Intended for healthcare professionals


Follow-up for patients with metal-on-metal hip replacements: are the new MHRA recommendations justified?

BMJ 2018; 360 doi: (Published 13 February 2018) Cite this as: BMJ 2018;360:k566
  1. Gulraj S Matharu, orthopaedic registrar1 2,
  2. Andrew Judge, professor of translational statistics1 2 3,
  3. Hemant G Pandit, professor of orthopaedic surgery1 4,
  4. David W Murray, professor of orthopaedic surgery1
  1. 1Nuffield Department of Orthopaedics, Rheumatology and Musculoskeletal Sciences, University of Oxford, Oxford, UK
  2. 2Musculoskeletal Research Unit, Translational Health Sciences, Bristol Medical School, University of Bristol, Bristol, UK
  3. 3MRC Lifecourse Epidemiology Unit, University of Southampton, Southampton, UK
  4. 4Leeds Institute of Rheumatic and Musculoskeletal Medicine, University of Leeds, Leeds, UK
  1. Correspondence to: G S Matharu gsm{at}

Recent MHRA guidance for the surveillance of people with metal-on-metal hip replacements are onerous, costly, and insufficiently evidence based, argue Gulraj Matharu and colleagues

Metal-on-metal hips were commonly used in young active patients with arthritis, with about 1.5 million implanted worldwide.123 However, the devices experienced high short term failure rates,45 and many patients have had to have revision surgery because of abnormal reactions to materials generated by the implants (collectively termed adverse reactions to metal debris).126 Metal-on-metal hip implants are now rarely used.12

Failure of conventional hip implants is associated with symptomatic bone damage and requires revision with a larger implant. However, adverse reactions to metal debris from metal-on-metal implants are associated with destruction of soft tissue (muscles, tendons, neurovasculature) as well as bone. The soft tissue destruction is often irreparable,78 resulting in many patients doing poorly after revision procedures.8 Furthermore, irreversible bone and soft tissue damage can occur without noticeable symptoms.9 A few patients also develop systemic symptoms from exposure to high levels of metal ions, though these usually resolve after revision surgery (box 1). Reassuringly, large cohort studies have reported that patients with metal-on-metal hips are not at increased risk of cancer, heart failure, or death compared with patients with conventional hip replacements.1314151617

Box 1

Systemic effects of high metal ion exposure101112


  • Hearing impairment or loss

  • Visual impairment or loss

  • Peripheral neuropathy

  • Cognitive impairment


  • Cardiomyopathy/heart failure

  • Breathlessness


  • Hypothyroidism

  • Malaise

  • Depression


Regulatory authorities worldwide therefore recommend regular follow-up of patients with metal-on-metal implants in order to identify and treat adverse reactions early, and hopefully improve outcomes.1819202122 Surveillance can include clinical review, measurement of blood metal ion levels (surrogate marker of implant wear), hip radiography, and cross sectional imaging (ultrasonography or magnetic resonance imaging (MRI)). However, the follow-up recommendations vary between countries (UK, USA, Europe, Australia, and Canada) and are not evidence based.23

In June 2017, the Medicines and Healthcare Products Regulatory Agency (MHRA) issued a planned update of its 2012 follow-up advice for patients with metal-on-metal implants.1824 These new recommendations, endorsed by professional orthopaedic bodies,25 advise more intensive follow-up, with most patients now requiring annual investigations for the implant lifetime.24 This will affect over 60 000 patients in the UK,2 as well as patients in many other countries that follow MHRA recommendations. Importantly most patients have no symptoms and have well-functioning implants.26 We consider whether the latest MHRA recommendations are justified by the available evidence and assess the financial implications and potential impact on patients.

What has the MHRA changed?

The 2012 MHRA guidance recommended all patients with recalled metal-on-metal implants and patients with non-recalled implants who have hip symptoms were reviewed annually with metal ion measurement and cross sectional imaging (ultrasonography or MRI).18 The 2017 guidance still recommends annual follow-up for all patients with symptoms but requires cross sectional imaging for low risk patients only (men with hip resurfacing femoral heads >48 mm and everyone with stemmed hip replacements with femoral heads <36 mm).

