Adv in Health Sci Educ DOI 10.1007/s10459-017-9770-5
Group interventions to promote mental health in health professional education: a systematic review and metaanalysis of randomised controlled trials Kristin Lo1 • Jamie Waterland1 • Paula Todd2 • Tanvi Gupta3 • Margaret Bearman4 • Craig Hassed5 Jennifer L. Keating1
•
Received: 16 August 2016 / Accepted: 8 March 2017 Springer Science+Business Media Dordrecht 2017
Abstract Effects of interventions for improving mental health of health professional students has not been established. This review analysed interventions to support mental health of health professional students and their effects. The full holdings of Medline, PsycINFO, EBM Reviews, Cinahl Plus, ERIC and EMBASE were searched until 15th April 2016. Inclusion criteria were randomised controlled trials of undergraduate and post graduate health professional students, group interventions to support mental health compared to alternative education, usual curriculum or no intervention; and post-intervention measurements for intervention and control participants of mindfulness, anxiety, depression, stress/distress or burnout. Studies were limited to English and short term effects. Studies were appraised using the PEDro scale. Data were synthesised using meta-analysis. Four comparisons were identified: psychoeducation or cognitive-behavioural interventions compared to alternative education, and mindfulness or relaxation compared to control conditions. Cognitive-behavioural interventions reduced anxiety (-0.26; -0.5 to -0.02), depression (-0.29; -0.52 to -0.05) and stress (0.37; -0.61 to -0.13). Mindfulness strategies reduced stress (-0.60; -0.97 to -0.22) but not anxiety (95% CI -0.21 to 0.18), depression (95% CI -0.36 to 0.03) or burnout (95% CI -0.36 to 0.10). Relaxation strategies reduced anxiety (SMD -0.80; 95% CI -1.03 to -0.58), depression (-0.49; -0.88 to -0.11) and stress (-0.34; -0.67 to -0.01). Method quality was generally poor. Evidence suggests that cognitive-behavioural, relaxation and mindfulness interventions
& Kristin Lo
[email protected] 1
Department of Physiotherapy, Monash University, Monash University Peninsula Campus Building B, McMahons Road Frankston, Melbourne, VIC 3199, Australia
2
Subject Librarian, Faculty of Medicine, Nursing and Health Sciences, Monash University, Melbourne, Australia
3
South Eastern Private Hospital, Melbourne, Australia
4
HealthPEER Monash University, Melbourne, Australia
5
Department of General Practice, Monash University, Melbourne, Australia
123
K. Lo et al.
may support health professional student mental health. Further high quality research is warranted. Keywords Mental health Health professional student Mindfulness Cognitivebehavioural Relaxation Psychoeducation Systematic review Meta-analysis
Introduction Health professional students experience psychological distress and challenges to personal wellbeing (Shapiro et al. 2000; Balogun et al. 2002). Distress is defined by Dyrbye et al. (2005), as an overarching concept including depression, anxiety and burnout [characterised by emotional exhaustion and cynicism (Maslach and Jackson 1981)]. A 2006 systematic review reported that medical students in America and Canada have a high prevalence of depression, anxiety and overall distress compared to age-matched peers (Dyrbye et al. 2006). Hope and Henderson (2014) conducted a systematic review into medical student distress outside Northern America and reported prevalence of anxiety (7.7–65.5%), depression (6–66.5%) and psychological distress (12.2–96.7%). The prevalence of burnout in medical students has also been reported in a systematic review to be 45–71% and may increase across time into professional life (IsHak et al. 2013). High rates of burnout have also been reported for nurses, physiotherapists and occupational therapists (Balogun et al. 2002; Rudman and Gustavsson 2012). Conditions such as distress and burnout can have serious consequences including sleeplessness, drug and alcohol use, family conflict, suicidal ideation and sickness (DiGiacomo and Adamson 2001; Struber 2003; Van Der Heijden et al. 2008; Dyrbye et al. 2012). Mental health issues can also result in increased rates of medication errors with depressed medical residents making 6.2 times as many medication errors per month than their non-depressed colleagues (Fahrenkopf et al. 2008). Thus improvements in the mental health of health professionals will not only have a positive impact on these individuals but also on the clients they support. Despite these statistics, only a small percentage of health professional students seek professional support or guidance for mental health (Roberts et al. 2000). Clinicians supervising students in the clinical environment report the common occurrence of student mental health issues and a lack of confidence and/or comfort with providing support (Lo et al. 2015). Health professional courses are demanding in terms of competition for specific career pathways, long contact hours and course length, knowledge retention requirements and the frequently confronting circumstances associated with service provision. These may amplify challenges to the mental health of learners. Consequently there has been a call for change, for health professional curricula to incorporate strategies to support student mental health and wellbeing (Seritan et al. 2012). Successful models have been reported. There have been five recent and relevant systematic reviews in this field, but none have focussed on the spectrum of learners who face the challenges associated with health service delivery. Conley et al. (2013) reviewed 83 controlled studies of programs to promote mental health in students undertaking a health professional course. These were stratified into psychoeducation (education regarding stress management, coping and relaxation strategies), cognitive-behavioural (monitoring cognition to modify behaviours/emotions), mindfulness (focusing on non-judgemental awareness of present circumstances), relaxation
123
Group interventions to promote mental health in health in health…
(including progressive muscle relaxation or autogenic training), meditation and other interventions (writing tasks/communication and problem solving skills). In improving social and emotional skills, self-perceptions, and emotional distress, mindfulness was most effective followed by cognitive-behavioural interventions then, relaxation and meditation (Conley et al. 2013). Psychoeducation was associated with modest significant effects on self-perception (Conley et al. 2015). To improve anxiety, mindfulness and cognitive-behavioural strategies were effective compared to controls (Regehr et al. 2013). Psychoeducation was associated with modest significant effects on anxiety (Conley et al. 2015). Computer-based CBT appeared to be more effective in targeting anxiety compared to control conditions (Farrer et al. 2013). Depression was improved by both mindfulness and cognitive-behavioural interventions while relaxation effects appeared similar to those of meditation (Regehr et al. 2013). In terms of burnout, relaxation methods or converting to a pass-fail grading system were found by Williams et al. (2014) to reduce burnout in comparison to control conditions. They also found that self-development groups reduced burnout in comparison to alternative education. The effects of mindfulness and on-duty time limits on burnout were mixed. No studies were found on preventing suicide. Academic measures were improved by psychoeducation (Conley et al. 2015). Conley et al. (2013) reported that interventions that included supervised practise were on average five times less likely to display psychological distress, poorer social/emotional skills and poorer self-perception than those without supervised practice. Consequently, interventions incorporating supervised skills practice were recommended for inclusion in curricula. Regehr et al. (2013) recommended further research into strategies for stress reduction in males. Our review and meta-analysis of recent randomised controlled studies testing interventions that address a broad range of mental health issues and include participants from a spectrum of health professions, fills a gap in the available evidence. The aim of this review was to identify the evidence that supports interventions suitable for embedding in usual health professional course curriculum and that could be delivered to groups of learners. This was due to the fiscal pressures associated with health professional courses. The question addressed was: ‘‘What is known about the effects of group interventions designed to enhance/maintain mental health of students enrolled in health professional education compared to alternative or usual curriculum, or no program?’’ We adapted a definition of mental health and defined it as ‘‘a state of well-being in which an individual can cope with the normal stresses of life, and can work productively and fruitfully’’ (World Health Organisation 2014). A secondary review aim was to describe program elements in included studies.
Methods Protocol This systematic review followed PRISMA standards. PRISMA is an evidence-based set of minimum standards for reporting systematic reviews and meta-analyses (Moher et al. 2009). It focuses on the evaluation of randomised controlled trials providing a checklist to ensure that the title, abstract, introduction, methods, results, discussion and funding are reported in a specific and systematic manner.
123
K. Lo et al.
Eligibility criteria To be included in the review, studies had to adhere to all of the following criteria:
Population Undergraduate or post-graduate health professional students.
Intervention Any group intervention designed to enhance or maintain mental health. There was no minimum or maximum sample size.
Comparison Any alternative approach to support mental health, usual curriculum content or no intervention control conditions.
Outcomes Must report at least post intervention measurements of mindfulness, anxiety, depression, stress/distress or burnout. Mindfulness is considered a valid indicator of attentional control which has been demonstrated to affect the perception of stress and other mental health indicators (Fjorback et al. 2011). Measurements of factors associated with mental health, such as sleep quality, did not meet this criterion.
