AS the new millennium approaches we can reflect with pride on advances in cardiac surgery and anesthesia. Many of us will focus less on the wonder of cardiac surgery and more on the question:“How did things change so fast and when will they stop?” The issues at hand are “whether all patients should be forced onto the fast track (FT)” and “whether FT is safe and effective.” Proving this in an era of evolving technology and economic competition is difficult. The prospective study by Wong et al. 1delineates “risk factors” for delayed extubation, prolonged length of stay (LOS) in the intensive care unit (ICU), and 30-day mortality in 885 patients undergoing coronary bypass graft (CABG), all treated on the FT pathway, and provides additional evidence that FT is “safe and effective”enough. The use of this qualifier may seem irreverent. However, the FT pathway is based on a volatile mix of economics, science, and human interactions. Fortunately, for most patients, FT has contributed to modest goals without appreciable impact on patient safety, facilitating a substantial reduction in hospital LOS over the past decade. 2
Fast Tracking and Original Investigations
The history of the FT pathway is economic in origin. Many credit a single hospital in Southern California with an aggressive strategy to “capture” the growing Health Maintenance Organization market via deeply discounted rates for CABG in the mid-1980s, increasing surgical volume from 250 in 1985, to 1,300 in 1989. Krohn et al. 3reported on 240 patients who underwent surgery between 1984 and 1986. Their approach included early extubation, rapid mobilization, fluid restriction, and steroid administration. The median postoperative LOS was 4 days, in-hospital mortality was 2%, and the 6-month readmission rate was only 2.5%. This report was very influential in the spread of FT. However, Anders 4purported a “dark lining,” given steadily increasing Medicare mortality rates at that center, suggesting adverse effects on quality of care.
By the early 1990s, Health Maintenance Organizations were now the economic paradigm for non-Medicare patients. Simultaneously, the federal government moved to contain escalating Medicare costs. The Medicare Participating Heart Bypass Center Demonstration, conducted between 1991 and 1996 at seven hospitals agreeable to a single global discounted rate for CABG, saved an estimated $50.3 million in 5 yr. The executive summary documents attempts to capture market share, realign financial incentives, and reduce costs and LOS by retooling processes of care. 5LOS and short- and long-term mortality all declined annually, despite increased severity of case mix. In-hospital complications increased marginally.
Our medical system remains expensive ($1,092 billion in 1997, or 13.5% of the gross national product). 6Medicare costs for CABG have increased from $2.8 billion in 1990 to $7.3 billion in 1996, whereas procedures increased only 40% to 180,000. Although spending has declined (only 4.8% increase in 1997), it is projected to double over the next decade. Thus, the pressure to curtail costs will continue indefinitely.
Associations of Clinical Factors with Timing of Extubation
The series by Wong et al. 1is the largest prospective, observational cohort analysis in which all patients undergoing CABG were managed on a FT pathway with the intention to extubate as soon as possible. Their findings are similar to our earlier retrospective series that documented a significant contribution of intraoperative process factors over preoperative risk factors alone in determining time to extubation. 7,8The data of Wong et al. extend these observations to ICU LOS as well.
The frequency of several variables associated with operative mortality is several percentage points lower than the Society of Thoracic Surgeons National Database for the corresponding period. 9Most notable are female gender, emergency surgery, and reoperation. The latter is greater still in many tertiary referral centers in the United States (i.e. , > 20% at the Cleveland Clinic 10) and in Veterans hospitals (10%). 11Mean patient age is several years younger. Age remains a complex risk factor with regard to mortality and resource consumption. 12However, most studies verify that elderly patients emerge more slowly from anesthesia, are more sensitive to sedation, and are less likely to be extubated early. The lower risk profile of this series may influence its generalizability. Case volume must be considered because, in general, efficiency increases with volume. 13Because many U.S. centers are low volume, extrapolation from this “efficient” high-volume center may be problematic.
