Readmission to Intensive Care Unit After Initial Recovery From Major ...

GENERAL THORACIC

Readmission to Intensive Care Unit After Initial Recovery From Major Thoracic Oncology Surgery Suk-Won Song, MD, Hyun-Sung Lee, MD, PhD, Jae-Hyun Kim, MD, Moon Soo Kim, MD, Jong Mog Lee, MD, and Jae Ill Zo, MD, PhD Yonsei Cardiovascular Hospital, Yonsei University College of Medicine, Seoul; Center for Lung Cancer, Research Institute and Hospital, National Cancer Center, Goyang, Gyeonggi, Korea

Background. Little has been published regarding outcomes subsequent to complications after thoracic surgery. The present study investigated outcomes and risk factors associated with mortality in patients admitted to an intensive care unit (ICU) after initial recovery from thoracic oncology surgery. Methods. From March 2001 to August 2005, 1,087 patients underwent major resection for lung or esophageal cancer. Ninety-four (8.6%) of those patients required ICU care after initial recovery, and were the subject of the present retrospective review. Results. The patient group included 85 males (90.4%), of mean age 66 years. Patients were classified as either survivors (n ⴝ 63, 67%) or nonsurvivors (n ⴝ 31, 33%). The most common reason for ICU readmission was pulmonary complication (n ⴝ 73, 77.7%). Sixty-four patients (68.1%) required mechanical ventilation and 42 (43.3%) required renal support. Multivariate analysis showed that the initial acute physiological assessment and chronic health evaluation (APACHE) III score at

readmission to ICU, duration of mechanical ventilation, and renal support were risk factors for in-hospital mortality. The overall three-year survival was 50.6%. Cox analysis showed that survivors who underwent tracheostomy had a poor prognosis (p ⴝ 0.011). Of 12 late mortalities in survivors who underwent tracheostomy, 9 (75%) were due to cancer-unrelated causes. Conclusions. The ICU readmission after thoracic oncology surgery was associated with high in-hospital mortality. Identification of patients with a high APACHE score and (or) prolonged ventilation at readmission may help predict the risk of mortality. Preemptive strategies designed to optimize treatment of such high-risk patients may improve outcomes. Survivors from ICU readmission after thoracic oncology surgery require meticulous and frequent follow-up due to a high risk of deterioration after discharge.

T

comes and requirements of patients who initially recovered from major thoracic oncology surgery but then required “step-up” care due to complications [5, 6]. Such care is associated with prolonged postoperative hospital stays, increased cost of care, and substantial surgical mortality. The identification of factors that predict which patients are at high risk for in-hospital mortality may lower the rate of mortalities through changes in ICU transfer strategies or intensive care. The purposes of this present study were to determine the predictive risk factors for ICU mortality after initial recovery from thoracic oncology surgery and to examine the outcomes in patients who initially recovered from major thoracic oncology surgery but then required ICUbased step-up care after complications.

he ultimate aim of thoracic oncology surgery is curative resection of the cancer without significant associated postoperative morbidity or mortality. However, a high proportion of patients undergoing major thoracic oncology surgery for lung or esophageal cancer suffer from postoperative complications that can only be managed in an intensive care unit (ICU). The factors associated with higher rates of ICU readmission for thoracic oncology surgery patients compared with other major oncology surgery patients are believed to be age, heavy smoking, high alcohol consumption, presence of underlying lung disease, and the extensive surgical procedure itself [1]. There are some published data regarding such mortalities and outcomes in cardiac, general surgery, and medical critical care patients, and also some attempts to elaborate predictive models for outcomes [2– 4]. However, little has been published regarding outAccepted for publication June 26, 2007. Presented at the Forty-third Annual Meeting of The Society of Thoracic Surgeons, San Diego, CA, Jan 29 –31, 2007. Address correspondence to Dr Hyun-Sung Lee, Center for Lung Cancer, Research Institute and Hospital National Cancer Center, 809 Madu1dong, Ilsandong-gu, Goyang, Gyeonggi, 411–769, Korea; e-mail: [email protected].

© 2007 by The Society of Thoracic Surgeons Published by Elsevier Inc

(Ann Thorac Surg 2007;84:1838 – 46) © 2007 by The Society of Thoracic Surgeons

Material and Methods The study retrospectively reviewed prospectively collected data from consecutive patients undergoing major resection for lung and esophageal cancer at the Center for Lung Cancer, National Cancer Center, Korea, between March 2001 and August 2005. A total of 1,087 consecutive patients underwent major thoracic oncology surgery for lung cancer and esophageal cancer. Of those, 0003-4975/07/$32.00 doi:10.1016/j.athoracsur.2007.06.074

SONG ET AL ICU READMISSION AFTER MAJOR THORACIC SURGERY

1839

Table 1. Preoperative Diagnosis and Type of Surgery Diagnosis Lung cancer (n ⫽ 57)

Esophageal cancer (n ⫽ 37)

Type of Surgery Pneumonectomy Lobectomy/ bilobectomy Bilateral simultaneous multiple wedge resections Ivor Lewis operation Esophagectomy with 3-FLND Transhiatal esophagectomy Palliative surgery

Survivors (n ⫽ 63) (%)

Nonsurvivors (n ⫽ 31) (%)

7 (11.1) 21 (33.3) 5 (7.9)

5 (16.1) 17 (54.8) 2 (6.5)

15 (23.8) 7 (11.1) 2 (3.2) 6 (9.5)

5 (16.1) 0 (0) 1 (3.2) 1 (3.2)

3-FLND ⫽ 3-field lymph node dissection.

