Proposed advantages of port-access cardiac surgery have yet to be substantiated. The authors retrospectively compared patients undergoing port-access cardiac surgery with a matched group undergoing conventional cardiac surgery.
Forty-six patients who underwent port-access cardiac surgery were matched with 46 who underwent conventional cardiac surgery. Absolute criteria for matching included morning-of-surgery admission, procedure undergone, and care being delivered by one of two surgeons. If possible, matching included care delivered by one of two anesthesiologists. Patients were matched as closely as possible for preoperative demographic and clinical characteristics.
All 46 pairs of patients were matched for procedure and admitted the morning of surgery. All 92 operations were performed by one of two surgeons, and 89% were performed by one of two anesthesiologists. Preoperative demographic and clinical characteristics were equivalent between groups. Compared with conventional cardiac surgery, port-access cardiac surgery increased surgical complexity (it almost tripled cardiopulmonary bypass time during coronary artery bypass grafting and increased it almost 40% during mitral valve procedures) and increased total operating room time (P < 0.0001). Port-access cardiac surgery had no beneficial effect on earlier postoperative extubation, decreased incidence of atrial fibrillation, or intensive care unit time, yet it decreased postoperative duration of stay (P = 0.029, all patients), a benefit observed primarily in patients undergoing coronary artery bypass grafting (P = 0.002).
This retrospective analysis revealed that port-access cardiac surgery increases surgical complexity, increases operating room time, has no effect on earlier postoperative extubation or decreased incidence of atrial fibrillation or intensive care unit time, and may facilitate postoperative hospital discharge (primarily in patients undergoing coronary artery bypass grafting). Properly designed prospective investigation is necessary to ascertain whether port-access cardiac surgery truly offers any benefits over conventional cardiac surgery.
MINIMALLY invasive cardiac surgery has gained prominence during the past several years and can be divided into two categories: beating-heart cardiac surgery without cardiopulmonary bypass (CPB) and Port-access cardiac surgery (PACS) with CPB. Port-access technology uses a system of five internal jugular vein and femoral artery or vein catheters and cannulae to provide closed-chest CPB, permitting the heart to be arrested and protected with cardioplegia in a manner similar to that used in conventional cardiac surgery (CCS). 1In place of a median sternotomy, a minithoracotomy is performed to allow surgical exposure. The proposed advantages of PACS include less postoperative pain, decreased hospital stay duration and rehabilitation periods, and reduced healthcare costs. 2These claims have not been substantiated in prospective randomized or retrospective clinical trials. 3–10Originally proposed by Peters in 1993, 11port-access technology subsequently was developed by Heartport (Redwood City, CA). Animal testing at Stanford and New York Universities in 1996 established safety and effectiveness of myocardial protection, feasibility and reproducibility of coronary artery bypass grafting (CABG), and applicability of the approach for mitral valve operations. 12–15After clearance from the Food and Drug Administration in October 1996, several institutions reported early experiences in humans in terms of safety and effectiveness for CABG and mitral valve operations. 2,16–29As of mid 1999, more than 8,000 patients had undergone PACS worldwide (Fischer, May 1999, written communication). This retrospective analysis compared perioperative variables of patients undergoing PACS with a matched group of patients undergoing CCS.
Materials and Methods
This investigation was approved by the Loyola University Medical Center Institutional Review Board for the Protection of Human Subjects. We retrospectively reviewed the charts of all patients undergoing PACS at this institution between July 1997 and March 1999. Of these 116 patients, we selected those who were admitted to the hospital on the morning of surgery and had their care delivered by one of two surgeons (M.B. or B.B.) and one of two anesthesiologists (M.C. or M.N.). Forty-six such patients were identified. Each patient undergoing PACS was matched with one who underwent CCS. Absolute criteria for matching included morning-of-surgery admission, type of procedure, and care being delivered by one of two surgeons (M.B. or B.B.). If possible, matching included care being delivered by one of two anesthesiologists (M.C. or M.N.). Because none of the PACS patients were undergoing reoperation or had renal failure necessitating dialysis, these two characteristics excluded CCS patients from being matched. Patients were matched as closely as possible for age, gender, height, weight, use of the left internal mammary artery, and time period. Case matching was performed by one of two individuals (E.M.F. or K.J.S.) who were blinded to postoperative outcome. Perioperative data then were recorded.