Study design Studies were randomised controlled trials (RCT) of interventions delivered to groups of students using face to face and/or online delivery methods.
Report characteristics The research was published in English as financial supports were not available to translate articles from languages other than English. The research was published in peer-reviewed journals, described interventions in enough detail to enable replication and presented data that enabled analysis of the effect of the intervention (point estimates and measures of variability for intervention and control groups after the intervention), or data that enabled estimates of these values.
Search strategy Databases searched were the full holdings of OVID Medline (1946 to present), PsycINFO (Ovid 1987–2016), EBM Reviews (Ovid 3rd quarter 2016), Cinahl Plus (Ebsco 1937–2016), ERIC (ProQuest 1945–2016), EMBASE (Elsevier 1957–2016) through to 15th April 2016. A combination of MeSH and keywords were used. The full search strategy for each database is available on request (‘‘Appendix’’). Two researchers searched
123
Group interventions to promote mental health in health in health…
the databases and their results were compared. Reference lists from the five systematic reviews published since 2013 were scanned for relevant articles. The search yield was imported into bibliographic management software (EndNote X7.1). Duplicates were removed and two researchers independently identified articles of potential relevance based on title and abstract. Full-text articles were retrieved and assessed against inclusion and exclusion criteria. Where required, authors were contacted to obtain additional data to confirm inclusion/exclusion criteria. Consensus on article inclusion was reached via discussion. Where disagreement occurred a third reviewer (JK) was consulted.
Data extraction Two researchers piloted data extraction tables to independently extract the following data: author/s, year and location of the study, participants’ demographic details (course of enrolment, age, gender), study details (number of participants, whether participants were volunteers, intervention and comparator), educational intervention (content, duration, frequency, homework, timing relative to coursework, intervention delivery mode, whether skill acquisition was supervised, assessment of compliance and program evaluation), and mental health outcomes (method of assessment, post-intervention measurements of relevant outcomes, and p values for tests of differences between groups). We also extracted details of control conditions. As the opportunity for supervised skill practice may impact on the effectiveness of interventions, studies were assessed for this inclusion. According to previously applied criteria (Conley et al. 2013), studies were considered to include supervised practice if the facilitator observed and provided constructive feedback across multiple sessions while students practised. Measures of mindfulness, anxiety, depression, stress/distress or burnout were the only outcomes of interest in this review. If authors collected data for the same type of outcome using a range of indices, outcome measures common across the included studies were preferentially extracted for analysis of effect. Key information was extrapolated from graphs where data were not reported elsewhere. Erogul and colleagues suggested that the use of volunteer participants may present a selection bias, favouring participation of reflective students. Thus data were extracted on whether participants volunteered. If data extraction discrepancies occurred, these were resolved through discussion. The accuracy of independent data extraction was calculated by comparing the data extracted by individual researchers (KL and TG) to data achieved by final consensus. A percentage of agreement was then calculated.
Risk of bias within studies Eligible studies were assessed independently by two researchers for potential bias using the PEDro scale a validated approach to systematic evaluation of the method quality of clinical trials (de Morton 2009). The scale consists of 11 items, ten of which are scored. If the item fulfils a particular criterion, it is given a score of one such that the total score adds to a maximum score of ten. Discrepancies were resolved through discussion. The PEDro scale assesses ten sources of bias including bias associated with random allocation of participants (selection bias), blinding (performance and detection bias), incomplete outcome data (attrition bias) and bias associated with intention to treat. Higher scores indicate higher quality papers with fewer potential sources of bias. As per protocol, items 2–11 on the 11
123
K. Lo et al.
item scale were used to assess study validity and bias (score 0–10). PEDro item decision rules were applied with the exception of item 2; ‘randomisation’ which was considered achieved if randomisation was reported, regardless of the methods used for randomisation. The review team considered that as item 3 assessed concealment, it assessed the study for the potential for the method of randomisation (e.g. using date of birth) to affect concealment. Item 8 ‘Attrition’ was calculated utilising the immediate post intervention scores. Where required, baseline measurements were tested for significant differences between groups with independent t tests.
Summary measures The principle summary measures were means and standard deviations for each outcome assessed and the number of participants. These were extracted or calculated separately for intervention and control groups.
Synthesis of results Interventions were categorised by two independent researchers into psychoeducation, cognitive-behavioural, mindfulness, relaxation and other categories (Conley et al. 2013). The content of the education interventions were thematically analysed by two independent researchers (Braun and Clarke 2006). This involved independent review of the intervention description, data extraction and coding and collating into themes. After themes were refined through consensus, researchers independently recoded the data according to the refined themes. A final phase of consensus was performed to check for consistency between reviewers in collation into themes. To enable meta-analysis, interventions were grouped into the six criteria proposed by Conley et al. (2013). Analysis of statistical data was performed using Cochrane Collaboration guidelines (Green and Higgins 2011). Where authors did not report between group differences required to estimate the intervention effect we calculated the Hedges g effect size using the formula sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ðn1 1ÞSD21 þ ðn2 1ÞSD22 SDpooled ¼ n1 þ n2 2 and associated 95% confidence intervals were calculated for each outcome. Outcome data for similar interventions and outcomes were pooled in meta-analysis using Review Manager Version 5.3. Conversion of standard errors (SE) of means to standard deviations (SD) was completed using the formula SE = SD/(Hn). Forest plots were generated. Hedges standardised mean difference (SMD) (Hedges 1981) and associated 95% confidence intervals were used to standardise data assessed on different scales. The convention was adopted that an SMD of\0.2 describes a small effect, 0.5 a moderate effect and[0.8 a large effect (Cohen 1988). The alpha level for the test of differences between groups was set at 0.05. A fixed effects meta-analysis model was initially applied. Trial heterogeneity was assessed using the I2 statistic which evaluates between study variability. Where statistical heterogeneity was concluded (I2 [ 50% or p value\0.1) (Higgins et al. 2003), data were re-analysed using random effects meta-analysis (Higgins et al. 2003).
123
Group interventions to promote mental health in health in health…
Risk of bias across studies Potential risk of bias across studies due to selective reporting of results was addressed (wherever possible) by comparing reported outcomes to those described in previously published protocols. The PEDro score was added to forest plots to facilitate visual association between results and risk of bias.
Results Study selection
Identification
From the initial search yield of 11,744 papers, 151 papers were retrieved for full text evaluation. Of these, 127 papers did not meet all inclusion criteria. Figure 1 summarises the process that resulted in the final yield of 24 articles.
Records idenfied through database searching (n = 11744)
Addional records idenfied from other sources (5 relevant recent systemac reviews) (n = 12)
Screening
Records aer duplicates removed (n = 7428)
Eligibility
Records screened (n = 7428)
Full-text arcles assessed for eligibility (n = 151)
Inclusion
Included studies (n = 24) Psychoeducaon (n = 2) Cognive-behavioural (n = 3) Mindfulness (n = 10) Relaxaon (n = 8)* Other (n = 3)
Records excluded (n = 7310)
Full-text arcles excluded (n = 127) Not RCT Not peer reviewed journal Not English language Clinical populaon Ineligible outcomes measured Ineligible intervenon Not only health professional students Intervenons not sufficiently described Insufficient data Not group intervenon
n = 14 n=9 n=2 n=7 n = 11 n=3 n = 61 n=3 n = 11 n=6
Included in meta-analysis (n = 19) Psychoeducaon (n = 2) Cognive-behavioural (n = 3) Mindfulness (n = 8) Relaxaon (n = 8)*
Fig. 1 PRISMA 2009 flow diagram describing the pathway of reports into the review. *Two studies included both relaxation and mindfulness interventions
123
K. Lo et al.
Data extraction Based on 1238 extracted data items, there was 96% agreement between independent reviewers. The sources of disagreement were commonly related to extraction of outcomes data. All disagreements were resolved with collaborative review of the source material and discussion. In two articles, key information was extrapolated from graphs (Shapiro et al. 1998; Jones and Johnston 2000).
Characteristics of included studies Population Publication demographics are summarised in Table 1. Most of the studies (19/24, 79%) were published after 2000 with 12/24 conducted in the USA and Canada. Included studies involved 2422 commencing participants who were primarily entry level students of medicine and nursing, typically in the first or second year of study. In one study gender was not reported; 68% of participants in the remaining studies were female. In 9/24 (38%) of the studies, participants were selected for inclusion in the study rather than advertising for interested volunteers.
