The efficacy and acceptance of online learning vs. offline learning in medical student education: a systematic review and meta-analysis
Original Article

The efficacy and acceptance of online learning vs. offline learning in medical student education: a systematic review and meta-analysis

Meng Gao1,2#, Yu Cui1,2#, Hequn Chen1,2, Huimin Zeng1,2, Zewu Zhu1,2, Xiongbing Zu1,2^

1Department of Urology, Xiangya Hospital, Central South University, Changsha, China; 2National Clinical Research Center for Geriatric Disorders, Xiangya Hospital, Central South University, Changsha, China

Contributions: (I) Conception and design: X Zu; (II) Administrative support: H Chen; (III) Provision of study materials or patients: None; (IV) Collection and assembly of data: M Gao, Y Cui; (V) Data analysis and interpretation: H Zeng, Z Zhu; (VI) Manuscript writing: All authors; (VII) Final approval of manuscript: All authors.

#These authors contributed equally to this work.

^ORCID: 0000-0002-3873-251X.

Correspondence to: Xiongbing Zu. No. 87, Xiangya Road, Changsha 410008, China. Email:

Background: To maintain the continuity of medical education during the COVID-19 epidemic, online learning has replaced traditional face-to-face learning. But the efficacy and acceptance of online learning for medical education remains unknown. This meta-analysis aimed to assess whether online learning improves learning outcomes and is more acceptable to medical students compared to offline learning.

Methods: Four databases were searched for randomized controlled trials (RCTs) and comparative studies (non-RCTs) involving online learning published from January 1900 to October 2020. A total of twenty-seven studies comparing online and offline learning in medical students were included. The Grading of Recommendations, Assessment, Development and Evaluations (GRADE) framework and Newcastle-Ottawa Scale (NOS) were used to assess the methodological quality of RCTs and non-RCTs respectively. The data of knowledge and skills scores and course satisfaction were synthesized using a random effects model for the meta-analysis.

Results: Twenty-one RCTs that were judged to be of high quality according to the GRADE framework and six non-RCTs studies which ranged from 6 to 8 (NOS) and can be considered high-quality were included in this meta-analysis. The revealed that the online learning group had significantly higher post-test scores (SMD =0.58, 95% CI: 0.25 to 0.91; P=0.0006) and pre- and post-test score gains than the offline group (SMD =1.12, 95% CI: 0.14 to 2.11, P=0.02). In addition, online education was more satisfactory to participants than the offline learning (OR: 2.02; 95% CI: 1.16 to 3.52; P=0.01). Subgroup analysis was performed on knowledge and skill scores at the post-test level. The selected factors included study outcome, study design and type, participants, course type and country. No significant factors were observed in the subgroup analysis except for course type subgroup analysis.

Discussion: Online learning in medical education could lead to higher post-test knowledge and skill scores than offline learning. It also has higher satisfaction ratings than offline education. In conclusion, online learning can be considered as a potential educational method during the COVID-19 pandemic. However, given the risk of bias of included studies such as the inclusion of non-randomized comparative studies, the conclusion should be made with cautions.

Trial Registration: CRD42020220295.

Keywords: Online learning; offline learning; meta-analysis; undergraduate medical education

Received: 20 January 2022; Accepted: 19 May 2022; Published: 30 June 2022.

doi: 10.21037/jxym-22-3


The World Health Organization classified COVID-19 as a pandemic on March 11, 2020, and the number of people infected with COVID-19 worldwide has been increasing sharply. Many educational institutions in the world, including schools and hospitals, had to suspend teaching activities. To maintain the continuity of medical education during the COVID-19 epidemic, online learning has replaced traditional face-to-face learning (1) because online technologies allow medical students to work at home between face-to-face classes and academic practices. Online learning is the act of teaching and learning through digital technology. As the core of online learning, digital technology has also become a strategy for improving the education and training of health workers (2) due to its wide application and continuous development and progress in various fields in recent years. Online learning is a general term for a variety of education approaches, concepts, methods and technologies that are constantly changing (3). It can include but is not limited to online computer-based digital education, large-scale open online courses, virtual reality (VR), virtual patients, mobile learning and basic conversion of content into a digital format (for example, PDF or HTML format for books) (3). Online learning can be used flexibly and unlimitedly with traditional methods (such as role-playing with standardized patients) so that students can practice their skills interchangeably. For educators, this educational approach can save time, effort, and space; automatically assess and record student learning progress; and obtain feedback from students (4). A series of studies have compared the effectiveness and feasibility of online and offline education for medical students, but the effect of online education is not particularly clear. Pei et al. (5) selected 16 published articles for meta-analysis and suggested that compared with offline learning, online learning has advantages in enhancing the knowledge and skills of medical students. However, He et al. (6) pointed out that online learning was not significantly different from traditional education in the effectiveness of knowledge and skills. The main reason for these inconsistent findings may be because the populations included in the two meta-analyses were different.

