Abstract
Background and objectives
Recent evidence indicates that fistula maturation and patency may be compromised in the elderly dialysis population compared with younger patients. The objective of this study was to characterize the short-term outcomes of arteriovenous fistulas and arteriovenous grafts for hemodialysis access in the Medicare population.
Design, setting, participants, & measurements
This was a retrospective cohort study performed using Medicare Part A and B claims data from 2006 through 2011. The study population included 16,464 dialysis-dependent patients age ≥66 years undergoing arteriovenous fistula and arteriovenous graft creation. The primary outcome measure was incidence of repeat fistula/graft creation and tunneled catheter placements in the 12 months after arteriovenous fistula and graft creation.
Results
In the 12 months postindex fistula/graft, the mortality in the fistula group was 28.2% versus 29.9% in the graft group (P=0.03). A repeat fistula/graft creation was required in 26.9% of patients in the fistula group and 16.7% in the graft group (P<0.001). There was no significant difference in the proportion of patients who required a tunneled hemodialysis catheter in the 12 months after an index fistula creation (fistula 28.4% versus graft 27.3%, P=0.19). In the index fistula group, 44.4% of patients required a repeat fistula/graft creation and/or a tunneled catheter, compared with 33.7% in the graft group (P<0.001). At 365 days after the index fistula/graft, the repeat fistula/graft/catheter-free survival was 39.7% in the fistula group versus 46.0% in the graft group (P<0.001). Index fistula was associated with a higher risk of loss of repeat fistula/graft/catheter-free survival with an odds ratio of 1.19 (95% confidence interval, 1.13 to 1.24).
Conclusions
Fistulas were associated with a somewhat lower mortality than grafts in the first 12 months after creation. However, the incidence of repeat fistula/graft creation and tunneled catheter placement is substantially higher in the first 12 months after fistula creation compared with grafts. One-year repeat fistula/graft/catheter-free survival is lower after fistula creation than grafts.
Keywords: arteriovenous shunt, ESRD, geriatric nephrology, Medicare claims, surgical failure
Introduction
Hemodialysis-dependent patients have two options for permanent upper-extremity vascular access: arteriovenous fistula and arteriovenous graft. Fistulas are created by surgically connecting a patient’s vein to an artery, whereas grafts use a prosthetic tube as a conduit between an artery and vein. Before a fistula can be used for hemodialysis access, it must dilate in diameter and become thick-walled. This process, termed maturation, is not required for grafts.
The National Kidney Foundation Dialysis Outcomes Quality Initiative Guidelines and the Fistula First Breakthrough Initiative both encourage the creation of fistulas over grafts (1,2). Similarly, the Center for Medicare and Medicaid Services ESRD Quality Incentive Program added a measure for 2014 that includes the percentage of patients who dialyze through a fistula and those using a catheter for >90 days. However, recent evidence indicates that fistula maturation and patency may be compromised in the elderly dialysis population (3,4). Furthermore, in the age ≥65 years hemodialysis population, all-cause mortality is 31% per year, suggesting that a significant proportion may not survive to experience the long-term benefits of fistulas over grafts (5). Therefore, the short-term implications of prioritizing fistulas over grafts merit careful consideration.
When a fistula or a graft is deemed unsalvageable, a new fistula/graft must be created. If immediate vascular access is necessary, a tunneled catheter must also be placed. As such, the occurrence of repeat fistula/graft creation and tunneled catheter placements after permanent vascular access creation serves as an indicator of failure of the index access. To characterize early failure rates of fistulas and grafts in the Medicare population, we investigated the incidence of repeat fistula/graft creation and tunneled catheter placements in dialysis-dependent patients in the 12 months after permanent vascular access creation.
Materials and Methods
Medicare outpatient, inpatient, and carrier files from 2006 through 2011 were queried for upper-extremity vascular access procedures identified by the current procedural terminology (CPT) codes for fistula and graft (Table 1). The first occurrence of the upper-extremity vascular access CPT code per beneficiary during the study period was defined as the index fistula/graft creation. Any new fistula/graft creation, as identified by the CPT codes referenced in Table 1, performed in the 12 months subsequent to the index fistula/graft was defined as a repeat fistula/graft creation, regardless of whether the index procedure was a fistula or a graft. Tunneled hemodialysis catheter placement procedures were also identified by CPT code (Table 1).