94 patients (8.6%) required readmission to the ICU after initial recovery. This study was approved by the Institutional Review Board and informed consent was obtained from the patients and next of kin for use of surgeryrelated clinical and serologic data.

Preoperative General Factors All patients underwent the following preoperative diagnostic procedures: posteroanterior radiography of the chest, chest and abdominal computed tomography scans, and fiberoptic bronchoscopy. Esophageal cancer patients underwent esophagogastroscopy with esophageal ultrasonography, and positron emission tomographycomputed tomographic (PET-CT) scans. Lung cancer patients underwent whole body bone scans. Routine blood tests were performed to determine the white blood cell count, the erythrocyte sedimentation rate, fibrinogen, albumin, and serum hemoglobin levels. Preoperative evaluation for pulmonary function included measuring partial arterial oxygen pressure (PaO2) and partial arterial carbon dioxide pressure (PaCO2). In addition, spirometry forced vital capacity and forced expiratory volume in 1 second (FEV1) were also determined and expressed as a percentage of prediction using standard prediction formulas. The definitions of preoperative risk factors were the following: (1) pulmonary dysfunction ⫽ %FEV1 less than 70% of predicted normal value, vital capacity less than 80% of predicted normal value, or hypoxia (PaO2 less than 80 mmHg); (2) cardiac dysfunction ⫽ history of ischemic heart disease, heart failure, or abnormal electrocardiographic findings; (3) renal dysfunction ⫽ history of renal disease, elevation of serum creatinine over 1.5 mg/dL, or 24 hours creatinine clearance less than 80 mL/minute.

Surgical Factors Major thoracic oncology surgery was defined as the operative procedures for lung (excluding single wedge resection) or esophageal cancers (Table 1). Four thoracic oncology surgeons (HSL, MSK, JML, and JIZ) in our institute performed the surgery. Various intraoperative factors were assessed including the amount of blood loss, perioperative blood transfusion, surgical duration, and pathological stage.

Postoperative Care Unless otherwise indicated, all patients were routinely extubated in the operating room at the completion of surgery. All patients were managed in the ICU for one night and then transferred to the general ward the next morning. We defined this immediate postoperative ICU stay as the first admission in the ICU. The length of stay in the ICU before first discharge from it was just a day. Postoperative pain control was routinely achieved using epidural analgesia or intravenous patient-controlled administration. Emphasis was placed on aggressive pulmonary care, early ambulation, and pain control to minimize postoperative pulmonary complications. Bronchoscopic aspiration of sputum was used liberally to aid expectoration of retained secretions and sputum. Low-dose steroid therapy was initiated if acute respiratory distress syndrome developed [7].

Postoperative Complications The definitions of postoperative complications were the following: (1) postoperative pulmonary complications, see paragraph below; (2) cardiac complication ⫽ heart failure or arrhythmia requiring medication; (3) renal complication ⫽ elevation of serum blood urea nitrogen over 30 mg/dL or serum creatinine over 2.0 mg/dL; (4) hepatic complication ⫽ elevation of bilirubin over 3.0 mg/dL or serum aspartate aminotransferase or serum alanine aminotransferase (AST/ALT) over 150 IU/L.

Definitions of Postoperative Pulmonary Complications Only postoperative pulmonary complications occurring within 30 days of surgery were included. The definitions of these complications have been previously described [8]. (1) ACUTE RESPIRATORY DISTRESS SYNDROME. (i) Acute onset with a PaO2/fraction of inspired oxygen 200 mm Hg or less. (ii) Bilateral infiltrates. The infiltrates may be patchy, diffuse, homogeneous, or asymmetrical, and should be consistent with pulmonary edema or fibrotic changes associated with fibroproliferation. Opacity due to pleural effusions or atelectasis should not be considered. Unilateral infiltrate was included for pneumonectomies. (iii) No evidence of left atrial hypertension. If measured, a pul-

GENERAL THORACIC

Ann Thorac Surg 2007;84:1838 – 46

GENERAL THORACIC

1840

SONG ET AL ICU READMISSION AFTER MAJOR THORACIC SURGERY

monary artery wedge pressure 18 mm Hg or less. (iv) The three previously cited criteria must occur together within a 24-hour interval. Same with acute respiratory distress syndrome (ARDS) except a PaO2/fraction of inspired oxygen 300 mm Hg or less. (2) ACUTE LUNG INJURY.

(3) PULMONARY INFECTION. (i) Pneumonia is diagnosed on the basis of a compatible chest radiograph and purulent sputum with Gram’s stain and sputum culture documenting the presence of microorganisms. (ii) Aspiration pneumonitis is defined as either the presence of bilious secretion or particulate matter in the tracheobronchial tree, or a postoperative chest radiograph with infiltrates not identified on a preoperative radiograph in patients who did not have tracheobronchial airways directly examined after regurgitation. (iii) Empyema and related findings. (4) SPUTUM RETENTION. This is defined as lobar or wholelung atelectasis based on chest radiography results, or sputum retention requiring aspiration bronchoscopy. (5) PROLONGED AIR LEAK. This is one requiring more than one

week of postoperative chest tube drainage.

