Endovascular Stroke Therapy Less Beneficial After Transfer

In-hospital mortality and functional outcome were worse in transferred patients. Arriving at a solution won’t be easy.

Endovascular Stroke Therapy Less Beneficial After Transfer

Acute ischemic stroke patients who undergo endovascular therapy have worse outcomes—specifically, higher in-hospital mortality and poorer function—if are transferred to another center before receiving the treatment, observational data from the United States show.

“The documented magnitude of benefit was substantial (~35 lives saved per 1,000 procedures, number needed to treat 28.6),” Mohamad Alkhouli, MD (Mayo Clinic, Rochester, MN), and colleagues point out. Their results, published online Wednesday in JACC: Cardiovascular Interventions, confirm earlier reports showing transfer’s ill effects, this time in US contemporary practice, and emphasize the need to rethink how stroke care is delivered, they say.

Importantly, the data set captured only patients who actually underwent endovascular stroke therapy (EST), they point out. “Patients who were candidates for EST at the receiving facility but became poor candidates for it later due to delays in transfer, interval bleeding, or clinical deterioration were not included,” meaning that the actual gap in outcomes is likely even larger.

The landscape of stroke care changed in 2015, when three large trials showed a benefit to mechanical thrombectomy in treating large-vessel occlusion (LVO) and guidelines formally endorsed an endovascular approach. Further positive findings widened the treatment window, in 2018, from within 6 hours to up to 24 hours, raising the question of whether the inherent delays of transfer would no longer be a concern.

Despite this evolution, “time is brain” remains true and concerted effort is required to figure out the best way to provide EST to the patients who need it, Alkhouli said. With the larger timeframe of up to 24 hours, “you now have more candidates. The number of patients who might benefit is very large and there is not the capacity to address all of those patients with the current manpower,” he explained. “So there is a need for expansion. Now, how does that expansion work? That’s the million dollar question.”

Uneven Outcomes

Using numbers from the Vizient clinical database, Alkhouli et al identified 22,193 acute ischemic stroke patients who underwent EST between October 2015 and September 2019. Around half had their procedures done at the hospitals where they presented and half were transferred for EST.

Propensity-matched analysis (n = 7,557 pairs) was done to account for differences in age, sex, race, comorbidities, hospital teaching status, institutional procedural volume, prior tissue plasminogen activator (tPA) use, and primary payer.

Even so, transferred patients had higher in-hospital mortality (14.7% vs 11.2%) and were more likely to have poor functional outcome, defined as death or discharge to hospice/long-term nursing care (40.7% vs 35.9%; both P < 0.001). Adjusted odds ratios were 1.38 (95%CI 1.26-1.51) and 1.26 (95%CI 1.18-1.34), respectively, for the two endpoints.

Secondary endpoints including intracranial hemorrhage and need for mechanical ventilation were significantly more common in the transferred group. Median cost, however, favored transfer over on-site EST ($35,179 vs $38,790; P < 0.001).

What to Do?

Over the 4-year period, neither the number of EST-capable centers nor the number of patients treated on-site changed, the investigators observe.

There are multiple approaches to optimizing access to stroke care, they note, but also many unknowns. It’s unclear whether expanding local access will improve outcomes and whether this would be feasible with the current infrastructure. Also, they ask, “how would that be planned, rolled out, and monitored as the availability of EST transcatheter skills locally does not equal the ability to operate and maintain a high-quality EST center? . . . Would expanding EST to more local hospitals lead to more patients receiving guideline-recommended lifesaving treatment, or would it dilute the operator’s experience and subsequently negatively impact low-performing sites?”

It may be necessary, Alkhouli noted to TCTMD, for other endovascular specialties, such as interventional cardiology and radiology, to become involved and undergo extra training. The process of improving stroke care delivery will require input from various professional societies, he said, calling for a “patient-centered approach more than a specialty-centered approach.”

In an editorial, Don Heck, MD (Triad Radiology Associates, Novant Health Forsyth Medical Center, Winston-Salem, NC), points out that current guidelines don’t just set standards for operators. “The guidelines also indicate optimal hospital requirements: adequate neuroangiography suites; expertise in neurocritical care, vascular neurology, and neurosurgery; and 24/7 access to computed tomography and MRI,” he writes. “All of which is to say a doctor with a catheter is not a stroke program, and a billboard does not make a stroke center.”

Agreeing with this message, Alkhouli said it’s important to have not only catheter skills but also expertise in imaging and other aspects of stroke care.

[We need] a patient-centered approach more than a specialty-centered approach. Mohamad Alkhouli

Apart from increasing the availability of EST, Heck suggests another option: simply bypassing the nearest hospital to go straight to a comprehensive stroke center (CSC). “The most important strategy for treating LVO patients faster is bypassing a primary stroke center (a hospital that gives tPA, but does not do EST) for a slightly more distant CSC initially,” he says. “For those LVO patients who are not tPA-eligible, failing to bypass will only cause a time delay. For those who are tPA eligible, even if slightly earlier tPA administration is possible at [closer hospital] it is not likely to make a difference compared with the [more-crucial] time delay to thrombectomy.”

Alkhouli countered that it’s hard for emergency medical services in the field to make the decision to travel further without knowing for sure if a patient actually has an LVO.

One caveat to their analysis, he acknowledged, is that it’s observational and may have residual confounding. One example of this, as pointed out by Heck, is the dramatically higher use of tPA in the transferred versus on-site group (38.9% vs 8.5%; P < 0.001). The researchers adjusted for this in their propensity matching, but to Heck the disparity hints that the off-site patients presented to the hospital earlier and that transfer might be even riskier than it appears.

“This is a descriptive study with limitations,” Alkhouli said, adding, “I don’t think that the study gives you all the answers. It definitely starts the conversation, though.”

  • Alkhouli reports no relevant conflicts of interest.
  • Heck reports having been a consultant for Stryker Neurovascular.