Key words: activities of daily living; complications; comprehensive geriatric assessment; mini-mental state examination; thoracic surgery elderly

Transcription

1 Importance of a Comprehensive Geriatric Assessment in Prediction of Complications Following Thoracic Surgery in Elderly Patients* Tatsuo Fukuse, MD; Naoki Satoda, MD; Kyoko Hijiya, MD; and Takuji Fujinaga, MD Background: The prevalence of comorbidities and functional impairment among elderly patients may enhance the risk of operation-related complications, but the importance of these conditions in elderly patients undergoing thoracic surgery remains unclear. Methods: One hundred twenty patients > 60 years of age who underwent thoracic surgery were registered prospectively and examined. A comprehensive geriatric assessment (CGA) that evaluated such diverse areas as functional status (ie, performance status and activities of daily living [ADLs] using the Barthel index), comorbidity, nutrition (ie, body mass index, arm-muscle circumference, albumin level, transferrin level, lymphocyte count, and cholinesterase level), and cognitive function (ie, mini-mental state examination [MMSE] and negative emotions for operation) was performed in the 2 weeks before patients underwent the operation. Results: The diseases of the 120 patients were as follows: lung cancer, 85 patients; mediastinal tumor, 14 patients; bullas, 12 patients; and other diseases, 9 patients. Postoperative complications developed in 20 patients (16.7%). The patients with dependence for performing the ADLs, and dementia were more likely to develop postoperative complications (p 0.041, and p , respectively). The patients who experienced longer operation times (ie, > 300 min; p 0.018) were more likely to have complications. The incidence of prolonged air leak in the patients with malnutrition increased seven-fold (p 0.045) and that of postoperative infectious diseases in those patients with obesity increased 24-fold (p ), while all patients who developed delirium had low scores in the MMSE preoperatively (p ). Using multiple logistic regression, the best model was obtained with a combination of MMSE (p 0.031) and the Barthel index (p 0.04). When the operation variables were added to this model, the operation time had the strongest effect (p 0.016). Conclusions: Dependence for the performance of ADLs and impaired cognitive conditions are important predictors of postoperative complications, especially when the operation time is long. CGA is necessary in addition to the conventional cardiopulmonary functional assessment in elderly patients. (CHEST 2005; 127: ) Key words: activities of daily living; complications; comprehensive geriatric assessment; mini-mental state examination; thoracic surgery elderly Abbreviations: ADL activity of daily living; AMC arm muscle circumference; BMI body mass index; CGA comprehensive geriatric assessment; MMSE mini-mental state examination; PS performance status; VATS video-assisted thoracic surgery Recently, the number of elderly people in the population has grown rapidly. Aging is associated with an increasing prevalence of multiple diseases and disabilities. It is also associated with a decline of the functional reserve of multiple organ systems, and *From the Department of Thoracic Surgery (Drs. Fukuse and Fujinaga), Faculty of Medicine, Kyoto University, Sakyo-ku, Kyoto, Japan; and the Department of Thoracic Surgery (Drs. Satoda and Hijiya), Otsu Red Cross Hospital, Otsu-City, Japan. Manuscript received January 9, 2004; revision accepted September 2, a progressive restriction in personal and social resources. 1 4 Aging has been demonstrated as a risk factor for lung resection by many authors, 5 7 due to the overall frailty and increased incidence of comor- Reproduction of this article is prohibited without written permission from the American College of Chest Physicians ( permissions@chestnet.org). Correspondence to: Tatsuo Fukuse, MD, Department of Thoracic Surgery, Faculty of Medicine, Kyoto University, 54 Shogo-in, Kawahara-cho, Sakyo-ku, Kyoto , Japan; fukuse@kuhp.kyoto-u.ac.jp 886 Clinical Investigations