ICU Transfer The decision to transfer a patient from the general ward to the ICU was made on a patient-by-patient basis by one of the four attending thoracic oncology surgeons and (or) one of two Fellows looking after the patient. The reasons for ICU transfer included signs of inadequate tissue perfusion, significant hemodynamic instability, requirement of invasive monitoring, use of inotropes, frequent nasotracheal suction, noninvasive ventilation, or mechanical ventilation. The same surgeons and Fellows looked after any one patient in both the general ward and the ICU. Patients readmitted to the ICU were identified by reviewing the ICU logbook. The reasons for readmission were determined by reviewing physicians’ progress notes, nurses’ progress notes, and the discharge summary. It is acknowledged that retrospective data collection regarding ICU readmission is subject to the detail and completeness of medical records.

Measurement of Acute Physiological Assessment and Chronic Health Evaluation (APACHE) Score The APACHE III prognostic system was developed in the United States based on data collected from 17,440 ICU admissions at 42 ICUs. The APACHE III prognostic system consists of two components: an APACHE III score, which can provide initial risk stratification for severely ill hospitalized patients within independently defined patient groups; and an APACHE III predictive equation, which uses the APACHE III score and reference data on major disease categories and treatment location immediately prior to ICU admission, to provide risk estimates for hospital mortality for individual ICU

Ann Thorac Surg 2007;84:1838 – 46

patients. A five-point increase in the APACHE III score (range, 0 to 299) is independently associated with a statistically significant increase in the relative risk of hospital death within each of the 78 major medical and surgical disease categories. Each of the 94 patients readmitted to the ICU were scored according to the APACHE III prognostic system as described by Knaus and colleagues [9]. The APACHE III scores were calculated by summing the acute physiological score, age score, and chronic health evaluation score. Acute physiological scores were calculated by summing scores for 17 variables over the first 24 hours of readmission to the ICU.

Statistical Analysis Data were analyzed using SPSS for Windows, version 12.0 (SPSS, Inc, Chicago, IL). Categoric variables were compared using the ␹2 or Fisher exact tests, and continuous variables were compared using the Student t or Mann-Whitney U tests as appropriate. The risk of ICU mortality associated with selected factors was evaluated using stepwise binary logistic regression analysis to estimate odds ratios (OR) and their 95% confidence intervals (CI). Continuous variables were dichotomized using the median values or the 60th percentile in the frequency distribution analysis as the cutoff value. A p value 0.05 or less, according to univariate analysis, was the criterion for submitting variables to the model. Goodness of fit was assessed using the Hosmer-Lemeshow ␹2 test. The relative risk, defined as the ratio of incidence among exposed to that among nonexposed subjects, was used to summarize the strength of the association between risk factors and ICU mortalities. Estimates of survival were obtained using the Kaplan-Meier method. Cox proportional hazards methodology was used to model the probability of survival as a function of time, and to examine differences in survival associated with various patient characteristics. Risk ratios (also referred to as hazards ratios) and 95% CIs are presented to indicate significance in multivariate models. Multivariate modeling was initially executed using forward selection, and then confirmed using backward selection.

Results A total of 94 (mean age, 66.0 ⫾ 7.3 years; range, 41 to 83 years; 85 men and 9 women) patients were readmitted to the ICU, yielding an ICU readmission rate of 8.6 % (6.7% [57 of 850] for lung cancer, 15.6% [37 of 237] for esophageal cancer patients). The preoperative diagnosis was lung cancer in 57 patients (60.6%) and esophageal cancer in 37 patients (39.4%) (Table 1). The readmission rate during the study period remained relatively constant. The most common reason for ICU readmission was pulmonary complication, which affected 73 of 94 (77.7%) patients, and the most common pulmonary complication was acute respiratory distress syndrome, which affected 44 (60.3%) patients. The second-most common reason for ICU readmission was cardiac-related problems (8.5%). The causes of ICU readmission are listed in Table 2. Eighteen patients were

SONG ET AL ICU READMISSION AFTER MAJOR THORACIC SURGERY

Table 2. Causes of Intensive Care Unit (ICU) Transfer Causes of ICU Transfer Pulmonary complications ARDS/ALI Pneumonia Atelectasis with dyspnea Cardiovascular complications Reoperation Cerebrovascular accident Others ALI ⫽ acute lung injury; syndrome.

Number

%

73 44/2 15 12 9 6 1 5

77.7

9.5 6.4 1.1 5.3

ARDS ⫽ acute respiratory distress

re-readmitted to ICU on three occasions, two on four occasions, one on five occasions, and one on six occasions. The ICU re-readmission was mainly associated with ARDS and its rebound phenomenon after steroid treatment. Mechanical ventilation immediately after ICU readmission was required for 64 patients (68.1%), and tracheostomies were performed in 40 patients (42.6%) at a median of three days after mechanical ventilation (mean, 3.4 ⫾ 6.2 days; range, 0 to 13 days). Of the 12 patients with atelectasis due to sputum retention, five required mechanical ventilation while the others were treated using frequent nasotracheal suction or aspiration bronchoscopy. Inotropic support such as epinephrine, norepinephrine, or high-dose dopamine infusion was used in 37 patients (39.4%), with ten receiving a combination of drugs. No patients required intraaortic balloon pump treatment. Renal failure was evident in 42 ICU patients (54%). No patient required ICU admission primarily to treat renal failure. Invasive nutritional support was used in almost all ICU patients. Thirty-two patients (34.0%) were diagnosed with a clinically significant methicillin-resistant Staphylococcus aureus (MRSA) infection at some stage during hospitalization.

