Mobile stroke units becoming more common despite cost effectiveness questions

March 14, 2017|Emergency Medicine

EXPERT ANALYSIS FROM THE INTERNATIONAL STROKE CONFERENCE

Early data on effectiveness

Dr. Nour reported some of the best evidence for the incremental clinical benefit of MSUs based on the reduced time for starting a tPA infusion. She used data collected by the Berlin group and published in September 2016 that compared the treatment courses and outcomes of patients managed with an MSU with similar patients managed by conventional ambulance transport for whom CT scan assessment and the start of TPA treatment did not begin until the patient reached a hospital.

The German analysis showed that, in the observational Pre-Hospital Acute Neurological Therapy and Optimization of Medical Care in Stroke Patients–Study (PHANTOM-S), among 353 patients treated by conventional transport, the median time from stroke onset to thrombolysis was 112 minutes, compared with a median of 73 minutes among 305 patients managed with an MSU, a statistically significant difference. However, the study found no significant difference for its primary endpoint: the percentage of patients with a modified Rankin Scale score of 0-1 when measured 90 days after their respective strokes. This outcome occurred in 47% of the control patients managed conventionally and in 53% of those managed by an MSU, a difference that fell short of statistical significance (Lancet Neurol. 2016 Sept;15[10]:1035-43).

Dr. Nour attributed the lack of statistical significance for this primary endpoint to the relatively small number of patients enrolled in PHANTOM-S. “The study was underpowered,” she said.

Dr. Nour presented an analysis at the meeting that extrapolated the results out to 1,000 hypothetical patients and tallied the benefits that a larger number of patients could expect to receive if their outcomes paralleled those seen in the published results. It showed that, among 1,000 stroke patients treated with an MSU, 58 were expected to be free from disability 90 days later, and an additional 124 patients would have some improvement in their 90 day clinical outcome based on their modified Rankin Scale scores when compared with patients undergoing conventional hospitalization.

“If this finding was confirmed in a larger, controlled study, it would suggest that MSU-based thrombolysis has substantial clinical benefit,” she concluded.

Another recent report looked at the first 100 stroke patients treated by the Cleveland MSU during 2014. Researchers at the Cleveland Clinic and Case Western Reserve University said that 16 of those 100 patients received tPA, and the median time from their emergency call to thrombolytic treatment was 38.5 minutes faster than for 53 stroke patients treated during the same period at emergency departments operated by the Cleveland Clinic, a statistically significant difference. However, this report included no data on clinical outcomes (Neurology. 2017 March 8. doi: 10.1212/WNL.0000000000003786).

Running the financial numbers

Nailing down the incremental clinical benefit from MSUs is clearly a very important part of determining the value of this strategy, but another very practical concern is how much the service costs and whether it is financially sustainable.

“We did a cost-effectiveness analysis based on the PHANTOM-S data, and we were conservative by only looking at the benefit from early tPA treatment,” Heinrich J. Audebert, MD, professor of neurology at Charité Hospital in Berlin and head of the team running Berlin’s MSU, said during the MSU session at the meeting. “We did not take into account saving money by avoiding long-term stroke disability and just considered the cost of [immediate] care and the quality adjusted life years. We calculated a cost of $35,000 per quality adjusted life year, which is absolutely acceptable.”

He cautioned that this analysis was not based on actual outcomes but on the numbers needed to treat calculated from the PHANTOM-S results. “We need to now show this in controlled trials,” he admitted.

During his talk at the same session, Dr. Grotta ran through the numbers for the Houston program. They spent $1.1 million to put their MSU into service in early 2014, and, based on the expenses accrued since then, he estimated an annual staffing cost of about $400,000 and an annual operating cost of about $100,000, for a total estimated 5-year cost of about $3.6 million. Staffing of the Houston MSU started with a registered nurse, CT technician, paramedic, and vascular neurologist, although, like most other U.S. MSUs, the onboard neurologist has since been replaced by a second paramedic, and the neurologic diagnostic consult comes via a telemedicine link.

Income from transport reimbursement, currently $500 per trip, and reimbursements of $17,000 above costs for administering tPA and of roughly $40,000 above costs for performing thrombectomy are balancing these costs. Based on an estimated additional one thrombolysis case per month and one additional thrombectomy case per month, the MSU yields a potential incremental income to the hospital running the MSU of about $3.8 million over 5 years, enough to balance the operating cost, Dr. Grotta said.

A key part of controlling costs is having the neurologic consult done via a telemedicine link rather than by neurologist at the MSU. “Telemedicine reduces operational costs and improves efficiency,” noted M. Shazam Hussain, MD, interim director of the Cerebrovascular Center at the Cleveland Clinic. “Cost effectiveness is a very important part of the concept” of MSUs, he said at the session.

The Houston group reported results from a study that directly compared the diagnostic performance of an onboard neurologist with that of a telemedicine neurologist linked in remotely during MSU deployments for 174 patients. For these cases, the two neurologists each made an independent diagnosis that the researchers then compared. The two diagnoses concurred for 88% of the cases, Tzu-Ching Wu, MD, reported at the meeting. This rate of agreement matched the incidence of concordance between two neurologists who independently assessed the same patients at the hospital (Stroke. 2017 Jan;48[1]:222-4), said Dr. Wu, a vascular neurologist and director of the telemedicine program at the University of Texas Health Science Center in Houston.

“The results support using telemedicine as the primary means of assessment on the MSU,” said Dr. Wu. “This may enhance MSU efficiency and reduce costs.” His group’s next study of MSU telemedicine will compare the time needed to make a diagnostic decision using the two approaches, something not formally examined in the study he reported.

However, telemedicine assessment of CT results gathered in a MSU has one major limitation: the time needed to transmit the huge amount of information in a CT angiogram.

The MSU used by clinicians at the University of Tennessee, Memphis, incorporates an extremely powerful battery that enables “full CT scanner capability with a moving gantry,” said Andrei V. Alexandrov, MD, professor and chairman of neurology at the university. With this set up “we can do in-the-field multiphasic CT–angiography from the aortic arch up within 4 minutes. The challenge of doing this is simple. It’s 1.7 gigabytes of data,” which would take a prohibitively long time to transmit from a remote site, he explained. As a result, the complete set of images from the field CT angiogram are delivered on a memory stick to the attending hospital neurologist once the MSU returns.