Intervention Delivery Duration of interventions varied with a mean of 79 min (SD 62.7, range 20–300). The mean number of sessions was 11 (SD 14.2, range 1–56). Thirteen studies prescribed activities for practise between sessions which varied from an unspecified practice period to 15 min each day for 133 days (Whitehouse et al. 1996). Half of the studies included supervised practise of target skills and one study provided online feedback. Design The design of the educational interventions varied (Table 1). Most studies compared interventions to no-intervention controls.
Psychoeducational interventions (Braithwaite and Fincham 2009; Aboalshamat et al. 2015) The results of psychoeducational interventions are summarised in Table 1. Results are pooled in meta-analysis in Fig. 2 which presents the comparison of psychoeducational to alternative education on the outcomes of anxiety, depression and stress. Two studies examining effects on anxiety (385 participants) were included. No significant differences between groups (standard mean difference 0.1; 95% CI -0.1 to 0.31) were observed. Heterogeneity assessment (97%) indicated high variability between studies and a random effects analysis was used. In the Braithwaite and Fincham study (2009) participants reported more anxiety in the intervention group. This was hypothesised to result from applying skills such as assertiveness to participant’s relationships which may have potentially raised issues that had previously not been attended to. Two studies assessed effects on depression (385 participants). No significant difference between groups was found in pooled analysis (SMD -0.2; 95% CI -0.4 to 0.00; p = .06).
123
Sample size and participants (year level)
77 UG introductory psychology (N/ A)
257 UG psychology (various)
292 UG psychology (various)
29 UG paramedics (2 = final year)
Braithwaite (2009) USA
Frazier et al. (2015) USA
Hintz et al. (2015) USA
Porter (2008) Canada
75%
70%
V
S
V
288 UG medicine and psychology (term 2 or 3)
58 UG medicine (1)
24 UG nursing (2)
de Vibe et al. (2013) Norway
Erogul et al. (2014) USA
Greene (1988) Canada
V
S
S
60 UG nursing (1)
75%
46%
76%
65%
44.8%
Mostly female
S
V
55%
% Female
S
Selected/ volunteer
Chen et al. (2013) China
Mindfulness
422 UG medical and dental (2 and 3)
Aboalshamat (2015) Saudi Arabia
Psychoeducational
First author (year) country
M
M
M
M
C
C
C
P
P
Intervention category#
Mindfulness meditation versus cognitive restructuring (learning to modify maladaptive thoughts)
Abridged MBSR versus no intervention control
MBSR versus usual curriculum
Mindfulness meditation versus waitlist control
Cognitive-behavioural counselling versus usual curriculum
Present control (PC) versus PC intervention with feedback (PC ? F) versus stress information only
Stress management intervention including cognitive behavioural intervention (present control) and biofeedback versus stress information only
ePREP versus usual curriculum
Self-development coaching with CD for motivation and relaxation versus control (placebo education)
Comparison
Table 1 Summary of mental health interventions for health professional students
3 9 60 min over ? weeks HW: ?(daily) 9 30 min
8 9 75 min over 8 weeks HW: 56 9 20 min
6 9 90 min 6 weeks ? 1 9 360 min session HW: 49 9 30 min
7 9 30 min 1 week HW: none
13 9 ? mins over 16 weeks HW: Nil
4 modules 9 ? over 2 weeks HW: ? sessions
4 modules 9 60 min over 2 weeks HW: 7 9 ? minutes
1 9 60 min, 8 weeks HW: 56 9 ? minutes
2 9 300 min, 1 week HW: ? sessions
Length of intervention (no. of sessions) 9 (minutes) 9 (week) Homework (HW): no. of sessions 9 (minutes)
Group interventions to promote mental health in health in health…
123
123
56%
V
V
S
89 UG nursing student (2)
78 UG/PG medicine/premedicine (various)
66 UG Medicine (final 2 years)
Raranasiripong (2015) Thailand
Shapiro et al. (1998) USA
Warnecke (2011) Australia
60 UG psychology (N/A)
40 UG nursing (1)
Dimou et al. (2014) Greece
Heaman (1995) USA
Relaxation
V
S
100%
V
58 Medicine (5)
Paholpak (2012) Thailand
100%
79%
64.6%
50%
51.5%
V
449 (229 full data set) PG medicine (1)
McGrady (2012) USA
81.5%
V
104 UG/PG medical, nursing, premedics, prehealth (various)
Jain (2007) USA
% Female
Sample size and participants (year level)
First author (year) country
Selected/ volunteer
Table 1 continued
R
R
M
M
M/R
M
M
M/R
Intervention category#
Stress management i.e.: quieting response: maintaining an alert mind with a relaxed body (Stroebel 1983) with EMG biofeedback versus control
Stress management (incuding diet, exercise, time management and safe sex education) versus no intervention control
Mindfulness versus usual curriculum
MBSR versus waitlist control
Biofeedback versus mindfulness meditation versus control
Breathing meditation versus non meditating activity (chatting, reading, napping)
Wellness program versus waitlist control
MBSR and somatic relaxation versus waitlist control
Comparison
5 9 90 min over 5 weeks HW: ? sessions
8 9 ? 8 weeks HW: 112 9 ? minutes
56 9 30 min over 8 weeks HW: 56 9 ? minutes (home diary)
7 9 150 min over 7 weeks HW: 7 9 ? minutes (assignments) ? 49 9 ? minutes (journals)
2 9 ? mins over 1 week HW: 84 9 ? minutes
28 9 20 min over 4 weeks HW: Nil
8 9 45 min over 16 weeks HW: Nil
4 9 90 mins over 4 weeks HW: ? sessions 9 30 min ? 1 9 360 min retreat
Length of intervention (no. of sessions) 9 (minutes) 9 (week) Homework (HW): no. of sessions 9 (minutes)
K. Lo et al.
76 UG nursing (sophomores and seniors)
79 UG nursing (1)
93 UG nursing (various)
35 UG medicine (1)
Johansson (1991) USA
Jones (2000) UK
Kanji (2006) UK
Kiecolt-Glaser (1986) USA
42 UG medicine (N/A)
128 UG nursing (all year levels)
20 UG pharmacy (5)
Dayalan (2010) India
Hughes (2003) USA
Unno (2013) Japan
Other
Sample size and participants (year level)
First author (year) country
Table 1 continued
S
V
V
V
V
30%
94%
50%
34%
91.4%
84.8%
100%
V
S
% Female
Selected/ volunteer
O
O
O
R
R
R
R
Intervention category#
5 9 120 min over 1 9 semester (? weeks) HW: nil 34 9 n/a over 2.5 weeks HW: nil
200 mg theanine supplements twice daily## versus placebo
42 9 25 min over 6 weeks HW: none
Minimum 5 9 30–45 min over 10 sessions HW: Encouraged but nil set guidelines
8 9 60 min over 8 weeks HW: Nil
6 9 120 min over ? weeks HW: Nil
6 9 50 min over 3 weeks HW: reported but no details provided
Length of intervention (no. of sessions) 9 (minutes) 9 (week) Homework (HW): no. of sessions 9 (minutes)
Peer group experience versus waitlist control
MST versus intervention control ‘‘let go of thoughts’’ (MST no practise)
Hypnosis/relaxation versus waitlist control
Autogenic training (AT)/laughter therapy (LT) versus waitlist control
Stress management (coping strategies, selfmonitoring, problem solving, situational reappraisal and time management skills) versus waitlist control ? distance version of education
Stress management (cognitive restructuring, breathing strategies and relaxation) versus education only control
Comparison
Group interventions to promote mental health in health in health…
123
123
N
N
Y
Braithwaite (2009) USA
Frazier et al. (2015) USA
Assessed HW compliance
Aboalshamat (2015) Saudi Arabia
Psychoeducational
First author (year) country
Table 1 continued
O
O
F&A
Delivery mode*
Y
N
N
S/V skill practise
DASS PSS
BECK dep BECK anx
DASS
Outcome measures
C
P
p
Meta-analysis^
DASS depression p < 0.05 (-0.18; -0.50 to 0.15)** DASS anx p < 0.01 (-0.18, -0.51 to 0.15)** DASS stress p < 0.01 p = 0.35** (-0.34, -0.67 to 0.02)** perceived stress p < 0.001 p = 0.45** (20.41, 20.74 to 20.08)**
No effect p = 0.33** (-0.01, -0.5 to 0.47)** p < 0.05 p = 0.31** (0.37, -0.12 to 0.86)**
Depression p = 0.00** (Hedges g 20.23, 95% CI 20.45 to 20.01) Anxiety p = 0.00** (0.23, 20.45 to 20.01)** Stress p = 0.00** (-0.12, -0.34 to 0.1)**.
