The Costs of VAM Complications
A surgeon's perspective on the Achilles' heel of our endovascular therapies.
What I (DEA) remember most about my first endovascular procedure 15 years ago–an iliac artery percutaneous transluminal angioplasty with the original Palmaz stent (Cordis Corporation, a Johnson & Johnson company, Miami, FL)–was the "minor nonsurgical" vascular access management (VAM) complication that occurred and the response from my interventional cardiology colleagues, "Don't worry about it, it goes with the territory, and it's still better than surgery." The patient stayed in the ICU for 2 days and in the hospital for 1 week recovering from a "nonsurgical" retroperitoneal hematoma, a four-unit packed cell transfusion, and a groin hematoma the size of a softball. The patient could not walk for a month. Since that time, I have been both intrigued and appalled by this "goes with the territory" attitude of VAM that oftentimes unfortunately prevails even today.
In the April 2003 issue of Endovascular Today, "Achieving Closure," we published our experience with VAM entitled "Manual Compression May Not Be Benign!"1 After 36 months, it is appropriate to update this work and analyze some very important recent data regarding not only the clinical but economic costs of VAM complications. Additionally, there is the mounting evidence of ergonomic, or physical costs, to our endovascular and catheterization lab staff relative to VAM. It is time to consider many of these work-related injuries a VAM complication. These VAM concerns must be underscored when considering that 60% to 70% of all diagnostics and interventional cases performed today continue to utilize traditional manual compression as their primary mode of VAM.
MANUAL COMPRESSION: THE GOLD STANDARD?
Since Seldinger's initial description in 1953, traditional manual compression has remained, by default, the gold standard of care for VAM with little modification of the "20 to 30 minutes hand-held pressure after catheter removal followed by overnight bed rest" technique, as described in the original publication.2 It remains remarkable that our gold standard has remained unchanged when considering the many other radical changes and improvements that have occurred in endovascular care. Besides the explosion of vascular closure devices (VCDs) on the market, other clinical factors that should challenge manual compression (or at least improve manual compression) as the standard include the older age of our patients today and the greater amount of complex cases, many of which require clopidogrel and significantly more anticoagulation than 50 years ago. Interventional cases today are also longer and often require larger sheaths, and patients have a higher incidence of peripheral vascular disease (PVD) requiring percutaneous peripheral interventions (PPIs). In general, patients are at higher risks for VAM bleeding complications than in the past. Just the simple fact that most of our patients are on clopidogrel would indicate that the same traditional manual compression technique described decades ago may be inadequate now. It seems intuitive that after more than 50 years of significant technological advancements in all other medical fields, we would have a different and better gold standard for VAM.
The Clinical Costs
VAM complications after manual compression remain a significant source of clinical and therefore economic costs (Figure 1).3,4 VCD trials have rarely demonstrated less complication versus manual compression, and VCD failure modes have created a new disease of complications primarily involving infection and thrombosis (Figures 2 and 3). These catastrophic complications have become a significant source of litigation in many practices and hospital facilities. VAM complications with manual compression have been reported between 0.4% to 27% after percutaneous coronary interventions (PCIs), and even higher after PPI.3-8 During the last 36 months, there have been attempts at standardization for reporting VAM complications, and there may be an overall slight trend toward decreased complications, but several very recent reports continue to underscore the significance of VAM complications in PCI and especially in PPI.8-11 Mishra et al recently reported a large PCI experience from the Washington Hospital Center. In more than 1,000 cases with VAM provided by manual compression utilizing either heparin or bivalirudin, the vascular complications, major bleeding, >15% hematocrit decrease, transfusion, and >4 cm hematoma rates were reported as 6.4%, 4.2%, 3.2%, 8%, and 7.8% for a significant (>15%) overall complication rate.12 Fewer complications were reported in the bivalirudin arm.12
Shammas et al recently reported a 16% VAM complication rate in 131 PPI patients but likely underreported "minor" events because they reported the following event rates: death (0.8%), limb loss (1.5%), major bleeding (4.6%), urgent revascularization (7.6%), and vascular complications (1.5%).8 A review of the literature shows that in-hospital complications during PPI range from 3.5% to 32.7%.8-11 Some insight into manual compression complications during PPI can be obtained from analyzing the Angiolink (Medtronic Vascular, Santa Rosa, CA) randomized multicenter pivotal trial and several trial subanalyses reported at the TCT conference in 2005.13,14 The overall manual compression complication rate in patients with PVD was 11.8% (1.7% major in-house, 3.5% major at 30 days, and 6.6% minor).14 In the highest-risk PPI subset during the Angiolink trial, the overall manual compression complication rate was reported at 22.9% (8.6% major and 14.4% minor). These multiple reports implicate manual compression as being inadequate and certainly not benign in VAM during PCI and especially during PPI.
