British Journal of Anaesthesia 104 (2): 16774 (2010)

doi:10.1093/bja/aep369

Advance Access publication December 30, 2009

Comparison of early cognitive function and recovery after

desflurane or sevoflurane anaesthesia in the elderly: a
double-blinded randomized controlled trial
D. Rortgen, J. Kloos, M. Fries, O. Grottke, S. Rex, R. Rossaint and M. Coburn*
Department of Anaesthesiology, University Hospital Aachen of the RWTH Aachen, Pauwelsstreet 30,
D-52072 Aachen, Germany
*Corresponding author. E-mail: mcoburn@ukaachen.de

Methods. Approved by the local ethical committee, 80 patients (aged 6575 yr) were enrolled
in this randomized, double-blinded study. Patients were allocated to either the desflurane (n40)
or the sevoflurane (n40) group. The primary outcome was the cognitive Test for Attentional
Performance with its subtests Alertness, Divided Attention, Visual Scanning, Working Memory,
and Reaction Change. In addition, PaperPencil Tests [Well-being Test BF-S, Recall of Digit Span
(DST), Digit-Symbol-Substitution Test, Trail Making Tests A and B, and Spielberg State-Trait
Anxiety Inventory] were measured. After baseline assessment 1224 h before operation, patients
were followed up 68 and 6672 h after operation. Among other outcome parameters, emergence times from anaesthesia and modified Aldrete scores were recorded.
Results. There was no difference in the incidence of POCD. However, according to the Paper
Pencil Tests, significant improvements for the desflurane group could be detected (Well-being
Test at 68 h, DST at 68 h, and Trail Making Test at 6672 h). Emergence was significantly faster
in the desflurane group for time to open eyes and time to extubation.
Conclusions. The total incidence of POCD showed no differences between the desflurane and
the sevoflurane groups. However, the tests Well-being scale, DST, and Trail Making Test, emergence times, and patients satisfaction were in favour of desflurane.
Br J Anaesth 2010; 104: 16774
Keywords: anaesthesia, geriatric; anaesthetics volatile, desflurane; anaesthetics volatile,
sevoflurane; recovery, cognitive; recovery, postoperative
Accepted for publication: November 23, 2009

The World Population Data Sheet of the year 2007

reports that 16% of the population in the industrial states
is older than 65 yr.1 With increasing life expectancy,
more and more patients aged 65 or older will receive
general anaesthesia. Patients older than 60 yr are particularly affected for postoperative cognitive dysfunction
(POCD) (26 41%).2 3 Alarmingly, patients experiencing
POCD are at an increased risk of death in the first year
after surgery.3 In conclusion, considering the increasing
numbers of the aged, POCD is a problem of growing
importance.

Desflurane and sevoflurane are widely used volatile

anaesthetics. Both are characterized with a low blood gas
partition coefficient [desflurane (0.42) and sevoflurane
(0.69)]. Concerning emergence times, desflurane has been
associated with shorter emergence times when compared
with sevoflurane, but early recovery of cognitive function
seems to be similar.4 In this study, we used Paper Pencil
Tests and complemented them with the cognitive Test for
Attentional Performance (TAP) with its subtests Alertness,
Divided Attention, Visual Scanning, Working Memory,
and Reaction Change.5 6

# The Author [2009]. Published by Oxford University Press on behalf of the British Journal of Anaesthesia. All rights reserved.
For Permissions, please email: journals.permissions@oxfordjournal.org

Downloaded from http://bja.oxfordjournals.org/ by guest on January 16, 2016

Background. Postoperative cognitive dysfunction (POCD) is being recognized as a complication contributing to perioperative morbidity and mortality of the elderly. We hypothesized
that the use of the shorter-acting volatile anaesthetic desflurane would be associated with less
incidence of POCD when compared with sevoflurane.

Rortgen et al.

The aim of this study was to compare early cognitive

function in the elderly after general anaesthesia with either
desflurane or sevoflurane using the computerized and
established TAP as the primary outcome. We hypothesized
that the use of the less soluble and faster-acting volatile
anaesthetic desflurane would be associated with less incidence of POCD when compared with sevoflurane.

