Intended for healthcare professionals

Letters Hip arthroplasty endpoints

Paper’s conclusions will only cause confusion

BMJ 2012; 345 doi: (Published 30 October 2012) Cite this as: BMJ 2012;345:e7005
  1. Simon S Jameson, orthopaedic research fellow1,
  2. Paul Baker, specialist registrar in trauma and orthopaedics2,
  3. David Deehan, consultant orthopaedic surgeon3,
  4. Mike Reed, consultant orthopaedic surgeon4,
  5. James Mason, director of research5
  1. 1Durham University, School for Medicine and Health, Stockton on Tees TS17 6BH, UK
  2. 2Northern Deanery, Newcastle upon Tyne, UK
  3. 3Newcastle Hospitals NHS Trust, Newcastle upon Tyne
  4. 4Northumbria Healthcare NHS Trust, Ashington, UK
  5. 5School for Medicine and Health, Durham University, Stockton on Tees, UK
  1. simonjameson{at}

McMinn and colleagues found that long term mortality was highest after cemented hip replacement and lowest with Birmingham hip resurfacing (BHR) after adjusting for age, sex, and American Society of Anesthesiology (ASA) grade.1 Mortality is initially high after hip replacement but returns to baseline 30-60 days after surgery.2 Multivariable modelling of associations with death is therefore appropriate only during this early period. Additionally, ASA grade is a crude measure of general health, and without data on comorbidities, social deprivation, and health scores survival analyses are flawed.3

The authors hypothesise that cementation (specifically high viscosity) is the cause of this difference in mortality, due to an increased risk of respiratory problems. Although other variables (including cement viscosity) are available in the National Joint Registry dataset, these were not analysed. If cementation accounted for the differences in mortality there would probably be an excess of perioperative deaths because of the “profound embolisation [during intra-operative cementation], which can persist for up to 20 minutes.”1

In a similar analysis conducted using NJR data, 10 813 men had BHR procedures for osteoarthritis,4 rather than the 8352 described by McMinn and colleagues. This discrepancy (almost 25%) is not explained. Furthermore, men with smaller component sizes have higher failure rates.4

Hip resurfacing may have a role in a small proportion of young active male patients. However, there is no evidence of superiority over other hip implants (despite their greater cost), and concerns regarding complications of metal-metal bearings remain.5

In our view, McMinn and colleagues’ paper will cause unnecessary confusion among patients, commissioners, and healthcare providers and will be used by companies to market their brand. We strongly encourage readers to consider our comments before embracing McMinn and colleagues’ conclusions.


Cite this as: BMJ 2012;345:e7005



View Abstract

Log in

Log in through your institution


* For online subscription