Stress Management for Medical Students: A Systematic Review

Tertiary education has always been regarded as highly stressful environment to students (Saipanish, 2003; Sherina et. al, 2003). Medical training further adds to the already stressful environment. Studies have revealed a high prevalence of psychological distress in medical students, ranging from 21.6% to 56% (Aktekin et al., 2001; Chandrasekhar et al., 2007; Dahlin et al., 2005; Firth, 1986; Guthrie et al., 1995; Miller & Surtees, 1991; Johari & Hashim, 2009; Saipanish, 2003; Sherina et al., 2003; Yusoff et al., 2011; Yusoff et al., 2010; Zaid et. al, 2007). Two studies in Malaysian government universities reported that 29.1 % to 41.9% of the medical students surveyed had psychological distress (Sherina et al., 2003; Yusoff et al., 2010) and another study in a Malaysian private medical school reported that 46.2% had psychological distress (Zaid et al, 2007). Apart from that, the stress level is higher in medical students compared to students in other courses. A study in Singapore reported that 57% of medical students had psychological distress compared to 47.3% of law students (Ko et al., 1999). Another study in Turkey reported that 47.9% of medical students had psychological distress compared to 29.2% of economic and physical education students as measured by GHQ (Aktekin et al., 2001). The alarming facts suggested that a sense of growing pressure on medical students.


Introduction
Tertiary education has always been regarded as highly stressful environment to students (Saipanish, 2003;Sherina et. al, 2003).Medical training further adds to the already stressful environment.Studies have revealed a high prevalence of psychological distress in medical students, ranging from 21.6% to 56% (Aktekin et al., 2001;Chandrasekhar et al., 2007;Dahlin et al., 2005;Firth, 1986;Guthrie et al., 1995;Miller & Surtees, 1991;Johari & Hashim, 2009;Saipanish, 2003;Sherina et al., 2003;Yusoff et al., 2011;Yusoff et al., 2010;Zaid et. al, 2007).Two studies in Malaysian government universities reported that 29.1 % to 41.9% of the medical students surveyed had psychological distress (Sherina et al., 2003;Yusoff et al., 2010) and another study in a Malaysian private medical school reported that 46.2% had psychological distress (Zaid et al, 2007).Apart from that, the stress level is higher in medical students compared to students in other courses.A study in Singapore reported that 57% of medical students had psychological distress compared to 47.3% of law students (Ko et al., 1999).Another study in Turkey reported that 47.9% of medical students had psychological distress compared to 29.2% of economic and physical education students as measured by GHQ (Aktekin et al., 2001).The alarming facts suggested that a sense of growing pressure on medical students.
Studies revealed that the stressors affecting medical students' well being seems to be related to the medical training especially related to academic matters (Aktekin et al., 2001;Guthrie et al, 1995;Kaufman et al., 1996Kaufman et al., , 1998;;Saipanish, 2003;Yusoff et al., 2011;Yusoff et al., 2010).They found that the top four stressors were tests and examinations, time pressure, too many content to be studied, and getting behind in work.Another three common stressors were conflicting demands, not getting work done within time planned and heavy workload.A small number of medical students suffer from personal problems, but the effect of this on medical students' psychological morbidity and academic success is unclear (Guthrie et al, 1995;Firth, 1986;Saipanish, 2003).Curriculum differences in medical schools may not necessarily cause differences in the overall pattern of stressors (i.e.most of the top stressors are related to academic matters), although frequency (rank) of some stressors may be significantly different (Kaufman et al., 1996(Kaufman et al., , 1998)).
It is worth to highlight that several medical education constituencies have emphasized the importance of teaching stress management and self-care skills to medical students (Steven et al., 2003;Susan et al., 2007).A recent literature review discovered that, although more than 600 articles addressed the importance of stress management programs in medical curricula, only 24 reported intervention programs with accompanying data; however none of the programmes provide convincing evidence of their effectiveness (Shapiro et al., 2000).Apart from that, their specific applications to medical education have been largely unexplored (Shapiro et al., 2000).Therefore a systematic review was done to evaluate the effectiveness of stress management specifically done on medical students with regard to five aspects which were 1) nature of participation, 2) research methods, 3) structure, facilitator and duration of intervention, 4) measured outcomes and instruments used to measure them and 5) outcomes of the intervention.On top of that we also categorized studies based on country.