Methods

168

Downloaded from http://bja.oxfordjournals.org/ by guest on January 16, 2016

The primary outcome parameter was the cognitive function

assessed with the computerized TAP (Version 1.7; Psytest,
2002), as reported previously.5 6 Of note, in this study, the
TAP was performed using an augmented set of subtests.
This was implemented to cover a broad range of cognitive
domains, including Alertness, Divided Attention, Visual
Scanning, Working Memory, and Reaction Change. In
addition to the TAP, a selection of classical cognitive tests
is assessed as secondary outcome parameters, in the course
called PaperPencil Tests, according to the way of assessment. This selection includes the Digit-Symbol-Substitution
Test (DSST), Recall of Digit Span (DST), and the Trail
Making Tests A and B.7 8 In addition, the short form of the
Spielberger State-Trait Anxiety Inventory (STAI) and the
Well-being Test BF-S are integrated in the PaperPencil
Tests to measure anxiety and depression in this study.9 10
POCD was defined as being a 20% decline in 20% of both
the PaperPencil Tests and the TAP.11 The TAP assessment
was completed by the study subjects in about 45 min.
Eighty patients, aged 65 75 yr, ASA I III, were
enrolled in this clinical trial. The trial was approved by the
local ethics committee. Randomization was computergenerated and allocation concealment was ensured by
enclosing assignments in sealed, sequentially numbered
envelopes. The anaesthetist could not be blinded owing to
the different administration methods of the anaesthetics;
however, both study investigators and patients were
blinded to group assignment (double-blinded observation).
The baseline measurements of the TAP and of the
Paper Pencil Tests were performed 12 24 h preoperative,
after obtaining written informed consent. The study subjects were visited by the investigator and were instructed
in a standard way on how to use the test battery (TAP) and
the Paper Pencil Tests. To prevent distraction from clinical routine and to guarantee a most identical setting, the
room-sharing patients were asked to leave the room and
all electronic devices such as telephone and TV were
switched off. The 6 8 and 66 72 h postoperative testing
were assessed in the same way for each patient. At all test
time points, the study subjects first performed the TAP followed by the Paper Pencil Tests.
For the secondary outcome parameters, the emergence
times from anaesthesia were recorded, including the time
of tracheal extubation. At a 20 s interval, starting from the
time of discontinuation of anaesthetic administration, a
blinded investigator asked each patient to open his or her

eyes, squeeze the investigators hand, and state time

(actual day) and spatial orientation until correct answers
were given. Every patient was discharged to the postanaesthetic care unit (PACU) for immediate recovery as
soon as they were able to be transported safely to the
PACU. At arrival, 5, 10, 15, 30, 45, 60 min, and at discharge, the modified Aldrete scores with a 10-point scale
were recorded in the PACU.12 The patients were asked for
nausea and pain at the same time intervals. Additionally,
vigilance, well-being, and energy were recorded at the
same time. Therefore, an independent investigator assessed
the patients vigilance; well-being and energy were evaluated by spontaneous patients statements. Vigilance was
rated as awake, tired, or sleeping. Well-being was classified as excellent, good, fair, and poor. Energy was divided
into normal, fair, and poor.
Further, the patients were questioned 6 8 h postoperative for patients self-evaluation of the anaesthesia
procedure, assessed by marks from one to six (excellent
unsatisfactory). Additionally, the patients were asked
whether they would choose the same anaesthesia if necessary for further operations.
The patients underwent elective surgery in traumatology, ear, nose, and throat surgery, gynaecology, urology,
or neurosurgery. Of note, neurosurgical interventions were
limited to operations on the spine. Chronic alcohol or drug
abuse, disturbed renal or liver function, diabetes mellitus,
disabling neuropsychiatric disorders, history of stroke, cardiopulmonary resuscitation, or brain trauma in the last 12
months, increased intracranial pressure, anaphylactic reaction to anaesthetics, emergency operations, history of myocardial infarction, adrenal insufficiency, congestive heart
failure, and lack of cooperation or legal incapacity were
excluded for eligibility. Duration of anaesthesia was
planned for at least 60 and up to 180 min. Standard monitoring included three-lead ECG, pulse oximetry, noninvasive arterial pressure measurements, temperature (AS/
3 monitor, GE-Datex Ohmeda, Helsinki, Finland), bispectral index monitoring (BIS model A 2000w, Software
Version 2.21, Aspect Medical Systems, Boston, MA,
USA), and the measurement of end-tidal carbon dioxide,
oxygen, and anaesthetic gas concentrations. All mentioned
parameters were recorded at fixed intervals of 5 min. For
induction of anaesthesia, patients were preoxygenated with
oxygen 100% for 3 min followed by a slow i.v. injection
of propofol (2 mg kg21) and simultaneously given remifentanil at 0.5 mg kg21 over 60 s. Rocuronium in a dosage
of 0.6 mg kg21 was used to facilitate tracheal intubation.
Depth of anaesthesia was controlled by adjusting desflurane or sevoflurane to age-adapted equipotent minimum
alveolar concentration (MAC) between 4.2 and 4.5 vol%
(desflurane) or 1.1 and 1.4 vol% (sevoflurane) in oxygen
30%13 (Cato Draeger, Lubeck, Germany) and administering remifentanil at a base rate of 0.15 mg kg21 min21.
During surgery, remifentanil was titrated due to clinical
needs. On the basis of the patients haemodynamic

