www.thelancet.com Published online September 28, 2016 http://dx.doi.org/10.1016/S0140-6736(16)31279-X
1
Articles
Interventions to prevent and reduce physician burnout:
a systematic review and meta-analysis
Colin P West, Liselotte N Dyrbye, Patricia J Erwin, Tait D Shanafelt
Summary
Background Physician burnout has reached epidemic levels, as documented in national studies of both physicians in
training and practising physicians. The consequences are negative e ects on patient care, professionalism, physicians’
own care and safety, and the viability of health-care systems. A more complete understanding than at present of the
quality and outcomes of the literature on approaches to prevent and reduce burnout is necessary.
Methods In this systematic review and meta-analysis, we searched MEDLINE, Embase, PsycINFO, Scopus, Web of
Science, and the Education Resources Information Center from inception to Jan 15, 2016, for studies of interventions
to prevent and reduce physician burnout, including single-arm pre-post comparison studies. We required studies to
provide physician-specifi c burnout data using burnout measures with validity support from commonly accepted
sources of evidence. We excluded studies of medical students and non-physician health-care providers. We considered
potential eligibility of the abstracts and extracted data from eligible studies using a standardised form. Outcomes
were changes in overall burnout, emotional exhaustion score (and high emotional exhaustion), and depersonalisation
score (and high depersonalisation). We used random-e ects models to calculate pooled mean di erence estimates for
changes in each outcome.
Findings We identifi ed 2617 articles, of which 15 randomised trials including 716 physicians and 37 cohort studies
including 2914 physicians met inclusion criteria. Overall burnout decreased from 54% to 44% (di erence 10%
[95% CI 5–14]; p<0·0001; I=15%; 14 studies), emotional exhaustion score decreased from 23·82 points to 21·17 points
(2·65 points [1·67–3·64]; p<0·0001; I²=82%; 40 studies), and depersonalisation score decreased from 9·05 to 8·41
(0·64 points [0·15–1·14]; p=0·01; I²=58%; 36 studies). High emotional exhaustion decreased from 38% to 24% (14%
[11–18]; p<0·0001; I²=0%; 21 studies) and high depersonalisation decreased from 38% to 34% (4% [0–8]; p=0·04;
I²=0%; 16 studies).
Interpretation The literature indicates that both individual-focused and structural or organisational strategies can
result in clinically meaningful reductions in burnout among physicians. Further research is needed to establish
which interventions are most e ective in specifi c populations, as well as how individual and organisational solutions
might be combined to deliver even greater improvements in physician wellbeing than those achieved with individual
solutions.
Funding Arnold P Gold Foundation Research Institute.
Introduction
Physician burnout, a work-related syndrome involving
emotional exhaustion, depersonalisation, and a sense
of reduced personal accomplishment,
1
has reached
epidemic levels, with prevalences near or exceeding 50%,
as documented in national studies of both physicians in
training
2,3
and practising physicians.
4–6
Consequences are
negative e ects on patient care,
7–9
professionalism,
10,11
physicians’ own care and safety (including diverse issues
such as mental health concerns and motor vehicle
crashes),
12,13
and the viability of health-care systems,
including reductions in physicians’ professional work
e ort.
14,15
Evidence has linked 1 point changes in burnout
scores with meaningful di erences in self-perceived
major medical errors,
8,9
reductions in work hours,
15
and
suicidal ideation.
12
These concerns have prompted calls
for increased attention to physician wellbeing, including
e orts targeting burnout.
16–18
Both individual-focused and
structural or organisational solutions are required.
16
A more complete understanding than at present of the
quality and outcomes of the literature on approaches to
prevent and reduce burnout is necessary to understand
the best evidence for e ective interventions and to
establish a strong foundation for further research to fi ll
gaps in this literature.
Previous reviews of physician distress have been limited
in their ability to inform these issues by a combination of
factors, such as an absence of focus on physicians
and burnout and inconsistent adherence to modern
methodological systematic review standards.
19–22
Therefore,
we did a systematic review and meta-analysis adhering to
methodological standards to examine the literature to date
on interventions to prevent and reduce physician burnout.
Methods
Search strategy and selection criteria
In this systematic review and meta-analysis (reported
according to the Preferred Reporting Items for Systematic
Published Online
September 28, 2016
http://dx.doi.org/10.1016/
S0140-6736(16)31279-X
See
Online/Comment
http://dx.doi.org/10.1016/
S0140-6736(16)31332-0
Division of General Internal
Medicine and Division of
Biomedical Statistics and
Informatics (Prof C P West MD),
Division of Primary Care
Internal Medicine
(Prof L N Dyrbye MD), Medical
Library (P J Erwin MLS), and
Division of Hematology
(Prof T D Shanafelt MD), Mayo
Clinic, Rochester, MN, US
Correspondence to:
Prof Colin P West, Division of
General Internal Medicine and
Division of Biomedical Statistics
and Informatics, Mayo Clinic,
Rochester, MN 55905, USA
Articles
2
www.thelancet.com Published online September 28, 2016 http://dx.doi.org/10.1016/S0140-6736(16)31279-X
Reviews and Meta-Analyses statement
23
), we did a
literature search to identify studies of interventions to
prevent and reduce physician burnout, with the aid of an
experienced medical librarian (PJE). We included studies
collecting comparative data to assess the e ect of an
intervention on physician burnout, excluding studies
of medical students and non-physician health-care
providers. We included single-arm pre-post comparison
studies. We required studies to provide physician-specifi c
burnout data using burnout measures with validity
support from commonly accepted sources of evidence,
consisting of the domains of content, response process,
internal structure, relations to other variables, and
consequences.
24
We searched MEDLINE, Embase, PsycINFO, Scopus,
Web of Science, and the Education Resources
Information Center from inception to Jan 15, 2016.
Search terms included “burnout” and “stress”, along
with numerous other wellbeing-related terms. We
applied no language restrictions. The full search
strategies are detailed in the appendix. We also
reviewed the reference lists of eligible studies and
previous evidence summaries to identify additional
literature.
Two reviewers (CPW and LND or CPW and TDS)
working independently considered the potential eligibility
of each of the abstracts generated by the search strategy.
Full-text articles were obtained unless both reviewers
decided that an abstract was ineligible. Each full-text
report was assessed independently for fi nal study
inclusion. Disagreements about inclusion of full-text
articles were resolved by consensus and we measured
agreement on inclusion of full-text articles with the
κ statistic.
Data analysis
We extracted data using a standardised form to enter
intervention descriptions, study participant charac-
teristics, study design, and study results according to the
burnout metric applied in each study. Outcomes were
di erences between intervention groups in overall
burnout, emotional exhaustion score (and high emotional
exhaustion), and depersonalisation score (and high
deperson alisation). We extracted the SE of each outcome
measure directly or calculated it from relevant reported
statistical results, such as p values and CIs. Data
extraction was assessed by two reviewers (CPW and LND
or CPW and TDS), and disagreements were resolved by
consensus. We contacted authors of studies to obtain
missing data. Duplicate data were not an issue because
we specifi ed the timepoint closest to the conclusion of
the intervention for each study, so multiple reports from
the same cohort were never included in the same
analysis.
