Ruth Salway

Sources of bias in ecological studies of non-rare events

Ecological studies investigate relationships at the level of the group, rather than at the level of the individual. Although such studies are a common design in epidemiology, it is well-known that estimates may be subject to ecological bias. Most discussion of ecological bias has focused on rare disease events, where the tractability of the loglinear model allows some characterization of the nature of different biases. This paper concentrates on non-rare events, where the Poisson approximation to the binomial distribution is not appropriate. We limit the discussion to bias that arises from within-area variability in exposures and confounders. Our aims are to investigate the likely sizes and directions of bias and, where possible, to suggest methods for controlling the bias or for addressing the sensitivity of inference to assumptions on the nature of the bias. We illustrate that for non-rare events it is much more difficult to characterize the direction of bias than in the rare case. A series of simple numerical examples based on a chronic study of respiratory health illustrate the ideas of the paper.

Estimating exposure response functions using ambient pollution concentrations

This paper presents an approach to estimating the health effects of an environmental hazard. The approach is general in nature, but is applied here to the case of air pollution. It uses a computer model involving ambient pollution and temperature input to simulate the exposures experienced by individuals in an urban area, while incorporating the mechanisms that determine exposures. The output from the model comprises a set of daily exposures for a sample of individuals from the population of interest. These daily exposures are approximated by parametric distributions so that the predictive exposure distribution of a randomly selected individual can be generated. These distributions are then incorporated into a hierarchical Bayesian framework (with inference using Markov chain Monte Carlo simulation) in order to examine the relationship between short-term changes in exposures and health outcomes, while making allowance for long-term trends, seasonality, the effect of potential confounders and the possibility of ecological bias. The paper applies this approach to particulate pollution (PM 10 ) and respiratory mortality counts for seniors in greater London (≥65 years) during 1997. Within this substantive epidemiological study, the effects on health of ambient concentrations and (estimated) personal exposures are compared. The proposed model incorporates within day (or between individual) variability in personal exposures, which is compared to the more traditional approach of assuming a single pollution level applies to the entire population for each day. Effects were estimated using single lags and distributed lag models, with the highest relative risk, RR = 1.02 (1.01-1.04), being associated with a lag of two days ambient concentrations of PM 10 . Individual exposures to PM 10 for this group (seniors) were lower than the measured ambient concentrations with the corresponding risk, RR = 1.05 (1.01-1.09), being higher than would be suggested by the traditional approach using ambient concentrations.

The freeze on mass media campaigns in England: a natural experiment of the impact of tobacco control campaigns on quitting behaviour.

AIMS To measure the impact of the suspension of tobacco control mass media campaigns in England in April 2010 on measures of smoking cessation behaviour. DESIGN Interrupted time series design using routinely collected population-level data. Analysis of use of a range of types of smoking cessation support using segmented negative binomial regression. SETTING England. MEASUREMENTS Use of non-intensive support: monthly calls to the National Health Service (NHS) quitline (April 2005-September 2011), text requests for quit support packs (December 2007-10) and web hits on the national smoking cessation website (January 2009-March 2011). Use of intensive cessation support: quarterly data on the number of people setting a quit date and 4-week quitters at the NHS Stop Smoking Services (SSS) (quarter 1, 2001 and quarter 3, 2011). FINDINGS During the suspension of tobacco control mass media spending, literature requests fell by 98% [95% confidence interval (CI) = 96-99], and quitline calls and web hits fell by 65% (95% CI = 43-79) and 34% (95% CI: 11-50), respectively. The number of people setting a quit date and 4-week quitters at the SSS increased throughout the study period. CONCLUSIONS The suspension of tobacco control mass media campaigns in England in 2012 appeared to markedly reduce the use of smoking cessation literature, quitline calls and hits on the national smoking cessation website, but did not affect attendance at the Stop Smoking Services. Within a comprehensive tobacco control programme, mass media campaigns can play an important role in maximizing quitting activity.

Large-scale Bayesian spatial modelling of air pollution for policy support

The potential effects of air pollution are a major concern both in terms of the environment and in relation to human health. In order to support environmental policy, there is a need for accurate measurements of the concentrations of pollutants at high geographical resolution over large regions. However, within such regions, there are likely to be areas where the monitoring information will be sparse and so methods are required to accurately predict concentrations. Set within a Bayesian framework, models are developed which exploit the relationships between pollution and geographical covariate information, such as land use, climate and transport variables together with spatial structure. Candidate models are compared based on their ability to predict a set of validation sites. The chosen model is used to perform large-scale prediction of nitrogen dioxide at a 1×1 km resolution for the entire EU. The models allow probabilistic statements to be made with regard to the levels of air pollution that might be experienced in each area. When combined with population data, such information can be invaluable in informing policy by indicating areas for which improvements may be given priority.

