Can the National Kidney Foundation Guidelines for First-Time Arteriovenous Fistula Creation Be Met in Underserved End-Stage Renal Disease Patients?
Article Outline
The aims of this study were to determine whether National Kidney Foundation (NKF) guidelines for native arteriovenous fistula (AVF) creation (at least 50% of all new end-stage renal disease [ESRD] patients and 40% of prevalent hemodialysis patients) could be met in an underserved population who presented in late stages of ESRD. We also sought to determine 1-year AVF patency rates and factors associated with early thrombosis. One hundred seventy-six patients underwent hemodialysis access surgery during the period 2003-2005 with a mean age of 51 years. Sixty-two percent were male, and 48% had diabetes mellitus. Ultrasound vein mapping was performed in only 37%. Temporary central venous access was necessary in 109 patients (62%) due to late presentation. Of the 160 patients who were first-time access, 137 (86%) received a native AVF and 23 (14%) had an arteriovenous graft. There was a higher rate of AVF creation in males (91% vs. 75% for females, p = 0.005). The 1-year primary patency was 90%. There were no differences in early thrombosis or 1-year patency rates with respect to gender, age, ethnicity, insurance status, presence of temporary access, or use of preoperative vein mapping. In an underserved population, NKF guidelines for native AVF for first-time access can be superseded with an excellent 1-year patency, despite late presentation.
Introduction
In recent years, the National Kidney Foundation (NKF) has recommended placing arteriovenous fistulas (AVFs) in at least 50% of all new end-stage renal disease (ESRD) patients, and 40% of prevalent hemodialysis patients.1 AVFs are favored over arteriovenous grafts (AVGs) due to higher patency rates, decreased need for corrective intervention, less infectious risk, and lower overall and costs. However, the Dialysis Outcomes Practice Patterns Study (DOPPS) reported a disappointing 15% AVF rate in incident ESRD patients in the United States and only a 24% AVF rate in prevalent patients.2 More recently, some centers have advocated the routine use of preoperative vein mapping in order to increase the rate of AVF formation.
In the underserved population with ESRD, it is potentially difficult to meet national guidelines as patients have poor access to routine medical care and subsequently present late in the course of renal failure. This delayed presentation leads to a greater need for temporary central venous access and a likely decrease in native AVF creation. Some authors have postulated that the use of central temporary access may adversely affect the success rate of native AVF formation.3 The purpose of our study was to evaluate whether NKF guidelines for AVFs could be met in an underserved population with ESRD who present in late stages of renal failure. In addition, we sought to determine short-term fistula primary patency rates as well as the validity of our selective approach to preoperative vein mapping.
Methods
Patient Data
A retrospective analysis was performed for all hemodialysis access operations between 2003 and 2005 at Harbor-UCLA Medical Center. The hospital is managed by the Los Angeles County Department of Health Services and primarily treats underserved and uninsured patients. All hemodialysis access procedures were performed by one of two vascular surgeons in conjunction with surgical residents. Clinical data collected included age, gender, race, payer status, history of diabetes mellitus (DM) and congestive heart failure (CHF), body mass index (BMI), history of prior AV access, presence of temporary access via central line, use of preoperative imaging, and site of access placement. First-time access was defined as the initial permanent AVF or AVG procedure, irrespective of whether central access was previously placed. For patients with a temporary access, the number of days from temporary access placement until the AVF/AVG procedure was recorded. The primary end point was 1-year primary patency of first-time AVF. Secondary end points included wound infection, cellulitis, or thrombosis of the AVF/AVG at 30 days and infection or thrombosis of the temporary access.
