Andy Murray and hip resurfacing: can a health plan deny it because it is not a total hip replacement?

By Ramon Martins Andrade, attorney (OAB/RJ 188.374)

English translation of an article first published in Portuguese: read the original.

In January 2019, Andy Murray underwent hip resurfacing surgery, the resurfacing of the femoral head. On 6 March, in an interview with the BBC, the Scottish tennis player, then aged 31, said that the operation had gone well and that he was walking without pain, “which hasn’t been the case for pretty much 18 months, two years”. He said that even tying his shoelaces had hurt and that he felt no pressure to return to playing: if the hip allowed it, great; if not, he was happy to be free of pain. In October of the same year, the BBC recorded that he had reached, at the Antwerp tournament, his first tour singles final since 2017, and noted that no player had returned to playing singles after a hip resurfacing. The account is his and concerns his hip; nothing is known, nor is anything asserted here, about the team, the implant or the clinical choices of that treatment.

The question the case leaves for those in Brazil is a different one: when the orthopaedic surgeon proposes resurfacing instead of a total hip replacement, can the health plan refuse on the ground that the technique is not the one listed? And can the patient demand the athlete’s technique? The article answers in layers: what the national implant registries have shown; how three countries have regulated the technique; and how Brazilian law separates the covered procedure, the chosen implant and the resolution of the disagreement between surgeon and health-plan operator.

Resurfacing is not a smaller total hip replacement

In a total hip replacement, according to the description by the British public health service (NHS), the “ball” of the joint, which is the end of the femur, and the “socket” in the pelvis are replaced with metal, ceramic or plastic parts; the head of the femur is cut off and removed. In resurfacing, the surface of the femoral head is removed and covered with a metal cap, instead of the whole joint being removed. The same page records two conditions: the technique does not work well in weaker bones and may be offered to people around 50 or younger. The promise of the method was to spare bone in young, active patients, who will probably need a second operation over the course of their lives.

That promise was tested at scale by the national implant registries, and the result divided patients into very different groups. The reference study is that of Smith, Dieppe, Howard and Blom, published in the Lancet in 2012 with data from the national registry for England and Wales (National Joint Registry, NJR): 434,560 primary hip arthroplasties performed between 2003 and 2011, of which 31,932 were resurfacings. In women, resurfacing had worse implant survival than conventional total hip replacement, whatever the head size: the predicted five-year revision rate for a 55-year-old woman was 8.3% with a 42 mm head and 6.1% with a 46 mm head, against 1.5% with a conventional cemented prosthesis. In men, size decided: 4.1% with a 46 mm head, 2.6% with a 54 mm head, 1.9% with a conventional prosthesis. Only 23% of the men operated on had a femoral head of 54 mm or more. The authors’ conclusion is direct: “We recommend that resurfacing is not undertaken in women”, and the pre-operative measurement should be used to assess the indication in men.

The later literature confirmed the picture without undoing it. With data from the Australian registry to the end of 2023, Wilson and colleagues (Journal of Orthopaedics, 2025) followed for up to 17 years the resurfacing model most used in that country today: survival of 94% at 17 years, comparable to that of another reference resurfacing and lower than that of the registry’s five best-performing conventional prostheses; in men under 65, however, the differences between the three groups were not significant. The authors themselves record that adverse reactions to metal “remain a concern”. A 2025 meta-analysis (Shanaa and colleagues, JBJS Reviews) found a crude revision rate of 7.8% for resurfacing against 16% for large-head metal-on-metal total hip replacement, without statistical significance, and the comparison is with an implant that has fallen out of use, not with today’s conventional prosthesis. And when resurfacing fails, conversion to a total hip replacement resolves the pain in most cases, but it is not simple surgery: the systematic review by Lari and colleagues (JBJS Reviews, 2026), covering 1,719 hips, found as the most frequent causes aseptic loosening (31.4%), adverse reaction to metal debris (27.1%) and fracture of the femoral neck (23.7%), complications in 10.5% and ten-year survival of between 69% and 92%.

The picture from the registries is not that of a technique better or worse in the abstract, but of a technique with a candidate of its own, generally a young, active man with a large femoral head and good bone, and with worse results outside that profile. Hence the legal consequence: neither may the health-plan operator treat resurfacing as a whim, nor may the patient treat it as a right independent of the indication.