Patient Outcomes Thirty-one patients died in the ICU, yielding a 33.0% in-hospital mortality rate and a 2.85% overall operative mortality rate in this period. These deaths comprised 24 lung cancer patients (42.1%) and seven esophageal cancer patients (18.9%). The causes of in-hospital mortality were primary respiratory failure (n ⫽ 22), sepsis from bacterial pneumonia (n ⫽ 4), invasive aspergillosis (n ⫽ 1), anaphylactic shock (n ⫽ 1), pulmonary thromboembolism (n ⫽ 1), ischemic bowel disease with systemic thromboembolism (n ⫽ 1), and acute myocardial infarction (n ⫽ 1). The remaining 63 patients were discharged from the ICU to the ward. The mean stay in the ICU was 5.9 ⫾ 8.3 days for survivors and 20.2 ⫾ 20.8 days for nonsurvivors (p ⫽ 0.001). There was no significant difference in the total hospital stay when comparing nonsurvivors with survivors (39.2 ⫾ 30.5 days for survivors and 33.2 ⫾ 27.7 days for nonsurvivors, p ⫽ 0.355). Total ICU

1841

costs were $9,976 and $18,560 for survivors and nonsurvivors, respectively (p ⬍ 0.0001).

Risk Factors for Hospital Mortality (n ⫽ 94) Twenty-three perioperative variables were examined for possible association with hospital mortality (Table 3). Variables first underwent univariate analysis and if found to be significant then underwent multivariate analysis, adjusting for confounders. Univariate analysis showed that the following seven variables were associated with in-hospital mortality: a diagnosis of lung cancer, a higher APACHE III score, mechanical ventilation, longer duration of mechanical ventilation, renal support, reintubation, and MRSA in sputum. Stepwise forward binary logistic regression analysis identified three perioperative variables independently associated with in-hospital mortality: an APACHE III score 50 or greater, mechanical ventilation during five or more days, and renal support (Table 4). These results were confirmed using stepwise backward binary logistic regression analysis. The ␹2 value for this model was 6.232 for 6 degrees of freedom, indicating that the model was an acceptable fit for the data set. Patients with an APACHE III score higher than 50 on the initial day of ICU readmission had an in-hospital mortality of 53.8% (28 of 52). The first-day APACHE III score accounted for the observed death rates (r2 ⫽ 0.83, p ⬍ 0.0001). Of 20 patients with these three factors, 17 died in ICU. Of the remaining three patients who survived in ICU and were discharged home, two patients died eight months after discharge due to cancer-unrelated causes and one patient remains alive 24 months after discharge.

Risk Factors for Survival (n ⫽ 63) Patient follow-up was complete (median follow-up, 11.5 months; range, 3 to 54.4 months). The overall survival at one and three years was 68.5% and 50.6%, respectively. The overall survival at one and three years in lung cancer was 71.43% and 57.0%, and in esophageal cancer was 65.2% and 42.8%, respectively. There was no significant difference between lung and esophageal cancer (p ⫽ 0.53). Twenty-seven patients (42.9%) died during the follow-up period. Thirteen (48.1%) died of cancerunrelated causes such as respiratory insufficiency (n ⫽ 7), emaciation (n ⫽ 5), or acute myocardial infarction (n ⫽ 1). Fourteen (51.9%) patients died from recurrence of the primary cancer. Kaplan-Meier survival analysis showed that mechanical ventilation (p ⫽ 0.045), having a methicillin-resistant Staphylococcus aureus infection (p ⫽ 0.022), and undergoing a tracheostomy (p ⫽ 0.009) were risk factors for poor survival. Cox proportional hazard regression analysis showed that survivors who had a tracheostomy had a poor prognosis (p ⫽ 0.011; OR, 2.81; CI 1.27 to 6.19) (Fig 1A). Of the 12 late mortalities in 19 patients who underwent tracheostomy, nine patients (75%) died from cancer-unrelated causes (p ⫽ 0.0016). Most mortalities due to respiratory insufficiency occurred within the first year of discharge (Fig 1B).

GENERAL THORACIC

Ann Thorac Surg 2007;84:1838 – 46

GENERAL THORACIC

1842

SONG ET AL ICU READMISSION AFTER MAJOR THORACIC SURGERY

Ann Thorac Surg 2007;84:1838 – 46

Table 3. Characteristics of Patients Readmitted to Intensive Care after Major Thoracic Oncologic Surgery Variables (1) Preoperative general factors (14 variables): 1. Age (years) 2. Male 3. Diagnosis Lung cancer Esophageal cancer 4. Laboratory data 4-1. White blood cell 4-2. Fibrinogen 5. Pulmonary function test 5-1. FVC (L) 5-2. FVC (%) 5-3. FEV1 (L) 5-4. FEV1 (%) 6. Smoking status Nonsmoker Former smoker Recent quitter/current smoker 7. Smoking amount (pack-years) 8. Cardiac dysfunction 9. Renal dysfunction 10. Neochemotherapy (2) Surgical factors (3 variables) 1. Surgical duration (minutes) 2. Blood loss (mL) 3. Perioperative transfusion (units) (3) ICU factors (6 variables) 1. APACHE III score 2. Mechanical ventilation 3. Duration of ventilation (days) 4. Renal support 5. MRSA on sputum culture 6. Reintubation a

Survivors (n ⫽ 63)

Nonsurvivors (n ⫽ 31)

p Valuea

Odds Ratio

95% Confidence Interval

65.2 ⫾ 7.2 59 (93.7)

67.4 ⫾ 7.3 26 (83.9)

0.158 0.150 0.025

0.35 3.12

0.088–1.420 1.174–8.275

33 (52.4) 30 (47.6)

24 (77.4) 7 (22.6)

