Early Tracheal Extubation after Coronary Artery Bypass Graft Surgery Reduces Costs and Improves Resource Use: A Prospective, Randomized, Controlled Trial

After we received approval from the Institutional Human Ethics Committee and patient consent, we enrolled 100 persons having elective CABG who were younger than 75 yr and had left ventricular function classified as grades I-IV and stable or unstable angina. Preoperative exclusion criteria included previous CABG or valvular heart surgery, history of allergy to propofol or its constituents, left bundle-branch block, documented myocardial infarction within the previous 3 weeks, active congestive cardiac failure, inotropic therapy within 24 h before the study or insertion of an intraaortic balloon pump, severe hepatic disease (alanine aminotransferase or aspartate aminotransferase > 150 IU/I), renal insufficiency (creatinine concentration > 180 micro gram/ml), severe chronic obstructive pulmonary disease (FEV¹< 0.8), or history of seizure or stroke.

Patients were randomly allocated to either the early extubation (early) or late extubation (late) groups according to a computer-generated randomization code. The participant randomization assignment was concealed in an envelope until anesthesia was induced. The surgeon and the research assistant collecting the data, but not the anesthesiologist providing the clinical care, were blinded to the group assignments. Preoperative sedation consisted of 1-3 mg lorazepam administered sublingually 1.5 h before surgery.

Anesthesia induction consisted of 15 micro gram/kg fentanyl and 0.15 mg/kg pancuronium given intravenously for tracheal intubation. Anesthesia was maintained with 0.5-2% isoflurane (inhalational agent) and oxygen before cardiopulmonary bypass (CPB). No benzodiazepine was used. After initiation of CPB, propofol infusion at 2-6 mg [centered dot] kg sup -1 [centered dot] h sup -1 was begun and maintained until 1-4 h in the cardiovascular intensive care unit (CVICU). Postoperative sedation with propofol was titrated to a Ramsay sedation score of 3-4. [7]Patient shivering was treated with 25-50 mg meperidine given intravenously. Persistent systemic hypertension (systolic blood pressure > 140 mmHg) was treated with nitroglycerin or nitroprusside (or both) infusion titrated to achieve systolic arterial pressure of 90-130 mmHg. Esmolol (20 mg) or propranolol (1 mg) given in intravenous boluses was used to control tachycardia (heart rate > 110 bpm). Indomethacin (50-100 mg) was given in suppository form for pain control when patients arrived in the CVICU. Patients were assessed for tracheal extubation within 1-6 h (see appendix 1). Analgesia was maintained by 1-4 mg/h morphine given intravenously after extubation.

Anesthesia was induced with 50 micro gram/kg fentanyl and 0.15 mg/kg pancuronium given intravenously for tracheal intubation. Midazolam (0.1 mg/kg) was administered intravenously before bypass. Isoflurane (0.2-1.5%) was used during CPB as required. In the CVICU, routine morphine (2-10 mg/h) and midazolam (1-3 mg/h) was titrated to the same Ramsay sedation score. Treatment for shivering, hemodynamic parameters, and cardiac rhythm control was the same as for the early extubation group. On the morning after surgery, sedation was discontinued and 50- to 100-mg indomethacin suppository were given to control pain. Patients were assessed for tracheal extubation at 7:00 AM according to conventional management protocols.

Radial and pulmonary arterial pressure were monitored in all patients. Standard CABG surgery was performed with surgeons blinded to group assignment. Patients had median sternotomy with harvesting of saphenous veins and internal thoracic arteries as conduits. Myocardial protection was achieved with intermittent antegrade cold blood cardioplegia through the aortic root, and systemic temperature was allowed to decrease to 33 degrees Celsius during CPB. Hematocrit concentrations were maintained between 20-25% and CPB flow between 2.0-2.5 l [centered dot] min sup -1 [centered dot] m sup -2. The mean perfusion pressure was kept at 50-60 mmHg with nitroprusside or phenylephrine titration. Patients were actively rewarmed to 38 degrees Celsius before removal of the aortic cross-clamp and separation from CPB.

