| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
Interact CardioVasc Thorac Surg 2008;7:582-585. doi:10.1510/icvts.2008.175208 © 2008 European Association of Cardio-Thoracic Surgery
Intraoperative graft flow measurements during coronary artery bypass surgery predict in-hospital outcomesDepartment of Surgery, Division of Cardiac Surgery, Dalhousie University, Halifax, NS, Canada Received 10 January 2008; received in revised form 3 April 2008; accepted 7 April 2008
*Corresponding author. New Halifax Infirmary, 1796 Summer St Rm 2269, Halifax, Nova Scotia, B3H 3A7, Canada. Tel.: +1-902-473-3808; fax: +1-902-473-4448.
Transit-time flowmetry enables immediate intraoperative assessment of blood flow parameters in coronary artery bypass grafts (CABG). The present study assesses the predictive value of measured graft flows on early and medium-term outcomes. All cardiac surgery patients with measured graft flows were included. The last intraoperative flow measurements recorded using the Medtronic Butterfly Flowmetry system were used for analysis. Patients were separated into two groups: patients with normal flow in all grafts or patients with abnormal flow  1 graft. Any pulsatility index (pulsatility index=min–max flow/mean flow)  5 was determined to be normal flow. The study population included 985 patients. Nineteen percent of patients had abnormal flow in 1 graft. Overall in-hospital mortality was 4.7% and not significant between the two groups. After adjusting for covariates, the in-hospital composite outcome for adverse cardiac events was more prevalent in the abnormal flow group (31% vs. 17%; P<0.0001) with an odds ratio of 1.7 (CI 1.1–2.7). Survivors to discharge had a mean follow-up of 1.8 years. However, abnormal flow was not an independent predictor of the medium-term mortality and readmission to hospital for cardiac reason following discharge. Our findings suggest that abnormal flows measured intraoperatively are independently associated with short-term in-hospital adverse outcome.
Key Words: Coronary artery disease; Cardiac surgery; Outcomes
Coronary artery bypass grafting (CABG) has established itself as an important therapeutic intervention for patients with symptomatic coronary artery disease [1]. While clinical results following CABG are generally undisputed, there are patients in which graft occlusion can occur resulting in the inevitable recurrence of symptoms, the need for readmission to hospital and coronary re-intervention and possibly even death [2, 3]. Estimated graft occlusion at one year for LIMA and saphenous vein, have been as high as 9% and 15–30%, respectively [2–4]. Some investigators have suggested that peri-operative and mid-term outcomes have a direct association with anastomotic quality of CABG [5]. Consequently, determining anastomotic quality appears critical at identifying and revising problematic grafts intraoperatively. A widely accepted technique of graft assessment is the transit time flowmeter (TTFM), which uses intraoperative flowprobes to measure blood flow transit time to calculate volumetric flow and derived parameters such as resistance through the graft. These parameters aid in identifying abnormal flow patterns and high resistance flow through grafts [6]. However, the clinical significance of abnormal flow, as measured with TTFM, remains unclear as there exists little data correlating graft flow parameters with patient outcomes early or late. The purpose of this study was to assess the predictive value of measured graft flows on early and medium-term postoperative outcomes.
