PULMONARY AND OTHER RESPIRATORY COMPLICATIONS AFTER CARDIOPULMONARY BYPASS AND INVESTIGATION OF AETIOLOGICAL FACTORS

G.D. GALE, S.J. TEASDALE, D.E. SANDERS,* P.J. BRADWELL, A. RUSSELL, B. SOLARIC, AND J.E. YORK

ATELECTASIS is a common respiratory complica- aetiological factors. This was considered neces- tion after open operations with an incidence sary to determine the effect of changes of tech- of 60 to 84 per cent (Table I). It is detrimental to nique, the haemodilution prime and the intra- patient recovery because it reduces lung com- aortic balloon on a continued high rate of atelec- pliance and increases the work of and tasis which was unresponsive to treatment with the alveolar-arterial oxygen gradient. This leads the incentive spirometer when used four times a to hypoxaemia if the arterial oxygen tension can- day.-~ not be maintained by an increase in the inspired oxygen concentration. METHOD This study was designed to survey respiratory complications after heart operations with car- Clinical data were collected prospectively in 50 diopulmonary bypass and to investigate possible consecutive patients following heart operations with cardiopulmonary bypass. The mean age was TABLE I 48.6 years (range 21-7I) and weight was 68.1 kg (range 40-94) with 34 males and 16 females. Two Per cent incidence patients had radiological signs of chronic of pulmonary . Operations are shown Year complications in Table II. The intra-aortic balloon was used in PROVAN, et al. 1 1966 61.5 one case after resection of left ventricular TEMPLETON, et al. 2 1966 74 aneurysm and in eleven cases with aorto- GAUERT, et al) 1971 70 coronary bypass. The high rate of use of the TURNBULL, et al. 4 1974 84.3 balloon is accounted for by the fact that its indi- GALE and SANDERS s 1977 84 cations had not been defined and it was being used prophylactically for stenosis of the left main From the Departments of Anaesthesia and Radiol- coronary artery and poor left ventricular func- ogy,* Toronto General Hospital, and University of To- tion. The mean stay in intensive care was four ronto, Toronto, Ontario, Canada. days. Reprints fi'om Dr. G.D. Gale, Department of Anaes- thesia, 2nd Floor Norman Urquhart Wing, Toronto Clinical examination of the chest has been General Hospital, 101 College Street, Toronto, On- found to under-estimate the frequency of atelec- tario, Canada, M5G 1L7. tasis s and changes of temperature, pulse and res-

TABLE II

Number of Patients with any Operations performed in this study patients type of atelectasis

Mitral valve replacement (One with double aorto-coronary bypass) 4 Aortic valve replacement (one with ventriculomyotomy) 8 Multiple valves (7 valves) 3 Aorto-coronary bypass (59 grafts and 11 balloons) 27 17 Septal defects (3 ventricular, 3 atrial) 5 3 Resection of left ventricular aneurysm (with balloon) 1 Ventriculomyotomy 1 Repair of descending thoracic aorta (redo) l Intra-aortic balloon 12 15 Canad. Anaesth. Soc. J., vol. 26, no. 1, January 1979 16 CANADIAN ANAESTHETISTS' SOCIETY JOURNAL

