The Medical Device Amendments of 1976: The Statute That Went Awry

By Jeffrey K. Shapiro –

The Medical Device Amendments  of 1976 (“MDA”), Pub. L. No. 94-295, 90 Stat. 539 (1976), are commonly described as the beginning of the modern era of device regulation.  In one sense, this description is absolutely correct.  The MDA established for the first time a comprehensive scheme for the premarket and postmarket regulation of devices.  But, there has been less recognition of how little of the MDA actually survived contact with reality.  Many of the most prominent features of the statutory scheme did not function as expected and many proved to be dead letters almost from the beginning.  At the same time, one of the features that came to be most central – the 510(k) procedure – was expected to play a minor and transitory role.  The system we know today really has to be seen as originating in the MDA but not emerging with its present day contours until the Safe Medical Devices Act of 1990 (“SMDA”), Pub. L. No. 101 629, 104 Stat. 4511 (1990).  However, our story for today is about the fate of the major provisions of the MDA as originally enacted.

The MDA decreed that FDA would review all types of medical devices existing in 1976 and by regulation place them in Class I, II or III.  See 90 Stat. 540-41.  Class I devices would be subject to various general postmarket controls (e.g., establishment registration, device listing, good manufacturing practice, or GMP).  Class II devices would be subject to FDA established performance standards and general postmarket controls.  See 90 Stat. 546 552.  Class III would need premarket application (“PMA”) approval or a completed product development protocol (“PDP”) and would adhere to general postmarket controls.  See 90 Stat. 553 59.

After the existing device types were classified, all new device types developed after 1976 were to be placed in Class III, unless FDA could be persuaded to reclassify them to Class I or II.  Flexible reclassification of device types would occur as new information emerged with experience.  See 90 Stat. 544 45, 547, 553, 572.  FDA was granted broad authority to issue regulations restricting the sale, distribution and use of specific device types as appropriate.  See 90 Stat. 567.  FDA also could ban unsafe devices and/or require mandatory recalls and repair, refund, and notification remedies when needed.  See 90 Stat. 560, 562-63.  In short, FDA was to have ample statutory authority to conduct flexible, risk based regulation of the three classes of devices from cradle to grave.

Almost nothing went according to the statutory plan.  The mere classification of existing device types into Class I, II or III, took 14 years to complete, far longer than envisioned.  FDA struggled to develop performance standards, which had been mandated as the centerpiece of regulatory control over Class II devices, issuing not a single performance standard prior to the SMDA, in which Congress mercifully downgraded performance standards to one of several optional “special controls” for Class II devices.  All of the device types placed in Class III were supposed to be subject to a prompt call for PMAs, but that process went so slowly that even today (i.e., 37 years later) it is not complete.  Thus, for many years, “preamendment” Class III devices have reached the market via 510(k) clearance rather than PMA review, and that is still true for about two dozen of them. 

Numerous other provisions failed to function as intended.  The PDP option, intended to be on par with PMA approval, was a complete flop, although the detailed PDP provisions remain on the books even today.  A procedural regulation for banning devices was issued, but the authority was invoked once in 1983 (banning prosthetic hair fibers), and never again.  See 21 C.F.R. Part 895 (1979).  The refund authority has never been used.  FDA has only issued two restricted device regulations – one for hearing aids and the other for analyte specific reagents.  See 21 C.F.R. § 801.421 (1990); 21 C.F.R. § 809.30 (1997).  The reclassification procedures proved burdensome and slow, and have been rarely used.  (A recent statutory reform attempts to make this process less burdensome.  Food and Drug Administration Safety and Innovation Act ("FDASIA"), Pub. L. No. 112-144, § 608(a) (2012) (modifying section 513(e) of the FDCA).)  There is also a rarely invoked and rather mischievous procedure allowing any person to seek FDA review and potential withdrawal of another person’s PMA approval.  See 90 Stat. 558 (section 513(g)).

At the same time, the section 510(k) provision in the MDA unexpectedly emerged as the dominant premarket pathway.  It provided that if a new device were “substantially equivalent” to a preamendment device, it could proceed to market with the same classification and controls (or, if the device type were not yet classified, it would proceed to market subject to whatever classification and controls were later applied).  The concern was that those already in the market in 1976 not gain a competitive advantage during the transition to the final regulatory scheme envisioned in the MDA.  By 1984, however, a noted commentator could write that 510(k) provisions had “in many ways eclipsed detailed statutory [PMA] provisions so painstakingly drafted” as the route to market for most medical devices.  J. Kahan, Premarket Approval Versus Premarket Notification:  Different Routes to the Same Market, 39 Food Drug Cosm. L.J. 510, 510 (1984). 

If there is a pattern to the failures of the original MDA, it was over confidence by Congress in the capabilities of centralized agency regulation.  Congress seems to have imagined that FDA would orchestrate the device industry from Washington D.C. with a degree of mastery and subtlety that proved impossible.  In particular, the idea that FDA would promulgate performance standards for hundreds of Class II devices proved impossible to carry out.  The establishment of a viable PMA process for many Class III devices was somewhat more successful, but it took decades to complete most of it, and some Class III devices still utilize the 510(k) pathway.  It also was impractical to have supposed, as Congress apparently did in 1976, that the resource intensive PMA process could be widely applied.  (Annually, about 2% of devices requiring premarket review reach the market via PMA approval.)  Fortunately, FDA developed the 510(k) program from an ill defined transitional measure to a robust pathway to market, and Congress had the good sense to ratify it in the SMDA.  If that had not happened, the inventive and vibrant device industry we have today might have been strangled in the crib.