7.35 ⫾ 2.45 387.2 ⫾ 143.1

7.27 ⫾ 2.73 417.8 ⫾ 186.3

0.894 0.482

3.00 ⫾ 0.63 83.2 ⫾ 16.3 2.14 ⫾ 0.60 84.3 ⫾ 21.9

2.96 ⫾ 0.59 85.4 ⫾ 12.2 2.14 ⫾ 0.42 89.6 ⫾ 14.7

0.720 0.500 0.993 0.177 0.547

8 (12.7) 3 (4.8) 52 (82.5) 38.2 ⫾ 24.2 17 (29.3) 1 (1.7) 5 (7.9)

5 (16.1) 3 (9.7) 23 (74.2) 38.4 ⫾ 27.9 12 (44.4) 0 (0) 5 (16.1)

0.959 0.221 1.000 0.289

1.93

0.749–4.972

2.23

0.594–8.377

289.9 ⫾ 114.7 508.5 ⫾ 439.6 1.0 ⫾ 3.7

257.0 ⫾ 83.0 461.3 ⫾ 330.3 0.6 ⫾ 1.2

0.185 0.599 0.464

45.8 ⫾ 14.9 34 (54.0) 5.2 ⫾ 12.6 19 (30.6) 14 (22.2) 7 (11.1)

70.2 ⫾ 31.4 30 (96.8) 20.8 ⫾ 22.5 23 (74.2) 18 (58.1) 11 (34.4)

⬍0.0001 0.001 ⬍0.0001 ⬍0.0001 0.001 0.007

25.59

3.284–199.371

6.51 4.85 4.40

2.469–17.145 1.915–12.261 1.500–12.910

p ⬍ 0.05 was considered to indicate a significant difference.

APACHE ⫽ acute physiological assessment and chronic health evaluation; FEV1 ⫽ forced expiratory volume in 1 second; capacity; ICU ⫽ intensive care unit; MRSA ⫽ methicillin- resistant Staphylococcus aureus.

Comment Intensive care is a standard component of postoperative treatment for most patients who undergo major thoracic oncology surgery. However, ICU readmission is believed to be associated with higher in-hospital mortality and may predict poor outcomes [1]. The rates of readmission to ICU after cardiac surgery have ranged from 3.7% to 5.5% [3, 10, 11]. A few studies have assessed the outcomes and requirements of the patients who initially recover from major thoracic oncology surgery and then require ICU care [5, 6]. Desiderio and Downey [12] reported a 3.1% ICU admission rate and that such patients had a hospital mortality rate of 29%. That series included general thoracic surgical patients, not just major thoracic oncology surgery patients. The major reason for those ICU admissions was pulmonary complication and the

FVC ⫽ forced vital

ICU duration had a broad range (1 to 68 days; median, 8 days), which is consistent with the findings of the present study. However, that study did not provide any postdischarge data. Hirschler-Schulte and colleagues [5] reported respiratory failure requiring mechanical ventilation in 4.4% of patients recovering from lobectomies and pneumonectomies for bronchial carcinoma. They reported an in-hospital mortality rate of 50% and a oneyear mortality rate of 81% (13 of 16 patients). Of the five patients who died within one year of discharge, all suffered from recurrent malignancy. The present study showed the rate of ICU readmission rate was 8.6% after major thoracic oncology surgery, and these patients had a hospital mortality rate of 33.0%. The rate of ICU readmission seems to be higher because our study includes major surgeries for esophageal cancer as well as

SONG ET AL ICU READMISSION AFTER MAJOR THORACIC SURGERY

1843

Table 4. Multivariate Logistic Regression Analysis of In-hospital Mortality Variables APACHE III score Duration of ventilation (d) Renal support

a

Category

No. of Patients

No. (%) of Nonsurvivors

Odds Ratio

95% Confidence Interval

p Valuea

ⱖ50 ⬍50 ⱖ5 ⬍5 Yes No

52 42 39 55 42 52

28 (53.8) 3 (7.1) 24 (61.5) 7 (12.7) 23 (54.8) 8 (15.4)

12.100 1 7.859 1 3.611 1

2.859–51.206

0.001

2.375–26.006

0.001

1.096–11.895

0.035

p ⬍ 0.05 was considered to indicate a significant difference.

APACHE ⫽ acute physiological assessment and chronic health evaluation.

lung cancer, and our policy for thoracic oncology surgery is to perform the complete mediastinal lymphadenectomy for lung cancer and total lymphadenectomy for esophageal cancer. Above all concerns after surgery, prevention of the ICU readmission is the most important. Meticulous patient selection through preoperative assessment can reduce the development of postoperative complication needing the ICU readmission. The preparation for surgery is not very different among hospitals. Our policy is for preoperative assessment of pulmonary function, requiring abstinence from cigarette-smoking over at least a week, improving nutrition through tube feedings if the patients are emaciated beforehand, and treating pneumonia with intravenous antibiotics for at least a week. If the patient has a positive response to the bronchodilator, we performed the operation after serum concentration of aminophylline reached the therapeutic level. Furthermore, appropriate surgical procedures, such as limited resection and video-assisted thoracic surgery (VATS), should be chosen for high-risk patients. Individualized treatment for thoracic oncology patients should be considered

through scrupulous risk analysis to prevent postoperative events. However, despite prudent patient selection we can meet inevitable postoperative events. Regarding the status of these 94 patients at the time of their initial discharge from the ICU, they were well enough to do vigorous activities and they actually did, although some had risk factors (such as old age, neoadjuvant therapy, idiopathic pulmonary fibrosis, etc) to develop postoperative complications. They had stable vital signs and satisfactory radiologic and laboratory data to transfer to a general ward for vigorous mobilization and pulmonary rehabilitation. It is reported that most of postoperative complications such as acute respiratory failure and cardiac arrhythmia are developed on the postoperative second or third day. We reported that postoperative pulmonary complications after lung cancer surgery were associated with higher preoperative fibrinogen level, longer surgical duration, and male gender [8]. Identifying patients who will require readmission and predicting postoperative complications are other issues. We attempted to discuss how to overcome these unpredictable or unavoidable situations and to focus on the factors that

