The elderly are one of the fastest growing segments of the American society. Individuals 65 years of age and older represent approximately 15% of the American public and this proportion is projected to increase to 26% by 2025.1 This trend indicates that an increasing number of patients undergoing surgery would belong to the elderly category leading to a projected 31% increase in general surgery workload.1,2
The vast majority of elderly patients who undergo surgery under general endotracheal anaesthesia are able to resume spontaneous respiration as soon as they recover from anaesthesia and are, therefore, extubated either in the operating room or in the ICU. However, extubation failure occurs in some patients who, therefore, require re-intubation. Furthermore, some patients develop severe post-operative complications that necessitate unplanned intubation. Therefore, unplanned intubation may reflect premature extubation or it could be a marker of severe physiologic derangement.
Unplanned post-operative intubation (UPI) may cause increased morbidity and/or mortality.3,4 UPI is a useful marker for quality improvement initiative in the peri-operative setting because it is a clear, definite, unambiguous occurrence.5 In many instances UPI is performed as an emergency which exposes patients to the physical risks of emergency intubation, including laryngeal trauma, oesophageal intubation and pulmonary aspiration.6,7
Although old age is commonly cited as a risk factor for post-operative respiratory failure and need for unplanned re-intubation,8 very few studies have explored the risk factors for UPI in an exclusively elderly surgical population. Therefore, the objectives of the present cross-sectional investigation were to describe the incidence of UPI following general and vascular surgery in the elderly, to identify perioperative variables associated with UPI and to examine the association of 30-day all-cause morbidity and mortality with UPI in the same group of elderly surgical patients.
Ethical approval for this study (IRB registration number: HUM0004487) was provided by the Institutional Review Board of the University of Michigan, Ann Arbor, Michigan, USA (Co-chair Dr John Weg) on 21 March 2006 with automatic annual renewal.
We used data from the Participant Use Data File of the American College of Surgeons (ACS) National Surgical Quality Improvement Program (NSQIP) from 2005 to 2008. This program is a multi-institutional (186 hospitals) reporting system designed to provide risk-adjusted surgical outcome data from various medical centres in the United States. The data are all prospectively collected. The methodology for collecting these data including their accuracy and reproducibility has been detailed in previous publications,9,10 but is briefly reviewed here. NSQIP-trained clinical nurse specialists collect and review the veracity of these data by using a standardised format and following strict definitions. Patients were followed through their hospital course and after discharge for 30 days post-operatively.
For the present investigation, we extracted data on all the elderly (age ≥ 65 years) patients (N = 190 986) in the database. We then excluded all outpatient procedures (N = 47 748) because these are typically associated with very low surgical morbidity or mortality rates.
We also excluded patients who were underweight (BMI ≤18.5 kg m−2), had disseminated malignancy, were comatose for more than 24 h pre-operatively (coded as ‘yes’ if patient is unconscious, or postures to painful stimuli, or is unresponsive to all stimuli prior to surgery) or had active physician do-not-resuscitate orders because these have high baseline risks of death.11 Patients who required mechanical ventilation during the 48 h preceding surgery (coded as ventilator-dependent in the database) were also excluded. Finally, after excluding patients with missing data on hospital length of stay (LOS), operation time, anaesthesia time and work relative value unit (RVU), the final study population of 115 692 constituted the study cohort.
The primary outcome was UPI, defined as a requirement for the placement of an endotracheal tube and mechanical or assisted ventilation because of the onset of respiratory or cardiac failure manifested by severe respiratory distress, hypoxia, hypercarbia or respiratory acidosis within 30 days of the operation. In patients who were intubated for their surgery, unplanned intubation occurs after they have been extubated after surgery. In patients who were not intubated during surgery, intubation at any time after their surgery was considered unplanned. Patients who were intubated because of repeat surgery were included only if they were extubated and then required re-intubation after their re-operation. Factors associated with UPI were explored after stratifying patients into two groups (UPI and no UPI).
Secondary outcome measures included the incidence of common post-operative site-specific complications including superficial and deep wound infections, wound disruptions, graft failures, organ space infections or bleeding requiring transfusion, as well as systemic complications including pneumonia, pulmonary embolism, delayed ventilator wean, myocardial infarction, cardiac arrest, septic shock, post-operative stroke, re-operation, urinary tract infection, sepsis, deep vein thrombosis and post-operative renal insufficiency not requiring dialysis. The incidences of these complications were compared between the two groups and the odds ratios (ORs) for their occurrence were calculated.