The Ischemic Costs After PCI May Cost a Life
Major and minor bleeding complications, as well as procedure-related blood transfusions, have been recently associated with a significant increase in adverse events, including ischemic events, as well as in-hospital and 1-year mortality after PCI.15-17 A retrospective analysis of more than 10,000 PCI patients found those who had VAM complications had a higher rate of in-hospital ischemic events (including death, myocardial infarction, and coronary artery bypass graft placement), and a multivariate analysis demonstrated that vascular complications were a significant predictor of 1-year mortality.16 Kinnaird et al reported that PCI patients who experienced VAM and vascular complications had a 1-year mortality rate of 7.5% compared to 1.1% in patients without complications.16 Adverse 1-year outcomes in PCI patients with VAM complications were also reported in the large multicenter REPLACE-2 trial and by the National Heart, Lung, and Blood Institute (NHLBI) dynamic registry.18,19 The NHLBI database of more than 6,500 PCI patients who had vascular access site hematomas showed that PCI patients requiring transfusions had a 30-day mortality rate of 9% versus a 1.5% mortality rate without transfusion.19 The death rates at 1 year for transfused versus nontransfused patients were 13% versus 5%.19
Clearly, even if the overall incidences of VAM complications during PCI are decreasing, the clinical and likely the economic costs to the patient and our health care society are significant when complications do occur. We contend that there is no such thing as a "minor complication" because it could be theorized that even a small hematoma or ecchymosis could elicit a significant local perivascular and even systemic inflammatory response responsible for increased ischemic events after PCI. Who is to say that inflammation and even the transfusion of PC after PCI or PPI is not responsible for other adverse long-term stent outcomes, including decreased mean luminal diameters, as compared to stent cases with no complications? Further complicating matters, it is likely that most "minor complications" are still not even recorded, much less reported.
The Economic Costs
Extrapolating economic costs to clinical complications has been challenging, but there is mounting evidence implicating the high financial costs of VAM complications, especially in PCI. Lauer et al reported that a single unit transfusion during PCI increased the overall costs of the hospitalization by $8,000.20 Independent predictors of hospital costs in the REPLACE-2 trial included both major and minor hemorrhage with those increased costs being $6,300 and $400, respectively. The length of hospital stay was found to be three times longer in REPLACE-2 patients with clinical hemorrhage (7.1 vs 23 days, P <.0001).18,21 Arora et al recently reported an evaluation of the local institutional costs associated with VAM from 5,844 PCI patients at Brigham and Women's Hospital.22 They reported an overall complication rate of 13.7% in the manual compression cohort, with the incidence of groin hematoma, bleeding, pseudoaneurysm, arteriovenous fistula, limb ischemia, and infection being 5.7%, 5.1%, 1.7%, 0.45%, 0.34%, and 0.29%, respectively. The cost per episode of the minor complications of hematoma, groin bleeding, and pseudoaneurysm were $2,974, $1,030, and $4,880, respectively. The total costs of groin bleeding and hematoma alone per 1,000 PCIs would calculate to approximately $210,000. The cost per episode of the major complications of limb ischemia, arteriovenous fistula, and infection were $14,772, $8,162, and $5,237. After a matched case control methodology, they concluded that the cost for vascular complications in PCI patients using manual compression was $388 per patient.22
Carpal tunnel syndrome (CTS) and other musculoskeletal injuries are becoming a significant source of concern regarding the issue of "injury and safety in the workplace." The US Department of Labor and the Occupational Safety and Health Administration (OSHA) has compiled shocking data on the subject of work-related musculoskeletal injuries. The US Department of Labor states that work-related injuries are "the nation's most common and costly occupational health problem affecting hundreds of thousands of American workers and costing more than $20 billion a year in workers' compensation."23,24
Wrist, elbow, shoulder injuries, and CTS are the source of more than two-thirds of these injuries. One of the most common repetitive tasks in the interventional lab that has been linked as a causative factor with CTS is our gold standard, manual compression for VAM. OSHA reports that CTS accounts for 50% of all work-related injuries and that more than 8 million Americans currently have CTS. There were approximately 850,000 work-related CTS cases reported in 1999 alone, and this number is likely increasing yearly.25,26 In a survey conducted by the Society for Cardiovascular Angiography, the likelihood of experiencing a musculoskeletal injury was directly related to the number of years active in an interventional lab.25 It is estimated that approximately 230,000 surgical procedures are performed annually for CTS and only about a quarter (23%) will return to work after surgery.25-27 Clearly, CTS and all work-related injuries are significant clinical and economic problems in our health care system today, and it is time they are considered VAM complications. This concern has recently been recognized by the US Department of Health and Human Services, which created several task forces and agencies to address ergonomic and work-related injuries. The National Institute of Occupational Safety and Health has created the National Occupational Research Agenda specifically to identify tasks in the workplace that are at high risk for causing repetitive stress injuries. It need not look much further than in our own labs at our gold standard for VAM.
Is What's Under the Skin Benign?
