Determinants of physical activity behaviour change in (online) interventions, and gender-specific differences: a Bayesian network model.

BACKGROUND: Physical activity (PA) is known to be beneficial for health, but adherence to international PA guidelines is low across different subpopulations. Interventions have been designed to stimulate PA of different target groups by influencing relevant psycho-social determinants, essentially based on a combination of the Integrated Model for Change, the Theory of Planned Behaviour, its successor the Reasoned Action Approach and the self-determination theory. The current study investigates the pathways through which interventions influence PA. Further, gender differences in pathways of change are studied. METHODS: An integrated dataset of five different randomised controlled trial intervention studies is analysed by estimating a Bayesian network. The data include measurements, at baseline and at 3, 6 (short-term), and 12 (long-term) months after the baseline, of important socio-cognitive determinants of PA, demographic factors, and PA outcomes. A fragment is extracted from the Bayesian network consisting of paths between the intervention variable, determinants, and short- and long-term PA outcomes. For each relationship between variables, a stability indicator and its mutual information are computed. Such a model is estimated for the full dataset, and in addition such a model is estimated based only on male and female participants' data to investigate gender differences. RESULTS: The general model (for the full dataset) shows complex paths, indicating that the intervention affects short-term PA via the direct determinants of intention and habit and that self-efficacy, attitude, intrinsic motivation, social influence concepts, planning and commitment have an indirect influence. The model also shows how effects are maintained in the long-term and that previous PA behaviour, intention and attitude pros are direct determinants of long-term PA. The gender-specific models show similarities as well as important differences between the structures of paths for the male- and female subpopulations. For both subpopulations, intention and habit play an important role for short-term effects and maintenance of effects in the long-term. Differences are found in the role of self-efficacy in paths of behaviour change and in the fact that attitude is relevant for males, whereas planning plays a crucial role for females. The average of these differences in subpopulation mechanisms appears to be presented in the general model. CONCLUSIONS: While previous research provided limited insight into how interventions influence PA through relevant determinants, the Bayesian network analyses show the relevance of determinants mentioned by the theoretical framework. The model clarifies the role that different determinants play, especially in interaction with each other. The Bayesian network provides new knowledge about the complex working mechanism of interventions to change PA by giving an insightful overview of influencing paths. Furthermore, by presenting subpopulation-specific networks, the difference between the influence structure of males and females is illustrated. These new insights can be used to improve interventions in order to enhance their effects. To accomplish this, we have developed a new methodology based on a Bayesian network analysis which may be applicable in various other studies.

A probabilistic framework for predicting disease dynamics: A case study of psychotic depression.

Unsupervised learning is often used to obtain insight into the underlying structure of medical data, but it is not always clear how to use such structure in an effective way. In this paper, we propose a probabilistic framework for predicting disease dynamics guided by latent states. The framework is based on hidden Markov models and aims to facilitate the selection of hypotheses that might yield insight into the dynamics. We demonstrate this by using clinical trial data for psychotic depression treatment as a case study. The discovered latent structure and proposed outcome are then validated using standard depression criteria, and are shown to provide new insight into the heterogeneity of psychotic depression in terms of predictive symptoms for different interventions.

Understanding disease processes by partitioned dynamic Bayesian networks.

For many clinical problems in patients the underlying pathophysiological process changes in the course of time as a result of medical interventions. In model building for such problems, the typical scarcity of data in a clinical setting has been often compensated by utilizing time homogeneous models, such as dynamic Bayesian networks. As a consequence, the specificities of the underlying process are lost in the obtained models. In the current work, we propose the new concept of partitioned dynamic Bayesian networks to capture distribution regime changes, i.e. time non-homogeneity, benefiting from an intuitive and compact representation with the solid theoretical foundation of Bayesian network models. In order to balance specificity and simplicity in real-world scenarios, we propose a heuristic algorithm to search and learn these non-homogeneous models taking into account a preference for less complex models. An extensive set of experiments were ran, in which simulating experiments show that the heuristic algorithm was capable of constructing well-suited solutions, in terms of goodness of fit and statistical distance to the original distributions, in consonance with the underlying processes that generated data, whether it was homogeneous or non-homogeneous. Finally, a study case on psychotic depression was conducted using non-homogeneous models learned by the heuristic, leading to insightful answers for clinically relevant questions concerning the dynamics of this mental disorder.

