Health-economic modelling of infectious disease diagnostics: current approaches and future opportunities

Objectives: Antimicrobial resistance (AMR) is a public health threat; infections with resistant organisms are estimated to cause over 650.000 infections and over 30.000 deaths in Europe. AMR is associated with antibiotic consumption: appropriate prescribing of antibiotics is key in combating AMR. To fight this threat, it has been suggested that point-of-care diagnostics to inform antibiotics prescribing are an important tool in reducing antibiotics prescriptions. With the objective of knowing the state of the art, we reviewed diagnostic cost-effectiveness analyses (CEAs) for infectious disease, focussing on model types, effects on prescriptions and AMR. 
Methods: We searched the literature comprehensively through the PUBMED, Web of Science and EMBASE databases, as well as grey literature for the period 2000-2018. We included economic evaluations for diagnostic strategies for infectious disease in all geographic areas. Data extraction was based on the Consolidated Health Economic Evaluation Reporting Standards statement, with an additional focus on the modelling type and inclusion of AMR.
Results: In general, CEAs of diagnostics for infectious disease fall in two categories: trial-based cost-effectiveness, focusing on short-term outcomes, and long-term models, focusing on outcomes extrapolated beyond the trial. Outcomes measured are for example prescriptions saved or hospitalizations saved. Cost-utility analyses seem to be difficult, as QALY differences between various diagnostic strategies are minor. Antimicrobial resistance is considered indirectly: as an ICER with prescriptions saved as an outcome; by calculating the threshold cost of resistance that would change the conclusion of cost-effectiveness; or as a point estimate; often, it is not considered at all.
Conclusions: CEAs of diagnostic strategies for infectious disease are relatively scarce and it can be challenging to prove the cost-effectiveness of diagnostic strategies, especially in the context of relatively inexpensive treatments, such as antibiotics. However, there is an important opportunity of including the societal costs of AMR in future CEAs.