To provide further evidence for the efficacy and acceptance of online teaching, the current meta-analysis aims to provide new perspectives for comparing the effects of online learning and offline learning interventions. Therefore, we designed this meta-analysis to further compare the effects of online learning and offline learning for medical students including clinical, nursing and pharmacy and to identify the factors that may lead to differences in the effectiveness of the two teaching methods. We present the following article in accordance with the PRISMA reporting checklist (available at


Search strategy

We developed comprehensive search strategies for the PubMed, Web of Science, Cochrane Controlled Trials Central Registry (CENTRAL) and Embase databases to identify research related to online learning. The search time of the database was from January 1, 1990, to October 2020; 1990 was chosen as the start year of the search because before that, the use of computers was limited to basic tasks (3). The search strategies were as follows: (“online learning” OR “digital education” OR “distance education” OR “Internet-Based Learning” OR “virtual education” AND “offline learning OR traditional education OR face-to-face learning OR classroom education OR usual teaching)”. The “Related Articles” function was also used to expand the search scope and supplement the computer search by manually searching all retrieved studies, reference lists of reference articles and conference abstracts. After completing all searches, we identified all potentially relevant articles, used Endnote X9 (reference management software) without language restrictions, and deleted duplicate studies. Two independent reviewers scanned the title, abstract, and even the full text of all records to identify potentially relevant studies.

Selection of studies

This meta-analysis has been registered at PROSPERO: CRD42020220295. According to the Preferred Reporting Items for Systematic Reviews and Meta-analysis and Meta-analysis of Observational Studies in Epidemiology recommendations for study reporting (7), the selection of the article was conducted independently by two reviewers. The inclusion criteria were as follows: all available randomized controlled trials (RCTs) and retrospective comparative studies (cohort or case-control studies) that compared any form of online learning online learning with offline learning (traditional learning) to medical students from all over the world, and that had at least one of the following outcomes: knowledge and skill outcomes measured by objective assessment tools. In addition, studies on blended learning models (online + offline learning) were excluded.

In addition, the included studies should meet the following criteria in adherence to the participant, intervention, comparison and outcome (PICO) search in the field of evidence-based medicine:Participants: medical undergraduate students including clinical, nursing and pharmacy.

Interventions: online computer-based digital education, large-scale open online courses, VR, virtual patients, mobile learning and basic conversion of content into a digital format (for example, PDF or HTML format for books).Comparisons: offline learning, especially referring to face-to-face teaching in a classroom, seminars, and reading text-based documents or books only. Outcomes: knowledge and skill outcomes measured by objective assessment instruments. The mean score and standard deviations of post-test, pre- and post-test gains.

Data extraction and assessment

The full texts of the included studies were screened twice, and data from these studies were also separately extracted by two authors in a standardized format. No duplicate publications were found during the data extraction process. The main outcomes were the knowledge and skill scores at post-test. The secondary outcomes were pre- and post-test gains (improvement), retention test scores and students’ overall satisfaction with the course format.

Randomized controlled trials were judged to be of high quality according to the Grading of Recommendations, Assessment, Development and Evaluations (GRADE) framework (8), which specifies four levels of evidence: high, moderate, low, and very low quality evidence. The methodological quality of RCTs was assessed by the Cochrane risk of bias tool, which included the following domains (I) random sequence generation, (II) allocation concealment, (III) blinding of participants and personnel, (IV) blinding of outcome assessment, (V) incomplete outcome data, (VI) selective reporting, and (VII) any other source of bias (8). The Newcastle-Ottawa Scale (NOS) was used to assess the methodological quality of those nonrandomized studies (9). The scores range from 0 to 9, and the scale includes: selection of patients, comparability of the study groups, exposure (Case Control Studies) or outcome (Cohort Studies).

Statistical analysis

All meta-analyses were performed using Windows Version 5.3 Review Manager (Cochrane Collaboration, Oxford, England) and STATA 12.0 (Stata Corp LP, University Town, Texas, USA). A random effects model was used due to differences in the expected population and course diversity (10). Standard mean differences (SMDs) were used for continuous parameter data, and odds ratios (ORs) were used for the dichotomous variables, with both types of data reported with 95% confidence intervals (CIs). For some studies that only reported continuous data as the means, 95% confidence interval, range and sample size, the standard deviations were converted using the technique described by Hozo et al. (11). The statistical heterogeneity between studies was evaluated using the χ2 test, and the significance was set to P=0.1, and I2 statistics were used to evaluate statistical heterogeneity (I2≥50% indicating there is heterogeneity) (12). The Z test was used to determine the pooled effects, and a P value <0.05 indicated the presence of statistical significance (13). Data are presented as forest plots, and a funnel plot was routinely constructed to assess publication bias (14).


Results of the search

We searched a total of 2,172 records in four databases: twenty-seven studies including 2,308 participants (1,191 participants for online learning and 1,117 participants for offline learning) met the final inclusion criteria and were full-text articles (Figure 1). Seven hundred fifty-seven records were excluded after screening the title and abstract, and 241 studies were excluded after reading the full text (Figure 1).

Figure 1 Flow diagram of study selection.

