Biomedical subjects

Kelvin K W Yau

Publications and source records attributed to Kelvin K W Yau.

At least 19 recordsLinked to original sources

PubMed2007

Assessing local influence in principal component analysis with application to haematology study data.

In many medical and health studies, high-dimensional data are often encountered. Principal component analysis (PCA) is a commonly used technique to reduce such data to a few components that includes most of the information provided by the original data. However, PCA is known to be very sensitive to some abnormal observations. Therefore, it is essential to assess such sensitivity in PCA. In this paper, the assessments of local influence based on generalized influence function are developed under the case-weights and additive perturbation schemes, along with a discussion of the perturbation scheme and the generalized influence function approach. When perturbing different variables of the data, it is noted that the directions of the largest joint local influence for the eigenvalues are all the same. Moreover, these directions are completely determined by the score values of the observations, to which an approximate cut-off point is given. The proposed methods are applied to analyse a set of haematology study data for illustration. Results add new insights in finding influential observations in the studied data set.

Health Status↗

PubMed2007

A score test for overdispersion in zero-inflated poisson mixed regression model.

Count data with extra zeros are common in many medical applications. The zero-inflated Poisson (ZIP) regression model is useful to analyse such data. For hierarchical or correlated count data where the observations are either clustered or represent repeated outcomes from individual subjects, a class of ZIP mixed regression models may be appropriate. However, the ZIP parameter estimates can be severely biased if the non-zero counts are overdispersed in relation to the Poisson distribution. In this paper, a score test is proposed for testing the ZIP mixed regression model against the zero-inflated negative binomial alternative. Sampling distribution and 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 applied to pancreas disorder length of stay that comprised mainly same-day separations and simultaneous prolonged hospitalizations.

Adult↗

PubMed2006

Multilevel logistic regression modelling with correlated random effects: application to the Smoking Cessation for Youth study.

A multilevel logistic regression model is presented for the analysis of clustered and repeated binary response data. At the subject level, serial dependence is expected between repeated measures recorded on the same individual. At the cluster level, correlations of observations within the same subgroup are present due to the inherent hierarchical setting. Two random components are therefore incorporated explicitly within the linear predictor to account for the simultaneous heterogeneity and autoregressive structure. Application to analyse a set of longitudinal data from an adolescent smoking cessation intervention that motivated this study is illustrated.

Adolescent↗

PubMed2006

Multiple-vehicle traffic accidents in Hong Kong.

'Multiple-vehicle traffic accident' refers to a crash between two or more moving objects. Unlike single-vehicle accidents, not all drivers involving in a multiple-vehicle accident are responsible for the occurrence of the event. Accordingly, variables such as road type, speed limit and number of vehicles involved in the accident are expected to play a much more important role in association with injury severity in multiple-vehicle accidents. To study the factors influencing injury severity of multiple-vehicle traffic accidents, a population-based study was conducted. The traffic accident data was obtained from the Traffic Accident Data System (TRADS), which was developed by the Transport Department, Police Force and Information Technology Services Department, Hong Kong. Multiple-vehicle traffic accidents (N = 10,630) occurring during the 2-year period 1999/2000 were considered. Potential risk factors such as district, human, vehicle, safety, environmental and site factors were examined. Categorizing injury severity into "fatal/serious" and "slight", a stepwise logistic regression model was applied to the population data set. The district board, time of the accident, driver's gender, vehicle type, road type, speed limit and the number of vehicles involved are significant factors influencing the injury severity. Identification of risk factors for severe traffic accidents provides valuable information to help with new and improved road safety control measures.

Accidents, Traffic↗

PubMed2006

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.

Aged↗

PubMed2006

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 render 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.

Adult↗

PubMed2006

Under-nutrition affects time to recurrence of gastroenteritis among Aboriginal and non-Aboriginal children.

This study investigated whether under-nutrition affected time to hospitalization for recurrence of gastroenteritis in Australian children. Linked hospitalization records of all infants, born in 1995 and 1996 in Western Australia, who were admitted for gastroenteritis during their first year of life (n=1001), were retrieved. A survival frailty model was used for determining the factors influencing the recurrent times over the subsequent seven years. Aboriginality and under-nutrition were significantly associated with an increased risk of recurrence (hazard ratios of 2.59 and 1.28). Hospitalizations due to gastroenteritis were common among Aboriginal children who had a higher mean re-admission rate and much shorter intervals between re-admissions than other patients. The proportion of patients with recurrence was also significantly higher for Aboriginals (38.5%) than for other patients (14.2%). Gastroenteritis remains a serious problem in Aboriginal children. This presents a complex challenge to be addressed with public-health principles, political determination and commitment, and adequate resources.

