Algorithms could help reduce avoidable antibiotic prescribing

Algorithms could be key to curbing global antibiotic resistance, with doctors who used the tool prescribing significantly less than those who did not.
 
A new study has analysed the data of more than 44,000 consultations of children under the age of 15 across 20 Tanzanian health facilities.
 
Doctors were randomly assigned to use a digital algorithm tool to inform their prescribing, or to carry out their usual care.
 
The study found those following the algorithm prescribed antibiotics in 23.2% of consultations, compared to 70.1% of consultations where the tool was not used.
 
The researchers, from the University of Lausanne in Switzerland, confirmed there was no difference in treatment success between the two groups and no harm was caused.
 
They found safety outcomes, including death and non-referred secondary hospitalisations, by day seven showed no significant difference.
 
And although Tanzania is a long way from Australia and offers a very different healthcare environment, the research comes as bacterial antimicrobial resistance grows, responsible for the deaths of around 1.27 million people across the world in 2019.  
 
Last year, the World Health Organization ranked antimicrobial resistance as one of the top 10 public health threats facing humanity.
 
The ePOCT+ digital clinical decision support algorithm combines an individual’s health information with the health worker’s knowledge, including point-of-care tests to provide advice on diagnosis.
 
However, Professor Mark Morgan, Chair of RACGP Expert Committee – Quality Care, told newsGP while there is a need to improve primary care antibiotic stewardship, he said ‘simple algorithms that score points for certain clinical features can be very hit and miss’.
 
‘In clinical practice the decision to prescribe is a complex one that involves numerous data points from the history and clinical examination,’ he said.
 
‘These are then overlaid by patient ideas, expectations, and social setting. It is often disappointing to reduce this to an over-simplified scoring system.
 
‘Any combination of point-of-care testing and clinical feature checklist would need to be carefully evaluated in many and varied Australian settings to understand how it works on the ground.’
 
Digital clinical decision support algorithms (CDSAs) have risen in popularity as doctors and researchers around the world become more aware of over-prescribing.
 
In 2022, more than one-third of Australians had at least one antibiotic prescription, and around 88% of antibiotics were prescribed by GPs.
 
‘GPs give antibiotics because they think patients want them, even when they might not be necessary. Although, in reality they are after symptom relief,’ newsGP previously reported.
 
Australian GPs have also warned against prescribing antibiotics as a ‘default position’, instead placing an emphasis on ensuring patients are informed of the available alternatives.
 
Recommended strategies include exploring patient understanding and expectations, offering delayed prescribing, and using appropriate non-antibiotic symptomatic management.
 
Aside from the overuse of antibiotics contributing to the rise of ‘superbugs’, Professor Morgan said resistance is also worsening as organisms are spread around the world by travel.
 
‘Avoiding low-value or too broad-spectrum antibiotic use is one of the ways to slow the development and spread of antimicrobial resistance,’ he said.
 
‘Antimicrobial stewardship is also important for the individual patient and their household because antibiotic use leads to selective advantage for any superbugs.’
 
Professor Morgan said there is no ‘magic bullet’ to improve primary care antibiotic stewardship, but there are many strategies GPs can utilise.
 
In 2022, he co-authored a review of successful evidence-based strategies to improve primary care antibiotic stewardship, ultimately finding a suite of strategies are needed.
 
‘These included pack size and prescribing restrictions, public and clinician awareness campaigns and clinical strategies such as clinical decision support, audits and patient information sheets,’ Professor Morgan said.
 
‘The use of point-of-care testing for C-reactive protein, procalcitonin and streptococcal infections can help identify which patients are most likely to benefit from antibiotics, but it costs time and money to do this.’
 
Log in below to join the conversation.

antibiotics antimicrobial resistance clinical decision tool overprescribing