A flexible AFT model for misclassified clustered interval-censored data.

Motivated by a longitudinal oral health study, we propose a flexible modeling approach for clustered time-to-event data, when the response of interest can only be determined to lie in an interval obtained from a sequence of examination times (interval-censored data) and on top of that, the determination of the occurrence of the event is subject to misclassification. The clustered time-to-event data are modeled using an accelerated failure time model with random effects and by assuming a penalized Gaussian mixture model for the random effects terms to avoid restrictive distributional assumptions concerning the event times. A general misclassification model is discussed in detail, considering the possibility that different examiners were involved in the assessment of the occurrence of the events for a given subject across time. A Bayesian implementation of the proposed model is described in a detailed manner. We additionally provide empirical evidence showing that the model can be used to estimate the underlying time-to-event distribution and the misclassification parameters without any external information about the latter parameters. We also provide results of a simulation study to evaluate the effect of neglecting the presence of misclassification in the analysis of clustered time-to-event data.

Non-alcoholic fatty liver disease and its association with obesity, insulin resistance and increased serum levels of C-reactive protein in Hispanics.

BACKGROUND: Non-alcoholic fatty liver disease (NAFLD) is a metabolic disorder of the liver, which may progress to fibrosis or cirrhosis. Recent studies have shown a significant impact of ethnicity on susceptibility to steatosis-related liver disease. AIMS: To estimate the prevalence of NAFLD among Chilean Hispanics as well as the clinical and biochemical variables associated with the disease. METHODS: Population-based study among Chilean Hispanics. The diagnosis of NAFLD was made on the basis of ultrasound evidence of fatty liver and absence of significant alcohol consumption and hepatitis C virus infection. RESULTS: A total of 832 Hispanic subjects were included. Ultrasound findings revealed diffuse fatty liver in 23% of the subjects. Variables associated with fatty liver in multivariate analysis were body mass index >26.9 [odds ratio (OR) 6.2; 95% confidence interval (CI) 3.3-11.5], abnormal aspartate aminotransferase levels (OR 14; 95% CI 8.2-23.7), presence of insulin resistance as measured by homoeostasis model assessment-insulin resistance (OR 3; 95% CI 1.8-4.8) and serum levels of high-sensitivity C-reactive protein (hs-CRP) greater than 0.86 mg/L (OR 2.9; 95% CI 1.6-5.2). Among subjects with NAFLD, levels of hs-CRP were similar regardless of the alanine aminotransferase (ALT) level. CONCLUSIONS: Chilean Hispanics exhibit a high prevalence of NAFLD. Obesity, insulin resistance, abnormal aminotransferase levels and elevated hs-CRP were independently associated with the presence of NAFLD. ALT elevation underestimates the presence of ultrasonographical fatty liver, whereas hs-CRP is a sensitive independent marker of NAFLD, which may be useful for detecting fatty liver in the general population.

Marginal Bayesian Semiparametric Modeling of Mismeasured Multivariate Interval-Censored Data.

Motivated by data gathered in an oral health study, we propose a Bayesian nonparametric approach for population-averaged modeling of correlated time-to-event data, when the responses can only be determined to lie in an interval obtained from a sequence of examination times and the determination of the occurrence of the event is subject to misclassification. The joint model for the true, unobserved time-to-event data is defined semiparametrically; proportional hazards, proportional odds, and accelerated failure time (proportional quantiles) are all fit and compared. The baseline distribution is modeled as a flexible tailfree prior. The joint model is completed by considering a parametric copula function. A general misclassification model is discussed in detail, considering the possibility that different examiners were involved in the assessment of the occurrence of the events for a given subject across time. We provide empirical evidence that the model can be used to estimate the underlying time-to-event distribution and the misclassification parameters without any external information about the latter parameters. We also illustrate the effect on the statistical inferences of neglecting the presence of misclassification.

Factors associated with prevalence and severity of caries experience in preschool children.