Cohort Studies↗

PubMed2005

Influence diagnostics for two-component Poisson mixture regression models: applications in public health.

In many medical and health applications, Poisson mixture regression models are commonly used to analyse heterogeneous count data. Motivated by two data sets drawn from public health studies, influence diagnostics are proposed for assessing the sensitivity of the fitted two-component Poisson mixture regression models. Under various perturbations of the observed data or model assumptions, influence assessments based on the local influence approach are developed for detecting clusters and/or individual observations that impact on the estimation of model parameters. Results from studies on recurrent urinary tract infections and maternity length of stay illustrate the usefulness of the influence diagnostics.

Cluster Analysis↗

PubMed2005

Evaluating the effectiveness of a participatory ergonomics approach in reducing the risk and severity of injuries from manual handling.

Manual handling is the greatest contributor to non-fatal injury and disease in the workplace, commonly accounting for one-third of national injury counts. Interventional strategies that have focused on selecting or modifying the worker have been ineffective in reducing injury risk. In recent times, participatory ergonomics has been widely adopted as a process to reduce the risk of injury from manual handling but it is not well validated as an intervention. This study evaluated the effectiveness of a participatory ergonomics risk assessment approach in reducing the rate and severity of injuries from manual and non-manual handling sustained by a cohort of 137 cleaners within a hospital setting. The date of injury and the workers' compensation claim cost and hours lost from work were obtained for each injury incurred during the 4-year pre-intervention and 3-year intervention period. The age, gender and hours worked were ascertained for every cleaner whether injured or not. Using generalized linear mixed modelling analysis, reductions of rate of injury by two-thirds, workers' compensation claim costs by 62% and hours lost by 35% for manual handling injuries were found to be associated with the intervention period. Although the cleaners experienced a significant intervention period reduction in non-manual handling injury rate, the corresponding changes in severity of injury were not significant. The success of the intervention supports the adoption of a participatory ergonomics approach in reducing the rate and consequence of injuries in the workplace.

Adult↗

PubMed2005

A robustified modeling approach to analyze pediatric length of stay.

PURPOSE: Length of stay (LOS) is an important measure of the cost of pediatric hospitalizations, but the guidelines developed so far are not rigorously evidence-based. This study demonstrates a robust gamma mixed regression approach to analyze the positively skewed LOS variable, which has implications for future studies of pediatric health care management. METHODS: The robustified approach is applied to analyze hospital discharge data on childhood gastroenteritis in Western Australia (n=514). The model accounts for demographic characteristics and co-morbidities of the patients, as well as the dependency of LOS outcomes nested within the 58 hospitals in the State. The method is compared with the standard linear mixed regression with trimming of extreme observations. RESULTS: For the empirical application, the linear mixed regression results are sensitive to the magnitude of trimming. The identified significant factors from the robust regression model, namely infection, failure to thrive, and iron deficiency anemia are resistant to high-LOS outliers. CONCLUSIONS: Robust gamma mixed regression appears to be a suitable alternative to analyze the clustered and positively skewed pediatric LOS, without transforming and trimming the data arbitrarily.

Child, Hospitalized↗

PubMed2005

Modelling bivariate count series with excess zeros.

Bivariate time series of counts with excess zeros relative to the Poisson process are common in many bioscience applications. Failure to account for the extra zeros in the analysis may result in biased parameter estimates and misleading inferences. A class of bivariate zero-inflated Poisson autoregression models is presented to accommodate the zero-inflation and the inherent serial dependency between successive observations. An autoregressive correlation structure is assumed in the random component of the compound regression model. Parameter estimation is achieved via an EM algorithm, by maximizing an appropriate log-likelihood function to obtain residual maximum likelihood estimates. The proposed method is applied to analyze a bivariate series from an occupational health study, in which the zero-inflated injury count events are classified as either musculoskeletal or non-musculoskeletal in nature. The approach enables the evaluation of the effectiveness of a participatory ergonomics intervention at the population level, in terms of reducing the overall incidence of lost-time injury and a simultaneous decline in the two mean injury rates.

Accidents, Occupational↗

PubMed2005

Multilevel modelling of hospitalisations for recurrent diarrhoeal disease in Aboriginal and non-Aboriginal infants and young children in Western Australia.

