A new study finds that using digitally supported algorithms to guide healthcare providers' diagnosis and treatment could significantly reduce the number of antibiotics prescribed to critically ill children without compromising treatment success or causing harm. It is well known that the growing problem of bacterial antimicrobial resistance (AMR) worldwide has resulted in a large number of pathogens becoming resistant to antibiotic treatments. One of the main causes is inappropriate and overprescription of antibiotics.
Electronic Clinical Decision Support Algorithms (CDSA) are digital health or mobile health tools that guide healthcare providers on which symptoms and signs to assess, which tests to recommend, and recommend appropriate diagnosis, treatment, and management. Researchers at the University of Lausanne (UNIL) in Switzerland developed CDSA to guide antibiotic prescribing in pediatric patients.
CDSA, called ePOCT+, is a digital clinical decision support algorithm used to guide healthcare providers in the management of acutely ill children under 15 years of age. It uses the results of basic tests (C-reactive protein (CRP), hemoglobin, and pulse oximetry) to provide recommendations on potential treatments.
To examine the impact of ePOCT+ on antibiotic prescribing compared with usual care, researchers conducted a randomized controlled trial in primary care settings in Tanzania. In Tanzania and many other resource-limited countries, more than 50% of sick children receive antibiotics at the time of medical consultation, and 80% to 90% of these are prescribed on an outpatient basis; most are considered inappropriate.
Over 11 months, 23,593 consultations from 20 ePOCT+ practices and 20,173 consultations from 20 usual care practices were included. Study participants ranged in age from 2 months to 14 years. The researchers found that the overall antibiotic prescribing rate at the initial consultation was 23.2% when using ePOCT+, compared with 70.1% in usual care settings that did not use the algorithm.
This equated to nearly three times less likelihood of receiving an antibiotic prescription for sick children when the algorithm was used than when it was not used. The reduction in antibiotic prescribing in ePOCT+ facilities did not result in an increase in patient adverse events, as no difference in treatment failure rates was observed.
The researchers noted that despite being in the intervention group, nearly 25% of patients were not managed with ePOCT+, which they attributed to low uptake of the tool due to the need for multiple entry of clinical data, including data from the electronic medical record (EMR), ePOCT+, paper logs, and longer consultation times. Researchers are assessing the issues that prevent health care providers from using ePOCT+ and similar tools so that these tools can be fully rolled out in Tanzania and other countries.
"Wide implementation of ePOCT+ could help address the urgent problem of antimicrobial resistance by reducing overprescribing of antibiotics in sick children while maintaining clinical safety," the researchers said.
The study was published in the journal Nature Medicine.