Characteristics and quality of the included studies

The main characteristics of the 27 included studies, such as participants, comparison, course and outcome are shown in Table 1. Except for 6 studies (15-20) that were nonrandomized controlled studies, the remaining 21 (21-41) studies were all RCTs that were judged to be of high quality according to the GRADE framework. All articles compared posttest scores; 16 articles compared both posttest scores and pre-test and posttest score gains on the same sample, but only 5 studies had sufficient pre- and posttest score gains for meta-analysis. One study compared retention test scores 22 days after the intervention, and 7 articles compared students’ overall satisfaction with the way they attended classes. Most studies were conducted in developed countries, and five studies were conducted in developing countries. The overall risk of bias assessed according to the Cochrane risk-of-bias tool for all included RCTs is shown in Figure 2. The framework of the Cochrane bias risk tool contains the seven abovementioned areas mentioned above. Most studies described the randomization process in detail, but few articles could achieve the true blinding of participants and outcome assessment. Only Phadtare et al. (23) achieved participant blinding by placing group assignments in sealed envelopes and revealing after participants had signed informed consent and Porter et al. (24) performed lecturing teacher blinding. For the rest of the nonrandomized studies, their scores ranged from 6 to 8 on the NOS, which can be considered high-quality. The assessments of detail were shown in Table 2.

Table 1

Characteristics of included studies

Author, year, country Comparison Samples (T/C, n) Participants Course Study design Assessment strategies Outcome Design [score]
Brettle et al., 2013, UK Online vs. face 70 (35/35) Undergraduate nurse Information literacy skill Pretest/post-test Skill test Pre- and post-session search skills score, follow up skill score RCT
Hu et al., 2016, USA 3D computer vs. text 100 (49/51) Medical students Laryngeal anatomy Post-test only Knowledge test Laryngeal anatomy test score and instructional materials motivation survey RCT
Phadtare et al., 2009, USA Online vs. standard 48 (24/24) Second- and third-year medical student Scientific writing Post-test only Skill test Manuscript quality and self-reported participant satisfaction RCT
Porter et al., 2014, USA Online vs. classroom 140 (71/69) Second- and third-year medical student Immunization course Post-test only Knowledge test Grades and evaluation and assessment of course RCT
Subramanian et al., 2012, USA Software vs. traditional 30 (15/15) third-year medical student Arrhythmia Pretest/post-test Knowledge test Post-test score, improvement and long-term retention RCT
Bjarne et al., 2013, Denmark e-learning vs. face 42 (21/21) Anesthesiology nurse Respiratory and pulmonary physiology Pretest/post-test Knowledge test Pre- and post-test score and improvement RCT
Bowdish et al., 2003, USA Virtual vs. text 112 (56/56) First-year medical students Human physiology Post-test only Knowledge test Teaching and Learning environment Questionnaire score and student achievement Quasi-experimental [8]
Chittenden et al., 2013 Web vs. written 74 (41/33) Third-year medical student Code status discussions Post-test only Skill test Student performance in conducting code status discussions and communication skills RCT
Soloman et al., 2014 Digital vs. live 29 (17/12) Third-year medical student CAD and renal failure Post-test only Knowledge test Exam score and feedback on the digital lecture format RCT
Moazami et al., 2014 Virtual vs. traditional 35 (15/20) Dental medical students Rotary instrumentation of root canals Post-test only Knowledge test Knowledge acquisition and its retention RCT
Alemánr et al., 2011 Computer vs. convention 41 (15/26) Second-year nurse student Medical-surgical nursing Pretest/post-test Skills and knowledge test Pre- and post-test score, evaluation of the students’ experience RCT
Portero et al., 2013 Virtual vs. convention 114 (71/43) Third-year medical student Radiology Post-test only Knowledge test Final oral examination and evaluation on image interpretation Case control [7]
Pusponegoro et al., 2015 Online vs. live 75 (39/36) Fifth-year medical student Gross motor screening method in infants Pretest/post-test Knowledge test Pre- and post-test score, improvement and satisfaction RCT
Bhatti et al., 2011 e-learning vs. face 148 (75/73) Third-year medical student Hemorrhoids Pretest/post-test Knowledge test Pre and post-test score, improvement and usefulness of website RCT
Dennis et al., 2003 Online vs. face 34 (17/17) Second-year medical student Problem-based learning Post-test only Knowledge test Learning outcomes, time on-task and generation of LIs RCT
Yeung et al., 2012 Computer vs. tradition 78 (43/35) Second-year medical student Cranial nerve anatomy Post-test only Knowledge test Post-test score and evaluation of participants’ experience RCT
Kaltman et al., 2018 Video vs. usual 99 (60/39) First-year medical student Communication Post-test only Skill test Simulation experience, OSCE communication behaviors and self-efficacy RCT
Morente et al., 2013 e-learning vs. tradition 73 (30/43) Undergraduate nursing student Pressure ulcer Pretest/post-test Knowledge test Pre- and post-test score and improvement RCT
Peine et al., 2016 e-learning vs. lecture 116 (61/55) Third-year medical student Modernized medical curricula Pretest/post-test Knowledge test Pre- and post-test score and self-assessment RCT
Nicklen et al., 2017 Online vs. face 38 (19/19) Third-year medical student Case‑based learning Post-test only Knowledge test Learning and self‑assessed perception of learning, satisfaction RCT
Clement et al., 2012 DVD vs. lecture 130 (71/59) Graduate nursing student Stigma and mental health Post-test only Knowledge test Knowledge, attitudes (cognitive and emotional) and behaviour RCT
Chao et al., 2012 Online vs. Lecture 167 (111/56) Fourth-year medical student Delirium Pretest/post-test Skill test Pre- and post-test score and improvement Case control [6]
Farahmand et al., 2016 Distance vs. Tradition 120 (60/60) Senior medical students Initial assessment of trauma Post-test only Knowledge and skill test Post-test score Quasi-experimental [8]
Taradi et al., 2005 WPBL vs. face 121 (37/84) Second-year medical student Acid-base physiology Post-test only Knowledge test Test scores and satisfaction survey results Case control [7]
Assadi et al., 2003 Video vs. traditional 81 (41/40) Undergraduate intern Basic life support instruction Pretest/post-test Knowledge and skill test Pre- and post-test score and satisfaction Prospective research [7]
Raupach et al., 2009 WPBL vs. face 143 (72/71) Fourth-year medical student Clinical reasoning skills Post-test only Knowledge test Post-test score, student activity and evaluation RCT
Alnabelsi et al., 2015 e-learning vs. face 50 (25/25) Fourth- and fifth-year medical student ENT Post-test only Knowledge test Pre- and post-test score, improvement and satisfaction RCT