Statistical Analysis
All patient data were collected in an Excel database (Microsoft Excel; Microsoft, Redmond, WA) and translated into a native SAS format using DBMS/Copy® (Dataflux; SAS Institute, Cary, NC). Descriptive statistics were calculated for all variables. Numerical variables were compared using the nonparametric Wilcoxon rank sum test and are reported as means. Categorical or nominal variables were compared using the chi-squared test or Fisher's exact test, as appropriate. Univariate and multivariate analyses were performed to determine factors associated with AVF creation, early thrombosis, and patency of first-time AVFs. For the purpose of statistical analysis of the early thrombosis rate and patency of first-time AVFs, the AVF sites were grouped as follows: radiocephalic-wrist AVF (including those that were created at the snuffbox, wrist, and forearm), brachiocephalic-elbow AVF, brachial artery-basilic vein antecubital AVF, and fistulas that were created at other locations that did not meet the aforementioned criteria (categorized as “other”).
Dialysis Access Algorithm
Upon presentation to the vascular surgery team, a physical examination, including a careful pulse exam and Allen's test, was performed. A tourniquet was then placed on the nondominant arm. If the cephalic vein was visualized from the wrist to the elbow or from the elbow to the upper arm and there was no history of ipsilateral central vein access or recent ipsilateral intravenous lines, the patient proceeded to surgery without preoperative vein mapping. Otherwise, a duplex scan was obtained. The order of preference for construction of hemoaccess was as follows: cephalic vein to radial artery AVF at the wrist, cephalic vein to brachial artery AVF at the elbow, basilic vein to brachial artery AVF at the elbow (with transposition), loop forearm AVG, and upper arm AVG.
Results
Patient Demographics
One hundred seventy-six patients underwent hemodialysis access surgery between 2003 and 2005. The mean age was 51 years, and 62% were male. The majority of patients were Hispanic (61%), followed by African American (21%) (Table I). Sixty-six percent of the cohort was uninsured. The average BMI was 29 kg/m2. Nearly half (48%) were diabetic, and 12% had a history of CHF. The most common risk factors for ESRD were hypertension (HTN, 52%) and DM (43%). Preoperative ultrasound vein mapping was performed in only 37%. Because of late presentation to the nephrologist, 109 (62%) patients required immediate hemodialysis access via a tunneled central line. Of the 109 patients with temporary access, seven (7%) developed infection of the temporary access line. The average length of time between placement of the temporary access and permanent AVF/AVG was 51 days.
Table I. Patient demographics (n = 176)
| Patient variables | |
 Mean age | 51 ± 15 |
| Male gender | 109 (62%) |
| Race | |
| Hispanic | 107 (61%) |
| African American | 37 (21%) |
| Asian | 16 (9%) |
| White | 10 (6%) |
| Other | 6 (3%) |
| Mean BMI | 29 ± 19 |
| No health insurance | 117 (66%) |
| Diabetes mellitus | 85 (48%) |
| Preoperative vein mapping | 65 (37%) |
| Presence of temporary access | 109 (62%) |
First-Time Access Surgery
One hundred sixty patients underwent first-time hemodialysis surgery. Of these, 137 (86%) received an AVF and 23 (14%) had an AVG placed. The types of access performed included wrist radiocephalic (23%), antecubital brachiocephalic (18%), forearm radiocephalic (17%), antecubital brachial artery-basilic vein (13%), forearm AVG (10%), “snuffbox” radiocephalic (9%), upper arm AVG (4%), and other fistula (6%).
Rates of AVF creation by patient variables are shown in Table II. Univariate analysis demonstrated a significantly higher rate of AVF placement in males (91% vs. 75% in females, p < 0.005) as well as for patients who were uninsured (90% vs. 77% in the insured, p = 0.03). The rate of AVF creation in patients aged <50 years (92%) approached significance compared to patients >50 years (81%) (Table II). On multivariate analysis, there was a significant independent inverse relationship between first-time AVF creation and older age (odds ratio [OR] = 0.976, confidence interval [CI] 0.935-1.001, p = 0.0584), female gender (OR = 0.34, CI 0.133-0.871, p = 0.0246). The OR for fistula creation in the uninsured was 2.242 (CI 0.874-5.749, p = 0.0929).
