) basic details, including the name of the first author and the year of publication; (b) study design, which encompasses the number of studies included, the study population, the duration of the study, the sample size and the interventions used in both the experimental and control groups; (c) tools used for assessing the risk of bias; and (d) the results and conclusions of the study.
2.5 Methodological quality assessmentThe methodological qualities of the included studies were assessed using the AMSTAR 2 tool ( Shea et al., 21 2017), which evaluates several dimensions, including reporting quality, risk of bias and statistical analysis. This tool consists of 16 items, with 7 key items (numbers 2, 4, 7, 9, 11, 13 and 15). Articles were rated as “yes,” “partially yes,” or “no” based on how well they met the criteria for each item. The status of methodology reporting was categorized into four levels: (1) High quality, indicating no non-compliance or only one non-critical item; (2) Medium quality, indicating noncompliance with multiple noncritical items; (3) Low quality, indicating noncompliance with one critical item, with or without noncritical items; and (4) Critically low quality, indicating noncompliance with multiple critical items.
Two investigators performed independent evaluations using AMSTAR 2 and cross-verified their results, discussing disagreements with a third investigator.
2.6 Classification of key indicatorsThe Kirkpatrick Model, proposed by Donald Kirkpatrick ( Kirkpatrick, 1998), evaluates learning effectiveness and learning outcomes, making it widely used in education. The model categorizes learning effectiveness into four levels: reactions level (learner satisfaction), learning level (knowledge and skill acquisition), behavioral level (ability to apply skills in practice) and results level (the impact of learning on the organization or individual) ( Praslova, 2010a). To enhance the applicability of the Kirkpatrick Model across various curricula and pedagogical areas, scholars like Barr ( Barr et al., 2005) and his colleagues, along with Steinert ( Steinert et al., 2006) and his team, have updated and adapted the model with more detailed and specific divisions within the four structural tiers. The application of the model has also been written about in higher education research ( Praslova, 2010a).
The assessment of the outcome indicators in this study is grounded in the updated version of the Kirkpatrick Model, as adapted by Leslie K ( Leslie et al., 2013). This model evaluates learning and learning effectiveness, categorizing it into seven distinct tiers:
- Learner Reaction (Level 1) 2. Change in Attitudes/Perceptions (Level 2a) 3. Knowledge/Skill Acquisition (Level 2b) 4. Behavioral Change (Level 3) 5. Change in Organizational Practices (Level 4a) 6. Benefits to Students/Residents (Level 4b) 7. Benefits to Patients/Community (Level 4c).
3.1 Search results and literature selection
In this study, 173 documents were retrieved from the database. After excluding 56 duplicates, 89 articles that did not meet the inclusion criteria were eliminated based on their titles and abstracts. Following a full-text screening, 17 studies were ultimately included ( McCutcheon et al., 2015; Abuejheisheh et al., 2023; Cheng et al., 2019; Du et al., 2022; Li et al., 2024; Niu et al., 2023; Vallee et al., 2020; Kegao and Pengxiang, 2020; Yue et al., 2022, 2024; Di et al., 2022; Haozhong et al., 2021; Yu et al., 2021; Yi, 2023; Yanjie et al., 2020; Xueqin et al., 2021; Yizhen et al., 2023). These comprised 11 meta-analyses, four systematic evaluations and meta-analyses, one systematic evaluation and 1 blended-methods systematic evaluation. Further searches conducted through Google Scholar and citation tracking did not result in any additional findings ( Fig. 1).
3.2 Basic characteristics of the included studiesAmong the 17 systematic reviews in this analysis, 16 were published between 2019 and 2024 and the original studies included were randomized controlled trials (RCTs). Only one publication from 2015 used a mixed methods review ( McCutcheon et al., 2015), with the majority of the articles included concentrating on nursing and clinical medical specializations. There are ten nursing course studies (such as basic nursing, internal medicine nursing, health assessment, CPR, education and clinical pathology), five clinical medicine specialty courses and one course each in Chinese medicine and basic medical studies, encompassing both theoretical and practical courses. The study population mostly consists of nursing students, college students enrolled in nursing courses, clinical medicine students, traditional Chinese medicine students, other health professionals (such as those in pharmacy and dentistry) and other groups.