T/C, test group/control group; RCT, randomized controlled trial; LIs, a key product that facilitates self-directed learning during the tutorial process; ENT, Ear, Nose and Throat.

Figure 2 The overall risk of bias for included RCTs. RCTs, randomized controlled trials.

Table 2

Risk of bias assessment for included non-RCT trials

Study Selection Comparability Exposure/outcome Score
a b c d e f g h i
Bowdish et al., 2003 8
Portero et al., 2013 7
Chao et al., 2012 6
Farahmand et al., 2016 8
Taradi et al., 2005, 7
Assadi et al., 2003 7

a, adequate case definition; b, representativeness of the cases; c, selection of controls; d, definition of controls; e, study controls for the most important factor; f, study controls for any additional factor; g, ascertainment of exposure; h, some methods of ascertainment for cases and controls; I, non-response rate. ★, a qualified identification, no special instructions are required. RCT, randomized controlled trial.


Knowledge and skill score at the post-test level

Data on knowledge or skill scores were available for all 27 studies, with a total sample size of 2,308 reported. The pooled results showed that the online learning group had significantly higher scores than the offline group (SMD =0.58, 95% CI, 0.25 to 0.91; P=0.0006) (Figure 3).

Figure 3 Forest plot for knowledge and skill score at the post-test level.

Pre- and post-test score gains

Five studies (20,25,26,31,41) including 278 students provided data on pre- and post-test score gains. There was a significant difference in the pre- and post-test score gains between the two groups (SMD =1.12, 95% CI, 0.14 to 2.11, P=0.02) (Figure 4). High heterogeneity was found, and a random-effects model was used (I2=92%).

Figure 4 Forest plot for pre- and post-test score gains.

Overall satisfaction

Overall satisfaction was reported in 7 eligible articles, but only three studies had suitable data for meta-analysis. A meta-analysis of these 3 studies (24,31,41) showed that online education was more satisfactory to participants than offline learning (OR: 2.02; 95% CI, 1.16 to 3.52; P=0.01). There was a moderate degree of heterogeneity, and a fixed effects model was used (P=0.12, I2=53%) (Figure 5). A summary of the outcomes and the results of the meta-analysis are shown in Table 3.

Figure 5 Forest plot for overall satisfaction at the post-test level.

Table 3

Results of meta-analysis comparison of online and offline learning

Outcome Studies No. Online group No. Offline group No. SMD/OR (95% CI) P value Study heterogeneity
χ2 df I2 (%) P value
Knowledge and skills (post-test) 27 1,191 1,117 0.58 (0.25 to 0.91) 0.0006 354.22 26 93 <0.00001
Knowledge gains (pretest/post-test) 5 141 137 1.12 (0.14 to 2.11) 0.02 50.04 4 92 <0.00001
Overall satisfaction 3 133 126 2.02 (1.16. to 3.52) 0.01 4.27 2 53 0.12

SMD/OR, standard mean deviance/odds ratio; df, degrees of freedom; CI, confidence interval.

Subgroup analysis

Subgroup analysis was performed on knowledge and skill scores at the post-test level (Table 4). The selected factors included study outcome, study design and type, participants, course type and country. There was a significant difference in course type subgroup analysis (Figure 6) compared with the original analysis (P=0.006), foundation course group analysis (SMD =0.07, 95% CI: −0.11 to 0.25, P=0.44) and other course group analysis (SMD =0.09, 95% CI: −1.10 to 1.28, P=0.88) were different from clinical course group (SMD =0.86, 95% CI: 0.41 to 1.31, P=0.0002) and original analysis (SMD =0.58, 95% CI: 0.25 to 0.91, P=0.0006). For the other selected factor subgroups, there was no significant difference between these subgroups (Figures S1-S5).