All costs were calculated in Canadian dollars. The total cost for the services provided for each patient was determined for both the early and late groups. All costs applicable to each of the services were summations of direct variables, direct fixed costs, and overhead (indirect variables and indirect fixed) costs. [8-11]The service unit workload cost was determined by accretion of direct variable and fixed costs used to provide that particular service. The total service cost was the summation of unit workload cost and overhead cost. This cost accounting per patient was adopted for the preoperative, intraoperative, CVICU, and postoperative ward stay periods. Departmental costs were divided into physician fees, operating room, CVICU, pharmacy, physiotherapy, laboratory medicine, radiology (including heart catheterization), respiratory therapy, clinical nutrition, and social services.

Direct variable costs are the costs related to direct patient care and varied with the duration or number of patients. Calculation of direct variable costs included physician fees (cardiologist, cardiovascular surgeon, anesthesiologist, surgical assistant, radiologist, and other consultants, such as the endocrinologist), and supporting services (nursing, pharmacy, clinical biochemistry, perfusion, respiratory therapy, physiotherapy, radiology, clinical nutrition, and social services departments). Professional fees were determined by the actual amount of fees reimbursed to medical staff through the Ontario Health Insurance Plan. Personnel costs for nursing (overtime or part-time costs) were based on the length of time the nurses spent performing a service multiplied by wages and fringe benefits. The supply costs for each supporting service was calculated individually. That is, pharmacy medication costs were detailed as preoperative (antianginal and antihypertensive drugs, lorazepam, and so on); intraoperative (fentanyl, pancuronium, propofol, midazolam, heparin, protamine, nitroglycerin, dopamine, and so forth); CVICU (propofol, morphine, midazolam, vasoactive drugs, and so on); and postoperative surgical ward (diuretic, antihypertensive, analgesic, laxatives, and so forth).

Direct fixed costs are those related to direct patient care and are relatively fixed regardless of the duration or number of patients. Fixed costs included assigned staff (nursing, pharmacy), nurse managers, clerical support, and housekeeping. Direct fixed costs in the operating room were calculated based on the cost-center budget allocation for each category. Fixed costs in the CVICU and the surgical ward were calculated based on the cost-center budget allocation for each category and the number of patient days for that period. A resource consumption cost was determined using these data and was calculated for each patient having CABG.

Overhead costs are those related to the indirect operation of a department or an institution. They were calculated by multiplying the service unit workload cost by a specific adjustment factor for each service department. The adjustment factor was calculated by considering the hospital's total operating budget with the individual departmental budget and was appropriated for unit of service for each department. Our adjustment factor was calculated by the Hospital Management Research Unit at the University of Toronto. The adjustment factors used to calculate overhead costs for different departments were as follows: surgical ward, 0.82; CVICU, 0.45; operating room, 0.30; clinical laboratory, 0.28; diagnostic imaging, 0.25; respiratory therapy, 0.22; clinical nutrition, 0.13; and physiotherapy, 0.30.

To assess CVICU use after early extubation and cardiac surgery, the following were calculated: number of cases and rates of readmission, operating room cancellation, and mortality. Admission data from one of two CVICUs (nine beds) were collected from May 1992 to April 1993 (period 1) when a moderate to large dose of opioid was administered during surgery followed by overnight sedation and late extubation (18-22 h). Similar data were compared between April 1993 to May 1994 (period 2) when the cardiac anesthesia practice changed to lower doses of opioid along with inhalational anesthetic, a short period of propofol sedation, and early extubation (1-6 h).

Data for each patient were collected by a research nurse and a research assistant during the surgical course from hospital admission to discharge. The preoperative, intraoperative, CVICU, postoperative costs, departmental costs, and total cost savings per CABG patient were compared for the early and late groups in all patients. The total service cost was further analyzed as total service cost per "uncomplicated" and "complicated" CABG surgery for both the early and late groups. "Complicated" was defined as those patients who failed to meet the extubation criteria at the postoperative set time period. Postoperative myocardial infarction was defined as (1) creatine kinase-MB concentration more than 50 IU/I and more than 8% of total creatine kinase; and (2) major 12-lead electrocardiographic changes from baseline in two or more leads: (a) new Q waves lasting at least 0.04 s and measuring 1 mm in depth, (b) S-T segment elevation or depression of more than 2 mm lasting 48 h, and (c) a symmetrical T-wave inversion persisting 48 h. [12,13]Cerebrovascular accidents were defined as the sudden onset of focal and/or symptoms of neurological deficit persisting more than 24 h, as documented by a neurologist.

