Breast implants, Hippocrates, and burden of proof.

The author argues that the burden of proof for the safety and efficacy of some medical devices properly lies with the manufacturers of the devices and with the providers who use them. The controversy over breast implants is used as a case for this shift in the burden of proof of safety and effectiveness.

The public and scientific debate regarding the safety of silicone gel breast implants is building. The implant is a "simple" device cleared for marketing through the 510K process at the Food and Drug Administration (FDA). Data on safety and effectiveness were not required by FDA to substantiate the request for marketing. Rather, the sponsor (manufacturer) was required only to substantiate the claim that the implant, was "substantially equivalent" in technical characterics and intended used to devices on the market prior to May 28, 1976. Four manufacturers were successful in persuading the FDA of this substantial equivalence, and, as a result, more than one million women have been implanted with the 510K devices.

The purpose of the following discussion is not to attempt to make a determination one way or another regarding the specific safety issue. Rather, the bigger issue is with whom does the burden of proof lie. Indeed, the case of the silicone gel breast implant argues persuasively that patients who receive such devices and payers who reimburse for them should insist that the burden of proof for establishing the safety and effectiveness of health care technologies liw with the manufacturers who market new devices and with the health practitioners who advocate and expose patients to new procedures.

FDA, with its tremendous work load, has done a yeoman's job over the years in ensuring the safety and effectiveness of the drugs and devices made available to practitioners and patients. However, for some devices, such as the implants in question, and for procedures in general, there are no rigorous requirements for demonstration of safety and effectiveness. In the case of breast implants, a combination of an accumulation of anectodal reports of harm and alleged inconsistencies in manufacturers' data generation and collection have led FDA to demand safety data that would argue that the devices be permitted to remain on the market. The response on the part of practitioners and manufacturers has been to assert that FDA has not presented sufficient data on the occurrence of adverse events to warrant a "voluntary moratorium" on the use of implants. Here, the question in public policy is who owns the burden of proof; the provider (practitioner and manufacturers), the community, or the regulatory body that represents the patient. During the 1980s, payers provided their own answer to the question. They established the coverage decision as a formidable hurdle for devices and procedures and identified cold, hard data substantiating safety and effectiveness as the only catapult available to surmount this new barrier to technology diffusion. For manufacturers and increasingly for practitioners, recognition of this demand for data has resulted in some proactive efforts to generate and collect such data.

The case of silicone gel breast implants points to the need for patients themselves to demand such data prospectively. Eighty percent of those one million women have had silicone gel breast implants inserted into their bodies for cosmetic reasons. If they had been fully informed, with a clear presentation of the "risks and benefits" that attended cosmetic implantation, what would their decisions have been? More to point, if and when these risks are clearly defined, what course of action will these women choose? Will significant expenditures of health care dollars be required to remove silicone gel breast implants should forthcoming safety data argue persuasively for such an intervention? What further risks will surgical removal of a cosmetically placed device pose? Whose responsibility will it be to pay for removal of cosmetically placed devices? What will we learn from this experience?

It is interesting that safety is in question here. So much has been written about the need for effectiveness research in the past five years that there is a tendency to subordinate the risk portion of the risk/benefit (safety/effectiveness) equation. Hippocrates himself if argued for such outcomes research long before Eddy, Donabedian, et al. As this public, scientific, and clinical issue is debated, the advice of Hippocrates to first "do no harm" takes on special significance. It reinforces the need to have a clear demonstration of safety and effectiveness before as wide a distribution as has been achieved by even such a seemingly inert and innocuous device as a silicone gel breast implant occurs.

FDA, through this experience and with the mandate of the Safe Medical Devices Act, very likely will establish more rigorous requirements for the so-called 510K devices. However, how can we ensure that the safety and effectiveness of procedures such as extracranial-intracranial bypass, carotid endarterectomy, laparascopic cholecystectomy, and high-dose chemotherapy-autologous bone marrow transplantation have been established before widespread diffusion occurs? This question is not easily answered. This is due to the lack of a major sponsor (manufacturer) with a profit motive and with a legal mandate to pass through a regulatory body's approval process. Notwithstanding this, the burden of proof will continue to appropriately shift to providers. Some parts of organized medicine have begun to respond by establishing formal registries to collect outcomes data or by cooperating with federally funded clinical trials. It is incumbent upon practitioners to more thoroughly assess and substantiate the practices and applications of technologies that they promote and apply.

The experience with and the debate over the silicone gel breast implant underscores the fact that precious little can be assumed with even the seemingly most simple device. Examples such silicone gel breast implants remind us not to ignore or presume the safety profile of an intervention. The admonitions of Hippocrates help in placing appropriate responsibility for assumption of the burden of proof.

William T. McGivney, PhD, is Head, Technology Assessment and Policy, Aetna Health Pans, Hartford, Conn. The views expressed in this article are those of the author and do not necessarily reflect those of Aetna Health Plans.