2 bid diseases. Previously, it was reported 8 13 that older age, smoking, greater extent of surgical resection, and reduced FEV 1 are associated with increased postoperative morbidity and mortality. In the multiinstitutional experience of the US Lung Cancer Study Group, 5 the mortality rate from pulmonary resection increased in a linear fashion with increasing age, as follows: patients 59 years of age, 1.3%; patients 60 to 69 years of age, 4.1%; patients 70 to 79 years of age, 7%; and patients 80 years of age, 8.1%. The Japanese Association for Chest Surgery 6 also showed that age was an important risk factor. Those patients who were 60 years old had a death rate of 0.4%, those aged 60 to 69 years had a death rate of 1.3%, those aged 70 to 79 years had a death rate of 2%, and those aged 80 years of age had a death rate of 2.2%. On the other hand, it has also been reported 9 11 that the risk of chronologic age per se does not appear to be very important. The appropriate assessment of geriatric patients undergoing thoracic surgery has become increasingly important. 1,2 Over the last few years, geriatricians have developed and validated an approach to care for the elderly called comprehensive geriatric assessment (CGA). 1 4 CGA evaluates the comorbid illnesses, mental status, nutritional status, living circumstances, social support systems, and polypharmacy. The goal of CGA is to provide information independent from the already used measures. We have, therefore, prospectively assessed the prognostic value of CGA in the prediction of postoperative outcome for older patients undergoing thoracic surgery. Materials and Methods This study was conducted at the Department of Thoracic Surgery of Otsu Red Cross Hospital. It was a prospective study of consecutive patients who were 60 years of age and had undergone thoracic surgery (including video-assisted thoracic surgery [VATS]) under general anesthesia from August 2000 to May All patients gave written informed consent, and the study was approved by the Ethics Committee of the Otsu Red Cross Hospital. The decision to undertake surgical resection was made by the surgeon performing the operation using the clinical, radiologic, and spirometric (except pneumothorax) criteria, without reference to the results of the investigations of mental status, nutritional status, or activities of daily living (ADLs) status. Preoperative spirometry was performed to determine FVC and FEV 1. The predicted postoperative FEV 1 was calculated with the following formula: preoperative FEV 1 (1 [the number of subsegments to be resected/the total number of subsegments of the lung]). The criteria for lung resection (except for pneumothorax) were considered when the postoperative FEV 1 was 800 ml/m Antiembolism stockings were used to prevent deep vein thrombosis during and 24 h after the operation in patients whose body mass index (BMI) values were 25. Single-lung ventilation was performed using a Robertshaw dual-line endobronchial tube during the operation. A thoracic paravertebral nerve block was performed and perioperative and postoperative analgesia were achieved with a routine continuous infusion of bupivicaine for 72 h. All patients were treated with IV antibiotics on induction of anesthesia and for at least 48 h after the operation. All operations were performed by one of three surgeons (ie, Drs. Fukuse, Fujinaga, or Satoda), and the postoperative care was supervised in the thoracic surgical ward or, if necessary, in the ICU. Data Collection The demographic, perioperative, and pathologic data were collected prospectively for all patients who were referred to the Department of Thoracic Surgery at Otsu Red Cross Hospital and were entered into a computerized database. Demographic data for patients, including age, gender, history of thoracotomy, and presence of serious medical illness, including digestive diseases, diabetes mellitus, liver diseases, metabolic diseases, hematologic diseases, collagen diseases, renal insufficiency, COPD, and cardiovascular disease (ie, history of infarction, angina pectoris, or congestive heart failure) were recorded on hospital admission. The functional status of patients at hospital admission was assessed using performance status (PS), defined according to the Eastern Cooperative Oncology Group and the Barthel index of ADLs, which measures the level of functional independence in the following six categories: bathing; dressing; using a bathroom; moving from one place to another; continence; and feeding. 14 The cognitive status was assessed using the Folstein mini-mental state examination (MMSE), a 30-item interviewer-administered assessment of several aspects of cognitive function. 