The total service cost per CABG patient in each group was calculated as "discharge criteria" and "actual discharge" costs. The discharge criteria were used to determine the time when the patient was medically suitable to be discharged from either the CVICU or the hospital (appendix 1). The actual discharge time was when the patient was actually discharged from either the CVICU or the hospital. The potential cost savings and actual cost savings per CABG patient for the early extubation group were compared with those of the late extubation group.

Results were expressed as mean +/- SD. Differences between the early and late group cost analysis and demographic data were calculated using two-tailed unpaired Student's t tests. The comparison of resource use between early and late extubation groups were analyzed using the Fisher exact test. Differences were considered significant if P < 0.05.

There were no differences in demographic data between the early and late extubation groups (Table 1). In the cost analysis, which included all complications, early extubation significantly reduced the actual CVICU cost by 53% and the total actual CABG surgery cost by 25% when compared with late extubation (Table 2). Forty-one patients (82%) had their tracheas extubated after operation within the defined 6 h in the early group; in 41 patients (82%), the tracheas were extubated after operation within the defined 22 h in the late group. The postoperative extubation time, CVICU, and hospital length of stay (LOS) were significantly lower in the early group (Table 3).

Total preoperative and intraoperative cost analysis revealed no difference between the two groups (Table 4). In particular, there was no difference in anesthetic costs between the two groups when all the intraoperative medications were compared.

The actual total CVICU cost was significantly less (18%) in the early group than in the late group because of substantial savings in CVICU nursing and supplies (23%) (Table 5). In addition, the CVICU discharge criteria total cost was significantly less (40%) in the early group than in the late group because of further savings achieved in use of CVICU nursing and supplies (51%), respiratory therapy and ventilator maintenance (29%), and physicians fees (17%) (Table 6).

The actual total postoperative surgical ward cost was significantly less (13%) in the early group than in late group, due to significant savings (13%) in use of nursing and supplies in the postoperative surgical ward (Table 7). In addition, the total discharge criteria cost on the surgical ward was substantially less (10.4%) in the early group than in the late group, with significant savings in nutrition service (11%) but with lesser savings in nursing and supplies (9%) (Table 8).

The total cost savings per CABG surgery was $1,699 (9%) in the early group ($17,640 +/- $1,677) when compared with the late group ($19,339 +/- $2,886) (Table 9). Further cost savings achieved in the discharge criteria per CABG surgery was $2,375 (13%) in the early group ($15,390 +/- $1,742) compared with the late group ($17,765 +/- $1,837) (Table 10).

The departmental cost savings per CABG surgery in the early group were significantly less by 23% in the CVICU and by 11% in surgical ward costs, and by 12% in respiratory therapy, when compared with the late group. No significant differences were found in physician fees, laboratory tests, and pharmacy costs between the two groups (Table 9).

The discharge criteria departmental cost savings per CABG surgery in the early group were significantly decreased by 51% in the CVICU, by 9% in surgical ward costs, and by 29% in respiratory therapy when compared with the late group. No significant differences were found in physician fees, laboratory tests, and pharmacy cost between the two groups (Table 10).

Nine patients (18%) in each group did not meet the extubation time criteria. Thus patients in the early group did not show an increase in postoperative complications and costs when compared with those in the late group. Major complications such as postoperative myocardial infarction, cerebrovascular accidents, sepsis, and death occurred primarily in patients in the late group. Table 11shows the costs per complicated CABG surgery for the early and late groups. Including all complicated patients, the total median CABG cost per patient was also significantly lower in early versus late groups ($17,269 vs. $19,327; P <0.02).

