#4606 innovative method to evaluate maturation of newly created arterio-venous fistulas: insights from a quality improvement project

Abstract Background and Aims Recently, we introduced central-venous oxygen saturation (ScvO2) and estimated upper-body blood flow (eUBBF) to monitor the maturation of newly created arterio-venous fistulas (AVF) in hemodialysis (HD) patients [1]. The approach was implemented in a clinical quality improvement project (QIP). Method The QIP involved two Renal Research Institute dialysis centers in New York City that use the Crit-Line monitor (CLM, Fresenius Medical Care, Waltham, MA, USA) as part of the standard of care. CLM measures automatically and non-invasively hematocrit and oxygen saturation (i.e., ScvO2 in patients with central venous catheter as vascular access). eUBBF was computed as described previously [1]. Patients with a with a newly created AVF were included, irrespective of their HD vintage. The QIP team comprised a registered nurse per clinic, two research physicians, and two data analysts. The team met weekly to assess the AVF maturation progress based on eUBBF and ScvO2 data that were displayed on a specific dashboard. Successful AVF maturation is typically accompanied by a rise in cardiac output and eUBBF that starts immediately after AVF creation. However, it is unknown what degree of eUBBF increase indicates a subsequently successful AVF maturation. To identify predictive eUBBF thresholds, we correlated eUBBF changes with the progress of AVF maturation in the six weeks following AVF creation. We defined maturation as unsuccessful when the newly created AVF required an intervention (e.g., angioplasty) or was abandoned. Results We studied 39 patients (age 56 ±18 years; 25 males). Twenty-eight of them were incident patients. Two patients had a second AVF created. In 15 patients AVF maturation was unsuccessful. In these patients, eUBBF increased to an average of 119.2% (SD 35.2) compared to pre-AFV creation. In patients with successful AFV maturation (N = 24), eUBBF increased to an average of 153.7% (SD 31.6) (Figure 1). The difference between the groups is 34.4 percentage points (95% CI: 12.4 to 56.4; p = 0.003). Using eUBBF as a diagnostic measure to discriminate successful vs. failed AVF maturation, an area under the receiver operating characteristics curve of 0.824 (95% CI: 0.673 to 0.925; p<0.0001) was obtained. An eUBBF increase of 25% had a sensitivity of 59% and a specificity of 92% to predict AVF maturation outcomes. Conclusion Computation of eUBBF provides a simple and non-invasive means to track AVF maturation and predict its outcomes.