Radiographic U. Jain,
Pulmonary
MD, PhD, T.L.K.
Abnormalities
Rao, MD, P. Kumar, B.M.
Blakeman,
After Different
MD, B.S. Kleinman, MD, M. Bakhos,
Types of Cardiac
MD, R.J. Belusko,
MD, and D.E. Wallis,
Surgery
MD, PhD, D.P. Kanuri,
MD,
MD
One aim of this study was to determine the incidence of new radiographic pulmonary abnormalities during hospitalization after cardiac surgery. Another aim was to determine if such abnormalities are more common among patients who had left internal mammary artery (LIMA) grafting. The predictive value of radiographic abnormalities for clinically important pulmonary morbidity was also determined. The anteroposterior chest radiographs of 152 patients obtained by portable equipment were evaluated to determine the incidence of new postoperative radiographic pulmonary abnormalities such as atelectasis, consolidation, infiltrate, and pleural effusion. Clinically important pulmonary morbidity was defined as a delay in tracheal extubation or discharge from the hospital because of a pulmonary reason. Among the 89 patients who had LIMA grafting and left pleurotomy, there was an 88% incidence of left-sided pulmonary abnormalities; a 73% incidence of left-sided atelectasis; and a 55% incidence of left-sided effusion. Among the 63 patients who had saphenous vein grafting only and/or valvular surgery, the respective incidences were 68%, 54%, and 35%. which were lower (P 5 0.05) than those in the patients who had LIMA grafting.
There was no significant difference in abnormalities between the saphenous vein grafting and the valvular surgery groups. The 35% incidence of left-sided pleural effusion when LIMA grafting and pleurotomy were not performed was unexpectedly high. There was no association between radiographic abnormalities and age, the duration of cardiopulmonary bypass, and the duration of aortic occlusion, indicating that cardiopulmonary bypass was not a primary etiology of these radiographic abnormalities. Among the patients who had LIMA grafting, pulmonary morbidity contributed to delayed extubation after surgery of one patient and delayed discharge from the hospital of three patients. Among the patients who did not have LIMA grafting, pulmonary morbidity delayed discharge of one patient from the hospital. In a large percentage of patients, chest radiography detected minor pulmonary abnormalities that were rarely associated with a worse clinical outcome after cardiac surgery. Radiographic abnormalities were significantly more common when LIMA grafting was performed.
MORBIDITY is a complication of coronary artery bypass grafting (CABG) surgery.’ Such morbidity is reported to be higher when a left internal mammary artery (LIMA) graft is used than when only saphenous vein grafting is performed.: ” Although pulmonary morbidity after valve surgery has been studied,“.” there are no studies comparing overall pulmonary morbidity between valve surgery and CABG patients. One aim of this study was to determine and compare the incidences of radiographic pulmonary abnormalities in the postoperative period in patients for valvular, vein grafting only, and LIMA and vein grafting operations. Another aim was to determine if such abnormalities are predictive of clinical morbidity. The previous studies primarily evaluated morbidity persisting beyond the first few postoperative days. The studies of pulmonary abnormalities during hospitalization following cardiac surgery have primarily used arterial and mixed venous blood gases” and spirometry. During this investigation, radiographic opacification of the pulmonary fields was studied in detail. The analysis of chest radiographs is of clinical importance because they are routinely obtained during postopcrativc hospitalization.