We used random-e ects models to calculate pooled
mean di erences using the generic inverse variance
method to incorporate heterogeneity related to di erent
interventions, settings, study designs, and burnout metrics
across studies. We scaled individual burnout domain scores
to the relevant full Maslach Burnout Inventory range (0–54
for emotional exhaustion score and 0–30 for de-
personalisation score).
1
When not reported, we calculated
SEs from available data.
Risk of bias was assessed by two reviewers (CPW and
LND or CPW and TDS) using Cochrane Collaboration risk
assessment tools for both randomised and observational
study designs.
25,26
Disagreements were resolved by
consensus. We measured heterogeneity using I. To
explore sources of heterogeneity, we prespecifi ed subgroup
See Online for appendix
Research in context
Evidence before this study
We searched MEDLINE, Embase, and PsycINFO from inception
to Jan 15, 2016, for previously published systematic reviews
and meta-analyses on interventions to prevent and reduce
physician burnout, using search terms including “burnout” and
“stress”, with no language restrictions. Previous reviews were
limited by an absence of focus on physicians or burnout and
inconsistent adherence to modern methodological systematic
review standards.
Added value of this study
Our study provides the most comprehensive systematic review
and meta-analysis to date of all studies assessing the eff ect of
interventions on burnout among physicians, summarising
results of 15 randomised controlled trials and 37 observational
studies. Our fi ndings emphasise that many individual-focused
and organisational interventions off er meaningful benefi t in
combating physician burnout. Eff ective individual-focused
strategies include mindfulness-based approaches, stress
management training, and small group curricula. Eff ective
organisational approaches include duty hour requirements and
locally developed modifi cations to clinical work processes.
Implications of all the available evidence
Our results substantiate that individual-focused interventions,
such as mindfulness, stress management, and small group
discussions, and structural or organisational interventions can be
eff ective approaches to reduce burnout domain scores. Duty hour
limitation policies appear eff ective, although, at present, these
results are derived only from observational studies in the USA.
The eff ect of individual-focused and structural or organisational
approaches in combination has not been studied, and which
classes of interventions might be most eff ective for specifi c groups
of physicians remains unknown. Additionally, further research is
needed to clarify optimal approaches to development and
implementation of interventions. Finally, sustainability of
intervention eff ects is poorly understood as few studies have
assessed long-term burnout outcomes.
Articles
www.thelancet.com Published online September 28, 2016 http://dx.doi.org/10.1016/S0140-6736(16)31279-X
3
analyses, consisting of comparisons of individual-focused
versus structural or organisational interventions, studies
of residents versus practising physicians, and di erent
study designs (randomised controlled trials vs observational
studies). Variability within studies is reported in the forest
plots and incorporated into the standard meta-analysis
statistics. We applied no other methods of within-study
variability assessment. We reported risk of bias, but did not
apply methods to account for potential bias beyond the
specifi ed subgroup analyses comparing results by study
design. We assessed publication bias by assessing funnel
plots for asymmetry.
Because of common interest in the e ect of duty hour
requirements on resident wellbeing and of mindfulness-
based and stress management-focused approaches, we
considered meta-analyses of these specifi c interventions
for each outcome. We based the primary analyses on the
fi rst study measurement after conclusion of the
intervention. However, because when the e ect of each
intervention might be maximised is unknown, for trials
reporting results at multiple timepoints, we did sensitivity
analyses including results from other timepoints for each
study in separate models. We used Review Manager 5.3
software for all analyses.
Role of the funding source
The funder of the study had no role in study design, data
collection, data analysis, data interpretation, or writing of
the report. The corresponding author had full access to all
the data in the study and CPW, LND, and TDS had fi nal
responsibility for the decision to submit for publication.
Results
Our search strategy identifi ed 2617 articles, of which
230 met the criteria for full-text review (fi gure 1). The
characteristics of the included studies are summarised in
the appendix. 15 randomised controlled trials including
716 physicians
27–41
and 37 unique cohort studies including
2914 physicians
42–79
met eligibility criteria. Agreement
between reviewers for study inclusion was high (κ=0·83).
Among the 15 randomised controlled trials, three involved
structural interventions within the work environment,
consisting of shortened attending rotation length,
33
various modifi cations to clinical work processes,
38
and
shortened resident shifts.
39
12 involved individual-focused
interventions, consisting of facilitated small group
curricula,
27,36,37,40,41
stress management and self-care
training,
28,30,32,34
communication skills training,
29,31
and a
so-called belonging intervention.
35
Four of these studies
indicated funding or coverage for physicians to participate
during the workday.
28,36,40,41
Seven studies involved resident
physicians (consisting of fi elds of internal medicine,
31,40
paediatrics,
30,32
and general surgery
35
[fi elds not reported
in two studies
37,39
]) and seven involved practising
physicians (consisting of fi elds of internal medicine or
primary care
33,36,38,41
and oncology
29
[fi elds not reported in
two studies
37,39
]).
Among the 37 cohort studies, 17 involved structural
interventions, consisting of USA duty hour require-
ments
45–47,50–52,54,61,63,78
and practice delivery chan ges.
49,53,65,66,68,71,79
20 involved individual-focused inter ventions, consisting
of facilitated and non-facilitated small group
curricula,
43,48,55,56,58–60,62,64,69,70,73,74
stress manage ment and self-
care training,
42,72
communication skills training,
43,58,69,73
and
mindfulness-based approaches.
44,57,67,75–77
Only four of the
cohort studies indicated funding or coverage for
physicians to participate during the workday.
49,60,64,76
19 studies involved resident phys icians (consisting of
fi elds of internal medicine,
46,47,65,78
surgical disciplines,
45,50,51,75
paediatrics,
54,61
obstetrics and gynaecology,
56,60
family
medicine,
44
neurology,
63
oncology,
64
and multiple spec-
ialties
52,72,76
[fi eld not reported in one study
67
]) and
20 involved practising physicians (consisting of fi elds of
internal medicine or primary care,
42,48,53,57,66,74,79
oncology,
43,58,73
intensive care,
49,71
surgical disciplines,
45,50
palliative
medicine,
77
and multiple specialties
55,59,62,67,70
).
All of the randomised controlled trials assessed results
immediately after the conclusion of the intervention.
Additional follow-up analyses were done in fi ve studies,
27–29,36,41
ranging from 19 weeks
27
to nearly 4 years
28
later. The
Maslach Burnout Inventory
1
was applied as the burnout
measure in all randomised studies but one (which used a
single-item measure assessing emotional exhaustion).
