28 interventions not to do


25 Feb 2015

The Choosing Wisely campaign that focuses on cutting down the numbers of unnecessary medical tests and treatments routinely used in primary care has just published its list of 28 recommendations of tests and treatments not to do. This list will be whittled down to 5 and published when the campaign launches this April in Australia.

In his blog Dr Justin Coleman, Chair of the RACGP working group for the NPS MedicineWise backed campaign writes:

To make the list:

  • the intervention must be done often – indeed, too often!
  • it must result in significant cost or harm, and
  • the recommendation against the intervention must be backed by evidence.

If you are an Australian GP, you can vote on which of the top ten  you think should make the cut, at the RACGP Choosing Wisely poll. You may not ever get to see the other 18, which is a pity, and why I’m publishing them here. They are not ‘official’ recommendations endorsed by anyone at this stage – we will work up the top five with full evidence and references, ready for official publication.

The 28 below are in rough order of what I think is important. As with all such lists, opinions will vary. I’d like to hear yours, so please do comment.

Because we have deliberately avoided the easy decisions and have concentrated on things which many GPs actually do, GP opinion is almost by definition divided for many of the recommendations. We don’t pretend to be adamant, but we do feel that the list is based on evidence.

This list is not trivial. Even if you disagree with a few here or there, I believe that if most GPs followed these recommendations in most cases, it would substantially improve patient safety, reduce harms caused by overtesting, overdiagnosis and overtreatment and, as a bonus, free up tens of millions of dollars annually within our health system.

List of recommendations

Top ten:

  1. Don’t commence medications for hypertension or hyperlipidaemia without first assessing CV risk, to guide the need for therapy.
  2. Don’t order screening lipid tests more often than 5-yearly in a person who is low-risk (i.e. absolute 5-yr CV risk < 10%).
  3. Don’t advocate routine self-monitoring of blood glucose for people with type 2 diabetes who are on oral medication only.
  4. Don’t use cardiovascular imaging (CT coronary calcium score, stress test, carotid u/s) to screen low-risk patients (i.e. absolute 5-yr CV risk < 10%, and asymptomatic).
  5. Don’t routinely use prostate specific antigen (PSA) as a screening tool. If ordering PSA, first inform the patient of the benefits and harms.
  6. Don’t order imaging for back pain (XR, CT, MRI) except to investigate fracture, tumour, infection or cauda equina syndrome.
  7. Individualise the HbA1c target in type 2 diabetes, and don’t aim too low in the elderly and those with a long duration of diabetes.
  8. Avoid, wherever possible, addictive medications (benzodiazepines, opioids) for chronic conditions, especially in those with a history of mental health conditions or substance abuse.
  9. Don’t use long-term anti-acid therapy without at least annual attempts to down-titrate or cease, unless Barrett’s oesophagus.
  10. Don’t use antipsychotics first-line in older patients for insomnia or dementia.

Eighteen more:

[n.b These are hitherto unpublished, so I may not have worded them quite as accurately, or included all caveats.)

  1. Don’t use shoulder imaging for straightforward shoulder pain.
  2. Don’t use antibiotics in asymptomatic bacteruria, except in pregnancy.
  3. Don’t order tumour markers as a general screening tool (e.g.CEA, PSA).
  4. Don’t order knee ultrasounds.
  5. Don’t order FBC, B12 or iron studies as a ‘routine annual screen’: order them if you have a reason to do so.
  6. Don’t use ultrasound guidance for shoulder joint injections.
  7. Don’t order upper abdominal ultrasound to investigate mildly abnormal LFTs.
  8. Don’t use antibiotics for otitis media (for age 2-12 years, non-Indigenous) where reassessment is a reasonable option.
  9. Don’t make ‘pre-disease’ diagnoses (e.g. pre-hypertension, pre-diabetes, osteopaenia) if doing so will tempt you to instigate medications.
  10. Don’t prescribe antibiotics for acute sinusitis (unless >7 days or worsening). Nor for pharyngitis or bronchitis.
  11. Avoid unnecessary ‘routine screening’ tests in pregnancy unless targeted reason. E.g. UEC, LFT, thalassaemia screen, full iron studies [ferritin alone is more useful], toxoplasma and listeria.
  12. Avoid early routine dating scans in pregnancy unless there is doubt about a woman’s dates.
  13. Don’t advocate screening mammography in women less than 50 years of age unless high risk, nor for women over 70 years of age.
  14. Don’t order ankle imaging unless the Ottawa ankle rules indicate it (ditto cervical spine imaging without applying Ottawa C-spine rules).
  15. Don’t treat osteoporosis without explaining risk-benefit to the patient in absolute terms, including a fracture risk calculation. Low BMD is not enough.
  16. Don’t order serological testing for herpes simplex.
  17. Don’t use anti-acid medication in unsettled infants.
  18. Don’t routinely do pelvic or breast examination before prescribing oral contraceptives.

The above list was developed along with RACGP members Drs Evan Ackermann, Jenny Doust, Simon Morgan and Rob Hosking. Thanks to Joanna Ong for logistics support, and to NPS MedicineWise for bringing this excellent initiative to Australia.


Already a member?

Login to keep reading.

Email me a login link