Operation complexity was measured using the work RVUs as well total operation and anaesthesia times. Work RVU is a scale (0–95) to quantify the amount of work involved in a specific surgery based on pre-procedural, intra-procedural and post-procedural time, technical skill, physical effort, mental effort and judgment and stress due to potential risk. It is the work portion of the Resource-based Relative Value System adopted by Medicare to quantify the amount of work involved in each medical procedure. A score of 0 is least complex and 95 is most complex.12
Dyspnea was coded as ‘yes’ in the database if the patient described difficult, painful or laboured breathing. Similarly, significant weight loss was codes as ‘yes’ for patients with a greater than 10% decrease in body weight in the 6-month interval immediately preceding surgery as manifested by serial weights in the chart, as reported by the patient or as evidenced by change in clothing size or severe cachexia. Patients who have intentionally lost weight as part of a weight reduction program do not qualify. Other definitions of the NSQIP-defined variables are detailed in previous reports.9,10
Data analysis was carried out with SPSS v.17.0 (SPSS Inc., Chicago, Illinois, USA). Basic descriptive statistics, including means, SDs and percentages were calculated for demographic and anthropometric data. Occurrences of complications and deaths are described as simple proportions. Pearson's χ2 (for categorical) and one-way ANOVA (for continuous) variables were used to examine baseline clinical and perioperative associations with UPI.
Prior to performing multiple logistic regression analyses, we examined the univariate predictors for multico-linearity by first creating a correlation matrix and scanning for highly correlated variables (≥0.7). None of the variables had sufficiently high correlation indicating a low likelihood of significant co-linearity.13 Multivariate analysis was performed using forward stepwise logistic regression to calculate the adjusted ORs for 30-day morbidity and all-cause mortality in the BMI categories with the normal group acting as reference. Variables were entered into the model on the basis of statistical significance and clinical relevance. A model fit was measured with the Hosmer and Lemeshow goodness-of-fit test which measures differences between actual and predicted values of the dependent variable. Good model fit is indicated by a non-significant and small χ2-value, indicating no differences in actual and predicted dependent values.13 All reported P-values were two-sided and a P-value of 0.05 was considered to be significant.
Among the 115 692 elderly patients studied (mean age 74.5 ± 6.4 years), 57 946 (50.1%) were men and the majority (87.9%) of the patients were functionally independent adults. Baseline demographic and clinical characteristics of the patients stratified according to the occurrence of UPI are detailed in Table 1. UPI occurred in 3866 (3.3%) of patients. Patients in whom UPI occurred were more likely to be older, male, functionally dependent, have pre-operative pneumonia, more likely to be current smokers and to consume more than two alcoholic drinks per day. Similarly, patients in the UPI group had a higher prevalence of chronic illnesses like chronic obstructive pulmonary disease (COPD), diabetes, hypertension and renal and cardiovascular diseases than those in the no UPI group.
Table 2 explores the operative characteristics of the study population according to the occurrence of UPI. In this regard, patients in the UPI group were more likely to have undergone emergency surgery, had a higher proportion of contaminated wounds and were more likely to have had general anaesthesia than patients in the no UPI group. Additionally, patients who required UPI were more likely to have undergone more complex (indicated by work RVU) and longer duration surgery and anaesthesia. This group also had a higher intra-operative transfusion rate than the no UPI group.
Analysis of the post-operative mortality and morbidity revealed important differences (Table 3). The overall 30-day mortality among those with UPI was 29.4% compared with 2.2% among patients with no UPI underscoring the importance of UPI. Stated another way, patients in the UPI group had an 18 times increased odds of dying than those in the no UPI group. All the measured post-operative complications were more frequent in the group with UPI. The complications with the highest odds of occurring in patients with UPI compared with the no UPI group were delayed ventilator wean (OR = 63.4), cardiac arrest requiring cardiopulmonary resuscitation (OR = 55.6), septic shock (OR = 36.4) and pneumonia (OR = 26.2). Finally, mean hospital LOS was three times as long in patients with UPI compared with those with no UPI (24.5 vs. 7.2 days; P < 0.001).
