Feature article

Does the addition of mesh improve outcomes in implant based breast reconstruction after mastectomy for breast cancer?

Shelley Potter, NIHR clinician scientist and, consultant senior lecturer in oncoplastic breast surgery1, Mairead MacKenzie, patient advocate2, Jane M Blazeby, professor of surgery3


What you need to know

  • The use of biological and synthetic meshes in implant based breast reconstruction has become standard care, but there is limited high quality evidence to support their safety or effectiveness.

  • Discuss with your patient the benefits and risks of using either types of mesh, including the lack of high quality long term comparative data, single stage compared with two-stage procedures, complications, and the need to consider the surgeon’s experience in the technique.

  • Encourage patients who opt for mesh based implant reconstruction to be involved in research studies contributing to evidence on long term outcomes

Of the 55 000 women diagnosed with breast cancer1 each year in the UK, more than 40% require a mastectomy2 as primary surgical treatment. To optimise quality of life and outcomes, the National Institute for Health and Care Excellence (NICE) recommends offering immediate breast reconstruction.3 In the UK4 and US,5 implant based breast reconstruction (IBBR) is the most commonly performed technique.

The implant is usually placed in a pocket under the pectoralis major muscle. This generally requires a two-stage approach as the pocket is not large enough to accommodate a fixed volume implant. A tissue expander is placed as a first stage and it is expanded by injecting fluid percutaneously until the desired size is achieved. The expander is then replaced by a fixed volume implant at a second operation. This technique is safe,6 but time consuming and uncomfortable. It is possible to use combined expander implants (“permanent expanders”) as a one-stage approach to avoid a second operation, but the results may be unsatisfactory.7

The introduction of biological and synthetic meshes in the early 2000s to augment the submuscular pocket has had a major impact on the practice of IBBR. The mesh can be used as a sling between the lower edge of the pectoralis muscle and the chest wall to provide coverage for the lower pole of the implant. This creates a much bigger submuscular pocket (fig 1) and enables a single stage procedure with the definitive and right sized implant placed at the time of mastectomy. It is thought that cosmetic outcomes are improved because of better lower pole projection and improved control of the inframammary fold, which creates a more natural looking result. A wide range of biological (eg, acellular dermal matrix) and synthetic (eg, titanium coated polypropylene) meshes are available. These differ in cost (£300 to £2000 (€340-€2270; $400-$2650) and in the absence of comparative evidence, usage is currently dependent on surgeon preference.

Fig 1

Types of implant based breast reconstruction

bmj-362-bmj-k2607-F1.jpg

Recently, practice has evolved. Surgeons may place the implant, fully or partially wrapped in mesh, on top of the pectoralis muscle in a subcutaneous position (fig 1). It is suggested that the subcutaneous technique may reduce postoperative pain as the muscle is not disturbed. Furthermore, it may help prevent implant “animation,” the distressing upwards movement of the implant that is seen when the chest muscles contract in standard submuscular techniques.8

It is unclear, however, whether mesh assisted procedures are a safe alternative to traditional IBBR techniques, and if patient reported outcomes including cosmetic outcomes, general well being, and postoperative pain are improved with it. There is also uncertainty about the best position for the implant when a mesh is used (under the skin or under the muscle) and the choice of mesh (biological or synthetic).

What is the evidence of uncertainty?

Sources and selection criteria

We searched PubMed, the Cochrane Library, and the clinicaltrials.gov databases to identify published and ongoing randomised clinical trials and systematic reviews that evaluated mesh in women undergoing IBBR after mastectomy for breast cancer or risk reduction.

There is a lack of high quality evidence comparing clinical and patient reported outcomes with mesh assisted IBBR and traditional IBBR. A recent systematic review noted a higher rate of infection with mesh assisted IBBR compared with traditional IBBR, but the quality of evidence is low.9 Most studies are small, single centre retrospective cohort studies and case series with methodological limitations. Few studies have assessed health related quality of life or cosmetic outcomes.10 A randomised controlled trial11 (142 patients) reported a higher rate of surgical complications including reconstructive failure and the need for re-operation at one year in patients undergoing mesh assisted IBBR compared with traditional IBBR. This is a small study, however, and at high risk of bias because of lack of blinding of outcome assessors.