Metal-on-metal: the 2010 recall and the follow-up

Resurfacing implants are, as a rule, metal-on-metal articulations, and it was in this type of implant that the recent history of orthopaedics had its most serious episode. On 7 September 2010, the British regulator (MHRA) published alert MDA/2010/069 on the recall of the entire DePuy ASR system, which included a resurfacing cap: the manufacturer had found that the five-year revision rate “is higher than expected across the entire size range”. The agency’s orders were not to implant, to return stock, to inform all patients and to follow them up with clinical examination “at least annually”.

Follow-up remains part of the treatment. The MHRA alert MDA/2017/018 expressly includes resurfacings, differentiates follow-up according to sex, femoral head diameter and model, and states that no single blood cobalt or chromium value predicts the outcome or requires revision; MRI or ultrasound carry more weight. For coverage, this matters because imaging and laboratory follow-up is part of the indicated treatment, and any replacement of the implant is supported by art. 15 of RN 465/2021 (the ANS normative resolution) and by the revision procedure listed in the ANS list of mandatory coverage (the “rol”). How long a prosthesis should last and what revision reveals about a product defect are questions dealt with in the article on how long hip and knee prostheses should last (in Portuguese).

The rol names procedures, not techniques

Annex I of the rol (RN 465/2021, in the version consolidated up to RN 678/2026) lists, in the pelvic girdle group, among others, the procedures “ARTHROPLASTY – SURGICAL TREATMENT”, “PARTIAL HIP ARTHROPLASTY – SURGICAL TREATMENT” and “REVISION OF HIP ARTHROPLASTIES WITH REMOVAL OF COMPONENTS AND IMPLANTATION OF PROSTHESIS”, all for the hospital segments and the reference plan, without a utilisation guideline. There is no line named resurfacing or “surface arthroplasty”. The rol describes procedures, not techniques or implant models. Lei 9.656/1998 (the Health Plans Act) says, in art. 10, § 12, that the rol “constitutes the basic reference” for health plans and “sets the guidelines for health care”; it does not say that each technique needs a line of its own.

The consequence is in the resolution itself. Art. 19, VI, of RN 465/2021 includes in the hospital plan the “orthoses and prostheses linked to the surgical acts listed in the Annexes to this Normative Resolution”, and art. 15 extends coverage to the removal, maintenance or replacement of those materials, “when necessary, as indicated by the treating professional”. Lei 9.656/1998 excludes only, in art. 10, VII, “prostheses, orthoses and their accessories not linked to the surgical act”. If resurfacing is a hip arthroplasty, and technically it is, the most natural reading is that it falls within the listed procedure and that the implant is an accessory to it. A denial that relies solely on the absence of the word resurfacing confuses the name of the procedure with the technique used to perform it.

A health-plan operator that nonetheless maintains that the technique is outside the rol shifts the discussion to § 13 of art. 10 of Lei 9.656/1998, which requires authorisation of the unlisted treatment when “there is proof of efficacy, in the light of the health sciences, based on scientific evidence and a therapeutic plan” or a recommendation from Conitec (the national health-technology assessment committee) or from a renowned international health-technology assessment body. The STF (Supreme Federal Court), in ADI 7.265 (18 September 2025), set five cumulative criteria, among them the absence of an adequate alternative in the rol, scientific proof of efficacy and safety, and registration with Anvisa (the Brazilian health regulator). This is where the route matters: treated as a technique of a listed procedure, resurfacing is a clinical question; treated as a procedure outside the rol, the criterion of an alternative in the rol weighs against the patient, because total hip replacement is, for many hips, an adequate alternative. That is why the medical report must explain why that hip is a candidate for resurfacing and why total hip replacement is not equivalent for it.

