PubMed Health⌕ Search

Biomedical subjects

Nicky Best

Publications and source records attributed to Nicky Best.

13 recordsLinked to original sources
PubMed2006

Improving ecological inference using individual-level data.

In typical small-area studies of health and environment we wish to make inference on the relationship between individual-level quantities using aggregate, or ecological, data. Such ecological inference is often subject to bias and imprecision, due to the lack of individual-level information in the data. Conversely, individual-level survey data often have insufficient power to study small-area variations in health. Such problems can be reduced by supplementing the aggregate-level data with small samples of data from individuals within the areas, which directly link exposures and outcomes. We outline a hierarchical model framework for estimating individual-level associations using a combination of aggregate and individual data. We perform a comprehensive simulation study, under a variety of realistic conditions, to determine when aggregate data are sufficient for accurate inference, and when we also require individual-level information. Finally, we illustrate the methods in a case study investigating the relationship between limiting long-term illness, ethnicity and income in London.

Data Interpretation, Statistical↗
PubMed2006

A Bayesian toolkit for genetic association studies.

We present a range of modelling components designed to facilitate Bayesian analysis of genetic-association-study data. A key feature of our approach is the ability to combine different submodels together, almost arbitrarily, for dealing with the complexities of real data. In particular, we propose various techniques for selecting the "best" subset of genetic predictors for a specific phenotype (or set of phenotypes). At the same time, we may control for complex, non-linear relationships between phenotypes and additional (non-genetic) covariates as well as accounting for any residual correlation that exists among multiple phenotypes. Both of these additional modelling components are shown to potentially aid in detecting the underlying genetic signal. We may also account for uncertainty regarding missing genotype data. Indeed, at the heart of our approach is a novel method for reconstructing unobserved haplotypes and/or inferring the values of missing genotypes. This can be deployed independently or, alternatively, it can be fully integrated into arbitrary genotype- or haplotype-based association models such that the missing data and the association model are "estimated" simultaneously. The impact of such simultaneous analysis on inferences drawn from the association model is shown to be potentially significant. Our modelling components are packaged as an "add-on" interface to the widely used WinBUGS software, which allows Markov chain Monte Carlo analysis of a wide range of statistical models. We illustrate their use with a series of increasingly complex analyses conducted on simulated data based on a real pharmacogenetic example.

Bayes Theorem↗
PubMed2005

Studying time to pregnancy by use of a retrospective design.

Biologic fertility can be measured using time to pregnancy (TTP). Retrospective designs, although lacking detailed timed information about behavior and exposure, are useful since they have a well-defined target population, often have good response rates, and are simpler and less expensive to conduct than prospective studies. This paper reviews retrospective TTP studies from a methodological viewpoint and shows how methodological problems can be avoided or minimized by appropriate study design, conduct, and analysis. Sensitivity analyses using data from four European retrospective TTP studies are presented to explore the issues. Although the identified biases tend to have small impacts, the effects are not systematic across studies, and sensitivity analyses are recommended routinely. Planning bias can be checked by comparing propensity to report contraceptive failures in different exposure groups. Medical intervention bias can be avoided by censoring and inclusion of unsuccessful pregnancy attempts. Truncation bias can be a serious problem if unrecognized, but it is avoidable with appropriate study design and/or analysis. Behavior change bias can be minimized by assessing the covariates at the beginning of unprotected intercourse. More complete inference is possible if the study design covers the whole population, not just those who achieve a pregnancy.

Bias↗
PubMed2005

Modelling exposure to disinfection by-products in drinking water for an epidemiological study of adverse birth outcomes.

We are conducting an epidemiological study on the association between disinfection by-product concentrations in drinking water and adverse birth outcomes in the UK, using trihalomethane (THM) concentrations over defined water zones as an exposure index. Here we construct statistical models using sparse routinely collected THMs measurements to obtain quarterly estimates of mean THM concentrations for each water zone. We modelled the THM measurements using a Bayesian hierarchical mixture model, taking into account heterogeneity in THM concentrations between water originating from different source types, quarterly variation in THM concentrations and uncertainty in the true value of undetected and rounded measurements. Quarterly estimates of mean THM concentrations plus estimates of the water source type (ground, lowland surface or upland surface) were obtained for each water zone. THM concentration estimates were typically highest from July to September (third quarter), and varied considerably between water sources. Our exposure estimates were categorized into 'low', 'medium' and 'high' THM classes. Our modelled quarterly exposure estimates were compared to a simple alternative: annual means of the raw data for each water zone. In all, 15-25% of exposure estimates were classified differently. The modelled THM estimates led to slightly stronger and more precise estimates of association with risk of still birth and low birth weight than did the raw annual means. We conclude that our modelling approach enabled us to provide robust quarterly estimates of ecological exposure to THMs in a situation where the raw data were too sparse to base exposure assessment on empirical summaries alone.