(Table 12) compares the resource use for cardiac surgical patients in one of the two CVICUs during the 1-y period before and after the early extubation program was instituted at the Toronto Hospital Cardiac Center. The data demonstrate that despite 10-day closure of the operating rooms for elective surgery by our provincial government for compulsory cost savings in health care spending during that year, 24 more cases (total 833 cases) were still performed compared with those of the previous year using conventional late extubation anesthesia (total 809 cases). Furthermore, the CVICU readmission rate did not increase, and the operating room cancellation rate due to a backlog of patients in the CVICU decreased from 2% to 0.3%.

This prospective, randomized, controlled trial is the first comprehensive cost comparison of early and late tracheal extubation after CABG surgery. Our data show that early tracheal extubation significantly reduces total costs per CABG surgery by 25% compared with late extubation. The savings are due primarily to two factors: (1) nursing and (2) ICU costs. There was no increase in the rate or costs of complications in the early group. Early tracheal extubation reduces CVICU and hospital LOS by accelerating the postoperative recovery phase and promoting earlier patient mobility. It shifts the high CVICU costs to the lower wards because of lower intensity care and cost savings in nursing, supplies, physician fees, laboratory, respiratory therapy, and other supporting services from early extubation after operation. This study also provided evidence for further potential cost savings of early tracheal extubation in using discharge criteria for CABG patients from the CVICU and the hospital. Furthermore, our data show that more cardiac cases can be done despite decreased resources from cost containment by reducing CVICU readmission and operating room cancellation rates.

The number of CABG surgeries continues to increase in our aging population and has an important effect on the health-care system. [1,4]This surgery consumes more health-care resources than any other single treatment. [14]In the United States, costs per CABG surgery vary greatly from $13,779 to $38,765 (in U.S. dollars), [6,10,11]whereas in Canada costs per CABG surgery range from $14,958 to $17,681 (in Canadian dollars). [8,9]The differences in CABG costs are due mainly to institutional accounting methods (e.g., charges versus costs, health maintenance organization or managed care payments, teaching versus nonteaching cardiac centers, inclusion of physician fees, or preoperative catheterization costs) and patient-specific factors (e.g., number of coronary vessels grafted, postoperative complications). [10]Although different aspects of resource use or cost containment in cardiac surgery have been addressed, [6,10,15-17]only limited data are available on the economic effect of early extubation. The preference and morbid complication issues in early and late extubation for CABG surgery have been raised and debated. [18-20]But the question remains, "Does early extubation really save money in cardiac surgical patients?" The comprehensive cost analysis in our study included preoperative catheterization costs, physician fees, and all departmental costs until patients were discharged home from the hospital. The actual costs per CABG surgery in early and late extubation were $17,640 and $19,339, respectively; however, if both the preoperative catheterization cost and physician fees were excluded, the actual costs were $13,044 and $14,629, respectively. In addition, we included discharge criteria costs to show the potential further cost savings in patients having cardiac surgery when postoperative management is maximized. The discharge criteria costs per CABG surgery after early and late extubation were $15,390 and $17,765, respectively; however, if both preoperative catheterization costs and physician fees were excluded, the actual costs would be $10,894 and $13,111, respectively.

In the United States, the mean duration of tracheal intubation in ICUs after CABG surgery ranges from 13.7-25.9 h, the ICU LOS is 1.9 days, and the hospital LOS ranges from 7.1-11.6 days.* In comparison, in Canada the mean ICU LOS varies from 2.3-5.2 days and the hospital LOS is 13.9 days after CABG surgery. [15]Our data showed a significant improvement in patients having early tracheal extubation in postoperative recovery, CVICU, and hospital LOS times when compared with those having late tracheal extubation (Table 3). It is important to realize that there is universal reluctance by physicians and nurses to transfer patients out of the ICU during the middle of the night, because of risks for complications and the inconvenience of shifting bed assignments. Our discharge criteria provide a uniform end point in medical fitness for patients to be discharged from both the ICU and the hospital.

