New paper by the Stracy group uses machine learning to predict individual risk and develop personalised treatment strategies
Antibiotics have saved countless lives but their widespread use has led to antibiotic resistance which is increasing throughout the world and threatens our ability to treat dangerous bacterial infections. Despite the significant health and economic burden, little is known about how antibiotic resistance emerges during treatment and we currently lack strategies to prevent it.
Combining whole genome sequencing and machine learning of patient records, new Dunn School PI Dr Mathew Stracy and the group of Professor Roy Kishony (Technion – Israel Institute of Technology) have developed an algorithm to predict individual patients’ risk of treatment-induced emergence of antibiotic resistance. The paper, published in Science, shows how a personalised antibiotic prescribing algorithm reduces the risk of emergence of antibiotic resistance in urinary tract and wound infections by 48%.
“We found that the antibiotic susceptibility of the patient’s past infections could be used to predict their risk of returning with a resistant infection following antibiotic treatment’ explained Dr Stracy. “Using this data, together with the patient’s demographics like age and gender, allowed us to develop the algorithm.”
It is hoped that this work will lead to a more personalised approach to treatment of bacterial infections which will limit the emergence of antibiotic resistance in patients and populations.
Written by Isabella Maudlin (Murphy Lab) @BellaMaudlin