Table 1.
Current procedural terminology codes and procedure descriptions
| CPT Code | Procedure Description |
|---|---|
| Fistula | |
| 36818 | Arteriovenous anastomosis, open; by upper-arm cephalic vein transposition |
| 36819 | Arteriovenous anastomosis, open; by upper-arm basilic vein transposition |
| 36820 | Arteriovenous anastomosis, open; by forearm vein transposition |
| 36821 | Arteriovenous anastomosis, open direct, any site |
| 36825 | Creation of arteriovenous fistula by other than direct arteriovenous anastomosis; autogenous graft |
| Graft | |
| 36830 | Creation of arteriovenous fistula by other than direct arteriovenous anastomosis; nonautogenous graft |
| Tunneled catheter | |
| 36565 | Insertion of tunneled centrally inserted central venous access device, requiring two catheters via two separate venous access sites; without subcutaneous port or pump (e.g., Tesio-type catheter) |
| 36558 | Insertion of tunneled centrally inserted central venous catheter, without subcutaneous port or pump; age ≥5 years |
CPT, current procedural terminology.
Inclusion criteria were age ≥66 years at the time of index fistula/graft creation, qualification for Medicare exclusively by virtue of age, continuous enrollment in fee-for-service Medicare for 12 months before and after the index fistula/graft creation, and dialysis dependence at the time of index fistula/graft creation (Figure 1). These criteria represent an effort to limit the cohort to patients who were undergoing dialysis vascular access creation for the first time. Patients who died within the 12 months after the index fistula/graft creation were included. Comorbidities were defined by having ever met the claims criteria of the Chronic Condition Data Warehouse as of the date of the index fistula/graft creation (6).
Figure 1.
Included and excluded patients.
Dialysis dependence was identified using Berenson-Eggers Type of Service (BETOS) Codes P9A (Dialysis Services Medicare Fee Schedule) and P9B (Dialysis Services Non-Medicare Fee Schedule). BETOS codes are assigned for each Health Care Financing Administration Common Procedure Coding System procedure code. A Health Care Financing Administration Common Procedure Coding System code can be assigned to only one BETOS code. BETOS code categories are designed to permit objective assignment and be stable over time (7). Our primary definition of chronic dialysis dependence was at least one occurrence of BETOS 9A or 9B during the 12 months before the index fistula/graft month or within 30 days after creation of the index fistula/graft (8). A sensitivity analysis defined dialysis dependence by occurrence of BETOS 9A or 9B in 3 consecutive months, with the third month being no later than 2 months after the index fistula/graft month (9).
Covariates included in the models were all comorbidities available from the Chronic Condition Data Warehouse (except CKD, which was used to select the cohort): race/ethnicity, total covered charges from outpatient, inpatient, and carrier files in the year before index fistula/graft creation date, inpatient setting for the index fistula/graft creation, age and its square, sociodemographics within the patient’s zip code (obtained by linking residential zip codes from Medicare enrollment files to the 2000 Census), index year, index month, and state of residence (10,11).
Statistical analysis was conducted using Stata, version 13.1 (StataCorp, College Station, TX). Two group comparisons for age were assessed by a t test of equality of means. Two group comparisons for dichotomous outcomes were assessed by chi-squared tests. Logistic regressions of postindex repeat fistula/graft creation and/or tunneled catheter placement were performed on the basis of index fistula versus graft, controlling for demographic, zip code, and comorbidity covariates (Table 2). Adjusted outcomes were calculated on the basis of predictions which varied the access procedure for each patient while holding all other covariates fixed, averaged across the sample (12). Survival analysis was also performed (13). Repeat fistula/graft/catheter-free survival was defined by the patient surviving for 12 months after the index fistula/graft creation without the occurrence of another fistula/graft creation or tunneled catheter placement. The proportional hazard assumption was assessed by Schoenfeld residuals (14). Statistical significance was defined by P=0.05, with a probability of 0.025 in each tail.
Table 2.