2.1. Patients All consecutive patients between July 2002–August 2005 undergoing CABG±valve with graft flow measurements were included. Indications for surgery were based on a weekly peer review process, involving cardiologists, cardiac surgeons and cardiac radiologists. Individual patients were queued for surgery based on objective criteria as previously described [8]. A median sternotomy was performed in all patients. CABG surgery performed with CPB was performed in a standardized fashion using ascending aortic cannulation and 2-stage venous cannulation of the right atrium. Intermittent cold blood cardioplegia was delivered antegrade via the aortic root. Off-pump or beating heart procedures were performed in a standard fashion as previously described [9]. The choice of conduits and/or construction of composite grafts was based on surgeon preferences rather than fixed criteria. Arterial conduits were harvested with minimal trauma (non-skeletonized IMA) and all treated with either a Papaverine® solution or nitroglycerine/calcium channel blocker (Verapamil®) solution prior to use. All postoperative cardiac surgery patients were taken to a dedicated cardiovascular intensive care unit (CVICU). Patients discharged from the CVICU were transferred to an intermediate care or general care ward under the care of the same team. All patients were monitored continuously for a minimum of 24 h. The Medtronic Butterfly Flowmeter system was used to measure intraoperative flows on all grafts, after surgical completion of proximal and distal anastomosis, normothermia, and termination of cardiopulmonary bypass were attained. All measurements were collected before and after administration of protamine sulfate. The data collected consisted of mean flow, diastolic flow (percent of time) and pulsatility index (PI). The data were then analyzed by comparing patients with normal flow patterns in all grafts (Group 1 PI<5) vs. abnormal flow in any one graft (Group 2 PI>I5). Patients who required intra-operative graft revision based on TTFM were analyzed separately.2.5. Data collection and analysis Peri-operative patient variables were collected prospectively by full-time research assistants. All analyses were performed with the Statistical Analysis Systems software package (SAS, Release 8.2, Cary, North Carolina). Descriptive statistics included continuous and discrete variables, which were analyzed accordingly with an unpaired t-test, Wilcoxon rank sum test, 2-test, and Fisher's exact test. Statistical significance was defined as a P-value of <0.05. In-hospital composite outcome was defined as postoperative myocardial infarction (MI), prolonged ventilation (>24 h), low cardiac output syndrome (low cardiac output requiring at least two inotropes and/or intra-aortic balloon pump), postoperative percutaneous coronary intervention, re-operation for graft occlusion, and in-hospital mortality. Freedom from mortality and readmission to hospital for cardiac reason were analyzed by the Kaplan–Meier method. Cox proportional hazards ratio model was generated to identify independent patient risk factors that predict the incidence of adverse cardiovascular event defined as mortality and readmission to hospital for cardiac reason. The variables included in the logistic regression and the Cox model include; PI>5, ejection fraction <40%, peripheral vascular disease/cerebrovascular disease, urgent status, preoperative renal failure, age>80 years, previous CABG, chronic obstructive pulmonary disease, myocardial infarction within 21 days, triple-vessel disease, diabetes mellitus, gender, and combined procedure (CABG+other).
3.1. Patient characteristics The study population included 985 patients out of 2431 patients operated during the same period of time. Patients were identified based on a TTFM dataset while undergoing CABG±valve/other cardiac procedure. Isolated CABG accounted for 87% of the study population while CABG+ valve and CABG+valve/other cardiac procedure accounted for 8% and 5%, respectively. Intra-operative graft revisions were performed in 20 (Table 1). Of all the revisions, 18 were found to have a significant graft problem that could be addressed surgically. The types of problems were classified as anastomotic quality, subclavian stenosis, conduit problems such as lie, dissection, or twist, and no identifiable problem. Patients with revisions were excluded from the remainder of the predictive model analysis.
The median PI for the study group was 2.9 (IQR 2.1–4.4). One hundred and eighty-four patients (19%) had abnormal flow defined as PI>5 in at least one graft and will be defined as the abnormal flow group. Preoperative patient demographics are illustrated in Table 2. Patients in the abnormal flow group were found to have more frequent preoperative renal impairment, triple-vessel disease, preoperative myocardial infarction, lower EF (<40%) and required urgent surgery.
