Survival Benefit of Stroke Centers Negated After 90 Minutes of Travel

July 28, 2016

Stroke patients treated at a primary stroke center have a lower case fatality rate than those treated in hospitals not designated as centers of excellence in stroke care, a new study of almost 900,000 Medicare patients shows.

However, the study also shows the mortality advantage was no longer present if it took the patient longer than 90 minutes reach a designated stroke center.

In addition, many stroke patients who lived within 90 minutes of a primary stroke center were not taken to that hospital and so missed out on best available care.

The study, published online on July 25 in JAMA Internal Medicine, was led by Kimon Bekelis, MD, Dartmouth-Hitchcock Medical Center, Lebanon, New Hampshire.

"Our results suggest that stroke patients have a better chance of survival if they are taken to a primary stroke center rather. However, if it will take more than 90 minutes to get to the stroke center then it is probably not worth it," Dr Bekelis told Medscape Medical News.

Largest Dataset

For the study, the researchers conducted a retrospective review of Medicare beneficiaries with stroke admitted to a hospital between 2010 and 2013.

Drive times were calculated on the basis of ZIP codes and street-level road network data. The investigators used an instrumental variable analysis based on the differential travel time to primary stroke centers to control for unmeasured confounding. The setting was a 100% sample of Medicare fee-for-service claims.

Among the 865,184 stroke patients included in the study (mean age, 78.9 years), 53.9% were treated at a primary stroke center. Results showed admission to a primary stroke center was associated with 1.8% lower case fatality rate at both 7 days and 30 days.

This translates into 56 patients needed to be treated at a primary stroke center to save 1 life.

'With almost 900,000 acute stroke patients, this is largest dataset used ever to look at the issue of outcomes at different hospitals," said Dr Bekelis.

Data on comprehensive stroke centers were not included.

"There were very few comprehensive stroke centers at the time this study was conducted, but some of the hospitals classified here as primary stroke centers may have actually been comprehensive stroke centers," Dr Bekelis noted.

He said the instrumental variable analysis created a "pseudo randomization," which simulates the effect of randomly assigning patients to primary stroke centers or a different hospital.

Reliable Results

"This technique allows us to harness the power of large retrospective datasets, but the standard errors are larger so the possibility of showing significance is less. But we still showed a significant result. We conducted multiple sensitivity analyses in many different areas and subgroups, and they all showed similar results so we think our results are reliable."

However, he cautioned that the results applied only to the population studied — that is, older Americans. "We can't apply these results to younger patients or to other countries."

Dr Bekelis also noted that the 90-minute timeframe should be taken only as a rough guide.

"This was the overall finding from a national view. There is a lot of subtlety in this message — many local factors have to be taken into account when making policies. For example, there may be intermediate hospitals which are not classified as primary stroke centers but have the capabilities of treating stroke patients well.

"In addition, there may be areas where telemedicine is used. This may change the 90-minute message. Each local area will have to make decisions on their particular circumstances," he said.

Noting that only half the patients in the dataset went to a primary stroke center, Dr Bekelis said this is far from optimum, given that only 16% of the dataset lived more than 90 minutes away from such a hospital.

"While we can expect fewer patients to go to a primary stroke center in rural areas where distances to hospitals are large, many patients who live in big cities still didn't go to a primary stroke center. "

Gap in Emergency Care

"Our data suggest that about 30% of individuals in this study lived within 90 minutes of a primary stroke center but went to a different hospital instead. It is curious why this happened," he continued.

He explained that the way patients are directed to which hospital varies in each area and depends on many local factors, including individual hospital and emergency networks.

"There is no national requirement for patients to go to a primary stroke center. Individual states dictate their own policy. It is likely that this would direct patients to a hospital in their own state even if there is another one closer but in a different state," he said.

"We need more primary stroke centers, but people also need to find out where their nearest stroke center is so they are prepared. Also emergency medical services need to do a better job of taking suspected stroke patients to hospitals capable of treating them properly.

"Local healthcare authorities and policymakers need to come together and formulate local decisions on where suspected stroke patients should be taken. Our data can provide some guidance on this."

He added that populations who live more than 90 minutes away from a stroke center need to be identified and other options developed for those patients, such as use of helicopters or telemedicine.

"While there are some very well-organized networks of stroke services in some regions, these are the exception rather than the rule at present. Our data shows we are lacking in taking care of the emergency medical needs of our older population. This needs to change, and we hope our data will help make that happen."

Need for Coordinated Networks

In an accompanying editorial, Lee H. Schwamm, MD, Massachusetts General Hospital, Harvard Medical School, Boston, suggests the following recommendation for emergency medical services: "[I]f it is a disabling stroke that started in the last 6 hours, then go to the highest-level stroke center that is within 30 to 45 extra minutes of drive time."

He points out that in the future, coordinated networks need to be developed with smartphone apps or similar methods that can factor in the onset time, degree of stroke severity, travel times, hospital door-to-needle and door-to-puncture times, rates of recanalization success, patient preference, and in-hospital mortality to determine the best possible destination for each patient.

Dr Schwamm adds that hospitals must be incentivized to report performance data to enable such smart triage, and stroke incidence and 90-day functional outcomes should become a reportable disease so that meaningful data can be collected.

He notes that such a prehospital system should adhere to national standards but be customized to reflect the local resources, prevalence of stroke, best available screening tools, acceptable levels of erroneous triage, and competing costs of the additional transport and reduced availability of emergency medical services.

"It will not be easy, but it is well worth doing," he concludes.

JAMA Intern Med. Published online July 25, 2016. Abstract Editorial

For more Medscape Neurology news, join us on Facebook and Twitter


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.