Outcomes in favour of intervention
K. Lo et al.
Assessed HW compliance
N
N
First author (year) country
Hintz et al. (2015) USA
Porter (2008) Canada
Table 1 continued
F
O
Delivery mode*
Y
N
S/V skill practise
MBI
DASS PSS
Outcome measures
C
C
Meta-analysis^
Emotional exhaustion p = 0.33** (-0.40; -1.23 to 0.43)** depersonalisation p = 0.41** (-0.14; -0.96 to 0.68)** personal accomplishment p = 0.06** (0.80, -0.05 to 1.65)**
PC versus control depression p = 0.02** (20.41, 20.75 to 20.06)** anxiety p = 0.00** (20.35, 20.69 to 0.00)** stress p = 0.04** (-0.12, -0.46 to 0.22)** PC ? F versus control depression p = 0.00** (-0.03, -0.64 to 0.03)** anxiety p = 0.01** (-0.27, -0.6 to 0.06)** stress p = 0.04** (-0.17, -0.5 to 0.16)**
Outcomes in favour of intervention
Group interventions to promote mental health in health in health…
123
123
N
Y
de Vibe et al. (2013) Norway
Assessed HW compliance
Chen et al. (2013) China
Mindfulness
First author (year) country
Table 1 continued
F
F
Delivery mode*
Y
N
S/V skill practise
MBI PMSS FFMQ
SAS SDS
Outcome measures
p = 0.25** (-0.43, -0.95 to 0.08)** p = 0.14** (0.09, -0.42 to 0.59)**Note: intervention baseline scores were higher than the control group. p = 0.204 (0.15, -0.8 to 38)**, p = 0.021 (0.17; -0.07 to 0.40)** non-reacting p < 0.001 (0.33; 0.10 to 0.56) ** non-judging p = 0.085 (0.17; -0.06 to 0.40)**, act aware p = 0.314 (0.15; -0.08 to 0.38)** describe p = 0.719 (0.06; -0.17 to 0.29)** observe p = 0.034 (0.17; -0.06 to 0.40)**
M
Outcomes in favour of intervention
M
Meta-analysis^
K. Lo et al.
Assessed HW compliance
Y
N
N
N
N
First author (year) country
Erogul et al. (2014) USA
Greene (1988) Canada
Jain (2007) USA
McGrady (2012) USA
Paholpak (2012) Thailand
Table 1 continued
F&A
F
F&A
F
A&F
Delivery mode*
N
Y
Y
Y
Y
S/V skill practise
SCL90 dep SCL90 anx
BECK anx BECK dep
BSI-total
SOSI-anx SOSI-dep
PSS
Outcome measures
M
M
R
–
M
Meta-analysis^
p = 0.41** (0.22, -0.30 to 0.73)** p = 0.05** (0.52, 0.00 to 1.05)**
p = 0.33** (-0.23, -0.49 to 0.04)** p = 0.09** (0.21, -0.56 to 0.14)**
Meditation group versus control p = 0.00** (0.63, 0.09 to 1.16) relaxation versus control p = 0.01** (-0.25, -0.79 to 0.29)
p = 0.39** (0.34, -0.46 to 1.15)** p = 0.69** (0.16, -0.64 to 0.96)**
p = 0.03 (20.6, 21.13 to 20.07)**
Outcomes in favour of intervention
Group interventions to promote mental health in health in health…
123
Assessed HW compliance
Y
N
First author (year) country
Raranasiripong (2015) Thailand
Shapiro et al. (1998) USA
Table 1 continued
123 F
F
Delivery mode*
Y
N
S/V skill practise
STAI SCL90R
STAI PSS
Outcome measures
M
M/R
Meta-analysis^
State p = 0.05** (-0.45, -0.92 to 0.01)** trait p < 0.02** (20.57, 21.04 to 20.11)** depression p = 0.04** (20.49, 20.95 to 20.02)
Biofeedback p = 0.00** (20.93, 21.46 to 20.40)** mindfulness p = 0.13** (0.48, -1.00 to 0.04)** biofeedback p = 0.038** (-0.23, -0.73 to 0.28)** mindfulness p = 0.09** (0.21, -0.31 to 0.73)**
Outcomes in favour of intervention
K. Lo et al.
N
Warnecke (2011) Australia
N
Y
Dimou et al. (2014) Greece
Heaman (1995) USA
Relaxation
Assessed HW compliance
First author (year) country
Table 1 continued
F&A
A&F
A
Delivery mode*
N
Y
N
S/V skill practise
STAI
PSS
DASS PSS
Outcome measures
R
R
–
Meta-analysis^
Group 1 state p = 0.00** (21.94; 23.00 to 20.88)** Group 2 state p = 0.23** (-0.55; -1.45 to 0.35)** Group 1 trait p = 0.32** (-0.43; -1.32 to 0.45)** Group 2 trait p = 0.62** (-0.22; -1.10 to 0.67)**
p = 0.04** (20.72; 21.43 to 20.01)**
stress p = 0.05 (p = 0.34**) depression p [ 0.05 (p = 0.52**) anxiety p [ 0.05 (p = 0.87**) stress p < 0.05 (p = 0.11**),^^
Outcomes in favour of intervention
Group interventions to promote mental health in health in health…
123
Assessed HW compliance
N
N
First author (year) country
Johansson (1991) USA
Jones (2000) UK
Table 1 continued
123 M
F
Delivery mode*
Y
Y
S/V skill practise
STAI-anx BECK-Dep
STAI-anx IPAT STAI-anx IPAT
Outcome measures
trait p = 0.05** (-0.44; -0.88 to 0.01)** state p = 0.02** (20.55, 21.00 to 20.1)** depression p = 0.06** (-0.43, -0.88 to 0.01)**
Sophomore anxiety p = 0.03** (20.70, 21.32 to 20.08)** Sophomore depression p = 0.03** (20.68, 21.30 to 20.06)** Senior anxiety p = 0.00** (21.06, 21.78 to 20.34)** Senior depression p = 0.04** (20.73, 2 1.43 to 20.04)**
R
R
Outcomes in favour of intervention
Meta-analysis^
K. Lo et al.
N
N
Kanji (2006) UK
Kiecolt-Glaser (1986) USA
N
N
Dayalan (2010) India
Hughes (2003) USA
Other
Assessed HW compliance
First author (year) country
Table 1 continued
F
F
F
F
Delivery mode*
Y
N
N
Y
S/V skill practise
STAI-anx State CES-dep
DUKES anx DUKES dep
BSI
STAI-anx (state and trait)
Outcome measures
–
–
R
R
Meta-analysis^
p = 0.52** (-0.15, -0.58 to 0.29)** p \ 0.62** (0.11, -0.55 to 0.33)**.
