Fig 1.
Causal loop diagram of the “Limits to Growth” archetypal structure.
Arrows with a + cause variables to change in the same direction, while arrows with a – cause variables to change in the opposite direction. Feedback loops are labelled R for reinforcing (positive feedback) or B for balancing (negative feedback).
Fig 2.
CLD representing AMR as a Limits to Growth archetype.
Archteype shown a) in its basic form and b) with surveillance providing an information connector between susceptibility and AMU.
Fig 3.
System Dynamics model structure representing AMR as a Limits to Growth archetype.
Structure pictured a) in its basic form, with prescribing determined by anecdotal success with the antimicrobial, and b) with surveillance providing an information connector between susceptibility and AMU.
Table 1.
Model variables and the biological or behavioural phenomena they represent.
Table 2.
Summary of sensitivity analysis results.
Fig 4.
Runs varying timescale parameters (time horizon of clinician judgment and timescale of microbial evolution) and relative fitness of the resistant form in presence and absence of the antimicrobial, for both variants of the model structure.
Fig 5.
Results varying timescale of microbial evolution for the surveillance-based prescribing variant of the model: a) sensitivity runs with microbial evolution timescale ranging 0.1-26 weeks, b) effect of parameter values ranging 0.1-2 weeks on number and amplitude of oscillations prior to stabilisation, and effect of parameter value on maximum amplitude of oscillations and time for susceptible fraction to stabilise with parameter ranges of c) 0.1-26 weeks and d) 0.1-2 weeks.