Machine learning for antidepressant treatment selection in depression.

Finding the right antidepressant for the individual patient with major depressive disorder can be a difficult endeavor and is mostly based on trial-and-error. Machine learning (ML) is a promising tool to personalize antidepressant prescription. In this review, we summarize the current evidence of ML in the selection of antidepressants and conclude that its value for clinical practice is still limited. Apart from the current focus on effectiveness, several other factors should be taken into account to make ML-based prediction models useful for clinical application.

The Adoption of a COVID-19 Contact-Tracing App: Cluster Analysis.

BACKGROUND: During the COVID-19 pandemic, there was limited adoption of contact-tracing apps (CTAs). Adoption was particularly low among vulnerable people (eg, people with a low socioeconomic position or of older age), while this part of the population tends to have lesser access to information and communication technology and is more vulnerable to the COVID-19 virus. OBJECTIVE: This study aims to understand the cause of this lagged adoption of CTAs in order to facilitate adoption and find indications to make public health apps more accessible and reduce health disparities. METHODS: Because several psychosocial variables were found to be predictive of CTA adoption, data from the Dutch CTA CoronaMelder (CM) were analyzed using cluster analysis. We examined whether subgroups could be formed based on 6 psychosocial perceptions (ie, trust in the government, beliefs about personal data, social norms, perceived personal and societal benefits, risk perceptions, and self-efficacy) of (non)users concerning CM in order to examine how these clusters differ from each other and what factors are predictive of the intention to use a CTA and the adoption of a CTA. The intention to use and the adoption of CM were examined based on longitudinal data consisting of 2 time frames in October/November 2020 (N=1900) and December 2020 (N=1594). The clusters were described by demographics, intention, and adoption accordingly. Moreover, we examined whether the clusters and the variables that were found to influence the adoption of CTAs, such as health literacy, were predictive of the intention to use and the adoption of the CM app. RESULTS: The final 5-cluster solution based on the data of wave 1 contained significantly different clusters. In wave 1, respondents in the clusters with positive perceptions (ie, beneficial psychosocial variables for adoption of a CTA) about the CM app were older (P<.001), had a higher education level (P<.001), and had higher intention (P<.001) and adoption (P<.001) rates than those in the clusters with negative perceptions. In wave 2, the intention to use and adoption were predicted by the clusters. The intention to use CM in wave 2 was also predicted using the adoption measured in wave 1 (P<.001, ß=-2.904). Adoption in wave 2 was predicted by age (P=.022, exp(B)=1.171), the intention to use in wave 1 (P<.001, exp(B)=1.770), and adoption in wave 1 (P<.001, exp(B)=0.043). CONCLUSIONS: The 5 clusters, as well as age and previous behavior, were predictive of the intention to use and the adoption of the CM app. Through the distinguishable clusters, insight was gained into the profiles of CM (non)intenders and (non)adopters. TRIAL REGISTRATION: OSF Registries osf.io/cq742; https://osf.io/cq742.

Investigating the understandability of XAI methods for enhanced user experience: When Bayesian network users became detectives.

In the medical domain, the uptake of an AI tool crucially depends on whether clinicians are confident that they understand the tool. Bayesian networks are popular AI models in the medical domain, yet, explaining predictions from Bayesian networks to physicians and patients is non-trivial. Various explanation methods for Bayesian network inference have appeared in literature, focusing on different aspects of the underlying reasoning. While there has been a lot of technical research, there is little known about the actual user experience of such methods. In this paper, we present results of a study in which four different explanation approaches were evaluated through a survey by questioning a group of human participants on their perceived understanding in order to gain insights about their user experience.

A comparison between discrete and continuous time Bayesian networks in learning from clinical time series data with irregularity.