Multivariate logistic regression indicated that some of the factors detailed in Table 4 were significantly associated with UPI. Re-operation was associated with the highest odds of UPI. Specifically, when controlling for the other covariates in the model, re-operation in this elderly surgical cohort was associated with four-fold higher odds of UPI (OR = 4.5; 95% confidence interval 4.29–4.86, P < 0.001). Furthermore, pre-operative dependent functional status, presence of ascites, emergency operation and intra-operative blood transfusion were significantly associated with a 1.8 times higher risks of UPI. Smoking and COPD history proved to be associated with only moderately increased risk of UPI in this elderly cohort. Presence of multiple medical diagnoses was not a significant independent predictor of the risk of UPI in our study cohort.
We present for the first time data on the incidence and risk factors as well as complications associated with UPI in a large cohort of elderly vascular and general surgery patients. As the proportion of elderly surgical patients increase, studies such as ours will become increasingly necessary. The main finding of this cross-sectional study is that although the incidence of UPI is low in the elderly, it is associated with high morbidity and mortality including considerably longer hospital LOS.
A large proportion of elderly surgical patients are able to resume and maintain spontaneous ventilation after surgery and once extubated do not require re-intubation. In most cases, UPI indicates the occurrence of a severe complication which may be organ-specific or systemic. Re-intubation is often associated with markedly increased morbidity and total cost of patient care.14,15
Several investigators have documented the incidence and risk factors associated with unplanned re-intubation.5 Many of these studies were on patients in the adult medical, surgical, trauma or burn ICU.3,4,12,15 Currently, there are no published studies on the incidence of UPI in an exclusively elderly surgical population. We found a UPI rate of 3.3% in this large, nationally representative cohort of elderly general and vascular surgery patients. This is similar to the published figures in other adult non-cardiac surgery patients.5,12 These previous investigators either described unplanned intubation as a component of post-operative respiratory failure5 or in adult patients in the trauma ICU.12 The latter report described UPI in trauma ICU patients who were significantly younger (mean age 37 years) than the elderly patients in the present study. Our study examined UPI because this is a more easily defined and measurable indicator. For example, several factors may be contributory to prolonged post-operative ventilation including difficult to measure factors such as local weaning parameters and availability of human and material resources.
Consistent with the findings of other investigators, we found that UPI is associated with a markedly increased risk of post-operative mortality. Because of diminished physiologic reserve along with clustering of co-morbid conditions associated with old age,16 complications such as UPI would be poorly tolerated in the elderly. Owing to the considerably increased risk of death and the protracted hospital LOS associated with UPI, it is very important to identify potentially preventable factors that may lead to UPI in the elderly surgical patient.
In order to effectively prevent UPI in the elderly, pre-disposing factors must be recognised. However, similar to the findings of previous investigators,5 many of the factors identified in our study are fairly intuitive but not necessarily modifiable. For example, factors such as COPD, smoking, ascites, renal failure, emergency surgery and intra-operative blood transfusion are either part of the pathological process or indicative of severity of the underlying disease. Despite these limitations, however, presence of the factors enumerated above in an elderly surgical patient may prove useful for alerting clinicians about the need for close follow-up and early intervention.
The present study confirms the previously documented high prevalence of pre-operative co-morbidities in the elderly8,17,18 and many of these conditions like cardiac disease; COPD, renal disease, hypertension and diabetes were more common in patients with UPI. Not surprisingly, clustering of medical diagnoses (≥3) was associated with a higher likelihood of UPI on univariate analysis. However, despite the apparent importance of pre-existing co-morbidities and the high prevalence of these in the elderly surgical patients, most patients did well post-operatively and did not require UPI. The usefulness of the American Society of Anesthesiologists (ASA) classification (a reflection of the severity of pre-operative co-morbidities) is again confirmed by our finding of high ASA status being a significant risk factor for UPI. Of all the pre-operative clinical features, history of dyspnea, COPD, high ASA status and poor pre-operative functional status were highly significant predictors of UPI, each one having a double-digit percentage difference between patients with UPI and those who did not require post-operative intubation.
