This page explores the concept of risk-based prescribing of antibiotics when treating common infections. This is a relatively novel idea, and takes into account a variety of factors (age, gender, patient history e.g. past hospitalisations) to determine a patient’s risk level for developing further complications, as recorded by hospitalisation in a 30 day window following the initial GP consultation.
Browse each plot below to find out more
Our research findings indicate that prescribing of antibiotics is largely unrelated to the risk level of the patient, suggesting that there is a significant opportunity to optimise antibiotic prescribing. Currently many very low risk patients, who have a strong chance of recovering on their own, receive prescriptions and equally there are a number of high risk patients who don't receive any treatment.
Does prescribing match patient risk levels?
This graph splits patients into five groups from the lowest to the highest risk (as assessed by a clinical risk prediction model), and shows the corresponding prescribing rate for each of the groups. Ideally, the majority of the highest risk patients should receive an antibiotic prescription, whereas the majority of the lowest risk patients should not. The red line shows a theoretical ideal prescribing situation, whereas the blue line shows the current state of prescribing in primary care.
Table showing prescribing by risk level
Table:
This table shows the raw data displayed in the plot opposite. Notice that prescribing is very uniform across all the risk groups, suggesting the chance of being given an antibiotic is independent of a patient's risk of developing further complications and suffering a hospitalisation.
Do we prescribe according to risk for different age-groups and by gender?
This graph shows how well prescribing matches patient risk levels for various groups (determined by age and gender). The risk score is a number between 0 and 100, which measures the accuracy of prescribing to patient risk levels. The higher the number, the more closely prescribing matches the risk level of patients.
How well do we prescribe according to risk for each indication?
This graph shows how well prescribing matches patient risk levels for each indication. The risk score (as above) is a number between 0 and 100, which measures the accuracy of prescribing to patient risk levels. The higher the number, the more closely prescribing matches the risk level of patients.