p = 0.00** (25.08, 26.32 to 23.84)** p = 0.00** (25.93, 27.34 to 24.53)**
anxiety p = 0.13** (-0.51, -1.20 to 0.17)** depression p = 0.88** (0.05, -0.62 to 0.72)**
AT: state p = 0.00** (1.73, 22.53 to 20.92)** AT: trait p = 0.00** (21.11, 21.73 to 20.49)** LT: state p = 0.13** (0.43, -0.13 to 1.00)** LT trait p = 0.47** (-0.20; -0.76 to 0.36)**
Outcomes in favour of intervention
Group interventions to promote mental health in health in health…
123
123
N
Unno (2013) Japan
N/A
Delivery mode* N
S/V skill practise STAI anx
Outcome measures
–
Meta-analysis^
stress p = 0.46** (0.32; -0.56 to 1.21)**
Outcomes in favour of intervention
Theanine is an amino acid found in tea and the authors considered that it produces relaxation without sedation
Category of intervention: C (Cognitive-behavioural), M (Mindfulness), O (Other), P (Psychoeducational), R (Relaxation)
^^ kindly provided on request
** these values were estimated based on post intervention means and standard deviations for comparison groups
^C (Cognitive-behavioural), M (Mindfulness), P (Psychoeducation), R (Relaxation) meta-analysis
##
#
? data were not reported
* A (Alone, students worked off campus), F (Face to face), O (Online)
AT Autogenic training a technique for relaxation involving auto-suggestion to tell your body to relax and control breathing, blood pressure, heartrate or body temperature; BSI Brief Symptom Inventory; BECK Beck Anxiety/depression inventory; DASS Depression Anxiety Stress Scale full version with 42 items; DASS-21 Depression Anxiety Stress Scale Shortened version with 21 items; DUKES Duke’s Anxiety or depression scale; EMG electromyogram; ePREP electronic Prevention and Relationship Enhancement Program; FFMQ Five Facet Mindfulness Questionnaire; HW Homework; IPAT Institute for Personality and Ability Testing Depression Scale; MBI Maslach Burnout Inventory; MBSR Mindfulness Based Stress Reduction which combines psychoeducation, mindfulness-based cognitive-behavioural intervention, breath awareness, body awareness and hatha yoga postures with additional homework practise; MST Mind Sound Technology involving pronouncing sounds slowly with awareness on particular areas of the brain to induce relaxation. For example, a component of technique 1 involved making the sound ‘‘Ah’’ while becoming aware of the right eye and then on the left brain); PG Postgraduate; PMSS Perceived medical school stress; POMS Profile of Mood States; PSS Perceived Stress Scale; S selected student; SAS Self-rating anxiety scale; SCL-90 Symptom CheckList 90; SDS Self-rating depression scale; SOSI Symptoms of Stress Inventory; STAI State-Trait Anxiety inventory, state anxiety, is a temporary response to perceived threat and trait anxiety, a personality characteristic rather than a temporary condition (Spielberger et al. 1970); S/V Supervised; UG Undergraduate; V Volunteer student
bold indicates the statistically significant results
Assessed HW compliance
First author (year) country
Table 1 continued
K. Lo et al.
Group interventions to promote mental health in health in health…
Fig. 2 Forest plot of psychoeducational interventions compared to alternative education. Note the number in parentheses after the author and year denotes the quality assessment score out of a maximum of 10
One study for the effect on stress (319 participants) also reported no significant difference between groups (SMD -0.07; 95% CI -0.29 to 0.15).
Cognitive behavioural interventions (Porter and Johnson 2008; Frazier et al. 2015; Hintz et al. 2015) The results of cognitive behavioural interventions are summarised in Table 1. Figure 3 presents the pooled comparisons of cognitive-behavioural interventions to control conditions on the outcomes of depression, anxiety, stress and burnout. No effects were found for burnout (23 participants, SMD 0.22; 95% CI -0.6 to 1.04). Effects favoured the
Fig. 3 Forest plot of cognitive behavioural interventions versus control
123
K. Lo et al.
intervention for anxiety (278 participants, SMD -0.26; 95% CI -0.5 to -0.02; p = .03), depression (278 participants, SMD -0.29; 95% CI -0.52 to -0.05; p = .02) and stress (278 participants, SMD 0.37 95% CI -0.61 to -0.13; p = .002).
Mindfulness interventions (Shapiro et al. 1998; Jain et al. 2007; de Vibe et al. 2013; Erogul et al. 2014) The results of mindfulness interventions are summarised in Table 1. Figure 4 presents the pooled comparisons of mindfulness to no intervention control conditions of the outcomes of anxiety, depression, burnout and stress. No significant effects were observed for anxiety (480 participants were included, SMD -0.06; 95% CI -0.25 to 0.12; p = .49), depression (418 participants, SMD -0.16; 95% CI -0.36 to 0.03; p = .10) or burnout (144 participants, SMD -0.13; 95% CI -0.36 to 0.10). Significant effects in favour of a mindfulness intervention were observed (175 participants, SMD -0.54; 95% CI -0.85 to -0.24; p = .0004) Relaxation interventions (Kiecolt-Glaser et al. 1986; Johansson 1991; Heaman 1995; Whitehouse et al. 1996; Jones and Johnston 2000; Kanji et al. 2006; Jain et al. 2007; Dimou et al. 2014; Ratanasiripong et al. 2015) The results of relaxation interventions are summarised in Table 1. Results are pooled in meta-analysis in Fig. 5. Heaman (1995) reported on two non-intervention control groups and two experimental groups that participated at two different time points, one after oncampus lectures and one during clinical placements. Johansson (1991) evaluated the sophomore (junior) and senior groups independently. These subgroups were added independently to the meta-analysis as data were independent.
Fig. 4 Forest plot of mindfulness intervention versus control
123
Group interventions to promote mental health in health in health…
Fig. 5 Forest plot of relaxation intervention versus control
Significant effects were observed for anxiety (334 participants, SMD -0.80; 95% CI -1.03 to -0.58; p \ .0001), depression (110 participants, SMD -0.49; 95% CI -0.88 to -0.11, p = .01) and stress (146 participants, SMD -0.34; 95% CI -0.67 to -0.01; p = .04).
Other interventions (Dayalan et al. 2010; Unno et al. 2013) The results of other interventions are summarised in Table 1. For the Dayalan et al. study (2010) individual baselines scores were not reported as the authors pooled pre-test measurements across groups. Consequently, we were unable to confirm baseline comparability. As the studies were heterogeneous with respect to intervention types, the results could not be pooled in meta-analysis.
Educational intervention content Interventions included a range of educational content (Table 2). Most reports specified that psychoeducational content (20/24, 83%) and meditation (16/24, 67%) were included. Breath awareness (13/24, 54%) and relaxation/guided imagery were also common program elements (9/24, 38%).
Outcomes A wide range of mental health assessment tools were utilised to assess intervention effectiveness. Anxiety was assessed in 19 studies using one of six tools. The State-Trait Anxiety Inventory (STAI) (Spielberger et al. 1970) was the most commonly used assessment tool for anxiety, utilised in eight studies. There were seven different measures of depression, reported by 17 studies. The Depression Anxiety and Stress Scale (DASS)
123
123
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Braithwaite (2009)
Chen et al. (2013)
Dayalan (2010)
de Vibe et al. (2013)
Dimou et al. (2014)
Erogul et al. (2014)
Frazier et al. (2015)
Greene (1988)
Heaman (1995)
Hintz et al. (2015)
Hughes (2003)
Jain (2007)
Johansson (1991)
Jones (2000)
Kanji (2006)
X
Abaolshamat (2015)
Psychoeducation
X
X
X
X
Cognitivebehavioural intervention
X
X
X
X
Mindfulness-based cognitivebehavioural intervention
Table 2 Overview of educational intervention content
X
X
X
X
X
X
X
Relaxation
X
X
X
X
X
X
X
Static awareness of present moment e.g.: breath awareness
X
X
X
X
X
X
X
Static awareness of present moment body scan
X
X
X
X
X
X
X
X
X
Meditation/ guided imagery
X
X
X
Dynamic awareness of present moment e.g.yoga
K. Lo et al.
Erogul et al. (2014)
Dimou et al. (2014)
X
Static awareness of present moment body scan
Awareness of sound/making sound
X
de Vibe et al. (2013)
X
X
Problem solving
X
X
Peer support
Chen et al. (2013)
X
Group discussion
X
X
X
X
X
X
Static awareness of present moment e.g.: breath awareness
Dayalan (2010)
Braithwaite (2009)
Abaolshamat (2015)
Warnecke (2011)
Communication
X
Shapiro et al. (1998)
Unno (2013)
X
Raranasiripong (2015)
X
X
Porter (2008)
X
X
Paholpak (2012)
X
X
X
Relaxation
McGrady (2012)
Mindfulness-based cognitivebehavioural intervention X
Cognitivebehavioural intervention
Kiecolt-Glaser (1986)
Psychoeducation
Table 2 continued
X
X
X
X
X
Hypnosis
Meditation/ guided imagery
X
X
X
X
Other e.g. nutritional supplement, study skills
Dynamic awareness of present moment e.g.yoga
Group interventions to promote mental health in health in health…
123
123 X
X
Problem solving
Awareness of sound/making sound
X
Hypnosis
X
Other e.g. nutritional supplement, study skills
X = included
Warnecke (2011)
Unno (2013)
Shapiro et al. (1998)
Raranasiripong (2015)
Porter (2008)
Paholpak (2012)
McGrady (2012)
Kiecolt-Glaser (1986)
Kanji (2006)
Jones (2000)
Johansson (1991)
Jain (2007)
Hughes (2003)
X
X
X
X
X
X
X
X
X
X
Peer support
X
X
Group discussion
Hintz et al. (2015)
X
Communication
Heaman (1995)
Greene (1988)
Frazier et al. (2015)
Table 2 continued
K. Lo et al.
Group interventions to promote mental health in health in health…
(Ahmed et al. 2009) was the most commonly used scale, utilised in four studies. Data were extracted on stress from eight studies. Stress was primarily assessed using the Perceived Stress Scale (PSS) (Cohen et al. 1983). Burnout was assessed in one study using Maslach’s Burnout Inventory (MBI) (Maslach and Jackson 1981). No studies presented data the cost or cost-effectiveness of interventions. We note that no article described cost issues.