Cognitive function, desflurane, and sevoflurane

Statistical analysis
Parametric data were tested with one-way ANOVA and presented as means and standard deviation, as means and
standard error of the means, as median and range, or as
means and 95% confidence intervals, accordingly.
Categorical data were analysed with the two-tailed
Pearsons x2 test and are given as numbers and per cents
of total. Statistical analysis was performed using SPSS
Software Version 16.0 (SPSS Inc., Chicago, IL, USA).
GraphPad PRISMw (GraphPad Software Inc., La Jolla,
CA, USA) was used to generate the figures.
The sample size was calculated for the primary outcome
parameters with a power of b0.9 and a significance level
of a0.05, considering a difference of 20% as relevant.
Mean values and standard deviations were taken from a
previous study in the elderly assessing the TAP ( patients
aged 65 75 yr) with the subtests Alertness, Divided
Attention, and Working Memory.6 The power was calculated for the subtests of the TAP (Alertness, Divided
Attention, and Working Memory) with n6 21 per
group. The trial size was determined with a total of 40
patients per group to compensate for possible drop-outs.
The power calculation was performed using nQuery
Advisorw, Version 4.0 (Statistical Solutions, Saugus, MA,
USA).

Results
A total of 80 patients were included in this study, 40 in
the desflurane group and 40 in the sevoflurane group. Both

study groups were comparable with regard to age, height,

weight, gender, education level, ASA status, and Apfel
score.14
At the 6 8 h postoperative assessment, the TAP was
completed by 31 patients in the desflurane group and 33
patients in the sevoflurane group, at the 66 72 h assessment by 27 and 30 patients. The reasons for drop-outs
were refusal of postoperative testing, not receiving the
allocated intervention, and discharge from hospital before
the second assessment.
The vital parameters systolic and diastolic arterial
pressure, heart rate, and Sao2 were comparable between the
groups, the latter being at least 98% at all times. The
mean remifentanil consumption was comparable in both
groups. The average anaesthetic gas concentration was
4.32 (0.26)% in the desflurane group and 1.27 (0.11)% in
the sevoflurane group. Type of surgery and duration of
anaesthesia did not differ between the groups. Baseline
assessment and the assessment of the TAP were performed
at comparable time points (Table 1).
Patients in the desflurane group recovered significantly
faster as indicated by the time to open eyes (P0.05)
and time to extubation (P0.05). No difference was seen
between the time to react on demand and time to time
(actual day) and spatial orientation. Length of stay in the
PACU and intra- and postoperative opioid consumption
were comparable between the groups (Table 1). In the
PACU, the modified Aldrete scores were similar in both
groups. The following scores were in favour of desflurane:
vigilance at arrival (P0.01), vigilance at 5 min
(P0.02), well-being at discharge (P0.03), and energy at
30 and 45 min (P0.05 and 0.04) (Table 2).
Baseline assessment of cognitive function showed no
difference between the two groups. The 6 8 and 66 72 h
postoperative tests for cognitive function and well-being
assessed with the Paper Pencil Tests and the ones
assessed with the TAP were normalized to the 12 24 h
preoperative baselines. The following subtests of the
Paper Pencil Tests showed significant differences in
favour of desflurane: Well-being at 6 8 h (P0.04), DST
at 6 8 h (P0.02), and Trail making Test A at 66 72 h
(P0.05) (Fig. 1). There was no difference for the TAP
between the two groups (Fig. 2).
Regarding both, the Paper Pencil Tests and the TAP,
the numbers with a decline of 20% from the 12 24 h preoperative baseline did not differ between the desflurane
and the sevoflurane groups at the 6 8 and 66 72 h postoperative assessment (Table 3). Furthermore, the incidence
of POCD (defined as being a 20% decline in 20% of both
the Paper Pencil Tests and the TAP)11 was similar at the
6 8 h (desflurane 39% and sevoflurane 36%; P0.67) and
66 72 h (desflurane 41% and sevoflurane 47%; P0.96)
time points.
The marks for patients self-evaluation of the anaesthesia procedure assessed 6 8 h after operation were significant in favour of the desflurane group (P0.03). Also the

