pulmonary complications following oncological surgery

(1)

pulmonary complications following oncological surgery

Çiğdem ÖZDİLEKCAN¹, Necla SONGUR¹, Bahadır M. BERKTAŞ², Meral DİNÇ¹, Emel ÜÇGÜL³, Uğur OK¹

1 SB. Ankara Onkoloji Eğitim ve Araştırma Hastanesi, Göğüs Hastalıkları, 2 Atatürk Göğüs Hastalıkları ve Göğüs Cerrahisi Eğitim ve Araştırma Hastanesi, 3 Gazi Üniversitesi Tıp Fakültesi, Medikal Onkoloji Anabilim Dalı, Ankara.

ÖZET

Onkolojik cerrahi sonrası gelişen postoperatif pulmoner komplikasyonlarla ilişkili risk faktörleri

Çalışmanın amacı, bir onkolojik cerrahi merkezinde farklı tipte elektif cerrahiler uygulanan hastalarda postoperatif pulmoner komplikasyonlar (PPC)’ın sıklığının saptanması ve bunlarla ilişkili risk faktörlerinin belirlenmesidir. Doksan beş erişkin hasta bir yıl süreyle prospektif olarak çalışmaya alındı. Çalışmaya alınan grupta, pulmoner komplikasyonların belirlenme- sinde yaş, cinsiyet, vücut kitle indeksi, eşlik eden akciğer hastalıkları (kronik obstrüktif akciğer hastalığı, astım, bronşek- tazi, restriktif akciğer hastalığı), operasyonun yeri ve tipi, “The American Society of Anesthesiologists (ASA)” sınıflandır- masına göre fiziksel durum, fizik muayene ve akciğer grafi bulguları, solunum fonksiyon testleri, anestezi süresi ve tipi, cerrahi insizyonun yeri ve uzunluğu, operasyon sonrası nazogastrik tüp varlığı dikkate alındı. Çalışma grubunun PPC oranı

%40 (38/95) olarak hesaplandı. En sık gözlenen komplikasyonlar atelektazi ve bronkospazmdı (%13.7). Çok değişkenli analizde tüm risk faktörleri içinde; abdominal insizyon (p= 0.008), anestezi süresi (p= 0.0001) ve FEV₁< %50 değerlerinin (p= 0.007) pulmoner komplikasyonları diğer değişkenlerden bağımsız olarak etkilediği saptandı. Diğer cerrahi girişimlere göre, çalışma grubunda pulmoner komplikasyon oranının daha yüksek olduğu görüldü. Bu sonuç, kanser hastalarına uygulanan majör rezeksiyon cerrahi seçeneklerinin varlığı ile açıklanabilmektedir. Yine de cerrahi süresinin kısaltılması ve se- çilmiş hasta gruplarında genel anesteziden kaçınılmasının PPC riskini azaltacağı kanısındayız.

Anahtar Kelimeler: Postoperatif pulmoner komplikasyonlar, risk faktörleri, onkolojik cerrahi.

SUMMARY

Risk factors associated with postoperative pulmonary complications following oncological surgery

Ozdilekcan C, Songur N, Berktas BM, Dinc M, Ucgul E, Ok U

Ministry of Health Ankara Oncology Training and Research Hospital, Chest Disease Department, Ankara, Turkey.

The purpose of our study was to determine the incidence of different postoperative pulmonary complications (PPCs) and their associated risk factors in patients who have undergone various elective surgical procedures in an oncological surgery

Yazışma Adresi (Address for Correspondence):

Dr. Çiğdem ÖZDİLEKCAN, Angora Evleri, Buluşmalar Caddesi E12 Blok No: 5, 06530 Beysukent, ANKARA - TURKEY

(2)

Pulmonary complications after surgery are a leading cause of postoperative morbidity and mortality. High incidence of pulmonary complications after surgery is the cause of prolonged hospital stay and costs as well as mortality.

Postoperative pulmonary complications (PPCs) have been reported to occur in 5%-10% of the general patient population and in 4%-22% of patients undergoing abdominal surgery (1,2).

Also the existence of underlying pulmonary disease such as chronic obstructive pulmonary disease (COPD) is typically considered as an important risk factor for postoperative complications (3).