TABLE III RADIOLOGICALSURVEY SHOWING NUMBER OF CASES WITH PULMONARY COMPLICATIONS IN FIFTY PATIENTS

Effusions: Right (3 small 1 medium) 4 Left (12 small, 1 large) 13 Bilateral (13 small, 3 medium, 1 large, 2 mixed) 19 Total Effusions 36 Atelectasis 32 Paratracheal haematomata 11 Gastric dilatation 7 Signs of congestive 5 Tension (with chronic obstructive lung disease) 1 Left haemothorax, bilateral , consolidation R.U.L. 1 Right basal consolidation 1 piration are poorly correlated with atelectasis signs of congestive heart failure post-operatively after cardiopulmonary bypass, s so respiratory in five patients including one with pulmonary complications were assessed radiologically. oedema. Preoperative chest films were reviewed. Films were then taken one to three hours after opera- Respiratory support and clinical progress tion, at 24 and 48 hours and every two days there- Seven patients were ventilated for three to six after unless required more frequently. A short hours after operation. Thirty-nine patients were time-exposure of one-twentieth of a second or ventilated overnight (from 12 to 24 hours). The less was used with a high kilovoltage technique to remaining four patients required respiratory sup- ensure good films. After extubation of the port for 37, 52 and 65 hours and 14 days respec- films were taken in the upright position with ad- tively. Thirty-two patients showed signs of re- ditional lateral views. duced chest expansion, rhonchi or r'~les at the Atelectasis was classified on radiological evi- bases. Only two patients were unable to maintain dence as plate, subsegmental, segmental, or lobar an arterial oxygen tension of 100ram Hg in type. Plate atelectasis (synonymous with disk (13.3 kPa) when the inspired oxygen proportion atelectasis) is described by Fraser and Par66 and was increased. was originally described by Fleischner. 7 It con- sists of linear densities almost always at the lung A telectasis bases, roughly horizontal or slightly oblique, varying in thickness from one to three millimeters Incidence and commonly several centimeters in length. The overall incidence of atelectasis was 64 per Lines due to plate atelectasis invariably extend to cent (Table IV). Plate and subsegmental atelec- the pleural surface. tasis was twice as common as segmental. There Statistical analysis was done using the chi- was only one patient with lobar atelectasis (in- squared test or Fisher's exact test of probability. complete right middle and lower) which was combined with a single left segmental atelectasis. RESULTS Five other patients showed two types of atelec- tasis. There were three with plate and subseg- Pulmonary complications: (Table III) mental and one each with plate and segmental The commonest finding was with an incidence of 72 per cent. Most effusions TABLE IV were small and there were more on the left side THE INCIDENCE OF ATELECTASIS IN FIFTY CONSECUTIVE than the right. Effusions were unrelated to the PATIENTS distribution of atelectasis and were more frequent in patients with a cardiac index of less than two Number of patients Per cent and one halflitres per minute (p < 0.02). Any type 32 64 Atelectasis was the second commonest finding Plate 15 30 and is considered below. Paratracheal hae- Subsegmental 15 30 matoma and gastric dilatation occurred in eleven Segmental 7 14 Lobar 1 2 and seven patients respectively. There were GALE, el al.: ATELECTASIS AFTER CARDIOPULMONARY BYPASS 17

