Earlier this month, the FDA released a letter announcing a new focus on radiation-emitting products. Here's the core of the letter:
In order to reduce the number of under-doses, over-doses, and misaligned exposures from therapeutic radiation the FDA is taking several steps to improve the safety and safe use of certain radiation therapy devices. Analyses of Medical Device Reports (MDRs) revealed device problems that appear to be the result of faulty design or use error that could be mitigated by the incorporation of additional safeguards. Between December 31, 1999, and February 18, 2010, FDA received 1,182 MDRs associated with the use of radiation therapy devices. Of these MDRs, linear accelerators accounted for 74%, radiation therapy treatment planning systems (RTP) accounted for 19%, and ancillary devices (e.g., radiation therapy simulators) accounted for 7%. The most frequently reported device problems were computer software issues, use of device, and incorrect display. In some reports, the manufacturer was unable to determine or identify the problem and reported the problem as “unknown.”
A separate analysis of these MDRs for software problems identified 362 MDRs. Of these MDRs, linear accelerators accounted for 66%, RTP accounted for 29%, and ancillary devices accounted for 5%.
It is important for manufacturers to investigate the cause of nonconforming product and analyze factors in addition to use error as part of Corrective and Preventive Actions (CAPA). Such analysis may include an assessment of the correlation between product user interface, controls, or user information and use error.
FDA plans to hold a public workshop on radiation therapy treatment planning, medical linear accelerators, and ancillary devices. Additional information will be published in a Federal Register notice. Topics will include:
- New safeguards and other special controls to improve safety;
- Possible changes in premarket device testing to provide appropriate assurances of safety and effectiveness, particularly for software; and
- Premarket review of all modifications to software.
FDA encourages manufacturers to attend the public workshop to discuss these issues.
The top issue identified by the FDA in this letter is software, and the proposed solution includes possible premarket testing of software (which FDA asserts comes within the definition of "device.") Issues concerning inappropriate radiation doses raised a couple of months ago extend to the use of diagnostic radiology as well, and have led to the proposal for the establishment of a central radiation dose registry. (Testimony at the FDA hearings on diagnostic radiology included a presentation from MITA's Executive Director, Dave Fischer, who was interviewed here at HealthBlawg the week before the hearing, in early March. A second public meeting was held March 30-31.)
The NY Times reported upon the issuance of the FDA letter:
Dr. Howard I. Amols, chief of clinical physics at Memorial Sloan-Kettering Cancer Center in New York, said the F.D.A.’s action did not address what he believed were more serious problems stemming from shortcomings in staffing, personnel competency and hospital quality assurance programs.
“I’d also caution that however commendable tougher standards for premarket approval of software may be, its not clear that F.D.A. has the expertise to police this,” Dr. Amols said. “In fact, I’m not sure anybody does. That’s one of the big problems with software. It comes down to a qualified user recognizing that something is amiss.”
Questions about the FDA's ability to regulate software are of great concern since it has asserted the right to regulate software as a device, not only in the context of radiology, but also in the context of electronic health records.
In the EHR context, Jeffrey Shuren, Director of FDA’s Center for Devices and Radiological Health, has signaled the agency's willingness to step back from asserting the right to require premarket approval of all EHR software — good news for the market, though maybe not so good for some of the big players in that market, since a PMA requirement could debilitate the market, and limit entrants to large organizations able to bankroll the PMA process and longer time-to-market. See Shuren's testimony at the February 25 HIT Policy adoption/certifciation workgroup meeting, and the transcript, too, for starters.
The idea that the FDA would exercise regulatory restraint and coordinate with the ONC, so that ONC could incorporate FDA requirements as part of the definition of Stage 2 Meaningful Use is very encouraging (two agencies within HHS coordinating their efforts … wipeout concept, man). I am encouraged that there can be use made of existing systems within government to capture and process incident reports — though, of course, Shuren makes the point that reporting is only the tip of the iceberg. There needs to be a commitment to addressing issues, and increased transparency. The point was made at the workgroup meeting that EHR license agreements constrain communications among providers about problems and potential improvements. It seems to me that the industry should be encouraging, not restricting, that sort of communication and opportunity for continuous improvement, rather than face more draconian governmental controls.
Bottom line: Both with respect to the software controlling radiation-emitting diagnostic and therapeutic technologies, and EHRs, there are alternatives to FDA full-bore PMA regulation of the software as devices — alternatives that can encourage a quality improvement mindset even in an arena where we need to have zero tolerance for errors. The challenge is to bring the quality home at a price point we can live with. The PMA approach (like the Pentagon's former build-it-from-scratch model vs. current buy it off-the-shelf model when it comes to computer hardware and software acquisitions) guarantees higher process costs without guaranteeing better outcomes.