A prescient commentary in 1910 by W. Pasteur pointed the direction to our current understan- ding of the etiology of postoperative pulmonary complications. He noted that ‘when the true history of postoperative lung complications comes to be written, active collapse of the lung, from deficiency of inspiratory power, will be found to occupy an important position among determi- ning causes (4). Despite many advances in me- dical and surgical practice, the incidence of PPCs has not changed appreciably over the past 35 years (5). Most PPCs develop as a result of changes in lung volumes that occur in response to dysfunction of muscles of respiration and other changes in chest wall mechanics (6). Vari- ous PPCs may occur in surgery patients including pneumonia, pulmonary embolism, bronchospasm, atelectasis, exacerbation of underlying chronic lung disease, respiratory insuffici- ency leading the need of mechanical ventilation

(1,4). The risk factors associated with PPCs are defined in two groups which are patient-related risk factors (smoking, general health status, age, obesity, underlying pulmonary disease) and procedure-related risk factors (surgical site, duration of surgical procedure, type of anesthesia, type of drug used during anesthesia) (7).

In this study, our purpose was to evaluate the incidence of PPCs of patients who were undergone various surgical procedures (abdomen, genitourinary, head and neck surgery) in an oncological surgery center. As predictors of possible PPCs, utility of patient-related and procedure-related risk factors were investigated.

MATERIALS and METHODS

Ninety-five adult patients referred to department of Surgeries at Ankara Ministry of Health Onco- logy Hospital for elective non laparascopic abdominal, genitourinary, head and neck surgery were studied prospectively for one year period.

All the patients were evaluated average 3-4 days before the operation and PPCs were observed within 10 days of hospitalization. The hospital’s Review Board approved the study and informed consent was obtained from patients.

Preoperative respiratory status of all patients was assessed by clinical examination. Detailed history of respiratory symptoms, accompanying pulmonary diseases such as; COPD, asthma, bronchiectasis, restrictive pulmonary disease is obtained from each patient. Patients with respiratory symptoms such as cough, sputum, dyspnea, wheezing were defined preoperatively as center. Ninety five adult patients were studied prospectively for one year period. For the study group, predictors of pulmonary complications of interest were determined as age, gender, body mass index, co morbid conditions (preexisting history of chronic obtructive pulmonary disease, asthma, bronchiectasis, restrictive lung disease), site and type of the operation, smoking history, The American Society of Anesthesiologists (ASA) physical status, physical examination and chest X- Ray findings, pulmonary function tests, type and duration of anesthesia, surgical incision site and length and presence of nasogastric tube suction. The PPC rate of our study group was 40% (38/95). Atelectasis and bronchospasm were the most frequently observed PPCs (13.7%) Among all the risk factors taken into consideration, only three were found to be significant independent predictors of pulmonary complications according to multivariate analysis as follows: incision location concerning abdomen (p= 0.008), duration of anesthesia per hour (p= 0.0001), values of FEV₁< 50% (p= 0.007). Our data revealed that the incidence of PPCs was high in our study group when compared to results of general population. Application of major resection surgeries for cancer patients can be an explanation for this result. Shortening the duration of surgery, avoiding general aneasthesia in selected group of patients may reduce the risk of PPCs.

Key Words: Postoperative pulmonary complications, risk factors, oncological surgery.

(3)

“symptomatic patients”; otherwise “asympto- matic” with the lack of respiratory symptoms.

Patients were classified into four groups according to The American Society of Anesthesiolo- gists (ASA) classification system, which is based on the presence of systemic disturbances.