Table 4

Subgroup analyses of online vs. offline education on knowledge and skills acquisitions at the post-test levels

Subgroup Studies No. Participants No. SMD/OR (95% CI) Study heterogeneity P value
χ2 df I2 (%) P value
All intervention 27 2,308 0.58 (0.25 to 0.91) 354.22 26 93 <0.00001 0.0006
Study outcome 0.76
   Knowledge 23 1,928 0.63 (0.26 to 1.00) 314.58 22 93 <0.00001 0.001
   Skills 5 444 0.77 (−0.05 to 1.59) 65.60 4 94 <0.00001 0.07
Study design 0.46
   Post-test only 16 1,415 0.47 (0.03 to 0.92) 232.13 15 94 <0.00001 0.04
   Pretest/post-test 11 893 0.73 (0.23 to 1.23) 111.68 10 91 <0.00001 0.004
Study type 0.09
   RCT 21 1,593 0.35 (0.05 to 0.66) 161.68 20 88 <0.00001 0.02
   Non-RCT 6 715 1.27 (0.25 to 2.28) 180.53 5 97 <0.00001 0.01
Participants 0.63
   Medical students 20 1,764 0.64 (0.23 to 1.04) 286.80 19 93 <0.00001 0.002
   Nurse students 5 356 0.27 (−0.43 to 0.98) 39.86 4 90 <0.00001 0.45
   Others 2 188 0.93 (−0.94 to 2.80) 23.82 1 96 <0.00001 0.33
Country 0.14
   Developed 22 1,876 0.34 (0.07 to 0.61) 163.03 21 87 <0.00001 0.01
   Developing 5 432 1.51 (−0.01 to 3.03) 173.46 4 98 <0.00001 0.05
Course type 0.006
   Clinical 18 1,586 0.86 (0.41 to 1.31) 280.74 17 94 <0.00001 0.0002
   Foundation 5 472 0.07 (−0.11 to 0.25) 0.4 4 0 0.98 0.44
   Other 4 250 0.09 (−1.10 to 1.28) 49.39 3 93 <0.00001 0.88

SMD/OR, standard mean deviance/odds ratio; df, degrees of freedom; CI, confidence interval.

Figure 6 Subgroup analysis of online vs. offline education on knowledge and skills acquisitions for course type at the post-test levels.

Publication bias

The research funnel chart (Figure 7) included in the meta-analysis was used to assess the publication bias in the knowledge and skill score at the post-test level. Most studies lay inside the 95% CIs, with a small number of studies having an uneven distribution, indicating that there was slight asymmetry.

Figure 7 Funnel plots illustrating meta-analysis of knowledge and skills acquisitions.

Sensitivity analysis

Twenty-one RCTs and 6 CCTs that scored six or more on the Newcastle-Ottawa scale were included in the sensitivity analysis. Leave-one-out cross validation was used in the sensitivity analysis to assess the stability of the meta-analysis results. There was no change in the significance of any of the outcomes except for overall satisfaction, which indicated that these meta-results were stable (Figures 8,9). When removing the article reported by Porter et al. (24), the result was no longer statistically significant (Figure 10) compared with the original meta-analysis (OR: 1.13; 95% CI: 0.51 to 2.53; P=0.77). This may be caused by a small sample and the forms of online learning and courses of learning were different for each study, there was heterogeneity between the included studies, which may influence the results of the meta-analysis.

Figure 8 Sensitivity analysis of knowledge and skill score at the post-test level.
Figure 9 Sensitivity analysis of pre- and post-test score gains.
Figure 10 Sensitivity analysis of overall satisfaction.


This meta-analysis of 21 RCTs and 6 CCTs including 2,308 students comparing the efficacy of online learning and offline learning showed that online learning was more effective for undergraduate medical students on post-test scores, pre- and post-test score improvement and overall satisfaction. No factors that significantly impacted the overall results were observed through subgroup analysis. Because the experimental design of the included articles was very different in participants, courses, examination format, and outcome measurement methods, there was considerable heterogeneity among the included studies. However, our sensitivity analysis showed that the results of the meta-analysis were robust.

