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.

Likelihood-based estimation of the regression model with scrambled responses

A significant problem in the collection of responses to potentially sensitive questions, such as relating to illegal, immoral or embarrassing activities, is non-sampling error due to refusal to respond or false responses. Eichhorn & Hayre (1983) suggested the use of scrambled responses to reduce this form of bias. This paper considers a linear regression model in which the dependent variable is unobserved but for which the sum or product with a scrambling random variable of known distribution, is known. The performance of two likelihood-based estimators is investigated, namely of a Bayesian estimator achieved through a Markov chain Monte Carlo (MCMC) sampling scheme, and a classical maximum-likelihood estimator. These two estimators and an estimator suggested by Singh, Joarder & King (1996) are compared. Monte Carlo results show that the Bayesian estimator outperforms the classical estimators in almost all cases, and the relative performance of the Bayesian estimator improves as the responses become more scrambled.

Finite mixture regression model with random effects: application to neonatal hospital length of stay

A two-component mixture regression model that allows simultaneously for heterogeneity and dependency among observations is proposed. By specifying random effects explicitly in the linear predictor of the mixture probability and the mixture components, parameter estimation is achieved by maximising the corresponding best linear unbiased prediction type log-likelihood. Approximate residual maximum likelihood estimates are obtained via an EM algorithm in the manner of generalised linear mixed model (GLMM). The method can be extended to a g-component mixture regression model with the component density from the exponential family, leading to the development of the class of finite mixture GLMM. For illustration, the method is applied to analyse neonatal length of stay (LOS). It is shown that identification of pertinent factors that influence hospital LOS can provide important information for health care planning and resource allocation. (C) 2002 Elsevier Science B.V. All rights reserved.

Hysterectomy fractions in New South Wales, 1971-2006

Hysterectomy fractions by age group for particular periods are of interest for: estimating proper population denominators for calculation of disease and procedure rates affecting the cervix and uterus; estimating the target population for Pap test programs, and response rates; and as a way of displaying the cumulative consequences of hysterectomies in a population. Hysterectomy fractions for populations can be determined by direct inquiry via a representative sample survey, or, as in this study, from prior hysterectomy rates of the cohorts of women which compose each age bracket. Hysterectomy data 1979-93 were obtained from the hospital In-patients Statistics Collection (ISC) which covers both public and private hospitals in NSW. Annual population denominators of women were obtained from Census data. Data were modelled by Poisson regression, using five.-year age group (15-greater than or equal to 85 years), annual period, and five-year birth cohort (APC model). Forward- and back-projection of the period effects were undertaken. The resultant NSW hysterectomy fractions by age and period are consistent with fractions obtained from modelled hysterectomy rates for Western Australia (1980-84), and fractions from national representative sample surveys (1989/90 and 1995) for younger women, but not for women aged greater than or equal to 70 years in 1995, which revealed higher hysterectomy fractions than modelled hysterectomy data would suggest. Hysterectomy fractions for NSW women by five-year age group for quinquennia centred on 1971 to 2006 are provided.

Anthropometric measures in relation to Basal Cell Carcinoma: a longitudinal study

Background: The relationship between anthropometric indices and risk of basal cell carcinoma ( BCC) is largely unknown. We aimed to examine the association between anthropometric measures and development of BCC and to demonstrate whether adherence to World Health Organisation guidelines for body mass index, waist circumference, and waist/ hip ratio was associated with risk of BCC, independent of sun exposure. Methods: Study participants were participants in a community- based skin cancer prevention trial in Nambour, a town in southeast Queensland ( latitude 26 degrees S). In 1992, height, weight, and waist and hip circumferences were measured for all 1621 participants and weight was remeasured at the end of the trial in 1996. Prevalence proportion ratios were calculated using a log- binomial model to estimate the risk of BCC prior to or prevalent in 1992, while Poisson regression with robust error variances was used to estimate the relative risk of BCC during the follow- up period. Results: At baseline, 94 participants had a current BCC, and 202 had a history of BCC. During the 5- year follow- up period, 179 participants developed one or more new BCCs. We found no significant association between any of the anthropometric measures or indices and risk of BCC after controlling for potential confounding factors including sun exposure. There was a suggestion that short- term weight gain may increase the risk of developing BCC for women only. Conclusion: Adherence to World Health Organisation guidelines for body mass index, waist circumference and waist/ hip ratio is not significantly associated with occurrence of basal cell carcinomas of the skin.

