The protocol of this systematic review adheres to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis Protocols (PRISMA-P 2015) [11] guidelines (see Additional file 1). The protocol is also registered in the PROSPERO database, the International prospective register of systematic reviews (CRD42022296113).

The PRISMA [12] guidance recommendations have been used to design the systematic review. The systematic review will apply the Consolidated Health Economic Evaluation Reporting Standards (CHEERS) [13] statement to extract relevant data from studies retrieved.

Decision problem review question

The objective of this review is to identify the modelling methods and assumptions used for colorectal cancer screening globally and to derive cost-effective analysis/cost-utility analysis/cost–benefit analysis outcomes. The following specific questions will be addressed in the review.

1.

What type of model (for example, Markov, semi-Markov, discrete event simulation) and model structure is used in a particular CRC screening programme?

2.

What are the model assumptions for each model structure, e.g. time horizons, LYGs, QALYs, costs and other parameters?

3.

What is the associated background mortality with the population used for the building of this model? Are other model parameters included, such as efficiency parameters?

Defining inclusion criteria

A PICOS (Participants, Interventions, Comparator, Outcomes and Study Design) framework was used for the inclusion and exclusion criteria. The study design was specified since randomised controlled trials and cohort studies were deemed unlikely to provide suitable reports on the cost-effectiveness modelling of screening strategies.

Participants—males and females at average risk of CRC, or individuals eligible to receive a CRC screening invitation for a given screening programme

Interventions—interventions for screening for CRC, for example, faecal immunochemical test (FIT), faecal occult blood test (FOBT), flexible sigmoidoscopy, colonoscopy, computed tomography colonography (CTC), double-contrast barium enema or any other screening methods, which are not included in this list

Comparators—we expect the main comparator in most papers to be no screening compared to various screening strategies, for example, varying age ranges of the screening population and thresholds associated with quantitative FIT testing and FOBT testing

Outcomes—incremental cost-effectiveness ratio (ICER), net benefit, life years gained (LYGs), quality-adjusted life years (QALYs), costs, colonoscopy utilisation or any other unit of health gain and screening harm

Study designs—all economic evaluation academic papers which are published and discuss an economic evaluation model are eligible for inclusion in the review

Defining exclusion criteria

Participants—only human participants are to be included in the papers and not animals

Interventions—interventions used outside of the interventions listed above

Study designs—to avoid potentially high bias, case reports and case series are excluded from this review. Randomised controlled trials and cohort studies are needed for the quality evaluation of models but are not directly necessary for this review

Time frame—academic papers published between January 2011 and December 2021

Identifying research evidence

The search strategy was optimised following advice from the Specialist Medical Librarian at the host institution, who suggested using several databases to meet the needs of the review. The databases used for the planned strategy were MEDLINE, EMBASE, Web Of Science and Scopus. Searches were conducted in December 2021 using the.

MESH search terms (see Additional file 2) and updated in January 2022 for the following:

1.

Colorectal Neoplasms OR colorectal cancer OR bowel cancer OR colorectal

2.

screen* OR Mass Screening

3.

Cost–Benefit Analysis OR cost-effective* OR cost-utility OR quality-adjusted life-years OR life-year* gain* OR economic evaluat* OR health technology assessment OR incremental cost-effective* ratio* OR ICER* OR cost analys* OR qaly* OR lyg*

Searches 1, 2 and 3 are then combined by AND. Only English papers were included in this review since there was no funding to translate non-English language papers. The studies which fulfil this criterion were then screened for inclusion by the PICOS criteria.

Study selection

There will be three stages in the study selection process.

1.

All results found from the initial search will be downloaded into the COVIDENCE systematic review manager [14] to facilitate the review management, including the removal of duplicated articles before starting the title and abstract review.

2.

The titles and abstracts of the papers will be screened by two reviewers following the PICOS criteria.

3.

The full text of the eligible papers will be reviewed to further confirm their eligibility.

If the two reviewers disagree on papers, a third reviewer will be introduced and a consensus of the three reviewers will be made. Also, in the case that multiple papers report the same model/method, the most recent paper with the overall model structure will be included, thus resulting in other papers with this model being discarded. The papers excluded will be deemed ineligible according to the PICOS criteria.

A sample of papers will be used to pilot the selection process using the inclusion criteria, to determine whether papers are being selected appropriately. If any noticeable or fundamental changes need to be addressed to the inclusion criteria, they can be iteratively improved by the reviewers to support the objectives.

Data extraction

The extracted data will use the CHEERS checklist as a benchmark to guide the collation of data, for example, extracting the study population, perspective, time horizon and discount rate from the papers. Data extraction will also focus on the model design, parameter selection and modelling methods to examine modelling practices and understand the sources of variation in methods, the likely impact on study quality and therefore its appropriateness for future application. Unfortunately, the CHEERS checklist does not include any items about potential bias within the studies. This is an important potential source of error in modelling and parameter assumptions can introduce unintended biases when reporting the results of the model [15]. This will be a limitation of the paper and will be addressed within the discussion of the review.

Quality assessment

This protocol aims to collate reported methods rather than examine their quality. Therefore, instead of using the Drummond checklist [16] which examines the methodological quality, the CHEERS checklist will be used for quality reporting of economic evaluations [17] as recommended by the EQUATOR Network [18]. Papers eligible will undergo quality assessment, where detailed information on study characteristics will be collected, based on the CHEERS guideline checklist.

Data synthesis

A PRISMA flow chart showing the number of papers remaining at each stage will be presented to document the study selection process.

As the purpose of this review is to examine the modelling methods used to evaluate the health economics of CRC screening, the results will be presented qualitatively. The studies included in the final review will be tabulated by the CHEERS checklist. Some key features will include the type of study, outcome measures and interventions. Thus, methodological study quality will be addressed, for example, the economic model choice, effectiveness measure and cost measures. Examination of the results will also collate any sensitivity analyses performed in the studies to help identify if there are any changes to the study conclusions due to the susceptibility of cost-effectiveness to key model parameters.

Dissemination

Results found from this systematic review will be published in a peer-reviewed journal and disseminated at international conferences and institutional academic workshops.