Procedures for PACS at our institution are similar to those detailed elsewhere. 1Briefly, a system of five internal jugular vein and femoral artery or vein catheters and cannulae provides closed-chest hypothermic CPB, permitting the heart to be arrested and protected with cardioplegia in a manner similar to that used in conventional open cardiac surgery. Arterial and venous cannulae are inserted through the femoral artery and vein. Two catheter introducers are inserted into the right internal jugular vein: an 11-French introducer for eventual placement of the coronary sinus catheter and a 9-French introducer for eventual placement of the pulmonary artery vent catheter. The coronary sinus catheter delivers retrograde cardioplegia, and the pulmonary artery vent catheter returns blood not drained by the femoral venous cannula. Finally, the endoaortic balloon functions as an aortic cross-clamp, delivers antegrade cardioplegia, decompresses the aortic root, and monitors aortic root pressure. Catheter and cannula placement are facilitated by transesophageal echocardiography or fluoroscopy. A submammary minithoracotomy is performed either on the left (for CABG) or on the right (for mitral valve surgery). A double-lumen endotracheal tube (or bronchial blocker) is necessary for lung separation.
Procedures for CCS at our institution follow: After induction of anesthesia and median sternotomy, surgical exposure (with or without left internal mammary artery harvesting) is performed. Cannulation of the ascending aorta and venae cavae then are performed in preparation for CPB. Hypothermic CPB with a membrane oxygenator and crystalloid prime are used (as in PACS). Nonpulsatile flows are maintained between 2.4 and 2.8 l · min−1· m−2, isoflurane is used by the perfusionist to maintain perfusion pressure in the range of 50–70 mmHg, and α-stat blood gas management is used (as in PACS).
At our institution, standard anesthesia for cardiac surgery (PACS or CCS) includes intravenous fentanyl (15–25 μg/kg), midazolam (0.15–0.20 mg/kg), and vecuronium or pancuronium with or without propofol (during CPB). If necessary, inhaled isoflurane or intravenous nitroglycerin are used for blood pressure control before initiation of CPB. During CPB, intravenous propofol may or may not given, and the lungs are allowed to deflate. Separation from CPB is facilitated with intravenous inotropic or vasoactive drugs (usually dobutamine or norepinephrine) at the discretion of the anesthesiologist managing the case.
After completion of cardiac surgery, patients are transferred to the intensive care unit (ICU). Postoperative care is standardized and tracheal extubation is accomplished at the earliest clinically appropriate time. Criteria for extubation in our ICU include an appropriate sensorium, normothermia, hemodynamic stability, adequate pulmonary function (partial pressure of oxygen [PO2] > 60 mmHg with inspired oxygen fraction = 0.4), adequate urine output, and minimal chest tube output. If hypertension, tachycardia, or excessive movement develops in a patient and tracheal extubation is not yet appropriate (for any reason), the ICU nurse is free to administer small amounts of intravenous midazolam or morphine. Our ICU and hospital do not follow strict criteria for discharging patients from the ICU and hospital, respectively.
All 46 matches of PACS and CCS patients were compared. Subgroup analysis of patients undergoing CABG and mitral valve procedures (mitral valve repair or replacement) also was performed. Continuous data were analyzed using the paired Student t test. If analyses were repeated using log-transformed variables or two nonparametric tests (Wilcoxon signed rank test and approximate permutation test), the qualitative results were exactly the same. Therefore P values from untransformed variables are presented. Dichotomous data were analyzed with using the McNemar chi-square test. P < 0.05 was considered to be statistically significant, and P values are reported only if significance was found. Results are expressed as the mean ± SD, unless otherwise indicated.
Primary outcome variables included in-room to incision time, CPB time, cross-clamp time, total operating room (OR) time, extubation time, ICU duration of stay, occurrence of postoperative atrial fibrillation, postoperative length (duration) of stay (LOS), and postoperative discharge day.