People without symptoms who have had metal-on-metal hip resurfacing (about 30 000 UK patients)2327 could previously be followed according to local protocol.18 Many hospitals discharged such patients either immediately or if initial reviews were satisfactory. However, the 2017 guidance has now added all women with hip resurfacings and men with small resurfacing implants (femoral head ≤48 mm) to the high risk group, which previously included only patients with withdrawn hip resurfacing implants and all stemmed hip replacements with femoral heads ≥ 36 mm. These high risk asymptomatic patients with hip resurfacings now require more intensive annual follow-up. The remaining asymptomatic low risk patients now require regular (annually to three yearly) radiography, metal ion tests, and Oxford Hip Score (OHS) assessment rather than follow-up as per local protocol.2829

Some of the changes that the MHRA made to its recommendations are supported by the evidence. These include more comprehensive risk stratification for patients who have had hip resurfacing and more intensive surveillance for groups found to have more problems (such as women with hip resurfacings).226 Men who have implants with an established track record, such as the Birmingham Hip Resurfacing (BHR),22630 now require less regular surveillance. Recommendations to use the Oxford Hip Score to monitor pain and function23 are also evidence based. It has been shown to detect suboptimally functioning implants, even in patients without symptoms.3132 However, some of the new MHRA recommendations seem inconsistent with the available evidence.

Which MHRA changes are not supported by evidence?

Investigation of symptomatic patients

All worldwide authorities, including the 2012 MHRA guidance, recommend symptomatic patients have ion testing and cross sectional imaging.23 This is because most patients with adverse reactions to metal debris have pain, with symptoms potentially signifying bone and soft tissue destruction that requires timely assessment and treatment.82633 The 2017 MHRA guidance still recommends ion testing and cross sectional imaging for symptomatic low risk patients. However, these recommendations have counterintuitively been downgraded for symptomatic high risk patients (such as women with recalled implants). High risk symptomatic patients now require imaging only if metal ion levels are abnormal. This change risks delayed diagnosis and treatment in high risk patients, poorer outcomes, and increased litigation.

Interpretation of metal ion measurements

The MHRA recommends that patients with metal ions of ≥7 parts per billion (ppb) require closer surveillance, based on early weak data.3435 Numerous studies show that metal ion levels are not useful for diagnosing adverse reactions,3637383940 so continuing to promote this ion threshold for identifying patients at risk of adverse reactions is not advisable. However, recent studies show that much lower ion levels (2-5.5 ppb) are effective at identifying patients at low risk of adverse reactions.363738 Therefore using these newer thresholds would help reassure many patients that they do not have adverse reactions.

Annual surveillance of patients with hip resurfacing implants

Longitudinal studies find little variation in metal ion levels in asymptomatic patients who have had hip resurfacing over the medium term (up to 10 years from surgery), even in those with high risk implants.41424344 Similarly, patients who have normal results on cross sectional imaging two to eight years after surgery experience few changes when imaging is repeated within three years4546 with any problems developing only after 7-11 years.47

A longitudinal study of 152 asymptomatic patients who had had hip resurfacing showed that no patient with normal initial test results (metal ions <2 ppb and no abnormality on ultrasound scans) at a minimum of three years postoperatively developed adverse reactions when tests were repeated within five years,31 suggesting that such patients do not need follow-up within five years. However, the new MHRA guidance recommends annual follow-up for many of these patients.

Registry data do not support new MHRA guidance

The MHRA was concerned that UK National Joint Registry data (the world’s largest arthroplasty registry) continued to show a risk of adverse reactions.2425 However, there will always be a risk of adverse reactions until all metal-on-metal implants are removed.26 Importantly, we are unaware of any new registry evidence suggesting the problem is getting significantly worse, especially for patients with resurfacings.24 For all hip resurfacings, the registry revision rates for the MHRA guidance in 2017 (12 years=13.6%)2 and 2012 (8 years=9.1%) were consistent with an average 1.1% annual rate.48 The Australian registry reports similar findings.1 Furthermore, the annual number of revisions of metal-on-metal implants for all causes and for adverse reactions to metal debris has continued to decrease since 2012.4950 Therefore it is unclear how the registry data support the increased MHRA surveillance.

