Utilization of arteriovenous grafts secondary to fistula failure in patients on dialysis

Issue: September 2018

ByZiauddin Ahmed, MD

ByNicole Evans, MD

ByShamik Bhadra, MD

BySandeep Aggarwal, MD

ByMaliha Ahmed, DO

Improvement of the clinical practice guidelines for chronic hemodialysis patients has become a priority for CMS.^1,2 The use of an arteriovenous fistula has become widely endorsed as the optimal vascular access device. The other two commonly used access modalities include an arteriovenous graft and a central venous catheter.

In 2003, CMS created the Fistula First Breakthrough Initiative (FFBI) with the purpose of providing better patient-centered care through quality improvements. The initiative encouraged an overall decrease in the use of central venus catheters (CVCs) for chronic vascular access, and an increase in the primary use of AVFs with a national prevalence goal of 66%.³ Consequently, from 2003 to 2012, the prevalence of AVF use increased from 32.2% to 60.4%.^4,5 However, catheter rates remain high due to late nephrology referrals for asymptomatic patients, many who were unwilling to undergo surgical fistula placement in anticipation of a worsening disease state.⁶

After the success of the FFBI, CMS implemented the End-Stage Renal Disease Quality Incentive Program (QIP) on Jan. 1, 2012 to further link quality to payment. This policy included two new sub-measures for vascular access type: fistula and catheter prevalence.4 Active fistula rates had to be at least 58%, while the catheter rate should be no more than 14%.² There were no guidelines established for the use of AVG placement. The consequence of not meeting these quality standards results in a financial penalty of up to 2% against the dialysis center for suboptimal patient care.⁴

Although there is significant evidence supporting the use of AVFs for maintenance hemodialysis, some believe the intent of the FFBI had its drawbacks and the strict fistula recommendation may have been harmful to some patients.^6,7 With the CMS goal of increasing AVF use, patients who have experienced AVF failure in one arm are subjected to starting a new AVF in the other arm instead of salvaging the same site with an AVG.⁸ Recent data from the Dialysis Outcomes and Practice Patterns Study (DOPPS) have shown that between 1996 and 2015, there was a large shift in the United States from the preferred lower arm placement of AVFs to upper arm AVFs, raising concerns about the long-term implications for available access sites.⁹ Even without subjecting the patient to using both arms, some patients have complained about body image concerns due to the unattractive dilated appearance of their fistula as compared to grafts and catheters.⁴ Additionally, fistulas may be more difficult to cannulate than grafts, especially during initial use, which can make them more problematic to use and tolerate.10 Fistula creation itself also can lead to significant alterations in cardiac structure and function, as well as endothelial dysfunction, resulting in increased morbidity and mortality in already high-risk individuals.¹¹ Finally, cost of placing and maintaining the AVF has increased, according to a recent study, because of the number of interventions involved in keeping the access open. In that study, the authors calculated that costs were two- to three-times higher for patients whose AVFs experienced primary or secondary patency loss and four-times higher for patients who never used their AVFs compared with patients whose AVFs maintained primary patency.¹²

PAGE BREAK

As the ESRD QIP has expand its use of quality indicators, there is an opportunity for patient and provider feedback. CMS has shifted its role from a passive payer to an active partner with quality care at the center of its agenda. In taking this new stance, it is critical that CMS be receptive to the advice of dialysis specialists and consider modifying its policies to provide its beneficiaries with the best care possible, including the right choice for access placement.

In this study, we look at the potential use of AV graft criteria in addition to the current fistula and catheter requirements to determine the best access.

Methods and results

Chart review was performed at a dialysis unit with 24 stations located in Philadelphia. This center has been providing comprehensive patient care for patients with ESRD since 1986, including hemodialysis and in-center peritoneal services. A retrospective analysis of 123 patients was conducted on the type of access (AVF, AVG and CVC) in place and as reported to the CMS in December 2014. Statistical analysis and calculations were done using IBM SPSS version 22, and z-score was calculated to compare between group proportions.

In the study group, 59 patients (47.97%) had AVFs only at the time of the analysis and 11 (8.94%) had a new fistula with a catheter in place, reaching a total prevalence rate of 70 patients (56.91%) with fistulas. This prevalence rate placed the clinic into the financial penalty category for not meeting the CMS criteria of at least 58% of patients having active/working AVFs.

The prevalence of patients with a placed AV graft only was 41 (33.33%) and a graft with catheter was three (2.44%), with a total prevalence of 44 grafts (35.77%). There were nine patients (7.32%) with catheters only (see Figure). The difference in concomitant catheter requirement in patients with graft vs. fistula was not significant statistically (z score 1.4088, two-tailed P = 0.15854).

Figure shows the access type in comparison to CMS policy guidelines. Source: Ziauddin Ahmed, MD; Nicole Evans, MD; Shamik Bhadra, MD; Maliha Ahmed, DO; and Sandeep Aggarwal, MD.