TABLE V 90 Under I hour NUMBER OF PATIENTS WITH ATELECTASlS CLASSI- 80 Z/Z///,, FIED BY ZONE, OUT OF A TOTAL OF FIFTY PATIENTS f/////~ /////// /////// 70 Basal Mid Apical Any Type 27 8 1 60 Plate 12 3 1 Subsegmental 11 6 0 50 Segmental 7 1 0 N Lobar 1 1 0 40 7////// ///////~ 30 TABLE VI e1111111 ,111/11, ¢1/11//# THE NUMBER OF PATIENTS WITH ATELECTASIS CLASSI- 20 I/1////, FIED BY SIDE OUT OF A TOTAL OF FIFTY PATIENTS: THE THREE CATEGORIES ARE MUTUALLY EXCLUSIVE I0 g/////~ Right only Left only Bilateral ~ 0 4o Any Type 11 7 14 I - 2 hours Plate 4 4 7 ~ 30 iiiiiiiii!ii!iiiiii!i!ii!~ Subsegmental 6 4 5 Segmental 2 3 2 t~ 20 7/~/~ iiiii!iiiiiiiii!!iiiii!iii Lobar 1 0 0 iiiiiiiiiiiiiiiiiiiiiiiiiiii!iiii~iiiiiiiii I0 and one with subsegmental and segmental atelectasis. 0 50 Over 2 hours Zofle Analysis by zone showed basal atelectasis to be roughly three times as common as mid zone 3O atelectasis (Table V). Apical atelectasis was rare and was present in only one patient who had a 20 single left apical plate lesion combined With single bilateral basal plate atelectasis, a moderate right I0 Neural effusion and a paratracheal haematoma. 0 PLATE SEGMENTAL Side SUBSEGMENTAL LOBAR The frequency of unilateral atelectasis was similar on each side. Bilateral atelectasis oc- FmURE 1 This shows the type ofatelectasis found postoperatively after cardiopulmonary bypass times of curred in 28 per cent of patients (Table VI). under one hour, one to two hours and over three hours. The number of patients in the three groups was respec- Aetiological factors tively 6, 32 and 12. Plate atelectasis was commoner in the group under one hour than in the other two groups (1) Preoperative (p < 0.05 and < 0.02). Analysis of data on age, obesity, preoperative pulmonary disease, productive cough, smoking judged by the last haematocrit on cardiopulmo- habits, treatment with , pulmonary nary bypass and the use of the intra-aortic balloon hypertension, poor left ventricular function were unrelated to the incidence of atelectasis. (Grade III and IV) and raised left ventricular Patients with a cardiopulmonary bypass time end-diastolic pressure above 12 mm Hg (1.6 kPa) of less than one hour had a higher incidence of failed to correlate these factors with the occur- plate atelectasis than intermediate and long rence of atelectasis. bypass groups (p < 0.05 and < 0.02). In the longer bypass groups the incidence of subsegmental (2) Peroperative factors atelectasis rose but failed to reach significance The use of positive end-expiratory pressure (Figure 1). Segmental atelectasis was signif- during operation, the degree of haemodilution as icantly more frequent if the positive fluid 18 CANADIAN ANAESTHETISTS' SOCIETY JOURNAL

40 Undlr :3 litres in nine out of eleven patients who lost more than one litre of blood in the first 24 hours after opera- 30 tion. ..:~IN There was one case of right basal consolidation 20 and nine cases of atelectasis out of eleven pa- tients who had radiological evidence of para- 10 g////d NINI \ ",d tracheal haematomata, but only four of these pa- 0 F////A \ 1 tients had over one litre of external blood loss. Only two patients with paratracheal hae- 40 matomata were re-opened for bleeding. 3-4 litres Atelectasis occurred in all seven patients with

•.:~i~~ ..:::.::::::::>'~.::>.:::::~>.~::...~.~~.:.~i.-;.~..:.~i:... ~~~ ~i~U.:..~ ~iWt.~ gastric dilatation, which was usually found on the morning after operation (p < 0.05). Of seven pa- ~:"~i:~::$ :~,~::~:.. tients with congestive heart failure four had l0 atelectasis and one developed a right basal con- //////~":::.:..~v~i:~:~'~'~k~ ~.~:i::;:'~#~:..~:~:. \ \ solidation secondary to pulmonary oedema. There was no correlation between the duration of postoperative ventilation and the incidence of Over 4 litres atelectasis.

40 Not re-operated for bleeding :50 30 20 20 I0 I0 o~ 0 0 PLATE SEGMENTAL I00 -~ SUBSEGMENTAL LOBAR 90- !!i!i!i!i!i!i!i!i!i!i!i!!!i!ii!i!i!i!i!i!i! FIGURE 2 This shows the type ofatelectasis found :!:!:!:!:!:!3!:!:!:!:!:!:!:!:!:!:!:!:!:! :::::::::::::::::::::: Re-operated for postoperatively in relation to the fluid load at the end of :!:!:!:!3!:!:!:!:!:!:!:!:i:!8!:i:??? 80- ,,,,,,,,,,,,,,,,,,,,, bleeding bypass, which is classified as under three litres, three to •_,,,,,,,,,,,,,,,. :,=,=,==,=== ..... four litres and over four litres. The number of patients .,,,,, .,,,,,,,,,,,,,,. in each of the four groups was 26, 16 and 7 respectively. 70 Segmental atelectasis was commoner in the group over four litres compared to those under three litres (p < 60 0.05). 50

