Acta Anaesthesiol Scand 1999; 43: 295–301 Copyright C Acta Anaesthesiol Scand 1999Printed in Denmark. All rights reserved

ACTA ANAESTHESIOLOGICA SCANDINAVICA

ISSN 0001-5172

Positive end-expiratory pressure prevents atelectasisduring general anaesthesia even in the presence of a highinspired oxygen concentration

P. NEUMANN1, H. U. ROTHEN4, J. E. BERGLUND2, J. VALTYSSON2, A. MAGNUSSON3 and G. HEDENSTIERNA2

Departments of 1Clinical Physiology, 2Anaesthesiology and Intensive Care Medicine, and 3Diagnostic Radiology, University Hospital, Uppsala, Sweden;4Department of Anaesthesiology and Intensive Care Medicine, University Hospital, Bern, Switzerland

Background: General anaesthesia impairs the gas exchange inthe lungs, and moderate desaturation (SaO2 86–90%) occurredin 50% of anaesthetised patients in a blinded pulse oximetrystudy. A high FiO2 might reduce the risk of hypoxaemia, butcan also promote atelectasis. We hypothesised that a moderatepositive end-expiratory pressure (PEEP) level of 10 cmH2O canprevent atelectasis during ventilation with an FiO2Ω1.0.Methods: Atelectasis was evaluated by computed tomography(CT) in 13 ASA I–II patients undergoing elective surgery. CTscans were obtained before and 15 min after induction of anaes-thesia. Then, recruitment of collapsed lung tissue was per-formed as a ‘‘vital capacity manoeuvre’’ (VCM, inspiration withPawΩ40 cmH2O for 15 s), and a CT scan was obtained at theend of the VCM. Thereafter, PEEPΩ0 cmH2O was applied ingroup 1, and PEEPΩ10 cmH2O in group 2. Additional CT scanswere obtained after the VCM. Oxygenation was measured be-fore and after the VCM.

GENERAL anaesthesia impairs pulmonary gas ex-change (1), and mild to moderate desaturation

of arterial blood (SaO2 86–90%) occurred in 53% of allpatients undergoing elective surgery and not moni-tored routinely by pulse oximetry (2). About 20% ofthese patients suffered from severe hypoxaemia(SaO2∞81%) (2). Thus, it seems reasonable to use highinspiratory oxygen concentrations during general an-aesthesia in order to minimise the risk of hypoxaemiaif pulse oximetry is not available. However, atelectasishas been proposed as a major cause of impaired oxy-genation during general anaesthesia (3, 4), and atel-ectasis is promoted by high inspiratory oxygen con-centrations during induction and maintenance of an-aesthesia (5–7). Increasing the inspired oxygenfraction (FiO2) during general anaesthesia mighttherefore increase the margin of safety only at the costof aggravating one cause of hypoxaemia.

A positive end-expiratory pressure of PEEPΩ10cmH2O has been shown to reduce, but not to eliminate,

295

Results: Atelectasis (±1 cm2) was present in 12 of the 13 patientsafter induction of anaesthesia. At 5 and 10 min after the VCM,atelectasis was significantly smaller in group 2 than group 1(P∞0.005). A significant inverse correlation was found betweenPaO2 and atelectasis.Conclusions: PEEPΩ10 cmH2O reduced atelectasis formationafter a VCM, when FiO2Ω1.0 was used. Thus, a VCM fol-lowed by PEEPΩ10 cmH2O should be considered when pa-tients are ventilated with a high FiO2 and gas exchange is im-paired.

Received 15 June, accepted for publication 20 October 1998

Key words: Atelectasis; lung density; PEEP; general anesthesia;computed tomography.

c Acta Anaesthesiologica Scandinavica 43 (1999)

the amount of atelectasis during general anaesthesia(8–10), but PEEP per se did not improve oxygenation inan unselected group of patients (10). Thus, PEEP alonemay not be the ideal tool for improving gas exchangeduring general anaesthesia. A vital capacity ma-noeuvre (VCM) (inflating the lungs to an airway press-ure of 40 cmH2O), in contrast, improves gas exchange(6, 11) and nearly eliminates atelectasis (6). However,atelectasis recurs within 5 min after a VCM (6), if thelungs are ventilated with 100% oxygen.