All the interventions involved BL, which combines online and offline learning methods. Some reviews summarized the evaluation of the effectiveness of BL with the help of SPOC (Small Private Online Course) ( Yue et al., 2022; Yu et al., 2021; Yanjie et al., 2020), BOPPPS ( Li et al., 2024), Rain Classroom ( Haozhong et al., 2021) and other online combined with offline learning and traditional face-to-face learning. Evaluation tools were used for methods assessment, including the non-randomized interventional study evaluation tool ROBINS-I ( Di et al., 2022), the Cochrane Risk of Bias Assessment Tool (version 5.1.03), the Newcastle-Ottawa Educational Scale ( Xueqin et al., 2021), the Modified Jadad Rating Scale ( Haozhong et al., 2021) and the Critical Appraisal of the Literature Project (CASP) from the Centre for Evidence-Based Medicine in Oxford, UK ( Yue et al., 2024). These tools helped determine the risk of bias and assess the quality of the original literature included in the study ( Table 1).
3.3 Methodological evaluation resultsThe overall quality of the systematic reviews included was predominantly low due to various factors. Nine articles were classified as critically low quality, four as low quality, four as moderate quality and none as high quality. No study indicated the source of funding for the included research, 15 did not clarify the rationale for the systematic reviews of the study design types included, 13 failed to register in advance or to establish the evaluation methodology prior to conducting the evaluation and 10 did not present a list of excluded studies or disclose potential conflicts of interest. Moreover, six studies failed to examine the influence of bias on the overall effect and did not disclose the effects of publication bias on the evaluation outcomes. ( Table 2).
3.4 Kirkpatrick model evaluation index classificationAccording to the modified Kirkpatrick assessment model by Leslie-K et al ( Leslie et al., 2013). and its application in higher education ( $author1$ et al., 2022), the indicators for assessing the effectiveness of teaching and learning identified in the literature—such as knowledge, skills, attitudes towards learning and satisfaction—were systematically categorized into various levels. These levels include satisfaction (Level-1 Reaction), interest in learning (Level-2a Modification of attitudes/perceptions), theoretical examination scores, clinical skills scores, degree of knowledge acquisition (Level-2b Acquisition of knowledge/skills), critical thinking (Level-3 Behavioral change), teamwork skills (Level-4a Changes in organizational practice) and independent learning skills (Level-4b Benefits to students/residents). Data related to each index, including MD value, SMD value, RR value, effect size, confidence interval, P-value, heterogeneity and other quantitative metrics, were extracted to evaluate the learning impact completely.
3.4.1 Short-term effects on studentsLevel-1 and Level-2 of the Kirkpatrick Model evaluate the immediate effects of instructional strategies on learners. The initial layer is the response tier, which employs "satisfaction" as an outcome measure, indicating students' subjective perceptions of blended learning. Six research examined the impact of instruction at this tier, comprising five nursing education studies ( Cheng et al., 2019; Du et al., 2022; Yue et al., 2024; Yu et al., 2021; Yizhen et al., 2023) and one medical education study ( Di et al., 2022). The effect indicator for the five nursing instruction studies was the relative risk (RR), with a 95 % confidence interval (CI) of up to 1.35 (1.23, 1.49), while for the single medical trial, the standardized mean difference (SMD) was reported, with a 95 % CI of 0.32 (0.05, 0.59). The results of these research indicate that both nursing and medical students exhibit a greater degree of satisfaction with blended learning.