This classification was defined as: Class 1 (ab- sent), Class 2 (mild), Class 3 (moderate), Class 4 (severe) Class 5 (almost certain to cause death) (8). Each patients’ smoking behavior was noted, by defining the pack-years and time of quit before the surgical procedure. The height and the weight of subjects were noted so as to calculate the body mass index (BMI) of each patient. Age, gender and type of the operation were also noted. Physical examination was focused especially on the respiratory system and presence of decreased breath sound, prolonged ex- piration, rales, rhonchi or wheezes and thoracal deformities observed during inspection were noted. Electrocardiogram (ECG) of each patient was considered as normal or abnormal (P and T-waves change, previous ischemia findings, ar- rithmias etc.) by the evaluation of Internal Medi- cine Department. Chest radiographs were cate- gorized as normal or abnormal for hyperinflati- on, vascular redistribution or edema, atelectasis, effusion, infiltrates and other parenchymal ab- normalities including the sequels of previous lung diseases. Blood levels of total protein and albumin before the operation were recorded. If any kind of accompanying pulmonary disease presenting with physical examination findings was obtained; bronchodilators, antibiotics, mu- colitics, theophylline, and oxygen therapy and for the appropriate patients oral steroids only for three days were started before the operation and the patients were called for a second visit after three days of treatment. Patients were defined as the ones who had been started medication before the operation, or the ones without any kind of medication as mentioned above. Patients with preexisting tracheostomies and thoracic surgery were excluded.

Spirometric tests were performed for each patient with Spirolab MIR in the sitting position before the operation. Forced expiratory maneuver was repeated three times and best values were

recorded on the composite envelope due to the criteria of European Respiratory Society (ERS) (9). The data of European Community for Coal and Steel (ECCS) were used as a reference for normal values. Forced vital capacity (FVC), forced expiratory volume during the first second of FVC (FEV₁), FEV₁/FVC ratio, mean forced expiratory flow during the middle of the FVC (FEF_25-75) and peak expiratory flow (PEF) values were me- asured and the results of these tests were stated as both percent and liter of predicted values.

Using proposed criteria as a guide, we defined severe airway obstruction as follows: FEV₁< 50%

of predicted and FEV₁/ FVC of < 70%; mild-moderate obstruction as follows: FEV₁from 50% to 80% of predicted and FEV₁/FVC of < 70%; and normal FEV₁ > 80% of predicted and FEV₁/FVC > 70% (10). Restrictive defect was defined as normal FEV₁/FVC, but low FEV₁% of predicted.

During the operation, the type of anesthesia and duration of the surgical procedure, type and site of the incision was recorded. Also the drugs used for the medication of anesthesia were noted.

After the surgical procedures patients were observed for the occurrence of pulmonary complications. Pulmonary complications were classified as death (defined as, death occuring during the same hospital stay), pneumonia (was defined as a new infiltrate on a chest radiograph combined with fever, leukocytosis and a positive sputum Gram stain or culture), pleurisy; (was defined radiologically with the evidence of pleural fluid by puncture), bronchitis (diagnosed if dyspnea, purulent sputum, wheezing, rhonchus developed in a stable patient preoperatively, asthma exacerbation (defined as the occurrence of asthma symptoms in a patient with or without history of asthma, but in stable status preoperatively), atelectasis (diagnosed if the patient had clinical and radiologic evidence of collapse), pulmonary embolism (diagnosed if the patient had clinical suspicion confirmed with the ventilation-perfusion lung scan).

The days of the occurrence of PPCs were also recorded. Appropriate treatments for each complicated patient were applied.

(4)

Statistical Methods

For descriptive analyses, we used 2 tests for ca- tegorical variables and Student’s t-tests for con- tinuous variables. To identify preoperative risk indicators of PPCs we did univariate logistic reg- ressions. The cut off values were chosen to ref- lect extreme values over normal range, or the clinical usefulness of these parameters. We then used Spearman correlation coefficients to scan for closely correlated, or collinear, variables to avoid multicollinearity in the final regression models. Collinear variables would be closely associated with both pulmonary complications and each other and so cause instability in multivariable regression models. To then identify independent risk factors of pulmonary complications we used multivariable conditional logistic regression analysis with preoperative variables most significantly associated with pulmonary complications on univariate analyses. A forward stepwi- se logistic approach was used with p< 0.05 as a limit for entering new variables. Crude and adjusted odds ratios (ORs) and their 95% confidence intervals were calculated using standard methods.

The utility of the model was assessed by correct classification, sensitivity, and specificity. To assess goodness of fit, the Hosmer-Lemeshowl statistic was calculated. The receiver operating characteristic (ROC) statistic (c statistic) was also calculated. The ROC describes the continu- ous trade-off between sensitivity and specificity, and ranges from 0.5 for random noise (a model no better than chance) to 1.0. Analyses were performed using SPSS software (SPSS Inc; Chi- cago, IL) (11).

