Table 1 Summary of characteristics of included articles

Alejandro-Salinas (2022)

(20)

Peru.

Cross -sectional.

281; medical students.

PTSD evaluated with the Impact of Event Scale - Revised (IES-R) a 22-item self-administered questionnaire.

Data exported to a Microsoft Excel spreadsheet from data obtained from Google Forms, then a statistical package STATA v15.0 (StataCorp, TX, USA) used for analysis.

Ali (2020) (21)

Pakistan.

Cross-sectional.

182; First, second- and third-year medical students.

Depression, Anxiety and Stress Scale (DASS).

Self-administered electronic questionnaires. Sample size calculated through non- probability consecutive sampling (50% of selected population). Frequency and percentages from the DASS were calculated to correlate the effects on daily routine. Chi square utilised to calculate association across different year groups. Pilot study conducted. No use of a control group.

Arima (2020) (8)

Japan.

Cross-sectional.

571; All medical students.

K-6 scale for psychological distress. Rosenberg Self Esteem Scale (RSES). General Self Efficacy Scale (GSES).

Self-administered electronic questionnaires. No sample size calculation. Demographics and variables were utilised to assess the determinant factor for psychological distress through logistic regression. Regression analysis utilised to evaluate RSES and GSES scores. Validity and reliability pertaining to measures of distress were assessed over four weeks prior to survey administration. No use of a control group.

Banerjee (2021)

[22]

Mauritius.

Cross -sectional.

663; medical students from 1st to 10th semester.

Questionnaire designed following a literature review measuring guilt, depression, time management, focus, sleep, comprehension, fear and motivation.

Eight items on questionnaire were used to assess the psychological impact on medical students due to COVID-19. The questionnaire was validated by five subject experts and initially tested via the use of a pilot study with 10 students. The reliability of the questionnaire was ascertained by Cronbach’s alpha.

Bolatov (2020) [18]

Kazakhstan.

Cross-sectional.

798 (intervention group). 619 (control group). First to fifth year medical students.

The Patient Health Questionnaire-9 (PHQ-9) scale. The Generalized Anxiety Disorder, the 7-item (GAD-7) Scale. The Patient Health Questionnaire-15 (PHQ-15) scale. Fear of COVID-19 was assessed using a 5-point adapted Snell’s questionnaire. The Copenhagen Burnout Inventory (CBI-S).

Self-administered electronic questionnaires. No sample size calculation. Burnout syndrome, depression, anxiety, somatic symptoms, and satisfaction with academic performance were analysed. The chi-squared test or independent sample t test was utilised to evaluate the differences between variables. In comparing results within one sample, ANOVA analysis and the Bonferroni post hoc test was utilised. In assessing independent variable associations, a logistic regression analysis was carried out. Control group included.

Dworkin (2021)

[16]

13 Countries.

Participants were asked to take 3–4 photographs over a two-week period depicting daily life during the pandemic, and then write brief reflections about each photograph.

26; Medical students.

A directed content analysis approach.

Qualitative study of photographic reflection taken.

Ferreira (2021)

[23]

Brazil.

Cross-sectional.

216; medical students from 9th to 12th semester.

Psychiatric diagnosis, use of psychotropic drugs, and legal or illegal psychoactive substances.

Questionnaires consisted of closed-answer questions (multiple-choice, single-answer, dichotomous-answer), matrix (Likert scale), and open-answer questions.

Frajerman ( 2022)

[24]

France.

Online national cross-sectional study

2623; Medical students.

Used Kessler’s K6, a validated scale in 6 items to evaluate psychological distress.

Statistical significance was tested in bivariate analyses using the chi-square test. Logistic regression models were performed for statistically significant associations in bivariate analyses for the K6 scale.

Kalok (2020)

[6]

Malaysia.

Cross-sectional.

627; Medical clinical year students.

Depression, Anxiety and Stress Scale-21 (DASS 21). Short Warwick Edinburgh Mental Well-Being Scale (SWEMWBS). Perception of the Effect of MCO* on Self-Wellbeing. Source of Social Support.

*MCO: Movement Control Order- Government instigated partial lockdown.

Self-administered electronic questionnaires. Sample size was calculated based on precision, confidence interval and dropout figures. Questionnaire Validation of Social Support and Perception of MCO on Self-Wellbeing performed. DASS 21, SWEMBS, and MCO effects were analysed for normality using the Kolmogorov–Smirnov test. Internal consistency for the effect of the MCO was evaluated using a reliability test. Correlations between depression and mental wellbeing, anxiety, stress, and the effects of the MCO were analysed using Pearson’s correlation whilst significant associations were highlighted through univariate analysis where statistically significant variables were utilised through multivariate logistic regression analyses. Student’s t-tests and one-way ANOVA was utilised to analyse the variability between groups for demographic variables and social support. No use of a control group.