Level-2 represented the learning level, with 14 studies reporting outcome indicators at this level, comprising nine studies in nursing education and three in medical education. Level-2a concentrated on alterations in students' attitudes or perceptions toward learning during the training process, with "interest in learning" serving as the outcome indicator. Only one nursing study ( Yu et al., 2021) addressed this level, indicating that BL increased students' interest in learning and motivated them to learn more. Level-2b concentrated on students' proficiency in knowledge, abilities and attitudes, evaluated through three indicators: theoretical performance, clinical competencies and knowledge attainment. Eleven studies reported "theoretical achievement," the predominant outcome indicator at this level, with the majority of studies use the Standardized Mean Difference (SMD) as the effect measure, accompanied with a 95 % Confidence Interval (CI) of up to 1.73 (1.32, 2.15) ( Di et al., 2022). Three studies addressed the "extent of knowledge acquisition," nine studies examined "clinical skills," although one study ( Cheng et al., 2019) determined that blended learning was not statistically significant in enhancing students' clinical skills.
3.4.2 Long-term effects on studentsLevel-3 and Level-4 of the Kirkpatrick Model assess the long-term effects of BL on students. Level-3 is the behavioral tier that evaluates changes in student conduct following the implementation of BL. Only two nursing studies ( Niu et al., 2023; Yi, 2023) reported findings at this level using "critical thinking" as an outcome indicator, implying that BL was advantageous for improving students' critical thinking skills. However, the effect indicators exhibited inconsistency across the two investigations.
Level-4 represents the results level, where assessment functions as the definitive expression of the long-term effects of the Kirkpatrick Model.Level-4a examines the influence of BL on students' internal collaboration processes. Research on this topic is limited, with only one nursing study ( Yu et al., 2021) reported a pedagogical effect of blended learning on enhancing students' collaborative skills. Level-4b focuses on what students gain while receiving BL and their continued growth and development in their future careers. Five nursing studies ( Yue et al., 2024; Yu et al., 2021; Yi, 2023; Yanjie et al., 2020; Yizhen et al., 2023) have reported teaching outcomes at Level-4b, concluding that BL improves students' self-directed learning. No research has reported the consequences of BL at Level-4c, hence it remains uncertain whether BL benefits nursing and medical students' future care objectives (patients) and the workplace.
4 Discussion4.1 Methodological quality assessment and recommendations
Using the AMSTAR 2 scale, we observed a generally low quality of methodology reports in the 17 included systematic reviews, mainly for failure to fulfill the requirements of entries 2, 3, 7, 10 and 13. None of the systematic reviews reported the funding source for the included studies. Authors should describe in the text the source of funding for the included studies and the extent of the funder's involvement in all aspects of the study (including protocol design, execution process, data interpretation and writing up of the results) so that readers can judge whether the financial sponsorship biases the systematic review. Although all studies reported the types of studies included, fifteen studies failed to clarify the rationale for incorporating systematic evaluation. Systematic reviews may also encompass non-randomized intervention studies to supplement data in cases of inadequate randomized controlled trials or absent outcome indicators. Systematic review articles must comprehensively detail the design of the studies included and the justification for their selection. Thirteen studies failed to register in advance or develop evaluation methods before conducting the evaluation. A detailed implementation plan should be developed prior to conducting the study and discrepancies between the implementation process and the plan should be documented to mitigate the risk of bias. Ten studies did not provide a list of the excluded studies and did not report potential conflicts of interest. Authors should list excluded literature and reasons for exclusion to reduce selective bias. In addition, six studies did not report the risk of bias effect of included studies on outcomes. When systematic evaluations make recommendations that may have implications for educational policy, authors should fully consider the impact of bias on outcomes and explain and discuss the magnitude of the risk of bias.
4.2 Kirkpatrick model hierarchy study statusUsing the modified version of the Kirkpatrick Model in higher education can provide institutions with a versatile tool for developing and refining their evaluation and assessment systems. It also offers a way to understand both short- and long-term organizational outcomes, moving beyond immediate reactions to specific courses and beyond mere scores on standardized tests ( Praslova, 2010b). Our analysis of 17 systematic reviews using the modified of the Kirkpatrick model revealed that the studies encompassed all levels of the model (Reaction, learning, behavior and results). The initial two levels indicated short-term effects on students, while the latter two levels demonstrated long-term effects, albeit with an uneven distribution of studies across the levels.