Low quality evidence from three small trials comparing different types of mesh suggests no difference in outcomes.121314 Of these, one very low quality underpowered study14 reported a higher rate of implant loss and more postoperative pain, fatigue, and impact on family life with biological mesh compared with synthetic mesh. Table 1 summarises findings from randomised controlled trials on mesh assisted IBBR. We found no trials comparing subcutaneous and submuscular implant reconstruction with mesh.

Table 1

Summary of randomised controlled trials evaluating mesh use in implant based breast reconstruction

Randomised controlled trials comparing implant based breast reconstruction with and without mesh
StudySample sizeInterventionComparisonOutcomes assessedMain findingsQuality of evidence*Uncertainty
McCarthy et al, 2012692 stage expander-implant reconstruction with human ADM (AlloDerm)
n=33
Standard 2 stage expander-implant reconstruction
n=36
1.     Pain using BREAST-Q; physical well being; chest and upper body scale and VAS postoperatively and during expansion phase.
    2. Rate of tissue expansion
No differences in pain postoperatively or during the expansion period. No differences in rate of postoperative expansionModerateUnderpowered. Study stopped by Data Safety Monitoring Board because of concerns about recruitment.
Use of human ADM in 2 stage expander-implant reconstruction does not reflect UK practice
Dikmans et al, 2016142Single stage direct to implant reconstruction with porcine ADM (Strattice)
n=59
Standard 2 stage expander-implant reconstruction
n=62
1.     HRQL assessed using the BREAST-Q and EQ-5D at 1 year.
    2. Safety: adverse events classified using CTCAE criteria at 1 year.
    3. Aesthetic outcomes based on photographs at 1 year
Significantly higher rates of surgical complications (OR 3.46, 95% CI 1.39 to 8.61); re-operation (OR 3.69, 95% CI 1.31to 10.42); and removal of implant (OR 16.82, 95% CI 2.44 to 115.94) in patients undergoing single stage reconstruction with ADM compared with 2 stage expander-implant procedures without ADM.
HRQL and aesthetic outcomes not reported
Very lowVery selected patient population (non-smokers, BMI <30, no post-mastectomy radiotherapy, small breasts).
High risk of bias due to lack of blinding.
Failed to account for surgeons’ learning curve with new technique.
Two-surgeon model (oncologic surgeon performing mastectomy and plastic surgeon performing reconstruction) not consistent with UK practice
Randomised controlled trials comparing different types of mesh
Mendenhall et al, 20151162 stage expander-implant reconstruction with human ADM (DermaMatrix)
n=59
2 stage expander-implant reconstruction with human ADM (AlloDerm)
n=57
    1. Incidence and grade of complications.
    2. Expander dynamics.
    3. Biointegration of ADM
No difference in overall complications (OR 1.24, 95% CI 0.64 to 2.40) or grade of complications (OR 1.33, 95% CI 0.75 to 2.35) between treatment groups. AlloDerm resulted in less time to complete expansion (42 v 72 days, P<0.001)LowAll procedures performed by single surgeon so not generalisable.
High risk of bias due to method of allocation concealment (sealed envelopes). No reported blinding of outcome assessors.
Use of human ADM in 2 stage expander-implant reconstruction does not reflect UK practice
Gschwantler-Kaulich et al, 201648Single stage direct to implant with porcine ADM (Protexa) n=23Single stage direct to implant with synthetic mesh (TiLOOP) n=251.     Cosmetic outcome assessed from photographs.
    2. Complications.
    3. HRQL assessed using EORTC QLQ C30 and BR23
No statistically significant difference in overall complications between patient groups (31.3% ADM, 24.0% TiLOOP, P=0.19) but significantly higher rate of implant loss and reconstructive failure in the ADM group (30.4% v 7.7%, P<0.0001).
Patients in ADM group reported significantly more arm pain (48% v 24%, P=0.04) and fatigue (35% v 12%, P=0.03) at the first post-operative visit and a more affected family life (17% v 0%, P=0.02) and less sexual interest (17% v 48%) at 6 months following reconstruction.
There was no difference in cosmetic outcome
Very lowReported as “pilot” randomised controlled trial but no feasibility endpoints assessed.
Underpowered trial.
No primary endpoint identified; no power calculation performed.
Insufficient details reported to assess risk of bias.
HRQL instruments not validated in breast reconstruction population.
ADM assessed not routinely used in UK
Hinchcliff et al, 2017302 stage expander-implant reconstruction with human ADM (AlloMaxTM)
n=15
2 stage expander-implant reconstruction with human ADM (AlloDerm)
n=15
1. Complication rate.
2. Patient satisfaction at 1 year using BRECON-31 questionnaire
No significant difference in complications between patient groups at 30 days or following implant exchangeVery lowAll procedures performed by single surgeon so not generalisable.
Very small sample size.
Insufficient details reported to formally assess risk of bias.
Use of human ADM in 2 stage expander-implant reconstruction does not reflect UK practice
  1. Using GRADE; ADM: acellular dermal matrix; CI: confidence interval; CTCAE: Common Terminology Criteria for Adverse Events (version 4.0); EORTC: European Organisation for Research and Treatment of Cancer; HRQL: Health-related quality of life; OR: odds ratio; UK: United Kingdom; VAS: visual analogueue scale