Patient demographics, zip code characteristics, and comorbidities
| Patient Characteristic | Index Fistula (n=12,384) | Index Graft (n=4080) | P Value |
|---|---|---|---|
| Mean age, y | 77.1 | 78.0 | <0.001 |
| Female | 5499 (59.4) | 2424 (44.4) | <0.001 |
| Race/ethnicity | <0.001 | ||
| White/non-Hispanic | 8399 (67.8) | 2110 (51.7) | |
| Black | 2074 (16.7) | 1261 (30.9) | |
| Asian | 466 (3.8) | 219 (5.4) | |
| Hispanic | 1205 (9.7) | 421 (10.3) | |
| American Indian/Alaskan | 122 (1) | 24 (0.6) | |
| Other/unknown | 118 (1) | 45 (1.1) | |
| Index fistula/graft performed in an inpatient setting | 2451 (19.8) | 1057 (25.9) | <0.001 |
| Mean total covered billed charges in 12 mo before index fistula/graft creation date (in thousands of dollars) | 189.8 | 251.6 | <0.001 |
| Year that index fistula/graft was performed | <0.001 | ||
| 2007 | 3142 (25.4) | 1249 (30.6) | |
| 2008 | 2982 (24.1) | 973 (23.8) | |
| 2009 | 3052 (24.6) | 1001 (24.5) | |
| 2010 | 3208 (25.9) | 857 (21.0) | |
| Month that index fistula/graft was performed | 0.02 | ||
| January | 1133 (9.1) | 418 (10.2) | |
| February | 1055 (8.5) | 390 (9.6) | |
| March | 1132 (9.1) | 383 (9.4) | |
| April | 1059 (8.6) | 389 (9.5) | |
| May | 1018 (8.2) | 335 (8.2) | |
| June | 1098 (8.0) | 329 (8.1) | |
| July | 995 (8.5) | 333 (8.2) | |
| August | 1052 (7.9) | 339 (8.3) | |
| September | 973 (8.6) | 320 (7.8) | |
| October | 1060 (8.6) | 325 (8.0) | |
| November | 942 (7.6) | 263 (6.4) | |
| December | 867 (7.0) | 265 (6.3) | |
| State of beneficiary residency (range) | 15–1132 | 1–537 | <0.001 |
| Beneficiary residency (including Washington DC, Puerto Rico, and Virgin Islands) (range) | 0.1–9.1 | 0–13.2 | |
| Patient zip code characteristics | |||
| Urban | 9441 (76.2) | 3272 (80.2) | <0.001 |
| Hispanic | 1428 (11.5) | 586 (14.3) | <0.001 |
| Single | 5335 (43.1) | 1845 (45.2) | <0.001 |
| Less than high school education | 2618 (21.1) | 948 (23.2) | <0.001 |
| Employed | 7205 (58.2) | 2325 (57.0) | <0.001 |
| Poor | 1609 (13.0) | 608 (14.9) | <0.001 |
| Sensory disability among older adults | 1797 (14.5) | 597 (14.6) | 0.19 |
| Noninstitutionalized older adults with physical disability | 3651 (29.5) | 1242 (30.4) | <0.001 |
| Mental disability | 1397 (11.3) | 492 (12) | <0.001 |
| Self-care disability | 1241 (10) | 438 (10.7) | <0.001 |
| Difficulty going outside the home disability | 2610 (21.1) | 912 (22.3) | <0.001 |
| Social security income (in thousands of dollars) | 11.2 | 10.9 | <0.001 |
| Older adults in an institution | 546 (4) | 181 (4) | 0.79 |
| Median household income (in thousands of dollars) | 42.9 | 41.5 | <0.001 |
| Patient comorbidities | |||
| Hypertension | 12,258 (99) | 4044 (99.1) | 0.45 |
| Diabetes | 9068 (73.2) | 3161 (77.5) | <0.001 |
| Atrial fibrillation | 3601 (29.1) | 1216 (29.8) | 0.38 |
| Heart failure | 9705 (78.4) | 3406 (83.5) | <0.001 |
| Stroke/transient ischemic attack | 3097 (25) | 1292 (31.7) | <0.001 |
| Ischemic heart disease | 9885 (79.8) | 3384 (82.9) | <0.001 |
| Acute myocardial infarction | 1874 (15.1) | 646 (15.8) | 0.28 |
| Chronic obstructive pulmonary disease | 5395 (43.6) | 1898 (46.5) | 0.001 |
| Hyperlipidemia | 10,659 (86.1) | 3445 (84.4) | 0.01 |
| Anemia | 12,126 (97.9) | 4032 (98.8) | <0.001 |
| Endometrial cancer | 109 (0.9) | 40 (1) | 0.56 |
| Prostate cancer | 1110 (9.0) | 324 (7.9) | 0.05 |
| Lung cancer | 212 (1.7) | 63 (1.5) | 0.47 |
| Breast cancer | 419 (3.4) | 220 (5.4) | <0.001 |
| Colorectal cancer | 546 (4.4) | 236 (5.8) | <0.001 |
| Alzheimer’s disease | 548 (4.4) | 239 (7.2) | <0.001 |