Fig 1. Overall survival in survivors from readmission to intensive care after initial recovery from major thoracic oncology surgery according to the tracheostomy as an independent prognosis factor. (A) Overall survival at one and three years was 77.0% and 62.8% for no-tracheostomy patients, and was 44.7% and 22.3% for tracheostomy patients, respectively (p ⫽ 0.011). (B) Cumulative mortality from cancer-unrelated causes in the patients who underwent the tracheostomy in the intensive care unit (p ⫽ 0.0016). (CI ⫽ confidence interval; OR ⫽ odds ratio.)

GENERAL THORACIC

Ann Thorac Surg 2007;84:1838 – 46

GENERAL THORACIC

1844

SONG ET AL ICU READMISSION AFTER MAJOR THORACIC SURGERY

affect the outcomes of patients who had been admitted to ICU repeatedly. How can we prevent ICU mortalities? This study revealed some factors that could predict ICU mortality. To improve the APACHE III score at the time of ICU readmission, an earlier transfer to the ICU would be helpful to break a vicious circle and reduce the duration of mechanical ventilation and renal support. Our policies for ICU management have changed in many aspects during the five years of this study. A prospective clinical trial of early low-dose steroid therapy for ARDS was begun in 2004 in our institute. We performed the early orotracheal intubation and full sedation for several days, especially in ARDS patients, with early low-dose steroid therapy as guideline-indicated. Regarding the decision of ICU transfer, we did not hesitate to transfer to ICU the patients who suffered from postoperative complication with hemodynamic compromise; in other words, we spent no time at the general wards for these kinds of patients. Low molecular weight heparin was subcutaneously injected to prevent thromboembolic events. If needed, early percutaneous tracheostomy or early percutaneous gastrostomy was performed. Ultimately, we have done an aggressive and early preemptive management to break the vicious circle. In this study we want to prove the change of ICU strategies and its effect by the analysis between ICU mortality and many variables. Our ICU mortality rate before 2004 was 37.0%, which was significantly reduced to 29.2% after 2004. Recently, Dulu and colleagues [13] reported that the earlier (rather than later) ICU admission of acute lung injury and ARDS after lung resection may lead to a more favorable outcome. Our data also showed that 94 patients were admitted to ICU again on the postoperative median fourth day due to mainly pulmonary complications. The median interval between first ICU discharge and readmission was three days (range, 1 to 29). The postoperative day at the time of ICU transfer in survivors (6.4 ⫾ 8.1 days) was slightly shorter than that (12.3 ⫾ 17.6 days) of nonsurvivors (p ⫽ 0.083). We guess that earlier transfer to the ICU when the patients are deteriorated leads to lower APACHE score at the time of admission in the ICU and improves the outcomes. While tracheostomies may be performed to ensure a safe and patent airway, most are performed to facilitate mechanical ventilation for respiratory failure [14]. One tenth of patients receiving mechanical ventilation undergo tracheostomy [15, 16]. The early use of tracheostomy has been advocated to promote more rapid liberation from mechanical ventilation and lower hospital costs, and this was the aim in the present patient group. Nevertheless, patients who require tracheostomy for respiratory failure are high consumers of resources, have the highest patient costs, and the highest hospital reimbursement. At our ICU, tracheostomy has been performed in ARDS patients with prolonged mechanical ventilation, bilateral vocal cord palsy after esophagectomy, old age with poor bronchial toilet, and so on. Although the overall three-year survival rate was 50.6%, if we consider only those who underwent tracheostomy

Ann Thorac Surg 2007;84:1838 – 46

the overall three-year survival rate was 22.3%. It is suggested that the severity of pulmonary dysfunctions requiring a tracheostomy leads to the poor survival, rather than a tracheostomy procedure itself. Tracheostomy might improve the immediate postoperative outcomes. However, the patient who underwent a tracheostomy suffered from recurrent pulmonary complications and had poor compliance for palliative chemo or radiation therapy when the tumor was recurred, which led to the limitation of long-term survival. Patients with renal failure are at greater risk for complications related to poor nutrition, fluid overload and respiratory dysfunction, metabolic and electrolyte abnormalities, and problems of drug overdose and toxicities, all of which can worsen the conditions in patients who are returned to the ICU. A limitation of this study was that it was based upon data from a single high volume institution and has typical institutional biases regarding patient selection. However, the study sought to analyze a relatively homogenous population by using major thoracic oncology surgery patients rather than general thoracic surgery patients. Survival was a primary endpoint in this study. Other significant patient outcomes, such as comfort, advancement of oral nutrition, improved patient-family communication, and simplified nursing care, which could not be easily abstracted and quantified, should have been considered. In conclusion, ICU readmission after major thoracic oncology surgery was found to be associated with high in-hospital mortality. Identification of patients with a high APACHE score and prolonged ventilation at readmission may help to predict the risk of mortality. Preemptive strategies designed to optimize high-risk patients may improve outcomes. Survivors from ICU readmission after thoracic oncology surgery require meticulous and frequent follow-up due to the risk of deterioration after discharge.