Mesh v no mesh

A 2015 systematic review evaluated the published evidence for acellular dermal matrix in IBBR.10 This included eight systematic reviews, one randomised controlled trial, 40 non-randomised comparative studies, and 20 case series. All studies evaluated clinical outcomes such as implant loss (the need to remove the implant because of post-operative complications leading to reconstructive failure), but few assessed patient outcomes and cosmesis. The data could not be meaningfully analysed because of the heterogeneity of the included studies, all of which were at high risk of bias. A more recent systematic review9 (28 non-randomised studies and 23 case series) compared overall complications and specific clinical outcomes including implant loss, infection, and capsular contracture in patients undergoing IBBR with and without mesh. Pooled analysis suggested a higher rate of infection in the mesh assisted group (risk ratio 1.55, 95% confidence interval 1.17 to 2.05) but no other important difference in complications; however, the quality of the included studies was low.9 Data on health related quality of life and oncological outcomes were lacking.15

Since the completion of these reviews, a multicentre Dutch randomised controlled study (142 women) comparing single stage direct-to-implant IBBR with acellular dermal matrix and traditional two stage expander-implant reconstruction has reported11 safety data at one year (table 1). Patients in the acellular dermal matrix group experienced more surgical complications (odds ratio 3.46, 95% confidence interval 1.39 to 8.61), complications requiring re-operation (odds ratio 3.69, 95% confidence interval 1.31 to 10.42), and had a higher incidence of reconstructive failure (odds ratio 16.82, 95% confidence interval 2.44 to 115.94) than those undergoing traditional reconstruction. While these results are concerning, the study is small and at high risk of bias because of a lack of blinding of outcome assessors. Importantly, it did not take account of the learning curve of participating surgeons,16 which has been shown to affect surgical complications in acellular dermal matrix assisted IBBR.17

A recent multicentre prospective North American cohort study18 (1297 women) found no difference in complications and patient reported outcomes between two stage IBBR with and without acellular dermal matrix. Complications were defined as adverse, surgery related, postoperative events requiring additional treatment at two years, including re-operation or readmission to hospital and reconstructive failure (removal of the implant). Patient reported outcomes were assessed preoperatively and at one week, three months, one and two years postoperatively using the validated BREAST-Q questionnaire. This is not standard practice in the UK, however, and would need to be evaluated within the context of a randomised study.

Type of mesh: biological or synthetic?

The 2015 systematic review10 included nine non-randomised studies comparing different types of acellular dermal matrix. These were small, retrospective, mostly single centre reports, and the findings of no differences between products are of limited value.

An updated literature review has identified two small single centre randomised controlled trials comparing complications of different human acellular dermal matrixes1213 and shows no substantial differences in outcomes between different products (table 1). These studies were not reported in sufficient detail for the risk of bias to be formally assessed, but as the trials were largely explanatory, the results are unlikely to be generalisable. A third small pilot randomised controlled trial compared biological and synthetic meshes.14 This study compared cosmetic outcomes using panel photographic assessment, complications, and quality of life using the EORTC QLQ C30 and BR23 questionnaires in patients receiving biological and synthetic mesh. There were no statistically significant differences in cosmetic outcome and overall complications between the patient groups, but patients in the acellular dermal matrix group experienced substantially higher rates of implant loss than those undergoing IBBR with synthetic mesh (n=7 v n=2, P<0.0001). Patients in the acellular dermal matrix group also reported more postoperative pain, fatigue, and disruption to their family life than those in the synthetic mesh group. Although reported as a “pilot” trial, this study is a small trial that is insufficiently designed to look at the target difference between the treatment groups.19 No primary outcome or power calculation are reported and there are insufficient details to enable formal assessment of bias risk. This study therefore represents very low quality evidence, the results of which cannot be relied on (table 1).