Adult↗
PubMed2005

A comparison of Bayesian spatial models for disease mapping.

With the advent of routine health data indexed at a fine geographical resolution, small area disease mapping studies have become an established technique in geographical epidemiology. The specific issues posed by the sparseness of the data and possibility for local spatial dependence belong to a generic class of statistical problems involving an underlying (latent) spatial process of interest corrupted by observational noise. These are naturally formulated within the framework of hierarchical models, and over the past decade, a variety of spatial models have been proposed for the latent level(s) of the hierarchy. In this article, we provide a comprehensive review of the main classes of such models that have been used for disease mapping within a Bayesian estimation paradigm, and report a performance comparison between representative models in these classes, using a set of simulated data to help illustrate their respective properties. We also consider recent extensions to model the joint spatial distribution of multiple disease or health indicators. The aim is to help the reader choose an appropriate structural prior for the second level of the hierarchical model and to discuss issues of sensitivity to this choice.

Bayes Theorem↗
PubMed2005

Relation of trihalomethane concentrations in public water supplies to stillbirth and birth weight in three water regions in England.

We investigated the association between total trihalomethanes (TTHMs) and risk of stillbirth and low and very low birth weight in three water regions in England, 1992-1998; associations with individual trihalomethanes (THMs) were also examined. Modeled estimates of quarterly TTHM concentrations in water zones, categorized as low (< 30 microg/L), medium (30-59 microg/L), or high (> or = 60 microg/L), were linked to approximately 1 million routine birth and stillbirth records using maternal residence at time of birth. In one region, where there was a positive socioeconomic deprivation gradient across exposure categories, there was also a positive, significant association of TTHM with risk of stillbirth and low and very low birth weight. Overall summary estimates across the three regions using a random-effects model to allow for between-region heterogeneity in exposure effects showed small excess risks in areas with high TTHM concentrations for stillbirths [odds ratio (OR) = 1.11; 95% confidence interval (CI), 1.00-1.23), low birth weight (OR = 1.09; 95% CI, 0.93-1.27), and very low birth weight (OR = 1.05; 95% CI, 0.82-1.34). Among the individual THMs, chloroform showed a similar pattern of risk as TTHM, but no association was found with concentrations of bromodichloromethane or total brominated THMs. Our findings overall suggest a significant association of stillbirths with maternal residence in areas with high TTHM exposure. Further work is needed looking at cause-specific stillbirths and effects of other disinfection by-products and to help differentiate between alternative (noncausal) explanations and those that may derive from the water supply.

Adult↗
PubMed2004

Tracing ingestion of 'novel' foods in UK diets for possible health surveillance--a feasibility study.

OBJECTIVE: To investigate the feasibility of using commercially available data on household food consumption to carry out food and nutritional surveillance. DESIGN: Taylor Nelson Sofres (TNS) collects information on foods brought home for consumption among a representative quota sample of the British population. In total, 33,177 households and 105,667 individuals provided data between 1991 and 2000. These were used to investigate sociodemographic, geographical and temporal trends in purchase patterns of the main macronutrients and four groups of marker products. RESULTS: Sociodemographic characteristics of the TNS sample were broadly consistent with those of the British population. Estimated energy intakes were slightly low (1667+/-715 kcal) in comparison with other national data. However, percentage energy contributions were consistent with national trends: e.g. consumption of alcohol in the home increased between 1991 and 2000 with higher intakes among more affluent households, while fat intakes decreased slightly over the same period. Significant temporal, geographic and socio-economic trends were found for all nutrients (P<0.0001). Intakes of marker products were sparse (purchased by <4% of households), but significant variations were detected in the proportion of households purchasing some or all of the marker products across temporal, geographic and socio-economic strata. CONCLUSIONS: A prospective nutrient surveillance system could be used to trace consumption patterns of foods or nutrients to inform nutritional surveillance. However, existing data sources would require a number of modifications to increase their suitability for such a project. Increasing surveillance to consider ingredients would require the development of a central coding system, with electronically linked barcode, ingredient and nutrient information.

Adolescent↗
PubMed2004

Interpreting posterior relative risk estimates in disease-mapping studies.