Of the 44 patients with grafts, 22 patients had an active fistula in the past or had a fistula attempted in the past before placement of an AV graft. Of these patients in the group, 13 patients (29.54%) had a primary fistula failure and nine patients (20.45%) had inadequate vessels for a fistula and consequently did not receive an AVF. Two of the 44 patients (4.54%) with grafts placed had unclear information since they had been transferred from another city/dialysis unit with no information regarding the reason for their graft placement or previous attempts of fistula placement. None of the grafts were placed without first attempting a fistula creation.

PAGE BREAK

Analysis

It is the goal of the dialysis team to place AVFs that can be successfully cannulated. However, we question why CMS does not consider AVG placement after fistula failure as an equally successful event if the staff believe that is a better alternative to placing another AVF. Our data show that 92 patients (74.79%) had a fistula as a primary vascular access at one point during their care, which surpassed the CMS performance standard. However, surgical placement of the AVF does not always result in a successfully working dialysis access. We propose two ways in which CMS can better account for the use of AVG placement after AVF failure.

1. If CMS were to exclude failed fistulas from the total number of cases, we would achieve a total of 70 of 101 patients, or a 69.31% AVF rate. That percentage would meet and exceed the CMS performance standard (58%); and

2. If CMS were to include patients with failed AVFs that were successfully grafted in the total AVF category, this would result in 92 of 123 patients with a successful access, or a 74.79% total AVF rate at our center. That would meet and exceed the CMS performance standard.

Discussion

Accountability and quality improvement have become increasingly emphasized and important topics in health care today. Chassin and colleagues13 have promoted quality in hospitals and health systems by defining accountability measures into four practical points, linking care process, intended care and outcomes. Those points are:

1. Strong scientific evidence demonstrates that the implementation of the care process improves patient outcomes;

2. Performing the care process requires a few intervening steps before intended care is delivered;

3. The accountability measure accurately assesses whether the intended care has been delivered when the care-process is performed; and

4. Implementing the care process has little or no chance of inducing adverse outcomes.

The ESRD QIP program may violate the fourth criteria. By failing to include AVG measures in the QIP criteria guidelines, AVF placement may be used inappropriately in certain patients, causing increased morbidity, mortality and health care costs. Placing a proximal AVF without attempting a distal AVG at a failed fistula site may lead to a burnout of AV access sites earlier. Additionally, appropriate patient-selective use of AVG, rather than AVF, is financially punishable by the QIP if standards criteria are not met. Primary graft use after AVF failure also may reduce catheter dependence especially in patients, with advanced CKD nearing dialysis, failing dialysis adequacy, high catheter-related infection risk and limited lifespan (elderly or terminally ill), which is not accounted for by ESRD QIP.

PAGE BREAK

Although AVF is widely accepted as the optimal vascular access device for maintenance hemodialysis, many of the original studies on this topic were non- randomized, single-center studies from the 1980s and early 1990s comparing the outcomes of AVF and AVG without controlling for external biasing factors or providing standardized definitions.^14-16 More recent studies have shown the actual benefits of AVF might be overestimated based upon patient comorbidities and vascular anatomy.¹⁷ Elderly patients and patients with diabetes and peripheral artery disease are more likely to receive an AVG, thus skewing the data in favor of AVF for better overall outcomes. Consequently, as more AVF are being used in elderly and patients with multiple comorbidities, we anticipate that AVF performance may be disappointing. A recent study by Lee and colleagues looking at the challenges of maintaining vascular access in the elderly population showed AVFs required more interventions in the first 6 months post-creation compared to AVGs.18 Patients who received an AVF depended on a catheter substantially longer before the permanent access became functional compared to those receiving an AVG.

As evidence-based research continues to define the arteriovenous access policy, standardized definitions must be used when describing the patency of a fistula. The current policy was based on studies that analyzed the functional patency of an AVF once successfully cannulated. By defining patency as the functional state, a critical phase between AVF creation and cannulation is excluded by the CMS. Success is dependent on the quality and compliance of a vessel and its ability to adapt to increased blood flow. During this period of time, failure rates are highest.¹⁹ When an AVF is created, especially in the upper arm where the vein is deeper, it is now recommended that the vein should be mobilized superficially to allow for easier needle access and to avoid hematoma formation.^20,21 As a result of this maneuver, the adventitial layer and vaso vesorum are removed, making the vein more vulnerable to sclerosis, stenosis and loss of viable circulation. This pathophysiologic mechanism has been described in numerous studies, including a large-scale meta-analysis which established an association between AVF and high risk of primary failure due to thrombosis, stenosis and maturation failure as compared to AVG.¹¹ In addition, the long unattractive scar from the fistula mobilization surgery is unacceptable to many patients, especially in those predisposed to keloid formation.