Therefore, we conducted this study to test the hypo-thesis that a moderate PEEP level of 10 cmH2O canprevent a renewed collapse of lung units after a VCMeven in the presence of an FiO2Ω1.0.

Material and methods

Study protocolIn patients scheduled for elective neurosurgery or eyesurgery, atelectasis was evaluated by computed tomo-

P. Neumann et al.

graphy (CT). CT scans were obtained with the pa-tients awake, and 15 min after induction of anaes-thesia. Then, a VCM was performed by inflating thelungs up to an airway pressure of 40 cmH2O (BOCmanometer, Ohmeda, attached to the endotrachealtube) for 15 s (12), and a CT scan was obtained at theend of the VCM. Thereafter, one group of patients wasventilated with PEEPΩ0 cmH2O, whereas a PEEPΩ10cmH2O was applied in the second group. Approxi-mately 20, 40 and 90 s as well as 5 and 10 min afterthe VCM additional CT scans were obtained. Thecompliance of the total respiratory system (Crs) andarterial blood gas samples were analysed 15 min afterinduction of anaesthesia, and 90 s and 10 min afterthe VCM.

Study populationAfter approval by the ethics committee of the Uni-versity Hospital of Uppsala, 13 ASA grade I or IIpatients without pulmonary disease were enrolled.Another 3 patients refused to participate in thestudy. All patients gave informed written consent.Patients were randomly assigned (sealed envelopes)to two study groups: After the VCM, PEEP waskept at 0 cmH2O in the ZEEP (zero end-expiratorypressure) group, whereas PEEP was increased to 10cmH2O in the PEEP group. The estimation of thesample size was based on previous studies (5, 6,13). Six patients were needed in each group to de-tect a 5 cm2 difference of atelectasis between the

Table 1

Demographic characteristics of the study population.

Group Patient no. Age (yr) Sex Smoking Weight (kg) Height (cm) BMI (kg ¡ mª2)

PEEP 0 cmH2O 1 66 M N 68 164 25.32 52 M S 75 174 24.83 44 M S 81 181 24.75 47 M S 70 180 21.67 51 M N 89 182 26.99 61 F N 71 167 25.5

Mean∫SD 54∫8 5 M, 1 F 3 N, 3 S 76∫8 175∫8 24.8∫1.7CI 95% 44–62 67–84 166–183 22.9–26.6

PEEP 10 cmH2O 4 48 F S 95 157 38.56 64 M N 67 170 23.28 74 M N 77 172 26.0

10 30 M N 72 180 22.212 36 M N 82 190 22.713 59 M N 85 176 27.4

Mean∫SD 52∫17 5 M, 1 F 5 N, 1 S 80∫10 174∫11 26.7∫6.2CI 95% 34–70 62–90 162–186 20.2–33.1

CI 95%: 95% Confidence interval. BMI: Body mass index. F: Female, M: Male. N: Non-smoker, S: Smoker. The different numbers of smokersand non-smokers in the 2 study groups is not significant (Fisher’s exact test). Patient 11 was excluded from analysis because he developedno atelectasis after induction of anaesthesia.

296

groups 10 min after the VCM with an a error ofÆ5% and a power of 90%. One patient of the PEEPgroup did not develop atelectasis after induction ofanaesthesia and was excluded from further analysis.Therefore, a total of 7 patients was allocated to thePEEP group. The characteristics of the study popu-lation are given in Table 1.