Methodology
The literature search was performed using the Google Scholar, PubMed database, EbscoHost databases, Cochrane Library database, Scopus database, and Science Direct database.Keywords used in searching include 'medical student', 'stress management', 'medical student wellbeing', and 'stress intervention'.No time limit was specified in searching.Abstracts of the searched articles were read through for relevance.Participants, sampling method, study design, intervention structure, content and technique, and outcomes were the key issues of inclusion criteria for in-depth study of the full articles.Articles must describe stress management specifically for medical students otherwise they were not included in this review.Some of the articles were searched from the reference lists of the articles of primary search.

Results
Based on the keywords stated in the method, our search found that Google scholar database yielded over 1000 articles, Pubmed database yielded 275 articles, Cochrane Library database yielded 99 articles, EBSCO host database yielded 408 articles, Scopus database yielded 324 articles and Science Direct yielded 14 articles.However, based on abstract reading we found 28 articles fulfilled our inclusion criteria and they were selected for in-depth review.After the in-depth review 22 articles were included for review, 6 articles were excluded due to irrelevant content for current review.A new article was found from the reference list of the primary search and it was included in this current review.Approximately 23 articles were appraised and the results were summarised in tables.The earliest study was found in 1978 and the latest study was found in 2011.The earliest study was reported in 1978 (Soskis, 1978) and the latest study was reported in 2011 (Yusoff, 2011).Results of this systematic review were tabulated based on the five areas which were 1) nature of participants, 2) research methods (table 1), 3) structure, facilitators and duration of intervention (table 2), 4) measure outcomes and instrumed used to measure them (table 3) and 5) outcomes of the interventions (table 4-8).
The smallest and biggest number of participants involved in an intervention were 9 (Zeitlin et al., 2000) and 315 (Hassed et al., 2008) respectively.However, majority of those studies (n=12) had involved 30 to 50 participants in an intervention ( 7 Questions covered on anxiety, depression and satisfaction (1) Michie & Sandu (1994).
Psychological distress level was measured by mostly by General Health Questionnaire 12item (n=2) and the Distress Subscale of Symptom Checklist Revised (n=2) followed by the Depression Anxiety Stress Scale (n=1) and other inventories (table 3).
Despite of the four most common measured outcomes (i.e.students' perception, anxiety, depression and psychological distress), they were other important outcomes to be considered in future research such as loneliness (n=3), mood states (n=3), academic performance (n=3), health biomarkers (n=3), quality of life (n=2) and general wellbeing (n=1) (table 3).
Outcomes of interventions were summarized based on five categories which were brief intervention (less than 2 days), short-duration intervention (2 days to 4 weeks), mediumduration intervention (more than 4 weeks and up to 8 weeks), long-duration intervention (more than 8 weeks) and other (duration was not mentioned in the articles).
There were three brief interventions reported by previous studies (table 4) and all of them had significant positive impacts on psychological health of medical students (table 4).The massage therapy improved immunologic and physiologic health marker (Zeitlin et al., 2000).While the Medical Student Wellbeing Workshop improved awareness of participants about stress, its effect and management as well as a well-accepted intervention by participants (Yusoff & Rahim, 2010).5).About three interventions were reported to have significant positive impacts on psychological health of medical students (Bughi et al., 2009;Jain et al., 2007;Michie & Sandhu, 1994) whereas other outcomes were different from each intervention (table 5).Nevertheless, these facts had provided evidence of positive impacts of short-duration intervention on medical students' psychological health, awareness and general wellbeing.

Name of intervention
There were six medium-duration interventions reported and most of them were well accepted by medical students as well as increased awareness of the students about handling The stress management course -Increased intrinsic and extrinsic satisfaction.
-Positive perception towards the intervention.
-Increase awareness on stress, its effects and management.
-Reduction of anxiety and depression symptoms pre and post intervention.
-Increased interest to consider a career in psychiatry.
-The intervention was very helpful in providing insight about stress and health.
-Reduced distractive thought and behaviour.
-Enhanced positive state of mind.
Nonetheless, these facts had provided evidence of positive impacts of the medium-duration intervention on medical students' psychological health, empathy, spirituality, awareness related to handling stress as well as general wellbeing (table 6).

Name of intervention
Outcome Summary of outcome (n) Mitchell et al (1983), US.