3.2. In-hospital outcomes The overall in-hospital mortality was 4.7% (4% vs. 7%; P=0.19) and the median length of hospitalization was six days (6 days vs. 7 days; P=ns). Unadjusted outcomes suggest that patients in the abnormal flow group had a higher incidence of peri-operative MI (0.7% vs. 1.0%; P=0.02), required more re-operation for graft occlusion (0% vs. 0.5%; P=0.04), and had more frequent low output syndrome postoperatively (10% vs. 18%; P=0.005). A composite adverse event outcome was used for the multivariable analysis as the outcome of interest and consisted of: postoperative MI, prolonged ventilation, low cardiac output syndrome, PCI, re-operation for graft occlusion, and in-hospital mortality. The composite outcome was more prevalent in the abnormal flow group (17% vs. 31%; P<0.0001). After adjustment, abnormal flow was found to be an independent predictor of the in-hospital composite outcome with an odds ratio of 1.8 (1.1–2.7). Additional independent predictors of the composite outcome were urgent surgery, elderly patients, low EF, triple-vessel disease, peripheral and cerebrovascular disease (Table 3).
3.3. Mid-term outcomes Survivors to discharge (n=789) were followed for a mean duration of 1.8 years (IQR 1.2–2.2). Follow-up of patients was 100% complete. All-cause mortality (8% vs. 14%; P=0.16) and incidence of readmission to hospital for cardiac reason (17% vs. 21%; P=0.26) appeared higher in the abnormal flow group but this failed to reach significance. Particular surrogate outcomes of graft patency, such as readmission for PCI/CABG and readmission for acute coronary syndrome, were not significantly different between both groups (1.87 vs. 0.7; P=0.3) and (12.6 vs. 18.1; P=0.08), respectively. Cox proportional hazard ratio analysis was performed to adjust for covariates using the composite outcome of all-cause mortality and readmission to hospital for cardiac reasons. In this analysis, abnormal flow was not found to be predictive of medium-term adverse event with a hazard ratio of 1.2 (0.8–1.7). Independent predictors of the composite mid-term outcome were low EF, renal insufficiency, peripheral and cerebrovascular disease (Table 4). Using Kaplan–Meier analysis, risk adjusted freedom from death or readmission to hospital for cardiac reason was better than 70% at two years follow-up in both groups of patients (Fig. 1).
We have shown that in nearly 1% of CABG±valve surgery patients abnormal flows measured by TTFM prompted surgical graft revision. In these patients, 18/20 had a surgically correctable problem identified. While our overall graft revision rate is lower than reported by D'Ancona et al., we also found that TTFM was capable of accurately finding important graft problems [10]. What is novel about our study is that we have shown that an abnormal PI (defined as >5) in any one graft measured intra-operatively, was independently associated with a significant risk of adverse in-hospital outcome (mortality and morbidity). TTFM has been validated as a reliable method of assessing intraoperative graft flow [11, 12], exactly why and how to use it in the clinical setting has yet to be fully understood. Many investigators have been unable to link abnormal flow to clinical outcomes or failed to adjust for clinical differences [7, 13, 14]. Taken together, there remains significant controversy as to the use of TTFM, particularly the impact of unrevised abnormal flow on patient outcomes which we have addressed in the present study. The present study was not designed to test the ability of TTFM to identify graft problems but, instead, to look at real life data on short and mid-term outcomes from a group of consecutive patients with TTFM intra-operative measurements. This explains why no attempts were made to report invasive or non-invasive graft assessment postoperatively. We selected a PI value of five in the present study for several factors such as clinical relevance, manufacturer's recommendations, measured distribution and previously published data [7, 15]. However, there may be PI values that could be more appropriate for certain coronary territory or particular graft arrangement, which we did not attempt to resolve. Finally, it is possible that our mid-term outcome analysis was underpowered to show a significant outcome at follow-up (<2 years). In summary, we have shown that TTFM measurements intra-operatively resulted in 0.9% graft revision with a majority having a correctable problem. In addition, we have shown that patients with a measured graft PI higher than five, that is not revised surgically, are more likely to experience in-hospital adverse events defined as prolonged ventilation, re-operation, postoperative PCI, peri-operative MI and mortality with an odds ratio of 1.8 (1.1–2.7). This would suggest that TTFM may be used not only to identify grafts with problems but a predictive tool to identify patients at increased risk of adverse events post-operatively.
Related Article
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
Top
Abstract
1. Introduction
