Received 8 February 2023, accepted 14 March 2023, date of publication 21 March 2023, date of current version 27 March 2023. Digital Object Identifier 10.1109/ACCESS.2023.3259972 VALS: Supporting Visual Data Analysis in Longitudinal Clinical Studies DUVÁN A. GÓMEZ 1,2 , (Member, IEEE), NATHALIE CHARPAK 3 , ADRIANA MONTEALEGRE 3,4 , AND JOSÉ TIBERIO HERNÁNDEZ 1,3 1 Department of Systems and Computing Engineering, Universidad de los Andes, Bogotá, Capital District 111711, Colombia 2 School of Engineering and Basic Sciences, Universidad EIA, Envigado, Antioquia 055428, Colombia 3 Fundación Canguro, Bogotá, Capital District 111321, Colombia 4 Medicine Faculty, Javeriana University, Bogotá, Capital District 110231, Colombia Corresponding author: Duván A. Gómez (da.gomez16@uniandes.edu.co) This work involved human subjects or animals in its research. Approval of all ethical and experimental procedures and protocols was granted by the Ethical Committee from the School of Medicine of Universidad Javeriana (Bogotá, Colombia) and the Ethical Committee of Fundación Santafé de Bogotá (Bogotá, Colombia). ABSTRACT Visual data analysis helps to understand different types of phenomena by allowing experts to explore for relationships, patterns, outliers, unexpected changes, and more. Experts need tools that help them find useful and actionable information in the data so that they can test their hypotheses and develop new ones. This need becomes more evident in longitudinal studies, where there are usually a large number of variables and the process being analyzed can be complex as well. We present VALS (Visual Analytics in Longitudinal Studies), a framework for visually exploring longitudinal clinical data. VALS includes a data model, a task categorization model, and an approach to guidance through feature engineering techniques and interactive visualizations, all of which help analysts perform their analysis tasks. VALS was designed in collaboration with healthcare experts with experience in longitudinal studies. We have also developed a tool prototype for a case study using real-world datasets. The evidence collected in the case study shows the usefulness of a VALS-based visual analytics tool. INDEX TERMS Cohort studies, exploratory data analysis, longitudinal clinical data, visual analytics. I. INTRODUCTION Longitudinal studies are a research method that involves measuring a phenomenon over a given time interval [1]. They are used to evaluate and track the development of a phenomenon in a sequential manner to learn more about the process of change [2]. They are commonly utilized in social sciences, psychology, sociology, statistics, and the health sciences. In the latter, longitudinal studies have application in domains such as epidemiology, clinical research, and therapeutic evaluation. Evidence from randomized controlled trials (RCT) is usually required to evaluate the efficacy of an intervention or a new drug. RCT are essentially longitudinal studies of cohorts of subjects [3]. The data recorded at the end of a longitudinal clinical study frequently exhibit high variability and volume. Therefore, the The associate editor coordinating the review of this manuscript and approving it for publication was Walter Didimo . applicable analysis techniques and associated software tools can become complex. Recently, several visual analysis solutions have been proposed to address data analysis problems in various contexts. However, work is still needed on how to support users in the analysis by integrating the different phases of the data analysis process, especially when dealing with data from longitudinal clinical studies. In this paper, we present VALS, which stands for Visual Analytics in Longitudinal Studies. VALS is a framework intended to support the exploration of data from longitudinal clinical studies. VALS is composed of a data model, a categorization of analysis tasks, and a first approach to guidance using feature engineering and information visualization tools. The data model seeks to represent the sequence of events of a longitudinal study and integrate expert knowledge in a way that can be used to provide guidance to analysts. The task 28820 This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4.0/ VOLUME 11, 2023