.=,,==,=,,,=,. .... balance at the end of operation exceeded four -:-:-:-:-:-:.:-:-:-:-:-:-:.:-:-:-:.:-:-:-: 40 Y::;;::;:::;:::;:::;. litres compared to those in which it was less than three litres (p < 0.05) (Figure 2). There was a :50 ::::::::::::::::::::: trend to less plate atelectasis and more subseg- ...... ii]ii]iii?iii?iiiiiii mental atelectasis in the group with a large fluid 20 load but this failed to reach significance. Subseg- I0 mental atelectasis was significantly more com- mon in patients who returned to the operating 0 iiii ii , I PLATE SEGMENTAL room for bleeding (p < 0.05) (Figure 3). Three SUBSEGMENTAL LOBAR patients in this group had thoracotomies, and another patient had a femoral balloon site FIGURE 3 This shows the type of atelectasis in rela- explored for haematoma. tion to re-operation for bleeding. Four patients out of 50 had re-operation for bleeding. Three had thoracotomies and one had exploration of a femoral balloon site. Sub- (3) Postoperative factors segmental atelectasis was commoner in the re-operated Bleeding and atelectasis. Atelectasis occurred group (p < 0.05). GALE, et al. : ATELECTASIS AFTER CARDIOPULMONARY BYPASS 19