CABS with LIMA and vein grafting, C’ABG with vein grafting only, any valvular surgery with or without vein grafting, aortic valve surgery only. and mitral valve surgery only. Similar techniques were used by the nine surgeons. Left-sided pleurotomy was performed in all patients undergoing LIMA grafting. The patients having vein grafting or valvular surgery rarely had pleurotomy. When pleurotomy was performed, all fluid was suctioned from the left pleural cavity prior to chest closure and a left pleural drain was placed. The lungs were compressed due to sternal retraction and due to manipulation during the placement of grafts to the coronaries. The left lung was compressed more because the left half of the sternum was retracted more during LIMA harvesting and because the heart lies primarily in the left chest cavity. During cardiopulmonary bypass (CPB) with a membrane oxygcnator. cold crystalloid potassium cardioplegia, systemic hypothermia to 28°C‘. and topical iced saline on the heart were used for myocardial protection. No insulation pad was used to protect the phrenic nerve. After CPB. the patients were ventilated with 100’4 oxygen in the operating room. After surgery, the inspired oxygen concentration was reduced while maintaining an arterial oxygen tension of greater than 60 mm Hg. The tidal volume was approximately 10 mL/kg. Positive end-expiratory pressure was generally 5 mm Hg with higher pressures being required in the presence 01 respiratory insufficiency. Postoperatively, physical therapists administered chest physical therapy and incentive spirometry every 4 hours while awake to all the patients in the intensive care unit. It was determined which patients stayed intubated longer than the day after surgery because of irespiratory insufficiency in the absence of residual effect of narcotics or muscle relaxants. and which patients were delayed in discharge from the hospital because of pulmonary infection or respiratory insufficiency. The clinical courses of these patients were studied. A preoperative upright posteroanterior chest radiograph was obtained. Anteroposterior chest radiographs were obtained in a semirecumbent position. by portable equipment. immediately on completion of the surgery, and daily until transfer out of the intensive care unit. Additional radiographs were obtained throughout the period of hospitalization for specific clinical indication\
P
ULMONARY
MATERIALS
AND METHODS
After approval from the Institutional Review Board, written informed consent to enter the study was obtained from 152 adult patients who underwent median sternotomy for cardiac surgery. The patients were divided into the following surgical categories:
From the Department ofAnesthesiology, Loyola University Medical Center, Maywood, IL. Address reprint requests to U. Jain, MD, PhD, Depamnent of Anesthesiology, Loyola University of Chicago, 2160 S First Ave, Maywood, IL 60153. Copyright G 1991 by W.B. Saunders Company 1053-077019110506-0013$03.00/0 592
Copyright
cc:-1991 by W.B. Saunders
Company
Journalof Cardiothoracic and Vascular Anesthesm, Vol5, No 6 (December), 1991: pp 592.595
POSTOPERATIVE
593
RADIOGRAPHIC ABNORMALITIES
Each radiograph was interpreted by one of six radiologists who were unaware of the type of cardiac surgery performed. All the radiologists used similar criteria described below. The presence of a new pulmonary abnormality on any of the postoperative radiographs was determined. Abnormalities considered to be primarily due to the cardiovascular changes were not included. Increased opacification of the pulmonary fields was classified into atelectasis, consolidation, and infiltrate. Atelectasis was defined as a loss of lung volume. Consolidation was defined as an area of radiographic opacity similar to that of the heart. The area could be irregular, nonsegmental, of ill-defined margins, and any size. Infiltrates were considered to be due to abnormal substances in alveolar or interstitial spaces with coexisting air. The distribution was usually lobar or segmental. Two or three abnormalities could be present in the same lung. The presence of a pleural effusion or noncardiogenic pulmonary edema was also determined. The association between the occurrence of radiographic pulmonary abnormalities and age, CPB period, and aortic occlusion period was studied in the entire group of patients. All the patients recovered to be discharged from the hospital. Records of follow-up visits after hospital discharge were available for all patients and any residual clinically significant pulmonary morbidity was noted. x2 Analysis was used to determine the statistical significance in categorical data. When the sample size was small, Fisher’s exact test was used. The difference between the means of a continuous variable for two groups was evaluated by the t test. The null hypothesis was rejected at P I 0.5.
? ? LIMA ? ?Vein Graft ? ?Valve Surgery
”
I
4
3
2
Radiographic
6
5
Abnormalities
Fig 1. The percent incidences of left-sided (1) radiological abnormalities, (2) atelectasis, (3) consolidation, (4) infiltrate, (5) effusion, and (6) bilateral effusion, for patients with LIMA grafting, vein grafting only, and any valvular surgery with or without vein grafting.