| Dementia | 1956 (15.8) | 988 (24.2) | <0.001 |
| Cataract | 8354 (67.5) | 2764 (67.7) | 0.73 |
| Glaucoma | 2963 (23.9) | 1125 (27.6) | <0.001 |
| Hip/pelvic fracture | 563 (4.5) | 296 (7.3) | <0.001 |
| Depression | 3438 (27.8) | 1338 (32.8) | <0.001 |
| Osteoporosis | 1652 (13.3) | 695 (17) | <0.001 |
| Rheumatoid arthritis/osteoarthritis | 6356 (51.3) | 2237 (54.8) | <0.001 |
| Asthma | 1790 (14.5) | 679 (16.6) | 0.001 |
| Acquired hypothyroidism | 2848 (23.0) | 1093 (26.8) | <0.001 |
| Benign prostatic hyperplasia | 3392 (27.4) | 798 (19.6) | <0.001 |
Values are n (%) or as otherwise indicated.
This research was deemed exempt from review by the institutional review board because the study was retrospective, the Medicare data used was deidentified, and there was no contact with human subjects for this research.
Results
During the study period, 16,464 index fistula and graft creations met the inclusion criteria. Of these, 12,384 (75.2%) underwent index fistulas and 4080 (24.7%) underwent index grafts. The mean age in the graft group was slightly older than the fistula group (Table 2). There was a smaller proportion of women and nonwhites in the index fistula group (Table 2). Patients in the graft group had significantly higher mean total covered billed charges in the 12 months before the index fistula/graft creation date (Table 2). Patients in the graft group had a higher incidence of living in a zip code associated with overall lower sociodemographics (Table 2). Patients in the graft group had a higher incidence of all comorbidities, with the exception of no difference in hypertension, atrial fibrillation, acute myocardial infarction, endometrial cancer, lung cancer, cataract, and a higher proportion in the fistula group of hyperlipidemia, prostate cancer, and benign prostatic hyperplasia (Table 2).
There were 4719 (28.7%) patients who died in the 12 months after index fistula/graft creation. Twelve-month mortality in the fistula versus graft groups was 27.3% versus 32.7%, respectively (P<0.001). After adjusting for all covariates, the mortality difference remained significant (P=0.03), but smaller in magnitude, with 28.2% mortality in the fistula group and 29.9% in the graft group (odds ratio [OR]=0.91; 95% confidence interval, 0.84 to 0.99).
In the fistula group, 26.5% required a repeat fistula/graft creation in the 12 months after the index fistula creation versus 17.5% in the graft group (P<0.001). There was no significant difference in the proportion of patients who required a tunneled hemodialysis catheter in the 12 months after the index fistula/graft creation (fistula 28.1% versus graft 28.4%, P=0.76). A higher proportion of patients underwent a repeat fistula/graft creation and/or tunneled catheter in the index fistula group versus the index graft group (43.8% versus 35.3%, P<0.001).
After adjusting for all covariates, in the fistula group, 26.9% required a repeat fistula/graft creation in the 12 months after the index fistula creation compared with 16.7% in the graft group (P<0.001) (Figure 2). There was no significant difference in the proportion of patients who required a tunneled hemodialysis catheter in the 12 months after an index fistula creation (fistula 28.4% versus graft 27.3%, P=0.19). A higher proportion of patients underwent a repeat fistula/graft creation and/or tunneled catheter in the index fistula group versus the index graft group (44.4% versus 33.7%, P<0.001).
Figure 2.
Fistula and graft outcomes adjusted for patient characteristics and comorbidities. †Outcomes are adjusted for all demographics listed in Table 2.
