| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Click here to read this article as a CME activity
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
a Cardiovascular Health Research in Manitoba (CHaRM) Investigator Group, Cardiac Sciences Program, St. Boniface General Hospital/I.H. Asper Clinical Research Institute, Winnipeg, Manitoba, Canada
b Department of Hematology/Medical Oncology, Cancercare Manitoba, Winnipeg, Manitoba, Canada
c Sections of Hematology and Critical Care, University of Manitoba, Winnipeg, Manitoba, Canada
Accepted for publication April 17, 2009.
* Address correspondence to Dr Arora, Cardiac Sciences Program, 369 Tache Ave, CR 3012, St Boniface General Hospital/I.H. Asper Clinical Research Institute, Winnipeg, Manitoba, R2H 2A7, Canada (Email: rakesh.arora{at}mac.com).
Presented at the Forty-fifth Annual Meeting of The Society of Thoracic Surgeons, San Francisco, CA, Jan 26–28, 2009.
| Cardiothoracic Anesthesiology:
The Annals of Thoracic Surgery CME Program is located online at http://cme.ctsnetjournals.org. To take the CME activity related to this article, you must have either an STS member or an individual non-member subscription to the journal.
|
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
Methods: A retrospective, propensity-matched, cohort study of all patients undergoing a cardiac surgical procedure at a single tertiary center was performed. The control cohort (n = 1,467) consisted of patients admitted to the traditional, mixed surgical intensive care unit (SICU) from January 2005 to January 2007. The intervention cohort (n = 1,089) consisted of patients admitted to a newly created "hybrid" cardiac surgery ICU (CICU) from January 2007 to January 2008, which was staffed by 24-hour in-house consultant intensivists and a daytime, fast track cardiac anesthesiologist. The primary outcomes were blood product utilization, requirement for ventilation, and ICU recidivism.
Results: The proportion of patients in the CICU cohort who received transfused red blood cells was decreased compared with the SICU cohort (30.2% versus 42.3%, p < 0.001). Similar reductions in platelets and fresh frozen plasma were also observed. The CICU patients were less likely to arrive to the ICU intubated (43.7% versus 66.5%, p < 0.001). There were no differences in postoperative complications. Overall hospital length of stay was reduced in the CICU cohort by a median of 1 day (6 days [interquartile range, 5 to 8] versus 7 days [5 to 9], p < 0.001). Significant reductions in mortality and ICU recidivism were not observed.
Conclusions: The current Manitoba CICU model of 24-hour intensive care physician/cardiac anesthesiologist staffing in postoperative cardiac surgery care is associated with reduced transfusion of blood components, decreased requirement for mechanical ventilation, and shorter hospital length of stay.
Over the past 2 decades, the risk profile of patients undergoing cardiac surgery has changed. Increasingly, cardiac surgical operations are performed on older patients with recent coronary syndromes, higher New York Heart Association classifications, lower left ventricular ejection fractions and cardiogenic shock preoperatively [1, 2]. These demographic changes within cardiac surgery are associated with increases in mortality, intensive care unit (ICU) length of stay (LOS), and resource utilization [3–8]. Despite the increasing complexity of post–cardiac surgical patients admitted to the ICU, the ideal intensive care organizational framework has yet to be determined.
Several centers have developed early extubation or "fast-track" clinical pathways to minimize the need for intensive care in appropriately selected patients [9–11]. Enders and colleagues [12] examined the safety of managing fast-track patients in a post anesthesia recovery unit, thereby avoiding ICU admission entirely. This study demonstrated that, in elective patients, direct admission to a post anesthesia recovery unit was safe and effective [12]. To address the increasing complexity of patients, other centers have created dedicated units for postoperative cardiac surgical care. Novick and coworkers [13] demonstrated that patients admitted to a specialized cardiac surgery recovery unit had reductions in their composite complication rate after the implementation of an early extubation protocol. Two studies have addressed the efficacy of 24-hour consultant intensive care physician "intensivist" staffing models in a mixed adult ICU [14, 15]. We are unaware of any studies that have examined this model of intensivist coverage in a dedicated cardiac surgery ICU.