169

Downloaded from http://bja.oxfordjournals.org/ by guest on January 16, 2016

(systolic arterial pressure or heart rate differing .20%

from baseline), autonomic (sweating, flushing, and salivating), and somatic signs (movement and swallowing), the
remifentanil infusion was increased successively until
symptoms were resolved. Standard treatment of blood loss
and fluid replacement strategy were used if necessary.
As part of the post-anaesthetic pain management, piritramide 0.05 mg kg21 i.v. and a short infusion of metamizole 15 mg kg21 were administered 20 min before the
estimated end of surgery as part of post-anaesthetic pain
management. Ten minutes before the end of surgery, the
inhaled anaesthetics were reduced to 0.5 age-adapted
MAC. Anaesthesia was maintained until all surgical interventions were completed, the haemodynamics (systolic
arterial pressure or heart rate) were stable, and full recovery of neuromuscular block was reached (TOF-Watch
SXw, Organon Teknika, Eppelheim, Germany). From this
point, return of spontaneous breathing was facilitated by
allowing end-expiratory carbon dioxide to increase up to
6.6 kPa. After full recovery of the upper airway reflexes
and return of sufficient spontaneous breathing [ventilatory
frequency .8 breaths min-1 and oxygen saturation (SaO2)
.95% at FIO2 100%], the patients tracheas were
extubated.

Rortgen et al.

Table 1 Anaesthesia data, emergence from anaesthesia, and time points of testing. Average anaesthetic gas concentration is presented as mean (SD). Type of
surgery is given in number and percentage of total in parentheses. Anaesthesia and PACU times are displayed as means with the upper and lower 95%
confidence interval. Intra- and postoperative piritramide consumption is presented as mean (SD). Emergence from anaesthesia: all time points are presented in
minutes as mean and lower and upper 95% confidence intervals in parentheses. Time points of testing are shown in hours and minutes (h:min) with the upper
and lower 95% confidence interval
Desflurane
4.32 (0.26)
6
4
8
7
12
127
66
5.2
7.2
7.4
7.7
7.7
10.8
18:12
15:26
15:16

1.27 (0.11)

(16%)
(11%)
(22%)
(19%)
(32%)
(105 149)
(54 78)
(1.9)
(4.1)
(6.2 8.6)
(6.5 8.9)
(6.5 8.9)
(9.3 12.2)
(17:47 18:37)
(13:54 16:57)
(13:48 16:45)

0.48

7 (18%)
2 (5%)
5 (13%)
8 (21%)
17 (44%)
152 (133 170)
71 (40 80)
5.9 (2.4)
7.9 (4.7)
9.2 (7.8 10.6)
9.2 (7.7 10.7)
9.4 (8.0 10.8)
13.4 (9.5 17.3)
18:36 (18:16 18:55)
14:43 (13:05 16:21)
15:15 (13:55 16:35)

0.09
0.58
0.20
0.62
0.05
0.12
0.05
0.21
0.13
0.52
0.98

Table 2 Post-anaesthetic care unit. All values are recorded after surgery at arrival (A) in the PACU, after 5 (5), 15 (15), 30 (30), 45 (45) min, and at discharge
from the PACU (D). Modified Aldrete scores are presented as mean (SD). Vigilance, well-being, and energy after anaesthesia are shown in numbers and per cent
of total (in parentheses). Vigilance is rated as awake (A), tired (T), and sleeping (S). Well-being is classified as excellent (E), good (G), fair (F), and poor (P).
Energy is divided into normal (N), fair (F), and poor (P)
AldreteA

Aldrete5

Aldrete15

Aldrete30

Aldrete45

Desflurane 9.5 (0.9)

Sevoflurane 9.3 (0.9)
P-value
0.23
VigilanceA (A/T/S)
Desflurane 4/27/5 (11/75/14)
Sevoflurane 4/17/18 (10/44/46)
P-value
0.01
Well-beingA (E/G/F/P)