38
Among the cohort studies, additional follow-up analyses
were done in four studies,
57,59,76,77
occurring between
1 month
76
and 2 years
59
after the conclusion of the
intervention. All but three of the cohort studies
75,76,79
applied
the Maslach Burnout Inventory.
2617 potentially eligible studies
identified by database search
2387 excluded after full-text
screening
2617 identified for screening
230 reviewed in depth
29 studies excluded
13 did not assess a validated
burnout metric
8 did not assess a comparison or
intervention
6 did not provide analysable data
2 contained duplicate data
149 studies excluded
62 did not assess a comparison
or intervention
56 did not assess a validated
burnout metric
18 did not provide analysable data
8 did not include physicians
5 contained duplicate data
44 randomised controlled trials
reviewed
15 eligible studies 37 eligible studies
186 observational studies reviewed
Figure 1: Study selection
Articles
4
www.thelancet.com Published online September 28, 2016 http://dx.doi.org/10.1016/S0140-6736(16)31279-X
For the fi ve randomised controlled trials and nine
cohort studies reporting di erences in overall burnout,
the pooled mean di erence estimate was a signifi cant
absolute reduction from 54% to 44% (di erence 10%
[95% CI 5–14]; p<0·0001; I=15%; fi gure 2). Results did
not di er for randomised controlled trials versus
observational studies (p=0·60; I=0%) or for residents
versus practising physicians (p=0·86; I=0%), but
structural or organisational interventions were more
e ective than were individual-focused ones (p=0·03;
I=79%; appendix). The six studies of duty hour
requirements
46,47,52,54,61,78
yielded a similar estimated pooled
burnout reduction among residents from 62% to 50%
(12% [6–17]; p<0·0001; I²=28%). Only two studies of
mindfulness-based or stress management-focused
interventions addressed overall burnout,
32,72
with a similar
but non-signifi cant estimated reduction from 34% to
28% (6% [–2 to 14]; p=0·14; I²=0%).
For the 12 randomised controlled trials and 28 cohort
studies reporting di erences in emotional exhaustion score
as a continuous variable, the pooled mean di erence
estimate was a signifi cant 2·65 point reduction (95% CI
1·67–3·64; p<0·0001; I²=82%) in emotional exhaustion
domain score from 23·82 points to 21·17 points (fi gure 3).
Results did not di er for randomised controlled trials
versus observational studies (p=0·55; I²=0%), residents
versus practising physicians (p>0·99; I²=0%), or structural
or organisational versus individual-focused interventions
(p=0·69; I²=0%; appendix).
Within the cohort studies,
heterogeneity was smallest across the four studies assessing
the e ect of the 2003 duty hour require ments,
45,47,50,51
with a
mean reduction in emotional exhaustion score from
22·98 points to 20·10 points (di erence 2·88 points
[95% CI 1·17–4·59]; p=0·0001; I²=5%). The fi ve studies of
duty hour requirements overall
45,47,50,51,63
yielded a similar
but non-signifi cant pooled emotional exhaustion score
reduction estimate from 23·01 points to 20·49 points
(2·52 points [–0·28 to 5·31]; p=0·08; I²=49%). The
11 studies of mindfulness-based or stress management-
focused interventions
28,30,32,34,42,44,57,67,75–77
yielded a somewhat
greater estimated pooled score reduction than did those of
other interventions from 24·64 points to 19·96 points
(4·68 points [2·84–6·51]; p<0·0001; I²=47%).
For the 11 randomised controlled trials and 25 cohort
studies reporting di erences in depersonalisation score
as a continuous variable, the estimated pooled mean
di erence was a signifi cant 0·64 point reduction
(95% CI 0·15–1·14; p=0·01; I²=58%) in depersonalisation
domain score from 9·05 points to 8·41 points (fi gure 4).
Results did not di er for randomised controlled trials
versus observational studies (p=0·51; I²=0%), residents
versus practising physicians (p=0·91; I²=0%), or
structural or organisational versus individual-focused
interventions (p=0·33; I²=0%; appendix). Within the
cohort studies, heterogeneity was lowest across the four
studies assessing the e ect of the 2003 duty hour
requirements,
45,47,50,51
with a mean reduction in de-
personalisation score from 12·89 points to 11·36 points
(di erence 1·53 points [95% CI 0·24–2·81]; p=0·02;
I²=0%). The ten studies of mindfulness-based or stress
management-focused interventions
28,30,32,34,42,57,67,75–77
yielded
a somewhat greater estimated pooled score reduction
than did those of other interventions from 8·54 points to
6·53 points (2·01 points [1·34–2·67]; p<0·0001; I²=0%).
Mean difference
(% [95% CI])
Intervention
(n)
Control
(n)
RCTs
Martins et al (2011)
32
37
Lucas et al (2012)
33
62
West et al (2014)
36
34
West et al (2015)
41
51
Ripp et al (2015)
40
21
Subtotal 205
p=0·37; I
2
=45%
Cohort studies
Goitein et al (2005)
46
115
Gopal et al (2005)
47
121
Martini et al (2006)
52
28
Landrigan et al (2008)
54
114
Kim and Wiedermann (2011)
61
56
Quenot et al (2012)
68
4
Weight et al (2013)
72
174
Kotb et al (2014)
74
31
Ripp et al (2015)
78
108
Subtotal 751
p<0·0001; I
2
=0%
Total 956
p<0·0001; I
2
=15%
36
62
34
56
17
205
111
106
23
93
202
4
358
31
123
1051
1256
–8% (–39 to 22)
–19% (–30 to –8)
–18% (–334 to 298)
1% (–14 to 16)
9% (–13 to 30)
–6% (–19 to 7)
–8% (–20 to 3)
–6% (–13 to 0)
–31% (–61 to –1)
–18% (–32 to –5)
–21% (–36 to –7)
0% (–60 to 60)
–6% (–14 to 2)
–10% (–46 to 27)
–10% (–20 to 1)
–9% (–13 to –5)
–10% (–14 to –5)
0–1·0 1·00·5–0·5
Favours control
Favours intervention
Figure 2: Overall burnout
RCT=randomised controlled trial.
Articles
www.thelancet.com Published online September 28, 2016 http://dx.doi.org/10.1016/S0140-6736(16)31279-X
5
For the eight randomised controlled trials and 13 cohort
studies reporting di erences in high emotional ex-
haustion, the pooled mean di erence was a signifi cant
absolute reduction from 38% to 24% (di erence 14% [95%
CI 11–18]; p<0·0001; I²=0%; appendix). Results did not
di er for randomised controlled trials versus observational
studies (p=0·79; I²=0%) or for structural or organisational
versus individual-focused interventions (p=0·97; I²=0%),
but interventions among practising physicians were more
e ective than were those among residents (p=0·006;
I²=87%; appendix). The four studies of duty hour re-
quirements overall
45–47,78
yielded a similar pooled estimated
high emotional exhaustion reduction from 37% to
25% (12% [5–19]; p<0·0001; I²=0%). The four studies of
mindfulness-based or stress management-focused inter-
ventions
30,34,44,72
yielded a similar but non-signifi cant
estimated pooled high emotional exhaustion reduction
from 27% to 17% (10% [–1 to 21]; p=0·07; I²=27%).