Conversely, registry data and other sources have shown clinical benefits since the 2012 MHRA guidance was implemented.18 Increased awareness of potential problems and regular surveillance has led surgeons to revise implants earlier and for less severe abnormalities than before.26 This has been associated with improved outcomes after revision surgery505152 compared with initial data.8

Given these observations and the limited evidence supporting the new MHRA proposal of increased surveillance, caution is warranted. A systematic review of 122 studies observed the prevalence of revision for adverse reactions was highly variable, even for similar implants, but was closely related to intensity of surveillance.53 The variability in surveillance is probably related to the inconsistent way follow-up guidance has been interpreted by different hospitals and surgeons.23 Increased surveillance can lead to overdiagnosis of problems that may never require revision and may put patients at unnecessary risks associated with revision procedures, especially given that abnormal reactions not requiring treatment occur in patients with well functioning conventional hip replacements.5455 Universal prostate cancer screening has similarly caused overdiagnosis and treatment related complications.56

Costs of implementing the new MHRA guidance

Previous work reported that the 2012 MHRA guidance would cost £8.2m (€9m; $11m) annually to implement in the UK.23 Implementing the 2017 guidance is likely to cost much more than this. Although there will be savings from less regular surveillance of low risk patients, the more intensive follow-up for asymptomatic patients with hip resurfacings will result in higher costs overall.

In addition to initial reviews and investigations, there will be other costs to health services, such as organising extra clinics, the resources required to perform and interpret tests, and the further investigations needed if abnormalities are identified. To review 60 000 patients just once will take about 1000 staffed clinic days (two doctors per session). Concerned patients have been directed back to primary care,2557 sometimes without hospitals providing instructions on how to advise them. General practitioners often do not know which hip implant a patient has received, and they should not be asked to measure metal ions since the results require careful interpretation. The follow-up costs for hip replacements that were performed privately in the UK may also fall on the NHS.

Finally, the cost of revision surgery is substantial. The 500-1000 revisions of metal-on-metal implants performed annually in the UK equate to £5m-£10m/year.50 The effect on hospital budgets of doing more revisions will be compounded by the NHS revision procedure tariff recently being reduced by nearly £3000.58

What do patients think about regular follow-up?

We spoke to five patients with metal-on-metal implants (box 2). All patients felt reassured they were having annual or two yearly clinical reviews, radiography, metal ion tests, and in some cases cross sectional imaging. Although none of the patients had any serious concerns about potentially being overinvestigated, some would prefer to have less regular reviews and tests in the future given there was now some evidence supporting this approach.

Box 2

Patients’ views on regular clinical surveillance and tests

We spoke to three women and two men aged 50-76 who had had metal-on-metal hip replacements 6-10 years previously at one of two centres in the UK. All were asymptomatic and were having some type of hospital follow-up every one to two years. Between them they had six hip resurfacing implants and one stemmed total hip replacement (two patients had bilateral implants).

Perspective on regular reviews and tests

  • All patients felt reassured that they were having regular clinical reviews, radiography, metal ion tests, and in some cases cross sectional imaging

  • Patients felt these regular check-ups and tests provided them with security despite being pain-free and would allow any issues to be identified early

  • No patient was concerned about the time needed to travel and attend these regular appointments

  • One patient highlighted their data may benefit others in the future by contributing to research

  • Patients did not feel they were being over investigated, though some recognised their views may have been different if the tests being performed were more invasive (eg, biopsies) or associated with more serious complications

  • Two patients said they would prefer to have less regular reviews and tests in the future given there was now evidence that asymptomatic hip resurfacing patients with normal tests can be safely discharged for five years. These patients did consider the resource implications in their responses (ie, “Someone else can have my appointment if I do not actually need it.”).

How follow-up could be improved

  • All patients were happy with their implants, the follow-up received, and the information provided over the years by their surgeons and hospitals regarding issues with metal-on-metal implants

  • They were also confident they could contact their clinicians between planned appointments if they developed any problems.

  • One patient said that they would like a simple questionnaire to be developed that could help stratify patients for subsequent follow-up


Protecting patients from poorly performing implants

The encouraging early results of the first hip resurfacing implant (BHR) were incorrectly extrapolated to apply to other metal-on-metal designs. These devices were subsequently widely marketed and implanted without robust evidence about their safety and effectiveness, which ultimately placed huge numbers of people at unnecessary risk.