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.

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.

Optimal crossover designs for logistic regression models in pharmacodynamics

Pharmacodynamics (PD) is the study of the biochemical and physiological effects of drugs. The construction of optimal designs for dose-ranging trials with multiple periods is considered in this paper, where the outcome of the trial (the effect of the drug) is considered to be a binary response: the success or failure of a drug to bring about a particular change in the subject after a given amount of time. The carryover effect of each dose from one period to the next is assumed to be proportional to the direct effect. It is shown for a logistic regression model that the efficiency of optimal parallel (single-period) or crossover (two-period) design is substantially greater than a balanced design. The optimal designs are also shown to be robust to misspecification of the value of the parameters. Finally, the parallel and crossover designs are combined to provide the experimenter with greater flexibility.

Suicide in urban New South Wales, Australia 1985-1994: Socio-economic and migrant interactions

Variation of suicide with socio-economic status (SES) in urban NSW (Australia) during 1985-1994, by sex and country or region of birth, was examined using Poisson regression analysis of vital statistics and population data (age greater than or similar to 15 yr). Quintiles of SES were defined by municipality of residence and comparisons of suicide by SES were adjusted for age and country (or region) of birth (COB), and examined by COB. Risk of suicide in females was 28% that of males for all adults and 21% for youth (age 15-24 yr). Suicide risk was lower in males from southern Europe, Middle East and Asia, and higher in northern and eastern European males, compared to the Australian-born. Risks for suicide increased significantly with decreasing SES in males, but not in females. The relationship of male suicide and SES was stronger when controlled for COB. For males, the relative risk of suicide, adjusted for age and COB, was 66% higher in the lowest SES quintile compared to the highest quintile, and 39% higher for youth (age 15-24 yr). For male suicide, the population attributable fraction for SES (less than the highest quintile) was 27%. Analysis of SES differentials in male suicide according to COB indicated a significant inverse suicide gradient in relation to SES for the Australian-born and those burn in New Zealand and the United Kingdom or fire. but not in non-English speaking COB groups, except for Asia. For Australian-born males, suicide risk was 71% higher in the lowest SES group (compared to the highest), adjusted for age. These findings indicate that SES plays an important role in male suicide rates among the Australian-born and migrants from English-speaking countries and Asia, and among youth; but not in female suicide, nor suicide in most non-English speaking migrant groups. Reduction in SES differentials through economic and social policies may reduce male suicide in lower SES groups and should be seen to be at least as important as individual level interventions. (C) 1998 Elsevier Science Ltd. All rights reserved.

Socio-economic, migrant and geographic differentials in coronary heart disease occurrence in New South Wales

Objective: This study examines the variation in coronary heart disease (CHD) mortality and acute myocardial infarction (AMI) by socio-economic status (SES), country of birth (COB) and geography (urban/rural) in the total population of New South Wales (Australia) in 1991-95. Method: CHD deaths and AMI are from complete enumerations of deaths and hospital admissions, respectively; and population denominators are from census information. Data are examined separately by sex, and comparisons of SES groups (based on municipalities), COB and region are analysed using Poisson regression, after adjustment for age. Results: The study identified higher risk for AMI admissions and CHD mortality in lower SES populations with significant linear trends, for both sexes, adjusted for age, region and COB. According to the population attributable fractions (PAF), 23-41% of the risk of CHD occurrence is due to SES lower than the highest quartile. The higher age-adjusted risk for CHD occurrence in rural and remote populations for both sexes, compared with urban communities, was lessened by adjustment for COB, and all but abolished when also adjusted for SES. COB analysis indicated significantly lower age-adjusted AMI admissions and CHD mortality compared with the Australian-born, Conclusions: Higher risks for CHD in rural populations compared with the capital city (Sydney) are due, in part, to lower SES, lesser migrant composition. Implications: Strategies for reducing CHD differentials should consider demographic factors and the fundamental need to reduce socio-economic inequalities, as well as targeting appropriate prevention measures.