For the six randomised controlled trials and ten cohort
studies reporting di erences in high depersonalisation,
the pooled mean di erence was a signifi cant absolute
reduction from 38% to 34% (di erence 4% [95% CI 0–8];
p=0·04; I²=0%; appendix). Results did not di er for
randomised controlled trials versus observational studies
(p=0·33; I²=0%), residents versus practising physicians
(p=0·34; I²=0%), or structural or organisational versus
Mean difference
(95% CI)
–8·30 (–17·14 to 0·54)
–5·83 (–11·79 to 0·13)
6·50 (–18·49 to 31·49)
–0·75 (–10·82 to 9·32)
0·90 (–2·20 to 4·00)
–2·67 (–6·49 to 1·15)
–7·47 (–15·29 to 0·35)
6·00 (–17·21 to 29·21)
–3·34 (–8·46 to 1·78)
–5·10 (–10·96 to 0·76)
–0·70 (–9·01 to 7·61)
0·22 (–3·07 to 3·51)
–2·06 (–3·86 to –0·27)
–4·50 (–8·62 to –0·38)
–6·51 (–11·23 to –1· 79)
1·14 (0·20 to 2·08)
–1·90 (–7·35 to 3·55)
–3·00 (–3·80 to –2·20)
–2·10 (–3·98 to –0·22)
0·20 (–0·37 to 0·77)
–5·20 (–10·41 to 0·01)
–6·00 (–10·82 to –1·18)
–6·00 (–9·63 to –2·37)
–4·23 (–5·46 to –3·00)
–6·80 (–8·80 to –4·80)
–1·00 (–3·84 to 1·84)
0·58 (–1·24 to 2·40)
4·00 (–7·39 to 15·39)
5·40 (–1·87 to 12·67)
–5·83 (–7·89 to –3·77)
–5·22 (–14·08 to 3·64)
–6·80 (–9·70 to –3·90)
1·48 (–2·71 to 5·67)
–1·40 (–3·07 to 0·27)
–0·60 (–2·72 to 1·52)
–6·30 (–9·77 to –2·83)
0·42 (–7·81 to 8·65)
–1·35 (–2·53 to –0·17)
–13·10 (–20·00 to –6·20)
0·65 (–2·82 to 4·12)
0·00 (–18·72 to 18·72)
–2·71 (–3·83 to –1·59)
–2·65 (–3·64 to –1·67)
Intervention
(n)
Control
(n)
RCTs
Oman et al (2006)
27
2 5
Rowe (2006)
28
5 3
Butow et al (2008)
29
16 14
Milstein et al (2009)
30
7 8
Bragard et al (2010)
31
49 47
Martins et al (2011)
32
37 36
Salles et al (2013)
35
13 14
Moody et al (2013)
34
5 5
West et al (2014)
36
34 34
Parshuram et al (2015)
39
14 13
Gunasingam et al (2015)
37
13 18
West et al (2015)
41
51 56
Subtotal 246 253
p=0·02; I
2
=15%
Cohort studies
Winefield et al (1998)
42
19 19
Ospina-Kammerer and Figley (2003)
44
14 10
Fujimori et al (2003)
43
58 58
Gelfand et al (2004)
45
26 26
Sluiter et al (2005)
49
4 4
Gopal et al (2005)
47
121 106
Barrack et al (2006)
50
* 23 20
Barrack et al (2006)
50
† 21 34
Hutter et al (2006)
51
35 35
Dunn et al (2007)
53
27 30
Rø et al (2008)
55
168 168
Krasner et al (2009)
57
56 56
Ghetti et al (2009)
56
17 17
Bragard et al (2010)
58
62 62
Winkel et al (2010)
60
18 18
Schuh et al (2011)
63
23 23
Meerten et al (2011)
62
79 79
Erler et al (2012)
66
6 6
Goodman and Schorling (2012)
67
40 40
Bar-Sela et al (2012)
64
15 15
Clayton et al (2013)
69
21 21
Giannini et al (2013)
71
71 71
Rosdahl and Kingsolver (2014)
75
5 5
Kotb et al (2014)
74
31 31
Fujimori et al (2014)
73
16 16
Warde et al (2015)
79
7 7
Podgurski et al (2015)
77
17 17
Goldhagen et al (2015)
76
30 30
Subtotal 1030 1024
p<0·0001; I
2
=87%
Total 1276 1277
p<0·0001; I
2
=82%
0–20 2010–10
Favours control
Favours intervention
Figure 3: Emotional exhaustion score
RCT=randomised controlled trial. *Staff . †Residents.
Articles
6
www.thelancet.com Published online September 28, 2016 http://dx.doi.org/10.1016/S0140-6736(16)31279-X
individual-focused interventions (p=0·61; I²=0%;
appendix). The four studies of duty hour requirements
overall
45–47,78
yielded a similar estimated pooled high
depersonalisation reduction from 54% to 48% (6% [0–13];
p=0·04; I²=0%). The three studies assessing the e ect
of mindfulness-based or stress management-focused
interventions on high depersonalisation
30,34,72
yielded a
similar but non-signifi cant pooled reduction estimate
from 21% to 16% (5% [–2 to 12]; p=0·13; I²=0%).