Who chooses the implant, and who decides when there is disagreement

Resolution 2.318/2022 of the CFM (Federal Council of Medicine) says, in art. 2, that it is for the treating physician to “determine the characteristics of the orthoses, prostheses and special implantable materials”; in art. 3, that the physician must give clinical justification for the indication; and, in art. 4, that the physician is “prohibited” from requiring an “exclusive supplier or commercial brand”, except with justification when the implant is produced by few manufacturers or by a single one. RN ANS 424/2017 follows the same line in art. 7: “the treating professional has the prerogative of determining the characteristics (type, raw material and dimensions)” of the material, and must justify the indication and offer “at least 3 (three) brands of products from different manufacturers, when available, from among those regularised with ANVISA”; if the physician does not indicate the three brands, or if the operator disagrees with them, the operator “shall convene a medical panel” (junta médica). The same reasoning has already been examined in the articles on the material for ACL surgery (in Portuguese) and on the screw in spinal surgery (in Portuguese).

The panel is the route that regulation offers for technical disagreement, and its procedure is detailed. Under art. 6 of RN 424/2017, the operator must guarantee the panel in every “technical-clinical disagreement over a procedure or health event to be covered”; it is formed by the treating professional, the operator’s professional and a tie-breaker, and “the tie-breaker’s opinion shall be accepted for coverage purposes” (§ 4). The notice under art. 10 must identify the operator’s professional, set out “the reasons for the technical-clinical disagreement” and indicate four names for the tie-break, with their respective qualifications; the treating professional has two working days to maintain the indication or accept the disagreement and, if maintaining it, chooses the tie-breaker from among the four (art. 11). Art. 9, § 1, is categorical: “Under no circumstances may the beneficiary be required to bear the tie-breaker’s expenses”. And art. 4, § 1, prevents the panel from serving as a pretext for delay: the procedure, or the tie-breaker’s conclusive opinion, may not exceed the guaranteed service time limits, now set in RN 566/2022, which replaced the RN 259/2011 cited in the original text.

Two consequences follow from that procedure. An operator that, on its own initiative, swaps resurfacing for a total hip replacement is not discussing brand: it is changing the technique, and that is the disagreement that RN 424 requires to be resolved by a panel, not by an authorisation under a different code. And the panel does not apply to everything: art. 3 rules it out in emergencies, in procedures outside the rol and the contract, and in the “indication of OPME or medication without registration with the National Health Surveillance Agency” (OPME being orthoses, prostheses and special materials). Whoever requests an implant without Brazilian registration does not even reach the panel.

Health registration is the other filter. Lei 6.360/1976 (the law on health surveillance of medicines and health products) says, in art. 12, that no product subject to health surveillance, “including imported ones”, may be “manufactured, offered for sale or delivered for consumption before being registered”. The useful question is not which brand an athlete used abroad, but whether the proposed model is registered with Anvisa. After implantation comes technovigilance, which Anvisa defines as “the system for the surveillance of adverse events and technical complaints involving health products in the post-marketing phase”. The agency also maintains the National Implant Registry (RNI), a system in which hospitals that perform hip and knee arthroplasty must record the procedures, with “data on the implanted products, the professional and the health service where they were performed”; according to the official page, the data “will be useful for improving the regulation of implantable products”. The verb tense says a lot: the page does not provide, as the British registry does, public revision reports by model.

And in the SUS?

The SUS (Brazil’s public health system) performs hip arthroplasty. A 2025 technical opinion from the Technical Advisory Unit for Health Lawsuits (NATJUS) that serves the Federal Court in Rio de Janeiro (No. 0842/2025) records in the SIGTAP table the “primary cemented total hip arthroplasty” (code 04.08.04.008-4) and describes access through the Medium and High Complexity Traumatology and Orthopaedics Network, by way of regulated referral, with Into (the National Institute of Traumatology and Orthopaedics) among the reference units in the state. That record shows the route for total hip replacement; it does not show that the network offers resurfacing. In the SUS, the technique available is the one in the table and in the service’s accreditation, and anyone who needs a technique outside it depends on a medical report explaining why the alternative offered does not meet that patient’s needs.

Three countries, three ways of deciding on the same implant

The United Kingdom decided by measured performance. NICE assessed resurfacing in 2002 (TA44) and, on 26 February 2014, replaced that assessment with TA304, whose recommendation 1.1 says that total hip replacement and resurfacing prostheses are treatment options for end-stage arthritis of the hip “only if the prostheses have rates (or projected rates) of revision of 5% or less at 10 years”. The rule chooses neither technique nor brand: it sets an outcome threshold and leaves to the national registry, the NJR, the task of showing who meets it.