AnaesthesiaAs premedication patients received 0.05–0.08 mg ¡kgª1 ketobemidon (KetoganA) intramuscularly. An-aesthesia was induced with 2–3 mg ¡ kgª1 fentanyl and2 mg ¡ kgª1 propofol, followed by a continuous in-fusion of 5–10 mg ¡ kgª1 ¡ hª1 propofol. During induc-tion the lungs were ventilated via a face mask with100% oxygen, and 0.1 mg ¡ kgª1 pancuronium wasgiven to facilitate endotracheal intubation. Patientswere mechanically ventilated (Servo Ventilator 900C,Siemens Elema, Lund, Sweden) with an FiO2Ω1.0, aconstant flow, a rate of 10 breaths ¡ minª1, an inspira-tory-to-expiratory ratio of 1 : 2, PEEPΩ0 cmH2O and atidal volume (VT) of initially 10 ml ¡ kgª1. Approxi-mately 5 min after induction of anaesthesia VT wasadjusted to maintain an end-tidal CO2 concentrationbetween 4.0 and 4.5% (CO2 Analyser Eliza, Engstrom,Bromma, Sweden). Thereafter, VT was kept constantuntil the end of the study. The peripheral arterial oxy-gen saturation (PaO2) was continuously monitored bypulse oximetry (Biox 3740, Ohmeda, Louisville, CO,USA).

PEEP prevents atelectasis during anaesthesia

Computed tomographyPatients were placed supine on the computed tomo-graphy table (Somatom Plus 4, Siemens, Erlangen,Germany). With the patient awake and at anaesthesiabaseline (15 min after induction), a frontal scout viewof the chest was obtained at end-expiration. For analy-sis of atelectasis, transverse scans were performed atend-expiration 1 cm above the right diaphragm witha scanning time of 750 ms at 58 mAs and 140 kV, slicethickness 8 mm. In order to scan approximately thesame lung area during anaesthesia at end-expirationand at the end of the VCM, the position for the latterscan was moved caudally by 60 mm based on thefindings of a previous study (12). The total x-ray ex-posure for each patient was approximately 1.1 mSv(total exposure in Sweden averages 5 mSv ¡ yearª1).

Image analysisImages were analysed with the computer program Si-enet-Magic View, Version VA30A (Siemens, Erlangen,Germany). The entire left and right lungs of each scanwere chosen as the region of interest (ROI) by draw-ing the external boundaries of the lungs at the inside

Fig. 1. Atelectasis formation. The amount of atelectasis after induction of anaesthesia is shown. A: Study group ventilated without PEEP (ZEEP)after the vital capacity manoeuvre (VCM). B: Study group ventilated with PEEPΩ10 cmH2O after the VCM. Data of individual patients areshown, as well as the mean value for each study group (bold line). CT scans were obtained with patients awake (AW), 15 min after induction ofanaesthesia (BA), at the end of VCM and 20, 40, 90 s, and 5 and 10 min after the VCM. Differences between the 2 study groups were highlysignificant 5 and 10 min after the VCM. In Fig. 1B only 5 patients seem to be shown before the VCM, because mean values (bold line) are nearlyidentical with the measurements of one patient. After the VCM 3 patients did not redevelop any atelectasis. Therefore only 4 separate lines areshown. Abbreviations: AW: awake; BA: baseline anaesthesia; VCM: vital capacity manoeuvre.

297

of the ribs and the internal boundaries along themediastinal organs. The total area of the lungs wasmeasured by including pixels with density values be-tween ª1000 and π100 HU (Hounsfield units). There-after, using the same ROIs, the following areas werecalculated:

HU: π100 to ª100, representing non-inflated tissue(14)HU: ª900 to ª1000, over-aerated parenchyma (15).

Densities, considered to reflect atelectasis, were iden-tified in dependent lung regions and delineatedmanually. Atelectasis was then calculated by includ-ing all pixels within these areas with HU betweenª100 and π100.

Compliance of the respiratory systemStatic compliance (Crs) was estimated by using aninspiratory hold manoeuvre (about 4 s). Pressureand flow were measured in the ventilator on the in-spiratory side and recorded on a personal computerfor on-line signal processing, taking gas com-pression within the ventilatory circuit into account

P. Neumann et al.

Table 2

Oxygenation, atelectasis and compliance.