Korobchansky (2021)

[25]

Ukraine.

Cross-sectional.

273; Medical students.

Examining factors including students’ lifestyle, day regime, features of organising distance learning and changes in physical and psycho-emotional state.

Self-administered electronic questionnaires. No sample size calculation. Assessing impact of adverse factors on health status and lifestyle of students through examining factors including students’ lifestyle, day regime, features of organising distance learning and changes in physical and psycho-emotional state. No use of a control group.

Kosendiak (2021)

[26]

Poland.

Cross-sectional study.

2920; Medical students.

Coping Orientation to Problems Experienced questionnaire (Mini-COPE), WHO’s Alcohol Use Disorders Identification Test, Fagerström Test for Nicotine Dependence.

Chi-square test was used with Bonferroni correction. To test statistically significant differences between groups the non-parametric Skillings–Mack test with the post-hoc Dunn’s test (for variables not meeting the conditions of normal distribution) was used. Spearman’s rank order test was used for correlation analysis.

Kosendiak (2022)

[27]

Poland.

Cross-sectional study.

225; Medical students in second year

Response to questions on alcohol intake, smoking and sleeping duration.

Chi-square test of independence was performed to assess statistically significant differences between expected and observed values contained in a contingency table.

Leroy (2021)

[28]

France.

Cross-sectional study.

4193; Medical students.

The mental health outcomes evaluated were suicidal thoughts, severe self-reported distress (as assessed by the Impact of Events Scale-Revised), stress (Perceived Stress Scale), anxiety (State-Trait Anxiety Inventory, State subscale), and depression (Beck Depression Inventory).

Multivariable logistic regression analyses were performed to test the association between the type of university studies (healthcare studies: medical and non-medical, and non-healthcare studies) and poor mental health outcomes.

Meo (2020) [7]

Saudi Arabia.

Cross-sectional.

530; First to fifth year medical students.

Psychological wellbeing, stress- allied queries and learning behaviour were analysed through Five Point Likert Scale.

Self-administered electronic questionnaires. Sample size obtained using simple random sampling and calculated using a power formula. Pilot study conducted. No use of a control group.

Miskulin (2020) [29]

Brazil.

Cross-sectional.

347; First to sixth year medical students.

Hospital Anxiety and Depression Scale (HADS).

Self-administered electronic questionnaires. No sample size calculation. Chi-square test and Mann-Whitney test were used to categorise and make comparisons of continuous variables respectively. Correlations were derived using Spearman correlation test. To evaluate the influence of year class and location of students during quarantine, binary logistic regression was utilised. No use of a control group.

Peng (2020)

[30]

China.

Cross-sectional study.

430; Medical students (442 non-medical students.

Questions on attitude toward COVID-19 e.g. “Do you hope the outbreak to stop quickly so you can return to school soon” and “Do you think you will be more capable to endure such public health emergence?“

Questionnaire study.

Pereira (2020) [19]

Brazil.

Cross-sectional.

860; First to fourth year medical students.

Prevalence of CMDs* analysed through Self-Reporting Questionnaire (SQR-20) designed by WHO** to screen for emotional distress. *CMDs: Common mental disorders**WHO: World Health Organisation

Self-administered electronic questionnaires. No sample size calculation. Groups were analysed through utilising the Chi-squared test and Kruskall-Wallis test for categorical variables and continuous variables respectively. Control group included.

Pravinraj (2022)

[31]

India.

Cross-sectional.

204; Medical students (Prefinal and final year).

Depression Anxiety Stress Scale (DASS 21).

Data was collected in preformed, self-administered, pretested questionnaires and Spearman`s correlation, Ordinal logistics regression was applied to find the predictors of Depression, Anxiety and Stress.

Qanash (2020) [4]

Saudi Arabia.

Cross-sectional.

362; First to sixth year medical students.

Four-Item Patient Health Questionnaire for Anxiety and Depression (PHQ-4).

Self-administered electronic questionnaires. Sample size was calculated using non- probability convenient sampling. Two Sample t-test used for continuous variables that had a normal distribution. Welch Two Sample t-test used when both groups had unequal variance. Wilcoxon rank sum test was utilised for continuous variables that were not normally distributed. Chi-square or Fisher’s exact test were utilised to analyse categorical variables. Kruskal-Wallis test was utilised for ordinal attributes. Excluded students with personal history of psychological illness. Pilot study was conducted. No use of a control group.

Rolland

(2022)

[32]

France.

Cross-sectional study.