There is currently much interest in conducting spatial analyses of health outcomes at the small-area scale. This requires sophisticated statistical techniques, usually involving Bayesian models, to smooth the underlying risk estimates because the data are typically sparse. However, questions have been raised about the performance of these models for recovering the "true" risk surface, about the influence of the prior structure specified, and about the amount of smoothing of the risks that is actually performed. We describe a comprehensive simulation study designed to address these questions. Our results show that Bayesian disease-mapping models are essentially conservative, with high specificity even in situations with very sparse data but low sensitivity if the raised-risk areas have only a moderate (less than 2-fold) excess or are not based on substantial expected counts (> 50 per area). Semiparametric spatial mixture models typically produce less smoothing than their conditional autoregressive counterpart when there is sufficient information in the data (moderate-size expected count and/or high true excess risk). Sensitivity may be improved by exploiting the whole posterior distribution to try to detect true raised-risk areas rather than just reporting and mapping the mean posterior relative risk. For the widely used conditional autoregressive model, we show that a decision rule based on computing the probability that the relative risk is above 1 with a cutoff between 70 and 80% gives a specific rule with reasonable sensitivity for a range of scenarios having moderate expected counts (approximately 20) and excess risks (approximately 1.5- to 2-fold). Larger (3-fold) excess risks are detected almost certainly using this rule, even when based on small expected counts, although the mean of the posterior distribution is typically smoothed to about half the true value.

Bayes Theorem↗
PubMed2003

Following Shipman: a pilot system for monitoring mortality rates in primary care.

As part of the investigations into the crimes of Harold Shipman, it has become clear that there is little monitoring of deaths in general practice. By use of data on annual deaths at family physician and practice level for five English health authorities for 1993-99, we investigate whether cumulative sum charts (a type of statistical process control chart) could be used to create a workable monitoring system. On such charts, thresholds for deaths can be set, which, if crossed, may indicate a potential problem. We chose thresholds based on empirical calculations of the probabilities of false and successful detection after allowing for multiple testing over physicians or practices. We also statistically adjusted the charts for extra-Poisson variation due to unmeasured case mix. Of 1009 family physicians, 33 (including Shipman) crossed the alarm threshold designed to detect a 2 SD increase in standardised mortality, with 97% successful detection and a 5% false-alarm rate. Poor data quality, plus factors such as the proportion of patients treated by these physicians in nursing homes or hospices are likely explanations for most of these additional alarms. If used appropriately, such charts represent a useful tool for monitoring deaths in primary care. However, improvement in data quality is essential.

Data Collection↗
PubMed2002

Geographical epidemiology of prostate cancer in Great Britain.

Prostate cancer incidence has increased during recent years, possibly linked to environmental exposures. Exposure to environmental carcinogens is unlikely to be evenly distributed geographically, which may give rise to variations in disease occurrence that is detectable in a spatial analysis. The aim of our study was to examine the spatial variation of prostate cancer in Great Britain at ages 45-64 years. Spatial variation was examined across electoral wards from 1975-1991. Poisson regression was used to examine regional, urbanisation and socioeconomic effects, while Bayesian mapping techniques were used to assess spatial variability. There was an indication of geographical differences in prostate cancer risk at a regional level, ranging from 0.83 (95% CI: 0.78-0.87) to 1.2 (95% CI: 1.1-1.3) across regions. There was significant heterogeneity in the risk across wards, although the range of relative risks was narrow. More detailed spatial analyses within 4 regions did not indicate any clear evidence of localised geographical clustering for prostate cancer. The absence of any marked geographical variability at a small-area scale argues against a geographically varying environmental factor operating strongly in the aetiology of prostate cancer.

Bayes Theorem↗
PubMed2002

Bayesian analysis of population PK/PD models: general concepts and software.

Markov chain Monte Carlo (MCMC) techniques have revolutionized the field of Bayesian statistics by enabling posterior inference for arbitrarily complex models. The now widely used WinBUGS software has, over the years, made the methodology accessible to a great many applied scientists, in all fields of research. Despite this, serious application of MCMC methods within the field of population PK/PD has been comparatively limited. We appreciate that for many applied pharmacokineticists the prospect of conducting a Bayesian analysis will require numerous alien concepts to be taken on board and it may be difficult to justify investing the time and effort required in order to understand them (especially since the approach is so computer-intensive). For this reason we provide here a thorough (but often informal) discussion of all aspects of Bayesian inference as they apply specifically to population PK/PD. We also acknowledge that while the WinBUGS software is general purpose, model specification for some types of problem, population PK/PD being a prime example, can be very difficult, to the extent that a specialized interface for describing the problem at hand is often a practical necessity. In the latter part of this paper we describe such an interface, namely PKBugs. A principal aim of the paper is to offer sufficient technical background, in an easy to follow format, that the reader may develop both the confidence and know-how to make appropriate use of the PKBugs/WinBUGS framework (or similar software) for their own data analysis needs, should they choose to adopt a Bayesian approach.

Bayes Theorem↗