The repeat fistula/graft/catheter-free survival at 12 months in the fistula group was 40.2% and 42.4% in the graft group (P=0.01). After adjusting for all covariates, the risk of loss of repeat fistula/graft/catheter-free survival was significantly higher in the fistula group with a hazard ratio of 1.19 (95% confidence interval, 1.13 to 1.24). The Cox proportional hazards estimate of the repeat fistula/graft/catheter-free survival at 12 months in the fistula group was 39.7% and 46.0% in the graft group (P<0.001) (Figure 3). The proportional hazard assumption with respect to index fistula versus graft could not be rejected (P=0.86). The results were very similar for the sensitivity analysis on the basis of the alternative definition of dialysis dependence described previously (Supplemental Appendix 1).
Figure 3.
Cox proportional hazards estimate of repeat fistula/graft/catheter-free survival.
Discussion
Older adults are the fastest growing segment of the ESRD population. The rate of prevalent ESRD for patients ages 65–74 years has increased 31% since 2000 and 48% for those age ≥75 years, compared with 20% for those ages 45–64 years (5). Recently, evidence has emerged suggesting that fistulas may not provide the same benefits over grafts in the elderly population compared with the younger population.
National Kidney Foundation Dialysis Outcomes Quality Initiative Guidelines state that fistula have the “lowest rate of thrombosis…providing longer survival of the access.” (2) This statement is based on two references: (1) a 2003 review article in which age was not a consideration and (2) a single institution study of 100 fistula and 109 graft patients with mean ages in the mid- to late 50s (15,16). More recently, the Dialysis Morbidity and Mortality Study demonstrated no patency advantage of fistula over graft in older adults (17). Likewise, a 2007 meta-analysis found that the 12-month primary and secondary patency rates of fistulas were lower in older adults (4).
Another major purported advantage of fistulas over grafts is resistance to infection (2). This benefit, however, can be realized only if the fistula matures such that it can be used for dialysis. Evidence suggests that fistula maturation is worse in older adults. A study of patients age ≥65 years demonstrated a 29% maturation failure rate with an increased OR (1.7) for maturation failure in the ≥65 year group versus the <65 year group (3). Similarly, the aforementioned meta-analysis found that elderly patients had a significantly higher rate of radial-cephalic fistula primary failure (OR=1.8) (4).
When a permanent access fails, the difficulty in creating a subsequent access is potentially increased because of the reduction in available sites for access creation. In an effort to limit our study population to patients undergoing first-time vascular access creation, we included only patients who were ≥66 years of age who qualified for Medicare exclusively by virtue of age. No patients included in the analysis initially enrolled in Medicare with a diagnosis of ESRD, and all were enrolled for at least 12 months before the creation of the index fistula/graft. The selection criteria presumably excluded patients who had a permanent access created before Medicare enrollment. The selection criteria also selected for patients who had the highest survival probabilities with regard to ESRD because mortality increases with the length of dialysis dependence (5).
Nevertheless, our results demonstrated high 12-month mortality in both the fistula and graft groups with a small 1.7% adjusted advantage in the fistula group. This is consistent with a recent decision and cost-utility analysis which demonstrated that fistula use was associated with a greater overall survival of 2.6 months and 3.6 quality-adjusted survival months (18). Although fistula was found to be superior to graft with regard to survival in the cost-utility analysis and our analysis, the differences were very modest in both studies and may not be clinically significant. Longer-term findings from a US Renal Data System cohort demonstrated that patients aged 67–80 years experienced improved survival up to 50 months from the time of initiation of hemodialysis when dialyzing through a fistula compared with a graft. However, this survival advantage was not observed in octogenarians and nonagenarians (19). Results from a smaller cohort of the Dialysis Morbidity and Mortality Study demonstrated no survival benefit of fistula over graft in those aged ≥65 years (17). Similarly, in a decision analysis, men with diabetes had a modest survival benefit at younger ages with a fistula first strategy, but this benefit progressively diminished with increasing age. For women with diabetes, there was no significant survival difference at all ages between fistula and graft (20).