4.2.1 Short-term effects on studentsAll 17 systematic evaluations indicated that BL has short-term effects on students. Six studies reached Kirkpatrick's level-1, evaluating student satisfaction with blended learning course design and found that students expressed greater satisfaction with blended learning compared with traditional instruction. However, it has also been noted ( Yu et al., 2021) that satisfaction is a subjective assessment metric susceptible to subjective judgments and thus, further research remains to be examined. Twelve studies achieved Kirkpatrick's level-2, with "theoretical results" being this level's most frequently used outcome indicator. Most of the studies judged the effectiveness of BL by examining the students' theoretical scores, operative scores and knowledge acquisition. The studies we included agreed that BL could improve students' theoretical scores and facilitate their knowledge acquisition. However, some studies reported that BL was not statistically significant in improving students' operational skills, which is a controversial point of discussion. More studies are needed to investigate whether BL genuinely improves students' operational skills.
4.2.2 Long-term effects on studentsThere is a lower frequency of studies addressing long-term impacts in comparison to those focusing on short-term impacts. Two studies achieved Kirkpatrick's level-3, concluding that critical thinking is essential to nursing practice and that blended learning enhances critical thinking skills in nursing students. Only one study reached Kirkpatrick's level-4a, concluding that BL improves undergraduate nursing students' ability to collaborate, but there were no findings regarding graduate and specialty nursing students. Furthermore, there is no evidence regarding the influence of BL on Level-4c levels, indicating insufficient investigation into the long-term competencies associated with students' behavioral and result levels. Few studies address the enhancement of students' collaborative competencies and critical thinking, indicating a lack of sufficient research to promote deeper learning and the advancement of higher-order thinking skills in students. Future research should enhance investigations at all levels of Level-3 and Level-4, particularly Level-4c. This includes exploring the influence of BL on students' comprehensive competency qualities and examining its effects on nursing students' future service recipients, specifically their patients and the community. Such research is essential for a more thorough assessment of the actual value of BL.
4.2.3 Applications and differences of BL in nursing and medical educationBased on the Kirkpatrick model, medical education has only explored the short-term effects of BL on improving students' theoretical and operative performance, whereas nursing education research has been more in-depth. With all four tiers of the model addressed, nursing education researches have reported not only the short-term effects of BL on students but also focuses on the long-term effects on students' critical thinking, collaborative skills and self-study skills. The current landscape of nursing and medical education concentrates on the short-term effects of BL on students, with nursing being the sole discipline to investigate the long-term effects on nursing students. Nonetheless, there exists a limited number of outcome indicators at each level, necessitating additional evidence. Although most of the studies included reported positive impacts, suggesting that BL significantly improves student learning outcomes and clinical skills and works well in terms of student satisfaction, some studies have indicated that BL may positively impact the health professions. However, this conclusion should be treated cautiously due to the high degree of heterogeneity ( Liu et al., 2016). In conclusion, BL and its outcome evaluation system in nursing education are more extensive and encompass more indicators than in medical education. The number of research suggesting that BL improves nursing students' cooperation, learning interest and critical thinking is modest. Future high-quality, multicenter and large-scale longitudinal studies can stimulate students' interest in learning and foster their ability to cooperate, as well as use pre-tests, post-tests, delayed tests and surveys to measure BL's role in nursing and medical education.
4.2.4 Recommendations for future research and practice based on the kirkpatrick modelIn nearly all research, the instruments used to collect and analyze data were not based on the same theoretical framework and the methods of data collection and analysis varied widely across the different levels of the Kirkpatrick Model. This inconsistency may contribute to discrepancies in the evaluation of the effectiveness of BL learning ( Liu et al., 2024). To improve future assessments, it is essential to establish standardized evaluation criteria while considering professional characteristics and cultural backgrounds that may influence outcomes. It is crucial to establish unified criteria and an evaluation system for theoretical achievement, professional skills, independent learning, satisfaction and learning quality. Additionally, we should use the Kirkpatrick Model for comprehensive assessments, combining quantitative and qualitative methods to create a unified data analysis standard. This will enhance the comparability and accuracy of evaluation results.
