Multi-level zero-inflated Poisson regression modelling of correlated count data with excess zeros

Count data with excess zeros relative to a Poisson distribution are common in many biomedical applications. A popular approach to the analysis of such data is to use a zero-inflated Poisson (ZIP) regression model. Often, because of the hierarchical Study design or the data collection procedure, zero-inflation and lack of independence may occur simultaneously, which tender the standard ZIP model inadequate. To account for the preponderance of zero counts and the inherent correlation of observations, a class of multi-level ZIP regression model with random effects is presented. Model fitting is facilitated using an expectation-maximization algorithm, whereas variance components are estimated via residual maximum likelihood estimating equations. A score test for zero-inflation is also presented. The multi-level ZIP model is then generalized to cope with a more complex correlation structure. Application to the analysis of correlated count data from a longitudinal infant feeding study illustrates the usefulness of the approach.

Assessing the impacts of grazing levels on bird density in woodland habitat: a Bayesian approach using expert opinion

Many studies on birds focus on the collection of data through an experimental design, suitable for investigation in a classical analysis of variance (ANOVA) framework. Although many findings are confirmed by one or more experts, expert information is rarely used in conjunction with the survey data to enhance the explanatory and predictive power of the model. We explore this neglected aspect of ecological modelling through a study on Australian woodland birds, focusing on the potential impact of different intensities of commercial cattle grazing on bird density in woodland habitat. We examine a number of Bayesian hierarchical random effects models, which cater for overdispersion and a high frequency of zeros in the data using WinBUGS and explore the variation between and within different grazing regimes and species. The impact and value of expert information is investigated through the inclusion of priors that reflect the experience of 20 experts in the field of bird responses to disturbance. Results indicate that expert information moderates the survey data, especially in situations where there are little or no data. When experts agreed, credible intervals for predictions were tightened considerably. When experts failed to agree, results were similar to those evaluated in the absence of expert information. Overall, we found that without expert opinion our knowledge was quite weak. The fact that the survey data is quite consistent, in general, with expert opinion shows that we do know something about birds and grazing and we could learn a lot faster if we used this approach more in ecology, where data are scarce. Copyright (c) 2005 John Wiley & Sons, Ltd.

Caregiver burden, time spent caring and health status in the first 12 months following stroke

Objective: To quantify time caring, burden and health status in carers of stroke patients after discharge from rehabilitation; to identify the potentially modifiable sociodemographic and clinical characteristics associated with these outcomes. Methods: Patients and carers prospectively interviewed 6 (n = 71) and 12 (n = 57) months after discharge. Relationships of carer and patient variables with burden, health status and time analysed by Gaussian and Poisson regression. Results: Carers showed considerable burden at 6 and 12 months. Carers spent 4.6 and 3.6 hours per day assisting patients with daily activities at 6 and 12 months, respectively. Improved patient motor and cognitive function were associated with reductions of up to 20 minutes per day in time spent in daily activities. Better patient mental health and cognitive function were associated with better carer mental health. Conclusions: Potentially modifiable factors such as these may be able to be targeted by caregiver training, support and education programmes and outpatient therapy for patients.

A Score Test for Zero-Inflation in Correlated Count Data

To account for the preponderance of zero counts and simultaneous correlation of observations, a class of zero-inflated Poisson mixed regression models is applicable for accommodating the within-cluster dependence. In this paper, a score test for zero-inflation is developed for assessing correlated count data with excess zeros. The sampling distribution and the power of the test statistic are evaluated by simulation studies. The results show that the test statistic performs satisfactorily under a wide range of conditions. The test procedure is further illustrated using a data set on recurrent urinary tract infections. Copyright (c) 2005 John Wiley & Sons, Ltd.

Trends for in situ and invasive melanoma in Queensland, Australia, 1982-2002

Objectives Queensland, the north-eastern state of Australia, has the highest incidence of melanoma in the world. Control measures started earlier here than probably anywhere else in the world; early detection programmes started in the 1960s and primary prevention in the 1980s. Data from the population-based Queensland Cancer Registry therefore provide an internationally unique data source with which to assess trends for in situ and invasive melanomas and to consider the implications for early detection and primary prevention. Methods We used Poisson regression to estimate the annual percentage change in rates across 21 years of incidence data for in situ and invasive lesions, stratified by age and sex. Joinpoint analyses were used to assess whether there had been a statistically significant change in the trends. Results In situ melanomas increased by 10.4% (95% CI: 10.1%, 11.1%) per year among males and 8.4% (7.9%, 8.9%) per year among females. The incidence of invasive lesions also increased, but not as quickly; males 2.6% (2.4%, 2.8%), females 1.2% (0.9%, 1.5%). Valid data on thickness was only available for 1991 to 2002 and for this period thin-invasive lesions were increasing faster than thick-invasive lesions (for example, among males: thin 3.8%, thick 2.0%). We found some suggestive evidence of lower proportionate increase for the most recent years for both in-situ and invasive lesions, but this did not achieve statistical significance. Among people younger than 35 years, the incidence of invasive melanoma was stable and there was a suggestion of a birth cohort effect from about 1958. Mortality rates were stable across all ages, and there was a suggestion of decreasing rates among young women, although this did not achieve statistical significance. Conclusion Age-standardised incidence is continuing to increase and this, in combination with a shift to proportionately more in situ lesions, suggests that the stabilisation of mortality rates is due, in large part, to earlier detection. For primary prevention, after a substantial period of sustained effort in Queensland, there is some suggestive, but not definitive, evidence that progress is being made. Incidence rates are stabilising in those younger than 35 years and the proportionate increase for both in situ and invasive lesions appears to be lower for the most recent period compared with previous periods. However, even taking the most favourable view of these trends, primary prevention is unlikely to lead to decreases in the overall incidence rate of melanoma for at least another 20 years. Consequently, the challenge for primary prevention programmes will be to maintain momentum over the long term. If this can be achieved, the eventual public-health benefits are likely to be substantial.

Habitat use by Darling Downs population of the grassland earless dragon: implications for conservation.

A population of the grassland earless dragon (Tympanocryptis pinguicolla) on the Darling Downs, Queensland, Australia, had been considered extinct until its recent rediscovery. We determined factors affecting grassland earless dragon abundance and prey availability in 3 habitats. Mean dragon body condition and prey numbers were higher in sorghum than grasslands and grass verges. Poisson regression analyses indicated that the dragon numbers were 10 times higher in sorghum, and that this may result from differences in prey numbers as well as other habitat conditions. Tracking data indicated selection of open versus closed microhabitat. Sorghum planted in rows provided alternating open and closed microhabitats for optimal thermoregulation conditions. Grasslands and grass verges were more uniformly shaded. Of individuals we tracked in the sorghum stubble, 85.7% used litter as overnight refuges. Litter was abundant in sorghum and sparse in grass habitats. The practices of minimum tillage and resting stubble strips possibly mitigate agricultural impacts on dragons and provide continuous access to suitable habitat. Changes in agricultural practices that affect the habitat suitability will potentially have detrimental impacts on the population. Our data suggest that conservation efforts be focused on maintaining suitability of habitats in crop fields. We recommend monitoring dragon abundance at control and trial sites of any new agricultural practices; this will provide opportunity to modify or stop undesirable practices before adoption by farmers. Conservation agencies may use our data as a baseline for monitoring long-term viability of the population.