Support group
-No measureable effect on academic performance.
-No measureable effect on anxiety or depression level.
-No measureable effect of stress symptomatology.
-No measureable effect on personality.
- Table 6.Outcomes of medium-duration stress management intervention (required duration of more than 4 weeks and up to 8 weeks).
There were seven long-duration interventions reported and most of them had significant positive effects on psychological health of medical students as well as they were well accepted by the students (table 7) while other outcomes were varied from each intervention ranging from increased awareness about stress management and it's important to improved immunologic health markers.Among these interventions, the Mind Body Medicine Skills showed very good impacts on medical students' stress biomarkers such as Cortisol, DHEA-S and testosterone (MacLaughin et al., 2010).In general, the outcomes of these interventions were related to improvement of psychological health, stress biomarkers, immunologic health marker, awareness about stress and its management, and general wellbeing (table 7).

Source and Country
Name of intervention Outcome Summary of outcome (n) Whitehouse et al (1996), US.

Self-hypnosis training
-Reduction of anxiety level during examination period.
-Improved quality of sleep.
-No measureable effect on loneliness state.
-Lowered number of T Lymphocyte at the late semester.
-Lowered stressful intensity perception towards stressful events.
-Rated as a very helpful intervention.
-Increased insight about their stresses.
-Increased confidence in handling stressful situations.Yoga exercise -Reduced distress symptoms.
-No measureable effect on depression.
-Rated as beneficial intervention.

Discussion
Without time limit, the literature search yielded 23 relevant articles reported on the effectiveness of stress management interventions on medical students.The earliest study was reported in 1978 (Soskis, 1978) and the most recent study was reported in 2011 (Yusoff, 2011).This section discussed on the effectiveness of those interventions with regards to five aspects which were 1) nature of participation, 2) research methods, 3) structure, facilitators and duration of intervention, 4) measured outcomes and instruments used to measure them, and 5) outcomes of the interventions.
In 2000, there were 15 studies reported stress reduction interventions on medical students (Shapiro et al., 2000) and in this present literature review there were 23 studies reported on the interventions (i.e.11 years after Shapiro et al (2000) systematic review, only 8 additional studies were reported on stress management interventions for medical students).These facts suggested that despite large number of articles criticized on the negative impacts of stress related to medical training on medical students and call for remedies to buffer the unwanted consequences yet very few have put on effort to study on specific effects of stress management interventions on medical students (Shapiro et al., 2000).Even fewer studies have provided convincing data on the effectiveness of stress management interventions on medical students' health.Therefore, now is the right moment for medical educators to put more effort to expand the body of evidence on effective interventions in buffering the negative consequences of stress related to medical training on medical students (Butterfield, 1988;Shapiro et al., 2000).

Nature of participation
This systematic review clearly showed that the biggest limitation of the reported studies was related to sampling method of the participants which was non-randomized.As a result, voluntary nature of participation to the interventions was more likely to attract students who were highly motivated to change and thus more sensitive to any intervention done; this may lead to inaccuracy of outcomes measured.Nevertheless, logically having students voluntarily participating in the intervention may be more practical and feasible (Finkelstein et al., 2007).Perhaps, random sampling method in selecting participants of stress management interventions should be considered in future research to minimise bias due to volunteer participation.Therefore more authentic and convincing outcomes could be measured.

Research methods
This systematic review clearly showed that very few studies used robust study designs in investigating impact of stress reduction interventions on medical students and all of them were conducted in United State (US) as shown in the table 1. Although, Shapiro et al (2000) recommended in previous literature review to incorporate rigorous study design such as randomized control trial, unfortunately this recommendation has not been addressed where 11 non-randomized studies were done whereas only one randomized control trial was done post-recommendation (table 1).A possible reason for researchers preferred to conduct nonrandomized experimental studies instead of randomized control trial may be due to issues related ethical, feasibility and practicality of randomizing participants into intervention and control groups (Finkelstein et al., 2007;Piaw, 2009;Katz, 2010).

Structure, facilitators and duration of intervention
This systematic review demonstrated that majority of interventions was conducted by psychologist/psychiatrist, offered as an elective course as well as seminar or workshop, consumed a duration of 360 to 540 minutes over 3 to 8 weeks and involved 30 to 50 participants.For RCT studies, majority involved year 1 medical students as study subjects, sample size for intervention groups ranged from 4 to 50 participants and relatively RCT used random sampling method had poorer response rate compared to non-random sampling (table 9).Perhaps stress management interventions should be conducted by general faculty members of medical schools instead of few experts so that the interventions can be implemented effectively to medical students.On top of that, most of the studies had not explained theoretical basis of the intervention was designed.Perhaps, future studies should describe the theoretical basis of stress management intervention was designed so that researchers could compare and come out with more effective intervention based on more robust theory of stress management intervention for medical students.It is worth highlighted that generally the interventions were categorised into brief, short-duration, medium-duration and long-duration stress management internvetions.

