MESSAGE FROM THE EDITOR Comparative Effectiveness Research in the Neurosciences There has been a lot of buzz about comparative effectiveness research recently. Defined as research that evaluates interventions already in clinical use, comparative effectiveness research has been touted as part of the answer to rising medical costs1 and is explicitly targeted to receive over $1 billion USD in recent versions of the economic stimulus package. How it will affect clinical researchers and ultimately impact the clinical neurosciences remains to be seen. Advancing science and health are the major goals driving the biomedical research agenda. As testimony to the success of this agenda, we have witnessed substantial gains in health in the US since the birth of the NIH extramural granting 63 years ago, with an increase in life expectancy from 67 in 1946 to 78 years in 2004,2 and with dramatic declines and even elimination of impactful diseases. Biomedical research is not responsible for all these gains, but its important role is undeniable.3 The health advances wrought by biomedical research have come at a cost. Most medical interventions are associated with increased cost to society, with only a handful actually producing cost savings, even when lost productivity is considered.3 As knowledge has increased thanks to research, clinical practice has become more evidence based with greater gains in health; however, the price tag has been largely ignored. At an average expenditure of $5,267 annually per person in 2002, healthcare costs are greater in the US than anywhere else in the world, with Switzerland a distant second at $3,446 per year.4 Greater spending has not translated to better healthcare overall, with a variety of measures, such as life expectancy and infant mortality, putting the US somewhere in the middle or even near the bottom compared to other highincome countries. Furthermore, if the rate of inflation in healthcare spending — recently 2.5% greater than overall annual inflation — continues, federal healthcare spending will require the entire current US federal budget by the year 2045.4 Clearly, as a society, we need to focus on healthcare value and not just absolute health improvements. The disconnection between healthcare costs and gains is the primary motivator for comparative effectiveness research. As far back as 2003, there has been a call to directly compare drugs in actual use, with a failed bill attempting to set up a modest fund for this purpose.5 Now 6 years later, healthcare costs have continued to increase with only modest measurable gains in health, and with a new administration and a focus on value, comparative effectiveness research has become a central component of healthcare reform. A6 Αnnals of Neurology Not everyone is in favor of a government-sponsored initiative for comparative effectiveness research. Some fear that a national program will usher in a new era of government intrusion into the physician-patient relationship, signal the beginning of healthcare rationing, and disincentivize the pharmaceutical, biotechnology and medical device industries. One argument goes as follows: what if a therapy is found to be only marginally effective at a population level but is more effective in some situations, for example in certain subgroups of patients or in some ethnic and racial groups? Medicare might use incomplete or flawed comparative effectiveness data to deny coverage of useful therapies, and private insurers would be tempted to follow Medicare’s lead. Although softening language has been added to the current bill, and its sponsors have stated that its purpose is not to support decisions about coverage, many Republican legislators and conservative groups remain in opposition. Although we often think of clinical trials as designed to introduce new therapies, many could qualify, at least in broad terms, as comparative effectiveness studies. To determine NINDS experience in funding trials that compared alternatives already used in clinical practice, we reviewed data from a prior study of the impact of the NINDS trial portfolio.6 Of the 31 trials with funding completed before 2000, 17 compared management options that were already being used in practice at the time, and therefore could qualify as comparative effectiveness research (Table). These trials accounted for $267 million of the total $335 million budget for the trial program, after inflation to 2004 dollars using the medical care component of the Consumer Price Index. Some of the ongoing large-scale trials in the NINDS portfolio are also comparative effectiveness studies, including the CREST trial comparing carotid endarterectomy to stenting, ARUBA comparing intervention to no intervention in patients with arteriovenous malformations, a trial performed in the Clinical Research Collaboration comparing propranolol and topiramate for migraine prophylaxis, and the CombiRx trial comparing the combination of glatiramer acetate and beta interferon to each agent alone. Thus, comparative effectiveness research is a major focus of trials funded by NINDS. The NINDS portfolio of trials comparing interventions already used clinically diverges from the target of proposed new programs in several respects. First, these trials all restrict entry to a small subset of the population treated in practice, and subjects are often monitored in ways that are inconsistent with routine care. This changes the focus to efficacy, rather than effectiveness, and potentially prevents generalizing re- Vol 65 No 2 February 2009 . Table. NINDS-Sponsored Trials Comparing Interventions Already in Clinical Practice Trial Title Randomized Indomethacin Germinal Matrix /Intraventricular Hemorrhage Prevention Trial Diazepam for Acute