15 The assessments of ADLs and MMSE, and the measurement of the mid-arm muscle circumference (AMC) were performed by the medical doctors in charge (Drs. Fujinaga, Satoda, and Hijiya), and they confirmed their data in the weekly conference of the thoracic surgical unit. The nutritional assessment was performed using data on anthropometrics (ie, height, weight, triceps skinfold, and AMC), and levels of serum albumin, serum transferrin, and cholinesterase. The markers of immunologic function (ie, total lymphocyte count) and other variables, such as levels of hemoglobin, sodium, potassium, BUN, and serum creatinine, were assessed. Complications and Outcomes Complications were assessed prospectively by the doctor in charge and were confirmed at the conference of the Division of the Thoracic Surgery in the hospital until hospital discharge and up to 90 days after the operation following successful hospital discharge. Complications included the following: prolonged air leak; persistent air leak for 7 days requiring intercostal drainage; chylothorax; the presence of nonclotting milky fluid obtained from chest tube drainage, and in which free microscopic fat was found; transient delirium; multiple acquired cognitive deficits (eg, memory, orientation, language, constructions, and executive control); arrhythmias (eg, atrial arrhythmia, fibrillation, flutter, tachycardia, or bradyarrhythmia) requiring treatment; bronchopneumonia; new shadow on a chest radiograph with purulent sputa treated with antibiotics; atelectasis; interstitial pneumonia seen on a chest radiograph and confirmed by bronchoscopy samples; new lung infiltrate seen on a chest radiograph; the distribution and description of the specific abnormalities noted on chest CT scan; pyelothorax; and purulent fluid from the pleural space with positive bacteriologic findings. The time from the operation to diminishing air leakage and the removal of all intercostal drains were noted. The time to hospital discharge was noted but was not assessed because some patients underwent adjuvant therapy consecutively. CHEST / 127 / 3/ MARCH,

3 Statistical analysis was performed utilizing a statistical software package (StatView, version 5.0; SAS Institute; Cary, NC). Univariate analysis was done utilizing the 2 test and Fisher exact test. A probability value of 0.05 was considered to be significant. Multivariate logistic regression analysis of complications was performed using variables with p 0.05 in univariate analysis. Results Patients From August 2000 to May 2003, 295 patients underwent thoracic surgery in our division. Of these patients, 120 patients were 60 years of age. Of those 120 patients, 60% had lung cancer and 52% had undergone lobectomy. In general, 65% of these 120 patients had one or more comorbid conditions (Table 1). Fifteen patients (12.5%) had a BMI of 18.5; 17 patients (14.2%) were obese with a BMI of 25 (26.7% in total). Similarly, 27.0% of the patients had a diminished AMC. Quite a few patients had low albumin levels (19.5%), transferrin levels (16.4%), cholinesterase levels (17.8%), and lymphocyte counts (15.0%). These patients presented with good functional and mental status, as measured by the PS score (PS 2, 91.8% of patients), the ADL scale (independent, 89.7% of patients), and MMSE (normal, 91.4% of patients). Table 1 Backgrounds of 120 Patients > 60 Years of Age Who Had Undergone Surgery Variable Values Age,* yr 70.3 (6.2)/60 84 Gender, No. Male 72 Female 48 Diagnosis, No. Primary lung cancer 69 Metastatic lung tumor 16 Pneumothorax (bullas) 12 Other lung diseases 17 Mediastinal tumor 6 Operation type Lobectomy 63 (60/3) Segmentectomy 4 Pneumonectomy 2 Partial resection 42 (20/22) Mediastinal or chest wall tumor extraction 9 (6/3) Preoperative complication, No. No 41 Yes 79 Myocardial disease, arrhythmia 13 Hypertension 43 COPD, emphysema 16 Diabetes mellitus 7 Renal, liver, or other diseases 17 *Values given as mean (SD)/range. Values in parentheses given as No. of patients who had undergone open surgery/no. of patients who had undergone VATS. Postoperative complications developed in 20 patients (16.7%). Seven patients had prolonged air leak ( 7 days), two patients had chylothorax (one needed thoracotomy), three patients had transient delirium, two patients had arrhythmias, three patients had bronchopneumonia, one patient had atelectasis, one patient had interstitial pneumonia, and one patient had pyelothorax. Of