Logistic regression was used to model associations with delayed extubation, prolonged ICU LOS, and mortality. A battery of variables was considered; many are defined, but some are not, which may account for the surprising (albeit weak) finding relating atrial arrhythmia to delayed extubation. Atrial fibrillation increases resource consumption and morbidity (especially stroke). It has a peak incidence on postoperative day 2. 14In our experience, it does not usually have an independent effect on timing of routine extubation. Total LOS, arguably the most important consideration, is not considered. This is unfortunate given that hospitals use monitored settings differently, but “everyone has to leave sometime.” There may be thresholds in timing of extubation that correlate best with changes in total LOS. 15Hospital-specific factors are likely to be critical. Of the safety outcomes, mortality is certainly critical. However, its low frequency in this study limits statistical modeling. The c-index of 0.66 is less than values reported by large database studies (approximately 0.75). 16However, the factors identified (female gender, emergency surgery, and poor left ventricular function) are consistent with larger reports. 16Female gender has also been associated with longer postoperative ventilation 1,17,18and LOS. 12
Of particular interest is the frequency of reintubation. Table 1
{tabft}FT = fast track; P = prospective; O = observational; RD = randomized; PreFT = before fast track; NA = not available. lists its frequency in several studies. 1,8,17–21Variation is related to the cohort and institution (i.e. , tertiary referral centers, 20valve surgery, 8,19,20,etc. ), the frequency of pulmonary comorbidity, 8or peculiar anesthesia techniques. 22The predictors of reintubation include patients with prolonged initial ventilation, consistent with a perioperative lack of cardiopulmonary reserve. 19,20
Readmission, either back to the ICU or after discharge, is not considered. There are little data in the literature on ICU readmission, and it is likely sensitive to hospital-specific practices. Although postdischarge readmission is difficult to track, reported data vary. One center recently reported a 20.9% readmission rate, 49% to outside hospitals. 23The authors of the latter study suggest this represents “cost shifting” spread out to other hospitals. Of note, the readmission rate was double in patients hospitalized ≥ 7 postoperative days. These rates are cause for concern.
An infrequent but feared complication is intraoperative recall. Dowd et al. recently published data from patients in this cohort. 24Of 617 consecutive cases, recall occurred in only 2 patients (0.3%). This number is lower than those previously reported, suggesting that fast tracking is safe given near-continuous use of volatile agents.
The Evolving Face of Fast Tracking
The use of off-pump CABG via standard median sternotomy (OPCAB) or smaller limited incisions (MIDCAB) is increasing dramatically. Although long-term outcome data are not available, experience suggests shorter time to extubation and LOS. Risk stratification models will likely differ significantly from cardiopulmonary bypass–based models. 25
Reports from several centers have indicated LOS similar to that with MIDCAB using the conventional cardiopulmonary bypass approach. Ott et al. 26reported a 29% discharge rate (of 100 patients) by postoperative day 3, whereas Walji et al. , 27coining the phrase “ultra-fast tracking,” reported a 56% discharge rate (of 258 patients) by postoperative day 4, with 23% discharged by postoperative day 2. Readmission rates for the early cohorts, which included both younger and lower-risk patients than the later groups, were 3.9% and 5.6%, respectively, with no early mortality. These reports are controversial, and there is little reason not to believe that other centers will adopt similar approaches.
The Realities of Fast Tracking
Cardiac surgery remains torn between science and regulatory pressures. The latter will likely remain the driving force behind clinical practices. Fast tracking is here to stay. Early extubation is a pivotal component, one in which anesthesiologists play a major role. Evidence suggests that 50% to 75% of all patients undergoing primary CABG (and most valves) can and should be safely extubated by 10 h postoperatively.
Although Wong et al. provide valuable information on variables associated with the timing of extubation, a respect for the consequences of “failure,” and a dose of common sense based on preoperative risk, intraoperative course, and institution-specific factors must prevail. Even a large study may not paint an accurate picture, and sophisticated model validation strategies do not automatically permit generalization. Logistic regression allows excellent discrimination between predictors with a high frequency in the cohort. Factors of suspected clinical relevance, if not present above a certain frequency, will not be appreciated. Risk factors for respiratory morbidity in other surgical settings (i.e. , morbid obesity, abnormal airway anatomy, renal or hepatic insufficiency, neurologic disease) may be overlooked.
Despite the aforementioned caveats, the evidence suggests that early extubation facilitates a measurable reduction in ICU and total LOS in nearly all hospitals, regardless of organizational structure. Each hospital should measure the safety and efficacy of its FT program to avoid cost shifting or patient harm. FT is safe and effective enough. Regardless of our individual opinions, it seems that most organizations paying for this health care have already made their decisions.