The lack of a standardized definition to distinguish functional from surgical patency has diminished the value of previously stated failure rates and biased the subsequent policies which guide patient care. Furthermore, a recent study showed AVF creation can cause widespread adverse hemodynamic consequences which may lead to worsening left ventricular hypertrophy, progressive high-output cardiac failure, atrial fibrillation, pulmonary hypertension, further atherosclerotic changes and overall increased morbidity and mortality in already high-risk patients.⁹

PAGE BREAK

Conclusions

It is our responsibility as health care providers to inform patients of the clinical trade-offs in choosing AVF vs. AVG. Fistulas have been proven to have better long-term survival and require fewer interventions overall as compared to grafts. However, fistulas have a higher primary failure rate, require more interventions to achieve maturation and require greater catheter dependence as compared to grafts.

As a result, the patients who are most likely to benefit from a graft are those who have already had a primary AVF failure, poor vascular anatomy or a short life expectancy. To ensure individualized patient care, patients should be thoroughly evaluated prior to vascular access placement with a comprehensive history and physical exam and careful preoperative vessel assessment. Instead of only focusing on the fistula creation, nephrologists should help patients preserve their existing site as long as possible. Ideally, the access should be started with a forearm radio-basilic fistula. If the patient experienced primary AVF failure or the access failed after a few years, it is still beneficial to use the same forearm site by converting it to a graft instead of abandoning the forearm by creating an upper arm fistula. Unlike fistulas, the secondary patency of a graft can be significantly extended by multiple revisions through jump grafts prior to moving on to a new site.

In our review, we do not seek to disqualify the benefits of AVF or the effort that the CMS QIP has put forth. Rather, we urge CMS to consider modifying their guidelines and performance standards to better account for individualized patient care. Through our single-center clinical practice we have identified a large percentage of patients who have failed AVF and successfully recovered the same access site with AVG. In the patient’s best interest it is usually beneficial to salvage a failed site rather than start another site or switch arms simply to meet a national quota. It is our recommendation that use of the AVG be included in the QIP evaluation of access placement and it be considered a viable alternative to an AVF.

References:
End-Stage Renal Disease (ESRD) Quality Initiative: Questions and answers. November 2004. www.cms.gov/medicare/end-stage-renal- disease/esrdqualityimproveinit/downloads/fffaqs.pdf.
End-State Renal Disease Quality Incentive Program 2014 Final Rule. www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/ESRDQIP/Downloads/UpdatedBaseline-2014-FR.pdf.
Clinical practice guidelines for vascular access. Am Kidney Dis. 2006;doi: 10.1053/j.ajkd.2006.04.040.
Fishbane S, et al. Am Kidney Dis. 2012;doi:10.1053/j.ajkd.2012.07.002.
Wish JB, et al. Sem in Dial. 2010;doi:10.1111/j.1525-139X.2010.00797.x
Gowda A, et al. Kidney Dis. 2014;8(3):218-224.
Lee HW, et al. Sem in Dial. 2013;doi:10.1111/sdi.12040.
Salman L, et al. Clin J Am Soc Nephrol. 2010;doi:10.2215/cjn.02380310.
Pisoni R, et al. Am Kidney Dis. 2018;doi:10.1053/j.ajkd.2017.09.012.
Leake AE, et al. J Vasc Surg. 2015;doi:10.1016/j.jvs.2015.02.018.
Dundon BK, et al. Int J Nephrol Renovasc Dis. 2014;doi:10.2147/ijnrd.s66390.
Thamer M, et al. Am Kidney Dis. 2018;doi:10.1053/j.ajkd.2018.01.034.
Chassin MR, et al. New Engl J Med. 2010;doi:10.1056/NEJMsb1002320.
Al-Jaishi AA, et al. Am Kidney Dis. 2014;doi:10.1053/j.ajkd.2013.08.023.
Allon M, et al. Kidney Int 2000;doi:10.1111/j.1523-1755.2000.00391.x
Disbrow DE, et al, J Am Coll Surg. 2013;doi:10.1016/j.jamcollsurg.2012.12.021.
Huijbregts HJ, et al. Clin J Am Soc Nephrol. 2008;doi:10.2215/cjn.02950707.
Rooijens PP, et al. J Vasc Surg. 2005;doi:10.1016/j.jvs.2005.05.025.
Lee T, et al. Am J Kidney Dis. 2018;doi:10.1053/j.ajkd.2018.03.023.
Beathard GA. AVFs and patient access. ESRD Network of Texas Inc. 2003:1-14.
Konner K, et al. J Am Soc Nephrol. 2003;doi:10.1097/01.ASN.0000069219.88168.39.

PAGE BREAK

For more information:
Ziauddin Ahmed, MD, is a professor of medicine at Drexel University College of Medicine. Nicole Evans, MS, MD, is a resident in internal medicine at University of Southern California/Los Angeles County Medical Center. Shamik Bhadra, MD, is a consultant in nephrology and is in private practice in Los Angeles. Maliha Ahmed, MD, is a PGY-1 in internal medicine at Drexel/Hahnemann University Hospital. Sandeep Aggarwal, MD, is an associate professor of medicine and associate fellowship program director at Drexel University College of Medicine. Disclosures: The authors report no relevant financial disclosures.