In our analysis, the fistula group had a nearly 10% higher failure rate in the first 12 months postoperatively compared with the graft group, when failure is defined as need for repeat fistula/graft creation or catheter placement. Vascular access failure is of much greater consequence in dialysis-dependent patients compared with predialysis patients. If a patient is predialysis, the patient will not require a tunneled catheter, and the practitioner may feel less urgency to create a functioning fistula/graft after the index fistula/graft fails. Accordingly, we limited the study population to patients who were dialysis dependent at the time of index fistula/graft creation as other authors have done when comparing outcomes of fistulas and grafts (21–26). Doing so helped to equalize the baseline risk of mortality associated with dialysis dependence and the baseline risk of requiring a repeat fistula/graft or catheter. Despite a relatively favorable health status in terms of comorbidities and higher sociodemographics within the fistula group, patients with an index fistula required more repeat fistula/graft creation. These results imply that, at least in the short term, fistulas may not be superior to grafts in reliability for hemodialysis access in dialysis-dependent older adults.
We are unable to discern from claims data whether functional failure is caused by failure to mature or poor patency. Regardless, the ultimate result is that patients with failed fistulas are unable to benefit from the advantages of fistulas over grafts. Furthermore, these results indicate that fistulas in older adults are associated with a higher number of access-related health care encounters compared with grafts, which effect quality of life and health care costs. These consequences are particularly relevant to the elderly ESRD population with poor 1-year life expectancy.
The optimal outcome for the patient after a fistula or graft operation is to survive without the need for an additional fistula/graft operation or tunneled catheter placement. To evaluate this, we assessed repeat fistula/graft/catheter-free survival. At 12 months, repeat fistula/graft/catheter-free survival was dismal for both the fistula and graft group but was 6.3% higher in the graft group. These sobering results further emphasize the need for guidelines that address optimal access type on the basis of individual patient characteristics versus a disease-based approach. Dialysis access recommendations need to be developed which take into account patient characteristics, including age and life expectancy.
The Centers for Medicare and Medicaid Services recently adopted the rates of hemodialysis access by fistula and long-term catheter as distinct quality metrics for payment penalties to outpatient dialysis facilities (27). Facilities receive scores on the basis of where they fall in the percentile of facilities for each measure, without any risk adjustment. The scores will be linked to payment reductions of up to 2%. Our analysis raises the possibility that dialysis facilities with higher proportions of elderly patients could be punished for apparently inferior performance simply because of the demographic composition of their patient population.
Limitations
This is an observational study. Although we adjusted for all readily available patient characteristics measured in claims data, it is possible that poorer vascular access reliability among patients who initially received a fistula was the result of other factors. Similarly, it is possible that the improved mortality associated with index fistula is related to unmeasured factors. Such a concern could arise from the lower mean total covered billed charges in the 12 months before index fistula/graft creation in the fistula group. This difference may indicate that patients in the fistula group have a lower medical acuity. Nevertheless, by including only individuals who were dialysis dependent at the time of index fistula/graft creation, we reduced heterogeneity in baseline risk of mortality and urgency of need for a functional vascular access. Future investigations should use a technique such as an instrumental variables approach to address unmeasured confounding.
We did not compare outcomes in the elderly population with a younger population. The major reason for this is that the life expectancy of the younger ESRD population is longer, and long-term outcomes may be more pertinent in the younger ESRD population than short-term outcomes.
Additionally, we did not address other access-related complications, such as infection and revision. Both would be important outcomes for future studies. We elected not to include access revisions because the descriptions of CPT codes for revision procedures are not as clear as the codes for fistula/graft creation. As such, the incidence of revisions is likely to be less accurate than the incidence of repeat fistula/graft creation. The total burden of vascular access-related health care encounters in older adults is likely greatly underestimated because of the exclusion of access revision procedures and access-related infection from the analysis. We also did not measure the number of days that patients who had catheters placed after fistula or graft creation were catheter dependent. This would be an important outcome to address in future studies. This analysis of older individuals addressed Medicare only. Future studies should address the implications of supplemental coverage on outcomes after vascular access creation.
