Tissue plasminogen activator (tPA), the only FDA approved therapy for acute ischemic attack — is known to be an effective, albeit, underused regimen. But is it cost-effective?
Yes, it is — for both short and long term care, a team of investigators concluded based on an analysis of data on 1657 acute ischemic stroke patients who were admitted to hospitals in the Netherlands within four hours of having an attack. Moreover, the hospitals that used a specific interventional program to improve timely stroke care showed a more than 10 percent increase in the thrombolysis rate compared with other participating (control) hospitals. The findings were reported in the July 11 online edition of Neurology.
Spending just $70 more per patient led to an “observed total cost savings of more than 5 percent, due to increased efficiency in care,” Louis Niessen, MD, PhD, one of the study authors, told Neurology Today. “Full-scale implementation of thrombolysis in acute stroke is long over-due,” said Dr. Niessen, an associate professor in the Health Systems Division of the Johns Hopkins Bloomberg School of Public Health in Baltimore, and a professor of public health economics at the University of East Anglia in Norwich, UK.
Commenting on the study, S. Claiborne Johnston, MD, PhD, professor of neurology and epidemiology, associate vice chancellor of research, and director of the Clinical and Translational Science Institute at the University of California-San Francisco, said: “There really aren't that many interventions that reduce costs and improve health, so this is a pretty remarkable finding — but also a believable one given previous research.” Dr. Johnston was not involved with the study.
DR. LOUIS NIESSEN sa...Image Tools
One of the strengths of the study, which pulled patients from the PRACTISE (Promoting Acute Thrombolysis in Ischemic StrokE) trial, is that the investigators were able to analyze data for patients in a real-life setting, Dr. Niessen said. The investigators collected information on resource use — number of CT scans, rt-PA dose, and length of hospital stay — and compared the costs, thrombolysis rate, and quality-adjusted life years (QALYs) years for all of the patients admitted within four hours of stroke to 12 participating hospitals — 880 patients were admitted to hospitals using a special interventional program, 777, to control hospitals. (The PRACTISE trial was a national multicenter cluster-randomized controlled trial that involved twelve hospitals, both urban and community, academic and nonacademic, in the Netherlands, with a total of 5,515 patients.)
For the intervention hospitals, teams of nurses and stroke neurologists were asked to note specific barriers to further tPA implementation in their own hospitals, and to set goals and action plans to address those challenges with realistic time frames. The study investigators monitored the results of these actions, and the teams had five half-day intervention meetings and one concluding session assessing obstacles and potential solutions in implementation, and peer-group consultations and reciprocal site visits, which took place over the course of two years. The teams used an Internet toolkit to aid them in the process, such as a hospital-wide dummy stroke patient simulation runs, detailing the indications and contra-indications to give tPA, and specifying hospital treatment protocols. In the control hospitals, tPA was administered as before — the hospitals only knew that they were participating in a registration project.
At baseline, in both groups, the mean NIHSS (National Institutes of Health Stroke Scale) was eight, indicating that they had mild/moderate symptoms. The thrombolysis rate in the intervention group was 44.3 percent versus 39.8 percent in the control group (-4.5%; 95%CI: 3.1%-5.9%). Mean costs per patient at three months were 9,192 dollars in the intervention group and 9,647 dollars in the control group (-455 USD; 95%CI: -232 to -679 USD). Overall, the cost-saving was about 1321 dollars: 22,994 dollars versus 24,315 dollars in the control group (-455 USD; 95%CI: -232 to -679 USD) and .05: 3.89 versus 3.84 in the control group (-0.05; 95%CI: -0.04 to 0.14) in QALYs. Costs were converted from Euros to US dollars (USD).
DR. S. CLAIBORNE JOH...Image Tools
“Our promotion strategy of appropriate care observes lower patient costs after three months due to reduction of hospital costs and of long-term residential costs,” Dr. Niessen told Neurology Today. He added that as the rates were already high in the study setting, “we do expect larger increases and larger cost reductions in settings where rates are considerably lower, as is the case in many US settings,” he said.
Shyam Prabhakaran, MD, associate professor of neurology at the Feinberg School of Medicine at Northwestern University in Chicago, IL, told Neurology Today that this study was an important piece of the puzzle for increasing tPA awareness and implementation. “They were able to get very detailed cost information… the results are impressive and show that while the intervention itself doubled costs per patient, that cost is offset by health care benefits and resulted in net savings overall,” he said.