BACKGROUND: Recently, mobile devices, such as smartphones, have been introduced into healthcare research to substitute paper diaries as data-collection tools in the home environment. Such devices support collecting patient data at different time points over a long period, resulting in clinical time-series data with high temporal complexity, such as time irregularities. Analysis of such time series poses new challenges for machine-learning techniques. The clinical context for the research discussed in this paper is home monitoring in chronic obstructive pulmonary disease (COPD). OBJECTIVE: The goal of the present research is to find out which properties of temporal Bayesian network models allow to cope best with irregularly spaced multivariate clinical time-series data. METHODS: Two mainstream temporal Bayesian network models of multivariate clinical time series are studied: dynamic Bayesian networks, where the system is described as a snapshot at discrete time points, and continuous time Bayesian networks, where transitions between states are modeled in continuous time. Their capability of learning from clinical time series that vary in nature are extensively studied. In order to compare the two temporal Bayesian network types for regularly and irregularly spaced time-series data, three typical ways of observing time-series data were investigated: (1) regularly spaced in time with a fixed rate; (2) irregularly spaced and missing completely at random at discrete time points; (3) irregularly spaced and missing at random at discrete time points. In addition, similar experiments were carried out using real-world COPD patient data where observations are unevenly spaced. RESULTS: For regularly spaced time series, the dynamic Bayesian network models outperform the continuous time Bayesian networks. Similarly, if the data is missing completely at random, discrete-time models outperform continuous time models in most situations. For more realistic settings where data is not missing completely at random, the situation is more complicated. In simulation experiments, both models perform similarly if there is strong prior knowledge available about the missing data distribution. Otherwise, continuous time Bayesian networks perform better. In experiments with unevenly spaced real-world data, we surprisingly found that a dynamic Bayesian network where time is ignored performs similar to a continuous time Bayesian network. CONCLUSION: The results confirm conventional wisdom that discrete-time Bayesian networks are appropriate when learning from regularly spaced clinical time series. Similarly, we found that time series where the missingness occurs completely at random, dynamic Bayesian networks are an appropriate choice. However, for complex clinical time-series data that motivated this research, the continuous-time models are at least competitive and sometimes better than their discrete-time counterparts. Furthermore, continuous-time models provide additional benefits of being able to provide more fine-grained predictions than discrete-time models, which will be of practical relevance in clinical applications.

Adaptation of clinical practice guidelines.

A rigorous development process of clinical practice guidelines through a systematic appraisal of available evidence is costly and time consuming. One way to reduce the costs and time, and avoid unnecessary duplication of effort of guideline development is by relying on a local adaptation approach of guidelines developed at the (inter)national level by expert groups. In this chapter we survey the work on guideline adaptation, which includes methodologies, case studies, assessment of effectiveness, and related work on guideline adaptation in the Artificial Intelligence community.

Formal methods for verification of clinical practice guidelines.

Formal methods play an important role in the development of software and hardware systems. In recent years, there has been a growing interest to apply these methods in the area of medical guidelines and protocols. This paper summarises these efforts, compares the approaches and discusses the role of formal methods in this area.

A constraint-based approach to medical guidelines and protocols.

Medical guidelines and protocols are documents aimed at improving the quality of medical care by offering support in medical decision making in the form of management recommendations based on scientific evidence. Whereas medical guidelines are intended for nation-wide use, and thus omit medical management details that may differ among hospitals, medical protocols are aimed at local use, e.g., within hospitals, and, therefore, include more detailed information. Although a medical guideline and an associated protocol concerning the management of a particular disorder are related to each other, one question is to what extent they are different. Formal methods are applied to shed light on this issue. A Dutch medical guideline regarding the treatment of breast cancer, and a Dutch protocol based on it, are taken as an example.

Exploiting Experts' Knowledge for Structure Learning of Bayesian Networks.

Learning Bayesian network structures from data is known to be hard, mainly because the number of candidate graphs is super-exponential in the number of variables. Furthermore, using observational data alone, the true causal graph is not discernible from other graphs that model the same set of conditional independencies. In this paper, it is investigated whether Bayesian network structure learning can be improved by exploiting the opinions of multiple domain experts regarding cause-effect relationships. In practice, experts have different individual probabilities of correctly labeling the inclusion or exclusion of edges in the structure. The accuracy of each expert is modeled by three parameters. Two new scoring functions are introduced that score each candidate graph based on the data and experts' opinions, taking into account their accuracy parameters. In the first scoring function, the experts' accuracies are estimated using an expectation-maximization-based algorithm and the estimated accuracies are explicitly used in the scoring process. The second function marginalizes out the accuracy parameters to obtain more robust scores when it is not possible to obtain a good estimate of experts' accuracies. The experimental results on simulated and real world datasets show that exploiting experts' knowledge can improve the structure learning if we take the experts' accuracies into account.

Multilevel Bayesian networks for the analysis of hierarchical health care data.

