Fig 1.
Study design for scoping review research (Prisma-ScR) with a topic modeling approach.
This design was adapted from Page et al. (2020) and purposes to identify the studies in databases and other methods through the query elaborated from the research question. After selecting studies by eligibility criteria, the titles, keywords, and abstracts were subjected to detecting terms using the topic modeling approach [23]. Terms were aggregated by semantic similarity into clusters (topics), summarized, and presented.
Fig 2.
Flow diagram for scoping review research (Prisma-ScR).
We identified 2,232 academic publications in initial search. After duplicates, and ineligible exclusions (n = 1,600), we included 41 gray literature records. In total, were selected 673 abstracts, titles, and keywords for topic modeling screening step.
Fig 3.
Detected terms and clustering.
A) 30 most cited terms detected by the topic modeling automation tool and each number of citations on the database. 26.9% of citations correspond to the neutral term reproductive. This was overestimated for being present in all titles and keywords and repeated in abstracts. Disregarding this term, we identified specific terms such as industry (586 citations) and surrogacy (345 citations); B) 10 Clusters of 121 detected terms aggregated by semantic similarity. We chose to organize the clusters based on the total number of citations (orange bars). The number of topics in each cluster is available in the blue bars.
Table 1.
Topics clusters of the detected terms.
Fig 4.
Location of the main themes of legislative changes within the main topics identified.
The main themes of recent legislative changes stated by the IFFS (2019) coincide with the terms identified by topic modeling. In this way, we identified in which topics the similar terms were inserted and prepared this graph to illustrate the location of the main themes of legislative changes in relation to the main topics in assisted reproduction. Note that each topic may be included in more than one topic due to the complexity of the market.