No published study has yet detailed or confirmed that early tracheal extubation reduces the cost of cardiac surgery. Intensive care unit LOS is a commonly used yardstick to measure cost indirectly. However, decreased ICU LOS does not necessary translate into cost savings unless variable costs are proportionately reduced. In a retrospective review, Klineberg and associates [21]claimed that within 5 h of ICU admission, 62.5% of their CABG patients were tracheally extubated and 46% of patients were discharged from ICU within 24 h. Cost saving was implied, but direct comparison between the two groups of patients was not possible. Prakash and colleagues [22]showed that 87% of patients having cardiac procedures could be tracheally extubated within 3 h after operation. On average, the successfully extubated patients stayed for 1.5 days in the ICU, and 4.1% of the patients extubated early required repeated tracheal intubation. However, no cost analysis was done and no comparison control group was included. In a randomized prospective study, Quasha and coworkers [23]failed to show any difference in ICU LOS. However, they suggested that indirect cost savings may be result from early mobilization and reduced need for sedation after early extubation. The financial savings for the early extubation group amounted only to $70.00 per patient. Foster and associates [24]found that the number of arterial blood gas measurements could be reduced by 42% when extubating the tracheas of patients early. Without support by any cost analysis, they claimed a reduction in hospital costs with fewer arterial blood gas measurements and reduced ventilator time. Westaby and coworkers [25]studied 1,000 patients having cardiac procedures, and promoted recovery of these patients in a less-expensive cardiac recovery area (CRA) rather than in a conventional ICU. In their study, more than 50% of the patients were tracheally extubated within 3 h of leaving the operating room. This allowed a greater turnover of CRA beds, thus removing the constraints of ICU bed availability. They implied cost savings with this approach based on reusable recovery bed costs ([pound sterling] 750/day), whereas a conventional ICU bed costs [pound sterling] 1,200/day. Of note in that study is that 85% of the cardiac operations were completed within 2.5 h from incision to closure. Patients with no complications were routinely transferred to the postoperative ward, which had a "medium-term area" to continue critical management of these patients with central venous catheters and vasoactive drug infusions. All patients with complications were transferred from the CRA to the ICU. The reduction in costs was based mainly on decreasing nursing staff costs in the CRA, although it has been acknowledged that British nurse:bed ratios are greater than in many other countries. Chong and colleagues, [26]working in the same institution, also published a prospective study of 198 patients admitted to the CRA compared with a retrospective cohort of 80 patients having cardiac procedures admitted to the ICU. This study confirmed that 53.5% of patients could be tracheally extubated within 2 h in the CRA. However, they failed to show any difference in hospital LOS when retrospectively compared with patients admitted to the ICU. In fact, some patients in these studies were discharged from the hospital to a referral hospital for recovery, but not always for home discharge. Arom and colleagues [27]and Engelman and associates [28]retrospectively analyzed their patients having cardiac surgery and concluded that early tracheal extubation shortened the postoperative LOS, resulting in reduced cost and resource use. However, as in any retrospective review, no cause-and-effect relation can be assumed. Obviously early tracheally extubated patients had fewer complications compared with those who required prolonged intubation. In addition, no cost-accounting analyses were done in their studies.

It is most important to have an accurate and comprehensive cost accounting for early extubation after cardiac surgery, if clinical decisions are based on economic factors. In addition, this will identify specific areas for quality improvement and cost reduction in cardiac surgery. Our data show that there was no difference between the early and late extubation groups in both preoperative and intraoperative costs (Table 4). The costs in the preoperative phase can be further reduced if a same-day admission program is available for these patients. The use of propofol in the early extubation group did not increase the perioperative anesthetic drug costs. As supported by our data, ICU costs rank second only to operating room costs in routine CABG surgery. [8]Early tracheal extubation significantly reduced the CVICU nursing and supplies costs when compared with late extubation (Table 5and Table 6). Similarly, early tracheal extubation also reduced the postoperative surgical ward costs when compared with late tracheal extubation (Table 7and Table 8). The costs savings after early extubation were achieved primarily by reducing the intensity of nursing care and LOS in the high-cost ICU setting and facilitated early mobilization in the ICU and on the ward for earlier hospital discharge, thus reducing the part-time and overtime costs of the nursing staff. It allows cost shifting from the high-cost ICU to lower-cost ward service (Table 9and Table 10).