In our center, a dedicated cardiac surgery intensive care unit was opened in January 2007 with a new "hybrid" clinical coverage model that incorporated the uninterrupted presence of trained intensivists and an early extubation pathway managed by cardiac anesthesiologists. The hypothesis of our study was that care provided in a hybrid dedicated cardiac surgery ICU staffed by 24-hour in-house intensivists would be associated with beneficial clinical outcomes.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
Description of the Intensive Care Physician Staffing Models
Saint Boniface General Hospital is one of two tertiary care centers in the province of Manitoba and became the sole provider of cardiac surgical procedures for the region (
1.2 million patient referral base) in 2007. Before 2007, patients were admitted to a traditional, mixed surgical intensive care unit (SICU) after cardiac surgery alongside patients from a variety of other surgical specialties. Daytime coverage of the SICU consisted of a consultant intensivist and a team of junior resident house-staff from various disciplines. Overnight coverage was provided by an in-house junior resident with the intensivist providing home-call back-up. After ICU discharge, the patients were transferred to a step-down for intermediate care, before the general cardiac surgical ward (Figure 1A). In the step-down unit, patients could retain invasive hemodynamic monitoring and receive restricted doses of a limited number of inotropic agents.
|
The decision to fast-track a patient postoperatively was made on a case-by-case basis in the operating room after consultation between the surgeon and the cardiac anesthesiologist. A dedicated CICU cardiac anesthesiologist, who was free from all other clinical and administrative duties, would direct care of patients suited for early extubation (fast-track candidates [10]) during daytime coverage (08h00–16h00). The criteria for extubation of fast-track candidates is summarized in Table 1. Patients who were not fast-track candidates (i.e. difficult airway, deep hypothermic arrest, significant pharmacological support, or ongoing bleeding due to coagulopathy), were managed by an intensivist within the same ICU. In addition to this daytime hybrid approach, the dedicated CICU model employed 24-hour, seven days a week in-house intensivist coverage. At 18h00, the daytime intensivist and cardiac anesthesiologist signed over care to an incoming intensivist who remained in-house, until 08h00 the following day. The nighttime intensivist served the same functional role as the daytime intensivist. The following morning, patients were either discharged directly to the ward or signed over to the daytime intensivist for continued CICU care. This model further streamlined patient care by eliminating the step-down unit and allowing direct transfer from the CICU to the ward (Figure 1B).
|
Outcomes Measured
To determine the safety and efficacy of the newly created CICU, requirement for mechanical ventilation, allogeneic blood product utilization, and ICU recidivism were the primary end-points as they have been shown to have negative prognostic impact on patients undergoing cardiac surgery [11, 16–18]. Recidivism was defined as readmission to the ICU during the same hospital stay [19]. Secondary end-points included common postoperative complications, ICU and hospital LOS, and mortality.
Statistical Analysis
Data are presented as mean ± SD or median and corresponding interquartile range where appropriate. Baseline characteristics of patients in both ICU cohorts were compared using Student's T test for continuous variables, or Chi Square testing for categorical variables. ICU mortality and 30-day mortality were both expressed as a proportion and as an odds ratio (OR) with 95% confidence intervals (95% CI). An odds ratio of less than 1 signifies decreased mortality in the CICU model compared to the former SICU model of care. The median ICU and hospital LOS were compared using the Wilcoxon rank-sum test. Categorical outcomes were compared using the Chi square test. All reported p values are 2-tailed, with alpha = 0.05.
As it was not possible to randomly assign the model of ICU care, a propensity analysis was undertaken to account for potential confounding factors and case selection biases. Methodology and the reporting of these data were based on a previous publication, which utilized propensity methods [20]. A propensity score derived from clinical and laboratory data was developed using multivariable logistic regression. These variables included demographic data (age, gender, body surface area, and Acute Physiology and Chronic Health Evaluation [APACHE] II score), and baseline co-morbidities including diabetes, chronic renal insufficiency, chronic obstructive pulmonary disease, and peripheral vascular disease. Cardiac history (CCS angina classification, previous myocardial infarction, left ventricular ejection fraction grade, cardiogenic shock, and preoperative acetylsalicylic acid, beta-blocker, and ACE inhibitor use), and operative data (preoperative hemoglobin, reoperation, surgical status, surgical procedure, intra-aortic balloon pump use, and cardiopulmonary bypass [CPB] and cross clamp times) were also included. Propensity scores were used to match patients in the CICU cohort with like patients in the SICU cohort. A greedy matching procedure selected match pairs initially identical to 5 decimal places of probability [21]. If no match existed at 5 decimal places, matching would occur at 4 decimal places and so on. If no match existed at 1 decimal place then that patient was excluded from the study. The high matching success was designed to reduce the possibility of introducing systematic biases. Statistical analyses were conducted by using SAS version 9.1 (SAS Institute, Cary, NC).
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
|
|
|
|
|
Comment |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
The implementation of an early extubation protocol to expedite care is a cost saving measure employed in a growing number of centers [11, 12]. Our hybrid model, which employs a cardiac anesthesiologist to manage the low-risk patients, facilitated a lower dependence on mechanical ventilation when leaving the operating room. However this ideology cannot be universally implemented in all patients as cardiac surgical procedures now extends to elderly, higher risk patients with increasing frequency [1, 2]. Therefore physicians with formal critical care training, versed in the latest concepts of "maximal supportive technologies" and therapies [23], are integral to a multidisciplinary team required for effective care of these complex patients. Familiarity and adherence with the evidence-based literature on blood conservation strategies is an example of the benefits of this approach.
