Medical device regulation

1. EDITORIAL
Rethinking medical device
regulation
Carl Heneghan • Mathew Thompson
Department of Primary Care Health Sciences, University of Oxford, Oxford, UK
Correspondence to: Carl Heneghan. Email: carl.heneghan@phc.ox.ac.uk
DECLARATIONS Complications with Poly Implant Prosthese (PIP) regulatory process in the USA, fewer than
breast implants1 and Metal on Metal (MoM) hip one-third had undergone randomized trials and
Competing interests
implants2 reflect systemic failings with the only half of the trials overall involved controls.8
None declared
current regulation of medical devices. Yet, these Whilst new drugs require at least randomized
Funding two cases highlight only a fraction of the burgeon- controlled trials to gain regulatory approval, for
The authors
ing increase in medical device safety alerts3 and medical devices even under the more stringent
problems with device recalls, and are leading to PMA approval process, only one controlled trial
received no funding
a rethink of the systems for regulatory approval (not necessarily randomized trial) is required.
Ethical approval in both Europe and the USA.4 However, an even more worrying issue with
Not required Therefore, having an understanding of medical device regulation in both the EU and US is the
device regulation is now an important require- use of ‘substantially equivalent’ in evidence sub-
Guarantor ment for doctors and healthcare professionals missions for regulatory purposes. The problems
Carl Heneghan alike. To aid this, French-Mowatt and colleagues with using ‘equivalence’ in the device approval
summarize the current medical device regulation process can be traced back several decades. In
Contributorship
in Europe,5 outlining the current requirements 1976, in the USA many devices were already on
Both CH and MT for CE regulation. Outside of the European the market, so a less burdensome alternative to
contributed to the Union, Susan Lamph describes the regulatory pro- PMA known as 510(k) provision was approved.
ideas, drafted the cesses across different countries and the lack of The 510(k) pathway did not require clinical
manuscript and harmonization with leads to wide variation in pre- trials; the manufacturer was only required to
approved the final market data requirements.6 demonstrate a device was ‘substantially equival-
version Analysis of medical-device recalls in the UK ent’ to another device already on the market. The
and the USA, and the device-regulation process, problem now is that the definition of equivalence
Acknowledgments reveals the increasing nature of the problems. is interpreted so loosely that the FDA admits
None From 2006 to 2010, the UK regulator, the MHRA they need to ‘clarify the meaning of “substantial
issued 2,124 manufacturer field safety notices, an equivalence.”’10
increase of 1,220% over this five-year period.3 In The predicate of equivalence is also used
the USA the number of recalls for moderate or within the European Union (EU) regulatory
high-risk devices more than doubled between system for device regulation. There are three Euro-
2007 to 2011.7 In addition, many recalled medical pean Directives related to device regulation.11 – 13
devices in the USA were originally cleared using These directives, which lead to CE marking and
a less stringent process called 510(k), or even access to the European market, state the extent
more problematic, recalled devices were con- and nature of clinical data required for approval.5
sidered to be so low-risk they were exempt from Problems occur because even for implantable
regulatory review in the first place. This situation devices, the scrutiny of evidence at the outset is
reflects a very low ‘bar’ currently for evidence left to private organizations known as Notified
requirements to gain regulatory approval, even Bodies;3 and secondly, clinical data required for
for high-risk devices. For instance, of 78 high-risk the equivalent route can involve as little as ‘a criti-
cardiovascular devices approved through the cal evaluation of the relevant scientific literature
more stringent Pre-Market Approval (PMA) currently available relating to the safety,
186 J R Soc Med 2012: 105: 186 –188. DOI 10.1258/jrsm.2012.12k030