PEEPΩ0 cmH2O PEEPΩ10 cmH2O

Mean∫SD CI 95% Mean∫SD CI 95%

Baseline anesthesiaPaO2 (kPa) 57.4∫8.7 48.3–66.5 50.1∫12.2 37.2–62.9Atelectasis (cm2) 10.1∫8.4 1.3–18.9 13.7∫9.7 3.5–23.9Compliance (m ¡ cmH2Oª1) 68∫16 52–84 71∫22 47–95

90 s after VCMPaO2 (kPa) 66.1∫6.0 59.8–72.4 65.2∫7.0 57.8–72.6Atelectasis (cm2) 2.1∫3 0–5.3 1.0∫1.6 0–2.7Compliance (m ¡ cmH2Oª1) 67∫14 52–82 107∫34 72–143

10 min after VCMPaO2 (kPa) 56.3∫8.7 47.3–65.4 62.9∫7.3 55.3–70.6Atelectasis (cm2) 10.0∫4.6 5.2–14.8 *** 1.9∫2.5 0–4.6Compliance (m ¡ cmH2Oª1) 60∫13 46–74 * 100∫31 67–132

CI 95%: 95% Confidence interval. VCM: Vital capacity manoeuvre. *Significant difference between PEEP and ZEEP (P∞0.05). ***Highlysignificant difference between PEEP and ZEEP (P∞0.001).

(software: LabVIEWA 3.1, C-O Sjoberg Engineering,32983 B 70 National Instruments, USA). The meanvalue of 2 inspiratory hold manoeuvres was usedfor each point. Crs was calculated as tidal volumedivided by end-inspiratory pressure minus end-ex-piratory pressure.

Blood gas analysisArterial blood gas samples were analysed with ABL300 and OSM 3 Hemoximeter (Radiometer, Copen-hagen, Denmark).

StatisticsAll data are presented as mean∫standard deviation(SD). Tests of significance between the 2 studygroups (PEEP group vs. ZEEP group) were per-formed with a Mann-Whitney U-test and for thedistribution of smokers with the Fisher’s exact test.For comparison of the same parameter within 1group (PaO2 before and after the VCM) a Wilcoxonmatched-pairs test was applied. P∞0.05 was chosenas the level of significance. Calculations were per-formed with the software package STATISTICAA ona personal computer.

Results

CT scans awake and 15 min after induction ofanaesthesiaPrior to anaesthesia, atelectasis (2.6 cm2) was detectedin only one patient, who was randomised to the PEEPgroup. Fifteen minutes after induction of anaesthesia(anaesthesia baseline) 1 of 13 patients showed virtu-

298

ally no atelectasis and was excluded from furtheranalysis. In the remaining patients, atelectasis aver-aged 10.1∫8.4 cm2 (CI95% 1.3–18.9 cm2) in the ZEEPgroup and 13.7∫9.7 cm2 (CI95% 3.5–23.9 cm2) in thePEEP group (Fig. 1). The total lung area at anaesthesiabaseline was 213∫36 cm2 in the ZEEP group and219∫47 cm2 in the PEEP group. Thus, prior to the re-cruitment manoeuvre atelectasis averaged 4.7% and6.3% of the total lung area in the ZEEP and PEEPgroups, respectively. Hyperinflated lung parenchyma(HU ª900 to ª1000) was almost absent (∞1.1 cm2)in both groups at anaesthesia baseline. There was nosignificant difference between the two groups in anyof the above parameters.

Vital capacity manoeuvre (VCM)Atelectasis was eliminated in all but one patient at theend of the VCM, and no clinically important adverseeffects were observed. Mean lung density decreasedfrom ª607∫60 HU at anaesthesia baseline toª853∫27 HU at the end of the VCM in the ZEEPgroup, and from ª597∫88 to ª835∫45 in the PEEPgroup. Lung tissue with attenuation values betweenª900 and ª1000 HU increased to 173∫104 and168∫102 cm2, while the increase of the transverselung area averaged 110∫43 cm2 and 119∫52 cm2 inthe ZEEP and PEEP groups, respectively.