4.3 Developments and perspectives of BL in theory and practiceTheoretical model and practical case studies of BL have gradually matured in recent years. Theoretically, BL integrates various advanced educational concepts, such as constructivism and situational cognition theory, which provides solid theoretical support for this model. In addition, technological progress has also created conditions for its realization, platforms such as SPOC (Small Private Online Course), MOOC (Massive Open Online Course) and Rain Classroom have significantly enriched learning resources and methods ( Yue et al., 2022; Haozhong et al., 2021). Regarding the study design and selection of subjects, most of the literature used randomized controlled trials (RCTs) or quasi-experimental design to collect data and summarize them by statistical methods such as Meta-analysis ( Kegao and Pengxiang, 2020; Yu et al., 2021). One study ( Margolis et al., 2017) has suggested ten instructional best practices for using BL from the student's perspective, such as consistency when team teaching, timeliness in posting materials, incorporation of student feedback into the course. It has also been suggested that Chat-GPT guidance mechanisms be implemented in BL to promote students' self-learning skills and critical thinking abilities ( Lee et al., 2024). Due to the limited sample size, most studies only focus on specific specialties or regions, which makes it difficult to reflect the overall situation and can easily bias the research results. In addition, due to different national conditions, different education policies, different teachers and different learning styles of each teacher, the specific implementation plans are also quite different ( McCutcheon et al., 2015). At the same time, some controversial issues are beginning to emerge, such as whether long-term use will weaken face-to-face communication opportunities remain to be explored ( Li et al., 2024). Students face self-regulating challenges to use technology effectively for learning; the main challenges for teachers are their unwillingness and negative perception of using technology for learning ( Rasheed et al., 2019). In summary, evidence already supports the superiority of blended learning in improving the quality of teaching and learning in nursing and clinical medicine, both based on theoretical models and practical case studies. However, we should not ignore the challenges and difficulties in conducting BL. When designing BL, more attention should be paid to improving student control, enhancing teachers' confidence and acceptance ( Li et al., 2025), stimulating social interactions and creating an emotional learning atmosphere ( Boelens et al., 2017). The education system should also develop a scientific and reasonable BL-related teaching system and program as soon as possible. Simultaneously, we observed a deficiency of BL-related research within the domains of Chinese medicine and pharmacy; however, this does not preclude the nursing profession from using BL as a conduit for fostering interdisciplinary teaching between nursing and other medical fields, thereby enhancing the scope and depth of the nursing education knowledge system.
5 ConclusionThis study employed the modified Kirkpatrick Model to categorize and summarize evaluation indicators of BL. The model was used to comprehensively analyze its effectiveness in nursing and medical education. BL integrates traditional face-to-face learning with digital learning. It augments students' theoretical knowledge, clinical abilities and competencies like independent learning, critical thinking and collaboration when executed proficiently. However, no research has confirmed whether BL facilitates the translation of knowledge into practical applications, especially in patient care. Simultaneously, challenges remain in data collection and analytical methodologies, inconsistency in evaluating teaching effectiveness indicators, time and financial expenditures, unequal distribution of educational resources and the design of curricula that match with the characteristics of each program and student. In the future, we should strengthen multidisciplinary collaboration and allocate more funds for teaching to advance the development of blended learning in nursing and medical education worldwide, particularly in the Asia-Pacific area.
CRediT authorship contribution statementGuiru Lin: Project administration, Formal analysis. Jianmei Yang: Visualization, Project administration. Liang Yinji: Writing – review & editing, Validation, Supervision, Funding acquisition, Conceptualization. Chenli Lin: Writing – review & editing, Conceptualization. Yanyan Hong: Writing – review & editing, Formal analysis, Data curation, Conceptualization. Li Minhui: Writing – review & editing, Writing – original draft, Formal analysis, Data curation, Conceptualization. Wanyi Ou: Writing – review & editing, Formal analysis, Data curation. Aiping Wu: Writing – review & editing, Formal analysis, Data curation, Conceptualization. Yanfei Ruan: Formal analysis, Project administration. Zumin Chen: Formal analysis, Project administration;
Ethics statementThe authors have nothing to report.
