Abstract BackgroundFrailty is a syndrome that is defined as an accumulation of deficits in physical, psychological, and social domains. On a global scale, there is an urgent need to create frailty-ready healthcare systems due to the healthcare burden that frailty confers on systems and the increased risk of falls, healthcare utilization, disability, and premature mortality. Several studies have been conducted to develop prediction models for predicting frailty. Most studies used logistic regression as a technique to develop a prediction model. One area that has experienced significant growth is the application of Bayesian techniques, partly due to an increasing number of practitioners valuing the Bayesian paradigm as matching that of scientific discovery. ObjectiveWe compared ten different Bayesian networks as proposed by ten experts in the field of frail elderly people to predict frailty with a choice from ten dichotomized determinants for frailty. MethodsWe used the opinion of ten experts who could indicate, using an empty Bayesian network graph, the important predictors for frailty and the interactions between the different predictors. The candidate predictors were age, sex, marital status, ethnicity, education, income, lifestyle, multimorbidity, life events, and home living environment. The ten Bayesian network models were evaluated in terms of their ability to predict frailty. For the evaluation, we used the data of 479 participants that filled in the Tilburg Frailty indicator (TFI) questionnaire for assessing frailty among community-dwelling older people. The data set contained the aforementioned variables and the outcome ”frail”. The model fit of each model was measured using the Akaike information criterion (AIC) and the predictive performance of the models was measured using the area under the curve (AUC) of the receiver operator characteristic (ROC). The AUCs of the models were validated using bootstrapping with 100 repetitions. The relative importance of the predictors in the models was calculated using the permutation feature importance algorithm (PFI). ResultsThe ten Bayesian networks of the ten experts differed considerably regarding the predictors and the connections between the predictors and the outcome. However, all ten networks had corrected AUCs 0.700. Evaluating the importance of the predictors in each model, ”diseases or chronic disorders” was the most important predictor in all models (10 times). The predictors ”lifestyle” and ”monthly income” were also often present in the models (both 6 times). One or more diseases or chronic disorders, an unhealthy lifestyle, and a monthly income below 1,800 euro increased the likelihood of frailty. ConclusionsAlthough the ten experts all made different graphs, the predictive performance was always satisfying (AUCs 0.700). While it is true that the predictor importance varied all the time, the top three of the predictor importance consisted of “diseases or chronic disorders”, “lifestyle” and “monthly income”. All in all, asking for the opinion of experts in the field of frail elderly to predict frailty with Bayesian networks may be more rewarding than a data-driven forecast with Bayesian networks because they have expert knowledge regarding interactions between the different predictors.
LINK
Journal of Physics: Conference Series Paper • The following article is Open access Exploring the relationship between light and subjective alertness using personal lighting conditions J. van Duijnhoven1, M.P.J. Aarts1, E.R. van den Heuvel2 and H.S.M. Kort3,4 Published under licence by IOP Publishing Ltd Journal of Physics: Conference Series, Volume 2042, CISBAT 2021 Carbon-neutral cities - energy efficiency and renewables in the digital era 8-10 September 2021, EPFL Lausanne, Switzerland Citation J. van Duijnhoven et al 2021 J. Phys.: Conf. Ser. 2042 012119 Download Article PDF References Download PDF 29 Total downloads Turn on MathJax Share this article Share this content via email Share on Facebook (opens new window) Share on Twitter (opens new window) Share on Mendeley (opens new window) Hide article information Author e-mails j.v.duijnhoven1@tue.nl Author affiliations 1 Building Lighting Group, Department of the Built Environment, Eindhoven University of Technology, Eindhoven, The Netherlands 2 Stochastics, Department of Mathematics and Computer Science, Eindhoven University of Technology, Eindhoven, The Netherlands 3 Research Centre Healthy and Sustainable Living, University of Applied Sciences Utrecht, Utrecht, The Netherlands 4 Building Healthy Environments for Future Users Group, Department of the Built Environment, Eindhoven University of Technology, Eindhoven, The Netherlands DOI https://doi.org/10.1088/1742-6596/2042/1/012119 Buy this article in print Journal RSS Sign up for new issue notifications Create citation alert Abstract The discovery of the ipRGCs was thought to fully explain the mechanism behind the relationship between light and effects beyond vision such as alertness. However, this relationship turned out to be more complicated. The current paper describes, by using personal lighting conditions in a field study, further exploration of the relationship between light and subjective alertness during daytime. Findings show that this relationship is highly dependent on the individual. Although nearly all dose-response curves between personal lighting conditions and subjective alertness determined in this study turned out to be not significant, the results may be of high importance in the exploration of the exact relationship.
MULTIFILE
Background: Early identification of older cardiac patients at high risk of readmission or mortality facilitates targeted deployment of preventive interventions. In the Netherlands, the frailty tool of the Dutch Safety Management System (DSMS-tool) consists of (the risk of) delirium, falling, functional impairment, and malnutrition and is currently used in all older hospitalised patients. However, its predictive performance in older cardiac patients is unknown. Aim: To estimate the performance of the DSMS-tool alone and combined with other predictors in predicting hospital readmission or mortality within 6 months in acutely hospitalised older cardiac patients. Methods: An individual patient data meta-analysis was performed on 529 acutely hospitalised cardiac patients ≥70 years from four prospective cohorts. Missing values for predictor and outcome variables were multiply imputed. We explored discrimination and calibration of: (1) the DSMS-tool alone; (2) the four components of the DSMS-tool and adding easily obtainable clinical predictors; (3) the four components of the DSMS-tool and more difficult to obtain predictors. Predictors in model 2 and 3 were selected using backward selection using a threshold of p = 0.157. We used shrunk c-statistics, calibration plots, regression slopes and Hosmer-Lemeshow p-values (PHL) to describe predictive performance in terms of discrimination and calibration. Results: The population mean age was 82 years, 52% were males and 51% were admitted for heart failure. DSMS-tool was positive in 45% for delirium, 41% for falling, 37% for functional impairments and 29% for malnutrition. The incidence of hospital readmission or mortality gradually increased from 37 to 60% with increasing DSMS scores. Overall, the DSMS-tool discriminated limited (c-statistic 0.61, 95% 0.56-0.66). The final model included the DSMS-tool, diagnosis at admission and Charlson Comorbidity Index and had a c-statistic of 0.69 (95% 0.63-0.73; PHL was 0.658). Discussion: The DSMS-tool alone has limited capacity to accurately estimate the risk of readmission or mortality in hospitalised older cardiac patients. Adding disease-specific risk factor information to the DSMS-tool resulted in a moderately performing model. To optimise the early identification of older hospitalised cardiac patients at high risk, the combination of geriatric and disease-specific predictors should be further explored.
