Article Text

Download PDFPDF

Practice makes perfect: establishing reasonable minimum thrombectomy volume requirements for stroke centers
  1. Kyle M Fargen1,
  2. David J Fiorella2,
  3. J Mocco3
  1. 1 Department of Neurological Surgery, Wake Forest University, Winston-Salem, North Carolina, USA
  2. 2 Department of Neurosurgery, Stony Brook University, Stony Brook, New York, USA
  3. 3 Department of Neurosurgery, Mount Sinai Hospital, New York, USA
  1. Correspondence to Dr Kyle M Fargen, Department of Neurological Surgery, Wake ForestUniversity, Winston-Salem, North Carolina 27109, USA; kfargen{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Recently, seven randomized controlled trials were published showing benefit to neurothrombectomy (NT) in the setting of anterior circulation emergent large vessel occlusion (ELVO).1–7 Accordingly, the American Heart Association/American Stroke Association has updated societal guidelines and provided a class 1, level of evidence A recommendation that NT be performed in patients with an ELVO who meet appropriate criteria.8 Subsequently, a recent national survey of neurointerventional physicians has suggested increasing procedural volumes, with many reporting increasing indications for patient selection since the publication of these trials.9 Further, studies have suggested that approximately 10,000 NT procedures were performed in the USA during the year 2015, with significant potential for increased volume in the future.10 11

The Joint Commission has recently proposed a new certification for hospitals entitled Thrombectomy-capable Stroke Center (TSC) Advanced Certification, intended as an intermediate classification between primary and comprehensive stroke centers.12 Most notably, TSC certification mandates the capacity to perform NT 24 hours a day, 7 days a week, the capability for other types of advanced imaging, and a written agreement for transfer with at least one comprehensive stroke center. The Joint Commission has requested comment on minimum NT volume requirements. Currently, the Joint Commission has suggested TSC certification eligibility requires that an institution has performed a minimum of 12 NTs during the previous year, or 24 NT over the past 2 years.12 A typical TSC would requires at least two to three experienced operators to provide the service 24/7. Thus, the present guidelines could translate to as few as four cases per year for each operator at some TSCs. A center performing such a low volume of NTs is highly unlikely to be performing any significant volume of other types of neuroendovascular procedures and, therefore, one would expect that the overall level of proficiency would be quite low. Thus, the criterion of 12 NTs per year represents such a low bar for proficiency in endovascular surgery that it clearly fails the ‘red face test.’

But what is a reasonable number? As systems of care evolve and our medical community evaluates how to provide access to this important evidence-based therapy for every eligible patient, the question of what constitutes an appropriate procedural volume is clearly a very important one. Herein, we review the current available literature on the topic, as well as relevant related literature, in order to provide an evidence-based estimation of an appropriate volume minimum.

Cerebrovascular procedural volume and outcomes

The available literature suggests a powerful direct correlation between endovascular volume and outcomes, both for individual practitioners and for institutions.13–16 Procedural morbidity and complications have been shown to be lower in patients undergoing endovascular or open cerebrovascular surgery at high-volume centers.17–25 Two studies have a demonstrated a significant correlation between aneurysm coiling case volumes and better patient outcomes.19 25 Similar volume–outcome relationships have been reported with carotid stenting. One study showed a 1.9 greater odds of 30-day mortality in patients treated by low-volume operators.23 A second nationwide inpatient sample database study demonstrated a lower rate of periprocedural stroke for every provider 10-unit carotid stent volume increase.24

Neurothrombectomy volume and outcomes

It is no surprise that lower-volume stroke centers report lower rates of good outcomes than higher-volume centers.26–28 However, the age of ‘modern neurothrombectomy’ essentially dates back only a few years (circa 2012) to the commercial availability of the stent retriever and larger-bore intracranial aspiration thrombectomy catheters. Thus there are limited referable data to document an association between operator or hospital NT volumes and patient outcomes from this time period. Studies predating this era are not as valid given the far less robust NT treatment effect documented in a series of randomized controlled trials reported before 2015 using older, inferior devices and techniques.