The greatest concerns for medical students’ online learning were knowledge acquisition and skill training. It is well known that undergraduate medical courses mainly focus on basic knowledge and skills. In this review, posttest knowledge and skill scores were reported differently in each included study. Therefore, we compared these two outcomes between the online and offline groups and found that the posttest scores of the online learning group were significantly higher. Considering prior knowledge or skill levels, the difference between the pre-intervention and post-intervention test scores for each student was calculated and designated as “improvement”. The pooled data of improvement included five studies that also showed that online learning students had a significantly higher improvement score. Subramanian et al. (25) reported that the average improvement score of the online group was nearly three times that of the offline group and demonstrated that not only was online learning an effective way of learning for medical students compared with the offline format, but it can also promote long-term retention. In most of the studies we included, multiple-choice questions (MCQs) were used as the posttest. The MCQ can not only objectively evaluate students’ test scores but also predict objective structured clinical examination (OSCE) scores, which in turn is a powerful predictor of clinical performance (42). The reasons why online learning works better are as follows. First, students can learn about medical knowledge and skills without participating in traditional classroom learning because they can access the information as many times as needed. Second, in addition to the same teaching materials used in online learning, good educational cases, such as representative patients, were also provided. This can prevent certain patients from being suitable for students due to ethical considerations, and there is no need to consider patients who refuse student participation in their care (25,43). In addition, as a novel instructional method, online learning can simulate and practice different clinical situations (experiential learning) (44). However, online learning also has some shortcomings and limitations, and technical problems have made students feel frustrated, so they need technical support related to learning (30). Hence, most of the studies we included were conducted in developed countries, and only five articles (18-20,29,31) were performed in developing countries. Additional problems included having no teacher present, learner isolation, and a lack of peer support and competition (45). These concerns are exacerbated when online methods are used to develop interpersonal and high-level clinical skills, where contextual clinical reasoning is the basis of competence (46).

In addition, although the included studies included medical students of all grades, the knowledge and skills taught in these studies actually only cover a small part of the learning objectives in medical education. Therefore, it is difficult to say that online learning is better than offline learning for topics that have yet to be studied. For online learning mainly composed of static and non-interactive learning resources, these learning resources are similar to offline learning to a large extent; usually, no significant difference was found when compared to offline learning (5). A study conducted by Nesterowicz et al. (47) reported that 92% of the subjects believed that online-learning was effective and that the subject of the course was the most important aspect.

In terms of subjective evaluation, contemporary medical students grew up in the Internet era. They are accustomed to the constant stimulation of e-mail, text, and social media, and their experiences affect their behaviour in the classroom. They prefer to listen to podcasts at twice the speed instead of attending lectures to use their time more effectively. They would rather choose a self-paced online training module learning method than using a rigorous 12-week course (22). Our meta-analysis of three studies also showed that the online learning group had a higher rate of overall satisfaction than the offline learning group. In addition to these three studies, Taradi (19) and Phadtare et al. (23) gained student satisfaction by surveys and showed that there was a statistically significant difference in the overall satisfaction with the course between the two groups; the online group had a higher overall satisfaction score. However, Raupach et al. (40) found that the overall satisfaction score with an online module was low; Nicklen et al. (38) also surveyed student satisfaction and showed that 63 percent of those in the intervention group reported a perception that online learning negatively impacted their learning. This variation in student satisfaction may be a result of the different online learning methods, and more similar studies are needed for further confirmation. When students encounter difficulties in using the online learning system, they need technical assistance and learn many things before they are able to use the system, which consumes their learning time and energy.

Currently, the number of people infected with COVID-19 disease is still rising sharply worldwide and there is no vaccine that can effectively prevent the infection of the virus. The global educational centre had not to force to close their classrooms and quickly make changes in medical education to ensure that all students still receive the absolute best level of education possible (48). Moreover, the world is changing, and the causes of education interruptions are not limited to epidemics; wars, regional conflicts, and various types of natural disasters are issues that should be kept on the future agenda as potential sources of interruption (49). Online learning has been the best choice to maintain regular teaching and learning (1). This review further confirms that online learning is more effective than offline learning in undergraduate medical education.

Despite the valuable conclusions drawn, the meta-analysis still has some limitations. First, our study included controlled clinical trials (CCTs), which may not be adequately powered. Second, educators who achieve good results with online learning tend to publish their results, which may result in potential publication bias. Third, because the forms of online learning and courses of learning were different for each study, there was heterogeneity between the included studies, which may influence the results of the meta-analysis. Random effects model can only address statistical heterogeneity but the heterogeneity caused by different ways of online learning cannot be addressed via statistical analysis. Last, the included studies in our review were not conducted under the circumstance of the COVID-19 pandemic. Therefore, it is difficult to conclude that online learning is more effective than offline learning for those courses influenced by COVID-19. More comparative studies conducted in the context of the COVID-19 pandemic are needed.


In summary, our meta-analysis demonstrates that online learning methods in medical education could achieve higher knowledge and skill scores at the posttest level than offline learning methods. In addition, it also has higher satisfaction ratings than offline education, indicating that contemporary medical students prefer this education mode. Through subgroup analysis, no significant factors were observed except the subject of the course, which indicates that not all courses are suitable for online learning.


The authors are grateful to the staff of Xiangya Hospital Central South University and all those who actively participated in this study.

Funding: The study was supported by Education Reform Research Project of Central South University (No. 2022JGB051).