The United States decided by the regulatory route of entry. On 18 February 2016, the FDA published a final order (Federal Register, docket FDA-2011-N-0661) requiring full premarket approval (PMA) for metal-on-metal total hip prostheses, with a deadline of 18 May 2016 for anyone wishing to continue selling them. The order itself records that resurfacings “are not the subject of this order”, because they were already under the more demanding regime: according to the agency’s page, there are two metal-on-metal resurfacing systems approved by PMA, the Birmingham Hip Resurfacing and the Cormet. On follow-up, the FDA acknowledges the importance of metal ion testing, but admits that “insufficient information exists to establish a definitive correlation between metal ion levels and clinical outcomes”. The filter is at the door.

Australia decided by continuous transparency. The national registry of the Australian Orthopaedic Association (AOANJRR) allows studies such as that of Wilson and colleagues, which compare a specific resurfacing model with the country’s five best conventional implants and with another reference resurfacing, separating the results by age and sex. The same technique can be defensible for one subgroup and inappropriate for another, and the difference only appears with public, comparable data. None of the three countries settles the question by the list of procedures: all treat the choice of implant as a clinical decision within a regulated and monitored universe, which RN 424 and Resolution CFM 2.318 attempt to reproduce in Brazil, still without a public registry of outcomes.

Liability: of the operator, the physician and the manufacturer

The discussion of a denial begins with its reasoning: the operator must say whether the refusal is one of coverage (procedure outside the rol) or technical-clinical (wrong indication), because the second necessarily goes through the panel. The Estatuto dos Direitos do Paciente (the Patients’ Rights Statute), Lei 15.378/2026, adds, in art. 18, the right to “seek a second opinion or the view of another professional or service on one’s state of health or the recommended procedures, at any stage of treatment, as well as to have sufficient time to make decisions”, which serves both those who doubt the proposed technique and those who want to confirm it.

If the problem appears after surgery, liability is divided. The Código de Defesa do Consumidor (the Consumer Protection Code) makes the manufacturer and the importer liable, “regardless of the existence of fault”, for harm caused by defects “in the design, manufacture, construction, assembly” of their products, “as well as for insufficient or inadequate information on their use and risks” (art. 12); a product is defective when it “does not provide the safety that may legitimately be expected of it” (§ 1). The hospital is strictly liable for its own services (art. 14), and the physician is liable on a finding of fault (art. 14, § 4), which includes indicating a technique for a patient outside the profile in which it works. The limitation period is five years, counted “from knowledge of the harm and of its authorship” (art. 27), which matters in metal-on-metal, where the reaction to wear can appear years later. One caveat: implant failure is not, in itself, a defect or medical error; only the report of the revision surgery and identification of the model make it possible to separate the causes.

Before deciding

Murray’s case serves to frame the right question, not to answer it for someone else. In my view, a denial based solely on the absence of the word resurfacing from the rol does not stand: the rol names hip arthroplasty, the prosthesis linked to the surgical act goes with the coverage, and the operator’s swapping of the technique is a technical-clinical disagreement that RN 424 requires to be resolved by a panel, with written reasons and a tie-breaker paid for by the operator. On the other side, the registries show that resurfacing has a candidate of its own, and anyone outside that profile, or who is offered an equivalent implant registered with Anvisa, has no right to the model requested. The first document is the orthopaedic surgeon’s written justification, with the femoral head measurement, the bone quality, the three regularised brands and the reason why total hip replacement is not equivalent; the second, the reasoned denial, stating whether the refusal is one of coverage or of indication. The overview of coverage denials is in the guide to health-plan denials (in Portuguese). If the health plan refused the technique or changed the surgery without a medical panel and you want to understand what can be requested, get in touch with the firm.

About the author

Ramon Martins Andrade (OAB/RJ 188.374) is a lawyer who graduated from UFRJ in 2011, with master’s degrees from Université Sorbonne Nouvelle and from the Naval War College (EGN, Brazilian Navy). He is currently a researcher in Human Rights and Health at ENSP/Fiocruz.

This content is for information and education only. It does not constitute legal advice, advertising of results or any guarantee of outcome, under Rule 205/2021 of the Brazilian Bar Association (OAB).