The impact of Type 2 diabetes prevention programmes based on risk-identification and lifestyle intervention intensity strategies: a cost-effectiveness analysis.

AIMS: To develop a cost-effectiveness model to compare Type 2 diabetes prevention programmes targeting different at-risk population subgroups with a lifestyle intervention of varying intensity. METHODS: An individual patient simulation model was constructed to simulate the development of diabetes in a representative sample of adults without diabetes from the UK population. The model incorporates trajectories for HbA1c , 2-h glucose, fasting plasma glucose, BMI, systolic blood pressure, total cholesterol and HDL cholesterol. Patients can be diagnosed with diabetes, cardiovascular disease, microvascular complications of diabetes, cancer, osteoarthritis and depression, or can die. The model collects costs and utilities over a lifetime horizon. The perspective is the UK National Health Service and personal social services. We used the model to evaluate the population-wide impact of targeting a lifestyle intervention of varying intensity to six population subgroups defined as high risk for diabetes. RESULTS: The intervention produces 0.0003 to 0.0009 incremental quality-adjusted life years and saves up to £1.04 per person in the general population, depending upon the subgroup targeted. Cost-effectiveness increases with intervention intensity. The most cost-effective options are to target individuals with HbA1c > 42 mmol/mol (6%) or with a high Finnish Diabetes Risk (FINDRISC) probability score (> 0.1). CONCLUSION: The model indicates that diabetes prevention interventions are likely to be cost-effective and may be cost-saving over a lifetime. In the model, the criteria for selecting at-risk individuals differentially impact upon diabetes and cardiovascular disease outcomes, and on the timing of benefits. These findings have implications for deciding who should be targeted for diabetes prevention interventions.

Temporal and farm-management-associated variation in faecal pat prevalence of Arcobacter spp. in ruminants.

In a 2-year longitudinal study of adult animals on 15 dairy farms and four sheep farms in Lancashire, UK, Arcobacter spp. were isolated from all farms although not at every sampling occasion. Faecal samples were collected and cultured using standard techniques for isolation of campylobacters. Assignment to species was via PCR assays. Apparent prevalence of Arcobacter spp. was higher in dairy cattle compared to sheep (40.1% vs. 8%, P < 0.001) and in housed cattle compared to cattle at pasture (50.1% vs. 20.9%, P < 0.001). This was reflected in the higher prevalence observed in herds that were housed (n = 4) all year compared to herds that grazed cattle on pasture in the summer and housed cattle in the winter (n = 11) (55.5% vs. 36%, P < 0.001). In the case of sheep, peak prevalence was observed in autumn with increased prevalence also being associated with improving pasture quality. There was an apparent inverse association between the faecal pat prevalence of Arcobacter spp. and Campylobacter jejuni although this may in part be an artefact of laboratory test method sensitivity, whereby a relative increase in the frequency of one bacterial species would reduce the sensitivity of detecting the other.

A multi-state spatio-temporal Markov model for categorized incidence of meningitis in sub-Saharan Africa.

Meningococcal meningitis is a major public health problem in the African Belt. Despite the obvious seasonality of epidemics, the factors driving them are still poorly understood. Here, we provide a first attempt to predict epidemics at the spatio-temporal scale required for in-year response, using a purely empirical approach. District-level weekly incidence rates for Niger (1986-2007) were discretized into latent, alert and epidemic states according to pre-specified epidemiological thresholds. We modelled the probabilities of transition between states, accounting for seasonality and spatio-temporal dependence. One-week-ahead predictions for entering the epidemic state were generated with specificity and negative predictive value >99%, sensitivity and positive predictive value >72%. On the annual scale, we predict the first entry of a district into the epidemic state with sensitivity 65∙0%, positive predictive value 49∙0%, and an average time gained of 4∙6 weeks. These results could inform decisions on preparatory actions.

A hierarchical model for real-time monitoring of variation in risk of non-specific gastrointestinal infections.

The AEGISS (Ascertainment and Enhancement of Disease Surveillance and Statistics) project uses spatio-temporal statistical methods to identify anomalies in the incidence of gastrointestinal infections in the UK. The focus of this paper is the modelling of temporal variation in incidence using data from the Southampton area in southern England. We identified and fitted a hierarchical stochastic model for the time series of daily incident cases to enable probabilistic prediction of temporal variation in risk, and demonstrated the resulting gains in predictive accuracy by comparison with a conventional analysis based on an over-dispersed Poisson log-linear regression model. We used Bayesian methods of inference in order to incorporate parameter uncertainty in our predictive inference of risk. Incorporation of our model in the overall spatio-temporal model, will contribute to the accurate and timely prediction of unusually high food-poisoning incidence, and thus to the identification and prevention of future outbreaks.

Temporal and farm-management-associated variation in the faecal-pat prevalence of Campylobacter jejuni in ruminants.

In a 2-year longitudinal study of adult animals on 15 dairy farms and four sheep farms in Lancashire, UK. C. jejuni was isolated from all farms, although not on every occasion. Faecal samples were collected and cultured using standard techniques for isolation of Campylobacter. Assignment to species was via PCR assays. Peak prevalence of C. jejuni in both cattle and sheep was observed during the summer and in cattle this apparent seasonality was associated with grazing pasture [odds ratio (OR) 2.14], while in sheep it was independent of grazing. Increased prevalence was associated with increased milk yield (OR 1.05) and herd size (OR 1.01) in dairy cattle, and with increased stocking density (OR 1.29) and pasture quality (OR 2.16) in sheep. There was considerable variation in prevalence between farms but no evidence of large-scale spatial variation. The association between C. jejuni prevalence and diet in dairy cattle deserves further investigation.