Of the 52 included studies, 47 (90%) reported no adverse
events associated with the examined interventions. One
study
33
reported negative e ects of short attending
rotations on resident assessments of faculty and four
studies
46,47,51,63
reported negative e ects of duty hour
requirements on subjectively assessed resident skills,
resident education, and patient care (table). The assessed
risk of bias for each included study is shown in the
appendix. The randomised studies inconsistently reported
details of randomisation processes and uniformly lacked
masking of participants to the interventions, as would be
expected because of the nature of the interventions. Other
potential biases were generally addressed well, and overall
risk of bias appeared similar between studies. The
observational studies were generally markedly limited by
potential for confounding as most of these studies
involved a pre-post assessment without a separate control
group. Publication bias was not evident for any outcome
as assessed through examination of symmetry in funnel
plots (data not shown). For the separate meta-analyses of
duty hour requirements and mindfulness-based and
stress management-focused approaches and for sensitivity
analyses including results from other timepoints for each
Mean difference
(95% CI)
–3·30 (–9·26 to 2·66)
–0·88 (–3·66 to 1·90)
–2·50 (–7·11 to 2·11)
3·75 (–2·99 to 10·49)
0·40 (–0·80 to 1·60)
–2·86 (–4·96 to –0·76)
1·50 (–7·46 to 10·46)
–1·10 (–2·77 to 0·57)
0·30 (–4·85 to 5·45)
–0·14 (–1·71 to 1·43)
–3·60 (–6·79 to –0·41)
–0·92 (–1·90 to 0·05)
–1·10 (–5·22 to 3·02)
1·11 (–0·71 to 2·93)
2·30 (–4·29 to 8·89)
–1·20 (–2·75 to 0·35)
2·73 (1·22 to 4·24)
–3·00 (–6·37 to 0·37)
–2·70 (–6·23 to 0·83)
0·00 (–15·56 to 15·56)
–0·80 (–1·17 to –0·43)
–2·50 (–3·60 to –1·40)
1·00 (–1·84 to 3·84)
0·00 (–6·33 to 6·33)
0·42 (–0·56 to 1·40)
–0·53 (–1·67 to 0·61)
4·10 (0·63 to 7·57)
–2·50 (–3·87 to –1·13)
–0·27 (–1·03 to 0·49)
–1·50 (–3·36 to 0·36)
–1·00 (–2·08 to 0·08)
–0·95 (–3·54 to 1·64)
2·61 (–0·53 to 5·75)
–4·13 (–8·44 to 0·18)
–2·00 (–5·16 to 1·16)
0·00 (–10·04 to 10·04)
–0·65 (–2·34 to 1·04)
–0·54 (–1·13 to 0·04)
–0·64 (–1·14 to –0·15)
Intervention
(n)
Control
(n)
RCTs
Oman et al (2006)
27
2 5
Rowe (2006)
28
5 3
Butow et al (2008)
29
16 14
Milstein et al (2009)
30
7 8
Bragard et al (2010)
31
49 47
Martins et al (2011)
32
37 36
Moody et al (2013)
34
5 5
West et al (2014)
36
34 34
Gunasingam et al (2015)
37
13 18
West et al (2015)
41
51 56
Parshuram et al (2015)
39
14 13
Subtotal 233 239
p=0·06; I
2
=31%
Cohort studies
Winefield et al (1998)
42
19 19
Fujimori et al (2003)
43
58 58
Gelfand et al (2004)
45
26 26
Gopal et al (2005)
47
121 106
Margalit et al (2005)
48
44 44
Hutter et al (2006)
51
35 35
Barrack et al (2006)
50
21 34
Dunn et al (2007)
53
27 30
Rø et al (2008)
55
166 166
Krasner et al (2009)
57
56 56
Ghetti et al (2009)
56
17 17
Winkel et al (2010)
60
18 18
Bragard et al (2010)
58
62 62
Meerten et al (2011)
62
79 79
Schuh et al (2011)
63
23 23
Goodman and Schorling (2012)
67
40 40
Bar-Sela et al (2012)
64
15 15
Erler et al (2012)
66
6 6
Giannini et al (2013)
71
71 71
Clayton et al (2013)
69
21 21
Kotb et al (2014)
74
31 31
Fujimori et al (2014)
73
16 16
Rosdahl and Kingsolver (2014)
75
5 5
Goldhagen et al (2015)
76
30 30
Podgurski et al (2015)
77
17 17
Subtotal 1024 1025
p=0·07; I
2
=65%
Total
p=0·01; I
2
=58% 1257 1264
0–20 2010–10
Favours control
Favours intervention
Figure 4: Depersonalisation score
RCT=randomised controlled trial.
Articles
www.thelancet.com Published online September 28, 2016 http://dx.doi.org/10.1016/S0140-6736(16)31279-X
7
study in separate models, none of the results di ered
substantially from the primary analysis results (data not
shown).
Discussion
Most studies in this systematic review and meta-analysis
reported on changes in burnout domain scores, fi nding a
signifi cant reduction in emotional exhaustion and
depersonalisation scores. Fewer studies reported on
changes in overall burnout or high burnout levels in each
domain than on changes in burnout domain scores,
fi nding a signifi cant reduction in absolute burnout
and in a high degree of emotional exhaustion and
depersonalisation. These e ects were consistent between
randomised controlled trials and observational studies,
allowing pooling of results across the full range of
eligible studies. Results were also similar for individual-
focused and structural or organisational interventions
for all outcomes other than overall burnout and for
practising physicians and residents for all outcomes
other than high emotional exhaustion. Heterogeneity
across all studies for each of these outcomes was low, but
the I values for these subgroup analyses were high,
suggesting that these results might refl ect genuine
subgroup di erences worthy of further exploration.
If applied to 2014 national data for US physicians,
6
an
absolute reduction in burnout of 10% (from 54% to 44%)
would represent an 18% relative risk reduction in
burnout. An absolute reduction in high degree of
emotional exhaustion of 14% (from 47% to 33%) would
represent a 30% relative risk reduction and an absolute
reduction in high degree of depersonalisation of 4%
(from 35% to 31%) would represent a 12% relative risk
reduction. These e ects would return burnout in each
domain to levels near or even below those previously
reported from 2011 national data.
5
Although the
magnitude of the reductions in burnout domain scores
appears modest, evidence has linked 1 point changes in
burnout scores with meaningful di erences in important
adverse outcomes.
8,9,12,15
Additionally, the cuto s between
average and high burnout scores span narrow ranges.
For example, a high depersonalisation burnout score is
10 or greater, with a fairly narrow range of average
depersonalisation burnout scores of 6–9.
1
Therefore,
reductions of only 1 or 2 points could o er benefi ts across
the full continuum of burnout scores and could signal
meaningful shifts in burnout severity category. This
point is illustrated by the three most precise studies
36,41,47
reporting di erences in both mean depersonalisation
score and high depersonalisation. Investigators of these
studies found reductions in mean depersonalisation
score of 1·2 points or fewer, which translated to 6–17%
reductions in absolute proportions of a high degree of
depersonalisation.
Our results substantiate that both individual-focused
and structural or organisational interventions can
reduce physician burnout. Although no specifi c
physician burnout interventions have been shown to be
better than are other interventions, both strategies are
probably necessary. However, their combination has
not been studied. The most commonly studied
interventions have involved mindfulness, stress
management, and small group discussions, and the
results suggest that these strategies can be e ective
approaches to reduce burnout domain scores. Duty
hour limitation policies also appear e ective, although,
at present, these results are derived only from
observational studies in the USA.
Various carefully planned approaches seem useful,
which is reassuring for individuals and organisations
contemplating tackling physician burnout. However,
this study makes clear that much additional research
into interventions for physician burnout is necessary.
