3.1 Study Selection Process

After removing duplicates, 3340 records were screened by title and abstract, leading to the exclusion of 2,512 and the advancement of 828 sources to full-text review. Of these, 463 were excluded, and 365 studies were deemed eligible for content evaluation. This final step excluded 339 studies, resulting in 26 studies meeting all inclusion criteria (Fig. 1 and Supplementary Methods 2; see the electronic supplementary material).

Figure 1.

PRISMA flow chart [25, 26]. PRISMA Preferred Reporting Items for Systematic Reviews and Meta-Analyses

3.2 Description of the Included Studies3.2.1 Structure and Design of the Economic Model3.2.1.1 Type of the Economic Analysis

Seventeen studies were full economic evaluations, while nine were partial economic evaluations and cost analyses. Among the full economic evaluations, ten were cost-utility analyses (CUAs) [34,35,36,37,38,39,40,41,42,43], three were cost-effectiveness analyses (CEAs) [44,45,46], two were cost-consequence analyses [47, 48], and other two deployed both CEA and CUA [49, 50]. The partial economic evaluations encompassed three cost-of-illness (COI) analyses [51,52,53], three burden-of-illness (BOI) analyses [54,55,56], one comparative cost analysis [57], one social return on investment (SROI) analysis, and one combined COI and budget impact analyses [58] (Supplementary Figure 1; see the electronic supplementary material).

3.2.1.2 Model Type and Methodological Framework

Among the 17 full economic evaluations, 16 studies reported the type of model used for conducting the economic analysis [34,35,36,37,38,39,40,41,42,43, 45,46,47,48,49,50]. Of these 16 studies, 11 relied on Markov models [34, 38, 40,41,42,43, 45, 47,48,49,50], three employed decision-tree models [36, 39, 46], one utilised the Markov decision-tree model [35], and one developed an epidemiological model [37]. The remaining full economic evaluation study did not report the model type used [44]. Additionally, of the nine partial economic evaluations seven followed a methodological framework for estimating economic and health outcomes. This framework was prevalence based in six studies [51,52,53, 56,57,59] and incidence based in one study [54] Two additional studies used alternative terminology to describe their methodological approach (‘population-based’ [55] and ‘forecasting’ [59]; Table 1).

Table 1 Characteristics of the included studies (n = 26)3.2.1.3 Time Horizon, Health States, and Cycle Length

All the studies conducted the economic analysis over a single or multiple time horizons, of which ten used a lifetime horizon [34, 37, 38, 40,41,42, 45, 46, 50, 54], while eight adopted a 1-year time horizon [39, 51,52,53, 56,57,58,59]. Three studies reported results for both the lifetime and other time horizons [47, 48, 50]. Three other studies used time horizons ranging from 5 to 15 years [35, 43, 55]. One study used a time horizon from 3 to 50 years of age [44], while the remaining study used a time horizon of 1 month [36] (Table 1). Among the 17 full economic evaluations, 14 studies reported the number of health states used in their models [34,35,36, 38,39,40,41,42,43, 46,47,48,49,50], ranging from two [41] to 80 [38], and 11 studies specified the model cycle length [34, 35, 38, 40,41,42,43, 47,48,49,50], varying from 1 week [43,