RESULTS

Of ninety-five patients included in the study, forty-two were female; fifty-three were male.

Mean age of patients were calculated as 60 ± 12 (min 20-max 85). The characteristics and demographics of patients due to the occurrence of PPCs are shown in Table 1. PPC rate was 40%

(38 of 95 patients).The duration of complications after the postoperative period was 4 ± 2 days. The types of pulmonary complications and their rates are shown in Table 2.

Univariate Analysis

As shown in Table 3, the following risk factors had significant crude ORs (p= 0.10): Abdominal incisions (p= 0.004), general anesthesia (p=

0.018), prolonged anesthesia duration (p=

0.0001), history of bronchiectasis (p= 0.099), presence of nasogastric tube (p= 0.0001), decreased FEV₁and FVC (p= 0.018 and 0.063 respectively), presence of air flow limitation on pulmonary function tests (p= 0.083).

Multivariate Analysis

The 7 risk factors with significant crude ORs were submitted to multivariate logistic regression analysis. Three risk factors were found to be significant independent predictors of pulmonary complications (Table 4).

Every additional hour of anesthesia duration placed patients at a risk 1.7 times higher. Abdo- minal incision set patients at a risk 11.4 times higher than genitourinary incisions. Reduced FEV₁than 50% of predicted value, situated patients at a risk 12 times higher than someone who had normal FEV₁ value. Even mildly reduced FEV₁caused 6.5 times higher risk of PPC.

DISCUSSION

PPCs are still an important health problem in patients undergoing surgical procedures. The ove- rall incidence of PPCs following abdominal surgery is approximately 20%, however estimations vary widely in literature (20 to 69%) (12). This variability is due primarily to the type of PPC studied, clinical criteria used in definition, and differing surgical populations.

In our study group, the incidence of PPCs was 40%. The most frequent type of complication observed was atelectasis and bronchospasm ha- ving the same percentage (13.7%). The current incidence of clinically significant atelectasis after abdominal surgery was reported approximately 15-20%, which is a common sequela of abdominal operations performed with general anesthesia (13). In our study group general anesthesia was preferred other than regional anesthesia (local, spinal, epidural) in great ma- jority of patients (82.1%-17.9% respectively).

(5)

In a previous study of Kocabaş et al concerning abdominal surgery patients, the incidence of PPCs were reported as 31% whereas Lawrence

et al reported 55 cases complicated among 82 cases in their cohort study for elective abdominal surgery (14,15).

Table 1. Characteristics and demographics of patients with and without postoperative pulmonary complications.

Characteristic No PPC (n= 57) PPC (n= 38) Total (n= 95)

Sex, No. (%)

Male 34 (59.7) 19 (50.0) 53 (55.8)

Female 23 (40.4) 19 (50.0) 42 (44.2)

Diagnosis, No. (%)

Malignancy 44 (57.9) 32 (84.2) 76 (80.0)

Non-malignancy 13 (68.4) 6 (15.8) 19 (20.0)

Presence of comorbid conditions, No. (%)

History of asthma 4 (7.0) 1 (2.6) 5 (5.3)

History of COPD 15 (26.3) 11 (28.9) 26 (27.4)

History of bronchiectasis 2 (3.5) 5 (13.2) 7 (7.4)

History of restrictive chest diseases 7 (12.3) 4 (10.5) 11 (11.6)

ASA physical status, No. (%)

I 32 (56.1) 18 (47.4) 50 (52.6)

II 23 (40.4) 17 (44.7) 40 (42.1)

III 2 (3.5) 3 (7.9) 5 (5.3)

Type of procedure, No. (%)*

Extremity 2 (3.5) 3 (7.9) 5 (5.3)

Abdomen 20 (35.1) 27 (71.1) 47 (49.5)

Head and neck 20 (35.1) 6 (15.8) 26 (27.4)

Genitourinary 15 (26.3) 2 (5.3) 17 (17.9)