DISCUSSION patients with congestive heart failure postopera- tively. Earlier papers agree roughly on the incidence The fi-equency of atelectasis was unrelated to of atelectasis after cardiopulmonary bypass but the type of operation and the use of the intra- they differ on the distribution of atelectasis and aortic balloon. The former finding agrees with aetiological factors. that of Gauert 3 but not with that of Provan.1 More segmental atelectasis was found when Incidence the residual fluid balance after bypass exceeded A frequency of atelectasis of 64 per cent is foul litres compared to those with less than three consistent with earlier observations which litres. This suggests that the increased lung water showed atelectasis occurring in 60 to 84 per cent found after bypass 9 causes atelectasis. of patients after cardiopulmonary bypass.l-S The variation in incidence of atelectasis reported may Bypass time and atelectasis be in part due to inconsistencies in frequency of Our study found that the overall frequency of radiological examination. The importance of a atelectasis was not increased in cases with a long lateral chest X-ray is illustrated in one of our bypass time in contrast to the findings of patients who showed atelectasis in one lateral others."3"4'~° More plate atelectasis was found in view only. There may also be inconsistencies in patients after a bypass time of less than one hour the classification of atelectasis that is used. Pro- compared to those with a longer bypass. This van ~ excluded minor radiological changes and finding is best explained by suggesting that many this may account for his lower incidence of cases developed plate atelectasis within the first atelectasis. hour on bypass, but in longer bypass groups, atelectasis progressed from plate to subsegmen- tal types (Figure 1). Distribution In the short bypass group post-bypass fluid The results of this study agree with others that load and blood loss were low, so that the main found more atelectasis at the bases than in the causes of atelectasis were reduced vital capacity upper zones. >s Three earlier studies found more and residual volume and supine immobilization atelectasis at the left base. >4 Our finding of the leading to small airway closure and basal atelec- same incidence on each side agrees with an ear- tasis. lier study 5 and suggests that in our patients there In the longer bypass groups additional factors were no important left unilateral aetiological after bypass were a larger fluid load, more bleed- factors, such as compression of the lung from a ing and thoracotomies in three patients, which large heart or surgical manipulation, uninten- may have converted developing plate atelectasis tional intubation of the right main or into the more extensive subsegmental and seg- failure to remove secretions from the left lung. mental types. These three factors illustrate pos- sible mechanisms of development of atelectasis. Aetiological factors An increased water load after bypass will cause a This study found that age, obesity and smoking rise in extra-vascular lung water which will re- habits were unrelated to the frequency of atelec- duce compliance, lung volume and ventilation, tasis. This might be explained partially by the fact causing airway closure. Bleeding may cause that our patients were neither very old (only one atelectasis by pressure on the lung and atelectasis was over 65 years) nor very obese. Gauert 3 found was present in nine out of eleven patients with that atelectasis increased with age after mitral but paratracheal haematoma. Patients who have not after aortic valve replacements. for bleeding are likely to get atelec- Our findings support the view of Hilbermans tasis from lung compression and trauma. that cardiac catheter data are a poor predictor of The mean duration of postoperative ventilation pulmonary function. The poor correlation be- was the same in both short and long bypass tween preoperative cardiac catheter data and groups, so that retention of bronchial secretions postoperative atelectasis might be expected, is a factor that would apply equally to both since surgery is designed to improve cardiac groups. function. The poor correlation between cardiac function and atelectasis continued into the recov- Hypoxaemia and atelectasis ery period, with atelectasis in only four of seven After discontinuing controlled ventilation only 2O CANADIAN ANAESTHETISTS" SOCIETY JOURNAL two patients out of fifty were unable to maintain segmental forms. The incidence of atelectasis an arterial oxygen tension of 100 mm Hg (13.3 was the same on each side and the site of atelec- kPa) with the use of an increased fraction of in- tasis was basal in three quarters of the patients. spired oxygen. This suggests that although Preoperative clinical and catheter data were atelectasis was common, hypoxaemia was pre- un,'elated to the incidence of atelectasis. vented in most patients. Atelectasis will matter There was a significant positive Correlation most clinically where there is a large alveolar- between a short cardiopulmonary bypass time arterial oxygen gradient. This may occur due to and plate atelectasis, between a large fluid load extensive atelectasis or due to interstitial pulmo- after bypass and segmental atelectasis, between nary oedema occurring in heart failure or fluid re-operation for bleeding and subsegmental overload. In these conditions arterial oxygena- atelectasis and between post-operative gastric tion cannot be maintained, lung compliance falls, dilation and atelectasis. the work of breathing is increased and the patient The type of operation, the use of the intra- goes into cardio-. Our results aortic balloon and the length of postoperative did not show a correlation between atelectasis respiratory ventilation were unrelated to the inci- and postoperative heart failure but when they dence of atelectasis. co-exist the problem of hypoxaemia is increased. The mechanism of development of atelectasis is discussed. CONCLUSIONS AND RECOMMENDATIONS RI~SUMt~ Although pleural effusion was the most fre- quent complication, most of these effusions were Les manifestations radiologiques des compli- small and of little clinical significance and were cations respiratoires apr~s chirurgie h coeur unrelated to the incidence ofatelectasis. ouvert, ainsi que les facteurs 6tiologiques possi- Atelectasis occurred in 64 per cent of patients. bles de ces complications, ont fait I'objet d'une Investigation of possible aetiological factors was 6tude chez 50 malades consdcutifs. disappointing in not indicating more clearly a re- L'at61ectasie s'est avdr6e la complication la lationship of these factors to atelectasis. plus fr6quente (apr~s les petites effusions pleu- Results suggest it is advisable to: rales) avec une incidence de 64pour cent. Plusieurs 1. Restrict fluid load at the end of cardiopul- types d'at61ectasie ont frequemment 6t6 observ6s monary bypass. en m~me temps chez un m~me individu. Les 2. Ensure adequate haemostasis and blood formes les moins importantes (atdlectasie clotting after bypass to prevent blood loss, discoi'de et at61ectasie sous-segmentaire) ont 6td paratracheal haematomata and re- rencontrdes les plus souvent. L'at61ectasie se re- operation. trouvait le plus souvent aux bases (3/4 des cas) et 3. Prevent gastric dilatation by the use of a elle se situait aussi fr6quemment h droite qu'h naso-gastric tube during and after opera- gauche. On retrouvait plus souvent la forme tion. discoide d'atdlectasie chez les malades ayant eu 4. Keep the time of cardiopulmonary bypass des circulations extracorporelles de courte to a minimum. duree; il y avait corrdlation significative entre la 5. Prevent and treat heart failure in the early forme segmentaire et une balance liquidienne postoperative period. positive importante en fin de C.E.C., entre l'at~lectasie de type sous-segmentaire et les in- SUMMARY terventions pour contr61e d'h6mostase: il y avait de m6me corr61ation significative entre Finci- Radiological evidence of pulmonary complica- dence d'atelectasie et la prdsence de dilatation tions and possible aetiological factors were in- gastrique. vestigated in 50 consecutive patients after heart Par ailleurs, I'on n'a pu etablir de corrdlation operations with cardiopulmonary bypass. entre l'incidence d'at61ectasie el les donn6es Atelectasis was the most frequent pulmonary cliniques et hemodynamiques pr6-0p6ratoires, la complication except for small pleural effusions, durde de C.E,C. ou I'emploi de contrepulsation with an incidence of 64 per cent. Several types of intra-aortique. atelectasis frequently co-existed, with a pre- Le travail discute du mecanisme de production dominance of the less extensive plate and sub- de l'atdlectasie. GALE, el al.: ATELECTASIS AFTER CARDIOPULMONARY BYPASS 21