Urban and rural suicide differentials in migrants and the Australian-born, New South Wales, Australia 1985-1994

We estimated risk of suicide in adults in New South Wales (NSW) by sex, country of birth and rural/urban residence, after adjusting for age; we also examined youth suicide (age 15-24 years). The study population was the entire population of NSW, Australia, aged greater than or equal to 15 years during the period 1985-1994. Poisson regression was used to examine the relationship between predictor variables and the risk of suicide, with the focus on migrant status and area of residence. A significantly higher risk of suicide was found in male migrants from Northern Europe and Eastern Europe/former USSR, compared to Australian-born males; a significantly lower suicide risk occurred in males from Southern Europe, the Middle East and Asia. In female migrants, those from UK/Eire, Northern Europe, Eastern Europe/former USSR and New Zealand exhibited a significantly higher risk of suicide compared to Australian-born females. A significantly lower risk of suicide occurred in females from the Middle East. Male migrants overall were at significantly lower risk of suicide than the Australian-born, while female migrants overall had a significantly higher risk of suicide than Australian-born females. Among migrant males overall, the rural-urban suicide risk differential was significantly higher for those living in non-metropolitan areas (RR = 1.9; 95% CI: 1.7-2.1). Suicide risk was significantly higher in non-metropolitan male immigrants from the UK/Eire (RR = 1.4; 95% CI: 1.1-1.7), Southern Europe (RR = 1.7; 95% CI: 1.2-2.4), Northern/Western Europe (1.5; 95% CI: 1.2-1.9), the Middle East (RR = 3.8; 95% CI: 1.9-7.8), New :Zealand (RR = 1.4; 95% CI: 1.0-1.8) and 'other' (RR = 2.6; 95% CI: 1.9-3.5), when compared to their urban counterparts. There was no statistically significant difference in suicide risk between rural and urban Australian-born males. For female suicide, significantly lower risk was found in female immigrants living in non-metropolitan areas who were from Northern/Western Europe (RR = 0.7; 95% CI: 0.4-0.96), as well as the Australian-born (RR = 0.7; 95% CI: 0.6-0.8), when compared to their urban counterparts. The non-metropolitan/metropolitan relative risk for suicide in female migrants overall was not significantly different from one. Among male youth there was a significantly higher suicide risk in non-metropolitan areas, with a relative risk estimate of 1.4 for Australian-born youth (95% CI: 1.2-1.5) and 1.7 for migrant youth (95% CI: 1.2-2.4), when compared with metropolitan counterparts. We conclude that suicide among migrant males living in non-metropolitan areas accounts for most of the excess of male suicide in rural NSW, and the significantly lower risk of suicide for non-metropolitan Australian-born women does not apply to migrant women. (C) 1999 Published by Elsevier Science Ltd. All rights reserved.