Measured outcomes and instruments used to measure them
This review revealed that various aspects of health outcomes were measured ranging from students' perception up to health biomarkers.Despite the variability of measured outcomes, there were three main outcomes related to psychological health as measured by most of the studies which were anxiety, depressive and psychological distress symptoms.

Outcomes of the interventions
For the past 24 years, regardless of the duration of stress management interventions, this literature review revealed the interventions done on medical students had important positive outcomes on several areas related to health.The outcomes ranged from positive students feedback up to improvement of health biomarkers.The reported positive outcomes were related to 1) positive student feedbacks, 2) improved psychological health, 3) improved loneliness and mood disturbances, 4) improved physiologic and immunologic health markers, 5) improved quality of life, spirituality, and empathy, 6) improved psychological states of mind, 7) increased awareness about stress, its effects and management, and 8) improved perceived ability to cope effectively and positively.Despite of these positive outcomes, none of studies demonstrated effectiveness of the interventions on clinical competencies, professionalism, doctor-patient relationships, attrition and suicidal thoughts.Perhaps these outcomes should be investigated in future researches.

Conclusion
This systematic review found that stress management interventions done on medical students were well-accepted and had important positive outcomes on several areas related to health.Despite these encouraging outcomes several limitations should be considered for future research which are (1) longer duration of follow up measurement on intended outcomes, (2) more robust research method, including proper sample size calculation, random sampling of subjects, randomised allocation of subjects to intervention group and comparable control group, (3) comparing impacts of intervention on different stages of medical training, (4) customized and personalized stress management, (5) investigate the impacts of stress management on professionalism, doctor-patient relationship and patient care in future, and (6) specify the theoretical basis of stress management was developed.
The implications of this review are significant on a few areas that might be worthwhile for further research.Future research must look at impacts of different duration and frequency of stress management interventions on students' health, personal and professional development; therefore optimal duration and frequency of these interventions to produce positive impacts can be determined.It is also worthy to explore which components of these interventions produce therapeutic effects and which are more effective.On top of that, future research must look at which of these interventions work best to which group of students, therefore personalized and customised stress management interventions can be designed accurately and effectively.Last but not least, future research must utilise rigour and robust research methodology to elicit real impacts of stress management interventions.Perhaps, the implications discussed in this review are not only confined to medical students, but it can be also utilised by researchers of other disciplines as a guideline to design, plan and conduct similar researches in their own setting.Utilization of similar health measurements for outcomes comparison in future researches is recommended.

Acknowledgment
We would like to express our deepest gratitude and thank to our parents, wife and family who gave great support and spared their time for the success of completing this chapter.We would like to express our thank to Professor Syed Hatim Noor, Associate Professor Mohd Jamil Yaacob, the Head Department of Medical Education, staff and colleagues for their continuous support.Last but not least my deepest appreciation to medical students who inspire me to write this chapter.Most and foremost our greatest gratitude to Universiti Sains Malaysia for providing funds under the Research University Grant 1001/PPSP/812086.

Table 1
8.7% (n=2) was offered as support group, 8.7% (n=2) was offered as a program built in the core curriculum and 4.3% (n=1) was offered as a volunteer program (table 2).

Table 2 .
Summary of structure, participants, facilitator and total duration of interventions of the 23 studies were conducted.

Table 4 .
Outcomes of brief stress management intervention (required duration of less than 2 days)Five short-duration interventions were reported by previous studies (table

Table 9 .
There were two interventions reported under other category.In general the outcomes of these interventions were related to improvement of psychological health, general wellbeing and awareness about coping strategies (table8).Summary of sample size of intervention and control groups for the randomized control trial studies.
These outcomes were mainly measured by established psychological health measurements such as the State-Trait Anxiety Inventory, Brief symptoms Inventory, Symptoms Checklist Revised, Depression Anxiety Stress Scale and Beck Depression Inventory.Other important outcomes that should be considered in future researches such as academic performance, patient-doctor relationship, loneliness state, health biomarkers, quality of life, and suicidal thoughts.