A perivascular scar score (PSS) was created during our original work in 2003 to objectively access the common femoral artery (CFA) after manual compression or VCD deployment during exposure for AAA endovascular aneurysm repair (EVAR).1 The PSS for manual compression was worse than for VCD use (both PSS >2.5 out of 3) during our initial study, and significant CFA scarring was consistently noted 2 to 3 weeks after what was considered successful VAM, raising questions about the healing of the CFA and timing, safety, and appropriateness of CFA reaccess. To date, no guidelines exist recommending the optimal time that safe CFA reaccess can and should be performed after manual compression or use of a VCD. These reaccess issues become especially important in PVD patients who frequently require multiple PPIs. Even more concerning is CFA reaccess after VCD use during which a wide variety of endoluminal and extraluminal foreign bodies are left behind, including metal staples, sutures, anchors, and collagen wads. We have continued to utilize the CFA cut-down method and the PSS evaluation over the last 36 months and recently during our experience with the Boomerang (Cardiva Medical, Mountain View, CA) VAM system reported in the April 2006 issue of Endovascular Today.
The historical manual compression PSS group (N = 55) from 2000 to 2003 had a mean PSS of 2.7. The corresponding suture-mediated VCD PSS was 2.65 for that time frame. Since 2003, an additional 40 manual compression and five suture-mediated CFAs were analyzed with both manual compression and suture-mediated PSS consistently remaining between 2.5 and 2.75, which underscores the significant perivascular inflammatory healing process observed during CFA cut downs, leading us to our original conclusion: manual compression is not benign!
We exposed four CFAs for EVAR 10 to 16 days (mean = 14) after prior Boomerang VAM. The mean time to exposure in the 2003 historical PSS group was also 14 days. The Boomerang PSS was 1.75, which represented significantly less scarring than the 2.75 with the historical manual compression group. It is scientifically unproven but clinically likely that the lower the PPS score is, the more optimal the CFA healing process and the overall long-term preservation of the CFA access will be. Even though these were few CFA exposures and this analysis may raise more questions than answers, it was obvious these groins and vessels were much less scarred than our experience with CFA dissections after traditional manual compression and use of a VCD (Figure 4).
PURSUIT OF THE IDEAL VAM SYSTEM?
It appears that after a decade and a tremendous amount of resources dedicated by industry to design the ideal VCD to replace traditional manual compression as our gold standard for VAM, success has been achieved in only 30% to 40% of the more than 7 to 8 million annual catheter-based treatments, considering that currently 60% to 70% of all diagnostic and interventional cases performed use traditional manual compression.23 We continue to contend that manual compression is not benign but have lowered our threshold to seek strategies to facilitate manual compression as we remain concerned about the limitations and catastrophic complication risks with our current VCDs.
A recent report by Guillinta et al described a patient with limb ischemia who had nine catheter-based procedures through the right CFA with multiple VCDs (manual compression = 4, Perclose [Abbott Vascular, Inc. Abbott Park, IL] = 3, Angioseal [St. Jude Medical, St. Paul, MN] = 1, and Vasoseal [Datascope Corporation, Montvale, NJ] = 1).24 We will likely see more of these reports and see more of these patients in our practice as we leave VCD debris and significant scar tissue on the CFA. We must protect the CFA and all routes of vascular access for future reaccess, and it is likely that leaving behind foreign bodies, VCD debris, and significant manual compression scarring simply is not benign!
Vascular access site complications remain the underappreciated Achilles' heel of our endovascular procedures. Traditional manual compression, our gold standard, has remained unchanged for more than 50 years and should be considered a tarnished gold standard when compared to the many advanced standards of modern health care. It is imperative that improved methods of VAM be developed, and it appears justified to pursue non-VCD and also nontraditional facilitated manual compression systems as alternatives to our current VAM strategies.
The authors thank Kelly Tilbe, NCMA, for her assistance with technical manuscript preparation.
David E. Allie, MD, is Director of Cardiothoracic and Endovascular Surgery at the Cardiovascular Institute of the South in Lafayette, Louisiana. He has disclosed that he holds no financial interest in any product or manufacturer mentioned herein. Dr. Allie may be reached at (800) 582-2435; email@example.com.
Ron Smalling, MD, is from St. John's Clinic-Cardiology, Springfield, Missouri. He has disclosed that he holds no financial interest in any product or manufacturer mentioned herein. Dr. Smalling may be reached at (417) 820-3911; firstname.lastname@example.org.
Chris J. Hebert, RT, RCIS, is Director of Cardiovascular Services at the Cardiovascular Institute of the South in Lafayette, Louisiana. He has disclosed that he holds no financial interest in any product or manufacturer mentioned herein. Mr. Hebert may be reached at (800) 582-2435; email@example.com.
Craig M. Walker, MD, is Medical Director and Founder, Cardiovascular Institute of the South in Houma, Louisiana. He has disclosed that he holds no financial interest in any product or manufacturer mentioned herein. Dr. Walker may be reached at (800) 445-9676; firstname.lastname@example.org.