Trends in the incidence, severity, and short-term outcome of stroke in Perth, Western Australia

Background and Purpose-This report describes trends in the key indices of cerebrovascular disease over 6 years from the end of the 1980s in a geographically defined segment of the city of Perth, Western Australia. Methods-Identical methods were used to find and assess all cases of suspected stroke in a population of approximately 134 000 residents in a triangular area of the northern suburbs of Perth. Case fatality was measured as vital status at 28 days after the onset of symptoms. Data for first-ever strokes and for all strokes for equivalent periods of 12 months in 1989-1990 and 1995-1996 were compared by age-standardized rates and proportions and Poisson regression. Results-There were 355 strokes in 328 patients and 251 first-ever strokes (71%) for 1989-1990 and 290 events in 281 patients and 213 first-ever strokes (73%) for 1995-1996. In Poisson models including age and period, overall trends in the incidence of both first-ever strokes (rate ratio = 0.75; 95% confidence limits, 0.63, 0.90) and all strokes (rate ratio = 0.73; 95% confidence limits, 0.62, 0.85) were obviously significant, but only the changes in men were independently significant. Case fatality did not change, and the balance between hemorrhagic and occlusive strokes in 1995-1996 was almost indistinguishable from that observed in 1989-1990. Conclusions-Our results, which are the only longitudinal population-based data available for Australia for key indices of stroke, suggest that it is a change in the frequency of stroke, rather than its outcome, that is chiefly responsible nationally for the fall in mortality from cerebrovascular disease.

Who fails to complete tuberculosis treatment? Temporal trends and risk factors for treatment interruption in a community-based directly observed therapy programme in a rural district of South Africa

SETTING: Hlabisa Tuberculosis Programme, Hlabisa, South Africa. OBJECTIVE: To determine trends in and risk factors for interruption of tuberculosis treatment. METHODS: Data were extracted from the control programme database starting in 1991. Temporal trends in treatment interruption are described; independent risk factors for treatment interruption were determined with a multiple logistic regression model, and Kaplan-Meier survival curves for treatment interruption were constructed for patients treated in 1994-1995. RESULTS: Overall 629 of 3610 surviving patients (17%) failed to complete treatment; this proportion increased from 11% (n = 79) in 1991/1992 to 22% (n = 201) in 1996. Independent risk factors for treatment interruption were diagnosis between 1994-1996 compared with 1991-1393 (odds ratio [OR] 1.9, 95% confidence interval [CT] 1.6-2.4); human immunodeficiency virus (HIV) positivity compared with HIV negativity (OR 1.8, 95% CI 1.4-2.4); supervised by village clinic compared with community health worker (OR 1.9, 95% CI 1.4-2.6); and male versus female sex (OR 1.3, 95% CI 1.1-1.6). Few patients interrupted treatment during the first 2 weeks, and the treatment interruption rate thereafter was constant at 1% per 14 days. CONCLUSIONS: Frequency of treatment interruption from this programme has increased recently. The strongest risk factor was year of diagnosis, perhaps reflecting the impact of an increased caseload on programme performance. Ensuring adherence to therapy in communities with a high level of migration remains a challenge even within community-based directly observed therapy programmes.

Dental set-vice provision in private general practice for an insured population in New South Wales, Australia

Objective To determine patterns of dental set-vices provided to a cohort of the insured population 18 years and over, in private general practice in New South Wales, Australia. Basic research design A cohort study using the person-years method and Poisson regression for analysis. Setting Data were derived from claims records submitted by members of a health insurance fund (Government Employees Health Fund-GEHF) for rebates during the study period 1 January 1992-31 December 1995. Participants There were 133,467 members aged 18 years and over from New South Wales. Main outcome measures To determine, by age group, for those members who used private general practice and made a claim (referred to as 'patients') the annual number of visits, total number of services received per year and number of services received at a visit, Results The mean number of visits per patient was 2.4 per year with patients under 45 years making fewer visits than the 45-54 age group reference category. Mean number of services utilised per patient-year was 5.9, with services provided increasing from 3.5 for the 18-24-year-old group, reaching a plateau of approximately 6.2 for those aged 45 years or more. The number of services received per visit was 2.4 and there were no differences by age. Service mix was dominated by restorative (35%), diagnostic (27%), and preventive services (18%); with age specific variations observed. Conclusions Age was found to be an important determinant in the use of dental services, independent of membership duration and gender.