FundingThis research was funded by the Program of China Scholarships Council (No. 202206785007), the Teaching Quality and Teaching Reform Project of Undergraduate University of Guangdong in China (2020), the Undergraduate Training Programs for Innovation and Entrepreneurship of Jinan University in China (No. S202310559058). The funding had no impact on the study protocol, design or other aspects of the study.
Declaration of Competing InterestThe authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
AcknowledgmentsWe thank the Program of China Scholarships Council, the Teaching Quality and Teaching Reform Project of Undergraduate University of Guangdong in China and the Undergraduate Training Programs for Innovation and Entrepreneurship of Jinan University in China for supporting this study financially.
Appendix A Supporting informationSupplementary data associated with this article can be found in the online version at doi:10.1016/j.nepr.2025.104421.
Appendix A Supplementary materialSupplementary material
| Author
(Year) | Included Studies | Curriculum Topic | Participants Type/age | Intervention/Control Measure | Methodological
Evaluation Tool | Outcomes | Main Conclusion | |||
| Study Design | Number | Sample | Interventions length | |||||||
| ( Huang et al., 2022) | Meta-Analysis | 66 | 13461 | NR | Clinical medicine | Clinical medicine students in China | blended learning / traditional learning | Non-randomized interventional study evaluation tool ROBINS-I | Theoretical knowledge, practical/operational skills and satisfaction | Compared with traditional learning, blended learning has a better effect on improving the theoretical knowledge and practical / operation skills of clinical medicine students in China and has a higher satisfaction degree. |
| ( Yizhen et al., 2023) | Meta-Analysis | 15 | 1193/ 1219 | NR | Nursing courses | Students in colleges and universities who are enrolled in or required to take nursing courses | Based on the rain classroom online and offline blended learning mode / traditional face-to-face learning mode | Cochrane Risk Bias Assessment Tool 5.1.0 | Theoretical assessment results, independent learning ability, satisfaction and operation results | The use of the rainy classroom model in nursing education improves students' final theoretical and operational exam scores, improves students' self-directed learning, increases students' satisfaction with the learning model and can be used as an adjunct to maintaining the quality of nursing learning during epidemics under epidemic normalization. |
| Author | Item 1 | Item 2# | Item 3 | Item 4# | Item 5 | Item 6 | Item 7# | Item 8 | Item 9# | Item 10 | Item 11# | Item 12 | Item 13# | Item 14 | Item 15# | Item 16 | Evaluation Final Rate |
| ( Yanjie et al., 2020) | Y | N | N | PY | Y | Y | N | Y | Y | N | Y | Y | Y | Y | Y | N | Critically low |
| ( Huang et al., 2022) | Y | N | Y | PY | Y | Y | Y | Y | Y | N | Y | Y | Y | Y | Y | N | Low |
| ( Yizhen et al., 2023) | Y | N | N | PY | Y | Y | Y | PY | Y | N | Y | Y | Y | Y | Y | N | Low |
