Real-world data investigating patient outcome after endovascular stroke therapy in high versus low volume centres finds a 30% increase in the likelihood of a good outcome for every 10 additional thrombectomy procedures performed at a particular hospital per year. The findings were presented by Sunil Sheth, assistant professor of neurology at McGovern Medical School at UTHealth, Houston, USA, at the International Stroke Conference (ISC; 5–8 February, Honolulu, USA), while the study was simultaneously published in Stroke.
“With the wave of clinical trials published in the last few years, endovascular therapy has dramatically changed the way we approach ischaemic stroke care,” began Sheth. He acknowledged that one of the key challenges burdening stroke systems currently is patient access to thrombectomy, as well as the appropriate screening and testing beforehand.
Sheth alluded to the fact that the data emerging from trials published in 2015 were from procedures performed at high volume tertiary care referral centres; centres with advanced neuroimaging, surgery and neuro-trained nurses. For this reason, he questioned whether the results could be translated to other clinical settings.
Addressing the ISC audience, Sheth argued: “There is a push to disseminate these treatments out in the community into local hospitals where patients may get treated faster because it is closer to where they live, and can be evaluated by the emergency medical services. However, the efficacy of this treatment in these settings is unknown.”
The current study aimed to gain a better understand of real-world practise and outcomes of endovascular stroke therapy. In order to do so, Sheth and colleagues looked into practise patterns of thrombectomy over a 10-year period (2006–2016) in a large cohort, and subsequently evaluated the association between clinical outcomes and hospital treatment in higher versus lower volume settings.
Sheth and his team performed a retrospective cross-sectional study, utilising both the Healthcare Cost and Utilization Project (HCUP) State Inpatient Database (SID), as well as the State Emergency Department Database (SEDD) on all discharges from non-federal acute care hospitals in Florida, from 2006–2016. Sheth noted that these datasets allowed the investigators to track individual patients through inpatient and emergency room visits. As Sheth et al also used the Nationwide Inpatient Sample (NIS) dataset from 2012–2016, they were able to assess the generalisability of the findings for the Florida cohort in a larger national sample.
The primary endpoint was good neurological outcome, defined as discharge to home or acute rehabilitation. Sheth and his colleagues included patients if they had endovascular stroke therapy, yet excluded those that had any diagnosis or treatment of arteriovenous malformation or fistula, prior intracerebral haemorrhage, or trauma.
In total, 3,890 patients from the Florida cohort were treated with thrombectomy, the median age being 73 (range: 61–82), while 51% were female and 44% received IV tPA. Sheth reported that the characteristics and past medical history of the patients treated at relatively lower volume centres (n=1,974) and those treated at relatively higher volume centres (n=1,916) were similar. One difference that Sheth mentioned was that in the higher volume centres, there was a lower volume of patients that received IV tPA; 35% compared with 50% in the lower volume centres.
Pointing to a series of graphs, Sheth described the annual trends that were observed over the 10-year period. The total number of thrombectomy procedures that were performed increased continuously, with a jump in the amount carried out in 2015, which, according to Sheth, coincided with the release of prominent endovascular stroke therapy trials. Further, the investigators observed a similar trend in the number of patients that were transferred for thrombectomy, while a continuous increase in the number of hospitals that were performing at least one thrombectomy procedure per year was also found.
In terms of the distribution of the all thrombctomy procedures performed in this time period, Sheth et al found that in 2008, the top three performing hospitals carried out almost 50% of the procedures, while the top eight carried out nearly 90%. However, come 2016 and a dramatic shift in the distribution of procedures took place; where the top three only carried out 20%, and the top eight nearer 40%. “By then, there were far more hospitals doing the procedure and the number of cases was distributed across a much larger number of hospitals,” remarked Sheth.
When validating these findings using their Nationwide cohort—hospitals across the USA—Sheth said: “Again, we saw a similar increase in the number of procedures that were being performed annually, with a similar jump in 2015.” When the authors subsequently looked at the number of procedures that were being performed in centres with a relatively lower volume—fewer than 20 procedures per year—they found that this pattern throughout the 10-year time period echoed the increase in the total annual thrombectomy procedures carried out at all hospitals, indicating that a substantial amount of procedures were performed in lower volume centres.
After adjusting the analysis for age, sex, and comorbidities, Sheth found that a continuously increasing likelihood of good outcome was found to be associated with an increase in the number of annual thrombectomy procedures performed in a Florida cohort, a finding of which was validated through the nationwide cohort (OR: 1.3; 95% CI: 1.2, 1.4).