Our analysis of Medicare claims data provides contemporary evidence in a large sample that suggests fistulas may be associated with a slightly lower mortality in the first 12 months after fistula creation compared with grafts among dialysis-dependent older adults. However, the reliability of fistulas for hemodialysis is significantly worse than grafts in the first 12 months. The small survival advantage of fistula must be balanced with the poor reliability of fistula in determining the most appropriate type of dialysis vascular access. Particularly in older adults, the choice of vascular access type should be individualized to each patient, taking into account comorbidities, demographics, and life expectancy. Because of the poor life expectancy of elderly patients with ESRD, short-term outcomes of dialysis access in this population merit careful consideration when developing guidelines and determining health policy. Further research is required to determine the effect of a high rate of early fistula failure in older adults on long-term morbidity, mortality, quality of life, and health care costs.
Disclosures
Dana Goldman is a partner at Precision Health Economics, a company providing consulting services to the life-sciences industry. John Romley has provided consulting services to Precision Health Economics. None of these activities has had any influence on the results or interpretation in this article. Karen Woo declares no conflicts of interest.
Supplementary Material
Acknowledgments
We thank the following colleagues for carefully reading the manuscript and providing valuable comments: Sanjay Arora, Steven Fox, Geoffrey Joyce, Ashwini Lakshmanan, Michael Menchine, Seth Seabury, Sophie Terp, and Julie Zissimopoulos.
This research was supported by the National Institutes of Health (USC-RCMAR Grant P30-AG043073).
Footnotes
Published online ahead of print. Publication date available at www.cjasn.org.
This article contains supplemental material online at http://cjasn.asnjournals.org/lookup/suppl/doi:10.2215/CJN.09040914/-/DCSupplemental.
References
- 1.Neumann ME: “Fistula first” initiative pushes for new standards in access care. Nephrol News Issues 18: 43–, 47–48., 2004 [PubMed] [Google Scholar]
- 2.III. NKF-K/DOQI Clinical Practice Guidelines for Vascular Access: Update 2006. Am J Kidney Dis 37[1 Suppl 1]: S137–S181, 2001 [DOI] [PubMed] [Google Scholar]
- 3.Lok CE, Oliver MJ, Su J, Bhola C, Hannigan N, Jassal SV: Arteriovenous fistula outcomes in the era of the elderly dialysis population. Kidney Int 67: 2462–2469, 2005 [DOI] [PubMed] [Google Scholar]
- 4.Lazarides MK, Georgiadis GS, Antoniou GA, Staramos DN: A meta-analysis of dialysis access outcome in elderly patients. J Vasc Surg 45: 420–426, 2007 [DOI] [PubMed] [Google Scholar]
- 5.US Renal Data System : USRDS 2013 Annual Data Report: Atlas of Chronic Kidney Disease and End-Stage Renal Disease in the United States, National Institutes of health, Bethesda, MD, National Institute of Diabetes and Digestive and Kidney Diseases, 2013 [Google Scholar]
- 6.Chronic Conditions Data Warehouse: Condition Categories. Available at: https://www.ccwdata.org/web/guest/condition-categories. Accessed April 6, 2015
- 7.Centers for Medicare and Medicaid Services: Berenson-Eggers Type of Service (BETOS). Available at: http://www.cms.gov/Medicare/Coding/HCPCSReleaseCodeSets/BETOS.html. Accessed April 6, 2015
- 8.Clement FM, James MT, Chin R, Klarenbach SW, Manns BJ, Quinn RR, Ravani P, Tonelli M, Hemmelgarn BR, Alberta Kidney Disease Network : Validation of a case definition to define chronic dialysis using outpatient administrative data. BMC Med Res Methodol 11: 25–31, 2011 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Cappell KA, Shreay S, Cao Z, Varker HV, Paoli CJ, Gitlin M: Red blood cell (RBC) transfusion rates among US chronic dialysis patients during changes to Medicare end-stage renal disease (ESRD) reimbursement systems and erythropoiesis stimulating agent (ESA) labels. BMC Nephrol 15: 116–127, 2014 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Fisher ES, Wennberg DE, Stukel TA, Gottlieb DJ, Lucas FL, Pinder EL: The implications of regional variations in Medicare spending. Part 2: Health outcomes and satisfaction with care. Ann Intern Med 138: 288–298, 2003 [DOI] [PubMed] [Google Scholar]