CT scans after VCMAll 6 patients ventilated without PEEP redevelopedatelectasis after the VCM. Already 5 min after theVCM, atelectasis was nearly as large as before theVCM, and it increased further during the next 5 min

PEEP prevents atelectasis during anaesthesia

(Fig. 1). In the PEEP group, in contrast, atelectasis av-eraged at the most 1.9 cm2 (CI95% 0–4.6 cm2), and 3 ofthe 6 patients did not redevelop any atelectasis afterthe VCM during the study period (P∞0.005 for differ-ences between groups).

Over-aerated lung tissue was minor in the ZEEPgroup (∞0.4 cm2, CI95% 0.01–0.21 cm2), but was sig-nificantly larger in the other group during ventilationwith PEEP, where it averaged 13.0 cm2 (CI95% 7.6–18.4cm2).

The total lung area was not significantly differentfrom anaesthesia baseline after the VCM in the ZEEPgroup, but it increased approximately 52 cm2 withPEEPΩ10 cmH2O (CI95% 28.7–75.3 cm2).

Oxygenation, compliance and atelectasisAt anaesthesia baseline oxygenation (Table 2) wassimilar in both study groups (57.4∫8.7 kPa ZEEP,50.1∫12.2 kPa PEEP), and 90 s after the VCM the PaO2

was equally and significantly (P∞0.05) improved, re-gardless of the application of PEEP in one group(66.1∫6.0 kPa ZEEP, 65.2∫7.0 kPa PEEP). However,

Fig. 2. Oxygenation and atelectasis. The x-axis shows atelectasis andthe y-axis gives the corresponding PaO2 measured at the same time forindividual patients. ò: ZEEP, 15 min after induction of anaesthesia;ô: ZEEP, 90 s after the recruitment manoeuvre; : ZEEP, 10 minafter the recruitment manoeuvre. P: PEEPΩ10 cmH2O, 15 min afterinduction of anaesthesia; S: PEEPΩ10 cmH2O, 90 s after the recruit-ment manoeuvre; : PEEPΩ10 cmH2O, 10 min after the recruitmentmanoeuvre. The following regression line was obtained for the relationbetween atelectasis and PaO2: PaO2Ωª1.0 ¡ Atelectasisπ66.2, rΩ0.76,P∞0.001.

299

10 min after the VCM, PaO2 had decreased to the pre-VCM level (56.3∫8.7 kPa) without PEEP, while no sig-nificant decrease occurred in the PEEP group (PaO2Ω62.9∫7.3 kPa). There was a highly significant inversecorrelation (P∞0.001) between arterial PaO2 and atel-ectasis during the course of anaesthesia (Fig. 2).

Crs was identical before and 90 s after the VCM inthe ZEEP group, but was significantly higher (P∞0.05)90 s after the VCM as compared to anaesthesia base-line in the group ventilated with PEEP (Table 2). Inboth groups, Crs decreased slightly between 90 s and10 min after the VCM.

Discussion

This study confirms earlier results (6) that atelectasisreappears within minutes after a recruitment ma-noeuvre during general anaesthesia, if 100% O2 isused. However, if PEEP Ω10 cmH2O is applied im-mediately after the VCM, re-expansion of atelectasismay have a sustained effect with respect to atelectasisand oxygenation. Thus, a recruitment manoeuvre fol-lowed by ventilation with PEEPΩ10 cmH2O shouldbe considered whenever gas exchange is impairedand a patient is ventilated with high oxygen concen-trations during general anaesthesia.

