June 15, 2011 (Vol. 31, No. 12)

Gail Dutton

Electronic Records May Reinvent Clinical Trial Operations and Personalized Medicine.

The convergence of commercial genomic kits and electronic medical records is helping to drive the healthcare industry’s focus on personalized medicine. Although it’s quite early, observers suggest this integration of technologies has the potential to change the way therapeutics are developed.

“We’re very, very early in the process, but we’re at the point of critical mass,” notes Mark Dente, M.D., chief medical informatics officer at GE Healthcare IT. GE has “50 or 60 different technologies from the operating room, to the emergency room, clinical research, and remote monitoring,” that are coming together to create what is being called the electronic patient.

“We’re at a nexus that will change the game,” improving quality of care as well as presenting new paradigms for the development of drugs and devices.

GE Healthcare IT has already developed an anonymized database of 18 million lives that complies with the strictures of the Health Information and Patient Portability Act (HIPPA). GE has both standardized the terminology within the database and normalized its data.

This database is already being used, according to Dr. Dente. For example, electronic medical records (eMRs) can be used early on in clinical trials. “We believe we can save four months between the times the first and last patients are enrolled in a trial,” he says. The time is saved by conducting protocol modeling to identify the available patient mix, and then adjusting the protocol accordingly. That simple change can dramatically alter the number of patients eligible for a given study.

As Dr. Dente points out, physicians can apply for federal funds to deploy eMRs in their practice, with the caveat that they meet the regulations for “Meaningful Use” set out by the Office of the National Coordinator under HHS.

ePatients also makes it possible to conduct virtual clinical trials, adds Tim Harris, Ph.D., director of the Advanced Technology Program at SAIC-Frederick. For example, a pharmaceutical company might work with physicians to identify anonymous patients in their electronic database who have tumors with a particular mutation and then compare the effectiveness of various therapies in different patients, either by following their progress or by making relevant clinical trials available.

Biosurveillance is another possible application of eMR data. Because the databases function in near real time, they can be used to trace adverse events as well as to identify off-label uses and track actual use of a given therapeutic. GE Healthcare IT has already used its database to help the Centers for Disease Control to identify patients with symptoms similar to H1N1 and to provide co-morbidities to identify outbreak areas. It also has identified food-borne illness and provided real-time alerting.

The availability of eMR may, eventually, cause researchers and regulators to think differently about clinical trials. “What if we could do a controlled release to 500 patients, monitor their health, and see the effectiveness of new medications, rather than enroll them in traditional Phase III trials?” Dr. Dente asks.

Repurposing existing monitoring technology could results in trials in which the participants use home-based technology such as electronic scales, spirometers, blood pressure monitors, or heart monitors to provide that information nightly, rather than visiting a physician weekly.

“That capability is available now,” he says. GE is talking with pharmaceutical companies and waiting for the right opportunity to put this into practice.

“To keep innovation going, the model for introducing innovation needs to change,” according to Harlan Weisman, M.D., chief science and technology officer for medical devices and diagnostics at Johnson & Johnson, and a member of the board of governors for the Patient-Centered Outcomes Research Institute (PCORI).

As he explains, “The FDA is raising the bar for needed clinical data before granting product approvals,” causing more patients to be studied during longer time frames. “These increasing requirements can discourage development of important new innovations.”

“New technologies such as eMRs and emerging methods for conducting outcomes research with large, real-world databases offer the possibility of an alternative approach, in which pre-approval trials demonstrate an acceptable level of safety and efficacy for a new product, based on the seriousness of the condition and the relative unmet need.”

Eventually, Dr. Dente adds, eMRs may play an active role in adverse event reporting in which events like a change of medicine class prompts the eMR to ask where this is triggered by an adverse event. If so, the eMR may populate the report form, which the physician verifies before sending an electronic report.

But, Mark Hoffman, Ph.D., vp of research solutions at Cerner, notes, “Genomic data is only beginning to be incorporated into formal medical records. It’s an important trend to monitor.”


