Knowledge Mapping of Multicriteria Decision Analysis in Healthcare: A Bibliometric Analysis

Abstract

Objective: Multicriteria decision analysis (MCDA) is a useful tool in complex decision-making situations, and has been used in medical fields to evaluate treatment options and drug selection. This study aims to provide valuable insights into MCDA in healthcare through examining the research focus of existing studies, major fields, major applications, most productive authors and countries, and most common journals in the domain.

Methods: A bibliometric analysis was conducted on the publication related to MCDA in healthcare from the Web of Science Core Collection (WoSCC) database on 14 July 2021. Three bibliometric software (VOSviewer, R-bibliometrix, and CiteSpace) were used to conduct the analysis including years, countries, institutes, authors, journals, co-citation references, and keywords.

Results: A total of 410 publications were identified with an average yearly growth rate of 32% (1999-2021), from 196 academic journals with 23,637 co-citation references by 871 institutions from 70 countries/regions. The United States was the most productive country (n = 80). Universiti Pendidikan Sultan Idris (n = 16), Université de Montréal (n = 13), and Syreon Research Institute (n = 12) were the top productive institutions. A A Zaidan, Mireille Goetghebeur and Zoltan Kalo were the biggest nodes in every cluster of authors' networks. The top journals in terms of the number of articles (n = 17) and citations (n = 1,673) were Value in Health and Journal of Medical Systems, respectively. The extant literature has focused on four aspects, including the analytic hierarchy process (AHP), decision-making, health technology assessment, and healthcare waste management. COVID-19 and fuzzy TOPSIS received careful attention from MCDA applications recently. MCDA in big data, telemedicine, TOPSIS, and fuzzy AHP is well-developed and an important theme, which may be the trend in future research.

Conclusion: This study uncovers a holistic picture of the performance of MCDA-related literature published in healthcare. MCDA has a broad application on different topics and would be helpful for practitioners, researchers, and decision-makers working in healthcare to advance the wheel of medical complex decision-making. It can be argued that the door is still open for improving the role of MCDA in healthcare, whether in its methodology (e.g., fuzzy TOPSIS) or application (e.g., telemedicine).

Keywords: CiteSpace; R-bibliometrix; VOSviewer; bibliometric analysis; healthcare; multicriteria decision analysis.

Figure 1

Flowchart of the study selection process.

Figure 2

Annual trend chart of publications.

Figure 3

The curve fitting results of annual publications.

Figure 4

Collaboration network of countries/regions. The size of a single node represents the number of publications of the countries/regions. The color of the nodes represents the publication in different countries with the change of time (yellow: earlier, orange: later).

Figure 5

Collaboration network of institutions. The size of a single node represents the number of publications of the institutions. The color of the nodes represents the average citation in different institutions.

Figure 6

Collaboration network of authors. The size of a single node represents the number of publications of the author. Each color represents a different cluster.

Figure 7

References' timezone network. The nodes indicate the co-citation references. The size of the nodes reflects the number of citation frequencies. The colors of the node and line represent different years. The X-axis represents the time.

Figure 8

Co-occurrence network of keywords according to the average publication year. The size of a single node represents the number of publications of each keyword. The color of the nodes represents the emergence time of different keywords (yellow: earlier, orange: later).

Figure 9

The cumulative growth keywords (top 20).

Figure 10

Trend topics. The X-axis represents the year, while the Y-axis is the cumulate occurrences of the keywords.

Figure 11

Thematic map showing clusters and the keywords from 1999 to 2021 identified by the co-occurrence network. The bubble size is proportional to the occurrences of the words in the cluster. The X-axis represents the centrality (i.e., the degree of interaction of a network cluster in comparison with other clusters) and gives information about the importance of a theme. The Y-axis symbolizes the density (i.e., measures the internal strength of a cluster network, and it can be assumed as a measure of the theme's development).