The authors are grateful to Bo Ryong Hwang, RN, and DongSeok Han, RN for their valuable contribution to collect the data in this article.

References 1. Rosenberg AL, Watts W. Patients readmitted to ICUs: a systematic review of risk factors and outcomes. Chest 2000; 118:492–502. 2. Engoren M, Buderer NF, Zacharias A, Habib RH. Variables predicting reintubation after cardiac surgical procedures. Ann Thorac Surg 1999;67:661–5. 3. Chen LM, Martin CM, Keenan SP, Sibbald WJ. Patients readmitted to the intensive care unit during the same hospitalization: clinical features and outcomes. Crit Care Med 1998;26:1834 – 41. 4. Durbin CG, Copel RF. A case-control study of patients readmitted to the intensive care unit. Crit Care Med 1993; 21:1547–53. 5. Hirschler-Schulte CJW, Hylkema BS, Meyer RW. Mechanical ventilation of acute postoperative respiratory failure after surgery for bronchial carcinoma. Thorax 1985;40:387–90.

6. Patel RL, Townsend ER, Fountain SW. Elective pneumonectomy: factors associated with morbidity and operative mortality. Ann Thorac Surg 1992;54:84 – 8. 7. Lee HS, Lee JM, Kim MS, Kim HY, Hwangbo B, Zo JI. Low-dose steroid therapy at an early phase of postoperative acute respiratory distress syndrome. Ann Thorac Surg 2005; 79:405–10. 8. Song SW, Lee HS, Kim MS, et al. Preoperative serum fibrinogen level predicts postoperative pulmonary complications after lung cancer resection. Ann Thorac Surg 2006; 81:1974 – 81. 9. Knaus WA, Wagner DP, Draper EA, et al. The APACHE III prognostic system. Risk prediction of hospital mortality for critically ill hospitalized adults. Chest 1991;100: 1619 –36. 10. Cohn WE, Selke FW, Sirois C, Lisbon A, Johnson RG. Surgical ICU recidivism after cardiac operations. Chest 1999; 116:688 –92.

SONG ET AL ICU READMISSION AFTER MAJOR THORACIC SURGERY

1845

11. Bardell T, Legare JF, Buth KJ, Hirsch GM, Ali IS. ICU readmission after cardiac surgery. Eur J Cardiothorac Surg 2003;23:354 –9. 12. Desiderio D, Downey R. Critical issues in early extubation and hospital discharge in thoracic oncology surgery. J Cardiothorac Vasc Anesth 1998;12(suppl 2):3– 6. 13. Dulu A, Pastores SM, Park B, Riedel E, Rusch V, Halpern NA. Prevalence and mortality of acute lung injury and ARDS after lung resection. Chest 2006;130:73– 8. 14. Heffner JE. Medical indications for tracheostomy. Chest 1989;96:186 –90. 15. Kollef MH, Ahrens TS, Shannon W. Clinical predictors and outcomes for patients requiring tracheostomy in the intensive care unit. Crit Care Med 1999;27:1714 –20. 16. Fischler L, Erhart S, Kleger GR, Frutiger A. Prevalence of tracheostomy in ICU patients. A nation-wide survey in Switzerland. Intensive Care Med 2000;26:1428 –33.

DISCUSSION DR DAVID H. HARPOLE, JR (Durham, NC): We have had two speakers now who both have been supportive of thoracic surgeons maintaining their own patients in the ICU (intensive care unit). I would just like to see a show of hands of people who, once your patients are readmitted to the ICU, send them to the care of intensivists and you are no longer responsible for the care of your patients until they come out of the unit. (A show of hands.) A handful. So I take it, then, reversing that, the vast majority of you, when your patients go back to the ICU, you are still the primary physician taking care of them. A show of hands. (A show of hands.) That’s what I want to see. Thank you. DR DAVID W. JOHNSTONE (Lebanon, NH): Did you specifically see whether your readmission to the ICU and your overall mortality correlate with preop FEV1 (forced expiratory volume in 1 second) or postop predicted FEV1 after lung resections? I wonder whether some of your results here are surrogates for poor pulmonary function. DR SONG: In this study we analyzed risk factors for ICU mortality for the patients who were readmitted to the ICU only. Preoperative pulmonary function was included as preoperative variables. The variables in detail were omitted in this presentation. DR MARK B. ORRINGER (Ann Arbor, MI): This was a very nicely delivered paper and I congratulate you. But in 2007, to be discussing the mortality of readmission to an ICU without discussing the more relevant issue—prevention—is a major deficit of this presentation. Thoracic surgical patients who are properly prepared for surgery in advance should not require postoperative intensive care. So I want to know what your policy is for preoperative assessment of pulmonary function, requiring abstinence from cigarette-smoking, and improving nutrition with tube feedings if the patients are emaciated beforehand? How we prepare our patients for major thoracic surgery prevents this kind of problem, and a 33% mortality for patients who go back to the intensive care unit is not reflective of modern thoracic surgery. I think you must address the issue of preoperative preparation of your patients and not just talk about what happens when they “fall into the hole” of going back to the intensive care unit. DR SONG: Thank you very much for your comments and question. I 100% agree with you. Prevention of postoperative