Reporting Checklist: The authors have completed the PRISMA reporting checklist. Available at

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See:


  1. Al-Balas M, Al-Balas HI, Jaber HM, et al. Distance learning in clinical medical education amid COVID-19 pandemic in Jordan: current situation, challenges, and perspectives. BMC Med Educ 2020;20:341. [Crossref] [PubMed]
  2. Crisp N, Gawanas B, Sharp I, et al. Training the health workforce: scaling up, saving lives. Lancet 2008;371:689-91. [Crossref] [PubMed]
  3. Car J, Carlstedt-Duke J, Tudor Car L, et al. Digital Education in Health Professions: The Need for Overarching Evidence Synthesis. J Med Internet Res 2019;21:e12913. [Crossref] [PubMed]
  4. Choules AP. The use of elearning in medical education: a review of the current situation. Postgrad Med J 2007;83:212-6. [Crossref] [PubMed]
  5. Pei L, Wu H. Does online learning work better than offline learning in undergraduate medical education? A systematic review and meta-analysis. Med Educ Online 2019;24:1666538. [Crossref] [PubMed]
  6. He L, Yang N, Xu L, et al. Synchronous distance education vs traditional education for health science students: A systematic review and meta-analysis. Med Educ 2021;55:293-308. [Crossref] [PubMed]
  7. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration. BMJ 2009;339:b2700. [Crossref] [PubMed]
  8. Shuster JJ. Review: Cochrane handbook for systematic reviews for interventions, Version 5.1.0, published 3/2011. Julian P.T. Higgins and Sally Green, editors. Research Synthesis Methods 2011;2:126-30.
  9. Stang A. Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. Eur J Epidemiol 2010;25:603-5. [Crossref] [PubMed]
  10. Higgins JP, Thompson SG, Deeks JJ, et al. Measuring inconsistency in meta-analyses. BMJ 2003;327:557-60. [Crossref] [PubMed]
  11. Hozo SP, Djulbegovic B, Hozo I. Estimating the mean and variance from the median, range, and the size of a sample. BMC Med Res Methodol 2005;5:13. [Crossref] [PubMed]
  12. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med 2002;21:1539-58. [Crossref] [PubMed]
  13. Zaykin DV. Optimally weighted Z-test is a powerful method for combining probabilities in meta-analysis. J Evol Biol 2011;24:1836-41. [Crossref] [PubMed]
  14. Egger M, Davey Smith G, Schneider M, et al. Bias in meta-analysis detected by a simple, graphical test. BMJ 1997;315:629-34. [Crossref] [PubMed]
  15. Bowdish BE, Chauvin SW, Kreisman N, et al. GTravels towards Problem Based Learning in Medical Education (VPBL). Instructional Science 2003;31:231-53. [Crossref]
  16. Sendra-Portero F, Torales-Chaparro OE, Ruiz-Gómez MJ, et al. A pilot study to evaluate the use of virtual lectures for undergraduate radiology teaching. Eur J Radiol 2013;82:888-93. [Crossref] [PubMed]
  17. Chao SH, Brett B, Wiecha JM, et al. Use of an online curriculum to teach delirium to fourth-year medical students: a comparison with lecture format. J Am Geriatr Soc 2012;60:1328-32. [Crossref] [PubMed]
  18. Farahmand S, Jalili E, Arbab M, et al. Distance Learning Can Be as Effective as Traditional Learning for Medical Students in the Initial Assessment of Trauma Patients. Acta Med Iran 2016;54:600-4. [PubMed]
  19. Taradi SK, Taradi M, Radic K, et al. Blending problem-based learning with Web technology positively impacts student learning outcomes in acid-base physiology. Adv Physiol Educ 2005;29:35-9. [Crossref] [PubMed]
  20. Assadi T, Mofidi M, Rezai M, et al. The Comparison between two Methods of Basic Life Support Instruction: Video Self-Instruction versus Traditional Method. Hong Kong Journal of Emergency Medicine 2015;22:291-6. [Crossref]
  21. Brettle A, Raynor M. Developing information literacy skills in pre-registration nurses: an experimental study of teaching methods. Nurse Educ Today 2013;33:103-9. [Crossref] [PubMed]
  22. Hu A, Shewokis PA, Ting K, et al. Motivation in computer-assisted instruction. Laryngoscope 2016;126:S5-S13. [Crossref] [PubMed]
  23. Phadtare A, Bahmani A, Shah A, et al. Scientific writing: a randomized controlled trial comparing standard and on-line instruction. BMC Med Educ 2009;9:27. [Crossref] [PubMed]
  24. Porter AL, Pitterle ME, Hayney MS. Comparison of online versus classroom delivery of an immunization elective course. Am J Pharm Educ 2014;78:96. [Crossref] [PubMed]
  25. Subramanian A, Timberlake M, Mittakanti H, et al. Novel educational approach for medical students: improved retention rates using interactive medical software compared with traditional lecture-based format. J Surg Educ 2012;69:449-52. [Crossref] [PubMed]
  26. Worm BS. Learning from simple ebooks, online cases or classroom teaching when acquiring complex knowledge. A randomized controlled trial in respiratory physiology and pulmonology. PLoS One 2013;8:e73336. [Crossref] [PubMed]