none of the radiographic abnormalities was associated with age or with the duration of CPB or aortic occlusion. Among the 89 patients who had LIMA grafting, pulmonary morbidity could be considered to have contributed to delayed extubation after surgery in one patient and delayed discharge from the hospital in three patients. One patient who had delayed extubation and discharge was a smoker who had preoperative chronic obstructive pulmonary disease and dyspnea. Postoperatively, he had left-sided atelectasis and consolidation, and was extubated on the third postoperative day. After extubation, he developed pneumonia, pulmonary edema, and a right pneumothorax and again required mechanical ventilation. The second patient was a brittle diabetic and smoker who had a preoperative upper respiratory infection with productive cough and dyspnea. Postoperatively, he had a left-sided pleural effusion and bibasilar rhonchi, dyspnea, and tachypnea. He also developed diabetic ketoacidosis and required reintubation and mechanical ventilation. He was discharged 24 days after surgery. The third patient had postoperative left-sided atelectasis, consolidation, effusion, and inadequate dia-
RESULTS
The incidences of various radiographic chest abnormalities that were new in the postoperative period are shown in Table 1 and Fig 1. Right-sided abnormalities were rare. None of the patients had noncardiogenic pulmonary edema. Left-sided atelectasis was the most common abnormality, followed by left-sided pleural effusion and then consolidation. The 35% incidence of left-sided pleural effusion when LIMA grafting was not performed was unexpectedly high. Only one of these patients, who underwent mitral valve replacement, had a left-sided pleurotomy. The effusions were first observed in the immediate postoperative radiograph and were reabsorbed over the next few days. One patient for LIMA grafting and one patient for vein grafting only had bilateral infiltrates. The incidence of abnormalities was not significantly different between the groups for vein grafting only and valvular surgery with or without vein grafting. Table 2 shows that in the entire group of patients,
Table 1. The Incidence of Various Types of Radiographic Abnormalities Associated With Different Types of Cardiac Surgery All Non-LIMA
LIMA (n = 89)
Vein Grafts Only
All Valves 2 Vein Grafts
Aortic Valve Only
(n = 31)
(n = 32)
(n = 8)
(n = 63)
%
NO.
ia
56
15
47
19
11
2
7
11
NO.
%
3
38
7
50
3
38
5
36
34
3
38
4
29
0
0
0
0
0
0
36
11
34
1
13
3
21
4 13 63% 11
5
16
0
0
1
%
NO.
%
NO.
%
NO.
Left-sided atelectasis, consolidation, infiltrate, or effusion
7a*
aa
43
68
25
81
Left-sided atelectasis
65*
73
34
54
19
61
Left-sided consolidation
37
42
17
27
6
Left-sided infiltrate
10
11
2
3
Left-sided effusion
49’
55
22
35
?? P
Pulmonary
MD, PhD, T.L.K.
Abnormalities
Rao, MD, P. Kumar, B.M.
Blakeman,
After Different
MD, B.S. Kleinman, MD, M. Bakhos,
Types of Cardiac
MD, R.J. Belusko,
MD, and D.E. Wallis,
Surgery
MD, PhD, D.P. Kanuri,
MD,
MD
One aim of this study was to determine the incidence of new radiographic pulmonary abnormalities during hospitalization after cardiac surgery. Another aim was to determine if such abnormalities are more common among patients who had left internal mammary artery (LIMA) grafting. The predictive value of radiographic abnormalities for clinically important pulmonary morbidity was also determined. The anteroposterior chest radiographs of 152 patients obtained by portable equipment were evaluated to determine the incidence of new postoperative radiographic pulmonary abnormalities such as atelectasis, consolidation, infiltrate, and pleural effusion. Clinically important pulmonary morbidity was defined as a delay in tracheal extubation or discharge from the hospital because of a pulmonary reason. Among the 89 patients who had LIMA grafting and left pleurotomy, there was an 88% incidence of left-sided pulmonary abnormalities; a 73% incidence of left-sided atelectasis; and a 55% incidence of left-sided effusion. Among the 63 patients who had saphenous vein grafting only and/or valvular surgery, the respective incidences were 68%, 54%, and 35%. which were lower (P 5 0.05) than those in the patients who had LIMA grafting.
There was no significant difference in abnormalities between the saphenous vein grafting and the valvular surgery groups. The 35% incidence of left-sided pleural effusion when LIMA grafting and pleurotomy were not performed was unexpectedly high. There was no association between radiographic abnormalities and age, the duration of cardiopulmonary bypass, and the duration of aortic occlusion, indicating that cardiopulmonary bypass was not a primary etiology of these radiographic abnormalities. Among the patients who had LIMA grafting, pulmonary morbidity contributed to delayed extubation after surgery of one patient and delayed discharge from the hospital of three patients. Among the patients who did not have LIMA grafting, pulmonary morbidity delayed discharge of one patient from the hospital. In a large percentage of patients, chest radiography detected minor pulmonary abnormalities that were rarely associated with a worse clinical outcome after cardiac surgery. Radiographic abnormalities were significantly more common when LIMA grafting was performed.