Quality assessment Study quality evaluations assessed using PEDro are presented in Table 3. Scores ranged from 1/10 to 7/10 (mean 3.58, S.D. 1.67). Key areas where rigour was lacking were in subject, therapist and assessor blinding, a limitation that may not be amenable to improvement in interventions of this nature. Only five studies met our criteria for attrition (at least one key outcome must be measured in more than 85% of participants who were initially allocated to groups). In five studies, data required to determine attrition were not reported and in 11 studies the attrition rate was higher than 15%. The range in overall attrition rate reported across the 13 studies was 6–50%. Two studies had high attrition rates: 49% from the control group in the study by McGrady et al. (2012) and 40% from the control group and 50% from the intervention group in the Dimou et al. study (2014). Only seven studies reported analysis by intention to treat or explicitly reported that participants received interventions as allocated. Eight studies adequately reported concealment of participants when determining group allocation. Fourteen studies reported between group differences. For the remaining studies we calculated these differences from reported data.
Data abstraction Risk of bias across studies Some studies (Greene and Hiebert 1988; Whitehouse et al. 1996) failed to provide data in a format to enable meta-analysis. Studies were also heterogeneous in terms of intervention and outcomes measures. Consequently only studies deemed adequately comparable could be included in the meta-analysis and this may have affected outcomes.
Discussion Summary of evidence Across the 24 papers included in this review, data were collected on 2491 health professional students relating to the impact of interventions to optimise mental health. Metaanalysis demonstrated that psychoeducational interventions appear to have no significant effect on anxiety, depression or stress in comparison to alternative education controls. This contradicts the meta-analysis into universal mental health prevention programs for higher education students performed by Conley et al. (2015) who concluded that psychoeducation was associated with modest significant effects on anxiety, stress, distress, self-perception and other academic measures but not on depression, socio-emotional skills or relationships. A key difference between these meta-analyses is that Conley, Durlak and Kirsch included
123
123
0
0
0
Jain (2007)
Johansson (1991)
Jones (2000)
1
1
1
Raranasiripong (2015)
Shapiro et al. (1998)
Unno (2013)
Warnecke (2011)
X = included, 0 = not included
0
0
Porter (2008)
0
0
McGrady (2012)
Paholpak (2012)
1
0
Hughes (2003)
0
0
Hintz et al. (2015)
Kiecolt-Glaser (1986)
0
Heaman (1995)
Kanji (2006)
0
Greene (1988)
1
0
Dimou et al. (2014)
1
1
de Vibe et al. (2013)
0
0
Dayalan (2010)
Erogul et al. (2014)
1
Chen et al. (2013)
Frazier et al. (2015)
1
0
1
0
1
1
0
1
0
1
1
1
1
1
1
1
1
1
1
1
0
1
0
1
1
Aboalshamat (2015)
Braithwaite (2009)
Baseline similarity
Concealment
Author
Table 3 Quality Assessment
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
1
Subject blinding
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Therapist blinding
1
1
1
0
0
1
0
0
0
0
0
0
1
0
0
0
0
0
0
1
0
1
0
1
Assessor blinding
0
0
1
1
0
0
0
1
0
1
0
0
0
1
1
1
0
1
0
1
1
0
0
0
Attrition
1
0
1
1
0
0
0
0
1
1
0
1
0
0
0
0
1
1
1
1
0
1
0
0
ITT
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Between group comparison
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Point estimates/measures of variation
7
5
7
5
2
4
2
4
5
5
3
4
4
4
4
4
4
6
4
7
3
6
2
6
TOTAL PEDro score
K. Lo et al.
Group interventions to promote mental health in health in health…
103 articles which comprised grey literature published between 1967 until the end of 2012 whereas we included only RCTs and ten of our included articles were published after 2012. We found that cognitive-behavioural interventions appear to be effective in modifying anxiety, depression and stress compared to alternative education controls. These findings are supported by other reviewers (Conley et al. 2015; Frazier et al. 2015; Hintz et al. 2015). The effect cognitive-behavioural interventions might have on reducing anxiety (SMD -0.77; 95% CI -0.97 to -0.57) was also supported by Regehr et al. (2013) and the effect on depression was supported by a Cochrane review of 15 trials (Jorm et al. 2008). Mindfulness interventions appear to have a significant effect in reducing stress in health professional students. No effects were seen on anxiety which may be due to reporting low baseline scores leaving potentially little room for improvement (McGrady et al. 2012) or the pre-intervention scores were not reported so differences in groups at baseline could not be established (Paholpak et al. 2012). It also may be that control conditions of nonmeditating activities such as napping, reading or chatting provided an intervention effect. No significant effect of mindfulness interventions were seen for depression or anxiety. In meta-analysis of MBSR in adults including both non-clinical and clinical-populations, Fjorback et al. (2011) concluded that MBSR can improve stress but can also impart additional health benefits such as reducing depression relapse and improving health related quality of life compared to control conditions. These were however mostly low quality studies. Our findings are not supported by Regehr et al. (2013) who conducted a metaanalysis of nine mindfulness interventions. When compared to control conditions they found significant reductions in anxiety (SDM -0.73; 95% CI -1.00 to -0.45) and depression (when combined with cognitive-behavioural interventions) (SDM -0.81; 95% CI -1.49 to -0.13). This may highlight the benefits of MBSR which includes mindfulness-based cognitive therapy rather than mindfulness alone (Regehr et al. 2013). The Regehr and colleagues research included both experimental and quasi-experimental methods. We included six studies that were published after the Regehr and colleagues’ review. Our meta-analysis differed in studying mindfulness in isolation while Regehr and colleagues pooled cognitive behavioural and mindfulness-based interventions for analysis when studying the effects on anxiety and depression. The authors justified this based on the argument that MBSR includes mindfulness based cognitive-behavioural interventions. We found that relaxation strategies may have a significant effect on anxiety, depression and stress in comparison to control conditions. This is supported by the meta-analysis by Manzoni et al. (2008) and a Cochrane systematic review of five trials (Jorm et al. 2008) that found reductions in self-reported depression for relaxation interventions compared to wait-list, no intervention control, or minimal intervention (SMD -0.59, 95% CI -0.94 to -0.24). Cognitive-behavioural and relaxation interventions may therefore be helpful in supporting anxiety, depression and stress, while mindfulness may help with stress. The implications of this review are that health professional programs might be encouraged to include cognitive-behavioural, relaxation strategies and mindfulness into curricula to support the mental health of health professional students.
Limitations This review was limited to published peer-reviewed articles that could result in publication bias. English language as an inclusion criterion also limits the review, particularly given that mindfulness and meditation practices originate from different cultures. It is possible that potentially relevant articles were not identified. The literature is limited by the number
123
K. Lo et al.
of RCTs in the health professional student population. Despite using the gold standard of RCTs, the quality of the articles varied widely with some low quality assessment scores thereby increasing the risk of bias and potentially compromising internal validity. We are therefore unable to draw robust conclusions from these articles. There are differences between the findings of systematic reviews in higher education and those of our study and such the findings may only be generalised to health professional students. Many papers were from Northern America and investigated primarily female medicine and nursing students, thus findings may not be generalizable to other health professional students. Studies often did not examine longer term effects. A number of studies advertised for volunteers and, given students with an interest in the area, this may bias outcomes. There were omissions of potential relevant information such as not including separate baseline data for control and intervention and not including attrition data. Most studies used selfrating psychological assessments and there may be a potential response bias favouring interventions given the participants were students of the institution in which the intervention was being studied. The number of studies with high attrition rates may indicate that interventions are not attractive to participants, and highlights the importance of designing programs that engage students. Despite the flaws, the included studies contain valuable information about the types of interventions that have been trialled. Studies with higher quality assessment scores of 7/10 (Shapiro et al. 1998; Warnecke et al. 2011; de Vibe et al. 2013) support positive effects of mindfulness on anxiety and depression but not on burnout and thus including only higher quality studies may have changed the conclusions of our meta-analysis.