To avoid repeating the failures associated with metal-on-metal implants, the introduction of new technologies must be rigorous and transparent. This should include usage by small expert groups, in randomised trials and independently controlled surveillance programmes. Any concerns must be reported to manufacturers and regulators early, and must be taken seriously to minimise any potential patient harm. The initial device experience should be interpreted in combination with registry and other data before any widespread introduction. Organisations such as Beyond Compliance and the Orthopaedic Data Evaluation Panel have been formed with the aim of safely introducing and monitoring the performance of new implants in clinical practice.5960


The new 2017 MHRA recommendations mean that many more patients with metal-on-metal hip implants will be subjected to annual reviews and testing for the implant lifetime, which means the burden of this problem for both patients and health systems will continue for many years. However, evidence increasingly supports less regular surveillance in many patients, which would allow better use of healthcare resources. Five yearly surveillance may help reduce anxieties and concerns in many asymptomatic patients with well functioning hip resurfacings, although patients should be consulted on the benefits and harms of different follow-up approaches.

It is unlikely that the substantial increase in follow-up costs will be offset by detecting the small proportion of asymptomatic patients with adverse reactions from hip resurfacing who would otherwise have been missed. Given the known pressures within health services there are serious concerns about how financially sustainable the new MHRA recommendations are. It remains to be seen whether implant manufacturers will contribute to some, or all, of the extra costs.61

Key messages

  • The MHRA’s updated advice recommends more intensive follow-up for a larger proportion of people with metal-on-metal hip replacements

  • This will affect over 60 000 patients in the UK with these implants, as well as many in other countries that follow MHRA advice

  • The recommendations do not reflect the evidence showing that many asymptomatic patients with normal test results do not require annual follow-up

  • Implementing the recommendations is likely to be associated with a substantial cost to health services


  • Contributors and sources: All authors have substantial expertise regarding metal-on-metal hip replacements and have published widely on this topic for a number of years. GSM recently successfully completed his DPhil on the investigation and management of metal-on-metal hip replacement patients, with AJ, HGP, and DWM acting as supervisors. This article arose from discussions between all the authors following the publication of the latest MHRA guidelines, with the evidence cited in this article identified during the course of GSM’s doctoral research. GSM proposed and developed this article, performed the analysis, conducted the patient interviews, and drafted and revised the manuscript. AJ contributed to the development of this article and revised the manuscript for important intellectual content. HGP and DWM are joint senior authors; they contributed equally to the development of this article and revised the manuscript for important intellectual content. All authors read and approved the final manuscript. DWM is the guarantor.

  • Competing interests: We have read and understood BMJ policy on declaration of interests and declare the following interests: GSM has received financial support for other research work from Arthritis Research UK, the Orthopaedics Trust, Royal College of Surgeons of England, and Royal Orthopaedic Hospital Hip Research and Education Charitable Fund. GSM has also received personal fees for undertaking medicolegal work for Leigh Day, which includes work relating to metal-on-metal hip replacements. AJ has received consultancy, lecture fees, and honorariums from Servier, UK Renal Registry, Oxford Craniofacial Unit, IDIAP Jordi Gol, and Freshfields Bruckhaus Deringer; he is a paid member of the data safety and monitoring board for Anthera Pharmaceuticals, and received consortium research grants from ROCHE. HGP is a paid consultant for Kennedys Law, Bristol Myers Squibb, Depuy Synthes, Medacta Int and Meril Life. He has received institutional research grants from UKIERI, Charnley Trust, Depuy Synthes, GSK and NIHR. HGP and DWM are paid consultants and speakers for Zimmer-Biomet, and both receive institutional research funding from Zimmer-Biomet and Stryker. DWM has received research funding related to metal-on-metal hip replacements from Depuy, Smith & Nephew, Wright Medical, and Zimmer-Biomet, and has been a consultant related to metal-on-metal hip replacements for Herbert Smith Freehills Lawyers and Wright Medical.

  • Patient involvement: We discussed the MHRA recommendations with five patients with metal-on-metal hip replacements to obtain their perspectives on various issues related to this article. We thank them for their important contribution to this article.

  • Provenance and peer review: Not commissioned; externally peer reviewed.


View Abstract

Log in

Log in through your institution


* For online subscription