MORBIDITY is a complication of coronary artery bypass grafting (CABG) surgery.’ Such morbidity is reported to be higher when a left internal mammary artery (LIMA) graft is used than when only saphenous vein grafting is performed.: ” Although pulmonary morbidity after valve surgery has been studied,“.” there are no studies comparing overall pulmonary morbidity between valve surgery and CABG patients. One aim of this study was to determine and compare the incidences of radiographic pulmonary abnormalities in the postoperative period in patients for valvular, vein grafting only, and LIMA and vein grafting operations. Another aim was to determine if such abnormalities are predictive of clinical morbidity. The previous studies primarily evaluated morbidity persisting beyond the first few postoperative days. The studies of pulmonary abnormalities during hospitalization following cardiac surgery have primarily used arterial and mixed venous blood gases” and spirometry. During this investigation, radiographic opacification of the pulmonary fields was studied in detail. The analysis of chest radiographs is of clinical importance because they are routinely obtained during postopcrativc hospitalization.
CABS with LIMA and vein grafting, C’ABG with vein grafting only, any valvular surgery with or without vein grafting, aortic valve surgery only. and mitral valve surgery only. Similar techniques were used by the nine surgeons. Left-sided pleurotomy was performed in all patients undergoing LIMA grafting. The patients having vein grafting or valvular surgery rarely had pleurotomy. When pleurotomy was performed, all fluid was suctioned from the left pleural cavity prior to chest closure and a left pleural drain was placed. The lungs were compressed due to sternal retraction and due to manipulation during the placement of grafts to the coronaries. The left lung was compressed more because the left half of the sternum was retracted more during LIMA harvesting and because the heart lies primarily in the left chest cavity. During cardiopulmonary bypass (CPB) with a membrane oxygcnator. cold crystalloid potassium cardioplegia, systemic hypothermia to 28°C‘. and topical iced saline on the heart were used for myocardial protection. No insulation pad was used to protect the phrenic nerve. After CPB. the patients were ventilated with 100’4 oxygen in the operating room. After surgery, the inspired oxygen concentration was reduced while maintaining an arterial oxygen tension of greater than 60 mm Hg. The tidal volume was approximately 10 mL/kg. Positive end-expiratory pressure was generally 5 mm Hg with higher pressures being required in the presence 01 respiratory insufficiency. Postoperatively, physical therapists administered chest physical therapy and incentive spirometry every 4 hours while awake to all the patients in the intensive care unit. It was determined which patients stayed intubated longer than the day after surgery because of irespiratory insufficiency in the absence of residual effect of narcotics or muscle relaxants. and which patients were delayed in discharge from the hospital because of pulmonary infection or respiratory insufficiency. The clinical courses of these patients were studied. A preoperative upright posteroanterior chest radiograph was obtained. Anteroposterior chest radiographs were obtained in a semirecumbent position. by portable equipment. immediately on completion of the surgery, and daily until transfer out of the intensive care unit. Additional radiographs were obtained throughout the period of hospitalization for specific clinical indication\
P
ULMONARY
MATERIALS
AND METHODS
After approval from the Institutional Review Board, written informed consent to enter the study was obtained from 152 adult patients who underwent median sternotomy for cardiac surgery. The patients were divided into the following surgical categories:
From the Department ofAnesthesiology, Loyola University Medical Center, Maywood, IL. Address reprint requests to U. Jain, MD, PhD, Depamnent of Anesthesiology, Loyola University of Chicago, 2160 S First Ave, Maywood, IL 60153. Copyright G 1991 by W.B. Saunders Company 1053-077019110506-0013$03.00/0 592
Copyright
cc:-1991 by W.B. Saunders
Company
Journalof Cardiothoracic and Vascular Anesthesm, Vol5, No 6 (December), 1991: pp 592.595
POSTOPERATIVE
593
RADIOGRAPHIC ABNORMALITIES
Each radiograph was interpreted by one of six radiologists who were unaware of the type of cardiac surgery performed. All the radiologists used similar criteria described below. The presence of a new pulmonary abnormality on any of the postoperative radiographs was determined. Abnormalities considered to be primarily due to the cardiovascular changes were not included. Increased opacification of the pulmonary fields was classified into atelectasis, consolidation, and infiltrate. Atelectasis was defined as a loss of lung volume. Consolidation was defined as an area of radiographic opacity similar to that of the heart. The area could be irregular, nonsegmental, of ill-defined margins, and any size. Infiltrates were considered to be due to abnormal substances in alveolar or interstitial spaces with coexisting air. The distribution was usually lobar or segmental. Two or three abnormalities could be present in the same lung. The presence of a pleural effusion or noncardiogenic pulmonary edema was also determined. The association between the occurrence of radiographic pulmonary abnormalities and age, CPB period, and aortic occlusion period was studied in the entire group of patients. All the patients recovered to be discharged from the hospital. Records of follow-up visits after hospital discharge were available for all patients and any residual clinically significant pulmonary morbidity was noted. x2 Analysis was used to determine the statistical significance in categorical data. When the sample size was small, Fisher’s exact test was used. The difference between the means of a continuous variable for two groups was evaluated by the t test. The null hypothesis was rejected at P I 0.5.