Despite this limitation, the available data do support an association between NT case volume and the likelihood of good patient outcomes. Two nationwide inpatient sample database surveys have provided evidence of a relationship between higher NT volumes and both reduced mortality and increased good clinical outcome rates.29 30 A multicenter retrospective study comparing high-volume (≥50 or greater NT procedures annually) and low-volume (<50 NTs annually) centers demonstrated shorter times from imaging to groin puncture, shorter total procedural times, better rates of reperfusion, and significantly better angiographic and clinical outcomes at high-volume centers.28

Most recently, a Vizient (University HealthSystem Consortium) database study by Rinaldo et al of over 8500 NT patients admitted to 118 institutions during 2012–2016 showed significant indirect associations between mortality and volume for both directly admitted and transferred patients.31 Using regression tree analysis, the authors categorized institutional NT volume into low, medium, and high strata based on natural thresholds in observed mortality index. The threshold for high-volume centers was calculated at ≥35.2 NT procedures per year. When these criteria were used, mortality rates were significantly lower for patients admitted to high-volume centers (9.8%) than for medium- (14.9%) or low- (19.7%) volume centers. Mortality was significantly lower for patients transferred to high-volume centers for NT than for those directly admitted to low-volume centers, suggesting that the benefit of treatment at a high-volume center may outweigh any detrimental consequences related to transfer delays.31

Evidence to support NT was derived from trials performed exclusively at high-volume centers

The series of randomized controlled trials that demonstrated a benefit for NT over medical management were conducted at high-volume stroke centers. This fact was underscored in an editorial written by several of the principal investigators of these positive NT trials, entitled ‘Neurothrombectomy trial results: stroke systems, not just devices, make the difference’. In this editorial, the authors emphasized that the NT randomized controlled trial results were generalizable only to other established, efficient stroke systems staffed by highly experienced neurointerventionalists.32 Thus, it is not at all valid for the medical community to expect that the same robust NT treatment effect will be observed in low-volume centers with less experienced practitioners. This question of generalizability needs further investigation, but until it is resolved, there are no data which would support directing patients with potential ELVO to very low-volume NT centers. By allowing low-volume centers to proliferate, rather than supporting high-volume regionalized systems of care, we run a very real risk of undermining the benefit we seek to provide.

Percutaneous coronary intervention as a corollary

There are plentiful data demonstrating an association of higher mortality rates, lengths of stay, and major adverse cardiac events with lower operator percutaneous coronary intervention (PCI) volumes.33 34 Initial US guidelines for minimum operator PCI volume, published in 1990, required at least 75 procedures annually to maintain privileges.35 This number was recently relaxed to a minimum of 50 procedures annually (averaged over 2 years), based on consensus agreements among multiple physician associations published in 2013.36 The reduction in case minimums for operators is largely secondary to a progressive global decline in PCI case volumes over the past decade.37 38

There are also strong data supporting better PCI outcomes at high-volume hospitals.34 39 40 Competency guidelines published in 2011 mandated a minimum US hospital case volume of 400 PCIs per year. Together with reductions in case minimums for individual operators, the consensus guidelines published in 2013 also reduced annual case minimums for hospitals to 200.36 Contrary to US guidelines, the 2015 British Cardiovascular Intervention Society recommendations remained steadfast with a minimum hospital volume of 400 PCI cases per year.41

‘Primary’ PCI (PPCI) refers to emergent PCI in the setting of acute ST-segment elevation myocardial infarction. PPCI is arguably most analogous to NT, as both procedures are performed emergently in critically ill patients. A large body of literature supports lower in-hospital mortality after PPCI among higher-volume PPCI centers39 42–50 and higher-volume operators.47 49 Based on these data, consensus documents from 2015 have recommended that PPCI for ST-elevation myocardial infarction be performed by operators that perform more than 11 PPCI procedures per year at centers that perform at least 36 PPCI procedures per year.36 These recommendations are similar to the Society of Vascular and Interventional Neurology recommendations, which state that CSC certification requires at least 25–30 NT procedures per center per year and a minimum of 10 NT procedures per operator per year.51 Furthermore, these recommendations are similar to natural NT mortality thresholds for high-volume centers (35.2 cases per year) calculated in regression tree analysis in the Rinaldo study.31

Conclusion: a more reasonable thrombectomy volume requirement

Given the clear associations between both high cerebrovascular and PPCI case volumes and good clinical outcomes, the lack of evidence of good outcomes from low-volume centers, and societal consensus recommendations, it would not be in the best interest of patients to set a low case volume threshold for the certification of NT centers. Rather, we should create thresholds that will direct patients to competent NT centers through improved prehospital efficiencies and better patient triage systems, as has been successfully demonstrated recently.52 53 To this end we would recommend aligning NT targets with those already established for PPCI centers, requiring a minimum of 36 NT procedures annually.



  • Competing interests None declared.

  • Provenance and peer review Not commissioned; internally peer reviewed.