D T Kopycka-Kedzierawski1, R J Billings. 1. Eastman Department of Dentistry, University of Rochester, Rochester, NY 14620, USA. dorota_kopyckakedzierawski@urmc.rochester.edu
Abstract
OBJECTIVES: Markov modeling is a useful mathematical procedure for calculating probabilities of disease prognosis. Increasingly, Markov models are being applied in medical and health services research and also in social sciences research. The purpose of our study was to use the Markov process to determine time-dependent transition probabilities for caries-free children to convert to a caries-active state and to assess the impact of salivary mutans streptococci (MS) levels on caries status. METHODS: Our analysis was based on data obtained from a 6-year longitudinal study of risk factors associated with caries onset in children. RESULTS: Based on a two-state Markov model, the probability that a caries-free child would convert to a caries-active state during the study ranged between 0.0046 and 0.0471. The highest probability of converting from a caries-free state to a caries-active state was 0.0471 at age 8.5 years. CONCLUSIONS: In addition to standard statistical methods of analyzing longitudinal caries data, Markov models show promise for use in the analysis of caries risk.
OBJECTIVES: Markov modeling is a useful mathematical procedure for calculating probabilities of disease prognosis. Increasingly, Markov models are being applied in medical and health services research and also in social sciences research. The purpose of our study was to use the Markov process to determine time-dependent transition probabilities for caries-free children to convert to a caries-active state and to assess the impact of salivary mutans streptococci (MS) levels on caries status. METHODS: Our analysis was based on data obtained from a 6-year longitudinal study of risk factors associated with caries onset in children. RESULTS: Based on a two-state Markov model, the probability that a caries-free child would convert to a caries-active state during the study ranged between 0.0046 and 0.0471. The highest probability of converting from a caries-free state to a caries-active state was 0.0471 at age 8.5 years. CONCLUSIONS: In addition to standard statistical methods of analyzing longitudinal caries data, Markov models show promise for use in the analysis of caries risk.