  27. Chittenden EH, Anderson WG, Lai CJ, et al. An evaluation of interactive web-based curricula for teaching code status discussions. J Palliat Med 2013;16:1070-3. [Crossref] [PubMed]
  28. Solomon DJ, Ferenchick GS, Laird-Fick HS, et al. A randomized trial comparing digital and live lecture formats ISRCTN40455708. BMC Med Educ 2004;4:27. [Crossref] [PubMed]
  29. Moazami F, Bahrampour E, Azar MR, et al. Comparing two methods of education (virtual versus traditional) on learning of Iranian dental students: a post-test only design study. BMC Med Educ 2014;14:45. [Crossref] [PubMed]
  30. Fernández Alemán JL, Carrillo de Gea JM, Rodríguez Mondéjar JJ. Effects of competitive computer-assisted learning versus conventional teaching methods on the acquisition and retention of knowledge in medical surgical nursing students. Nurse Educ Today 2011;31:866-71. [Crossref] [PubMed]
  31. Pusponegoro HD, Soebadi A, Surya R. Web-Based Versus Conventional Training for Medical Students on Infant Gross Motor Screening. Telemed J E Health 2015;21:992-7. [Crossref] [PubMed]
  32. Bhatti I, Jones K, Richardson L, et al. E-learning vs lecture: which is the best approach to surgical teaching? Colorectal Dis 2011;13:459-62. [Crossref] [PubMed]
  33. Dennis JK. Problem-based learning in online vs. face-to-face environments. Educ Health (Abingdon) 2003;16:198-209. [Crossref] [PubMed]
  34. Yeung JC, Fung K, Wilson TD. Prospective evaluation of a web-based three-dimensional cranial nerve simulation. J Otolaryngol Head Neck Surg 2012;41:426-36. [PubMed]
  35. Kaltman S, Talisman N, Pennestri S, et al. Using Technology to Enhance Teaching of Patient-Centered Interviewing for Early Medical Students. Simul Healthc 2018;13:188-94. [Crossref] [PubMed]
  36. Morente L, Morales-Asencio JM, Veredas FJ. Effectiveness of an e-learning tool for education on pressure ulcer evaluation. J Clin Nurs 2014;23:2043-52. [Crossref] [PubMed]
  37. Peine A, Kabino K, Spreckelsen C. Self-directed learning can outperform direct instruction in the course of a modern German medical curriculum - results of a mixed methods trial. BMC Med Educ 2016;16:158. [Crossref] [PubMed]
  38. Nicklen P, Keating JL, Paynter S, et al. Remote-online case-based learning: A comparison of remote-online and face-to-face, case-based learning - a randomized controlled trial. Educ Health (Abingdon) 2016;29:195-202. [PubMed]
  39. Clement S, van Nieuwenhuizen A, Kassam A, et al. Filmed v. live social contact interventions to reduce stigma: randomised controlled trial. Br J Psychiatry 2012;201:57-64. [Crossref] [PubMed]
  40. Raupach T, Muenscher C, Anders S, et al. Web-based collaborative training of clinical reasoning: a randomized trial. Med Teach 2009;31:e431-7. [Crossref] [PubMed]
  41. Alnabelsi T, Al-Hussaini A, Owens D. Comparison of traditional face-to-face teaching with synchronous e-learning in otolaryngology emergencies teaching to medical undergraduates: a randomised controlled trial. Eur Arch Otorhinolaryngol 2015;272:759-63. [Crossref] [PubMed]
  42. Wilkinson TJ, Frampton CM. Comprehensive undergraduate medical assessments improve prediction of clinical performance. Med Educ 2004;38:1111-6. [Crossref] [PubMed]
  43. Zary N, Johnson G, Boberg J, et al. Development, implementation and pilot evaluation of a Web-based Virtual Patient Case Simulation environment--Web-SP. BMC Med Educ 2006;6:10. [Crossref] [PubMed]
  44. Ruiz JG, Smith M, van Zuilen MH, et al. The educational impact of a computer-based training tutorial on dementia in long term care for licensed practice nursing students. Gerontol Geriatr Educ 2006;26:67-79. [Crossref] [PubMed]
  45. Vaona A, Banzi R, Kwag KH, et al. E-learning for health professionals. Cochrane Database Syst Rev 2018;1:CD011736. [PubMed]
  46. Grimmer-Somers K, Milanese S, Chipcase L. Research into Best Practices in e-Learning for Allied Health clinical education and training. Brisbane: Clinical Education and Training Queensland, 2011.
  47. Nesterowicz K, Librowski T, Edelbring S. Validating e-learning in continuing pharmacy education: user acceptance and knowledge change. BMC Med Educ 2014;14:33. [Crossref] [PubMed]
  48. Theoret C, Ming X. Our education, our concerns: The impact on medical student education of COVID-19. Med Educ 2020;54:591-2. [Crossref] [PubMed]
  49. Bozkurt A, Sharma RC. Emergency remote teaching in a time of global crisis due to CoronaVirus pandemic. Asian Journal of Distance Education 2020;15:1-4.
doi: 10.21037/jxym-22-3
Cite this article as: Gao M, Cui Y, Chen H, Zeng H, Zhu Z, Zu X. The efficacy and acceptance of online learning vs. offline learning in medical student education: a systematic review and meta-analysis. J Xiangya Med 2022;7:13.

Download Citation