I read with interest the article by Jones et al. (1) in the March 10, 2015, issue of the Journal. I am concerned that, based on the methodology and data presented, the conclusions may be misleading and suggest that interventionalists are making treatment decisions for peripheral vascular interventions (PVIs) and the choice of atherectomy, particularly, based on financial remuneration rather than on scientific evidence, depth of clinical experience, and interest in improving outcomes for patients.

The article presented data for Medicare fee-for- service beneficiaries between 2006 and 2011, showing no statistical increase in overall rate of PVIs, but with significant shifts in site of services from inpatient hospitalization to outpatient and office settings. Additionally, increases in atherectomy procedures during that time period were presented as two-fold in the hospital outpatient setting and 50-fold in the office setting. The authors concluded that changes in reimbursement intended to result in cost savings to Medicare inadvertently drove the shift in PVI site of service and the increase in atherectomy procedures in outpatient and office settings, thereby neutralizing cost savings. However, several points of clarification are necessary for accurate interpretation of the presented data.

First, the article presents an increase in atherectomy procedures up to 50-fold in the office setting during the study period. However, unlike percutaneous trans- luminal angioplasty (PTA) and stenting, atherectomy was essentially nonexistent in the office setting until 2011, when it first became eligible for reimbursement for similar Medicare beneficiaries as the study population. Ignoring this development and reporting a trend over the entire time period of 2006 to 2011 is misleading.

Moreover, Table 1 shows that from 2006 to 2011, the majority of atherectomies (95.7%) were per- formed in the hospital setting versus the office (4.2%). By comparing absolute numbers of atherectomy procedures in the office setting in 2010 (n 1⁄4 w0 pro- cedures) versus 2011 (n 1⁄4 292 procedures) (see Table 3 in Jones et al. [1]), this relatively small absolute in- crease becomes overmagnified when quantified as a percentage. Atherectomies performed in the office in 2011 represented only 16% of the total number of atherectomies performed for that year in all 3 settings. Second, mean costs of atherectomy procedures are presented as exceeding those of stenting and PTA procedures during the study period. However, in the hospital outpatient setting, atherectomy was reimbursed at the same level as PTA until 2008, and less than or equal to stent procedures since 2008 (Table 1). In addition, the utilization and costs of atherectomy are likely overestimated relative to those of stenting as the PVIs in the article were categorized as angio- plasty, stenting, or atherectomy, without regard to procedures involving more than one treatment modality, and the costs of procedures using both atherectomy and stenting were only attributed to
atherectomy in the analysis.

Third, the analysis excluded patients undergoing expensive surgical or hybrid revascularization procedures (n 1⁄4 8,901 [20.6%]) from the entire pool of 39,339 patients who underwent revascularization. Therefore, the conclusion about the erosion of savings due to shift in site of service and outpatient reimbursements is based on incomplete information that ignores the substantial reduction in the rate of surgical bypass procedures (33%; p < 0.001) during the study period. Considering that lower extremity bypass surgery is an expensive, inpatient-only procedure typically requiring a 3.8- to 10-day length of stay (LOS) and with Medicare costs ranging from $17,215 to $28,983 per bypass procedure (FY 2011 rates), the significant reduction in surgical procedures which likely resulted in significant cost savings to Medicare is not represented fairly in this analysis. (Medical provider analysis and review data for fiscal years [FY] 2009–2012, and FYs 2011–2014 Final Rules also are available at the Centers for Medi- care and Medicaid Services website.)

Fourth, the patients treated with atherectomy in the study represented a sicker population than those who received stents. Atherectomy patients were more likely to be older black males (p < 0.002) with diabetes, hypertension, renal failure, ischemic heart disease, heart failure, and stroke, compared to patients who underwent stenting (p < 0.001). These comorbidities are associated with critical limb ischemia (CLI) and amputation as well as longer, complex, and more heavily calcified lesions in the infrapopliteal arteries, which may result in less effective treatment with balloons and stents alone (2). The differences in patient populations could also indicate an improved access to care (i.e., patients were treated who would otherwise have been treated surgically or who would not have undergone revascularization at all). These clinical considerations in treatment
choices were not addressed in the authors’ conclusions, nor did the authors assess the patient outcomes.