ACKNOWLEDGEMENTS diopulmonary bypass. A clinical study in adults. Canad. Anaesth. Soc. J.21:181 (1974). 5. GALE, G.D. & SANDERS, D.E. The Bartlett- We are most grateful for support and encour- Edwards incentive spirometer: a preliminary as- agement in writing this paper from Professor sessment of its use in the prevention of atelectasis A.A. Scott, for statistical analysis by Ms. B. after cardio-pulmonary bypass. Canad. Anaesth. Kuzin, and for financial assistance from the Soc. J. 24:408 (1977). 6. FRASER, R.G. & PARI~, J.A. Diagnosis of diseases Marion Webster Taylor Fund of Toronto General of the chest. Volume 1,301. W.B. Saunders Com- Hospital. pany, Toronto (1970). 7. FLEISCHNER, F. Uber das Wesen der basalan REFERENCES horizontalen shattenstreifen in Lungenfield. Wien, Arch. finn. Med. 28:461 (1936). 1. PROVAN,J.L., AUSTEN, W.G., & SCANNELL, J.G, 8. HILBERMAN, M., KAMM, B., LAMY, M., DIET- Respiratory complications after open-heart RICH, H.P., .MARTZ, K., & OSBORN, J.J. An surgery. J. Thoracic and Cardiovasc. Surg. 51 : 626 analysis of potential physiological predictors of (1966). respiratory adequacy following cardiac surgery. 2. TEMPLETON, A.W., ALMOND, C.H., SEABER, A., J. Thorac. Cardiovasc. Surg. 7/:711 (1976). SIMMONS, C., & MACKENZIE, J. Postoperative 9. BYRICK, R.J., KAY, J.C., & NOBLE, W.H. Ex- pulmonary patterns following cardiopulmonary travascuhn lung water accumulation in patients bypass. Am. J. Cardiol. 961 : 1007 (1966). following coronary artery surgery. Canad. 3. GAUERT, W.B., ANDERSON, D.S., REED, W.A., & Anaesth. Soc. J. 24:332 (1977). TEMPLETON, A.W. Pulmonary complications fol- 10. ANDERSON, N.B. & Cma, J. Pulmonary function, lowing extracorporeal circulation. Southern Med. cardiac status, and postoperative course in relation J. 64:679 (1971). to cardiopulmonary bypass. J. Thorac. Car- 4. TURNBULL, K.W., MIYAGISHIMA, R.T., & GE- diovasc. Surg. 59:474 (1970). REIN, A.N. Pulmonary complications and car-