9.5 (0.9)
9.3 (0.9)
0.46
Vigilance5 (A/T/S)
4/27/6 (11/73/17)
5/17/17 (13/44/44)
0.02
Well-being5 (E/G/F/P)

9.5 (0.7)
9.4 (0.99)
0.38
Vigilance15 (A/T/S)
4/27//6 (11/73/16)
5/19/15/ (13/49/38)
0.07
Well-being15 (E/G/F/P)

9.5 (0.8)
9.5 (0.9)
0.67
Vigilance30 (A/T/S)
3/27/5/ (9/77/14)
6/20/13 (15/51/33)
0.06
Well-being30 (E/G/F/P)

Desflurane

0/18/16/3 (0/49/43/8)

1/20/11/5 (3/54/39/14)

1/19/12/3 (3/54/34/9)

9.6 (0.6)
9.4 (0.9)
0.55
Vigilance45 (A/T/S)
2/21/4 (7/78/15)
5/17/10 (16/53/31)
0.14
Well-being45 (E/G/
F/P)
0/15/10/2 (0/56/37/7)

Sevoflurane 0/13/20/6 (0/33/51/15)

0/13/20/6 (0/33/51/15)

0/13/21/5 (0/33/54/13)

0/13/22/4 (0/33/56/10)

P-value

0.33
Energy5 (N/F/P)
3/16/17 (8/44/47)
1/11/27 (3/28/69)
0.13

0.13
Energy15 (N/F/P)
4/18/14 (11/50/39)
0/16/23 (0/41/59)
0.04

0.17
Energy30 (N/F/P)
4/21/9 (12/62/26)
2/16/21 (5/41/54)
0.05

0/16/18/2 (0/44/50/6)

0.31
EnergyA (N/F/P)
Desflurane 2/17/17 (6/47/47)
Sevoflurane 1/11/27 (3/28/69)
P-value
0.15

question of repetition of the same anaesthesia for further

operations showed more positive answers for the desflurane group (P0.05).

Discussion
In the present study, we compared early cognitive function
in the elderly after desflurane or sevoflurane anaesthesia.
The TAP showed no differences in incidence and severity
between the desflurane and the sevoflurane groups at the
6 8 and 66 72 h assessment. Compared with the baseline, the incidence of POCD was high in both groups (desflurane 39% and sevoflurane 36% at the 6 8 h assessment

AldreteD

9.8 (0.5)
9.9 (0.4)
0.55
VigilanceD (A/T/S)
4/27/5 (11/75/14)
9/23/7 (23/59/18)
0.29
Well-beingD (E/G/
F/P)
0/23/10/3 (0/64/28/
8)
0/10/19/3 (0/31/59/9) 0/16/22/4 (0/41/56/
10)
0.16
0.03
Energy45 (N/F/P)
EnergyD (N/F/P)
2/17/7 (8/65/27)
5/23/8 (14/64/22)
1/12/19 (3/38/59)
3/18/18 (8/46/46)
0.04
0.08

and at 66 72 h assessment desflurane 41% and sevoflurane 47%). However, in various subtests of the Paper
Pencil Tests, significant improvements for desflurane
(Well-being scale, DST, and Trail making Test A) could
be detected. Emergence times such as time to open eyes
and time to extubation were faster after desflurane anaesthesia. The scores for vigilance, well-being, and energy in
the PACU turned out beneficial to some extent in the desflurane group. Patients self-evaluation of anaesthesia was
in favour of desflurane.
The power was calculated considering a difference of
20% as relevant. However, a difference of 10 15% might
be clinically more realistic. Taking this into account, the

170

Downloaded from http://bja.oxfordjournals.org/ by guest on January 16, 2016

Average anaesthetic gas concentration (%)

Type of surgery
Trauma
ENT
Gynaecology
Urology
Neurosurgery
Anaesthesia time (min)
PACU time (min)
Intraoperative piritramide (mg)
Postoperative piritramide (mg)
Time to open eyes (min)
Time to react on demand (min)
Time to extubation (min)
Time to time (actual day) and spatial orientation (min)
Baseline assessment (h:min)
6 8 h postoperative assessment (h:min)
66 72 h postoperative assessment (h:min)