The Appendixpresents the 46 matched pairs of PACS patients and CCS patients. Of the 46 matches, 23 (50%) underwent CABG, 10 (22%) underwent mitral valve repairs, 9 (20%) underwent mitral valve replacements, 2 (4%) underwent aortic valve replacements, and 2 (4%) underwent atrial septal defect repairs. All 92 patients were admitted on the morning of surgery. No patient had undergone previous cardiac surgery or dialysis for renal failure. Of the 92 cardiac operations, all were performed by either M.B. (53%) or B.B. (47%), and anesthesia was performed by M.C. (54%), M.N. (35%), or a small group of seven other anesthesiologists (11%). Of the CABG patients, all but one PACS patient and all CCS patients underwent left internal mammary artery grafting. Only four matches (9%) were unmatched for gender (two mitral valve repairs and two mitral valve replacements).
Tables 1 and 2present demographic and clinical characteristics and perioperative data, respectively, for all patients. The two groups were equivalent regarding all important preoperative demographic and clinical characteristics. Compared with patients undergoing CCS, those undergoing PACS had significantly increased in-room to incision times, significantly increased CPB times, significantly increased total OR times, and significantly decreased postoperative LOS. Although postoperative discharge day was decreased in PACS patients, the difference between groups was not statistically significant (P = 0.090). There was no difference between groups regarding postoperative use of dobutamine (eight PACS patients, seven CCS patients), norepinephrine (two PACS patients, five CCS patients), nitroglycerin (two PACS patients, zero CCS patients), epinephrine (zero PACS patients, one CCS patient), or use of an intraaortic balloon pump (zero PACS patients, two CCS patients). There was no difference between groups regarding OR return (four PACS patients, zero CCS patients), postoperative reintubation (one PACS patient, two CCS patients), postoperative myocardial infarction (zero PACS patients, zero CCS patients), postoperative stroke (one PACS patient, one CCS patient), or death (zero PACS patients, three CCS patients).
Table 3presents perioperative data for the 23 matched pairs of CABG patients. Compared with CCS CABG patients, PACS CABG patients had significantly increased in-room to incision time, significantly decreased number of saphenous vein grafts, significantly increased CPB time per graft, significantly increased total OR time, significantly decreased postoperative LOS, and significantly decreased postoperative discharge day.
Table 4presents perioperative data for the 19 matched pairs of patients undergoing mitral valve repair or replacement. Compared with patients undergoing mitral valve repair or replacement using CCS procedures, those operated on using PACS procedures had significantly increased in-room to incision time, significantly increased CPB time, significantly increased cross-clamp time, and significantly increased total OR time.
Proposed advantages of PACS include improved cosmesis, less postoperative pain, decreased hospital stay duration and rehabilitation periods, and reduced healthcare costs. 2Preliminary experience indicates that the technique may be associated with acceptable morbidity and mortality rates, 2,21,25decreased postoperative extubation time 16,29and incidence of atrial fibrillation, 2decrease ICU time, 29and earlier postoperative discharge 16,29and recovery. 22These claims are controversial and have not been substantiated by properly designed investigation. 3–10Retrospective analysis revealed that, compared with CCS, PACS increases surgical complexity (increased CPB time per graft during CABG and increased CPB time during mitral valve procedures); increases OR time; has no effect on earlier postoperative extubation, atrial fibrillation, or ICU time; and may facilitate postoperative hospital discharge (primarily in patients undergoing CABG).
