TY - JOUR
T1 - Bayesian random-effect model for predicting outcome fraught with heterogeneity
T2 - An illustration with episodes of 44 patients with intractable epilepsy
AU - Yen, A. M.F.
AU - Liou, H. H.
AU - Lin, H. L.
AU - Chen, Tony Hsiu Hsi
PY - 2006
Y1 - 2006
N2 - Objective: The study aimed to develop a predictive model to deal with data fraught with heterogeneity that cannot be explained by sampling variation or measured covariates. Methods: The random-effect Poisson regression model was first proposed to deal with over-dispersion for data fraught with heterogeneity offer making allowance for measured covariates. Bayesian acyclic graphic model in conjunction with Markov Chain Monte Carlo (MCMC) technique was then applied to estimate the parameters of both relevant covariates and random effect. Predictive distribution was then generated to compare the predicted with the observed for the Bayesian model with and without random effect. Data from repeated measurement of episodes among 44 patients with intractable epilepsy were used as an illustration. Results: The application of Poisson regression without taking heterogeneity into account to epilepsy data yielded a large value of heterogeneity (heterogeneity factor = 17.90, deviance = 1485, degree of freedom (df) = 83). After taking the random effect into account, the value of heterogeneity factor was greatly reduced (heterogeneity factor = 0.52, deviance = 42.5, df = 81). The Pearson χ2 for the comparison between the expected seizure frequencies and the observed ones at two and three months of the model with and without random effect were 34.27 (p = 1.00) and 1799.90 (p <0.0001), respectively. Conclusion: The Bayesian acyclic model using the MCMC method was demonstrated to have great potential for disease prediction while data show over-dispersion attributed either to correlated property or to subject-to-subject variability.
AB - Objective: The study aimed to develop a predictive model to deal with data fraught with heterogeneity that cannot be explained by sampling variation or measured covariates. Methods: The random-effect Poisson regression model was first proposed to deal with over-dispersion for data fraught with heterogeneity offer making allowance for measured covariates. Bayesian acyclic graphic model in conjunction with Markov Chain Monte Carlo (MCMC) technique was then applied to estimate the parameters of both relevant covariates and random effect. Predictive distribution was then generated to compare the predicted with the observed for the Bayesian model with and without random effect. Data from repeated measurement of episodes among 44 patients with intractable epilepsy were used as an illustration. Results: The application of Poisson regression without taking heterogeneity into account to epilepsy data yielded a large value of heterogeneity (heterogeneity factor = 17.90, deviance = 1485, degree of freedom (df) = 83). After taking the random effect into account, the value of heterogeneity factor was greatly reduced (heterogeneity factor = 0.52, deviance = 42.5, df = 81). The Pearson χ2 for the comparison between the expected seizure frequencies and the observed ones at two and three months of the model with and without random effect were 34.27 (p = 1.00) and 1799.90 (p <0.0001), respectively. Conclusion: The Bayesian acyclic model using the MCMC method was demonstrated to have great potential for disease prediction while data show over-dispersion attributed either to correlated property or to subject-to-subject variability.
KW - Bayesian acyclic graphic model
KW - Heterogeneity
KW - Markov Chain Monte Carlo (MCMC)
KW - Predictive model
KW - Random effect
UR - http://www.scopus.com/inward/record.url?scp=33845917898&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=33845917898&partnerID=8YFLogxK
U2 - 10.1055/s-0038-1634127
DO - 10.1055/s-0038-1634127
M3 - Article
C2 - 17149504
AN - SCOPUS:33845917898
VL - 45
SP - 631
EP - 637
JO - Methods of Information in Medicine
JF - Methods of Information in Medicine
SN - 0026-1270
IS - 6
ER -