1712; Medical students (also recruited non-medical students.

Kessler’s K6 scale

Descriptive information was provided as percentages. Statistical significance was tested in bivariate analyses using the Chi2 test and Fisher’s exact test to compare between groups’ prevalence. Subsequently, logistic regression models adjusted for age and sex were performed for statistically significant associations in bivariate analyses.

Ross (2021)

[33]

South Africa.

Cross-sectional study.

256; 5th year medical students

Question about being in a good ‘headspace’ to engage in online learning.

Data were downloaded from Google forms onto an Excel spreadsheet, cleaned and analysed descriptively by using Statistical Package for Social Sciences (SPSS) version 27 to determine central tendency, variation and associations and calculate odds ratios.

Sam (2022)

[14]

Malaysia.

Qualitative study.

13; medical students.

The recorded interview data were thematically analysed using the six phases of Braun and Clarke’s Thematic Analysis.

In-depth individual interview via Microsoft Teams (Microsoft Corp.) with semi-structured questions.

Šimić (2021)

[34]

Bosnia and Herzegovina.

Cross-sectional study.

246; medical students.

The impact scale of the traumatic event (IES - Impact of Event Scale). Furthermore, the questionnaire contained six questions on the negative impact of a pandemic on mental health.

Participants completed a modified anonymous online questionnaire.

Soltan (2021)

[35]

Egypt.

Cross -sectional.

282; medical students.

Psychometric tools for the assessment of depression, anxiety and stress (Depression Anxiety Stress Scales DASS-21) and the Impact of Event Stress Scale-Revised (IES-R)

The variables were expressed in number, percentage, mean and standard deviation. Association between qualitative variables was assessed using Chi-square test. Fisher’s exact test was used in the case that any of the expected cells were less than five. A logistic regression was performed to ascertain the effects of possible risk factors on depression, anxiety, and other outcomes.

Tahir (2022)

[36]

Pakistan.

Cross-sectional.

1100; Medical students from 5 medical schools.

The influence of COVID-19 Pandemic on sleep, physical activity, and nutrition, substance abuse, dealing with finances, spirituality and family life using Likert scales.

Convenience sample. Self-administered online questionnaires.

Thind (2021)

[37]

Saint Kitts and Nevis.

Cross-sectional.

104; Medical students (2nd, 3rd and 4th years).

2 questions on the survey; (1) How anxious did you feel during the lockdown? (2) How depressed did you feel during the lockdown?

Survey distributed using co-students, friends’ circle, and through social media platforms.

Wang (2021)

[38]

China.

Cross-sectional.

403; Medical students.

Perceived Stress Scale (PSS-10).

The sample size was computed by conducting linear multiple regression prior to power analysis using the G* Power 3.1 software. The potential stressors in this study were adapted from the Source of Stress Questionnaire.

Wurth (2021)

[15]

Switzerland.

Mixed methods.

803; Medical students (2nd to 6th years).

Perceived Stress Scale (PSS).

A survey containing on one hand open-ended questions, yielding qualitative data, and on the other hand, Likert type items and Yes-No responses to closed questions

Xiao (2020)

[39]

China.

Cross-sectional.

933; Medical students.

Patient Generalized Anxiety Disorder-7 and Health Questionnaire-9.

Multivariable logistic regression analyses were performed to test the association between the type of university studies (healthcare studies: medical and non-medical, and non-healthcare studies) and poor mental health outcomes.

Zhao (2021) [40]

China.

Cross-sectional.

666; First to third year medical students.

Depression measured through Patient Health Questionnaire-9 (PHQ-9). Simplified Coping Style Questionnaire utilising a Likert scale. Ego Resilience 89 Scale.

Self-administered electronic questionnaires. Sample size was calculated using stratified sampling. Assessing prevalence of depression and exploring the role of coping styles as facilitators between resilience and depression. Comparison among groups was analysed through two-tailed t-tests and one-way analysis of variance tests. Hierarchical linear regression was utilised to assess the mediating role of coping styles alongside resilience and depression. Structural equation modelling was utilised to depict the role that coping styles had in the relationship between resilience and depression. Validity and reliability of questionnaire was analysed.

Žuljević (2021)

[17]

Croatia.

Cross-sectional pre and post survey.

437 in pre survey and 235 after; Medical students.

Oldenburg Burnout Inventory and Copenhagen Burnout Inventory.

Data were collected before lockdown in December 2019 and January 2020 and again after the end of lockdown in June 2020. Study was initially planned in 2019 with the aim of comparing medical student study satisfaction and burnout between clinical and preclinical study years, as well as using a follow-up survey to assess possible changes as the academic year goes on.