Future recommendations Future recommendations include the need for high quality randomised control trials to determine long-term effects of interventions. Interventions that significantly modify burnout warrant attention. Future research might benefit from including records of student attendance and compliance with home practice to establish the potential confounding or influential effects this may have on outcomes. Studies may also benefit from selecting student participants rather than advertising for volunteers. As group interventions were investigated in non-clinical populations, conclusions may be augmented by research into effectiveness of interventions for learners with specific mental health diagnoses. Agreement on consistent outcome measures including physiological measures of stress would be beneficial to enable pooling of data in meta-analysis. Exploring strategies to support male students is an area that requires further development. Although we were interested in the cost-effectiveness of programs, studies did not report information about the cost of interventions and this is a target for consideration in future work. Two articles triangulated findings with physiological measures of stress such as salivary cortisol which decreases the response bias however, as yet there is no gold standard for physiological measures of stress. This would be an avenue for further investigation.
Conclusions Strategies are required to support student mental health. We explored what is known about the effects of group interventions enrolled in health professional education. Evidence suggests that interventions incorporating cognitive-behavioural interventions and
123
Group interventions to promote mental health in health in health…
relaxation may be valuable additions to health professional curricula to support a number of aspects of student mental health. Interventions that include mindfulness may improve stress. This review is a call to arms: given the importance of health professionals with robust mental health, we need more quality studies involving well matched control conditions, high quality trial design and long term outcome assessment. Collaboration is also required to agree on a suite of outcome measures as comparison of interventions is challenging with diverse outcome measures. Acknowledgements We acknowledge the Traditional Custodians of our land and pay our respects to their Elders, past and present. Thank you to Tonya Jones for her assistance in data extraction.
Appendix Medline search strategy (Ovid 1946–2016), April 2016 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.
student$.mp. mindful$.mp. resilien$.mp. burnout.mp. (wellness or wellbeing).mp. exp Mental Health/ calm.mp. exp Performance Anxiety exp ‘‘Quality of Life’’ 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 or 10 1 and 11
[mp = title, abstract, original title, name of substance word, subject heading word, keyword heading word, protocol supplementary concept, rare disease supplementary concept, unique identifier] PsycINFO (Ovid 1987–2016), 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.
student$.mp. mindful$.mp. resilien$.mp. burnout.mp. exp Mental Health (distress or stress).mp. calm.mp. exp Performance Anxiety exp ‘‘Quality of Life’’ 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 or 10 1 and 11
ALL EBM Reviews (1st Quarter 2016) 1. 2. 3. 4. 5.
student$.mp. mindful$.mp. resilien$.mp. burnout.mp. (wellness or wellbeing).mp.
123
K. Lo et al.
6. 7. 8. 9. 10. 11. 12.
exp Mental Health (distress or stress).mp calm.mp. exp Performance Anxiety exp ‘‘Quality of Life’’ 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 or 10 1 and 11
[mp = ti, ab, tx, kw, ct, ot, sh, hw] CINAHL Plus (Ebsco 1937–2016) 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.
mindful* distress stress wellness wellbeing calm ‘‘mental health’’ resilien* burnout ‘‘performance anxiety’’ ‘‘quality of life’’ 1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 or 10 or 11 student* 11 and 12 limit to RCT
ERIC (ProQuest 1945–2016), 1
(mindful* OR stress OR distress OR wellness OR wellbeing OR calm OR ‘‘mental health’’ OR resilien* OR burnout OR ‘‘performance anxiety’’ OR ‘‘quality of life’’) AND student* Limit to research OR numerical/quantitative data EMBASE (Elsevier 1957–2016). #18.1 mindful* #18.2 ‘distress’/exp OR distress OR ‘stress’/exp OR stress #18.3 ‘wellness’/exp OR wellness OR ‘wellbeing’/exp OR wellbeing #18.4 calm #18.5 ‘mental health’/exp OR ‘mental health’ #18.6 resilien* #18.7 ‘burnout’/exp OR burnout #18.8 ‘quality of life’/exp OR ‘quality of life’ #18.9 ‘performance anxiety’/exp OR ‘performance anxiety’ #18.10 (#18.1 OR #18.2 OR #18.3 OR #18.4 OR #18.5 OR #18.6 OR #18.7 OR #18.8 OR #18.9) #18.11 student* #18.12 (#18.10 AND #18.11) AND ‘randomized controlled trial’
123
Group interventions to promote mental health in health in health…
References Aboalshamat, K., Hou, X.-Y., & Strodl, E. (2015). The impact of a self-development coaching programme on medical and dental students’ psychological health and academic performance: A randomised controlled trial. BMC Medical Education, 15, 134. Ahmed, I., Banu, H., Al-Fageer, R., & Al-Suwaidi, R. (2009). Cognitive emotions: Depression and anxiety in medical students and staff. Journal of Critical Care, 24, e1–e7. Balogun, J. A., Titiloye, V., Balogun, A., Oyeyemi, A., & Katz, J. (2002). Prevalence and determinants of burnout among physical and occupational therapists. Journal of Allied Health, 31, 131–139. Braithwaite, S. R., & Fincham, F. D. (2009). A randomized clinical trial of a computer based preventive intervention: Replication and extension of ePREP. Journal of Family Psychology, 23, 32–38. Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3, 77–101. Chen, Y., Yang, X., Wang, L., & Zhang, X. (2013). A randomized controlled trial of the effects of brief mindfulness meditation on anxiety symptoms and systolic blood pressure in Chinese nursing students. Nurse Education Today, 33(10), 1166–1172. Cohen, J. (1988). Statistical power analysis for the behavioral sciences. 2nd edn. New Jersey: Hillsdale. Cohen, S., Kamarck, T., & Mermelstein, R. (1983). A global measure of perceived stress. Journal of Health and Social Behavior, 24, 385–396. Conley, C. S., Durlak, J. A., & Dickson, D. A. (2013). An evaluative review of outcome research on universal mental health promotion and prevention programs for higher education students. Journal of American College Health, 61, 286–301. Conley, C. S., Durlak, J. A., & Kirsch, A. C. (2015). A meta-analysis of universal mental health prevention programs for higher education students. Prevention Science, 16, 487–507. Dayalan, H., Subramanian, S., & Elango, T. (2010). Psychological well-being in medical students during exam stress-influence of short-term practice of mind sound technology. Indian Journal of Medical Sciences, 64, 501–507. de Morton, N. A. (2009). The PEDro scale is a valid measure of the methodological quality of clinical trials: A demographic study. Australian Journal of Physiotherapy, 55, 129–133. de Vibe, M., Solhaug, I., Tyssen, R., Friborg, O., Rosenvinge, J. H., Sorlie, T., et al. (2013). Mindfulness training for stress management: A randomised controlled study of medical and psychology students. BMC Medical Education, 13, 107. DiGiacomo, M., & Adamson, B. (2001). Coping with stress in the workplace: Implications for new health professionals. Journal of Allied Health, 30, 106–111. Dimou, P. A., Bacopoulou, F., Darviri, C., & Chrousos, G. P. (2014). Stress management and sexual health of young adults: A pilot randomised controlled trial. Andrologia, 46, 1022–1031. Dyrbye, L. N., Harper, W., Moutier, C., Durning, S., Power, D., Massie, F. S., et al. (2012). A multiinstitutional study exploring the impact of positive mental health on medical students’ professionalism in an era of high burnout. Academic Medicine, 87, 1024–1031. Dyrbye, L. N., Thomas, M. R., & Shanafelt, T. D. (2005). Medical student distress: Causes, consequences, and proposed solutions. Mayo Clinic Proceedings, 80, 1613–1622. Dyrbye, L. N., Thomas, M. R., & Shanafelt, T. D. (2006). Systematic review of depression, anxiety, and other indicators of psychological distress among U.S. and Canadian medical students. Academic Medicine, 81, 354–373. Erogul, M., Singer, G., McIntyre, T., & Stefanov, D. G. (2014). Abridged mindfulness intervention to support wellness in first-year medical students. Teaching and Learning in Medicine, 26, 350–356. Fahrenkopf, A. M., Sectish, T. C., Barger, L. K., Sharek, P. J., Lewin, D., Chiang, V. W., et al. (2008). Rates of medication errors among depressed and burnt out residents: Prospective cohort study. BMJ, 336, 488–491. Farrer, L., Gulliver, A., Chan, J. K., Batterham, P. J., Reynolds, J., Calear, A., et al. (2013). Technologybased interventions for mental health in tertiary students: Systematic review. Journal of Medical Internet Research, 15, e101. Fjorback, L. O., Arendt, M., Ornbol, E., Fink, P., & Walach, H. (2011). Mindfulness-based stress reduction and mindfulness-based cognitive therapy: A systematic review of randomized controlled trials. Acta Psychiatrica Scand, 124, 102–119. Frazier, P., Meredith, L., Greer, C., Paulsen, J. A., Howard, K., Dietz, L. R., et al. (2015). Randomized controlled trial evaluating the effectiveness of a web-based stress management program among community college students. Anxiety, Stress & Coping: An International Journal, 28, 576–586. Greene, Y., & Hiebert, B. (1988). A comparison of mindfulness meditation and cognitive self-observation. Canadian Journal of Counselling, 22, 25–34.