Spatial and spatio-temporal analysis of Salmonella infection in dairy herds in England and Wales.

Using data from a cohort study conducted by the Veterinary Laboratories Agency (VLA), evidence of spatial clustering at distances up to 30 km was found for S. Agama and S. Dublin (P values of 0.001) and borderline evidence was found for spatial clustering of S. Typhimurium (P=0.077). The evolution of infection status of study farms over time was modelled using a Markov Chain model with transition probabilities describing changes in status at each of four visits, allowing for the effect of sampling visit. The degree of geographical clustering of infection, having allowed for temporal effects, was assessed by comparing the residual deviance from a model including a measure of recent neighbourhood infection levels with one excluding this variable. The number of cases arising within a defined distance and time period of an index case was higher than expected. This provides evidence for spatial and spatio-temporal clustering, which suggests either a contagious process (e.g. through direct or indirect farm-to-farm transmission) or geographically localized environmental and/or farm factors which increase the risk of infection. The results emphasize the different epidemiology of the three Salmonella serovars investigated.

Second-order analysis of inhomogeneous spatial point processes using case-control data.

Methods for the statistical analysis of stationary spatial point process data are now well established, methods for nonstationary processes less so. One of many sources of nonstationary point process data is a case-control study in environmental epidemiology. In that context, the data consist of a realization of each of two spatial point processes representing the locations, within a specified geographical region, of individual cases of a disease and of controls drawn at random from the population at risk. In this article, we extend work by Baddeley, Møller, and Waagepetersen (2000, Statistica Neerlandica54, 329-350) concerning estimation of the second-order properties of a nonstationary spatial point process. First, we show how case-control data can be used to overcome the problems encountered when using the same data to estimate both a spatially varying intensity and second-order properties. Second, we propose a semiparametric method for adjusting the estimate of intensity so as to take account of explanatory variables attached to the cases and controls. Our primary focus is estimation, but we also propose a new test for spatial clustering that we show to be competitive with existing tests. We describe an application to an ecological study in which juvenile and surviving adult trees assume the roles of controls and cases.

Clustering of equine grass sickness cases in the United Kingdom: a study considering the effect of position-dependent reporting on the space-time K-function.

Equine grass sickness (EGS) is a largely fatal, pasture-associated dysautonomia. Although the aetiology of this disease is unknown, there is increasing evidence that Clostridium botulinum type C plays an important role in this condition. The disease is widespread in the United Kingdom, with the highest incidence believed to occur in Scotland. EGS also shows strong seasonal variation (most cases are reported between April and July). Data from histologically confirmed cases of EGS from England and Wales in 1999 and 2000 were collected from UK veterinary diagnostic centres. The data did not represent a complete census of cases, and the proportion of all cases reported to the centres would have varied in space and, independently, in time. We consider the variable reporting of this condition and the appropriateness of the space-time K-function when exploring the spatial-temporal properties of a 'thinned' point process. We conclude that such position-dependent under-reporting of EGS does not invalidate the Monte Carlo test for space-time interaction, and find strong evidence for space time clustering of EGS cases (P < 0.001). This may be attributed to contagious or other spatially and temporally localized processes such as local climate and/or pasture management practices.

A mixture model for sucking patterns of breast-fed infants.

Previous studies of sucking patterns have mainly been on bottle-fed babies and have assumed that the babies' sucks occur within bursts separated by gaps of predetermined minimum length which is fixed over the feed. This study considers babies that are breast-fed, a more complex and natural process than bottle-feeding, and develops a more sophisticated model for the pattern of bursts and gaps which allows the parameters of the process to vary over the feed. We consider data from four breast feeds of each of 32 babies. We develop a two-component mixture model based on an underlying Markov chain model for the switching between bursts and gaps. We use the model to provide summary statistics for each feed and give estimates of the normal range of each of the model's parameters.

Non-parametric estimation of spatial variation in relative risk.

We consider the problem of estimating the spatial variation in relative risks of two diseases, say, over a geographical region. Using an underlying Poisson point process model, we approach the problem as one of density ratio estimation implemented with a non-parametric kernel smoothing method. In order to assess the significance of any local peaks or troughs in the estimated risk surface, we introduce pointwise tolerance contours which can enhance a greyscale image plot of the estimate. We also propose a Monte Carlo test of the null hypothesis of constant risk over the whole region, to avoid possible over-interpretation of the estimated risk surface. We illustrate the capabilities of the methodology with two epidemiological examples.

Second-order analysis of space-time clustering.

We consider the problem of detecting and describing space-time interaction in point process data. We extend existing second-order methods for purely spatial point process data to the spatial-temporal setting. This extension allows us to estimate space-time interaction as a function of spatial and temporal separation, and provides a useful reinterpretation of a popular test, due to Knox, for space-time interaction. Applications to simulated and real data indicate the method's potential.

Second-order analysis of spatial clustering for inhomogeneous populations.

Motivated by recent interest in the possible spatial clustering of rare diseases, the paper develops an approach to the assessment of spatial clustering based on the second-moment properties of a labelled point process. The concept of no spatial clustering is identified with the hypothesis that in a realisation of a stationary spatial point process consisting of events of two qualitatively different types, the type 1 events are a random sample from the superposition of type 1 and type 2 events. A diagnostic plot for estimating the nature and physical scale of clustering effects is proposed. The availability of Monte Carlo tests of significance is noted. An application to published data on the spatial distribution of childhood leukaemia and lymphoma in North Humberside is described.