Prevalence of arteriovenous fistulas in incident hemodialysis patients: correlation with patient factors that may be associated with maturation failure.

BACKGROUND

Lok et al previously reported a risk equation for arteriovenous fistula (AVF) maturation failure. It is unclear whether this model or a more comprehensive model correlates with incident AVF use in the US hemodialysis population.

STUDY DESIGN

Cross-sectional study.

SETTING & PARTICIPANTS: 195,756 adult patients initiating outpatient hemodialysis therapy in the United States between July 1, 2005, and December 31, 2009, with 6 months or more prior nephrology care.

PREDICTOR

Patient characteristics (age, peripheral vascular disease, coronary artery disease, and race) populating the AVF maturation failure risk equation and other demographic and clinical variables from the Centers for Medicare & Medicaid Services (CMS) Medical Evidence Report (CMS 2728).

OUTCOMES & MEASUREMENTS: AVF use at first outpatient dialysis treatment as recorded on the CMS 2728.

RESULTS

Using the risk categories defined by Lok et al, AVF use varied from 19.0% (very high risk) to 25.6% (low risk). In a model using only these risk categories, logistic regression showed lower ORs for moderate-, 0.90 (95% CI, 0.88-0.93); high-, 0.80 (95% CI, 0.78-0.83); and very high-risk patients, 0.68 (95% CI, 0.63-0.73) compared with low risk. In the expanded model, odds were lower for women, blacks, Hispanics, age older than 85 years, diabetes, peripheral vascular disease, congestive heart failure, other cardiac disease, and underweight. Odds were higher for hypertension, overweight, obesity, 12 months or more nephrologist care, most insurance types, and each successive year after 2005. Despite associations, the C statistic for the expanded model was 0.64.

LIMITATIONS

This analysis is limited by lack of access creation history before dialysis therapy initiation and minimal external validation of CMS 2728 data.

CONCLUSIONS

Clinical risk factors identified by Lok and expanded in this analysis have limited ability to predict incident AVF use. Even patients judged at highest risk can have successful AVF construction and initiate dialysis therapy through a functioning AVF.