P-value

Sevoflurane

Cognitive function, desflurane, and sevoflurane

A 400

P=0.005

P=0.014

300

P=0.032
120

P=0.003

110

200

100

100

90
P=0.013
80

0
Baseline

C 130

68 h

Baseline

6672 h

D 120

P=0.000

120

68 h

6672 h

P=0.001
P=0.03

P=0.010

110

100
100
90

$
P=0.012

90

P=0.001
P=0.008

80

80
Baseline

E 120

68 h

6672 h

Baseline

F 120

P=0.003

110
100

100

90

90

P=0.000
Baseline

68 h

P=0.008

80

P=0.015

70

6672 h

110

P=0.05

80

68 h

P=0.000

70
6672 h

Baseline

68 h

6672 h

Fig 1 Cognitive function and well-being assessed with Paper Pencil Tests. All values are shown as mean (SEM) and display the change in per cent at
68 and 6672 h postoperative normalized to the 1224 h preoperative baseline. All ordinates are in per cent. (A) Well-being Test BF-S (an increase
in per cent presents a decline in well-being). For (B F) an increase in per cent shows an increase in outcome: (B) DST, (C) DSST, (D) Trail Making Test
A, (E) Trail Making Test B, and (F) six-item short-form of the Spielberg State-Trait Anxiety Inventory (STAI). Black lines and open diamonds display
the sevoflurane group and grey lines and open circles present the desflurane group; *P0.04; P0.02; $P0.05. An increase in baseline might be due
to a learning effect in (B F).

study might be underpowered. Further limitations of the

study are: patients above 75 yr and those with cardiac and
endocrine disease are excluded. However, these patients
are at high risk to develop POCD. This fact weakens the
external validity of the study. In addition, if the
Bonferroni correction is applied, as the study has several
outcome parameters, all test results fail to be significant.
The average anaesthetic gas concentration was 4.32
(0.26)% in the desflurane group and 1.27 (0.11)% in the
sevoflurane group. Both gas concentration ranges are in
line with the age-adapted MAC described by Nickalls and

Mapleson.13 As the anticipated operation is a stress factor

for the patients, the timing of the baseline assessment is
critical and was therefore assessed in a standardized
manner 12 24 h before operation in the hospital. Heavner
and colleagues4 used a baseline assessment just before premedication was given, but this could have influenced the
baseline testing even more.
Learning by repeated postoperative assessments may
represent a strong confounder. Retest reliability for the
TAP has been shown before, but not for such a short interval as used in our study.6 Looking at the Paper Pencil

171

Downloaded from http://bja.oxfordjournals.org/ by guest on January 16, 2016

110

Rortgen et al.

A 130
120

B 120

P=0.012

115

P=0.000

110
110

95
100

100

P=0.037

95

90

90
Baseline

68 h

6672 h

C 110

Baseline

68 h

6672 h

Baseline

68 h

6672 h

68 h

6672 h

D 110
P=0.015

105

110

100
100
95
P=0.02

90

85

85
Baseline

68 h

6672 h

E 140

F 105

P=0.024

100
120
95
100
90
P=0.01
85

80
Baseline

68 h

6672 h

G 140
130

Baseline

H 250
P=0.000

P=0.046

200

P=0.007

120

150

110

100

100

50

90

P=0.038
P=0.000

0
Baseline

68 h

6672 h

Baseline

I 120

J 115

110

110

100

105

90

100

68 h

6672 h

P=0.013
P=0.013

P=0.02
P=0.015

95

80
Baseline

68 h

6672 h

Baseline

68 h

6672 h

Fig 2 Cognitive function assessed with the TAP. All values are mean (SEM) and display the change in per cent at 6 8 and 66 72 h postoperative
normalized to the 12 24 h preoperative baseline. All ordinates are in per cent. For (A J), an increase in per cent shows an increase in outcome. All
figure pairs show first reaction time and then valid reaction. (A and B) Alertness; (C and D) Divided Attention; (E and F) Visual Scanning; (G and H)
Working Memory; (I and J) Reaction Change. Black lines and black open symbols stand for the sevoflurane group and grey lines and grey open circles
display the desflurane group.