Table II. Rates of first-time AVF creation and 1-year patency by patient variables
| Patient factors | First-time AVF rates | p | 1-year patency | p |
|---|---|---|---|---|
| Age (years) | 0.07 | 0.46 | ||
| >50 | 72/89 (81%) | 93% | ||
| <50 | 65/71 (92%) | 94% | ||
| Gender | 0.005 | 0.17 | ||
| Male | 94/103 (91%) | 94% | ||
| Female | 42/57 (75%) | 83% | ||
| Ethnicity | 0.73 | 0.17 | ||
| Asian | 11/14 (89%) | 100% | ||
| African American | 31/35 (89%) | 75% | ||
| Hispanic | 83/95 (87%) | 92% | ||
| White | 8/10 (80%) | 100% | ||
| Insurance status | 0.03 | 0.23 | ||
| Insured | 43/56 (77%) | 94% | ||
| Uninsured | 94/104 (90%) | 88% | ||
| BMI | 0.47 | 0.25 | ||
| <25 | 56/63 (89%) | 94% | ||
| >25 | 94/104 (90%) | 88% | ||
| History of DM | 0.11 | |||
| Yes | 63/78 (81%) | 95% | 0.41 | |
| No | 74/82 (90%) | 87% | ||
| Location of access | 0.16 | |||
| Radiocephalic-wrist | 98% | |||
| Brachiocephalic-elbow | 79% | |||
| Brachial-basilic antecubital | 88% | |||
| Other | 100% | |||
| Temporary access | 1.0 | 0.54 | ||
| Yes | 80/93 (86%) | 89% | ||
| No | 57/67 (85%) | 91% | ||
| Use of preoperative imaging | 0.096 | 0.20 | ||
| Yes | 44/56 (79%) | 82% | ||
| No | 93/104 (89%) | 94% |
The primary AVF patency rates at 1 year, analyzed by patient variables, are shown in Table II. The overall 1-year patency was 90%. The 95% CI for primary patency was 82-97%. The mean follow-up was 321 days. There were no differences in 1-year patency rates with respect to any of the variables analyzed, including the use of preoperative ultrasound vein mapping, presence of temporary central venous access, age, gender, and ethnicity. The long-term fistula patency curve is shown in Figure 1.
Fig. 1
Life-table graph for primary AVF patency. Line numbers are fistulas at risk for a given interval. Solid line demonstrates the actuarial survival curve, and dotted lines represent the 10% standard error limits.
Sixteen patients (10%) had complications following first-time access surgery including cellulitis (five patients, 3%), wound hematoma (three patients, 2%), and early thrombosis (eight patients, 5%). The early AVF thrombosis rates by type of access were as follows: brachiocephalic AVF (13%), radiocephalic AVF (1.5%), brachial artery-basilic vein (0%), and other AVF (0%) (p = 0.0365). There was no significant association found between early thrombosis and gender, BMI, DM, history of prior access, or use of preoperative imaging.
Second-Time Access Surgery
Of the 16 patients who had second-time access, seven (44%) received an AVF and nine (56%) an AVG. Types of access placed included antecubital brachiocephalic (19%), antecubital radiocephalic (13%), wrist radiocephalic (6%), forearm radiocephalic (6%), forearm AVG (38%), and upper arm AVG (13%). Following second-time access surgery, six patients had complications, including five with cellulitis and one with a pseudoaneurysm. None had early thrombosis. Given the small number in this cohort, further analysis was not performed.
Discussion
The present study demonstrates that the NKF guidelines for fistula placement can be superseded in an underserved population. The rate of AVF placement was 86% for first-time access patients and 44% for second-time access patients. The majority of patients presented in late stages of ESRD, as highlighted by the fact that 62% required temporary access via tunneled central line for immediate hemodialysis. AVF creation was lower in women (75%) than in men (91%) and tended to be lower in patients >50 years (81%) than those <50 years (92%) of age. For first-time native AVF, the early thrombosis rate was 5% and the 1-year primary patency was 90%. However, it must be noted that this patency rate does not reflect functional patency. Early thrombosis and 1-year patency rates were not influenced by ethnicity, insurance status, use of preoperative ultrasound, female gender, need for temporary access, or the presence of DM.
