Implementation partnerships in a community-based intergenerational oral health study.

BACKGROUND: University-community partnerships are a common strategy used in implementing community-based health promotion trials, yet few published studies report these interactions in detail. "Baby Smiles" was a five-year intervention study in Oregon, USA. The study involved 400 low-income women during and after pregnancy across four rural counties. In this report, we describe and assess four university-community health partnerships formed to support the intervention. METHODS: A community health partnership advisory group for the study was established in each of the four participating counties. Group membership ranged from 9 to 23 individuals. A survey was administered to the groups five times in a 2.5 year period. The survey asked members' opinions of the intervention's goals, scientific basis and relevance to their organisation. Questions also asked about members' knowledge of oral health, beliefs about access to dental care for low-income pregnant women and children in their county and how their organisation functioned. RESULTS: There was strong overall support by each partnership group despite differences in the groups' structure, foci and turnover in membership during intervention period. Responses to specific survey items indicating misinformation or negative opinions about oral health care were used to address weaknesses in study implementation throughout the conduct of the study. CONCLUSION: Systematic monitoring of community support for a multi-year oral health intervention is feasible and can identify potential barriers to address while the study is underway.

Efficiency issues among statistical methods for demonstrating efficacy of caries prevention.

Although repeated tooth-surface-specific information is commonly collected during a longitudinal caries clinical trial, traditional methods often make limited use of the repeated measures. Newer methods of analysis, such as methods based on time-to-event and methods for longitudinal or clustered data, have the potential to increase the efficiency of the statistical analysis. We compare a range of analytical methods from the traditional analysis based only on the number of caries onsets to newer methods that incorporate time at risk and surface-specific information, such as Poisson regression methods for clustered data, with respect to the efficiency of treatment comparisons. Under most circumstances, the greatest gain in efficiency associated with time-to-event methods will be due to the ability of subjects to contribute caries onsets to the analysis until they are lost from the study. Incorporating the number of surfaces at risk, the surface time at risk, and surface-specific characteristics will typically produce only a modest gain in efficiency.

Overbite and overjet are not related to self-report of temporomandibular disorder symptoms.

Overbite and overjet, especially high or low values, have been found in some studies to be associated with temporomandibular disorders (TMD). This study evaluates the relationship between overbite/overjet and three TMD self-report measures (pain, joint noises, limited mouth-opening). Subjects were from two population-based cross-sectional studies (3033 subjects). After adjustment for age and gender, high or low values of overbite were not associated with an increased risk of self-reported TMD pain as compared with a reference category of a normal overbite of 2 to 3 mm (-8 to -1 mm, odds ratio = 0.36, 95% confidence interval = 0.05-2.76; 6 to 15 mm, odds ratio = 1.08, 95% confidence interval = 0.68-1.72). Similar non-significant results were found for overjet and TMD pain, and for the association of overjet/overbite and joint noises or limited mouth-opening. This study provides the strongest evidence to date that there is no association between overbite or overjet and self-reported TMD.

A covariance estimator for GEE with improved small-sample properties.

In this paper, we propose an alternative covariance estimator to the robust covariance estimator of generalized estimating equations (GEE). Hypothesis tests using the robust covariance estimator can have inflated size when the number of independent clusters is small. Resampling methods, such as the jackknife and bootstrap, have been suggested for covariance estimation when the number of clusters is small. A drawback of the resampling methods when the response is binary is that the methods can break down when the number of subjects is small due to zero or near-zero cell counts caused by resampling. We propose a bias-corrected covariance estimator that avoids this problem. In a small simulation study, we compare the bias-corrected covariance estimator to the robust and jackknife covariance estimators for binary responses for situations involving 10-40 subjects with equal and unequal cluster sizes of 16-64 observations. The bias-corrected covariance estimator gave tests with sizes close to the nominal level even when the number of subjects was 10 and cluster sizes were unequal, whereas the robust and jackknife covariance estimators gave tests with sizes that could be 2-3 times the nominal level. The methods are illustrated using data from a randomized clinical trial on treatment for bone loss in subjects with periodontal disease.

Between-subject and within-subject statistical information in dental research.

The evaluation of risk factors in dental research frequently uses observations at multiple sites in the same patient. For this reason, statistical methods that accommodate correlated data are generally used to assess the significance of the risk factors (e.g., generalized estimating equations, generalized linear mixed models). In applications of these methods, it is typically assumed (implicitly, if not explicitly) that between-subject and within-subject comparisons will produce the same estimated effect of the risk factor. When between- and within-subject comparisons conflict, the statistical methods can give biased estimates or results that are difficult to interpret. For illustration, we present two examples from periodontal disease studies in which different statistical methods give different estimates and significance levels for a risk factor. Statistical analyses in dental research should assess whether different sources of information give similar conclusions about risk factors or treatments.

