CDRH instructs sponsors on how to collect patient preference information, and then incorporate that data into device submissions seeking FDA approval in a May 13 draft guidance.

Simultaneously, the public-private Medical Device Innovation Consortium issued a report outlining strategies for incorporating patient preference data into their device application submissions. (See Also see "Device Consortium Tries To Shed Light on Patient Risk-Benefit Preferences" - Medtech Insight, 13 May, 2015..)

The FDA guidance specifically addresses appropriate mechanisms for applying patient tolerance for risks and benefits, in PMA, humanitarian device exemption or de novo submissions (but not 510(k)s), as well as how to incorporate the information into device labeling. It supports the device center's growing emphasis on formally leveraging patient input as a factor for regulatory decisions, even to the point of allowing patient preferences to push a device over the approval line when FDA reviewers don't otherwise conclude that the benefits outweigh the risks.

“We might feel the risks outweigh the benefits, or it is a very tough call [on a device]," explained CDRH Director Jeff Shuren at a May 13 Medical Device Innovation Consortium meeting in Washington, D.C. "But when there is a clearly identifiable patient population or subpopulation for whom the technology would be indicated, and they have a different value preference … then we would approve that device.”

CDRH took this approach in its January approval of EnteroMedics Inc.'s Maestro obesity device, which did not meet all of its pivotal endpoints but it was supported by validated patient survey data showing a higher tolerance for risk in exchange for weight loss.(See Also see "FDA Weighs Patients' Risk Tolerance in Approving Obesity Device" - Medtech Insight, 14 January, 2015..)

"Now the question is how can we do this more systematically," Shuren said; the guidance, he noted, was a key step to make that possible.

Considering Patient Differences

When considering patient preference data in a submission, FDA will examine study quality factors such as the representativeness of a sample, the differences between, or heterogeneity, of patient preferences, established good research practices and patient centeredness, the guidance says.

“We list out some examples of what we see as potential preference-sensitive decisions in this guidance,” said Anindita Saha, CDRH director of external expertise and partnerships, at the MDIC meeting.

She added, “This is a nascent field and there’s still of lot of work to be done in terms of the methodologies so we’re not prescribing you need to use method A or method B … but rather, these are the types of qualities we’d like to see in your patient preference studies.”

For example, sponsors need to use an adequate size sample to ensure that study results can be generalized to the population of interest, according to the guidance. In an example from CDRH’s study of patient preference for weight loss devices, the sample included 500 patients drawn from an online panel that represented a cross section of the U.S. population.

Also, individual circumstances of different patients vary. Besides gender, age, race, socioeconomic and cultural background, “a patient’s own experience of his/her disease may influence the patient’s personal risk tolerance attitude,” the guidance states. It recommends that sponsors use either a sufficient sample size, or discrete choice experiments to effectively identify and quantify preference heterogeneity.

“Patient-Centeredness” means that the patient – not just the health care professional – is the central part of the study, FDA writes in its guidance.

Studies are more likely to be accepted by FDA if they follow good research practices set by a recognized professional organization. The agency pointed to good research practices established by the International Society for Pharmacoeconomics and Outcomes Research for use of discrete choice experiments to determine patient differences, for example.

FDA also wants sponsors to be certain that the patient – rather than the health care professional – is the central part of the study, which should measure “inherent values attitudes and values of well-informed patients,” and also caregivers’ and parents’ preferences.

Effective communication, and clear health numeracy – being certain that patients can understand and use the numbers presented by their health care provider about device choices – figures prominently in design of patient preference studies, FDA wrote.

To help patients understand risks and benefits of device choices, sponsors and providers should avoid presenting options using fractions and decimals, and if possible, describe uncertainty in both positive and negative frames -- such as a 20 percent chance of adverse events, or 80 percent chance of good outcomes. Also, providers should be encouraged to use an array of formats when presenting risk options to a patient, including verbal frequency, percentages, and icons or pictograph.

Labeling Considerations

FDA focuses on describing outcomes of patient preference studies in a section of the guidance on “general labeling recommendations.” Among label elements sponsors may want to consider, is a specialized informed consent section, so that patients may explicitly accept the probable risks of the device, in exchange for probable benefits, the agency said.

FDA states that health care professional labeling should include a summary of the patient preference study describing the population studied, the method used to elicit patient preferences, attributes and levels of benefits and risk included in the study design, as well as its results.

A label might also assist patients in understanding about the device if it includes elements such as benefits, and risks and complications.

Additional Patient Labeling Recommendations

FDA’s draft patient preference guidance states that good patient labeling should tell patients:

If they might benefit from use of the device;
The potential benefits from use of the device;
The potential risks or complications from use of the device, and the likelihoods of each;
Any relevant contraindications, warnings and precautions;
If they share characteristics with the group of patients who view benefits as outweighing risks; and
Any additional information about what is known and not known about patient outcomes.

Among the examples that FDA provides in the draft illustrating patient preference choices and its role in decision-making is a scenario with a permanently implanted aesthetic device is intended to improve body appearance, and is studied in a healthy patient population.

For this device, data from the clinical trial suggests similar body improvement benefit as marketed alternatives, but faster recovery from implantation surgery. But a higher rate of meaningful adverse events was observed, including the need for reoperation to replace it, with lesser improvement in body appearance. Patient preference data shows some patients place a high value on appearance enhancement, and would accept the higher level of risk.

But FDA may conclude the device poses an unreasonable risk of illness or injury, so may decide NOT to approve the device despite the positive patient preference information. The agency may recommend that the sponsor explore design and manufacturing process changes, to improve the benefit-risk profile, and win device approval.

FDA wants feedback on the draft guidance, said Saha, and has set a 90-day comment period. “We’ll also be working on training, because we are going to be getting more of these studies," she added. "We want to make sure the rest of our reviewers and other folks at the center understand the studies because we do want to make sure we are able to scrutinize [them] … and make sure it’s a valid study.”

Guidance: What The Patient Prefers, And How To Make It Part Of Your FDA Submission

More from Regulation

More from Policy & Regulation