Subcutaneous v submuscular implant reconstruction with mesh

Our updated search did not identify any randomised controlled trials or systematic reviews in this field. We found one narrative review8 which includes case series, with few studies directly comparing submuscular and subcutaneous techniques.

Is ongoing research likely to provide relevant evidence?

We searched ClinicalTrials.gov, the ISRCTN registry, and the Cochrane Library and identified several small ongoing randomised trials: two multicentre studies comparing two stage expander-implant and single stage direct-to-implant reconstruction with acellular dermal matrix in Europe (NCT02061527, n=120) and Canada (NCT00956384, n=189) and six studies comparing meshes; three comparing human (NCT03145337; NCT02891759) and non-human acellular dermal matrix (NCT02521623; n=60); and three comparing biological and synthetic mesh in submuscular (NCT02985073; n=40) and subcutaneous (NCT02830685; NCT02831426) IBBR. A single small study (NCT03143335) compares subcutaneous and subpectoral techniques.

While these trials will add to the evidence base, they are unlikely to be sufficiently large or pragmatic to determine whether mesh is safe or if it improves patient reported and cosmetic outcomes in IBBR; which mesh should be used; or where the implant should be placed. We are conducting a non-randomised prospective multicentre cohort study in the UK (iBRA study; ISRCTN37664281) to inform the feasibility, design, and conduct of a pragmatic randomised controlled trial in IBBR.20

What should we do in light of this uncertainty?

In light of the lack of evidence and recent issues with other mesh based procedures,21 surgeons and specialist nurses involved in decision making for breast reconstruction should ensure that patients are fully informed that there is limited short and long term safety and patient reported outcome data for mesh assisted IBBR, and that surgeons may have limited experience with the technique. As large numbers of women are electing to undergo these procedures, the degree to which this information is currently shared with patients is unknown, raising questions about the quality of information provision and informed consent.

Published guidelines2223 are largely based on poor quality evidence and expert opinion, but offer sensible advice regarding current best practice. The American Society of Plastic Surgeons recommends that mesh use should be considered on a per patient basis.22 Careful patient selection and performing mesh assisted IBBR with caution in high risk groups (such as current smokers, patients who have had previous breast radiotherapy, and those with a high BMI) is recommended by the UK professional associations, the Association of Breast Surgery, and the British Association of Plastic, Reconstructive and Aesthetic Surgeons,23 together with a robust prospective audit of surgical outcomes to generate data to support practice.

Recommendations for further research

Surgeons should commit to perform new techniques within the context of well designed protocolled early phase evaluation studies or registries using standardised outcome measures, and should encourage patients electing to undergo mesh assisted procedures to take part in research studies. Multicentre, pragmatic trials are needed to determine the best and most cost effective approach to IBBR.

  • Population: Women aged 16 or over undergoing mastectomy for breast cancer or risk reduction who elect to undergo immediate implant based breast reconstruction

  • Intervention and comparisons: It may be possible to address the questions within a single trial with an adaptive or factorial design:

    1. Single stage direct-to-implant reconstruction with mesh versus standard two stage expander-implant reconstruction

    2. Biological versus synthetic mesh

    3. Subcutaneous versus submuscular implant reconstruction with mesh

  • Outcome: Patient centred outcomes including patient satisfaction and safety outcomes such as rates of implant loss. We recommend the use of the breast reconstruction core outcome set. Adequate follow-up and appropriate timing of outcome assessment are essential to understand the final cosmetic result achieved. Economic evaluation will be an important component of any future trial.

How patients were involved in the creation of this article

A patient advocate is a co-author of this article, and patients are involved on the steering group of our ongoing IBBR studies.

They expressed concerns that patients were unaware of the degree of uncertainty surrounding the use of mesh, and remarked on the difficulties patients face in getting clear advice on what approach would be right for them.

The “what patients need to know” box includes questions for patients to ask their surgeon, to help them make more informed decisions.

What patients need to know

  • Mesh assisted implant based breast reconstruction is increasingly performed; however, there is no good quality published research that it has improved patient satisfaction and cosmetic outcomes.

  • Short term complications of mesh assisted procedures may be higher than traditional implant reconstruction, and there is no evidence on long term outcomes of mesh procedures including the need for further surgery over time.

  • Ask your surgeon for information about the technique they will perform and their experience with it to decide whether or not you would like to choose this option.