Port-access cardiac surgery is associated with unique challenges not associated with CCS. The technique necessitates specialized expertise from perfusionists, anesthesiologists, and surgeons. Perfusionists must be aware of the unique circuitry and potential problems associated with port-access technology. 30Anesthesiologists must be proficient with transesophageal echocardiography to guide proper placement of the coronary sinus catheter, pulmonary artery vent catheter, venous drainage cannula, and endoaortic balloon catheter. 23,28Proper placement of a double-lumen endotracheal tube (or bronchial blocker) with one-lung ventilation also is necessary. Surgeons must operate through small incisions, and the quality of the surgical results may be suboptimal. In our analysis, CPB time per graft in PACS patients was almost three times that in CCS patients during CABG, and CPB and cross-clamp times in PACS patients were almost 40% longer in duration than in CCS patients during mitral valve surgery, indicating increased surgical complexity. Investigations involving small numbers of patients undergoing PACS CABG indicate that early graft patency rate (assessed angiographically within 3 months of operation) is acceptable (similar to CCS) 18,24,25; long-term patency has yet to be assessed. Also, port-access mitral valve replacement may be associated with an increased incidence of reoperation for paravalvular leak. 20Port-access technology is relatively new; long-term outcome and quality of surgical results have yet to be assessed and remain a concern. 3–10It already has been shown that minimally invasive cardiac surgery without CPB (in which surgeons must operate through similar small incisions) is associated with increased risk of early vein graft thrombosis 31and technically suboptimal long-term outcome. 31,32In one striking case report, the wrong coronary artery was bypassed. 33Furthermore, the “completeness” of revascularization (patients undergoing minimally invasive CABG with or without CPB tend to receive fewer total grafts) and its effect on long-term outcome are being debated. 3,4,31,32In our analysis, PACS patients received significantly fewer saphenous vein grafts than did CCS patients (1.0 ± 0.8 vs. 2.8 ± 1.0, respectively;P = 0.0001).
We found that in-room to incision time was almost doubled, and total OR time was increased more than 30% in PACS compared with CCS (irrespective of whether the procedure was CABG or mitral valve surgery). The increased in-room to incision time in PACS patients is probably secondary to the extra time necessary for correct positioning of a double-lumen endotracheal tube (or bronchial blocker), insertion of the two catheter introducers into the right internal jugular vein, and correct positioning of the coronary sinus and pulmonary artery vent catheters. Although these catheters were not used in all patients, most received them (34 coronary sinus catheters and 22 pulmonary artery vent catheters were inserted). The increased total OR time is probably secondary to the increased in-room to incision time, increased CPB time, and increased time necessary for correct positioning of the venous cannula and endoaortic balloon. Other investigators have shown that PACS is reliably associated with increased OR time (increased preparation, CPB time, cross-clamp time, total OR time, and so forth). 20,22,24
Claims that PACS decreases postoperative pain, allows for earlier extubation, decreases postoperative atrial fibrillation, and decreases ICU time are unsubstantiated. Observational reports indicating that PACS may allow for earlier extubation, 16,29decreases postoperative atrial fibrillation, 2and decreases ICU time 29are limited by design, 2involve very small numbers of patients, 16or were published in the form of an abstract. 29We found no difference between patients undergoing PACS and those undergoing CCS regarding extubation time, postoperative atrial fibrillation, and ICU time. We made no attempt to assess postoperative pain retrospectively. Previous investigations involving PACS have yet to reveal any benefits regarding quality of postoperative pain relief. 20,22Postoperative pain associated with a minithoracotomy may be greater than that associated with a median sternotomy. 34
Our analysis revealed that PACS may facilitate postoperative hospital discharge (primarily in CABG patients). If data from all patients are analyzed, postoperative LOS was significantly decreased (P = 0.029) in PACS patients. Although postoperative discharge day was earlier in PACS than in CCS patients (4.8 ± 3.7 vs. 6.1 ± 2.6, respectively), the difference was not statistically significant (P = 0.090). Breakdown of the data between CABG and mitral valve procedures reveals that PACS primarily facilitates early postoperative hospital discharge in patients undergoing CABG. In CABG patients, postoperative LOS (P = 0.002) and postoperative discharge day (P = 0.004) both were decreased significantly in patients undergoing PACS. In patients undergoing mitral valve procedures, postoperative LOS and postoperative discharge day were essentially equivalent. These results indicate that time of hospital discharge may be influenced more by surgical procedure (CABG vs. mitral valve surgery) than by how the surgery was performed (PACS vs. CCS). Previous observational reports indicating that PACS may decrease discharge time 16,29and enhance recovery 22involve small numbers of patients 16,22or were published in the form of an abstract. 29Furthermore, one recent investigation revealed that “ultrafast tracking” (discharge within postoperative days 1 to 4) of the majority (70%) of a wide variety of cardiac surgical patients who receive a median sternotomy is possible. 35
Previous observational reports involving PACS have not substantiated claims of decreased cost. Data presented in abstract form in May 1999 at the second annual meeting and scientific sessions of the International Society for Minimally Invasive Cardiac Surgery revealed that PACS may increase cost. 36Seventy-five patients undergoing conventional mitral valve repair or replacement were compared with 61 patients undergoing similar procedures with port-access technology. 36PACS patients had a slightly shorter postoperative hospital stay but significant increases in OR and perfusion costs. 36Potential sources of increased cost in PACS include increased OR use and cost of the specialized port-access equipment. At our institution, OR costs are $281/30 min, fluoroscopy costs are $74/30 min, and the cost of the port-access equipment is $7,200.