Anthony S. Kim, MD, assistant clinical professor of neurology and director of the UCSF Stroke Center, added, “it's a ‘proof of concept’ that for a relatively small expenditure you can get enough of a benefit to justify that expenditure. Because stroke can cause such long-term disabilities — which can incur costs over time — anything that can improve this is very likely to be cost-effective, it's just a matter of making that investment upfront to realize these benefits.” Dr. Kim co-authored an accompanying editorial in Neurology.
Dr. Prabhakaran expressed concerns, however, about the applicability of the results to US hospitals. “I think one limitation is the authors mention that imaging tests after tPA are not recommended in the Netherlands. That may be why there was less utilization of CT in the intervention group and could have led to savings. I wonder how translatable that is to other countries where aggressive treatment, tPA included, often leads to an increase in cost of care related to the tests that are performed in those patients.”
The cost of hospital stay is also higher in the US on a per-patient basis, according to the investigation. In the US, the mean cost of hospital stay is “1000 USD, versus 365 USD in the UK, and 500 USD in the Netherlands,” they wrote, adding that this may lead to “large cross-national differences in the lifetime costs of stroke especially due to differences in residence costs.”
James C. Grotta, MD, professor and chairman of the department of neurology at the University of Texas Medical School in Houston (UTHealth), commented on the importance of the study and its use of actual patient data rather than modeling, as well as its demonstration of cost savings for hospitals — whereas previous studies had shown that savings were mainly due to reduced post-hospital costs.
But he cited some distinctions between hospital practice in the Netherlands and the US. For example, he said, they used 20 and 50 ml vials of tPA in the study, “whereas we use 100 cc vials in the US. They also only allocated 15 minutes of time for the neurologist to oversee the administration of treatment and then get back to their office — I think it would take neurologists more than 15 minutes to oversee a tPA treatment.” Some of these economic assessments may not be generalizable, he said. Dr. Grotta serves on the Neurology Today editorial advisory board.
Dr. Johnston, however, said he was convinced by these findings. The cost of the drug is pretty comparable, and although overall costs of healthcare are lower in the Netherlands, that's not because of the acute episodes of care; it has to do more with overall utilization, he said. “I think the study is transplantable and it teaches us about the cost-effectiveness on a global scale.”
DR. SHYAM PRABHAKARA...Image Tools
Dr. Prabhakaran observed that the initiative (PRACTISE) mostly focuses on “the hospital barriers such as strategies to improve thrombolysis at these individual hospitals who received the intervention. But the biggest barrier to thrombolysis in the US and worldwide occurs in the pre-hospital setting: it's related to patient awareness, knowledge of stroke symptoms and calling 911,” he said; there needs to be more work done earlier in the process before the patients even get into the hospital setting.
Of course, there are also hospital barriers — even among patients who are presenting within the four-hour time window, and the article points that out, Dr. Prabhakaran said. But not calling 911 early after symptom onset — “that is probably the number one reason why thrombolysis rates are as low as they are.”
Dr. Niessen told Neurology Today that these results now “open the field for joint quality improvement studies in the US, involving leading clinicians and support staff as well as insurance companies, HMOs and other health care management organizations such as the American Heart Association's Get With The Guidelines program and its Stroke Toolbox.”
Listen as Shyam Prabhakaran, MD, associate professor of neurology at the Feinberg School of Medicine at Northwestern University in Chicago, discusses in-hospital and pre-hospital barriers to stroke care. http://bit.ly/Pdms3L
• Dirks M, Baeten SA, Niessen LW, et al. Real-life costs and effects of an implementation program to increase thrombolysis in stroke. Neurology 2012; E-pub 2012 July 11.
• Dirks M, Niessen LW, Dippel DWJ, et al. Promoting Thrombolysis in Acute Ischemic Stroke for the PRomoting ACute Thrombolysis in Ischemic StrokE (PRACTISE) Investigators. Stroke 2011; E-pub 2011 Mar 10.
