Modeling of resistance to antibiotics

Modeling of resistance to antibiotics

Pills in a boxThe efficiency of antibiotics has over a long period led to overuse and misuse, causing the bacteria to develop resistance to one or more antibacterial agents. Infections and diseases that were successfully treated by antibiotics are now becoming an increasing public health problem. 

We have conducted a retrospective study aimed at examining the connection between antibiotics resistance and use of different antibiotics, and also the connection between the level of bacteria and use of different antibiotics as such use can increase the number of certain bacteria.

Our data stems from three hospitals in Norway, a country with low levels of antibiotic use. We have developed statistical models to capture such low signals. To reduce noise, we conducted our study at a hospital ward level. We proposed random effect Poisson and binomial regression models, with a reparametrisation that allowed us to reduce the number of parameters to be estimated.

Through scenario simulation, we have studied the potential effects of reduced or increased antibiotic use. When we for example considered the impact of antibiotic consumption on antibiotic-resistant Pseudomonas aeruginosa we found effects of consumption on resistance under conditions with a relatively low use of antibiotic agents. This strengthens the recommendation of prudent use of antibiotics.

Publications

Berild, Dag; Abrahamsen, Tore G.; Andresen, Stein; Bjørløw, Egil; Haug, Ola; Kossenko, Irina M.; Kubar, Olga I.; Lelek, Michaela; Mintchenko, Svetlana I.; Pyasetskaya, Maria F.; Ringertz, Signe H.; Sysenko, Galina A. (2008). A controlled intervention study to improve antibiotic use in a Russian paediatric hospital. International Journal of Antimicrobial Agents, Vol. 31, pp. 478-483. DOI: 10.1016/j.ijantimicag.2008.01.009

Aldrin, M., Raastad, R., Tvete, I.F., Berild, D., Frigessi, A., Leegaard, T., Monnet, D.L., Walberg, M. and Müller, F. (2013). Antibiotic resistance in hospitals: a ward specific random effect model in a low antibiotic consumption environment.  Statistics in Medicine, Vol. 32, Issue 8, pp. 1407–1418. DOI: 10.1002/sim.5636.

Department

Financing

Sykehuset Asker og Bærum HF