48 / 240
of EE were low in most participants (71%), with 15% and
19% indicating moderate and high levels, respectively.
Levels of DP were also low in the majority (56%), with
17% exhibiting moderate and 21% with high levels. Results
are listed in
Table 2
for the number of respondents meeting
criteria for the true syndrome of burnout characterized by
high levels of EE and DP combined with low levels of PA,
as well as those with high EE and DP alone, irrespective of
PA. Both have been used in the literature to classify individ-
uals as demonstrating high levels of burnout.
2,11,12
The use
of the EE and DP indices to measure burnout independent
of PA is based on findings from the development of the
MBI-HSS showing strong correlation between levels of EE
and DP regardless of PA.
10
On the basis of these 2 criteria,
3.5% exhibited burnout syndrome and 16% demonstrated
high levels of burnout.
Table 3
lists the mean MBI-HSS sub-
scores for our survey participants. The mean (SD) EE score
fell into the low range at 16.5 (11.5), the mean (SD) DP
score was moderate at 6.2 (5.4), and the mean (SD) PA score
was high at 41.2 (5.8) (scale: low EE 18, high EE 27;
low DP 5, high DP 10; high PA 40, low PA 33).
Correlation and linear regression modeling were per-
formed to determine predictors of burnout. In keeping with
similar studies of burnout,
3,13
our analysis concentrated on
EE and DP, which had the strongest associations among the
3 burnout subscales.
Table 4
summarizes significant results.
Age showed an inverse relationship with EE (
r
= –0.39,
P
\
.0001) and DP (
r
= –0.28,
P
\
.0041). The length of time
married also showed similar negative correlations with
EE (
r
= –0.33,
P
= .0007) and DP (
r
= –0.33,
P
= .0045).
The number of children in the home was correlated with
EE (
r
= 0.22,
P
= .0275) and DP (
r
= 0.23,
P
= .0235).
With regard to practice-related factors, the number of
hours worked per week showed an association with EE (
r
=
0.31,
P
= .0016). Likewise, the number of years on the job
was also related with EE but showed an inverse relationship
(
r
= –0.25,
P
= .0108). There was no statistically significant
relationship between practice setting (ie, academic or pri-
vate, solo or group) and EE or DP.
Discussion
Physician burnout continues to be a widespread problem
with many deleterious sequelae. The negative impacts of
physician burnout on the health care landscape are well
documented and include such effects as dissatisfied and less
compliant patients, riskier prescribing profiles, lower produc-
tivity, and increases in medical errors, to name a few.
14-18
Although several studies have recently begun to address this
phenomenon in otolaryngologists,
7,8,19
we have yet to attain a
thorough understanding of the risk factors leading to its
occurrence. Herein we report a study of burnout in practicing
otolaryngologists with correlation to potentially modifiable
risk factors.
Burnout was not very prevalent in our survey population.
Of those surveyed, only 3.5% experienced the composite
syndrome of burnout with high scores on all 3 indices, and
16% had burnout according to subscale measurements of
EE and DP. In addition, analysis of the subscale results
shows a more favorable picture of practicing otolaryngolo-
gists’ health with respect to burnout. Both emotional
exhaustion and depersonalization scores on average were in
the low range. High levels of EE and DP were found in
only 19% and 21% of respondents, respectively. These
results are in contrast to other published surveys of burnout
in academic otolaryngologists and department chairs, which
showed moderate levels of burnout in the majority of
respondents.
7,8
Prior studies
7
have also demonstrated lower
levels of burnout among otolaryngologists when compared
to other surgical specialties such as general surgery and OB/
GYN, and our results are in keeping with this. Our respon-
dents also had lower mean burnout scores than were
reported in the normative data for the medicine subscale of
the MBI-HSS, which showed mean (SD) EE, DP, and PA
levels of 22.19 (9.53), 7.12 (5.22), and 36.53 (7.34), respec-
tively.
10
When compared to large surveys of burnout such
as the one by Shanafelt et al
20
of 7905 members of the
American College of Surgeons, our population also had a
lower level and degree of burnout. This may reflect a sam-
pling bias of our study in that those surgeons experiencing
higher levels of burnout may have been less likely to com-
plete and return our survey because of a lack of interest or
time. Therefore, it is possible that the extent of burnout was
underreported in our study population. The study population
in Shanafelt et al also comprised 41% general surgeons
compared to 4.7% otolaryngologists. This difference in
study population may account for the observed difference in
burnout reported, in light of the fact that the general surgery
population tends to have higher degrees of burnout than
otolaryngology.
Both the prevention and treatment of burnout rely heav-
ily on the recognition of its manifestations. Recognition can
be difficult in professionals with high stress such as physi-
cians, who frequently demonstrate poor insight into their
own mental and professional health.
21
This has contributed
Table 2.
Percentage of Participants Meeting Criteria for Burnout
Burnout
No. (%)
"
EE/DP
19/115 (16)
"
EE/DP,
#
PA
4/115 (3.5)
Abbreviations: EE, emotional exhaustion; DP, depersonalization; PA, per-
sonal accomplishment,
"
, high level;
#
, low level.
Table 3.
Mean Maslach Burnout Inventory Subscores for Survey
Participants
Mean Standard Deviation Range
Emotional exhaustion
16.5
11.5
Low
Depersonalization
6.2
5.4
Moderate
Personal accomplishment 41.2
5.8
Low
Otolaryngology–Head and Neck Surgery 146(2)
26