Type of anesthesia, No. (%)^#

General 42 (73.7) 36 (94.7) 78 (82.1)

Other (local, spinal, epidural) 15 (26.3) 2 (5.3) 17 (17.9)

Mean age, yr (± SD) 60.9 (13.4) 60.3 (12.5) 60.7 (13.0)

Mean BMI 25.0 (5.8) 24.6 (5.6) 24.8 (5.7)

Mean anesthesia duration, hour* 2.7 (2.0) 5.1 (2.2) 3.6 (2.4)

Mean incision length, cm 17.2 (8.2) 20.3 (6.9) 18.8 (7.7)

Mean history of smoking, pack-years 17.7 (23.3) 18.1 (23.1) 17.9 (23.1)

Spirometry

Mean FEV₁, L^# 2.9 (0.7) 1.8 (0.7) 2.0 (0.7)

Mean percent predicted FEV₁^# 86.9 (20.5) 70.4 (21.5) 77.9 (21.7)

Mean FVC, L 2.8 (1.0) 2.5 (0.9) 2.7 (0.9)

Mean percent predicted FVC^# 86.0 (20.3) 77.0 (20.2) 82.4 (20.6)

FEV₁/FVC 78.9 (13.2) 75.5 (11.4) 77.5 (12.5)

Mean serum total protein g/dL 6.7 (1.0) 6.6 (0.8) 6.7 (1.0)

Mean serum albumin g/dL 3.9 (0.7) 3.8 (0.7) 3.8 (0.7)

* p< 0.01, ^#p< 0.05, PPC: Postoperative pulmonary complication, No PPC: Without any postoperative pulmonary complication, FVC: Forced vital capacity, FEV₁: Forced expiratory volume in one second, BMI: Body mass index, SD: Standart deviation, FEV₁: Forced expiratory volume in one second, ASA: American Society of Anesthesiologist, COPD: Chronic obstructive pulmonary disease.

(6)

Our study group included not only abdominal, but head-neck, genitourinary and extremity surgeries as well. From this point of view, we did not evaluate just a specified group of patients;

however the evaluation was based on great ma- jority of cancer patients undergoing abdominal resection surgeries, since the results of multivariate analysis indicated abdominal surgery as an independent risk factor. This factor can be con-

sidered as a potential weakness in this study with lack of control group and relatively small sample size, but a larger sample would have re- quired a longer study period or multicenter trial.

At this point we share the same idea with Wong et al who emphasized the same handicap in their trial (1). When all the predictors of pulmonary complications were taken into consideration, only three risk factors were independently associated with pulmonary complications in logistic regression analysis. These were abdominal incisions, duration of anesthesia, and FEV₁values.

Neither age, ASA physical status, BMI, type of anesthesia, presence of malignant disease, accompanying comorbid conditions (such as history of asthma, COPD, and restrictive chest disease) nor history of smoking and cessation time were predictive results of PPCs. These datas were really surprising when evaluating previous study results.

Table 2. Types of pulmonary complications.

Types of complications No (%)

Atelectasis 13 (13.7)

Bronchospasm 13 (13.7)

Pneumonia 11 (11.6)

Effusion 8 (8.4)

Respiratory failure 5 (5.3)

Thromboembolism 2 (2.1)

Table 3. Variables significantly associated with pulmonary complications on univariate analysis (p< 0.10).

Variable p Crude OR (95% CI)

Incision location

Abdomen 0.004 10.11 (2.07-49.24)

Head and neck 0.360 2.25 (0.40-12.71)

Genitourinary, index Anesthesia type

General 0.018 6.42 (1.38-29.93)

Other, index

Anesthesia duration, per hour 0.0001 1.69 (1.34-2.13)

History of bronchiectasis 0.099 4.16 (0.76-22.69)

Nasogastric tube 0.0001 6.43 (2.49-16.58)

FEV₁, per L increment 0.018 0.46 (0.24-0.88)

FEV₁percent of predicted

≥ 80%, index

≥ 50% and < 80% 0.001 5.26 (2.00-13.80)

< 50% 0.030 4.67 (1.16-18.82)

FVC, per L increment 0.063 0.64 (0.40-1.02)

FVC percent of predicted 0.041 0.98 (0.96-0.99)

Pulmonary function test category Normal, index

Air flow limitation 0.083 2.36 (0.89-6.26)

Restrictive defect 0.506 1.50 (0.46-4.95)

OR: Odds ratio, FVC: Forced vital capacity, FEV₁: Forced expiratory volume in one second.