“Personalized medicine is an ongoing process that goes beyond genetics and helps drive more effective therapies. We’ll be seeing more genetic-based diagnostics,” Dr. Hoffman says. The convergence of IT and genomics will accelerate that process. “If you can work with an organization that has electronic medical records, there would be a potential advantage,” Dr. Harris adds. “I think that would make it much easier to acquire phenotype data. The economic benefits should be very large.”

There are other benefits, too. “As we refine the information we’ll identify gaps in real-world uses, in terms of gaps in diagnostics and therapeutics, which will present opportunities to developer companies,” Dr. Weisman predicts.


Realizing the promise of eMRs depends to some extent upon continually developing standards, allowing greater interoperability of data exchange among competing eMR systems. Standards have not yet evolved to make all data portable across systems.

At GE, “We are increasing our nomenclature and terminology research for our customers as well as making that part of the public domain. We look forward to presenting that to standards organizations,” Dr. Dente notes.

What, exactly, constitutes an electronic medical record has yet to be determined. “There’s much less debate about what type of information we want to include in the database than about the ramifications of having eMRs,” Dr. Harris says.

At their most basic, “eMRs are electronic ways of holding data so patients don’t have to fill in forms again. Even having a summary level of information will save the healthcare system time and money.” They also can be much more. “How deep do we want to go?” Dr. Harris asks, pointing out that eMRs could store detailed information, or could link to other sources—such as hospital picture archiving and communication systems (PACS)—for images and test results. “eMRs needn’t contain all the information about an individual,” Dr. Harris says, but they may, in the future, store DNA sequence information.

For electronic clinical trials, they may be more valuable if they do contain comprehensive, yet de-identified, information. Organizations already access insurers’ databases for information, but an eMR would provide immediacy.

The PCORI is being formed “to add the patient to the equation. As a cardiologist, I can speak about clinical trial results, but people want to know what’s right for them rather than statistical averages,” Dr. Weisman stresses. PCORI, once operational, will develop systems that provide information in near real time based upon medical conditions and such additional factors as lifestyle and geographic and occupational risks. Eventually, “You (as a patient) will be able to query and get information about people with profiles similar to yours, and what happens to them when they choose different options for prevention, diagnosis, and treatment,” he predicts.

Organizations like 23andMe and others that provide genetic testing directly to consumers, are filling in some of the information gaps. As Anne Wojcicki, CEO, explains, the information allows patients to tell their physicians they have the markers for particular conditions, but 23andMe doesn’t sequence the genome. “While there’s huge government pressure to put medical record information online, there’s not yet a movement toward including genetic information in electronic medical records.”

23andMe is a retail DNA testing service providing information and tools for consumers to learn about and explore their DNA. The first step is to order a kit from the company (shown). Once the kit is received, it should be registered, and the tube spit into and returned to the lab. A CLIA-certified lab analyzes the DNA in six to eight weeks and then users can start exploring their genomes.

Probability Analysis

Making data available electronically is only one part of the challenge, though. “The critical skill is the integration of genotypic and phenotypic information,” stresses Don Rucker, M.D., CMO at Siemens Healthcare USA. The convergence of IT and genomics involves far more than adding gene sequencing information to individuals’ medical files. “The whole nature/nurture boundary is far more confused than anyone thought.” Therefore, for the convergence to have meaning, Dr. Rucker says a probability analysis is vital.

That analysis must be based upon all the available information—genomics, proteomics, imaging, blood tests, etc.—to develop a better sense of what is being expressed and what it means to individual patients. For example, Dr. Rucker points out that while the presence of the 9p21 gene increases the probability of heart disease to about 50%, a cardiac CT angiogram can now determine the probability of coronary disease to a level of 98% or 99% accuracy.

“For the pharmaceutical industry, the bad news is that there is more data to sift through. The good news is that there is great confidence about the data and the ability to develop more targeted datasets, potentially lowering clinical trial size,” Dr. Rucker concludes.

According to Siemens’ researchers, adding sequencing information to individuals’ medical files is not enough. A probability analysis based on all the available information, including imaging and blood tests, is essential in order to get a better sense of what is being expressed and what it means to individual patients.

Previous articleScientists Find Ovarian Cancer Cells Barge Through Mesothelium to Invade New Organs
Next articleAre Early Clinical Successes Enough to Bring RNAi Back from the Brink?