Methodological aspectsRandomisation caused an unequal distribution ofsmokers in the two study groups (ZEEP group: 3smokers; PEEP group: 1 smoker, difference n.s.). In aprevious study analysing 45 patients (22 smokers and23 non-smokers), a smoking history had no influenceon the amount of atelectasis or gas exchange disturb-ances observed after induction of anaesthesia (16).Likewise, in the present study the three smokers ofthe ZEEP group had slightly less atelectasis beforeand after the VCM as compared to the three non-smokers of the ZEEP group. Thus, a smoking historyof our patients has most likely not affected the resultsof this study.

Atelectasis has been evaluated by CT-scanning aspreviously described (14). In order to avoid excessiveradiation exposure, we chose to scan only 1 segment,located 1 cm above the diaphragm. This slice is notrepresentative for the whole lung, but showed a re-gion where most atelectasis can be expected to occurafter induction of anaesthesia (8, 12).

Atelectasis was delineated manually, which hasbeen demonstrated to result in a small bias as com-pared to computerised evaluation (14). For individualdata points, we found the following linear relation be-tween the manually enclosed area defined as atelecta-

P. Neumann et al.

sis and the total area of pixels with HU between π100and ª100: Area of HU π100 to ª100Ω1.04 atelectasisπ 2.94 cm2; rΩ0.97. Thus, computerised and manualevaluation lead to very similar results, with a slightlyhigher area obtained for the computerised approach.

During ventilation with PEEPΩ10 cmH2O as com-pared to ZEEP, the diaphragm moves caudally, but wedid not try to adjust the scanning position in order toensure that no parts of the diaphragm were scanned.Thus, lung tissue located more caudally was scannedin the PEEP group before rather than after the VCM.This might have caused some bias because theamount of atelectasis increases from the lung apex tothe diaphragm (5, 17). However, such an effect wouldonly be small (∞1.0 cm2 of atelectasis (5, 17) for lungtissue that is expected to be 1.5 cm apart (12)), anddoes not explain the difference observed between thePEEP and ZEEP groups.

Our recruitment manoeuvre, that inflated the lungsup to an airway pressure of 40 cmH2O for 15 s, mightbe questioned, since it can possibly cause barotraumaand most likely reduces cardiac output for the timeperiod of inflation. However, no immediate or late ad-verse effects were noted during the hospital stay ofthe study patients. In addition, lower airway press-ures up to 30 cmH2O have failed to restore normaloxygenation (11, 12) and resulted in incomplete re-cruitment (12). Whether a time period of 15 s is re-quired for a successful recruitment manoeuvre, orwhether shorter inflation times may be equally suc-cessful, needs to be established.

ImplicationsThe finding that 12 of 13 patients developed atelecta-sis within 15 min after induction of anaesthesia is inaccordance with previous studies (6, 8, 13, 14), evenwhen lower oxygen concentrations were used. Atel-ectasis has been demonstrated already after 5 min ofanaesthesia (8). This may suggest that compression,rather than absorption of gas behind occluded air-ways, is of major importance for atelectasis formation,since in a computer model the time taken for lungunits to collapse when filled with pure oxygen hasbeen estimated to be 7.85 min (18). However, in a pre-vious study atelectasis re-occurred already 5 min aftera recruitment manoeuvre during ventilation withpure oxygen, but not with 40% O2 in nitrogen (6). Inthe present study atelectasis increased rapidly be-tween 90 s and 5 min after the VCM, if no PEEP wasapplied (Fig. 1a), but not with PEEPΩ10 cmH2O (Fig.1b). Thus, atelectasis is most likely caused by a combi-nation of lung compression and absorption of gasfrom poorly ventilated lung areas. Compression atel-

300

ectasis can be prevented by a PEEP level which isequal to or greater than the superimposed hydrostaticpressure of that lung region (19).