  • Discuss with your surgeon their personal complication rates, in particular how many patients need a second operation for complications and how many patients need their implant removed, and when and why this may be needed, to make an informed decision about the surgery.

Education into practice

  • Find out how many patients in your practice have undergone implant based breast reconstruction in the last five years and what type of surgery they had. What outcomes would you ask for in them to assess patient satisfaction and safety?

  • Describe how you will explain to a patient considering implant based breast reconstruction about the types of procedure, and evidence for these.

Notes

  1. Competing interests: All authors declare: no support from any organisation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years; no other relationships or activities that could appear to have influenced the submitted work.
  2. SP is an NIHR clinician scientist and chief investigator for the iBRA study, which is funded by the NIHR Research for Patient Benefit Programme. JMB is an NIHR senior investigator. This work was undertaken with the support of the MRC ConDuCT-II (Collaboration and innovation for difficult and complex randomised controlled trials in invasive procedures) Hub for Trials Methodology Research (MR/K025643/1) and the NIHR Biomedical Research Centre at the University Hospitals Bristol NHS Foundation Trust and the University of Bristol. The views expressed are those of the authors and not necessarily those of the UK National Health Service, National Institute for Health Research, or Department of Health.
  3. SP and JMB identified the topic; SP conducted the review and wrote the first draft of the manuscript; SP, MM, and JMB critically reviewed and revised the manuscript and all authors approved submission. SP is the guarantor.
  4. The authors would like to thank Chris Holcombe for critically reviewing the manuscript and Professor Chris Metcalfe for advice on trial design.
  5. Provenance and peer review: commissioned; externally peer reviewed.