these patients, one with diabetes mellitus, pulmonary emphysema, and chronic renal failure died of interstitial pneumonia on postoperative day 25. The mean ( SD) duration of air leak was days, and the mean intercostal drainage time was days. Univariate Analysis for Complications The patients who experienced a longer operation time (ie, 300 min vs 300 min; p 0.018) were more likely to have complications after the operation. The specific operative procedures (eg, lobectomy, partial resection, under thoracotomy, or VATS) or combined resections did not correlate with the complications assessed (Table 2). Patients with a disability in the performance of ADLs were more likely to develop postoperative complications (p 0.041). Dementia also strongly correlated with the postoperative complications (p ) [Table 3]. The single nutritional indexes such as albumin level, transferrin level, or lymphocyte count did not predict the postoperative prolongation of air leak; however, the abnormalities of all these factors correlated with prolonged air leak (p 0.045). Obese patients (ie, those with a BMI of 25) were more likely to have postoperative infectious diseases such as pneumonia and wound infection (p ). Octogenarians or patients with dementia showed a significant relationship to transient delirium (p and p , respectively) [Table 4]. Multivariate Analysis for Complications The preoperative nutritional status (ie, albumin level, transferrin level, lymphocyte count, cholinesterase level, BMI, and AMC), pulmonary function (ie, FEV 1 and FVC), comorbid diseases, smoking, functional status (ie, PS and Barthel index), cognitive function (ie, depression for operation by MMSE), chronologic age, and gender were assessed using the multiple logistic regression model, and the best model was obtained with a combination of MMSE (p 0.031) and Barthel index (p 0.04). When the operation variables (ie, operation time, volume of blood loss, extended resection, and the manner of thoracotomy) were added to this model, the operation time had the strongest effect (p 0.016). Larger volume of blood loss and larger extent of resection also had an additional explanatory effect but were not statistically significant (Table 5). 888 Clinical Investigations

4 Table 2 Relationship of Smoking, Disease, Preoperative Pulmonary Function, and Procedure Performed to Postoperative Complication* Variable No. Complication, % p Value Total Disease Lung cancer Others Smoking Never smoker Ex- or current smoker Preoperative pulmonary function VC, % predicted 80% % FEV 1, % predicted 70% % 20.7 Operative procedure Lobectomy Pneumonectomy Combined resection Yes No Open surgery or VATS Open VATS Pleural adhesion Yes No Operative duration, min Operative bleeding, ml *VC vital capacity. p Discussion Table 3 Univariate Analysis of Preoperative CGA Related to Postoperative Complications Variable No. Complication, % p Value Total Age, yr Preoperative complications Yes No PS 0, , ADLs, Barthel index score * BMI 18.5 or Mid-AMC, cm Female 17/male Female 17/male Albumin, g/dl Transferrin, g/dl Lymphocyte count, cells/ L Hemoglobin, g/dl Negative emotions for operation Yes (0, 1, 2) No (3) Dementia, MMSE score * *p strong; 1 moderate; 2 mild; 3 none. This study demonstrated that the patients who underwent operations with longer time and larger volume of blood loss were more likely to develop postoperative complications. In addition, patients who were ex-smokers or current smokers and had comorbid diseases tended to develop complications. These results are compatible with those of previous reports. 9,16 Unlike the previous studies, however, reduced FEV 1 did not correlate with the high incidence of postoperative complications. This is most likely due to the careful selection of patients according to the predicted postoperative FEV 1 criteria in our department, as described by Olsen et al. 13 Consequently, only 6% of patients in the present study had a preoperative FEV 1 of 70% predicted. Of course, performing further pulmonary function tests, such as carbon monoxide diffusion capacity, maximum inspiratory pressure, or maximum expiratory pressure, might have proved the association between pulmonary function and postoperative morbidity. The patients with dementia estimated by MMSE developed postoperative complications fourfold more than the patients without dementia (56% vs 14%, respectively) in this study. The multivariate analysis confirmed that dementia strongly correlated with the postoperative complications. Kaneko et al 17 studied the relationship between the preoperative cognitive state and the occurrence of postoperative delirium in 36 patients who were 70 years of age undergoing GI operations and reported that the patients who developed postoperative delirium demonstrated preoperative cognitive impairment. To the best of our knowledge, the present study is the first to show that the preoperative dementia correlated CHEST / 127 / 3/ MARCH,

