NN&I - July 2010
National Subscribe to our free eNewsletter at www.nephronline.comJuly 2010 Nephrology News & Issues 11 LettersIn the article, "Is there a need to mandate access surveil-lance in the dialysis clinic?" (June 2010, NN&I), one of us (A. Besarab)1 signed on with 16 others against the mandate in the Conditions for Coverage to provide vascular access "surveillance." CMS has mandated that monitoring and surveillance be performed on all permanent non-catheter accesses.2 The signers believed the medical evidence for the mandate for surveillance was premature. Although many practitioners use the two terms of moni-toring and surveillance interchangeably, they should not. The goal of both monitoring and surveillance is the detec-tion of an anatomically severe stenosis that is physiologi-cally significant and likely to thrombose over time. The detection of a stenotic lesion, however, is only the first step in the diagnostic algorithm and therapeutic process for access care. It was clearly the intent of the authors of the NN&I article1 to state that the mandate for surveillance was not sufficiently evidence based. The article cited a number of studies and position papers arguing that surveillance, fol-lowed by angioplasty if stenosis is detected, does not alter the primary outcome of longevity or the secondary outcome of thrombosis-rate reduction of the access.3-7 All of these studies, however, can be criticized on many levels, including small sample size, lack of data on important co-unfounders (control of which is impossible with sample sizes under 400 patients), and whether the intervention produced the desired improvement in physiologic function. Success of angioplasty only occurs if the procedure restores both the anatomical and clinical/physiological parameters used to detect the stenosis in the first place into acceptable limits. In a properly conducted randomized controlled trial, defining exactly what constitutes a successful angioplasty is crucial. Dilatation of a stenotic lesion with an angioplasty balloon must also have a lasting beneficial effect. Without satisfying these two components, the angioplasty may actu-ally cause harm, as stated by the authors of the article. So does surveillance have a benefit? Not mentioned by the authors were valuable and often conflicting observa-tions obtained within the same studies they cite4,6 and important follow up reports using data from the same cohorts.8,9 Also not mentioned were studies demonstrat -ing a benefit of duplex ultrasound surveillance10 and access blood flow11 on access patency. In 1993, Harold Feldman, MD, using data A. Besarab pro-vided,12 calculated a study cohort size of 572 patients would be necessary to demonstrate an increase in average "in use access age" from two to three years. This said, we do not believe the outcome in access care should focus on prolon-gation of access survival. The more important outcome for us is a reduction in thrombosis rate. Thrombosis has effects on many levels. Unfortunately, as with access survival cohort size, statistical estimates by Dr Edward Peterson, head of biostatistics at Henry Ford Health System (using two different designs/models) indicated that > 500 patients would be needed to demonstrate a 50% reduction in thrombosis rate based on recently published data.13One of us (A. Besarab) was in the final phases of design-ing a large randomized trial with Dialysis Clinics Inc. to compare the value of derived intra-access static pressure surveillance combined with "clinical monitoring" versus clinical monitoring alone as practiced at the facility level. The cohort size needed to demonstrate a 50% reduction in the thrombosis rate was 600 patients. DCI determined that none of their units would agree to a study that would include a "control" group because of the Conditions for Coverage mandate requiring surveillance. These new con-ditions by CMS had the effect of precluding definitive scien-tific investigation in this area. We believe that "evidence-based medicine is the integra-tion of best research evidence with clinical expertise and patient values," as formulated by Sackett in 1996.14 To do what is best for the patient, all three are important; one alone is not sufficient. Clinical judgement is needed to interpret surveillance data. Not all stenoses are progressive and not all lesions need to be treated. We also believe that good physical exams and good temporal trending of clini -cal problems [monitoring] works for arteriovenous fistulae, less well in grafts. When done properly by well trained dialy-sis staff, as was done in many of the studies cited, 3-7 it would be difficult to show a benefit to conduct surveillance as well. In the real world, however, the average nurse/patient care technician and nephrologist have limited technical knowl -edge about access anatomy and function. We believe in surveillance because with modern staffing patterns, consistent access evaluations are not performed. As a result, one obtains a low positive predictive value of "clinical monitoring" as practiced in the typical dialysis unit. We believe that the confluence of patient issues, best medi-cal evidence, and clinical expertise mandate the continued ability to study the issue scientifically. Thrombosis is a major event; it needs to be avoided. Surveillance applied to a population results in decreased adverse events, and we believe it has significant value because at a minimum, it fosters the ability to salvage National_NNI0710_7.indd 11 6/16/10 2:34:06 PM
You must have JavaScript enabled to view digital editions.