| ( Yu et al., 2021) | Y | Y | N | PY | Y | Y | Y | Y | Y | N | Y | Y | Y | Y | Y | N | Moderate |
| ( Yi, 2023) | Y | N | N | PY | Y | Y | N | Y | Y | N | Y | N | N | Y | N | N | Critically low |
| ( Yue et al., 2022) | Y | N | N | PY | Y | Y | N | Y | Y | N | Y | N | N | Y | N | N | Critically low |
| ( Kegao and Pengxiang, 2020) | Y | N | N | PY | Y | Y | N | PY | Y | N | Y | Y | Y | Y | Y | N | Low |
| ( Yue et al., 2024) | Y | N | N | PY | Y | Y | N | PY | PY | N | Y | Y | Y | Y | Y | N | Low |
| ( Haozhong et al., 2021) | Y | N | N | PY | Y | Y | N | Y | N | N | N | N | N | N | N | N | Critically low |
| ( Xueqin et al., 2021) | Y | N | N | PY | Y | Y | N | N | N | N | Y | N | N | Y | Y | N | Critically low |
| ( Abuejheisheh et al., 2023) | Y | N | N | Y | Y | Y | N | Y | Y | N | N | N | Y | N | N | Y | Critically low |
| ( Niu et al., 2023) | Y | Y | N | PY | Y | Y | Y | Y | Y | N | Y | Y | Y | Y | Y | Y | Moderate |
| ( McCutcheon et al., 2015) | Y | N | N | Y | Y | Y | N | Y | Y | N | N | N | Y | N | N | Y | Critically low |
| ( Cheng et al., 2019) | Y | N | Y | PY | Y | Y | Y | Y | Y | N | Y | Y | N | Y | N | Y | Critically low |
| ( Du et al., 2022) | Y | Y | N | Y | Y | Y | N | Y | Y | N | Y | Y | Y | Y | Y | Y | Moderate |
| ( Li et al., 2024) | Y | Y | N | Y | Y | Y | Y | Y | N | N | Y | Y | Y | Y | Y | Y | Moderate |
| ( Vallee et al., 2020) | Y | N | N | N | Y | Y | Y | N | Y | N | Y | Y | Y | Y | Y | Y | Critically low |
| Kirkpatrick Model | Effect evaluation index | Author, Year | Intervention/ Control Population | Number of studies | Effect metrics | Random/Fixed effects summary estimate (95 % CI) | Random/Fixed effects p-value | I² |
| Level 1(Reaction) | degree of satisfaction | Huang Di, 2022 | / | 19 | RR | 1.35(1.23,1.49) | P < 0.05 | 88.20 % |
| Zhang Yizhen, 2023 | 393/405 | 5 | RR | 1.32(1.21,1.43) | P < 0.00001 | 0.00 % | ||
| Peng Yu, 2022 | 314/316 | 3 | RR | 1.20(1.10,1.31) * | P < 0.001* | 0.00 % | ||
| Li Yue, 2024 | 331/333 | 5 | RR | 1.28(1.18,1.40) | P < 0.001 | 33.00 % | ||
| Cheng Li.2019 | 122/117 | 4 | SMD | 0.33(0.08,0.59) | P = 0.01 | 24.00 % | ||
| Lin Du,2022 | / | 12 | SMD | 0.32(0.05,0.59) | p = 0.019 | 82.30 % | ||
| Level 2a (modification of attitudes) | Learning interest/motivation | Peng Yu, 2022 | 234/240 | 3 | MD | 1.48(0.93,2.04) | P < 0.001 | 31.00 % |
| Level 2b (acquisition of knowledge/skills) | Theoretical results | Zhang Yizhen, 2023 | 877/899 | 11 | MD | 5.52(5.39,5.66) | P < 0.00001 | 34.00 % |
| Tian Yi, 2023 | 806/802 | 5 | MD | 6.18(5.57,6.78) * | P < 0.00001* | 0.00 % | ||
| Deng Kegao, 2022 | 3409/3446 | 26 | MD | 4.00(3.41,4.59) | P < 0.00001 | 72.00 % | ||
| SMD | 0.59(0.49,0.68) | P < 0.00001 | 72.00 % | |||||
| Peng Yu, 2022 | 1530/1542 | 13 | SMD | 0.7(0.56,0.84) | P < 0.001 | 69.00 % | ||
| Wang Yanjie, 2020 | 588/610 | 9 | SMD | 0.63(0.42,0.84) | P < 0.00001 | 68.00 % | ||
| Huang Di, 2022 | / | 57 | SMD | 1.73(1.32,2.15) | P < 0.05 | 98.90 % | ||
| Li Yue, 2024 | / | 17 | SMD | 0.80(0.71,0.88) | P < 0.001 | 85.10 % |
©2025. The Authors