Using association analysis on the literature to identify the use of informatics in public health
Abstract
Background: With growing applications of informatics in various domains of public health, there is a need to examine the evolving scope of public health informatics in its development and advancement toward achieving the goals of public health. The primary objectives of this study were, through data mining on the literature, to characterize the increasing and expanding use of informatics in public health and to identify public health domains that have demonstrated greater use of informatics for their research and practice.
Methods: We conducted a literature search of MEDLINE from 2004 to 2020 using PubMed and search terms were derived by two main concepts: public health and informatics, along with core domains of public health. To retrieve relevant articles, a variety of search features was used including the hierarchy of MeSH (Medical Subject Headings) terms and use of MeSH main and subheadings. From each article retrieved, MeSH terms were extracted and collated for frequency and co-occurrence analyses to delineate underlying public health domains that showed the status and trends in the use of informatics. To help guide the analysis and interpretation, frequent terms located in a hierarchical relationship in the MeSH tree structure were manually collapsed to a broader, higher-level term in the hierarchy.
Results: The co-occurrence analysis was performed to identify any distinct patterns of public health domains and informatics occurring together, thereby suggesting a range of multidisciplinary efforts benefited from the field of public health informatics. We further analyzed frequent patterns based on the data partitioned by four-year interval, indicating an upward trend in using the informatics with varying frequencies across the domains over time.
Conclusion: Our approach was suitable for describing the use of informatics, elicited by indexing terms, in diverse domains of public health and would be useful in leveraging the full potential of public health informatics.
Keywords
public health informatics, association analysis, MeSH terms, co-occurrence analysis
Using association analysis on the literature to identify the use of informatics in public health
Background: With growing applications of informatics in various domains of public health, there is a need to examine the evolving scope of public health informatics in its development and advancement toward achieving the goals of public health. The primary objectives of this study were, through data mining on the literature, to characterize the increasing and expanding use of informatics in public health and to identify public health domains that have demonstrated greater use of informatics for their research and practice.
Methods: We conducted a literature search of MEDLINE from 2004 to 2020 using PubMed and search terms were derived by two main concepts: public health and informatics, along with core domains of public health. To retrieve relevant articles, a variety of search features was used including the hierarchy of MeSH (Medical Subject Headings) terms and use of MeSH main and subheadings. From each article retrieved, MeSH terms were extracted and collated for frequency and co-occurrence analyses to delineate underlying public health domains that showed the status and trends in the use of informatics. To help guide the analysis and interpretation, frequent terms located in a hierarchical relationship in the MeSH tree structure were manually collapsed to a broader, higher-level term in the hierarchy.
Results: The co-occurrence analysis was performed to identify any distinct patterns of public health domains and informatics occurring together, thereby suggesting a range of multidisciplinary efforts benefited from the field of public health informatics. We further analyzed frequent patterns based on the data partitioned by four-year interval, indicating an upward trend in using the informatics with varying frequencies across the domains over time.
Conclusion: Our approach was suitable for describing the use of informatics, elicited by indexing terms, in diverse domains of public health and would be useful in leveraging the full potential of public health informatics.