Repetitive Seizures Active Intervention Indomethacin Control Intervention Placebo Target Population Very low birth weight neonates Superior Intervention Active Cost of Trial $8,875,271.77 Diazepam rectal gel Placebo Active $1,563,303.38 Asymptomatic Carotid Artery Stenosis Collaborative Study Carotid endarterectomy Medical therapy Active $43,320,427.97 Stroke Prevention In Atrial Fibrillation I North American Symptomatic Carotid Endarterectomy Trial Warfarin or aspirin Placebo Acute repetitive seizure Asymptomatic internal carotid artery stenosis Atrial fibrillation Active $16,093,548.04 Carotid endarterectomy Medical therapy Active $64,033,234.16 Phenytoin Placebo Symptomatic internal carotid artery stenosis Post traumatic seizures Acute spinal cord injury Acute spinal cord injury Early Parkinson's disease Active $2,278,077.76 Active $6,330,642.45 Active $12,639,012.97 Active $34,015,598.08 Patients undergoing carotid endarterectomy Atrial fibrillation Control $3,920,504.47 Control $18,695,305.37 Control $29,198,825.66 No Difference $5,475,726.35 No Difference $5,753,116.22 No Difference $7,941,149.59 No Difference $3,105,610.96 No Difference $3,559,806.67 Dilantin for Seizure Prophylaxis after Brain Trauma National Acute Spinal Cord Injury Study II National Acute Spinal Cord Injury Study III Deprenyl/Tocopherol in Parkinson Disease (DATATOP) Methylprednisolone Placebo Methylprednisolone, long duration Tocopherol/ deprenyl Methylprednisolone, short duration Placebo Aspirin and Carotid Endarterectomy High-dose aspirin Low-dose aspirin Low-dose warfarin + aspirin Superficial temporalmiddle cerebral artery bypass Valproate Standard warfarin High-dose epinephrine Warfarin Low-dose epinephrine Aspirin Ischemic stroke or transient ischemic attack Post traumatic seizures Cardiopulmonary resuscitation Atrial fibrillation Methylprednisolone, high dose Percutaneous discectomy Methylprednisolone, low dose Conventional discectomy Acute spinal cord injury lumbar disc herniation Stroke Prevention In Atrial Fibrillation III Extracranial/Intracranial Arterial Anastomosis Study Valproate for Seizure Prophylaxis after Brain Trauma Brain Resuscitation Clinical Trial III Stroke Prevention In Atrial Fibrillation II National Acute Spinal Cord Injury Study I Conventional vs. Percutaneous Discectomy - A Clinical Trial Summary (total) Medical therapy Phenytoin $266,799,161.85 Trial costs were inflated to 2004 dollars based on the medical care component of the Consumer Price Index. sults to the full population actually treated in practice. New methods beyond traditional randomized trials are being developed to better evaluate treatments in routine use while attempting to control the difficult problem of confounding.7 Second, the primary focus of NINDS-sponsored trials remains the advancement of more effective therapies. Rather than comparing equally accepted alternatives, NINDS trials often compare one intervention that is used infrequently (often considered riskier or more invasive) with another that dominates practice. The agenda for comparative effectiveness research will have different priorities. With cost and value major foci, we will see more studies attempting to establish that a cheaper intervention is no less effective than a more expensive one. As demonstrated by our current and past research agenda, there are many questions in the clinical neurosciences that are excellent targets for comparative effectiveness research. In addition to the topics previously or currently studied, there are numerous clinical questions that come up daily and could impact neurological care. For example, when should new generation anticovulsants be favored over their predecessors? When is an echocardiogram worth the cost after stroke? Are newer sleep medications better tolerated than their predecessors and worth the cost? Who really benefits from lumbar diskectomy? These are the kinds of questions our patients and trainees ask frequently, and it will be nice to have evidence to answer them rather than relying on messaging from marketers and on our own personal experience. Clinical neuroscientists must be ready to offer questions and research plans in comparative effectiveness or the agenda may not adequately represent us. As the caretakers of diseases with tremendous impact on individuals and at a popu- © 2009 American Neurological Association . Published by Wiley-Liss, Inc., through Wiley Subscription Services A7 lation level, we have a responsibility to represent our patients during the birth of this field, particularly given the substantial investment expected as part of the economic stimulus package. Perhaps now is the time to begin to come together with a plan for participation and leadership. S. Claiborne Johnston, MD, PhD and Stephen L. Hauser, MD Editors References 1. Orszag PR. Research on the Comparative Effectiveness of Medical Treatments: A Congressional Budget Office Paper. Washington: Congress of the United States, 2007. 2. Arias E. United States life tables, 2004. Natl Vital Stat Rep A8 Message from the Editor 2007;56:1-39. 3. Murphy KM, Topel RH, eds. Measuring the Gains from Medical Research. Chicago: University of Chicago, 2003. 4. Wilensky GR. Developing a Center for Comparative Effectiveness Information. Health affairs (Project Hope) 2006;25:w572-585. 5. Lyles A. Comparative Effectiveness Research: NICE for the NHS, but False Starts for the US. Clinical Therapeutics 2008;30:1702-1703. 6. Johnston SC, Rootenberg JD, Katrak S, et al. Effect of a US National Institutes of Health Programme of Clinical Trials on Public Health and Costs. Lancet 2006;367:1319-1327. 7. Lohr KN. Emerging Methods in Comparative Effectiveness and Safety: Symposium Overview and Summary. Med Care 2007;45:S5-8.