deterioration is the most important thing. Like you, I really think that the preoperative meticulous assessment and preoperative pulmonary rehabilitation are needed. In our institute we preoperatively always do rehabilitation for the patients who are going to undergo major thoracic oncology surgery. Individualized treatment for thoracic oncology through prudent risk analysis would reduce the readmission to the ICU after surgery. DR THOMAS J. WATSON (Rochester, NY): I very much enjoyed your talk. I am curious about your recommendation for early percutaneous gastrostomy. As most of your complications were pulmonary, I would think one would be reluctant to put feedings into the stomach versus, say, placing a nasoduodenal tube in the postpyloric position. Could you comment on whether you might be getting more pulmonary complications from the G-tube feedings? DR SONG: Thank you. We concluded in this study that tracheostomy is an independent prognostic factor for poor survival. However, I think the conditions at the time of performing a tracheostomy lead to poor survival rather than a tracheostomy procedure itself. Consideration for early tracheostomy in the postoperative course may prevent patients from deteriorating to the point of no return. DR WATSON: No, I’m talking about gastrostomy, not tracheostomy. DR SONG: Regarding gastrostomy, since 2004 some patients, especially in ARDS patients, underwent percutaneous gastrostomy to prevent aspiration during and after weaning from prolonged mechanical ventilation. During the usage of nasogastric tube, we experienced frequent aspiration by gastric contents. Recently, we have a tendency to do the percutaneous bedside gastrostomy to prevent the aspiration pneumonia, especially for ARDS (acute respiratory disease syndrome) patients, if available. DR JOE B. PUTNAM, JR (Nashville, TN): I enjoyed your presentation and your thoughtful evaluation of these patients who are readmitted to the ICU. Acute lung injury is indeed a very morbid and often mortal problem in these patients. I have two questions concerning your consistencies of patient care. Do you and your staff routinely use guidelines for routine care such that every patient who has a pulmonary resection or esophageal

GENERAL THORACIC

Ann Thorac Surg 2007;84:1838 – 46

GENERAL THORACIC

1846

SONG ET AL ICU READMISSION AFTER MAJOR THORACIC SURGERY

resection is managed in as identical a fashion as possible based upon guidelines in your unit? And secondly, are other guidelines used in the intensive care unit? For example, were guidelines used for maintaining and weaning mechanical ventilation, and for glucose control? DR SONG: Thank you very much for your comment and question. We routinely provide preoperative and postoperative guidelines in lung and esophageal cancer, respectively. Also, we have critical pathways for simple lobectomy and VATS (videoassisted thoracic surgery) lobectomy. Regarding ventilator care, I think low pressure with low volume is important. We have a trend to maintain the pressure less than 20 cm H2O if possible. Recently, regarding sugar control, our goal of blood sugar is

Ann Thorac Surg 2007;84:1838 – 46

between 80 and 130 during ICU environment. In the past, blood sugar level, in ICU patients, tried to maintain less than 200 mg/dL. I think that some patient selection algorithms need to be designed for the early management in the ICU. The indications of tracheostomy in the ICU were ARDS patients with prolonged ventilation, bilateral vocal cord palsy after esophagectomy with any sign of aspiration or airway problems, in elderly patients with poor bronchial toilet, and so on. Tracheostomy might improve the immediate postoperative outcomes. However, the patients with tracheostomy had a poor survival due to recurrent pulmonary complications or their inability to receive palliative chemotherapy or radiotherapy. In the ICU a model should be designed for the individualized protocol for these patients.

Requirements for Maintenance of Certification in 2008 Diplomates of the American Board of Thoracic Surgery (ABTS) who plan to participate in the Maintenance of Certification (MOC) process which will begin in 2008 must hold an unrestricted medical license in the locale of their practice and privileges in a hospital accredited by the JCAHO (or other organization recognized by the ABTS). In addition, a valid ABTS certificate is an absolute requirement for entrance into the Maintenance of Certification process. If your certificate has expired, the only pathway for renewal of a certificate is to take and pass the Part I (written) and the Part II (oral) certifying examinations. The names of individuals who have not maintained their certificate will no longer be published in the American Board of Medical Specialties directories. Diplomates’ names will be published upon successful completion of the Maintenance of Certification process. The CME requirements are 30 Category I credits earned during each year prior to application. At least half of these CME hours need to be in the broad area of thoracic surgery. Category II credits are not allowed. Interested individuals should refer to the Booklet of Information for Maintenance of Certification for a complete description of acceptable CME credits. Diplomates in the Maintenance of Certification process will be required to complete all sections of the SESATS

© 2007 by The Society of Thoracic Surgeons Published by Elsevier Inc

self-assessment examination. It is not necessary for Diplomates to purchase SESATS individually because it will be sent to them after their application has been approved. Diplomates may apply for Maintenance of Certification in the year their certificate expires, or if they wish to do so, they may apply up to two years before it expires. However, the new certificate will be dated 10 years from the date of expiration of their original certificate or most recent recertification certificate. In other words, going through the Maintenance of Certification process early does not alter the 10-year validation. Diplomates certified prior to 1976 (the year that time-limited certificates were initiated) are also required to participate in MOC if they wish to maintain valid certificates. The deadline for submission of application for the Maintenance of Certification is May 10 of each year. All ABTS diplomates will receive a letter from the Board outlining their individual timeline and MOC requirements. A brochure outlining the rules and requirements for Maintenance of Certification in thoracic surgery is available upon request from the American Board of Thoracic Surgery, 633 North St. Clair St, Suite 2320, Chicago, IL 60611; telephone (312) 202-5900; fax (312) 202-5960; e-mail: [email protected]. This booklet is also published on the website: www.abts.org.

Ann Thorac Surg 2007;84:1846

•

0003-4975/07/$32.00