(7)

In the study of Brooks et al; age ≥ 60 years, BMI

≥ 27, presence of malignant disease, impaired cognitive function in the preoperative setting and positive smoking history within the past 8 weeks were risk factors significantly associated with PPCs. Only site of incision (upper abdominal or both upper and lower abdominal) was a common predictive risk factor when compared with our study (12). Also in the study of Smeta- na, the site of the surgical incision found to be the most important predictor of PPCs (15).

In fact, our study’s univariate analysis results revealed 7 risk factors as predictors of PPCs: Ab- dominal incisions, general anesthesia, prolonged anesthesia duration, history of bronchiectasis, presence of nasogastric tube, decreased FEV₁, FVC, and presence of airflow limitation on pulmonary function tests. However, among all of these factors only prolonged anesthesia duration, incisions of abdomen, and reduced FEV₁values remained as independent factors in multivariate analysis.

Smoking history is frequently cited in univariate analysis, but does not remain as an independent risk factor in multivariate analysis just like our study’s risk factor evaluation results (1,5,16,17). Recent trials have shown that, patients with severe COPD suffered more high gra- de pulmonary complications and had high incidence of PPCs with poor long term survival (2,11). Our study group included 26 (27.4%) COPD patients, but COPD was not an indepen-

dent risk factor according to our data. However determination of FEV₁values and airflow limitation as an independent factor may point out that, our COPD patient group mostly contained both mild and moderate forms, and avoid the severe forms from general anesthesia.

Early studies suggested that, spirometric results were predictive of PPCs. Just similar to our results, these studies demonstrated an association between abnormal pulmonary function or forced expiratory volume in one second (FEV₁) and the incidence of various postoperative complications (12). However, other studies have pointed out that although, spirometry is widely used to assess pulmonary risk, if used alone; it had limi- ted clinical value as a screening test to predict PPCs (3,18). These data show that, the clinical utility of preoperative spirometry as a predictor of PPCs therefore remains unclear.

History of cancer has been cited as a risk factor in other studies, however not as an independent risk factor (19). Of the 95 patients included in the study, 76 (80%) of them had history of malignant disease. Since the history of cancer was not an independent risk factor in our study group, the explanation for this nearly high PPC rate (40%) could be the result of major cancer surgeries which require long time and application of general anesthesia. For instance, our study population included 12 (12.6%) surgical resections for esophagus cancers and PPCs were observed 58.3% of them. PPCs occur in 25-50% of pati- Table 4. Variables independently associated with pulmonary complications by multivariate conditional logistic regression.

Variable Adjusted OR 95% CI p

Incision location

Abdomen 11.40 1.87-69.51 0.008

Head and neck 2.68 0.35-20.51 0.344

Genitourinary, index 0.011

Anesthesia duration, per hour 1.70 1.28-2.26 0.0001

FEV₁percent of predicted

≥ 80%, index 0.005

≥ 50% and < 80% 6.47 1.78-23.52 0.005

< 50% 12.04 1.96-74.09 0.007

OR: Odds ratio, CI: Confidence interval, FEV₁: Forced expiratory volume in one second.

(8)

ents after esophagectomy (6). These complications arise from a number of factors, including the type of incision used, the extend of mediastinal dissection, the development of recurrent la- rengeal nerve injury that may impair coughing efficiency postoperatively, and the presence of an intrathoracic reconstructive organ or pleural effusion (6,20).

According to the literature, the duration of anesthesia reflects more the underlying disease process of patients rather than an independent risk factor (1). However, our data revealed ‘duration of anesthesia’ as an independent factor.

Surgical procedures involving genitourinary and head and neck systems usually require a short duration of anesthesia which could be considered as low risk for PPCs when compared to the abdominal operations. We have observed that, both type of incision and duration of anesthesia can adversely affect lung function.