For example, although heterogeneity in results across
intervention types was generally modest, data are
insu cient to fully delineate which classes of inter-
ventions might be most e ective. Randomised studies of
structural or organisational interventions have been
uncommon, with only three
33,38,39
reported in the
literature to date. Additional studies are particularly
Intervention Physician sample demographics Adverse event description
Lucas et al (2012)
33
Short (2 week) inpatient attending
rotation (4 week rotation control)
n=62 (crossover trial); 52% men; median age 38 years
(range 29–55); general medicine inpatient attending
physicians; USA
Decreased perceived ability by residents and medical students of
attendings to fairly assess trainees and decreased summary
assessments of attendings by medical students
Goitein et al (2005)
46
2003 USA DHR pre-DHR n=115; post-DHR n=111; 47% men; mean age NR;
internal medicine residents; USA
Increased negative reported eff ects on patient care and resident
education
Gopal et al (2005)
47
2003 USA DHR pre-DHR n=121; post-DHR n=106; pre-DHR 48% men;
mean age NR; post-DHR 42% men; internal medicine
residents; USA
Decreased overall resident satisfaction with the training programme
Hutter et al (2006)
51
2003 USA DHR n=35; sample demographics NR; surgical residents; USA Decreased assessment by faculty of residents’ technical skills, clinical
judgment, effi ciency, and professionalism
Schuh et al (2011)
63
2008 USA DHR (2003 USA DHR
control); 1 month under each DHR
set
n=23; mean age 30 years (SD 3); 44% men; neurology
residents; USA
Decreased resident assessment of ability to provide continuity of care
and of knowledge of patients and decreased faculty assessments of
residents’ clinical skills and patient care
DHR=duty hour requirements. NR=not reported.
Table: Adverse events
Articles
8
www.thelancet.com Published online September 28, 2016 http://dx.doi.org/10.1016/S0140-6736(16)31279-X
needed in this domain. Which interventions o er the
greatest value to physicians and their organisations
remains unclear, as well as whether or not the processes
involved in development and deployment of inter-
ventions could infl uence their e ectiveness. For
example, relative to externally developed approaches,
interventions for which physicians in the local work
environment are engaged in design and implementation
might heighten their sense of control and engagement,
which might be expected to e ectively reduce burnout.
80
Our data are not adequate to address this hypothesis,
however. Future research into organisational inter-
ventions to reduce physician burnout should address the
optimal approaches to development and implementation
of burnout reduction strategies, along with assessment
of the feasibility and costs associated with these
interventions.
Additionally, few studies have assessed long-term
or post-intervention e ects. The results of these
assessments generally suggest sustained or even aug-
mented benefi ts for many months after completion of
the studied intervention,
28,36,41,55,57,59,77
but this fi nding is not
universal.
27
Whether or not potentially benefi cial
interventions require periodic re-exposure to sustain or
maximise their e ects or how frequently such re-
exposure should occur is unknown.
This study has limitations. Data for participant
demographics were only sporadically reported in the
included studies, and the possibility of di ering inter-
vention e ects for di erent participant subgroups
remains largely unaddressed. Also, many of the
included cohort studies had substantial risk of bias,
largely due to low ability to control for potential
confounding factors. However, the overall quality of the
randomised trials in this review was moderate, and
despite their methodological di erences and limitations,
the observational studies and randomised trials yielded
statistically similar results.
Additional research is needed to clarify categories of
benefi cial interventions to reduce physician burnout,
which interventions or combinations of interventions
might be most e ective, and optimal approaches to
development and implementation of these interventions.
Rigorous, well-designed, generalisable studies addressing
these questions are now needed to build on this early
foundation of evidence to expand understanding of inter-
ventions to address the pervasive problem of physician
burnout.
Contributors
CPW, LND, and TDS designed the study and acquired, analysed, and
interpreted data. PJE did the literature search and interpreted data.
CPW drafted the manuscript, with critical revisions for important
intellectual content from all authors.
Declaration of interests
We declare no competing interests.
Acknowledgments
This study was supported by a grant from the Arnold P Gold Foundation
Research Institute.
References
1 Maslach C, Jackson SE, Leiter MP. Maslach burnout inventory
manual, 3rd edn. Palo Alto: Consulting Psychologists Press, 1996.
2 Dyrbye LN, Thomas MR, Massie FS Jr, et al. Burnout and suicidal
ideation among U.S. medical students. Ann Intern Med 2008;
149: 334–41.
3 West CP, Shanafelt TD, Kolars JC. Quality of life, burnout,
educational debt, and medical knowledge among internal medicine
residents. JAMA 2011; 306: 952–60.
4 Shanafelt TD, Balch CM, Bechamps GJ, et al. Burnout and career
satisfaction among American surgeons. Ann Surg 2009;
250: 463–71.
5 Shanafelt TD, Boone S, Tan L, et al. Burnout and satisfaction with
work-life balance among US physicians relative to the general
US population. Arch Intern Med 2012; 172: 1377–85.
6 Shanafelt TD, Hasan O, Dyrbye LN, et al. Changes in burnout and
satisfaction with work-life balance in physicians and the general
US working population between 2011 and 2014. Mayo Clin Proc 2015;
90: 1600–13.
7 Fahrenkopf AM, Sectish TC, Barger LK, et al. Rates of medication
errors among depressed and burnt out residents: prospective cohort
study. BMJ 2008; 336: 488–91.
8 West CP, Tan AD, Habermann TM, Sloan JA, Shanafelt TD.
Association of resident fatigue and distress with perceived medical
errors. JAMA 2009; 302: 1294–300.
9 Shanafelt TD, Balch CM, Bechamps G, et al. Burnout and medical
errors among American surgeons. Ann Surg 2010; 251: 995–1000.
10 West CP, Shanafelt TD. Physician well-being and professionalism.
Minn Med 2007; 90: 44–46.
11 Dyrbye LN, Massie FS Jr, Eacker A, et al. Relationship between
burnout and professional conduct and attitudes among US medical
students. JAMA 2010; 304: 1173–80.
12 Shanafelt TD, Balch CM, Dyrbye LN, et al. Special report: suicidal
ideation among American surgeons. Arch Surg 2011; 146: 54–62.
13 West CP, Tan AD, Shanafelt TD. Association of resident fatigue
and distress with occupational blood and body fl uid exposures and
motor vehicle incidents. Mayo Clin Proc 2012; 87: 1138–44.
14 Dyrbye LN, Shanafelt TD. Physician burnout: a potential threat to
successful health care reform. JAMA 2011; 305: 2009–10.
15 Shanafelt TD, Mungo M, Schmitgen J, et al. Longitudinal study
evaluating the association between physician burnout and changes
in professional work e ort. Mayo Clin Proc 2016; 91: 422–31.
16 Wallace JE, Lemaire JB, Ghali WA. Physician wellness: a missing
quality indicator. Lancet 2009;
374: 1714–21.
17 Bodenheimer T, Sinsky C. From Triple to Quadruple Aim:
care of the patient requires care of the provider. Ann Fam Med 2014;
12: 573–76.
