SYSTEMATIC LITERATURE REVIEW AND CRITICAL APPRAISAL OF COST-EFFECTIVENESS MODELS OF ANTIMICROBIAL TREATMENTS
Author(s)
You-Li Ling, PhD1, Alec Morton, BSc, MSc, PhD2, Lotte Steuten, MSc, PhD3, Claudio Jommi, MSc4, Maria Gheorghe, PhD5, James W. Dennis6, Chris Little, MPharm7, Edward Ivor Broughton, BSc, MPH, PhD8.
1Pfizer, New York, NY, USA, 2National University of Singapore, Singapore, Singapore, 3Office of Health Economics, London, United Kingdom, 4Universita del Piemonte Orientale, Milano, Italy, 5Pfizer, Bucharest, Romania, 6Senior Medical Writer, HEOR Ltd, Cardiff, United Kingdom, 7Pfizer Ltd., London, United Kingdom, 8Pfizer, Muenchen, Germany.
1Pfizer, New York, NY, USA, 2National University of Singapore, Singapore, Singapore, 3Office of Health Economics, London, United Kingdom, 4Universita del Piemonte Orientale, Milano, Italy, 5Pfizer, Bucharest, Romania, 6Senior Medical Writer, HEOR Ltd, Cardiff, United Kingdom, 7Pfizer Ltd., London, United Kingdom, 8Pfizer, Muenchen, Germany.
OBJECTIVES: Health economic evaluations typically compare new antimicrobials with older, low-cost agents that are increasingly undermined by rising resistance. This has led to undervaluation, reduced investment, and an insufficient clinical pipeline to address antimicrobial resistance. To address this, a framework recognising the population-level benefits of antimicrobials, STEDI (spectrum, transmission, enablement, diversity, and insurance), has been proposed. However, the extent to which these broader value elements have been integrated into economic evaluations remains unclear.
METHODS: A systematic literature review (SLR) identified all published cost-effectiveness models of antimicrobials approved within the last ten years (2015-2025). Searches were conducted in MEDLINE® and Embase, via Ovid, from database inception to 28 October 2025, supplemented by hand-searching conference proceedings and Health Technology Agency (HTA) websites. Models were critically appraised with respect to structure, analytical approach, and inclusion of STEDI value elements.
RESULTS: In total, electronic database searches identified 485 records, and an additional 15 were identified through supplemental searches. Of the 83 included records, 34 unique models were extracted for data synthesis (conference abstracts were searched for completeness but not extracted). Decision tree models were the most common (11 standalone; 10 combined with a long-term Markov model), followed by patient-level microsimulation models (n=7), disease transmission models combined with a decision tree (n=4), and Markov models (n=2). Only four models described methods for estimating transmission and diversity value; one also included enablement. No models included spectrum or insurance value components.
CONCLUSIONS: Despite consensus on the broader value of antibiotics and the need to reflect this in value estimates, only a small number of studies have operationalised these concepts to estimate population-level value of antibiotics. Consensus-based guidance from regulatory agencies, health economists, and HTA bodies can direct activities to address the scientific and methodological challenges, thereby improving the consistency and appropriate representation of antimicrobial value.
METHODS: A systematic literature review (SLR) identified all published cost-effectiveness models of antimicrobials approved within the last ten years (2015-2025). Searches were conducted in MEDLINE® and Embase, via Ovid, from database inception to 28 October 2025, supplemented by hand-searching conference proceedings and Health Technology Agency (HTA) websites. Models were critically appraised with respect to structure, analytical approach, and inclusion of STEDI value elements.
RESULTS: In total, electronic database searches identified 485 records, and an additional 15 were identified through supplemental searches. Of the 83 included records, 34 unique models were extracted for data synthesis (conference abstracts were searched for completeness but not extracted). Decision tree models were the most common (11 standalone; 10 combined with a long-term Markov model), followed by patient-level microsimulation models (n=7), disease transmission models combined with a decision tree (n=4), and Markov models (n=2). Only four models described methods for estimating transmission and diversity value; one also included enablement. No models included spectrum or insurance value components.
CONCLUSIONS: Despite consensus on the broader value of antibiotics and the need to reflect this in value estimates, only a small number of studies have operationalised these concepts to estimate population-level value of antibiotics. Consensus-based guidance from regulatory agencies, health economists, and HTA bodies can direct activities to address the scientific and methodological challenges, thereby improving the consistency and appropriate representation of antimicrobial value.
Conference/Value in Health Info
2026-11, ISPOR Europe 2026, Vienna, Austria
Value in Health, Volume 29, Issue 12S
Code
SA68
Topic
Economic Evaluation, Health Policy & Regulatory, Study Approaches
Topic Subcategory
Literature Review & Synthesis
Disease
Infectious Disease (non-vaccine)