Journal of the National Cancer Institute Advance Access originally published online on March 11, 2008
JNCI Journal of the National Cancer Institute 2008 100(6):384-385; doi:10.1093/jnci/djn070
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
© Oxford University Press 2008.
NEWS |
CLINICAL TRIALS THAT COST LESS
Electronic Health Records Help Recruit Trial Participants and Track Drug Safety
Recruiting patients for clinical trials can take months, and sometimes researchers still end up short. But what if they could enlist enough patients in half the time and conduct trials at a fraction of the cost?
Electronic health records may enable researchers to do exactly that. Researchers in Scotland will use Scottish and Danish electronic health records to recruit between 13,000 and 16,000 patients to compare traditional nonsteroidal anti-inflammatory drugs (NSAIDs), such as ibuprofen and diclofenac, with the newer NSAID celecoxib. The trial is called the Standard Care Versus Celecoxib Outcome Trial (SCOT). No one has ever compared each NSAID head to head in a long-term trial.
"Electronic health records offer intriguing possibilities for clinical trials," said John Norrie, director of the Centre for Healthcare Randomised Trials at the University of Aberdeen in Scotland. Countries with national health services are extremely interested in knowing how effective one treatment is over another. But they don't have the resources to fund huge trials, so electronic health records enable them to answer questions about treatments at a reasonable cost, he said.
Pfizer, the maker of celecoxib, is footing the bill for the SCOT trial—an unusual move because drug makers traditionally don't fund studies that compare already marketed drugs or existing treatments, said Norrie. Because of potential side effects from NSAIDs, such as heart attack, stroke, and gastrointestinal bleeding, the European Medicines Agency asked Pfizer to sponsor a European study of NSAIDs in patients without cardiovascular disease.
Arthritis patients identified through electronic records will be assigned to receive any one of the NSAIDs at random. Using Scotland and Denmark's electronic databases will speed up the search for suitable participants, and follow-up will be done electronically, making the trial less expensive than a traditional trial. Researchers will track the number of NSAID prescriptions filled, whether a patient has been hospitalized for cardiovascular disease or a gastrointestinal illness, and whether the patient died.
In the United States, Pfizer is also sponsoring a sister trial to compare NSAIDs in 20,000 patients with established cardiovascular disease who are being treated for arthritis. The trial, called Precision, began in 2006 and is being led by researchers at the Cleveland Clinic. Patients are randomized to receive any of the NSAIDs, and neither doctor nor patient knows who is taking what.
Although both trials have the same goal, they are complementary, said Annlouise R. Assaf, Ph.D., senior director of epidemiology at Pfizer. Both trials will reflect patient care in the real world, she said, "but the SCOT trial is even more real world" because the patients are not blinded. Both doctors and patients know which drug is being prescribed, and their care is delivered by general practitioners. Patients may also be taking other drugs, too, which is as close to natural care as possible.
This approach is an advantage because most drug trials select participants carefully and take place in tightly controlled environments, said Tom MacDonald, M.D., Ph.D., lead investigator of the SCOT trial and professor of clinical pharmacology at Dundee University in Scotland. "But we don't know how translatable those are to the real world," he said.
|
Using electronic health records to conduct clinical trials is not a new idea. In fact, just last year researchers at the University of Glasgow in Scotland published a 10-year study that compared the drug pravastatin to placebo for reducing the risk of death from heart attack. That study used the nation's electronic health records.
But few countries have a standard platform for all health records, said MacDonald. Since the 1960s, Scotland's doctors have kept electronic health records on all patients. One large central computer was all that was necessary to manage all the information because Scotland's population is only 5 million. "Today you could probably store that information on a single desktop," MacDonald said. The nation's small size has made it relatively easy to create an integrated system whereby all computer databases, including those from hospitals, practitioners, and pharmacists, talk to each other. In Western Europe, only Denmark and Holland have integrated systems. Sweden, which also maintains extensive electronic health records, is linking them into one system.
In 2004, President Bush called for all health care workers to adopt electronic health records to create a National Health Information Network in the U.S. by 2014. Since then, health care organizations, community health professionals, and technology vendors have turned their attention to creating the strategies, protocols, and computer information technology infrastructures necessary to make his vision a reality.
But progress has been slow. Part of the reason is money. Experts estimate that wiring the country's entire health care system would cost more than $160 billion, and many small physician practices and hospitals have balked at the cost. They also fear that adopting electronic health records will disrupt their work and increase malpractice liability because of data loss, mistakes during data entry, or unauthorized use of patient records.
Nevertheless, a few small-scale electronic health record networks do exist at so-called integrated health systems that combine hospitals, clinics, laboratories, doctors, and nurses. Some, like the U.S. Veterans Affairs health system and Kaiser Permanente, a health maintenance organization, both insure people and provide health care. Such networks have been used to conduct observational studies that examine the risks and benefits of certain drugs.
But these networks aren't always useful, Assaf said. They don't necessarily include all drugs or treatments that are available but rather only those that they cover. Also, some haven't existed long enough to provide adequate numbers of patients and outcomes. Such databases weren't conceived with clinical research in mind, and so important data, such as social or behavioral information, are often lacking.
Existing electronic health databases do, however, lend themselves to certain types of trials, MacDonald said, such as comparing competing or alternative therapies for treatable diseases. Moreover, such research costs about one-fifth as much as trials in which patients are recruited and followed up, MacDonald added, because they speed up patient recruitment and all follow-up is done electronically.
They can also track a drug's long-term safety, which has been an important issue for regulators, MacDonald explained. "Electronic health records provide a mechanism to study long-term safety, and it's a mechanism that both regulators and companies appear comfortable with and can actually be done at a reasonable cost."
But electronic records aren't the clinical trial panacea. "While useful to answer certain questions, electronic health records could never replace rigorous de novo research because they don't consistently take measurements that we expect in clinical trials," Assaf said. Moreover, because only small countries have integrated systems, each database contains too few patients to study treatments for rare diseases.
That limitation may change in the future, however, as more nations work toward digitizing health records and devising standards for software. Pfizer, in fact, is participating in several initiatives that aim to adopt standards so that the world health community would one day be able to share information by using a common language.
Ultimately, patients are the ones who will benefit the most from these studies. By the time the SCOT trial is over in 2012, Pfizer's patent on the drug will expire within 2 years.
"This is a big deal for anyone who will prescribe NSAIDs in the future," said MacDonald. "We're going to know which is the better treatment when the drugs become generic."
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||