The strengths of the present investigation are its large cohort of patients from several institutions and the robustness of the data collection process, as well as the unambiguous, categorical-dependent variable. However, there are some limitations to the study. The cross-sectional nature of the study design did not mandate standardised care across the various institutions. Therefore, we cannot determine the causal relationship between UPI and mortality or risk of complications. Additionally, it was not possible to determine the process leading to UPI (accidental extubation, duration of intubation prior to initial extubation, failed planned extubation or physiologic deterioration) in these patients. However, whatever the final common pathway leading to UPI, it is an important complication that is accompanied by significant morbidity and mortality in the elderly surgical patient. To this end, UPI may be a useful outcome measure for quality improvement initiatives5 and efforts to reduce its incidence are justified.
In conclusion, this large, cross-sectional study of elderly general and vascular surgery patients revealed that the incidence of UPI is low in this group of patients. Despite a high prevalence of chronic medical conditions, most patients did well post-operatively and did not require intubation. However, when UPI did occur, it was associated with high morbidity and mortality. It is, therefore, essential to identify modifiable factors that may help reduce the occurrence of UPI in the elderly.
No financial support or conflict of interest declared by any of the authors on this manuscript.
1 Etzioni DA, Liu JH, Maggard MA, et al
. The aging population and its impact on the surgery workforce. Ann Surg 2003; 238:170–177.
2 Pofahl WE, Pories WJ. Current status and future directions of geriatric general surgery. J Am Geriatr Soc 2003; 51:S351–S354.
3 Epstein SK. Extubation failure: an outcome to be avoided. Crit Care 2004; 8:310–312.
4 Seymour CW, Martinez A, Christie JD, Fuchs BD. The outcome of extubation failure in a community hospital intensive care unit: a cohort study. Crit Care 2004; 8:R322–R327.
5 Snyder CW, Patel RD, Roberson EP, Hawn MT. Unplanned intubation
after surgery: risk factors, prognosis, and medical emergency team effects. Am Surg 2009; 75:834–838.
6 Rashkin MC, Davis T. Acute complications of endotracheal intubation. Chest 1986; 89:165–167.
7 Williamson JA, Webb RK, Szekely S, et al
. The Australian Incident Monitoring Study. Difficult intubation: an analysis of 2000 incident reports. Anaesth Intensive Care 1993; 21:602–607.
8 Ergina PL, Gold SL, Meakins JL. Perioperative care of the elderly
patient. World J Surg 1993; 17:192–198.
9 Daley J, Khuri SF, Henderson W, et al
. Risk adjustment of the postoperative morbidity rate for the comparative assessment of the quality of surgical care: results of the National Veterans Affairs Surgical Risk Study. J Am Coll Surg 1997; 185:328–340.
10 Khuri SF, Henderson WG, Daley J, et al
. The patient safety in surgery study: background, study design, and patient populations. J Am Coll Surg 2007; 204:1089–1102.
11 Shepardson LB, Youngner SJ, Speroff T, Rosenthal GE. Increased risk of death in patients with do-not-resuscitate orders. Med Care 1999; 37:727–737.
12 Daley BJ, Garcia-Peres F, Ross SE. Reintubation as an outcome predictor in trauma patients. Chest 1996; 110:1577–1580.
13 Hair JF, Anderson RE, Tatham RL, Black WC. Multivariate data analysis. 5th edition London: Prentice Hall International; 1998.
14 Rajakaruna C, Rogers CA, Angelini GD, Ascione R. Risk factors for and economic implications of prolonged ventilation after cardiac surgery. J Thorac Cardiovasc Surg 2005; 130:1270–1277.
15 Listello D, Sessler CN. Unplanned extubation: clinical predictors for reintubation. Chest 1994; 105:1496–1503.
16 Vanitallie TB. Frailty in the elderly
: contributions of sarcopenia and visceral protein depletion. Metabolism 2003; 52:22–26.
17 Rosenthal RA, Kavic SM. Assessment and management of the geriatric patient. Crit Care Med 2004; 32:S92–S105.
18 Carson JL, Duff A, Poses RM, et al
. Effect of anemia and cardiovascular disease on surgical mortality and morbidity. Lancet 1996; 348:1055–1060.