Port-access cardiac surgery is associated with risks not normally associated with CCS, such as aortic dissection, 20,24,25aortic valve trauma, 20coronary sinus trauma, 16right ventricular perforation, 17decreased cerebral perfusion from endoaortic balloon migration, 19endoaortic balloon rupture, 20,27pulmonary embolus, 18,20,26femoral artery or vein trauma, 24and pericardial tamponade. 17,18The technique has been associated with increased risk of postoperative bleeding 16,18and increased mortality rate, 20and difficulty in venting air from the heart after mitral valve procedures may initiate postoperative neurologic dysfunction. 20,24In our analysis, more PACS patients than CCS patients (four vs. zero, respectively) returned to the OR for reexploration secondary to bleeding, but the difference was not statistically significant (P = 0.125). Furthermore, PACS does not avoid the known problems associated with conventional hypothermic CPB (e.g. , neurologic dysfunction, pulmonary dysfunction, renal dysfunction, coagulation disorders). Increased CPB time associated with PACS theoretically increases risk of complications associated with CPB. PACS also exposes the patient and OR personnel to radiation from fluoroscopy. Debate has occurred regarding whether the “learning curve” for the technique is acceptable. 16,20,37
Finally, although preliminary experience indicates that PACS may be associated with morbidity and mortality rates equivalent to those of CCS, 2,21,25in comparing outcomes one must keep in mind that PACS patients are generally “healthier” than average CCS patients and therefore should be expected to fare better. For example, a recent multicenter investigation involving 804 PACS CABG patients revealed that fewer PACS patients had cerebrovascular disease (4 vs. 10%), peripheral vascular disease (3 vs. 13%), diabetes (21 vs. 32%), previous myocardial infarction (32 vs. 49%), and decreased ejection fraction (31 vs. 44%) compared with 1,094 patients with a history of CCS. 29
The limitations of retrospective analysis are well-known, but properly designed prospective investigation comparing PACS and CCS has not been (and may never be) performed. We are left with observational or retrospective analysis to compare the two techniques. The strength of our investigation is the careful matching of each PACS patient with a CCS patient who underwent the same surgical procedure. All 92 patients were admitted to the hospital on the morning of surgery (which somewhat standardized “healthiness”) and were operated on by one of two surgeons during the same time period. Furthermore, 89% of the patients were cared for by one of two anesthesiologists, and the two groups were equivalent regarding important preoperative demographic and clinical characteristics (age, gender, height, weight). We are confident that retrospective matching of patients was appropriate and that perioperative care was relatively “standardized.”
The proposed advantages of PACS are controversial and have yet to be substantiated. 3–10This retrospective analysis revealed that, compared with CCS, PACS increases surgical complexity, increases OR time, has no effect on postoperative extubation or atrial fibrillation or ICU time, and may facilitate postoperative hospital discharge (primarily in patients undergoing CABG). The technique does not avoid the known problems associated with conventional CPB and is associated with risks not normally associated with CCS. Long-term outcome and quality of surgical results have yet to be assessed and remains a concern. 3–10Properly designed and powered prospective, randomized investigation is necessary to ascertain whether PACS truly offers any benefits over CCS.