A multilevel proportional hazards model was used to determine the prognostic factors affecting hospitalisations for recurrent diarrhoeal disease in infants and young children. All infants born in 1996 who had index gastroenteritis admission to Western Australian (WA) hospitals during their first year of life were included in our study cohort (n = 486). There were 618 hospital admissions for infectious diarrhoea over the 6-year follow-up period. Aboriginal children stayed significantly longer in hospital than their non-Aboriginal peers, and comorbidities such as dehydration, gastrointestinal sugar intolerance, failure to thrive, iron deficiency anaemia and certain infections (genitourinary, scabies and/or otitis media) were all significantly associated with the delayed discharge of patients. Substantial variations among patients (variance = 0.660) and between health service regions of WA (variance = 0.296) were found. Over 30% of the total variation could be attributed to the heterogeneity between health districts. For any two patients in the same health district, the within-region (intraclass) correlation was estimated to be 0.309. In the absence of detailed socio-demographic data, application of the standard survival procedure may lead to incorrect inferences due to regional clustering and repeated observations on individuals. By accounting for latent patient and regional effects, the multilevel analysis clearly confirmed the high burden of infectious diarrhoea among Aboriginal infants and children, and their much longer hospital stays. Coexisting morbidities contributed to the prolonged and recurrent hospitalisations. Findings of this epidemiological study indicated the need of multifaceted clinical disease prevention and hygiene promotion strategies to control the disease.

Anemia, Iron-Deficiency↗

PubMed2004

Modelling the distribution of ischaemic stroke-specific survival time using an EM-based mixture approach with random effects adjustment.

A two-component survival mixture model is proposed to analyse a set of ischaemic stroke-specific mortality data. The survival experience of stroke patients after index stroke may be described by a subpopulation of patients in the acute condition and another subpopulation of patients in the chronic phase. To adjust for the inherent correlation of observations due to random hospital effects, a mixture model of two survival functions with random effects is formulated. Assuming a Weibull hazard in both components, an EM algorithm is developed for the estimation of fixed effect parameters and variance components. A simulation study is conducted to assess the performance of the two-component survival mixture model estimators. Simulation results confirm the applicability of the proposed model in a small sample setting.

Age Factors↗

PubMed2004

Risk factors for ischaemic stroke recurrence after hospitalisation.

OBJECTIVE: To determine risk factors for ischaemic stroke recurrence among patients admitted to hospital for a first-ever occurrence of ischaemic stroke. DESIGN, SETTING AND PATIENTS: Retrospective study involving linked hospitalisation and death records. The cohort comprised 7816 people who were hospitalised for first-ever ischaemic stroke between July 1995 and December 1999 in Western Australia. Cox's proportional hazards model was used to identify risk factors for stroke recurrence. MAIN OUTCOME MEASURES: Time to first recurrence; cumulative recurrence risk; risk factors for recurrence. RESULTS: The median time to first stroke recurrence was 255 days. The cumulative probability of first recurrence was 5.1% (95% CI, 4.6%-5.7%) at 6 months, 8.4% (95% CI, 7.6%-9.1%) at 1 year and 19.8% (95% CI, 18.1%-21.4%) at 4 years. The risk of first recurrence was increased by advancing age (hazard ratio [HR], 1.03; 95% CI, 1.02-1.04), Aboriginality (HR, 1.50; 95% CI, 1.02-2.22), diabetes (HR, 1.27; 95% CI, 1.07-1.51), a history of cardiac conditions (HR, 1.18; 95% CI, 1.01-1.38), post-stroke urinary incontinence (HR, 1.27; 95% CI, 1.03-1.57) and transfer to another hospital on index admission (HR, 1.26; 95% CI, 1.08-1.46). Admission at first stroke occurrence to a hospital maintaining a stroke unit reduced the risk of recurrence (HR, 0.84; 95% CI, 0.72-0.99). CONCLUSION: The risk factors identified in our study have implications for planning secondary prevention strategies. In particular, Aboriginality and transfer to another hospital upon admission for first-ever ischaemic stroke were important risk factors. Research into the level of compliance and access to stroke treatment by Aboriginal patients to prevent further strokes is required.

Aged↗

PubMed2004

Recurrent ischaemic stroke hospitalisations: a retrospective cohort study using Western Australia linked patient records.

