Modeling antimicrobial cycling and mixing: Differences arising from an individual-based versus a population-based...

Area of Science:

• Mathematical modeling
• Infectious disease epidemiology
• Clinical microbiology

Background:

• Antibiotic resistance is a major threat to hospital infection management.
• Treatment strategies like "mixing" and "cycling" are debated but interpreted differently by modelers and clinicians.
• Standard models often overlook individual patient drug use, potentially limiting clinical relevance.

Purpose of the Study:

• To compare a novel, patient-centric mathematical model with a traditional population-based model for hospital bacterial infection spread.
• To investigate how accounting for individual patient drug use affects predictions of infection management strategies.
• To identify scenarios where the patient-centric model yields different optimal strategies compared to the standard model.

Main Methods:

• Developed two deterministic mathematical models using differential equations to simulate bacterial infection spread in hospitals.
• Model 1: Population-based perspective (traditional approach).
• Model 2: Patient-centric perspective, incorporating individual drug use patterns.

Main Results:

• The patient-centric model produced different predictions compared to the standard population-based model.
• Examples demonstrated that the optimal strategy could be reversed in the new model.
• The patient-centric approach identified strategies that maximized uninfected patients or minimized the spread of double resistance differently.

Conclusions:

• Traditional models offer valuable insights but require careful interpretation due to their population-level focus.
• A patient-centric modeling approach provides a more nuanced understanding of infection dynamics and treatment efficacy.
• Incorporating individual patient drug use is crucial for accurate prediction and effective management of hospital-acquired infections.