It is most important to recognize that control of perioperative costs must include not only improved efficiency but also no increase in morbidity. The economic consequences of complications after CABG are far more costly than an uncomplicated recovery. [6]No study in the literature has addressed the costs of complications of early extubation after cardiac surgery. Our morbidity outcome study indicates that early extubation in patients having CABG does not increase the incidence of postoperative cardiorespiratory complications. [29] 

In our cost study, the percentages of patients who could not have tracheal extubation within the defined time parameters after operation were equal (18%) in both early and late groups. As demonstrated, early tracheal extubation did not increase the risks for major complications when compared with late extubation (Table 11). The four patients who had postoperative myocardial infarction also had atrial dysrhythmias. Nonetheless, major postoperative complications were costly. When all complications were included for cost analysis in both groups (Table 2), early tracheal extubation significantly reduced ICU (53%) and total CABG (25%) costs when compared with late extubation.

Other cost savings can be achieved by fewer cardiac surgery cancellations because of a backlog of occupied ICU beds leading to loss in operating room time. In addition, delay in cardiac surgery may lead to deterioration of a patient's medical condition and increase their perioperative cardiac morbidity and mortality risks. Our data indicate that early tracheal extubation improved CVICU use, with significantly fewer cardiac surgery cancellations (0.3% versus 2%) when compared with our earlier practice.

One limitation of our study is that the anesthesiologists were not blinded to the group allocation. Nevertheless, the surgeons and the research assistant were blinded to the group allocation. All end-point measurements were objective. The ICU and ward management personnel were blinded and independent of the investigators.

A second limitation is that our results do not apply to patients older than 75 y or those who have preoperative risks as defined by our exclusion criteria. Finally, this costs study is not designed to determine differences in the incidence of complications such as postoperative myocardial infarction, cerebrovascular accident, and death.

Early tracheal extubation after CABG surgery has cost benefits and improves resource use when compared with late tracheal extubation. Early tracheal extubation 1-6 h after surgery reduces total costs per CABG surgery without increasing the rate or costs of complications in patients younger than 75 y. The predominant savings are from reductions in nursing and ICU costs.

The authors thank Hildo Bolley, MSc, of the Department of Health Administration, University of Toronto, for his advice and assistance in calculating the departmental cost adjustment factors; Peter Lewycky, PhD, of the Center of Cardiovascular Research, for his advice in study design and statistical analysis; Linda Jussaume, RN, and Doreen Craig, RN (managers in the CVICU); Ann Tattersall, RN, and Marion Ryujin, RN (managers in the surgical ward); Jerry Hanimyan, of the respiratory therapy department; and Terry Hawn of the physiotherapy department, for their enthusiastic cooperation in this study; David Young, DBA, of the Health Care Management Program, Boston University, and Dr. Lawrence Rotstein, associate professor of surgery, whose constructive review of the manuscript was invaluable; and Lindy Stringer for assistance in preparing the manuscript.

Patient responsive and cooperative

Negative inspiratory force > -20

Vital capacity > 10 ml/kg

Pa^O2> 80 on FiO²less or equal to 0.5

CI > 2.01/min/m²

Temp > 36.5 degrees Celsius

pH > 7.3

Chest tube drainage < 100 ml/h for 2 h

Absence of uncontrolled dysrhythmia

Patient alert and cooperative

No inotropic support and no significant dysrhythmia

Maintain adequate ventilation Pa^O2> 80 mmHg, Pa^CO2< 60 mmHg, SpO²> 90%, while breathing 50% oxygen via face mask)

Chest tube drainage < 50 ml/h for 2 h

Urine output > 0.5 ml [centered dot] kg sup -1 [centered dot] h sup -1

No active seizure

Hemodynamically stable

Stable cardiac rhythm

Noninfected incisions and afebrile

Patient able to void and has bowel movement

Patient is independent in ambulating and feeding

*(Baxter Healthcare, unpublished data, 1995).