VaxiMap: optimal delivery of vaccinations for housebound patients

BackgroundThroughout the UKs Covid-19 vaccination campaign, responsibility for vaccinating housebound patients has rested with individual GP surgeries, posing them a difficult logistical challenge (the travelling salesman problem). In response to demand from GPs, and a lack of existing solutions tailored specifically to vaccination, VaxiMap was created. This tool provides optimal routes for vaccine delivery and has been free to all users since its inception in January 2021. MethodsVaxiMap generates optimal routes subject to the constraint that the number of patients per route should be fixed. This ensures that a known quantity of vaccine can be set aside for each route and minimises wastage. The user need only upload an Excel spreadsheet of patient postcodes to be visited. A divide-and-conquer approach of iterative k-means clustering followed by within-cluster route optimisation is used to generate the routes. FindingsWe find substantial savings in the time taken to plan vaccinations, as well as savings in the time taken to visit housebound patients. We estimate total savings to date of 4,700 hours of practitioner time, equivalent to 2.5 work-years, or approximately {pound}91k at typical practitioner salaries. InterpretationThe adoption of VaxiMap yielded both time and cost savings for GP surgeries and accelerated the UKs Covid-19 vaccination campaign at a critical moment. FundingFinancial support for VaxiMap was provided by Magdalen College, Oxford, Oxford University Innovation, and JHubMed, part of UK Strategic Command. These parties were not involved in the preparation of this manuscript.