172

Downloaded from http://bja.oxfordjournals.org/ by guest on January 16, 2016

90

Cognitive function, desflurane, and sevoflurane

Table 3 Twenty per cent deviation to the 12 24 h preoperative assessment in cognitive function and well-being assessed with both the Paper Pencil Tests and
the TAP. Values present numbers and per cent of total in parentheses with a deviation .20% to the normalized 12 24 h preoperative values. BF-S, Well-being
Test BF-S; DST, Recall of Digit Span; DSST, Digit-Symbol-Substitution Test; TMT A, Trail Making Test A; TMT B, Trail Making Test B; STAI, short-form of
the Spielberg State-Trait Anxiety Inventory; RT, mean reaction time; VR, valid reaction
6 8 h postoperative >20%
Desflurane

Sevoflurane

P-value

Desflurane

Sevoflurane

P-value

11
4
4
6
8
9
12
3
1
4
4
0
11
6
2
0

19
7
6
5
7
5
9
0
1
4
9
1
6
7
5
0

0.07
0.51
0.73
0.75
0.55
0.24
0.42
0.11
1.0
1.0
0.23
1.0
0.14
1.0
0.43

5
3
0
5
2
2
5
2
1
4
2
1
4
3
1
0

8
7
2
1
3
4
3
0
1
4
5
0
9
1
2
0

0.21
0.31
0.50
0.09
1.0
0.68
0.45
0.22
1.0
1.0
0.43
0.49
0.21
0.33
1.0

(37)
(13)
(14)
(21)
(33)
(27)
(40)
(10)
(3)
(13)
(15)
(0)
(44)
(24)
(7)
(0)

(61)
(22)
(19)
(16)
(25)
(15)
(28)
(0)
(3)
(13)
(29)
(3)
(22)
(27)
(16)
(0)

Tests, we used complementing versions when available

(e.g. Trail Making Test).
In this study, elderly patients underwent surgery known
to be associated with a high risk of POCD. The total level
of POCD seems to be similar to earlier studies performed
in the same age group.2 3 However, these data are limited
since different methods have been used to measure and
define POCD.
By using the DSST, Heavner and colleagues4 could not
find any difference between general anaesthesia with desflurane or sevoflurane in the elderly. Heavner and colleagues attributed their findings to the limitations of the
DSST. We used enhanced testing instruments and added
the computerized TAP with its subtests Alertness, Divided
Attention, Visual Scanning, Working Memory, and
Reaction Change in our investigation. The TAP is established to measure attention and was used in previous
studies.5 6 It was designed to analyse cognitive functions.
The increase in the subtest Alertness indicates temporary
deficits in attention whereas general slowness can be distinguished.15 16 Performing two or more different duties simultaneously is often needed in daily routine. This is
examined by dual-tasks in the subtest Divided Attention.17
The subtest Visual Scanning is important because it
measures continuous attention. In the concept of Working
Memory, the short-term memory is the essential instance
to control information flow for good attentional performance.18 21 Using the subtest Working Memory, a continuous control of the information flow through short-term
memory is tested. As mental flexibility seems to be an
important aspect of cognitive functioning, this is assessed
by the subtest Reaction Change.20 21
In this study, emergence times for time to open eyes and
time to extubation for desflurane were significantly faster
than those for sevoflurane. The study by Chen and

(19)
(11)
(0)
(19)
(8)
(7)
(20)
(8)
(4)
(15)
(8)
(4)
(18)
(14)
(4)
(0)

(32)
(23)
(7)
(3)
(11)
(13)
(10)
(0)
(3)
(13)
(19)
(0)
(36)
(4)
(7)
(0)

colleagues22 described a similar recovery of cognitive function for desflurane and sevoflurane anaesthesia in the
elderly.
In conclusion, there was no difference in Alertness,
Divided Attention, Visual Scanning, Working Memory, and
Reaction change 68 h and 6672 h after operation between
desflurane and sevoflurane anaesthesia. Despite some advantages for desflurane in three subtests of the PaperPencil
Tests (Well-being scale at 68 h, DST at 68 h, and Trail
making Test at 6672 h), anaesthesia in the elderly with desflurane or sevoflurane seems to be similar regarding recovery. Emergence times from anaesthesia for time to open
eyes and time to extubation in the elderly with desflurane
were faster than those with sevoflurane. Patients satisfaction
after anaesthesia was in favour of desflurane.

Acknowledgements
We would like to thank Jan-Hinrich Baumert and Oliver Kunitz who
helped in initiating this trial.

Funding
This work was funded in parts by Baxter Germany GmbH
(financial support). None of the authors received any corporate support, honoraria, etc. from the above-mentioned
sponsor of this study.