Within our hybrid model, familiarity and adherence with the evidence-based blood conservation strategies was evident by the significant reduction in transfusion rates. Since 2001, the average proportion of patients transfused red cell units at our institution was approximately 45%. Since the implementation of the CICU, that proportion has dropped to slightly over 30%; which represents a relative risk reduction of nearly 30%. This reduction is not only a cost savings, but may also impact patients' short and long-term outcomes [16, 24].
Although our study was not powered, or designed to detect a difference in 30-day mortality with respect to red cell transfusions, reduced transfusions may partially explain the trends in reductions seen in ICU (1.29% vs. 1.50%) and 30-day (1.71% vs. 2.14%) mortality within the CICU cohort as compared to the SICU cohort. Despite various perioperative strategies to reduce blood transfusions, some patients will continue to require blood products. Having dedicated intensivists, with expertise in transfusion management, may ultimately benefit the patient.
The current recommendation by the Society of Critical Care Medicine for Level I critical care centers is to have 24-hour in-house certified intensivist staffing [25]. Though 24-hour staffing may be associated with improvements in care processes, staffing satisfaction, and reductions in ICU complication rates [15], this issue has not been investigated in the postoperative cardiac surgery ICU setting. It is well appreciated that the first 24 hours of care after cardiac surgery are the most important for the majority of patients [26]. Guru et al. examined the nature of preventable deaths in patients undergoing coronary artery bypass surgery and revealed that 61% of deaths were related to ICU problems, including 35% of deaths due to improper diagnosis of life threatening events [27]. We speculate that the CICU intensivist staffing model could potentially limit these errors by promptly recognizing and managing serious issues that occur "after-hours." Kogan et al. in their study found that respiratory issues (43.4%), atrial fibrillation (13.2%), and renal insufficiency (11.3%) were the primary cause for ICU recidivism in fast-track patients [28]. These insidious issues can be promptly dealt with by experienced in-house intensivist coverage. In addition, in-house intensivist staffing may allow for continual optimization of patients, thus preparing them for ICU discharge the next morning. This is supported by our data, which shows no change in ICU recidivism despite the elimination of the step-down unit.
The LOS in our traditional SICU was similar to those reported by others [1, 10]. After implementation of the new CICU model, the overall hospital LOS was reduced by one full median day. Reductions in ICU LOS were not seen and is likely a partial reflection of our current institutional policy mandating that all postoperative cardiac surgery patients stay the first night in the ICU.
Limitations
A randomized control trial comparing the two ICU care models was not possible. As such, an observational before-and-after study design, as used by others [12, 13, 15], was utilized to address the question. Employing a propensity score analysis allowed for the examination of a large patient cohort (n = 2,556), matched on more than 20 clinical, physiological, and surgical variables. This approach permits valid comparison despite the lack of a priori patient randomization and should serve to both reduce variability and increase overall precision of outcome estimates.
It is possible that changes in overall care occurred over time, which were unrelated to the ICU models. During the study period (2005–2007) between both models of care, there were no major operative, ICU, or anesthetic treatment or policy changes that were implemented. The intra-operative anesthetic management and the extubation guidelines remained similar within both models of care. The findings from the Transfusion Requirements in Critical Care trial formed the basis for our institution's general transfusion practices and did not change throughout the study period [29]. Furthermore, the findings from the BART trial had limited effect on our study [30]. Though all patients received some form of anti-fibrinolytic therapy, only high-risk patients in our center received aprotinin. The majority of patients received either e-aminocaproic acid or tranexamic acid as the anti-fibrinolytic of choice. Up until the withdrawal from the market, late October 2007 (at the end of our study period), the use of aprotinin at our institution remained similar throughout the years.
Although the major change being the staffing model, the routine day-to-day ICU patient management had not changed and was carried out by nursing staff and allied health individuals that were familiar with postoperative cardiac cases. One variable that we were unable to control for was the addition of cardiac surgeons and cardiac anesthesiologists with formal critical care training that routinely staffed the new CICU. The presence of these highly qualified personnel may have generated a culture of more tenacious observance to process of care protocols on a 24-hour/7 days a week basis.
We did not capture reintubation of patients within either care model. This would be useful data to examine given that reintubation has been linked with increased morbidity and mortality [22]. Similar ICU length of stays between the two models indirectly suggests that reintubation was probably not significantly greater in the CICU model.
A 1-day shorter hospital LOS, in the absence of an appreciable change in the duration of ICU LOS may also be due to changes in patient care over time. The use of universal telemetry monitoring on all postoperative patients (a current feature in our institution), rather than limiting monitoring to patients in the step-down unit, may have had an impact and will be the subject of future study.