5 Table 4 Univariate Analysis of Preoperative CGA Related to Specific Postoperative Complications Complications Total Complication, % p Value Prolonged air leak Albumin, g/dl Transferrin, g/dl Lymphocyte counts, cells/ L Albumin 3.7, g/dl; * transferrin 229, g/dl; lymphocyte count 999 cells/ L Others Infectious disease BMI * Transient delirium Age, yr * Dementia, MMSE score * *p with the postoperative complications in patients who were undergoing thoracic surgery. Additionally, the functional dependency estimated by the Barthel index strongly correlated with the postoperative complications seen in this study. For the assessment of functional status, the Karnofsky or Eastern Cooperative Oncology Group PS scales have been utilized, and it is well-known that PS correlates with the prognosis of lung cancer. 1,2 However, PS was not a good indicator for predicting the postoperative complications in our study. Repetto et al 3 also reported that the ability to perform ADLs estimated by Barthel index was more sensitive than that assessed Table 5 Best Models From Logistic Regression Analysis of Postoperative Complications* Variable Model Coefficient p Value R 2 Value Complication Intercept Barthel index MMSE Intercept Barthel index MMSE Operation time *Barthel index scale: 0 to 100. MMSE score: 0, 24; 1, 23. by PS alone, and they concluded that many aspects of functional impairment are not fully recognized by PS. These results suggest that the assessment of ADLs by the Barthel index is more sensible than the assessment of PS for the prediction of postoperative complications. In contrast, Dales et al 8 studied 117 consecutive patients undergoing thoracotomy and reported that the quality-of-life measurement was not a good predictor of morbidity. The controversy in the results of their study and ours may be partly because their study subjects were younger patients (ie, 10% of their patients were 50 years of age), and partly because their study included patients with severely impaired pulmonary function (ie, more than half of their patients had an FEV 1 of 60% predicted. The logistic regression equation also suggested that low scores on the Barthel index and MMSE, and longer operation time are risk factors for developing postoperative complications in our study. Consequently, the dependence on others for performing the ADLs and impaired cognitive conditions are important predictors of postoperative complications, especially when the operation time is long. Our study also showed that specific comorbidities predicted specific complications. First, the patients with malnutrition (ie, low levels of albumin, transferrin, and peripheral lymphocyte) were more likely (sevenfold) to have prolonged air leak. This was probably due to the retardation of tissue healing as McClave et al 18 reported that protein deficiency proved to contribute to poor healing rates with less collagen formation and wound dehiscence. Second, octogenarians and those with dementia are more likely (ie, 20-fold to 30-fold) to develop delirium. It is notable that one third of patients with dementia developed delirium postoperatively. Third, obese patients are more likely (ie, 20-fold) to develop postoperative infection. This is consistent with the cohort study reported recently by Dindo et al 19 indicating that obesity was a risk factor for wound infection after surgery. Accordingly, these comorbid conditions should be monitored closely for specific postoperative complications. Our study has some drawbacks. First, various thoracic surgeries were included in this study. However, even the duration and extent of the manipulations were variable, the general anesthesia, thoracotomy, and compression of the lung during operation were performed more or less for every patient. In addition, because the operations were performed by the same team, the procedures were similar for each type of operation. Accordingly, we think that the results of this study were sensitive and reliable for estimating postoperative morbidity. Second, the threshold age of 60 years is not high enough to be a 890 Clinical Investigations