123
K. Lo et al. Green, S.& Higgins, J. P. T. (2011). In Cochrane handbook for systematic reviews of interventions version 5.1.0. (Vol. pp.) The Cochrane Collaboration Version. Heaman, D. (1995). The quieting response (QR): A modality for reduction of psychophysiologic stress in nursing students. Journal of Nursing Education, 34, 5–9. Hedges, L. V. (1981). Distribution theory for glass’s estimator of effect size and related estimators. Journal of Educational and Behavioral Statistics, 6, 107–128. Higgins, J. P., Thompson, S. G., Deeks, J. J., & Altman, D. G. (2003). Measuring inconsistency in metaanalyses. British Medical Journal, 327, 557. Hintz, S., Frazier, P. A., & Meredith, L. (2015). Evaluating an online stress management intervention for college students. Journal of Counseling Psychology, 62, 137. Hope, V., & Henderson, M. (2014). Medical student depression, anxiety and distress outside North America: A systematic review. Medical Education, 48, 963–979. Hughes, L. C., Romick, P., Sandor, M. K., Phillips, C. A., Glaister, J., Levy, K., et al. (2003). Evaluation of an informal peer group experience on baccalaureate nursing students’ emotional well-being and professional socialization. Journal of Professional Nursing, 19, 38–48. IsHak, W., Nikravesh, R., Lederer, S., Perry, R., Ogunyemi, D., & Bernstein, C. (2013). Burnout in medical students: A systematic review. The Clinical Teacher, 10, 242–245. Jain, S., Shapiro, S. L., Swanick, S., Roesch, S. C., Mills, P. J., Bell, I., et al. (2007). A randomized controlled trial of mindfulness meditation versus relaxation training: Effects on distress, positive states of mind, rumination, and distraction. Annals of Behavioral Medicine, 33, 11–21. Johansson, N. (1991). Effectiveness of a stress management program in reducing anxiety and depression in nursing students. Journal of American College Health, 40, 125–129. Jones, M. C., & Johnston, D. W. (2000). Evaluating the impact of a worksite stress management programme for distressed student nurses: A randomised controlled trial. Psychology & Health, 15, 689–706. Jorm, A. F., Morgan, A. J., & Hetrick, S. E. (2008). Relaxation for depression. The Cochrane Library. Kanji, N., White, A., & Ernst, E. (2006). Autogenic training to reduce anxiety in nursing students: Randomized controlled trial. Journal of Advanced Nursing, 53, 729–735. Kiecolt-Glaser, J. K., Glaser, R., Strain, E. C., Stout, J. C., Tarr, K. L., Holliday, J. E., et al. (1986). Modulation of cellular immunity in medical students. Journal of Behavioral Medicine, 9, 5–21. Lo, K., Curtis, H., Francis-Cracknell, A., Maloney, S., Keating, J., & Bearman, M. (2015). Plan B: Clinicians’ perceptions of an innovative Fitness to Practise program. In Association of medical educators Europe (AMEE) conference. (Ed.), (Vol. pp.) Glasgow September 5–9. Manzoni, G. M., Pagnini, F., Castelnuovo, G., & Molinari, E. (2008). Relaxation training for anxiety: A 10-years systematic review with meta-analysis. BMC Psychiatry, 8, 1–12. Maslach, C., & Jackson, S. E. (1981). The measurement of experienced burnout. Journal of Organizational Behavior, 2, 99–113. McGrady, A., Brennan, J., Lynch, D., & Whearty, K. (2012). A wellness program for first year medical students. Applied Psychophysiology & Biofeedback, 37, 253–260. Moher, D., Liberati, A., Tetzlaff, J., & Altman, D. G. (2009). Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. PLoS Med, 6, e1000097. Neumann, M., Edelha¨user, F., Tauschel, D., Fischer, M. R., Wirtz, M., Woopen, C., et al. (2011). Empathy decline and its reasons: A systematic review of studies with medical students and residents. Academic Medicine, 86, 996–1009. Paholpak, S., Piyavhatkul, N., Rangseekajee, P., Krisanaprakornkit, T., Arunpongpaisal, S., Pajanasoontorn, N., et al. (2012). Breathing meditation by medical students at Khon Kaen University: Effect on psychiatric symptoms, memory, intelligence and academic achievement. Journal of the Medical Association of Thailand, 95, 461–469. Porter, S., & Johnson, A. (2008). Increasing paramedic students’ resiliency to stress: Assessing correlates and the impact of intervention. College Quarterly, 11, 13. Ratanasiripong, P., Park, J. F., Ratanasiripong, N., & Kathalae, D. (2015). Stress and anxiety management in nursing students: Biofeedback and mindfulness meditation. Journal of Nursing Education, 54, 520–524. Regehr, C., Glancy, D., & Pitts, A. (2013). Interventions to reduce stress in university students: A review and meta-analysis. Journal of Affective Disorders, 148, 1–11. Roberts, L. W., Warner, T. D., Carter, D., Frank, E., Ganzini, L., & Lyketsos, C. (2000). Caring for medical students as patients: Access to services and care-seeking practices of 1027 students at nine medical schools. Academic Medicine, 75, 272–277. Rudman, A., & Gustavsson, J. P. (2012). Burnout during nursing education predicts lower occupational preparedness and future clinical performance: A longitudinal study. International Journal of Nursing Studies, 49, 988–1001.
123
Group interventions to promote mental health in health in health… Seritan, A., Hunt, J., Shy, A., Rea, M., & Worley, L. (2012). The state of medical student wellness: A call for culture change. Academic Psychiatry, 36(1), 7–10. Shapiro, S. L., Schwartz, G. E., & Bonner, G. (1998). Effects of mindfulness-based stress reduction on medical and premedical students. Journal of Behavioral Medicine, 21, 581–599. Shapiro, S. L., Shapiro, D. E., & Schwartz, G. E. (2000). Stress management in medical education: A review of the literature. Academic Medicine, 75, 748–759. Spielberger, C. D., Gorsuch, R. L., Lushene, R. E., & Vagg, P. R. (1970). State-trait anxiety inventory (STAI). BiB, 2010, 180. Stroebel, C. F. (1983). QR, the quieting reflex. Berkley Trade Pub: Berkley. Struber, J. C. (2003). Physiotherapy in Australia—Where to now. The internet Journal of Allied Health Sciences and Practice, 1, 2. Unno, K., Tanida, N., Ishii, N., Yamamoto, H., Iguchi, K., Hoshino, M., et al. (2013). Anti-stress effect of theanine on students during pharmacy practice: Positive correlation among salivary (alpha)-amylase activity, trait anxiety and subjective stress. Pharmacology, Biochemistry and Behavior, 111, 128–135. Van Der Heijden, F., Dillingh, G., Bakker, A., & Prins, J. (2008). Suicidal thoughts among medical residents with burnout. Arch Suicide Research, 12, 344–346. Warnecke, E., Quinn, S., Ogden, K., Towle, N., & Nelson, M. R. (2011). A randomised controlled trial of the effects of mindfulness practice on medical student stress levels. Medical Education, 45, 381–388. Whitehouse, W. G., Dinges, D. F., Orne, E. C., Keller, S. E., Bates, B. L., Bauer, N. K., et al. (1996). Psychosocial and immune effects of self-hypnosis training for stress management throughout the first semester of medical school. Psychosomatic Medicine, 58, 249–263. Williams, D., Tricomi, G., Gupta, J., & Janise, A. (2014). Efficacy of burnout interventions in the medical education pipeline. Academic Psychiatry, 39, 47–54. World Health Organisation (2014). Mental health: A state of wellbeing. In (ed.), (Vol. pp.).
123