In conclusion, a dedicated hybrid cardiac surgery intensive care unit, which employs an intensive physician staffing model comprised of highly qualified personnel well-versed in both multisystem and fast-track approaches, is associated with reductions in the need for postoperative mechanical ventilation, blood component transfusion, and hospital length of stay. Further studies are needed to ascertain the cost-effectiveness of this model and to identify which aspects of this new paradigm are responsible for the most significant improvement in clinical outcomes.
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
One of the challenges in cardiovascular critical care is the early diagnosis and management of acute kidney injury. I notice that you did not have an important change in this, and I am wondering, since this is an ongoing challenge, how you are going to address this.
Again, I want to congratulate you. I think this is a very important area that we should be developing within thoracic surgery, and, I might say, thoracic surgeons over the last several years have played a major role in developing this specialty. Thank you.
DR KUMAR: Thank you for your comments. Our group agrees that tailored and focused ICU care postoperatively is essential in the modern era of cardiac surgery. As patient risk profiles change and surgical procedures increase in complexity, having individuals who understand the operation and are able to deliver appropriate ICU care is crucial. Our hybrid CICU model addresses this by having cardiac care specialists formally crossed-trained in critical care medicine delivering 24-hour in-house care.
With respect to your comment on acute kidney injury, it is important to recognize that individuals with renal impairment after surgery have a worse prognosis. With such a low event rate and the limited number of study patients, it is difficult to show a reduction in renal dysfunction after surgery. We are currently performing a post hoc analysis of high-risk patients, including those with renal impairment preoperatively and postoperatively, to see the extent of improvement in outcomes with this model of care. Having certified intensivists delivering 24-hour multisystem care allows for early diagnosis and intervention of insidious postoperative complications, such as pending renal impairment or cardiac dysfunction. In addition, with critical care training, decisions to utilize maximal supportive therapies, such as renal replacement therapy, can be delivered at anytime during the day.
DR VASSYL A. LONCHYNA (Chicago, IL): Doctor Kumar, a very nice presentation and my congratulations to the group at St. Boniface. Also I would like to recognize and greet Dr Barwinsky, who created your surgical unit, your retired former chief.
As a surgeon who now practices critical care, I am very interested in your findings and the way you have rededicated your unit along the various tracks there. You have rightfully stated at the end that one of the weaknesses is the fact that you don't have the data such as the hours to extubation, because that certainly would tell us the impact of your intervention in the intensive care unit, whether the fast track or the more complex track. Granted, also, we know that we have very excellent nurses in the cardiac units that also contribute significantly to the intervention of the protocols that allow for early extubation.
I would like to also just present another option for a hybrid unit, and that would be that of an electronic ICU. In the Chicagoland area, we have at least three major medical hospital systems that have this electronic ICU, which adds another layer of care in overseeing the various problems in ICUs, and this could perhaps add greater safety to taking care of patients in the postoperative period. Congratulations again for a beautiful presentation.
DR KUMAR: Thank you for your comments. With respect to additional early extubation data, that is a deficiency in our current database system, which we are looking into. Although we are unable to give a true denominator of potential fast-track candidates, given that the majority of patients are actually extubated intraoperatively, the proportion fast-tracked is in fact probably higher than recorded.
Regarding the two tracks of ICU care, since our study was inclusive of all patients undergoing surgery, our results can be generalized and not limited to isolated CABG patients as in previous studies. At this point, we are currently separating the two cohorts to further analyze the impact of our hybrid model of care. At our institution, fast-track anesthesia is more than a process, it's a mindset of our anesthesiologists. If a patient is healthy before the operation, receives a technically sound surgery, they should be healthy on recovery, and are treated as such. In addition, because of the hybrid model with the presence of a cardiac anesthesiologist in the ICU, there is a continuity of anesthetic management that flows from the operating room to the ICU.
As you mentioned, having dedicated critical care nurses familiar with post–cardiac surgical patients is essential in developing an effective and efficient ICU model. Nursing background and experience with cardiac patients was similar between the two cohorts; therefore, this was not a significant confounding factor within the study.
With regard to an electronic ICU, as our institution is the only referral center within the district, employing an electronic model would be of less benefit in our system. Additionally, although an electronic ICU is potentially more efficient, we feel the presence of a bedside intensivist 24-hours a day to assess the patient, direct care, and perform any necessary invasive procedures is crucial. Post–cardiac surgical patients are fundamentally different from general ICU patients. Our institution believes that in order to address the needs of all cardiac patients in an efficient manner, on-going optimization with in-house care is advantageous. With our hybrid 24-hour model of care and direct ICU to ward transfers, we were able to demonstrate a significant decrease in the total hospital length of stay by 1 median day—not only advantageous for the patient, but also a cost savings for the health care system.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