18 Sikka R, Morath JM, Leape L. The Quadruple Aim: care, health,
cost and meaning in work. BMJ Qual Saf 2015; 24: 608–10.
19 Thomas NK. Resident burnout. JAMA 2004; 292: 2880–89.
20 Awa WL, Plaumann M, Walter U. Burnout prevention: a review of
intervention programs. Patient Educ Couns 2010; 78: 184–90.
21 Regehr C, Glancy D, Pitts A, LeBlanc VR. Interventions to reduce
the consequences of stress in physicians: a review and
meta-analysis. J Nerv Ment Dis 2014; 202: 353–59.
22 Ruotsalainen JH, Verbeek JH, Marine A, Serra C.
Preventing occupational stress in healthcare workers.
Cochrane Database Syst Rev 2015; 4: CD002892.
23 Moher D, Linerati A, Tetzla J, Altman DG. Preferred reporting
items for systematic reviews and meta-analyses: the PRISMA
statement. BMJ 2009; 339: b2535.
24 Cook DA, Beckman TJ. Current concepts in validity and reliability
for psychometric instruments. Am J Med 2006; 119: 166e7–16.
25 Higgins JP, Altman DG, Gotzsche PC, et al. The Cochrane
Collaboration’s tool for assessing risk of bias on randomised trials.
BMJ 2011; 343: d5928.
26 Sterne JA, Higgins JP, Reeves BC, for the development group for
ACROBAT-NRSI. A Cochrane Risk Of Bias Assessment Tool: for Non-
Randomised Studies of Interventions (ACROBAT-NRSI), Version 1.0.0.
September, 2014. http://www.riskofbias.info (accessed May 29, 2016).
27 Oman D, Hedberg J, Thoresen CE. Passage meditation reduces
perceived stress in health professionals: a randomized, controlled
trial. J Consult Clin Psychol 2006; 74: 714–19.
Articles
www.thelancet.com Published online September 28, 2016 http://dx.doi.org/10.1016/S0140-6736(16)31279-X
9
28 Rowe MM. Four-year longitudinal study of behavioral changes in
coping with stress. Am J Health Behav 2006; 30: 602–12.
29 Butow P, Cockburn J, Girgis A, et al. Increasing oncologists’ skills
in eliciting and responding to emotional cues: evaluation of a
communication skills training program. Psychooncology 2008;
17: 209–18.
30 Milstein JM, Raingruber BJ, Bennett SH, Kon AA, Winn CA,
Paterniti DA. Burnout assessment in house o cers: evaluation of
an intervention to reduce stress. Med Teach 2009; 31: 338–41.
31 Bragard I, Etienne A, Merckaert I, Libert Y, Razavi D. E cacy of
a communication and stress management training on medical
residents’ self-e cacy, stress to communicate and burnout.
J Health Psychol 2010; 15: 1075–81.
32 Martins AE, Davenport MC, Del Valle MP, et al. Impact of a brief
intervention on the burnout levels of pediatric residents.
J Pediatr (Rio J) 2011; 87: 493–98.
33 Lucas BP, Trick WE, Evans AT, et al. E ects of 2- vs 4-week
attending physician inpatient rotations on unplanned patient
revisits, evaluations by trainees, and attending physician burnout:
a randomized trial. JAMA 2012; 308: 2199–207.
34 Moody K, Kramer D, Santizo RO, et al. Helping the helpers:
mindfulness training for burnout in pediatric oncology—a pilot
program. J Pediatr Oncol Nurs 2013; 30: 275–84.
35 Salles A, Nandagopal K, Walton G. Belonging: a simple, brief
intervention decreases burnout. J Am Coll Surg 2013; 217: S116.
36 West CP, Dyrbye LN, Rabatin JT, et al. Intervention to promote
physician well-being, job satisfaction, and professionalism:
a randomized clinical trial. JAMA Intern Med 2014; 174: 527–33.
37 Gunasingam N, Burns K, Edwards J, Dinh M, Walton M.
Reducing stress and burnout in junior doctors: the impact of
debriefi ng sessions. Postgrad Med J 2015; 91: 182–87.
38 Linzer M, Poplau S, Grossman E, et al. A cluster randomized trial
of interventions to improve work conditions and clinician burnout
in primary care: results from the Health Work Place (HWP) study.
J Gen Intern Med 2015; 30: 1105–11.
39 Parshuram CS, Amaral AC, Ferguson ND, et al. Patient safety,
resident well-being and continuity of care with di erent resident
duty schedules in the intensive care unit: a randomized trial.
CMAJ 2015; 187: 321–29.
40 Ripp J, Fallar R, Korenstein DR. A randomized controlled trial to
decrease job burnout in fi rst-year internal medicine residents using
a facilitated discussion group intervention. J Gen Intern Med 2015;
30: S90.
41 West CP, Dyrbye LN, Satele D, Shanafelt TD. A randomized
controlled trial evaluating the e ect of COMPASS
(Colleagues Meeting to Promote And Sustain Satisfaction) small
group sessions on physician well-being, meaning, and job
satisfaction. J Gen Intern Med 2015; 30: S89.
42 Winefi eld H, Farmer E, Denson L. Work stress management for
women general practitioners: an evaluation. Psychol Health Med
1998; 3:
163–70.
43 Fujimori M, Oba A, Koike M, et al. Communication skills training
for Japanese oncologists on how to break bad news. J Cancer Educ
2003; 18: 194–201.
44 Ospina-Kammerer V, Figley CR. An evaluation of the Respiratory
One Method (ROM) in reducing emotional exhaustion among
family physician residents. Int J Emerg Ment Health 2003; 5: 29–32.
45 Gelfand DV, Podnos YD, Carmichael JC, Saltzman DJ, Wilson SE,
Williams RA. E ect of the 80-hour workweek on resident burnout.
Arch Surg 2004; 139: 933–40.
46 Goitein L, Shanafelt TD, Wipf JE, Slatore CG, Back AL. The e ects
of work-hour limitations on resident well-being, patient care, and
education in an internal medicine residency program.
Arch Intern Med 2005; 165: 2601–06.
47 Gopal R, Glasheen JJ, Miyoshi TJ, Prochazka AV. Burnout and
internal medicine resident work-hour restrictions. Arch Intern Med
2005; 165: 2595–600.
48 Margalit AP, Glick SM, Benbassat J, Cohen A, Katz M.
Promoting a biopsychosocial orientation in family practice: e ect of
two teaching programs on the knowledge and attitudes of practising
primary care physicians. Med Teach 2005; 27: 613–18.
49 Sluiter JK, Bos AP, Tol D, Cal M, Krijnen M, Frings-Dresen MH.
Is sta well-being and communication enhanced by multidisciplinary
work shift evaluations? Intensive Care Med 2005; 31: 1409–14.