We found a highly significant inverse correlation be-tween atelectasis and PaO2 (Fig. 2). This is in accordancewith previous findings, that shunt (and PaO2) is corre-lated to the size of atelectasis (10, 20, 21). Inflating thelungs up to an airway pressure of 40 cmH2O for 15 seliminated atelectasis in 11 of 12 patients, and im-proved oxygenation significantly. However, this effectwas only sustained if PEEPΩ10 cmH2O was appliedafter the VCM. Thus, a recruitment manoeuvre duringgeneral anaesthesia with a high FiO2 seems to be oflittle value, if lung collapse is not avoided thereafter, asby using PEEP. The observation period of 10 min fol-lowing the VCM in the present study might seem tooshort to draw the definite conclusion that PEEPΩ10cmH2O prevents the recurrence of atelectasis, but evenif the increase of atelectasis with PEEPΩ10 cmH2O be-tween 5 and 10 min after the VCM were extrapolated toa 60-min period, atelectasis would still not averagemore than 4.5 cm2. Since gas uptake from an unventi-lated lung area has been calculated to occur even fasterwhen Ø30% oxygen in nitrous oxide is used instead ofpure oxygen (18), our results might also be valid for themajority of general anaesthesias when a mixture of O2

and N2O is used.Ventilation with PEEPΩ10 cmH2O has been shown

to diminish atelectasis in combination with doubledtidal volumes (9), and without any other changes ofthe ventilatory pattern (8). Since a higher airwaypressure is necessary to open up collapsed alveolithan to keep recruited alveoli open, a PEEP level of∞10 cmH2O might prevent lung collapse equally ef-fectively after a recruitment manoeuvre as PEEPΩ10cmH2O. A lower PEEP level should be beneficial forthe patients in terms of haemodynamics, and mightreduce the amount of over-aerated tissue, which aver-aged 13 cm2 (approximately 5% of total lung area)after PEEPΩ10 cmH2O was applied.

In accordance with previous findings (6), compliancedecreased to a similar degree in the PEEP and ZEEPgroups after the VCM, although less atelectasis re-curred with PEEP. Likewise, compliance was un-changed in the ZEEP group 90 s after the VCM as com-pared to baseline anaesthesia (Table 2), although atel-ectasis was significantly reduced by the VCM. Thus, noclose relationship exists between compliance and theamount of lung collapse during general anaesthesia.

ConclusionsOxygenation and atelectasis are strongly and in-versely correlated during general anaesthesia. A vital

PEEP prevents atelectasis during anaesthesia

capacity manoeuvre can eliminate atelectasis in thevast majority of patients, and it restores normal gasexchange. However, atelectasis reappears within min-utes after a VCM, if 100% O2 is used, but not if PEEPΩ10 cmH2O is applied simultaneously. Thus, a vital ca-pacity manoeuvre followed by ventilation withPEEPΩ10 cmH2O should be considered whenever gasexchange is impaired and a patient is ventilated withhigh oxygen concentrations during general anaes-thesia.

Acknowledgements

We thank the X-ray laboratory team (Ms Marianne Almgren, MsAnn Erikson, Ms Ewa Larsson) for skilful technical help, andwe thank the staff of the Department of Neuroanaesthesia fortheir support during the study.

This study was supported by grants from the Swedish Medi-cal Research Council (No 5315), the Swedish Heart-Lung-Foundation, and the Datex-Engstrom-Company.

References

1. Sykes MK, Young WE, Robinson BE. Oxygenation duringanaesthesia with controlled ventilation. Br J Anaesth 1965:37: 314–325.

2. Moller JT, Johannessen NW, Berg H, Espersen K, Larsen LE.Hypoxemia during anaesthesia – an observer study. Br JAnaesth 1991: 66: 437–444.

3. Bendixen HH, Hedley-Whyte J, Laver MB. Impaired oxy-genation in surgical patients during general anesthesia withcontrolled ventilation: A concept of atelectasis. N Engl J Med1963: 269: 991–996.

4. Laver MB, Morgan J, Bendixen HH, Radford EP. Lung vol-ume, compliance, and arterial oxygen tension during con-trolled ventilation. J Appl Physiol 1964: 19: 725–733.