6 criterion for older age, and 70 years would be better. However, two large-scale studies 5,6 demonstrated that the patients who are 60 years of age have significantly higher morbidity and mortality compared with those 60 years of age. From another point of view, our study supports the idea that older age (ie, between the 60s and 80s) alone is not a risk factor for operations. Third, the present results were not cross-validated on additional samples to ensure that they are reliable, because 120 cases did not constitute a large enough sample to do this. In the future, it will be necessary to provide the crossvalidation of the predictors in this study. Besides, it also should be noted that the multiple logistic regression model would be unstable if the sample size is small and the ratio of the number of cases to the number of parameters is If the significance level had been set to 0.05, but not to 0.01, in this study, and the ratio of the cases to the number of parameters were 10 (120/12 10), the results of this study would be reliable and reproducible. In conclusion, the dependence of ADLs and impaired cognitive conditions were important predictors of postoperative complications. Because CGA adds substantial information to the functional assessment of elderly patients undergoing thoracic surgery, the intensive care of elderly patients after thoracic surgery might be improved by CGA, such as with the Barthel index and MMSE, in addition to conventional cardiopulmonary functional assessment. References 1 Reuben DB. Geriatric assessment in oncology. Cancer 1997; 80: Zagonel V. Importance of a comprehensive geriatric assessment in older cancer patients. Eur J Cancer 2001; 37(suppl): S229 S233 3 Repetto L, Fratino L, Audisio RA, et al. Comprehensive geriatric assessment adds information to Eastern Cooperative Oncology Group performance status in elderly cancer patients: an Italian Group for Geriatric Oncology Study. J Clin Oncol 2002; 20: Extermann M, Overcash J, Lyman GH, et al. Co-morbidity and functional status are independent in older cancer patients. J Clin Oncol 1998; 16: Ginsberg RJ, Hill LD, Eagan RT, et al. Modern thirty-day operative mortality for surgical resections in lung cancer. J Thorac Cardiovasc Surg 1983; 86: Wada H, Nakamura T, Nakamoto K, et al. Thirty-day operative mortality for thoracotomy in lung cancer. J Thorac Cardiovasc Surg 1998; 115: Damhuis RA, Schutte PR. Resection rates and postoperative mortality in 7,899 patients with lung cancer. Eur Respir J 1996; 9: Dales RE, Dionne G, Leech JA, et al. Preoperative prediction of pulmonary complications following thoracic surgery. Chest 1993; 104: Ebner H, Sudkamp N, Wex P, et al. Selection and preoperative treatment of over-seventy-year-old patients undergoing thoracotomy. Thorac Cardiovasc Surg 1985; 33: Naunheim KS, Kesler KA, D Orazio SA, et al. Lung cancer surgery in the octogenarian. Eur J Cardiothorac Surg 1994; 8: Osaki T, Shirakusa T, Kodate M, et al. Surgical treatment of lung cancer in the octogenarian. Ann Thorac Surg 1994; 57: Krasna MJ, Deshmukh S, McLaughlin JS. Complications of thoracoscopy. Ann Thorac Surg 1996; 61: Olsen GN, Block AJ, Swenson EW, et al. Pulmonary function evaluation of the lung resection candidate: a prospective study. Am Rev Respir Dis 1975; 111: Mahoney FI, Barthel DW. Functional evaluation: the Barthel Index. Md State Med J 1965; 14: Folstein MF, Folstein SE, McHugh PR. Mini mental state: a practical method for grading the cognitive state of patients for the clinicians. J Psychiatr Res 1975; 12: Jagoe RT, Goodship TH, Gibson GJ. The influence of nutritional status on complications after operations for lung cancer. Ann Thorac Surg 2001; 71: Kaneko T, Takahashi S, Naka T, et al. Postoperative delirium following gastrointestinal surgery in elderly patients. Surg Today 1997; 27: McClave SA, Snider HL, Spain DA. Preoperative issues in clinical nutrition. Chest 1999; 115(suppl):64S 70S 19 Dindo D, Muller MK, Weber M, et al. Obesity in general elective surgery. Lancet 2003; 361: Vasconcelos AGG, Almeida RMV, Nobre FF. The path analysis approach for multivariate analysis of infant mortality data. Ann Epidemiol 1998; 8: CHEST / 127 / 3/ MARCH,