Using EHRs in research poses privacy concerns
Overcoming barriers for use
Electronic health records (EHRs) offer a rich resource for facilitating clinical research by identifying patients who fit the eligibility requirements for a study and allowing researchers to collaborate with primary care physicians to recruit them.
But taking advantage of this resource requires researchers to navigate a complicated path of regulations and institutional policies intended to protect patient privacy. And those policies are not always consistent, says J. Thomas Bigger, MD, a professor of medicine and pharmacology at Columbia University in New York City, who serves on the faculty of the university's Center for Bioethics.
"The regulations are not that crystal clear," Bigger says. "They tend to be interpreted different ways in different institutions. It depends on who's doing the local regulations, whether their perspective is to protect the institution at all costs or whether they have a broader view of it."
Bigger and Chunhua Weng, PhD, MS, an assistant professor of biomedical informatics at Columbia, say human subjects protections concerns can be addressed while still making it easier for researchers to find and approach potential research subjects.
They say policymakers can assist by clarifying HIPAA privacy rules and harmonizing them with other federal human subjects protection regulations. IRBs, Weng and Bigger say, can do their part by ensuring that local policies are not so stringent that they preclude commonsense research activities, particularly those preparatory to the research itself.
"We observe a lot of inconsistencies regarding who has access to patient data for trial pre-screening," Weng says.
Their recommendations for facilitating the use of EHRs in research are included in an article in a recent issue of the Journal of the American Medical Informatics Association1.
Identifying potential subjects
Weng and Bigger note that IRB concerns are not the only barrier to more effective use of EHRs in research. They identify other problems, such as the completeness and accuracy of the records and the drain on physicians' time involved in assisting researchers. Bigger, himself a clinical trial researcher, says the days are gone when he could buttonhole a colleague at the elevators and pitch a trial to him without interruption.
"Now, when the first elevator came, they'd be gone," he says. "They're late for something. They're always late for something. They don't have the luxury of time they used to have."
But even that problem, he says, can be eased by relaxing the restrictions on who can view medical records in order to determine whether patients are eligible for trials.
Currently, Weng says, her own institution's IRB can grant waivers for access to EHRs to search for potential subjects. This information is usually de-identified, but can still prove useful — with physicians' cooperation, Bigger says.
"You can search the medical records to see how many eligible patients each clinician might have. [The clinicians] get the list with names and IDs on them, while the researcher just gets the counts — Clinician A has 60 patients, Clinician B has 12.
"You can see that some clinicians have so few eligible patients it wouldn't be productive for anybody concerned to pursue it any further," Bigger says.
And he says it is important for researchers to be able to communicate with physicians to determine if patients who are found electronically are really a good fit with a clinical trial. Physicians often know information about patients — terminal illness, alcohol use, non-compliance with medications — that isn't in the record but may make them ineligible.
"A lot of doctors won't put sensitive stuff in electronic medical records," Bigger says. "But they know the reasons this patient wouldn't be suitable for the trial."
He and Weng say HIPAA regulations would allow this sort of preparatory activity without patient authorization or an IRB waiver, as long as a researcher is a member of the covered entity, but that many institutions interpret the rules more strictly.
Weng and Bigger say some IRBs also impede research by requiring that researchers obtain permission from the patient's primary care provider (PCP) before they approach a subject. Again, they say this is not required under federal regulations.
This can cause problems, particularly if it is not clear from the EHR who is the patient's PCP. Weng says she knows of a researcher who ran into difficulty because a subject had more than one PCP.
"One patient would be involved in a clinical trial study, and then two weeks later, an unknown primary care provider called the researcher and got angry, saying this patient should not be participating in the study," she says.
"There's no clear rule regarding how to resolve those conflicting opinions of different primary care providers regarding research participation."
She and Bigger argue that while a PCP's authorization may be a reasonable requirement in higher-risk studies, it's not necessary for all research.
'Falling into a crack'
Once a patient is actually enrolled in a trial, the privacy rules continue to cause difficulty, they say, making it hard for researchers to access scheduling in order to coordinate research and clinical visits. Just as importantly, researchers often cannot enter data in EHRs, and so can't show, for example, the study medications a patient is taking.
"We see patients falling into a crack, separated by the clinical care team and the clinical research team," Weng says. "Each team only takes care of a patient's certain activities. So it's hard to detect drug/drug interactions between an experimental drug and the regular treatments."
Bigger says there needs to be a better method of sharing information in both directions, between the researchers and the clinicians.
"You may have to do it a little differently with every physician in every study," he says. "But if you don't come to a wide-open communication among all the parties, you're just asking for trouble, of one kind or another — misunderstandings or really serious interactions with the treatments being studied."
They say informaticians could create sophisticated means of allowing two-way communication while protecting patient privacy, but that it likely will require policy changes at the U.S. Department of Health and Human Services level to accomplish.
While much of the burden of solving problems associated with research use of EHRs lies with policymakers, they say IRBs can help by showing flexibility.
"I hope the local IRBs can appreciate the diversity of research settings," Weng says. "Some researchers are clinicians themselves, so they're enrolling their patients or their colleagues' patients. Other researchers may not have a pre-existing relationship with patients. If we just have one single, one-size rule trying to fit it all, that becomes kind of difficult."
Reference
- Weng C, Appelbaum P, Hripcsak G et al. Using EHRs to integrate research with patient care: promises and challenges. J Am Med Inform Assoc 2012 Apr 29 (epub).
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content