5. Reber A, Engberg G, Wegenius G, Hedenstierna G. Lungaeration. The effect of pre-oxygenation and hyperoxygen-ation during total intravenous anaesthesia. Anaesthesia 1996:51: 733–737.

6. Rothen HU, Sporre B, Engberg G, Wegenius G, Hogman M,Hedenstierna G. Influence of gas composition on recurrenceof atelectasis after a reexpansion maneuver during generalanesthesia. Anesthesiology 1995: 82: 832–842.

7. Rothen HU, Sporre B, Engberg G, Wegenius G, Reber A,Hedenstierna G. Prevention of atelectasis during general an-aesthesia [see comments]. Lancet 1995: 345: 1387–1391.

8. Brismar B, Hedenstierna G, Lundquist H, Strandberg A,Svensson L, Tokics L. Pulmonary densities during anes-thesia with muscular relaxation – a proposal of atelectasis.Anesthesiology 1985: 62: 422–428.

301

9. Hedenstierna G, Tokics L, Lundquist H, Andersson T,Strandberg A, Brismar B. Phrenic nerve stimulation duringhalothane anesthesia. Effects of atelectasis. Anesthesiology1994: 80: 751–760.

10. Tokics L, Hedenstierna G, Strandberg A, Brismar B, Lund-quist H. Lung collapse and gas exchange during generalanesthesia: effects of spontaneous breathing, muscle paral-ysis, and positive end-expiratory pressure. Anesthesiology1987: 66: 157–167.

11. Nunn JF, Bergman NA, Coleman AJ. Factors influencing thearterial oxygen tension during anaesthesia with artificialventilation. Br J Anaesth 1965: 37: 898–914.

12. Rothen HU, Sporre B, Engberg G, Wegenius G, Hedenstier-na G. Re-expansion of atelectasis during general anaes-thesia: a computed tomography study. Br J Anaesth 1993: 71:788–795.

13. Rothen HU, Sporre B, Engberg G, Wegenius G, Hedenstier-na G. Reexpansion of atelectasis during general anaesthesiamay have a prolonged effect. Acta Anaesthesiol Scand 1995:39: 118–125.

14. Lundquist H, Hedenstierna G, Strandberg A, Tokics L, Bris-mar B. CT-assessment of dependent lung densities in manduring general anaesthesia. Acta Radiol 1995: 36: 626–632.

15. Gould GA, Macnee W, McLean A, Warren PM, Redpath A,Best JJK et al. CT measurements of lung density in life canquantitate distal airspace enlargement: an essential definingfeature of human emphysema. Am Rev Respir Dis 1988: 137:380–392.

16. Gunnarsson L, Tokics L, Gustavsson H, Hedenstierna G. In-fluence of age on atelectasis formation and gas exchangeimpairment during general anaesthesia. Br J Anaesth 1991:66: 423–432.

17. Reber A, Engberg G, Sporre B, Kviele L, Rothen HU, Wegen-ius G et al. Volumetric analysis of aeration in the lungs dur-ing general anaesthesia. Br J Anaesth 1996: 76: 760–766.

18. Joyce CJ, Baker AB, Kennedy RR. Gas uptake from an un-ventilated area of lung: computer model of absorption atel-ectasis. J Appl Physiol 1993: 74: 1107–1116.

19. Gattinoni L, D’Andrea L, Pelosi P, Vitale G, Pesenti A, Fuma-galli R. Regional effects and mechanism of positive end-ex-piratory pressure in early adult respiratory distress syn-drome [see comments] [published erratum appears in JAMA1993: 270: 1814]. JAMA 1993: 269: 2122–3127.

20. Hedenstierna G, Tokics L, Strandberg A, Lundquist H, Bris-mar B. Correlation of gas exchange impairment to develop-ment of atelectasis during anaesthesia and muscle paralysis.Acta Anaesthesiol Scand 1986: 30: 183–191.

Address:G. Hedenstierna MD, PhDDepartment of Clinical PhysiologyUniversity of UppsalaS-75185 UppsalaSweden