- 11.Romley JA, Jena AB, Goldman DP: Hospital spending and inpatient mortality: Evidence from California: An observational study. Ann Intern Med 154: 160–167, 2011 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Graubard BI, Korn EL: Predictive margins with survey data. Biometrics 55: 652–659, 1999 [DOI] [PubMed] [Google Scholar]
- 13.Cox DR: Regression models and life-tables (with discussion). J R Stat Soc, B 34: 187–220, 1972 [Google Scholar]
- 14.Grambsch PM, Therneau TM: Proportional hazards tests and diagnostics based on weighted residuals. Biometrika 81: 515–526, 1994 [Google Scholar]
- 15.Huber TS, Carter JW, Carter RL, Seeger JM: Patency of autogenous and polytetrafluoroethylene upper extremity arteriovenous hemodialysis accesses: A systematic review. J Vasc Surg 38: 1005–1011, 2003 [DOI] [PubMed] [Google Scholar]
- 16.Perera GB, Mueller MP, Kubaska SM, Wilson SE, Lawrence PF, Fujitani RM: Superiority of autogenous arteriovenous hemodialysis access: Maintenance of function with fewer secondary interventions. Ann Vasc Surg 18: 66–73, 2004 [DOI] [PubMed] [Google Scholar]
- 17.Chan MR, Sanchez RJ, Young HN, Yevzlin AS: Vascular access outcomes in the elderly hemodialysis population: A USRDS study. Semin Dial 20: 606–610, 2007 [DOI] [PubMed] [Google Scholar]
- 18.Xue H, Lacson E, Jr, Wang W, Curhan GC, Brunelli SM: Choice of vascular access among incident hemodialysis patients: A decision and cost-utility analysis. Clin J Am Soc Nephrol 5: 2289–2296, 2010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.DeSilva RN, Patibandla BK, Vin Y, Narra A, Chawla V, Brown RS, Goldfarb-Rumyantzev AS: Fistula first is not always the best strategy for the elderly. J Am Soc Nephrol 24: 1297–1304, 2013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Drew DA, Lok CE, Cohen JT, Wagner M, Tangri N, Weiner DE: Vascular access choice in incident hemodialysis patients: A decision analysis. J Am Soc Nephrol 26: 183–191, 2015 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Lok CE, Sontrop JM, Tomlinson G, Rajan D, Cattral M, Oreopoulos G, Harris J, Moist L: Cumulative patency of contemporary fistulas versus grafts (2000-2010). Clin J Am Soc Nephrol 8: 810–818, 2013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Oliver MJ, McCann RL, Indridason OS, Butterly DW, Schwab SJ: Comparison of transposed brachiobasilic fistulas to upper arm grafts and brachiocephalic fistulas. Kidney Int 60: 1532–1539, 2001 [DOI] [PubMed] [Google Scholar]
- 23.Fitzgerald JT, Schanzer A, McVicar JP, Chin AI, Perez RV, Troppmann C: Upper arm arteriovenous fistula versus forearm looped arteriovenous graft for hemodialysis access: A comparative analysis. Ann Vasc Surg 19: 843–850, 2005 [DOI] [PubMed] [Google Scholar]
- 24.Gibson KD, Gillen DL, Caps MT, Kohler TR, Sherrard DJ, Stehman-Breen CO: Vascular access survival and incidence of revisions: A comparison of prosthetic grafts, simple autogenous fistulas, and venous transposition fistulas from the United States Renal Data System Dialysis Morbidity and Mortality Study. J Vasc Surg 34: 694–700, 2001 [DOI] [PubMed] [Google Scholar]
- 25.Rooijens PP, Burgmans JP, Yo TI, Hop WC, de Smet AA, van den Dorpel MA, Fritschy WM, de Groot HG, Burger H, Tordoir JH: Autogenous radial-cephalic or prosthetic brachial-antecubital forearm loop AVF in patients with compromised vessels? A randomized, multicenter study of the patency of primary hemodialysis access. J Vasc Surg 42: 481–486; discussions 487, 2005 [DOI] [PubMed] [Google Scholar]
- 26.Woods JD, Turenne MN, Strawderman RL, Young EW, Hirth RA, Port FK, Held PJ: Vascular access survival among incident hemodialysis patients in the United States. Am J Kidney Dis 30: 50–57, 1997 [DOI] [PubMed] [Google Scholar]
- 27.Fishbane S, Miller I, Wagner JD, Masani NN: Changes to the end-stage renal disease quality incentive program. Kidney Int 81: 1167–1171, 2012 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.