Fifth, the authors cited lack of efficacy and comparative evidence to justify the increased use of atherectomy. However, during the study period, paucity of data was true for the entire field of peripheral artery disease (PAD) treatment. Often, practice patterns change in advance of published data, based on clinical experience to support therapeutic decision making. In the last several years, numerous clinical device trials in PAD patients have reached completion, and evidence has been published.

In the Zilver PTX trial, data clearly demonstrate that PTA alone in the femoropopliteal artery segment is not a satisfactory procedure due to high restenosis rates, resulting in a poor 12-month primary patency rate as low as 32.8%. Similar poor results were seen in the clinically driven target lesion revascularization (TLR) rate at 12 months post-PTA of 17.5% (3). Although some bare-metal and drug-eluting stents have improved patency and TLR rates significantly (3), concerns remain regarding the high cost of treating in-stent restenosis (ISR).

Atherectomy (plaque removal) is a therapeutic mo- dality that preserves the native vessel for future treatment options without leaving a permanent implant (metal stent) behind. In addition, atherectomy reduces the need for costly revision of in-stent reste- nosis, avoids or reduces the amount of barotrauma to the vessel, and has a lower dissection rate than PTA, while reducing the rates of target lesion revascularization. One of the largest, prospective, multicenter studies, the DEFINITIVE LE study (Determination of EFfectiveness of the SilverHawk PerIpheral Plaque ExcisioN System [SIlverHawk Device] for the Treat- ment of Infrainguinal VEssels/Lower Extremities), included 800 patients and confirmed the fact that directional atherectomy is safe and effective (78% overall primary patency, 95% prevention of major amputation in CLI patients) in a variety of lesions, in patients with and without diabetes, and in claudicants and CLI patients (4). Additionally, the EXCITE ISR (EXCImer Laser Randomized Controlled Study for Treatment of FemoropopliTEal In-Stent Restenosis) trial demonstrated a significant reduction in TLR that favored laser atherectomy over PTA for femoropopliteal ISR in a randomized, controlled trial with long lesions (5). The ongoing large LIBERTY 360 trial is attempting to demonstrate real-world comparative outcomes for multiple atherectomy, PTA, and stent modalities in diverse lesion subsets involving the popliteal and infrapopliteal arteries.

Devices and techniques for revascularization have evolved rapidly since 2006. Several different atherectomy devices, including directional, orbital, rotational, excisional, and laser atherectomy devices, have been approved in the United States since 1998. However, some devices only gained significant adoption in recent years due to technological evolution, clinical evidence, and improved understanding of endovascular techniques. Of note, the article sug- gests that improved outpatient reimbursement led to growth in use of atherectomy in the outpatient setting (1); in fact, inpatient atherectomy procedures increased 30.3% from 2010 to 2011 (see Table 3 in Jones et al. [1]), whereas outpatient atherectomy increased by 20.4%, potentially indicating that other (nonfinancial) factors may have influenced the in- crease in atherectomy use. Although significant challenges exist in designing and funding random- ized comparative effectiveness trials for all patient and lesion subsets, we continue to gain new information regarding performance and durability of procedures.

Although there are economic drivers that dictate care in all locations of service, outliers exist in every specialty and site of service. It is believed that most physicians strive to do what is in the best interest of their patients. There is concern with singling out a specific therapy from a retrospective dataset without consideration of all the inherent episodic costs, especially when excluding data representing probable significant savings to Medicare by reduction in more costly inpatient surgical services. In the interests of the patients who present with this challenging dis- ease state, the interventionalists who are thought- fully trying to care for them, and the payers, such as Centers for Medicare and Medicaid Services, who are interested in reducing overall costs, additional financial analyses and assessments of true costs are critical for accurate and informed decision making. Only then can conclusions be made about value for specific therapies to treat PAD in each site of service, with the overall aim of improving patient access to quality care.

*Jeffrey G. Carr, MD
*Tyler Cardiac and Endovascular Center