Following this, Sheth said that the study investigators looked into whether they could account for the fact that these larger thrombectomy volume centres may also be more sophisticated. Thus, they investigated the likelihood of good discharge outcome for patients that had acute ischaemic stroke but did not get thrombectomy in hospitals. However, for these particular patients, no relationship was observed between outcome and thrombectomy volume of the hospital. This finding argued against the fact that higher endovascular volume centres were simply better at caring for patients with acute ischemic stroke, and that outcome differences may be directly related to endovascular treatment volumes.
On discussion of the findings, Sheth maintained that: “The challenge that is facing these stroke systems is: How to ensure that every patient can have rapid access to high quality thrombectomy.” He acknowledged the current solutions to this particular challenge: traditional hub-and-spoke models, tip-and-treat models (where the physician and team travels to the patient), thrombectomy-capable models (smaller hospitals that perform thrombectomy with fewer requirements), and mobile stroke units—that help with the prehospital triage.
“Through these real-world data, we observed a significant effect of hospital volume on discharge outcomes, with the odds ratio indicating a 30% increase in the likelihood of good outcomes for every 10 additional annual endovascular stroke treatments per year,” said Sheth. Of importance, he highlighted that these findings also support the idea that the results of the recent endovascular stroke trials may not be generalisable to every clinical setting.
However, Sheth noted particular limitations, the main being that beyond thrombectomy volume, there are numerous hospital-specific factors that can determine outcome, such as time of onset, occlusion location, recanalisation grade and infarct volume.
Yet, in summary, Sheth said that three important take-aways of the current findings still stand. In the large population-level study of patients treated with thrombectomy between 2006–2016, Sheth and colleagues observed a continuous increase in the annual treatment rates and the number of thrombectomy performing hospitals. Further, they witnessed a shift in procedural volume across a “substantially greater number of hospitals,” while the patients treated with thrombectomy in hospitals with greater annual procedural volume had better discharge outcomes. Sheth reported that in conclusion, these findings support the need for further study on the efficacy of endovascular stroke therapy outside of specialised centres.
Speaking to NeuroNews, Sheth contextualises the importance of this treatment: “There is no country in the world where the absolute number of people living with or died from stroke has declined between 1990 and 2013. In the USA, approximately 795,000 people experience a stroke each year with nearly 90% being acute ischaemic stroke, which remains the leading cause of adult disability in the USA. In 2015 landmark clinical trials demonstrated that endovascular stroke treatments for patients with large vessel occlusion leads to dramatic improvements in patient outcomes.
“However, in the wake of these results, stroke systems of care around the globe are now faced with the daunting task of ensuring that patients with acute ischaemic stroke have access to appropriate screening and therapy.
“The evidence of benefit for endovascular stroke therapy that emerged from these trials was derived from treatments rendered almost exclusively at high volume stroke centres, with specialised neuro-imaging, neuro-intensive care, neuro-rehabilitation and neuro-nursing. However, since the publication and adoption of these findings into guidelines, it has become well-established that the likelihood of good neurologic outcome for these patients remains dependent on minimising delays in treatment. Even 15-minute delays in endovascular reperfusion have been associated with quantifiable decrements in clinical outcomes. As such, there has been an increase in demand for the procedure as well as calls for the dissemination of the treatment away from tertiary-care referral centres into the community, to avoid the costly delays associated with inter-hospital transfer.
“On the other hand, transferring endovascular stroke therapy patients to higher volume centres has also been associated with reduced mortality. In the absence of clear data on the relative efficacy of treatment in lower volume centres, this lack of clarity on the optimal distribution of endovascular stroke therapy resources had led to considerable confusion, with stroke centre certifying agencies such as The Joint Commission initially requiring physician and hospital minimal endovascular stroke therapy volume requirements for certification, and then very recently revoking and then reinstating that criterion.
“Given the need to structure stroke systems of care in the modern treatment era, as well as the poorly characterised effect on endovascular stroke therapy outcomes away from tertiary-care referral centres, understanding the trends in treatment patterns as well as outcomes in relation to treatment volumes and inter-hospital transfer is of vital importance. The study described here provides for the first time large-scale data on the utilisation of the procedure as well as the finding that its outcomes are directly tied to annual volumes.”