A retrospective cohort study was undertaken to determine factors that affect the frequency of recurrent ischaemic stroke hospitalisations. Linked hospitalisation records of all Western Australian patients admitted for ischaemic stroke for the first time during July-December 1995 were retrieved until December 2000 to derive the number of readmissions for recurrent strokes, patient medical conditions and co-morbidities at the index episode. A negative binomial regression model adjusting for inter-hospital variations was used to determine the prognostic factors influencing recurrent stroke hospitalisations. Of the 678 patients in the cohort, 124 (18.3%) experienced repeated episodes of ischaemic stroke. Rural residence and carotid endarterectomy procedure were positively associated with the recurrence frequency, the adjusted incidence rate ratio being 1.66 (95% CI: 1.17-2.36) and 3.96 (95% CI: 2.30-6.82), respectively. Rural patients contributed to 18% of the patients in the cohort yet they accounted for 27% of those sustaining repeated episodes of stroke. Readmissions were also related to the presence of diabetes at the index episode. The effect of diabetes, with adjusted incidence rate ratio 1.35 (95% CI: 1.01-1.79), was only evident after accounting for within hospital correlations. These findings have implications on hospital resource planning and secondary preventive strategies to reduce the burden of stroke.

Aged↗

PubMed2004

Ischemic stroke hospital stay and discharge destination.

PURPOSE: Stroke hospitalizations are among the most expensive, with a mean length of stay (LOS) higher than other hospitalizations. This retrospective study assesses factors influencing ischemic stroke LOS taking into consideration the discharge destination of patients. METHODS: Linked hospital separation records between July 1995 and December 1999 were extracted to determine the first admission for ischemic stroke in Western Australia. Multiple hospitalization records for the same patient were screened to obtain the total duration of hospitalization. Demographic characteristics, hospital type, and medical history of patients were also retrieved. In the presence of censoring and without prior assumption on the time-to-discharge distribution, the Cox's proportional hazards model was used to assess the factors affecting LOS. RESULTS: During the study period, 6469 patients with a first-ever admission for ischemic stroke were identified, with average LOS being 28 days (95% CI, 26-30 days). Hospital stays were significantly longer for females and patients directly admitted to hospitals maintaining a specialist stroke unit, whereas patients residing in rural areas had shorter stays. CONCLUSIONS: The impact of stroke severity and placement in nursing homes after discharge need to be understood to manage LOS and the cost of acute care. Accurate diagnosis upon initial presentation would benefit both the efficiency of hospitals and the outcomes of rural ischemic stroke patients.

Age Distribution↗

PubMed2004

Risk factors affecting the severity of single vehicle traffic accidents in Hong Kong.

A population-based case-control study was conducted to examine factors affecting the severity of single vehicle traffic accidents in Hong Kong. In particular, single vehicle accident data of three major vehicle types, namely private vehicles, goods vehicles and motorcycles, which contributed to over 80% of all single vehicle accidents during the 2-year-period 1999-2000, were considered. Data were obtained from the newly implemented traffic accident data system (TRADS), which was developed jointly by the Transport Department, Police Force and Information Technology Services Department, Hong Kong. The effect of district, human, vehicle, safety, environmental and site factors on injury severity of an accident was examined. Unique risk factors associated with each of the vehicle types were identified by means of stepwise logistic regression models. For private vehicles, district board, gender of driver, age of vehicle, time of the accident and street light conditions are significant factors determining injury severity. For goods vehicles, seat-belt usage and weekday occurrence are the only two significant factors associated with injury severity. For motorcycles, age of vehicle, weekday and time of the accident were determined to be important factors affecting the injury severity. Identification of potential risk factors pertinent to the particular vehicle type has important implications to relevant official organisations in modifying safety measures in order to reduce the occurrence of severe traffic accidents, which would help to promote a safe road environment.

Accidents, Traffic↗

PubMed2004

Modeling zero-inflated count series with application to occupational health.

A zero-inflated Poisson mixed autoregression model is presented for analyzing time series of count events with excess zeros. The model is motivated by the evaluation of a participatory ergonomics intervention intended to reduce manual handling workplace injuries over a specified time period. Random effects are introduced into the linear predictor of the model to account for serial correlation between successive observations. Parameter estimation is achieved by maximizing an appropriate log-likelihood function to obtain approximate residual maximum likelihood estimates. The method enables the evaluation of occupational intervention using population level aggregated count data series containing extra zeros.

Occupational Health↗