IMPROVEMENT STRATEGIES FOR STROKE CARE
Shyam Prabhakaran, MD, associate professor of neurology at the Feinberg School of Medicine at Northwestern University in Chicago, IL, noted that there are already several large initiatives in the US working to improve overall quality of care for stroke: the development of Primary Stroke Centers, Get With The Guidelines, and Target: Stroke. “What would be of interest would be some formal study of these initiatives and their effect on outcomes and health care costs as they've done here in the Netherlands. There should be some organized way of assessing whether those strategies and the implementation of an intervention like Target: Stroke or Get With The Guidelines has a downstream benefit,” Dr. Prabhakaran said.
S. Claiborne Johnston, MD, PhD, professor of neurology and epidemiology, associate vice chancellor of research, and director of the Clinical and Translational Science Institute at the University of California-San Francisco, said that the specific intervention in this study may not be the answer to improving stroke care in the US, but it does provide some valuable in-sight. “When you look at what the American Heart Association's program, Get With The Guidelines, does just by encouraging hospitals to track their performance and cre-ate an improvement plan — it seems to be enough to improve care.
“I'm not sure that the specifics of the intervention are so important, but a com-mitment to improvement and a system for tracking the process measures, quality metrics — that's crucial,” he told Neurology Today. There are likely many effective intervention models — some which may work better than others depending on the setting, he said.
At his home institution, Anthony S. Kim, MD, assistant clinical professor of neurology and director of the UCSF Stroke Center said they've been using a real-time tracking system that will contribute to the improvement of hospital-administered stroke services. “We've been trying to leverage information technology to provide rapid feed-back for quality improvement in stroke care….We find that real-time feedback about the percentage of patients getting tPA, real-time feedback of door-to-needle time — this is directly linked to improvements because the first step in improving a process is to measure it,” he said.
In the May 23 issue of Neurology, a paper by Atte Meretoja, MD, PhD, and colleagues, offered an analysis aimed at cutting this in-hospital delay (or door-to-needle time) for acute stroke patients down to 20 minutes. Patients in this study were prospectively regis-tered in the Helsinki Stroke Thrombolysis Registry, and a series of interventions to reduce treatment delays were implemented between 1998 and 2011 on a total of 1,860 patients.
The key to reducing the time to treatment, the authors wrote, “is to do as little as pos-sible after the patient has arrived at the emergency room and as much as possible before that, while the patient is being transported.” For example, lack of reliable history, they wrote, is a major cause of delay in evaluation of tPA patients; if patient history is taken while in transit, this could majorly reduce delays. A 20 minute target is feasible, they concluded.
The biggest reason for exclusion from tPA is the failure to arrive at the hospital in time, Dr. Kim said. “This study by Dirks et al. focuses on what we can do at the hospital level — and there are certainly improvements that can be made there, but there's also likely greater impact to be made by focusing on the factors that limit patients getting to the hospital in time.”
Communities where ambulances triage stroke patients to stroke centers, Dr. Johnston pointed out, tend to benefit more than the communities where ambulances just take patients to the near-est hospital; “One implication of this is that we ought to think about ways of supporting better use of those centers that make tPA a priority — as well as other elements of stroke care.”
James C. Grotta, MD, professor and chairman of the department of neurology at the University of Texas Medical School in Houston, said that in a community intervention study he conducted for the T.L.L. Temple Foundation several years ago in East Texas, they provided a public edu-cation intervention to patients, hospital, and EMS workers about signs of stroke and early re-sponse. “We found that we were able to increase community-wide tPA treatment rates, and that this was sustained.” But almost all of the improvement, he added, was due to better responsiveness of paramedics in the emergency department, not, for example, from the individual patients calling 911 earlier.
Our efforts, Dr. Grotta, who is also director of the Stroke Service at the Mischer Neurosci-ence Institute at the Memorial Hermann Hospital –Texas Medical Center, said, should be focused on working with EMS and paramedics to make sure we have a “coordinated system of pre-hospital stroke recognition and triage so that the right pa-tients are taken to the right hospitals and that those stroke centers utilize programs like this to improve their efficiency.”
• Meretoja A, Strbian D, Kaste M, et al. Reducing in-hospital delay to 20 minutes in stroke thrombolysis. Neurology 2012; E-pub 2012 May 23.
• Morgenstern LB, Staub L, Grotta JC, et al. Improving Delivery of Acute Stroke Therapy: The TLL Temple Foundation Stroke Project. Stroke