Systematic literature reviews (SLRs) are fundamental for aggregating published evidence, identifying knowledge gaps and informing health economic evaluations, especially in the field of novel diagnostics where randomised controlled trials are commonly absent [1]. The global publishing output within the field of novel diagnostics is increasing rapidly owing to increasing efforts towards precision medicine. Simultaneously, the global publishing output within the field of health economics is increasing rapidly owing to increasing efforts towards value-based healthcare and present budget constraints. The number of publications focussing on the health economic impact of diagnostics almost doubled from 86,244 in 2010 to 152,404 in 2025 (PubMed search using MeSH terms ‘Diagnostic Techniques and Procedures’ and ‘Health Care Economics and Organizations’). Consequently, the workload of performing SLRs is increasing as the number of articles that requires screening grows larger. The average time to complete an SLR is over 15 months, while the proportion of truly relevant articles for data extraction may be as low as 1% of the total search results [2]. The high workload may reduce researchers’ willingness to conduct an SLR or may lead to search strategies that are too narrow when prioritising time constraints over review quality. Moreover, it may render an SLR outdated by the time it is published.

Background and Objective
This systematic literature review addresses model-based cost-effectiveness studies for therapy response monitoring with positron emission tomography (PET) generally combined with low-dose computed tomography (CT) for various cancer types. Given the known heterogeneity in therapy response events, studies should consider patient-level modelling rather than cohort-based modelling because of its flexibility in handling these events and the time to events. This review aims to identify the modelling methods used and includes a systematic assessment of the assumptions made in the current literature.

Methods
This study was conducted and reported following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 statement. Information sources included electronic bibliographic databases, reference lists of review articles and contact with experts in the fields of nuclear medicine, health technology assessment and health economics. Eligibility criteria included peer-reviewed scientific publications and published grey literature. Literature searches, screening and critical appraisal were conducted by two reviewers independently. The Consolidated Health Economic Evaluation Reporting Standards (CHEERS) were used to assess the methodological quality. The Bias in Economic Evaluation (ECOBIAS) checklist was used to determine the risk of bias in the included publications.

Results
The search results included 2959 publications. The number of publications included for data extraction and synthesis was ten, representing eight unique studies. These studies addressed patients with lymphoma, advanced head and neck cancers, brain tumours, non-small cell lung cancer and cervical cancer. All studies addressed response to chemotherapy. No study evaluated response to immunotherapy. Most studies positioned PET/CT as an add-on modality and one study positioned PET/CT as a replacement for conventional imaging (X-ray and contrast-enhanced CT). Three studies reported decision-tree structures, four studies reported cohort-level state-transition models and one study reported a partitioned survival model. No patient-level models were reported. The simulation horizons adopted ranged from 1 year to lifetime. Most studies reported a probabilistic analysis, whereas two studies reported a deterministic analysis only. Two studies conducted a value of information analysis. Multiple studies did not adequately discuss model-specific aspects of bias. Most importantly and regularly observed were a high risk of structural assumptions bias, limited simulation horizon bias and wrong model bias.

Conclusions
Model-based cost-effectiveness analysis for therapy response monitoring with PET/CT was based on cohorts of patients instead of individual patients in the current literature. Therefore, the heterogeneity in therapy response events was commonly not addressed appropriately. Further research should include more advanced and patient-level modelling approaches to accurately represent the complex context of clinical practice and, therefore, to be meaningful to support decision making.

Registration
This review is registered in PROSPERO, the international prospective register of systematic reviews funded by the National Institute for Health Research, with CRD42023402581.