Despite NKF guidelines, only about 20% of hemodialysis patients in the United States use an AVF for dialysis access.4 The low rates of AVF creation in the United States have been ascribed to several factors. Sands and Perry5 identified five areas that contribute to the difficulty in maintaining NKF guidelines, including “availability and funding of pre-end-stage renal disease care, limited use preoperative imaging, financial disincentives, lack of accountability, and educational deficits.” Friedman et al.6 demonstrated the importance of early referral to a nephrologist for pre-ESRD vascular access placement. Patients followed by nephrologists and with gradual progression of ESRD had a 46% AVF rate for first dialysis compared to 21% in patients with rapid progression of ESRD and only 2% in patients with no pre-ESRD care. The present study suggests that AVFs can be created even if patients are referred late and are unknown to the nephrologist. Other factors associated with a lower likelihood of native AVF creation include female gender, older age, obesity, peripheral vascular disease, African American race, and low income.7, 8 We also observed a lower rate of AVF creation in females, which is likely due to smaller vessels, although the 1-year patency rates were unaffected by gender. Interestingly, AVF rates were higher in the uninsured than the insured, though we suspect that this is a reflection of the fact that the insured patients were older and, thus, had poorer-quality veins.
More recently, several groups in the United States have reported marked improvement in the rates of AVF creation. Sands and Miranda9 increased AVF prevalence from 28% to 44% with the adoption of a multidisciplinary AVF creation program. Using a similar approach, Gibson et al.10 increased AVF prevalence from 41.3% to 73.7%. By the last year of study, 95% of accesses placed were native fistulas. Primary patency for AVF was 72% and 61% at 1 and 2 years, respectively. The AVF rate (86%) and 1-year patency (90%) in the present study compare favorably with these reports. The ability to achieve a high AVF rate in the present study of an underserved population may have several explanations. The age of our population was on average 51 years, perhaps resulting in better-quality veins. The average age of ESRD patients in other studies has ranged 55-65 years. The ethnicity of our population, with 61% Hispanic and only 21% African American, may also be contributory. Several reports have commented on the less favorable vascular anatomy of African American patients.11, 12 Few studies, however, are available regarding the success of AVF formation in the U.S. Hispanic population. Another explanation for the high rate of AVF creation may be the relatively high operative volume at our institution. O'Hare et al.13 demonstrated that patients undergoing vascular access surgery at centers performing more than 30 first-time access procedures annually were more than three times as likely to receive an AVF in comparison to centers performing 10 or fewer procedures per year.
The use of temporary hemoaccess via a central venous catheter is a major source of morbidity, mortality, and expenditure in ESRD patients. Such catheters are prone to thrombosis, infection, central venous stenosis, and inadequate dialysis.14, 15 O'Hare et al.13 demonstrated that patients requiring temporary hemoaccess had significantly lower odds of AVF placement. Other studies have suggested that temporary access is associated with reduced patency of subsequent AVF and AVG.3 The reason for the reduced patency is unclear. Some have postulated that the decreased patency may be due to central vein injury leading to stenosis, premature cannulation of the AVF, or other unknown factors that may inhibit AVF maturation. Importantly, in the present study, an aggressive approach was taken for AVF creation even though the majority of patients (62%) presented late for access placement. Comparatively, other studies have noted that 25-44% of patients present less than a month before dialysis needs to be initiated.16, 17 Furthermore, the 1-year patency rate was not affected by the presence of temporary central venous access. Although we do not have longer-term patency data, these preliminary results suggest that such an aggressive policy of AVF creation is warranted even in this type of patient population.
Many studies support the routine use of ultrasound vein mapping in all patients prior to fistula placement. Reasons cited include confidence in the vascular anatomy,18 improved short-term fistula patency,19 and detection of occult venous stenoses or occlusions.18 However, our belief is that a careful physical examination may be adequate to determine suitability for AVF in most patients. We feel that ultrasound vein mapping should be reserved for patients with visually inadequate upper and/or lower arm veins, patients with an ipsilateral central access, or those with a recent ipsilateral intravenous line. A study by Wells et al.20 supports our view that ultrasound is unnecessary in most patients and a careful examination is sufficient for determining site of dialysis access. Overall, only 37% of patients in the present study underwent ultrasound vein mapping, and the use of preoperative venous mapping was not associated with a decrease in AVF rates or patency. It should be noted, however, that the average age of the patients in the present study was younger than in other studies; thus, patients may have had more favorable veins for fistula creation.