OBJECTIVE: Large health care datasets normally have a hierarchical structure, in terms of levels, as the data have been obtained from different practices, hospitals, or regions. Multilevel regression is the technique commonly used to deal with such multilevel data. However, for the statistical analysis of interactions between entities from a domain, multilevel regression yields little to no insight. While Bayesian networks have proved to be useful for analysis of interactions, they do not have the capability to deal with hierarchical data. In this paper, we describe a new formalism, which we call multilevel Bayesian networks; its effectiveness for the analysis of hierarchically structured health care data is studied from the perspective of multimorbidity. METHODS: Multilevel Bayesian networks are formally defined and applied to analyze clinical data from family practices in The Netherlands with the aim to predict interactions between heart failure and diabetes mellitus. We compare the results obtained with multilevel regression. RESULTS: The results obtained by multilevel Bayesian networks closely resembled those obtained by multilevel regression. For both diseases, the area under the curve of the prediction model improved, and the net reclassification improvements were significantly positive. In addition, the models offered considerable more insight, through its internal structure, into the interactions between the diseases. CONCLUSIONS: Multilevel Bayesian networks offer a suitable alternative to multilevel regression when analyzing hierarchical health care data. They provide more insight into the interactions between multiple diseases. Moreover, a multilevel Bayesian network model can be used for the prediction of the occurrence of multiple diseases, even when some of the predictors are unknown, which is typically the case in medicine.

Multilevel temporal Bayesian networks can model longitudinal change in multimorbidity.

OBJECTIVES: Although the course of single diseases can be studied using traditional epidemiologic techniques, these methods cannot capture the complex joint evolutionary course of multiple disorders. In this study, multilevel temporal Bayesian networks were adopted to study the course of multimorbidity in the expectation that this would yield new clinical insight. STUDY DESIGN AND SETTING: Clinical data of patients were extracted from 90 general practice registries in the Netherlands. One and half million patient-years were used for analysis. The simultaneous progression of six chronic cardiovascular conditions was investigated, correcting for both patient and practice-related variables. RESULTS: Cumulative incidence rates of one or more new morbidities rapidly increase with the number of morbidities present at baseline, ranging up to 47% and 76% for 3- and 5-year follow-ups, respectively. Hypertension and lipid disorders, as health risk factors, increase the cumulative incidence rates of both individual and multiple disorders. Moreover, in their presence, the observed cumulative incidence rates of combinations of cardiovascular disorders, that is, multimorbidity differs significantly from the expected rates. CONCLUSION: There are clear synergies between health risks and chronic diseases when multimorbidity within a patient progresses over time. The method used here supports a more comprehensive analysis of such synergies compared with what can be obtained by traditional statistics.

Probabilistic causal models of multimorbidity concepts.

Multimorbidity, i.e., the presence of multiple diseases within one person, is a significant health-care problem for western societies: diagnosis, prognosis and treatment in the presence of of multiple diseases can be complex due to the various interactions between diseases. A literature review reveals that there is a variety of definitions that describe different concepts with respect to multimorbidity, both for the cause of multimorbidity as well as the implications of multimorbidity. To be able to aid computerized decision support systems within patient care, e.g. electronic clinical guidelines that can be personalized given the patient's problems, these multimorbidity aspects need to be defined rigorously in a formal language. In this paper, we employ causal Bayesian networks to define and analyze a novel framework that can be used to model a spectrum of aspects related to multimorbidity. We conclude that this framework provides a solid basis for modeling interactions between multiple diseases.

Using model checking for critiquing based on clinical guidelines.

OBJECTIVE: Medical critiquing systems compare clinical actions performed by a physician with a predefined set of actions. In order to provide useful feedback, an important task is to find differences between the actual actions and a set of 'ideal' actions as described by a clinical guideline. In case differences exist, the critiquing system provides insight into the extent to which they are compatible. METHODS AND MATERIAL: We propose a computational method for such critiquing, where the ideal actions are given by a formal model of a clinical guideline, and where the actual actions are derived from real world patient data. We employ model checking to investigate whether a part of the actual treatment is consistent with the guideline. RESULTS: We show how critiquing can be cast in terms of temporal logic, and what can be achieved by using model checking. Furthermore, a method is introduced for off-line computing relevant information which can be exploited during critiquing. The method has been applied to a clinical guideline of breast cancer in conjunction with breast cancer patient data.