Could MRI be causing overdiagnosis and overtreatment of knee pathology in Australia?  — ASN Events
  1. Schedule
  2. Poster Viewing II (even numbers)
  3. Could MRI be causing overdiagnosis and overtreatment of knee pathology in Australia? 

Could MRI be causing overdiagnosis and overtreatment of knee pathology in Australia?  (#138)

Katy Bell 1 2 , Leon Smith 1 , Rachelle Buchbinder 3 4 , Ian Harris 5 6 , Jenny Doust 2 , Paul Glasziou 2 , Alexandra Barratt 1
  1. School of Public Health, University of Sydney, Sydney, NSW, Australia
  2. Bond University, Gold Coast
  3. Cabrini Institute , Melbourne
  4. Monash University , Melbourne
  5. Ingham Institute , Sydney
  6. University of NSW , Sydney

Aim To apply markers of overdiagnosis to data on use of knee MRI in Australian population

Methods Using the published literature and Medicare Benefits statistics we applied the following markers of overdiagnosis:

  1. earlier detection of the target condition (disease) 
  2. new disease definition includes high proportion of biologically indolent conditions 
  3. increasing diagnosis and treatment of the target condition
  4. balance of benefit: harm of diagnosis and treatment of the target condition is borderline or unfavourable

Results

  1. Earlier detection of target condition: MRI is sensitive  for knee pathology as it can visualise all joint components (cartilage, menisci, bone, ligaments and other soft tissues)  in contrast to Xray. Use of MRI may be implicitly lowering the diagnostic threshold for knee “pathology” or enabling the creation of a new disease definition of “pre-osteoarthritis”
  2. High proportion of biologically indolent conditions: 90% of adults (symptomatic or asymptomatic) 50 years and over have knee “abnormalities” detectable by MRI
  3. Increasing test usage and treatment rates: Utilization of knee MRI increased strongly per 100,000 population from 95/100 000 in 1999 (knee MRI first listed on MBS 1998) to 513 in 2013. A further increase was seen to 971/100,000 in 2016 (following subsidy for use by primary care doctors in 2013). For those aged over 55, it increased from 80/100 000 in 1999 to 1391/100 000 in 2016.                                      In contrast, knee arthroscopy, measured by anaesthesia rates for this, has decreased from 256/100000 in 2002 (anaesthesia for arthroscopy first listed on MBS 2001) to 226/100 000 2016. For those aged over 55, it decreased from 375/100 000 in 2002, to 345/100 000 in 2016.
  4. Balance of net benefit/harm:  High quality RCTs have shown that arthroscopic intervention for common degenerative conditions in the knee (often diagnosed by MRI) has no benefit over placebo.

Conclusions These results suggest that the increased utilisation of knee MRI in Australia has not translated into increased rates of knee arthroscopy. Further research is needed to  investigate whether there has been changes in the (age adjusted) rates of knee joint replacements (total and partial) and of other possible treatments for MRI detected abnormalities. Qualitative studies may provide more information on how knee MRIs are being used for patient care, and whether they offer any potential benefits to patients.  Definitive evidence on the incremental value of knee MRI above and beyond plain X-ray will require new RCTs which include MRI in the diagnostic work-up of knee pain.