In conclusion, in an underserved, predominantly Hispanic population, NKF guidelines for native AVF creation for first-time access patients were superseded despite the fact that many patients presented late and required temporary central venous access. Acceptable rates of early thrombosis and 1-year primary patency were achieved, even though preoperative ultrasound was used sparingly. Early thrombosis and 1-year patency rates were not affected by the use of vein mapping, gender, BMI, or the need to initiate preaccess hemodialysis via a temporary catheter. Increasing AVF prevalence is crucial to improving patient outcomes and decreasing expenditures.9 The low rate of AVF creation nationally is a systemic problem that requires modification of our dialysis care system. A multidisciplinary approach must be taken toward caring for patients with ESRD, particularly those who are indigent.
References
- . K/DOQI clinical practice guidelines for vascular access, 2000. Am J Kidney Dis. 2001;37:S137–S181
- . Vascular access. Current practice and practical aspects of management, ASN Renal Week 2000. Am Soc Nephrol. 2000;377–385
- Vascular access use in Europe and the United States: results from the DOPPS. Kidney Int. 2002;61:305–316
- National surveillance of dialysis-associated diseases in the United States, 1997. Semin Dial. 2000;13:75–85
- . Where are all the AV fistulas?. Semin Dial. 2002;15:146–148
- First hemodialysis access selection varies with patient acuity. Adv Ren Replace Ther. 2000;7(Suppl. 1):S4–S10
- Factors associated with the prevalence of arteriovenous fistulas in hemodialysis patients in the HEMO study. Kidney Int. 2000;58:2178–2185
- Predictors of type of vascular access in hemodialysis patients. JAMA. 1996;276:1303–1307
- . Increasing numbers of AV fistulas for hemodialysis access. Clin Nephrol. 1997;48:114–117
- Assessment of a policy to reduce placement of prosthetic hemodialysis access. Kidney Int. 2001;59:2335–2345
- . Increasing arteriovenous fistulas in hemodialysis patients: problems and solutions. Kidney Int. 2002;62:1109–1124
- National profile of practice patterns for hemodialysis vascular access in the United States. J Am Soc Nephrol. 2002;13:2117–2124
- Impact of surgeon and surgical center characteristics on choice of permanent vascular access. Kidney Int. 2003;64:681–689
- National Kidney Foundation-Dialysis Outcomes Quality Initiative. NKF-DOQI clinical practice guidelines for vascular access. Am Kidney Dis. 1997;30(Suppl. 3):S150–S191
- . Hemodialysis vascular access morbidity. J Am Soc Nephrol. 1996;7:523–535
- United States Renal Data System. The USRDS Dialysis Morbidity and Mortality Study: wave 2. Am J Kidney Dis. 1997;30:S67–S85
- Is there a rationale for rationing chronic dialysis? A hospital based cohort study of factors affecting survival and morbidity. BMJ. 1999;318:217–223
- . Vascular ultrasonography prior to dialysis access surgery. Am J Surg. 2002;184:568–572
- A strategy for increasing use of autogenous hemodialysis access procedures: impact of preoperative noninvasive evaluation. J Vasc Surg. 1998;27:302–308
- Selective use of ultrasonographic vascular mapping in the assessment of patients before hemodialysis access surgery. Br J Surg. 2005;92:1439–1443
Presented at the 2006 Southern California Vascular Surgery Society Meeting, Temecula, California, May 2006.
PII: S0890-5096(07)00359-7
doi:10.1016/j.avsg.2007.10.004
© 2008 Annals of Vascular Surgery Inc. Published by Elsevier Inc All rights reserved.