? ? LIMA ? ?Vein Graft ? ?Valve Surgery
”
I
4
3
2
Radiographic
6
5
Abnormalities
Fig 1. The percent incidences of left-sided (1) radiological abnormalities, (2) atelectasis, (3) consolidation, (4) infiltrate, (5) effusion, and (6) bilateral effusion, for patients with LIMA grafting, vein grafting only, and any valvular surgery with or without vein grafting.
none of the radiographic abnormalities was associated with age or with the duration of CPB or aortic occlusion. Among the 89 patients who had LIMA grafting, pulmonary morbidity could be considered to have contributed to delayed extubation after surgery in one patient and delayed discharge from the hospital in three patients. One patient who had delayed extubation and discharge was a smoker who had preoperative chronic obstructive pulmonary disease and dyspnea. Postoperatively, he had left-sided atelectasis and consolidation, and was extubated on the third postoperative day. After extubation, he developed pneumonia, pulmonary edema, and a right pneumothorax and again required mechanical ventilation. The second patient was a brittle diabetic and smoker who had a preoperative upper respiratory infection with productive cough and dyspnea. Postoperatively, he had a left-sided pleural effusion and bibasilar rhonchi, dyspnea, and tachypnea. He also developed diabetic ketoacidosis and required reintubation and mechanical ventilation. He was discharged 24 days after surgery. The third patient had postoperative left-sided atelectasis, consolidation, effusion, and inadequate dia-
RESULTS
The incidences of various radiographic chest abnormalities that were new in the postoperative period are shown in Table 1 and Fig 1. Right-sided abnormalities were rare. None of the patients had noncardiogenic pulmonary edema. Left-sided atelectasis was the most common abnormality, followed by left-sided pleural effusion and then consolidation. The 35% incidence of left-sided pleural effusion when LIMA grafting was not performed was unexpectedly high. Only one of these patients, who underwent mitral valve replacement, had a left-sided pleurotomy. The effusions were first observed in the immediate postoperative radiograph and were reabsorbed over the next few days. One patient for LIMA grafting and one patient for vein grafting only had bilateral infiltrates. The incidence of abnormalities was not significantly different between the groups for vein grafting only and valvular surgery with or without vein grafting. Table 2 shows that in the entire group of patients,
Table 1. The Incidence of Various Types of Radiographic Abnormalities Associated With Different Types of Cardiac Surgery All Non-LIMA
LIMA (n = 89)
Vein Grafts Only
All Valves 2 Vein Grafts
Aortic Valve Only
(n = 31)
(n = 32)
(n = 8)
(n = 63)
%
NO.
ia
56
15
47
19
11
2
7
11
NO.
%
3
38
7
50
3
38
5
36
34
3
38
4
29
0
0
0
0
0
0
36
11
34
1
13
3
21
4 13 63% 11
5
16
0
0
1
%
NO.
%
NO.
%
NO.
Left-sided atelectasis, consolidation, infiltrate, or effusion
7a*
aa
43
68
25
81
Left-sided atelectasis
65*
73
34
54
19
61
Left-sided consolidation
37
42
17
27
6
Left-sided infiltrate
10
11
2
3
Left-sided effusion
49’
55
22
35
?? P
