Modelling Challenge
As part of the Summer School, we would like to split you into groups to do some interesting / fun pilot research projects.
Project 1. Hospital Screening
AMR tends to be a major problem in the health-care setting; for this reason individuals are screened before scheduled operations and only undergo treatment if they are clear of AMR bacteria. As a theoretical / modelling project, it would be interesting to consider the optimal management of hospital screening.
You may wish to assume that transmission occurs randomly both within and outside the hospital setting; screening takes time and has an associated cost; cancelling treatment is costly unless a (screened and negative) replacement can be found; an AMR outbreak within the hospital is extremely costly. It would be interesting to find the optimal screening time for incoming patients as well as those remaining within hospital. Could screening be reactive?
Project 2. Antimicrobial Usage.
Prescribing of Antimicrobials is the biggest challenge to the development of AMR. In this project we take a statistical look at prescribing across the UK. Links to raw data sources are at the side, but to get you started here are some useful data sources:
- GP Prescriptions: January, February, March 2016. Jan 2016, 2015, 2014, 2013, 2012, 2011
- GP Locations: February 2016 (last 4 columns are X,Y and Lat, Long)
- GP Patients by Age: March 2016
- Key to BNF codes (Codes beginning 0501 are antibiotics)
Multiple questions can then be adressed using this data.
- Trends overall, and trends within practices
- Differences between practices
- Relationship between patient-mix (age and gender) vs. prescribing
- Relationship between overall levels of prescribing (i.e., all drugs) and antibiotic prescribing. Some GPs might be more prone to prescribe and could be a sign of a less healthy population
- Shifts in prescribing (which antibiotic is chosen)
- Seasonality (will require you to download extra files)