50 Barrack RL, Miller LS, Sotile WM, Sotile MO, Rubash HE. E ect of
duty hour standards on burnout among orthopaedic surgery
residents. Clin Orthop Relat Res 2006; 449: 134–37.
51 Hutter MM, Kellogg KC, Ferguson CM, Abbott WM, Warshaw AL.
The impact of the 80-hour resident workweek on surgical residents
and attending surgeons. Ann Surg 2006; 243: 864–75.
52 Martini S, Arfken CL, Balon R. Comparison of burnout among
medical residents before and after the implementation of work
hours limits. Acad Psychiatry 2006; 30: 352–55.
53 Dunn PM, Arnetz BB, Christensen JF, Homer L. Meeting the
imperative to improve physician well-being: assessment of an
innovative program. J Gen Intern Med 2007; 22: 1544–52.
54 Landrigan CP, Fahrenkopf AM, Lewin D, et al. E ects of the
Accreditation Council for Graduate Medical Education duty hour
limits on sleep, work hours, and safety. Pediatrics 2008;
122: 250–58.
55 Rø KE, Gude T, Tyssen R, Aasland OG. Counselling for burnout
in Norwegian doctors: one year cohort study. BMJ 2008;
337: a2004.
56 Ghetti C, Chang J, Gosman G. Burnout, psychological skills, and
empathy: Balint training in obstetrics and gynecology residents.
J Grad Med Educ 2009; 1: 231–35.
57 Krasner MS, Epstein RM, Beckman H, et al. Association of an
educational program in mindful communication with burnout,
empathy, and attitudes among primary care physicians. JAMA 2009;
302: 1284–93.
58 Bragard I, Libert Y, Etienne A, et al. Insight on variables leading to
burnout in cancer physicians. J Cancer Educ 2010; 25: 109–15.
59 Isaksson Ro KE, Tyssen R, Ho art A, Sexton H, Aasland OG,
Gude T. A three-year cohort study of the relationships between
coping, job stress and burnout after a counselling intervention for
help-seeking physicians. BMC Public Health 2010; 10: 213.
60 Winkel AF, Hermann N, Graham MJ, Ratan RB. No time to think:
making room for refl ection on obstetrics and gynecology residency.
J Grad Med Educ 2010; 2: 610–15.
61 Kim H, Wiedermann B. Have Accreditation Council for Graduate
Medical Education duty hour limits made a di erence?
A re-examination of resident sleep, mental health, education,
and safety seven years later. J Invest Med 2011; 59: 630–31.
62 Meerten M, Bland J, Gross SR, Garelick AI. Doctors’ experience of
a bespoke physician consultation service: cross-sectional
investigation. Psychiatrist 2011; 35: 206–12.
63 Schuh LA, Khan MA, Harle H, et al. Pilot trial of IOM duty hour
recommendations in neurology residency programs: unintended
consequences. Neurology 2011; 77: 883–87.
64 Bar-Sela G, Lulav-Grinwald D, Mitnik I. “Balint group” meetings
for oncology residents as a tool to improve therapeutic
communication skills and reduce burnout level. J Cancer Educ
2012; 27: 786–89.
65 Cheng T, DeOliveira MC, Ramani S, Krishnamurthy R, Jackson A.
3+1: a new model for resident training. J Gen Intern Med 2012;
27: S544–45.
66 Erler A, Beyer, M, Welbers G, Gerlach FM. Merging rural family
practices into the SCHAAZ: e ects on work satisfaction and risk of
burnout. Z Allg Med 2012; 88: 303–12 (in German).
67 Goodman MJ, Schorling JB. A mindfulness course decreases
burnout and improves well-being among healthcare providers.
Int J Psychiatry Med 2012; 43: 119–28.
68 Quenot JP, Rigaud JP, Prin S, et al. Su ering among carers
working in critical care can be reduced by an intensive
communication strategy on end-of-life practices. Intensive Care Med
2012; 38: 55–61.
69 Clayton JM, Butow PN, Waters A, et al. Evaluation of a novel
individualised communication skills training intervention to
improve doctors’ confi dence and skills in end-of-life
communication. Palliat Med 2013; 27: 236–43.
70 Dembitzer A, Wang B, Grask A, et al. Burnout in
clinician-educators and the importance of lifelong learning:
fi ndings from a medical education faculty development program.
J Gen Intern Med 2013; 28: S38.
71 Giannini A, Miccinesi G, Prandi E, Buzzoni C, Borreani C,
for the ODIN Study Group. Partial liberalization of visiting policies
and ICU sta : a before-and-after study. Intensive Care Med 2013;
39: 2180–87.
Articles
10
www.thelancet.com Published online September 28, 2016 http://dx.doi.org/10.1016/S0140-6736(16)31279-X
72 Weight CJ, Sellon JL, Lessard-Anderson CR, Shanafelt TD,
Olsen KD, Laskowski ER. Physical activity, quality of life, and
burnout among physician trainees: the e ect of a team-based,
incentivized exercise program. Mayo Clin Proc 2013; 88: 1435–42.
73 Fujimori M, Shirai Y, Asai M, et al. Development and preliminary
evaluation of communication skills training program for
oncologists based on patient preferences for communicating bad
news. Palliat Support Care 2014; 12: 379–86.
74 Kotb AA, Mohamed KA, Kamel MH, Ismail MA, Abdulmajeed AA.
Evaluating the e ect of an educational program on level of
professional burnout among family physicians in the Faculty of
Medicine—Suez Canal University. Arab J Psychiatry 2014; 25: 52–60.
75 Rosdahl JA, Kingsolver KO. Mindfulness training to increase
resilience and decrease stress and burnout in ophthalmology
residents: a pilot study. Invest Ophthalmol Vis Sci 2014; 55: 5579.
76 Goldhagen BE, Kingsolver K, Stinnett SS, Rosdahl JA. Stress and
burnout in residents: impact of mindfulness-based resilience
training. Adv Med Educ Pract 2015; 6: 525–32.
77 Podgurski L, Claxton R, Greco C, Croom A, Arnold R.
Brief mindfulness-based self-care curriculum for an interprofessional
group of palliative care providers. J Pain Symptom Manag 2015;
49: 384–85.
78 Ripp JA, Bellini L, Fallar R, Bazari H, Katz JT, Korenstein D. The impact
of duty hours restrictions on job burnout in internal medicine residents:
a three-institution comparison study. Acad Med 2015; 90: 494–99.
79 Warde C, Ching WT, Sohmer R. Enhancing resilience to promote VA
patient-centered medical home transformation and the untoward
e ects of the national access crisis. J Gen Intern Med 2015; 30: S516.
80 Swensen S, Kabcenell A, Shanafelt T. Physician-organization
collaboration reduces physician burnout and promotes
engagement: the Mayo Clinic experience. J Healthc Manag 2016;
61: 105–27.