References

173

1 Haub C. 2007 World Population Data Sheet. The 2007 World

Population Data Sheet. Washington: Population Reference Bureau,
2007; 6

Downloaded from http://bja.oxfordjournals.org/ by guest on January 16, 2016

BF-S
DST
DSST
TMT A
TMT B
STAI
Alertness RT
Alertness VR
Divided Attention RT
Divide Attention VR
Visual Scanning RT
Visual Scanning VR
Working Memory RT
Working Memory VR
Reaction Change RT
Reaction Change VR

66 72 h postoperative >20%

Rortgen et al.

12 Aldrete JA. The post-anesthesia recovery score revisited. J Clin

Anesth 1995; 7: 89 91
13 Nickalls RWD, Mapleson WM. Age-related iso-MAC charts for
isoflurane, sevoflurane and desflurane in man. Br J Anaesth 2003;
91: 170 4
14 Apfel CC, Greim CA, Haubitz I, et al. A risk score to predict the
probability of postoperative vomiting in adults. Acta Anaesthesiol
Scand 1998; 42: 495 501
15 Jasper H, Sharpless S. Habituation of the arousal reaction. Brain
1956; 79: 655 80
16 Fimm B. Analyse und Standardisierung der neuropsychologischen
Aufmerksamkeitsbatterie, 1. Fassung. Freiburg: Psychologisches
Institut der Universitat, 1989
17 Zimmermann P, Fimm B. Testbatterie zur Aufmerksamkeitsprufung
(TAP), Version 1.02. Freiburg: Psytest, 2002
18 van Zomeren AH, Brouwer WH. Head injury and concepts of
attention. In: Levin HS, Grafman J, Eisenberg HM, eds.
Neurobehavioral Recovery from Head Injury. New York: Oxford
University Press, 1987
19 Atkinson RC, Shiffrin RM. The control of short-term memory.
Sci Am 1971; 225: 82 90
20 Baddeley AD, Hitch G. Working memory. In: Bower GH, ed. The
Psychology of Learning and Motivation. New York: Academic Press,
1974; 8
21 Sohlberg MM, Mateer CA. Introduction to Cognitive Rehabilitation:
Theory and Practice. New York: The Guildford Press, 1989
22 Chen X, Zhao M, White PF, et al. The recovery of cognitive
function after general anesthesia in elderly patients: a
comparison of desflurane and sevoflurane. Anesth Analg 2001; 93:
1489 94

174

Downloaded from http://bja.oxfordjournals.org/ by guest on January 16, 2016

2 Moller JT, Cluitmans P, Rasmussen LS, et al. Long-term postoperative cognitive dysfunction in the elderly: ISPOCD1 study. Lancet
1998; 351: 857 61
3 Monk TG, Weldon BC, Garvan CW, et al. Predictors of cognitive
dysfunction after major noncardiac surgery. Anesthesiology 2008;
108: 18 30
4 Heavner JE, Kaye AD, Lin BK, King T. Recovery of elderly patients
from two or more hours of desflurane or sevoflurane anaesthesia. Br J Anaesth 2003; 91: 502 6
5 Coburn M, Henzler D, Baumert JH, et al. Influence of a 30 min
break on divided attention and working memory in resident
anaesthetists on daily routine. Br J Anaesth 2006; 97: 315 9
6 Coburn M, Baumert JH, Roertgen D, et al. Emergence and early
cognitive function in the elderly after xenon or desflurane anaesthesia: a double-blinded randomized controlled trial. Br J Anaesth
2007; 98: 756 62
7 Reitan RM. Validity of the Trail Making test as an indicator of
organic brain damage. Percept Mot Skills 1958; 8: 271 6
8 Wechsler D. Wechsler Adult Intelligence Scale-Revised. New York:
Psychological Corporation, 1981
9 Marteau TM, Bekker H. The development of a six-item shortform of the state scale of the Spielberger State-Trait Anxiety
Inventory (STAI). Br J Clin Psychol 1992; 31: 301 6
10 von Zerssen D. Clinical self-rating scales (CSRS) of the Munich
Psychiatric Information System (PSYCHIS Munchen). In: Sartorius
N, Ban TA, eds. Assessment of Depression. Berlin, Heidelberg,
New York: Springer, 1986; 270 303
11 Grichnik KP, Ijsselmuiden AJ, DAmico TA, et al. Cognitive
decline after major noncardiac operations: a preliminary prospective study. Ann Thorac Surg 1999; 68: 1786 91