OBJECTIVES: The aim of the present study was to examine the prevalence and severity of caries experience in the primary dentition of preschool children and to assess the association of disease distribution with oral hygiene levels, reported oral health behaviours and socio-demographic factors. METHODS: Study samples comprised 1250 3-year-old and 1283 5-year-old pre-school children from four distinct geographical areas in Flanders. Information on oral hygiene and dietary habits, oral health behaviours and socio-demographic variables was collected using questionnaires completed by the parents. Clinical examinations were performed using standardized criteria. Caries experience was recorded at the level of cavitation (d(3) level). Simple as well as multivariable logistic regression analyses were performed in order to identify factors associated with prevalence and severity of caries experience. RESULTS: Visible plaque was present in 31% of 3-year-olds and 37% of 5-year-olds. In 3-year-olds, 7% presented with caries experience while this was the case in 31% of 5-year-olds. Multivariable logistic regression revealed significant associations, in 3-year-olds, of caries experience with presence of dental plaque (OR = 7.93; 95% CI: 2.56-24.55) and reported consumption of sugared drinks at night (OR = 7.96; 95% CI: 1.57-40.51). In 5-year-olds, significant associations were seen with age (OR = 7.79; 95% CI: 2.38-25.43), gender (OR = 0.37 with 95% CI: 0.19-0.71 for girls), presence of visible dental plaque (OR = 3.36; 95% CI: 1.64-6.89) and reported habit of having sugar-containing drinks in between meals (OR = 2.60 with 95% CI: 1.16-5.84 and OR = 3.18 with 95% CI: 1.39-7.28, respectively for 1x/day and > 1x/day versus not every day). In 5-year-olds with caries experience (30.8% of total sample), the severity of disease was further analysed (d(3)mft between 1 and 4 versus d(3)mft 5 or higher). Multivariable analyses showed a significant association with gender [girls more likely to have higher disease levels; OR = 4.67 (95% CI: 1.65-13.21)] and with presence of plaque (OR = 3.91 with 95% CI: 1.23-12.42). CONCLUSIONS: Presence of visible plaque accumulation and reported consumption of sugared drinks were associated with prevalence of caries experience in Flemish preschool children. Severity of disease was associated with gender and with presence of plaque. Results underline the importance of plaque control and diet management from very young age on.

Variables associated with longitudinally registered plaque accumulation in a cohort of Flemish schoolchildren.

OBJECTIVES: To analyze the change in pattern of plaque accumulation on buccal and occlusal surfaces of permanent teeth and associated variables in a cohort of 4,468 children examined on a yearly basis between the ages of 7 and 12 years. METHOD AND MATERIALS: Oral hygiene level on buccal surfaces was assessed using the Plaque Index of Silness and Loe; for occlusal surfaces, a simplified version of the index as described by Carvalho et al was used. Data on oral health habits were collected using questionnaires completed by the parents and by the school health care center. Multiple ordinal logistic regression models using first-order generalized estimating equations were fitted to estimate population average effects taking into account the correlated structure of the data. RESULTS: Girls brushed significantly more frequently than boys (as reported by the parents) and presented with significantly less dental plaque. In all survey years, starting to brush at a young age, no daily consumption of sugar-containing drinks, and brushing at least twice a day were significantly associated with lower plaque accumulation scores. Parental help did not seem to influence the accumulation of occlusal plaque, but it did influence the amount of buccal plaque on incisors and premolars at older ages. The presence of sealants was significantly associated with less dental plaque. CONCLUSION: Regarding future policies for preventive strategies in schoolchildren, help with brushing at older ages can be recommended. Application of sealants can be encouraged, but further research is needed to confirm whether the presence of sealants improves oral cleanliness.

Reported methodological quality of split-mouth studies.

BACKGROUND/AIM: Hujoel & Moulton previously questioned the reported quality of split-mouth studies. Since then, there has been little enquiry into the methodology of this study design. The aim was to conduct a systematic review of the reported methodology of clinical studies using a split-mouth design published in dental journals over a 1-year period (2004). MATERIAL AND METHODS: An extension of the CONSORT guidelines for cluster-randomized designs was used to evaluate quality. We evaluated the methods used and quality of reporting split-mouth studies. RESULTS: Thirty-four studies were eligible for this review. The results showed that many papers lack essential qualities of good reporting, e.g. five of 34 papers gave the rationale for choosing a split-mouth design, 19 of 34 (56%) used appropriate analytical statistical methods and only one of 34 presented an appropriate sample size calculation. Of the five studies that used survival analysis, none of them used a paired approach. CONCLUSIONS: Despite some progress in statistical analysis, if the reporting of studies represents the actual methodology of the trial, this review has identified important aspects of split-mouth study design and analysis that would benefit from development.

Gallbladder disease is associated with insulin resistance in a high risk Hispanic population.

BACKGROUND/AIMS: We tested whether cholesterol gallstone disease (GS) is associated to insulin resistance and serum C-reactive protein (CRP) in a high risk population. METHODS: This was a nested case-control study on 881 Chilean subjects that included clinical examination, abdominal ultrasound and blood chemistries. Insulin resistance was determined by the homeostasis model assessment index (IR-HOMA). RESULTS: Compared to controls (n = 582), age and sex adjusted median IR-(HOMA) was significantly elevated in subjects with GS (n = 299) (P < 0.001). Risks of GS in subjects with insulin resistance and different body mass index (BMI) were: BMI < 25, 2.2 (1.1-4.7); BMI 25-30, 1.5 (0.9-2.5) and BMI > 30, 1.7 (1.0-2.9). Risk of GS in subjects with metabolic syndrome was 1.7 (CI, 1.2-2.5) and with fatty liver, 1.5 (1.1-2.2). Risk of GS in subjects with CRP > 1 mg/dL was 1.0 (0.7-1.7). CONCLUSIONS: GS is associated to insulin resistance, fatty liver and to metabolic syndrome, but not to serum CRP in a high risk Hispanic population. Insulin resistance could have a major role in the pathogenesis of GS favoring the production of cholesterol supersaturated bile and altering gallbladder function.