2. Rethinking Medical Device regulation
performance, design characteristics and intended without a substantial battle with the medical
purpose of the device’. The use of equivalence is device industry.
therefore left to the manufacturer and the Notified Sprange and Clift’s analysis of manufacturers’
Bodies to determine, without any outside scrutiny submission challenges, to the NICE medical
of the decision making process centrally or within technology program, in the supplement edition
each EU country. J R Soc Med 2012;105 (S1) reveals there are
Even for the more stringent PMA process, there significant issues in relation to basic and general
are profound differences in evidence requirements research skills that need to be addressed
between the US and EU. For example, EU amongst manufacturers.17 In addition, interviews
approval of a ‘GuardWire’ developed by Percu- with manufacturers highlight the current status
Surge for use during angioplasty, required a 22 quo: ‘pharmaceutical and medical technologies
patient study with no control group. Yet, in the were also considered very different by manufac-
USA, FDA regulators required an 800 patient mul- turers.’ As such, the widespread belief is that
ticentre randomized controlled trial.14 devices do not require the same level of evidence
Perhaps what is even more concerning than the as drugs to gain access to a market and be used
device recalls and high profile cases (such as the in clinical practice. Sprange’s study highlights
MoM hips and PIP implants), is that many the need for education and research ‘tools,’
medical device problems go unnoticed. For which will facilitate higher quality evidence sub-
example, women participating in a breast cancer missions for approval in the future.
study were left with hundreds of tiny particles of The European Directive on medical devices
the heavy metal tungsten in their breast tissue will be revised later this year. The European Com-
due to a faulty device cleared under the 510(k) mission has stated it will use this opportunity to
processes.9 Similarly, recalled device notices strengthen existing legislation, particularly pro-
often go unheeded. In 2006 the maker of a surgical visions relating to market surveillance, vigilance
clip, the Hem-o-lok issued an urgent recall notice and the functioning of notified bodies. In the
warning surgeons to stop using the clips on living UK, the House of Commons Science and Technol-
kidney donors. However, three years later a ogy Committee plans to examine whether current
surgeon used one of the clips to tie off a 29-year- legislation and regulations on safety and efficacy
old male’s renal artery during an operation in of medical implants are fit for purpose.18
which he donated a kidney to his wife. He bled Failures of medical devices cause harm and
to death twelve hours later when the clip malfunc- cost money. More stringent requirements to
tioned.10 – 15 Currently we have limited ability to provide evidence from clinical trials for the effi-
trace most patients in whom medical devices cacy and safety medical devices before they are
have been used (or implanted), so when problems approved should therefore be welcomed by
or recalls occur, it can be impossible to know the patients, clinicians and the medical device indus-
magnitude of the problem. try. Evidence for new devices must also be open
However, it seems as though the tide is to scrutiny by patients in individual countries, as
turning in terms of regulatory requirements. well as healthcare providers and researchers. The
The American system is coming under increased potential risk of a new device should match the
scrutiny with calls for the removal of the 510(k) type of evidence required prior to approval for
process. The influential Institute of Medicine use in clinical settings. Without these changes to
has recommended the FDA do away with the current systems, it is likely we will continue to
510(k) approval process and replace it ‘with an see substantial complications arising from faulty
integrated premarket and post-market regulatory devices.
framework that effectively provides a reasonable
assurance of safety and effectiveness throughout
the device life cycle.’16 It is possible that all
References
implantable devices in particular will require
1 Heneghan C. The saga of Poly Implant Prosthese breast
PMA approval and thus clinical trial data in
implants. BMJ 2012;344:e306
the future. However, more stringent regulations 2 Heneghan C, Langton D, Thompson M. Ongoing problems
are unlikely to be passed into law in the US with metal-on-metal hip implants. BMJ 2012;344:e1349
J R Soc Med 2012: 105: 186 –188. DOI 10.1258/jrsm.2012.12k030 187

3. Journal of the Royal Society of Medicine
3 Heneghan C, Thompson M, Billingsley M, Cohen D. com/medical-devices-active-implantable.html (last
Medical-device recalls in the UK and the device-regulation accessed 2 April 2012)
process: retrospective review of safety notices and alerts. 12 www.ce-marking.com (General) Medical Device Directive,
BMJ Open 2011;1:e000155. Epub 2011 May 15 MDD (93/42/EEC). http://www.ce-marking.com/
4 Zuckerman DM, Brown P, Nissen SE. Medical device recalls medical-devices.html (last accessed 2 April 2012)
and the FDA approval process. Arch Intern Med 13 www.ce-marking.com In Vitro Diagnostic Medical Device
2011;171:1006 –11. Epub 2011 Feb 14 Directive, IVDMDD (98/79/EC). http://www.ce-marking.
5 French Mowat E, Burnett J. How are medical devices com/medical-devices-in-vitro-diagnostic.html (last
regulated in the European Union? J R Soc Med 2012;105(S1): accessed 2 April 2012)
S22– S28 14 Kaplan A Baim DS, Smith JJ, et al. Medical device
V,
6 Lamph S. Regulation of medical devices outside the development: from prototype to regulatory approval.
European Union. J R Soc Med 2012;105(S1):S12 –S21 Circulation 2004;109:3068 –72. Review. No abstract available
7 See http://www.citizen.org/documents/substantially- 15 No system tracks faulty medical devices in U.S. http://
unsafe-medical-device-report.pdf (last accessed 2 April www.msnbc.msn.com/id/33184398/ns/
2012) health-health_care/t/no-system-tracks-faulty-medical-
8 Dhruva SS, Bero LA, Redberg RF. Strength of study devices-us/#.T3m9OdUxZeE (last accessed 2 April 2012)
evidence examined by the FDA in premarket approval of 16 Romza-Kutz DJ, Browne RE Jr, Costello JF Jr, Morrissey MM.
cardiovascular devices. JAMA 2009;302:2679 –85 FDA Taking Comments on Potential New Hurdles for
9 Denise Grady. Riddled With Metal by Mistake in a Approval of Medical Devices. http://www.
Study: New York Times http://www.nytimes.com/2011/ troutmansanders.com/fda-taking-comments-on-potential-
03/22/health/22breast.html?_r=3&scp=1&sq=Axxent% new-hurdles-for-approval-of-medical-devices-09-08-2011/
20FlexiShield%20Mini&st=cse. (last accessed 2 April (last accessed 8 May 2012)
2012) 17 Sprange K, Clift M. The NICE Medical Technologies
10 US Food and Drug Administration. 510(k) Working Group Evaluation Programme (MTEP): manufacturer submission
preliminary report and recommendations. August 2010. challenges. J R Soc Med 2012;105(S1):S4 –S11
http://www.fda.gov/downloads/aboutfda/centersoffices/ 18 Committee announce new inquiry into the Regulation of
CDRH/cdrhreports/UCM220784.pdf (last accessed 2 April medical implants. http://www.parliament.uk/business/
2012) committees/committees-a-z/commons-select/
11 www.ce-marking.com Active Implantable Medical Device science-and-technology-committee/news/new-inquiry-
Directive, AIMDD (90/383/EEC). http://www.ce-marking. medical-implants/ (last accessed 2 April 2012)
188 J R Soc Med 2012: 105: 186 –188. DOI 10.1258/jrsm.2012.12k030