Antigenic variation and cross-reactivity in Bacteroides forsythus clinical isolates detected by western blot.

Bacteroides forsythus is one of the etiologic agents of destructive periodontal diseases. Determining which antigenic components of the bacterium are recognized in the immune response of periodontitis patients is an important step in assessing strategies for vaccine development. The aim of this study was to identify the major strain-variable and cross-reactive antigens of B. forsythus clinical isolates recognized by serum IgG from patients with early-onset rapidly progressive periodontitis. Ten patient sera with measurable IgG against antigenic components of the species were identified by Western blot. Positive sera were tested by checkerboard ELISA to identify those most responsive to strain-variable antigens in nine clinical isolates and ATCC strain 43037. Correlation analysis of the ELISA data suggested that different subsets of isolates were preferentially recognized by different sera. Western blots revealed that certain sera also recognized major shared components across all the isolates, but preferential recognition of different isolate subsets by different patients was clearly confirmed. To determine if the variable antigens recognized were nonprotein, proteinase K-digested isolates were compared to undigested controls by Western blot. The main strain-variable antigens were proteinase resistant, while proteins at 200 and 210 kDa were identified as the major shared components. Two-dimensional SDS-PAGE revealed that these proteins are the quantitatively dominant heat-modifiable components of the cell envelope. Even though variable antigens are prominent in the immune response of patients, a cross-protective vaccine based on the shared envelope proteins of B. forsythus seems feasible in light of these observations.

Antigenic variation in Bacteroides forsythus detected by a checkerboard enzyme-linked immunosorbent assay.

Evidence indicating that multiple serotypes of Bacteroides forsythus participate in rapidly progressing periodontal infections has not been reported previously. Our aim was to develop an assay for detecting subsets of B. forsythus clinical isolates which differ in serogroup membership and subsets of patients with immunoglobulin G (IgG) responses which differ in serogroup recognition. A checkerboard enzyme-linked immunosorbent assay (ELISA) was used to assess variation in the IgG binding profiles of 22 clinical isolates in sera from 28 patients with early-onset rapidly progressive periodontitis. To accommodate the maximum number of isolates and sera in a given assay run, a multiplate assay grid with standard 96-well microtest plates was established. Single dilutions of individual sera were placed in rows crossing columns of isolate-coated wells, and antigen-specific IgG immobilized in the wells was measured as ELISA absorbance. Pooled sera and isolates were assayed in parallel to serve as negative controls for variation in IgG binding profiles. Correlation and hierarchical cluster analysis of the absorbance data matrix showed that the isolates could be sorted into at least four clusters based on variations in their IgG binding profiles across different sera. Furthermore, at least two patient clusters were defined by variations in their serum IgG antigen recognition profiles across different isolates. We conclude that multiple serogroups of B. forsythus exist and that different serogroups are dominant in the antibody response of different patients. The method applied here could be used to serologically classify clinical isolates of other species which evoke a serum antibody response in patients.

Efficiency of regression estimates for clustered data.

Statistical methods for clustered data, such as generalized estimating equations (GEE) and generalized least squares (GLS), require selecting a correlation or convariance structure to specify the dependence between observations within a cluster. Valid regression estimates can be obtained that do not depend on correct specification of the true correlation, but inappropriate specifications can result in a loss of efficiency. We derive general expressions for the asymptotic relative efficiency of GEE and GLS estimators under nested correlation structures. Efficiency is shown to depend on the covariate distribution, the cluster sizes, the response variable correlation, and the regression parameters. The results demonstrate that efficiency is quite sensitive to the between- and within-cluster variation of the covariates, and provide useful characterizations of models for which upper and lower efficiency bounds are attained. Efficiency losses for simple working correlation matrices, such as independence, can be large even for small to moderate correlations and cluster sizes.

Statistical issues in periodontal research.

In the past 10 to 12 years, there have been several statistical issues identified in periodontal research which require and have generated non-standard or new statistical approaches. The purpose of this paper is to give an overview of these issues and approaches. Three general categories of issues are described: (i) statistical methods for detecting when disease progression occurs, and biological theories and corresponding statistical models which attempt to describe how the disease progresses; (ii) design issues in studies of therapeutic efficacy; and (iii) analytic issues arising from periodontal data analysis.

Exploration of prediction models for caries risk assessment of the geriatric population.

The purpose of this study was to demonstrate a method for establishing a model designed to predict the caries risk of elderly individuals. Twenty-three patients over the age of 65 and living in a semi-independent retirement center were examined; several variables were collected and analyzed for their ability to predict the development of new carious lesions. The analysis was performed using logistic regression where the proportion of new decay was used as the dependent variable. The model for prediction of combined coronal and root caries included the variables flow rate, gender, and root caries index. The proposed method has the advantages of easily collected data, individualized criteria, and the ability to order patients as to the relative risk of developing decay.