References

  1. Cancer Research UK. http://www.cancerresearchuk.org/health-professional/cancer-statistics/statistics-by-cancer-type/breast-cancer2016.
  2. G LawrenceO KearinsC Lagord. The Second All Breast Cancer Report, 2011. http://www.ncin.org.uk/search/all+breast+cancer+report
  3. National Institute of Clinical Excellence. Breast cancer (early & locally advanced): diagnosis and treatment. 2009. https://www.nice.org.uk/guidance/CG80?UNLID=
  4. JC MenniePN MohannaJM O’DonoghueR RainsburyDA Cromwell. National trends in immediate and delayed post-mastectomy reconstruction procedures in England: A seven-year population-based cohort study. Eur J Surg Oncol2017;43:52-61. 10.1016/j.ejso.2016.09.019 27776942
  5. American Society of Plastic Surgeons. 2015 Plastic Surgery Statistics Report, 2015. https://www.plasticsurgery.org/news/plastic-surgery-statistics?sub=2015+Plastic+Surgery+Statistics
  6. PTR ThiruchelvamF McNeillN JallaliK Hogben. Post-mastectomy breast reconstruction. BMJ2013;347:f5903.
  7. C EriksenEN LindgrenJ FrisellB Stark. A prospective randomized study comparing two different expander approaches in implant-based breast reconstruction: one stage versus two stages. Plast Reconstr Surg2012;130:254e-64e.22842422
  8. MK TasoulisFM IqbalS CawthornF MacNeillR Vidya. Subcutaneous implant breast reconstruction: Time to reconsider?Eur J Surg Oncol2017;43:1636-46. 10.1016/j.ejso.2017.04.008. 28528191
  9. H HallbergS RafnsdottirG Selvaggi. Benefits and risks with acellular dermal matrix (ADM) and mesh support in immediate breast reconstruction: a systematic review and meta-analysis. J Plast Surg Hand Surg2018;52:130-47. 10.1080/2000656X.2017.1419141. 29320921
  10. S PotterD BrowningJ SavovićC HolcombeJM Blazeby. Systematic review and critical appraisal of the impact of acellular dermal matrix use on the outcomes of implant-based breast reconstruction. Br J Surg2015;102:1010-25. 10.1002/bjs.9804. 26109277
  11. REG DikmansVL NegenbornM-B Bouman. Two-stage implant-based breast reconstruction compared with immediate one-stage implant-based breast reconstruction augmented with an acellular dermal matrix: an open-label, phase 4, multicentre, randomised, controlled trial. Lancet Oncol2017;18:251-8. 10.1016/S1470-2045(16)30668-4. 28012977
  12. KM HinchcliffH OrbayBK BusseH CharvetM KaurDE Sahar. Comparison of two cadaveric acellular dermal matrices for immediate breast reconstruction: A prospective randomized trial. J Plast Reconstr Aesthet Surg2017;70:568-76. 10.1016/j.bjps.2017.02.024. 28341592
  13. SD MendenhallLA AndersonJ Ying. The BREASTrial: stage I. Outcomes from the time of tissue expander and acellular dermal matrix placement to definitive reconstruction. Plast Reconstr Surg2015;135:29e-42e. 10.1097/PRS.0000000000000758 25539349
  14. D Gschwantler-KaulichP SchrenkV Bjelic-Radisic. Mesh versus acellular dermal matrix in immediate implant-based breast reconstruction - A prospective randomized trial. Eur J Surg Oncol2016;42:665-71. 10.1016/j.ejso.2016.02.007. 26947961
  15. H HallbergR LewinS Rafnsdottir. Acellular dermal matrix support for breast reconstruction after mastectomyhttps://www2.sahlgrenska.se/upload/SU/HTA-centrum/HTA-rapporter/HTA-report%20Acellular%20dermal%20matrix%20support%20for%20breast%20reconstruction%20after%20mastectomy%202017-02-15.pdf2017.
  16. S PotterRL WilsonJ HarveyC HolcombeCC Kirwan. Results from the BRIOS randomised trial. Lancet Oncol2017;18:e189. 10.1016/S1470-2045(17)30164-X. 28368251
  17. A ColwellB DamjanovicB ZahediL Medford-DavisC HertlWG Jr Austen. Retrospective review of 331 consecutive immediate single-stage implant reconstructions with acellular dermal matrix: indications, complications, trends, and costs. Plast Reconstr Surg2011;128:1170-8.
  18. M SorkinJ QiHM Kim. Acellular dermal matrix in immediate expander/implant breast reconstruction: a multicenter assessment of risks and benefits. Plast Reconstr Surg2017;140:1091-100.
  19. S PotterC HolcombeJ Blazeby. Response to: Gschwantler-Kaulich et al. Mesh versus acellular dermal matrix in immediate implant-based breast reconstruction—a prospective randomized trial. Eur J Surg Oncol2016;42:1767-8. 10.1016/j.ejso.2016.05.041 27424788
  20. S PotterEJ ConroyPR WilliamsonBreast Reconstruction Research Collaborative. The iBRA (implant breast reconstruction evaluation) study: protocol for a prospective multi-centre cohort study to inform the feasibility, design and conduct of a pragmatic randomised clinical trial comparing new techniques of implant-based breast reconstruction. Pilot Feasibility Stud2016;2:41. 10.1186/s40814-016-0085-8 27965859
  21. R Coombes. Patients harmed by mesh implants address emotional parliamentary meeting. BMJ2017;358:j3585. 10.1136/bmj.j3585 28743686
  22. American Society of Plastic Surgeons. Evidence-based clinical practice guideline: breast reconstruction with expanders and implants. In: American Society of Plastic Surgeons T,ed., 2013. https://www.plasticsurgery.org/Documents/Health-Policy/Guidelines/guideline-2013-breast-recon-expanders-implants.pdf
  23. L MartinJM O’DonoghueK HorganS ThrushR JohnsonA GandhiAssociation of Breast Surgery and the British Association of Plastic, Reconstructive and Aesthetic Surgeons. Acellular dermal matrix (ADM) assisted breast reconstruction procedures: joint guidelines from the Association of Breast Surgery and the British Association of Plastic, Reconstructive and Aesthetic Surgeons. Eur J Surg Oncol2013;39:425-9. 10.1016/j.ejso.2012.12.012 23321393

Journal Information

Article Information

Author Affiliations

MEDICAL DISCLAIMER NOTICE: To the fullest extent permitted by law, the material and information displayed in The BMJ is provided "as is" without any guarantees, conditions or warranties as to accuracy. We rely on our authors of articles, contractors and third party data providers to confirm the accuracy of information and advertisements presented and to describe generally accepted practices and therefore we as the publisher and editors cannot warrant its accuracy. Differences may occur also between the print and online text of articles and advertisements. Readers should be aware that professionals in the field may have different opinions. Because of this fact and also because of regular advances in medical research we strongly recommend that readers independently verify any information that they chose to rely upon. Ultimately it is the reader's responsibility to make their own professional judgements. Opinions posted on Rapid Responses, the Advice Zone, International Experience and any other parts of the sites are those of the individuals posting them and not the views of BMJ.


This article was written by BMJ from The BMJ and was legally licensed through the NewsCred publisher network. Please direct all licensing questions to legal@newscred.com.