Rapid Time to Thrombolysis Does Not Cut Diagnostic Accuracy

July 20, 2018

A stroke center in Helsinki, Finland, which has an average door-to-needle time for thrombolysis of less than 20 minutes, has reported reassuring data on low rates of misdiagnosis, suggesting prompt treatment times do not sacrifice diagnostic accuracy.

"We are so fast that we wanted to check that we are not skipping anything," senior author Perttu J. Lindsberg, MD, Helsinki University Hospital, Finland, told Medscape Medical News.

"Many people think we will be sacrificing diagnostic accuracy to get our door- to-needle times down to under 20 minutes, but this paper shows that this is not the case. We had a misdiagnosis rate of 15% — that is a very low figure. And the correct diagnosis of acute cerebral ischemia was 91%." 

The study was published online July 11 in Neurology. 

19-Minute Door-to-Needle Time

For the study, the investigators analyzed data on the accuracy of admission diagnosis in a cohort of 1015 stroke-code patients arriving by ambulance to Helsinki University Hospital, which is the only 24/7 neurology service within the district and receives more than 1000 acute-stroke-code patients annually.

Lindsberg noted that the label of "stroke code" means these patients had been identified by emergency medical services as a potential acute stroke case and a possible candidate for thrombolysis and had thus been urgently transported to the central stroke treatment hospital.

Median time from admission to the start of thrombolysis (door-to-needle time) was just 19 minutes.

Results showed the rate of correct admission diagnosis was 91.1% (604/663) for acute cerebral ischemia (ischemic stroke/transient ischemic attack), 99.2% (117/118) for hemorrhagic stroke, and 61.5% (144/234) for stroke mimics.

Among the 150 (14.8%) patients misdiagnosed, initial imaging showed no acute findings in 135 (90.0%) and 100 (67.6%) presented with a National Institutes of Health Stroke Scale score of 0 to 2.

Misdiagnosis altered medical management in 70 patients, including administration of unnecessary treatments (thrombolysis, n = 13; other, n = 24), omission of thrombolysis (n = 5), delays to specific treatments of stroke mimics (n = 13; median, 56 hours), and delays to antiplatelet medication (n = 14; median, 1 day).

Misdiagnosis extended emergency department stay (median, 6.6 vs 5.8 hours) and led to unnecessary stroke unit stay for 10 patients. For eight patients (0.8%), misdiagnosis was possible or likely to have worsened outcomes, but no patients died as a result of misdiagnosis.

"These findings support the safety of gradually optimized DNTs [door-to-needle times] in a high-volume neurologic ED [emergency department], when sufficient physician training and support are in place," the authors write.

More Research, Resources Warranted

The investigators note that the results warrant increased "attention, research and resourcing to improve rapid admission diagnostics of acute neurologic patients."

Lindsberg said he was pleased with the findings.  Asked how these figures on misdiagnosis compare to those in other centers, he said it was difficult to know because this information is not often published.

"The importance of this paper is that this is a large cohort of patients with full diagnostic follow-up. These statistics are difficult to establish as it requires an in-depth retrospective analysis as to the final diagnosis of all patients.   I haven't seen a comparable study, but now that we have published this can be used as a benchmark."

He noted that many stroke guidelines recommend a goal of less than 1 hour for door-to-needle time in stroke but that "not many hospitals can do it."

"We have been training for this for 20 years. We are constantly honing our processes to reduce our door-to-needle time. We are now probably at the optimal time. I can't see that we can get down much further than 19 minutes. We have to thoroughly clinically assess the patient and make sure they absolutely do have an ischemic stroke. We don't want to jump the tPA [tissue plasminogen activator] gun."    

Lindsberg said the key to achieving such rapid treatment times is largely due to having a coordinated integrated stroke system that includes emergency medical services.

Centralized Stroke Service

"The main focus of our approach is to do as much as we can before the patient arrives. The key thing is that we know they are coming as emergency services have alerted us that they are on their way with a stroke code patient," he said.

This means the team can access the patient's digital medical records and lab results and assess candidacy for thrombolysis. It also means the imaging service can be ready to go as soon as the patient arrives, he added.

"If they are found to be a candidate for thrombolysis, then this is started right there in the CT room," said Lindsberg.

All of this is possible because of a highly centralized stroke service, he said. His center is the main hospital in the city for acute stroke and, as such, it has a high stroke patient volume and a highly skilled and experienced team on hand at all times.

"The emergency services are well organized and integrated into the stroke chain, so that anyone with a suspected stroke will immediately be brought to our hospital. Unlike the situation in the US, we don't have different ambulance companies or hospitals competing with each other or different insurance requirements to think about," Lindsberg said.

He said the next challenge is to see whether these impressive times can also be achieved with endovascular therapy.

"That's a whole new ball game. It involves adding a CT angiogram but more difficult is the need to transit the patient to the angio suite and assemble the whole interventional team, but we are hopeful we can also be very efficient in this regard too," he said.

Strengths, Limitations

In an accompanying editorial, Klaus Fassbender, MD, Saarland University Medical Center, Homburg, Germany, and Kevin N. Sheth, MD, Yale School of Medicine, New Haven, Connecticut,  note the results show "high" rates of correct admission diagnoses and rates of over- and undertreatment "were within normal ranges."

However, they also point out some of the study's limitations.

These include the lack of a comparator, such as a parallel group or a historical control group. In addition, the laboratory workup included only international normalized ratio, and tPA was administered apparently without awaiting other laboratory markers, despite their rapid availability with a point-of-care laboratory test.

"In the last decade, in parallel with such in-hospital improvements, strategies to streamline prehospital stroke management have arisen, with the use of mobile stroke units, allowing diagnostic workup, treatment, and correct triage decisions, starting at the emergency site," they note.

"Considering the stroke chain of survival as a continuum beginning with symptom onset and ending with causal treatment, a synergistic combination of both pre- and in-hospital stroke management improvements, such as those studied here, may be the key for fastest possible treatment times and best possible outcomes," they conclude.

The study was funded by the Maire Taponen Foundation, The Finnish Medical Foundation, The Finnish Foundation for Cardiovascular Research, Helsinki University Hospital governmental research funds, Jane and Aatos Erkko Foundation, Signe and Ane Gyllenberg Foundation, and Sigrid Juselius Foundation.  Lindsberg and Fassbender have disclosed no relevant financial relationships. Sheth has served on the editorial boards of Neurology Today, Neurosurgery, Stroke, and Neurology; holds a provisional patent for stroke detection technology; has received research support from Remedy Pharmaceuticals, Bard, Stryker, Novartis, National Institutes of Health, American Academy of Neurology Clinical Research Fellowship, the American Heart Association, and the Passano Foundation; and has served as a legal consultant.

Neurology. Published online July 11, 2018.  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.