We have concluded that, history of concomitant malignant disease was not an independent risk factor for prediction of PPCs in the study group.

However, the application of major surgeries that require long time could be a secondary relative risk factor in these group of patients. Spirometry (especially the values of FEV₁and FVC) and presence of airflow obstruction are the predictors of PPCs. Spirometry may be helpful in patients with COPD or asthma to predict PPCs. On the other hand, the evaluation of spirometric results by chest physicians may guide the anesthesiologists for the selection of anesthesia type or may indi- cate a close postoperative pulmonary care to reduce PPCs.

We underlie that spirometric results can deny a surgery especially concerning patients with FEV₁ values < 50%. Because preoperative pulmonary assessment involves identifying patients at risk for significant complications, assessing the magnitude of the risk, and identifying the factors that may be modified to decrease risk.

We also emphasize that as a risk-reduction stra- tegy, shortening the duration of surgery, avoiding general aneasthesia in selected group of patients may be helpful reduce the risk of PPCs.

REFERENCES

1. Wong DH, Weber EC, Schell MJ, et al. Factors associated with postoperative pulmonary complications in patients with severe chronic obstructive pulmonary disease.

Anesth Analg 1995; 80: 276-84.

2. Hall JC, Talala RA, Mander J. A multivariate analysis of the risk of pulmonary complications after laparatomy.

Chest 1991; 99: 923-7.

3. Kroenke K, Lawrence VA, Theroux JF, et al. Postoperati- ve complications after thoracic and major abdominal surgery in patients with and without obstructive lung disease. Chest 1993; 104: 1445-51.

4. Pasteur W. Active lobar collapse of the lung after abdominal operations. Lancet 1910; 2: 1080-3.

5. Gass GD, Olsen GN. Preoperative pulmonary fuction testing to predict postoperative morbidity and mortality.

Chest 1986; 89: 127-35.

6. Ferguson MK. Postoperative assessment of pulmonary risk. Chest 1999; 115: 58-63.

7. Powell CA, Caplan CE. Pulmonary function tests in preoperative pulmonary evaluation. Clin Chest Med 2001;

22: 703-14.

8. Dripps DR, Lamont A, Eckenfoff JE. The role of anesthesia in surgical mortality. Jama 1961; 178: 261-6.

9. European Respiratory Society. Standardised Lung Func- tion testing. Lung volumes and forced ventilatory flows.

Eur Respir J 1993; 6(Suppl 6): 5-40.

10. Hughes JMB, Pride NB. Lung Function Tests. London:

WB Saunders, 2000: 4-25.

11. Hosmer DW, Lemeshow S. Applied logistic regression.

New York: John Wiley, 1989: 140-5.

12. Brooks JA, Brunn DNS. Predictors of postoperative pulmonary complications following abdominal surgery.

Chest 1997; 111: 564-71.

13. Platell C, Hall JC. Atelectasis after abdominal surgery. J Am Coll Surg 1997; 185: 584-92.

14. Kocabaş A, Kara K, Özgür G, et al. Value of preoperative spirometry to predict postoperative pulmonary complications. Respir Med 1996; 90: 25-33.

15. Lawrence WA, Dhanda R, Hilsenbeck SG, Page CP. Risk of pulmonary complications after elective abdominal surgery. Chest 1996; 110: 744 -50.

16. Smetena GW. Preoperative pulmonary assessment of the older adult. Clin Geriatr Med 2003; 19: 35-55.

17. Fernandez ALG, Pereira EDB, Ancao MS. Evaluation of predict risk factors for pulmonary complications in patients who were undergoing elective upper abdominal surgery. Am J Respir Crit Care Med 1995; 151: 490.

18. Oleh TC, Hnatiuk MC, Corcoran PC, et al. Spirometry in surgery for anterior mediastinal masses. Chest 2001;

120: 1152-6.

19. Velanovich V. The effects of age, gender, race and concomitant disease on postoperative complications. R Coll Surg Edinb 1993; 38: 225-30.

20. Ferguson MK, Martin TR, Reeder LB, Olak J. Mortality after esophagectomy: Risk factor analysis. World J Surg 1997; 21: 599-604.