A common framework for ecological inference in epidemiology, political science and sociology

Ecological studies arise within many different disciplines. This chapter describes common approaches to ecological inference in an environmental epidemiology setting, and compares these with traditional approaches in political science and sociology. These approaches vary considerably, both in their use of terminology and notation, and in the relative importance of the various issues that make ecological analyses problematic. The aims of this chapter are twofold. Firstly, we describe ecological inference in an epidemiology setting, where the interest is in the relationship between disease status and exposure to some potential risk factor. We concentrate on those issues which are of particular concern in epidemiology, for example the presence of additional (possibly unmeasured) covariates, termed confounders. Secondly, we seek to unite the current work in epidemiology, political science, and sociology by clarifying differences in terminology, by describing commonly used approaches within a common statistical framework, and by highlighting similarities and differences between these approaches. Often different models can be attributed to different sets of underlying assumptions; we emphasize that such assumptions are crucial in the conclusions drawn from ecological data, and their appropriateness should be carefully considered in any specific situation. Combining approaches from all three disciplines gives a broad range of possible assumptions and available techniques from which to choose.

Bayesian latent variable modelling in studies of air pollution and health

This paper describes the use of Bayesian latent variable models in the context of studies investigating the short-term effects of air pollution on health. Traditional Poisson or quasi-likelihood regression models used in this area assume that consecutive outcomes are independent (although the latter allows for overdispersion), which in many studies may be an untenable assumption as temporal correlation is to be expected. We compare this traditional approach with two Bayesian latent process models, which acknowledge the possibility of short-term autocorrelation. These include an autoregressive model that has previously been used in air pollution studies and an alternative based on a moving average structure that we describe here. A simulation study assesses the performance of these models when there are different forms of autocorrelation in the data. Although estimated risks are largely unbiased, the results show that assuming independence can produce confidence intervals that are too narrow. Failing to account for the additional uncertainty which may be associated with (positive) correlation can result in confidence/credible intervals being too narrow and thus lead to incorrect conclusions being made about the significance of estimated risks. The methods are illustrated within a case study of the effects of short-term exposure to air pollution on respiratory mortality in the elderly in London, between 1997 and 2003.

Interpreting Long-Term Trends in Time Series Intervention Studies of Smoke-Free Legislation and Health

Background: Numerous studies have investigated the impact of smoke-free laws on health outcomes. Large differences in estimates are in part attributable to how the long-term trend is modelled. However, the choice of appropriate trend is not always straightforward. We explore these complexities in an analysis of myocardial infarction (MI) mortality in England before and after the introduction of smoke-free legislation in July 2007. Methods: Weekly rates of MI mortality among men aged 40+ between July 2002 and December 2010 were analysed using quasi-Poisson generalised additive models. We explore two ways of modelling the long-term trend: (1) a parametric approach, where we fix the shape of the trend, and (2) a penalised spline approach, in which we allow the model to decide on the shape of the trend. Results: While both models have similar measures of fit and near identical fitted values, they have different interpretations of the legislation effect. The parametric approach estimates a significant immediate reduction in mortality rate of 13.7% (95% CI: 7.5, 19.5), whereas the penalised spline approach estimates a non-significant reduction of 2% (95% CI:-0.9, 4.8). After considering the implications of the models, evidence from sensitivity analyses and other studies, we conclude that the second model is to be preferred. Conclusions: When there is a strong long-term trend and the intervention of interest also varies over time, it is difficult for models to separate out the two components. Our recommendations will help further studies determine the best way of modelling their data.

The effectiveness of mass media campaigns in reducing smoking in England: an observational study

Abstract Background Little evidence exists for the effectiveness of mass media campaigns in the UK despite campaigns running regularly since 1999. In 2010, the UK Government ceased spending on national public health mass media campaigns in England. Tobacco control (TC) campaigns were reintroduced in 2011, but at a lower level of funding. We need to know whether campaigns are effective to ensure continued funding. We aimed to assess the effect of government-funded televised campaigns for TC shown in England during the 2000s on adult smoking behaviours. Methods We measured exposure to government-funded television advertisements for TC using television ratings (TVRs) and coded the advertisements for emotional content using a theoretically based framework. Recall of advertisements from 2005 to 2009 was ascertained from participants of the International Tobacco Four Country Survey and analysed with generalised estimating equations, comparing recall between advertisements with positive and negative emotional content. Data for smoking prevalence and cigarette consumption from 2002 to 2010 were obtained from Opinions and Lifestyle surveys. Measures of smoking cessation behaviour (calls to the NHS quitline 2005–11, text requests for quit support packs 2007–10, smoking cessation website hits 2009–11, attendance and 4-week quitters at NHS Stop Smoking Services [SSS] 2001–11) were obtained from the Department of Health and NHS Information Centre. Data were analysed with generalised additive models with adjustments for seasonality, time trends, and other tobacco policies. Findings Between 2002 and April, 2010, 35 282 TVRs for TC were broadcast (mean 42 advertisements per year). An increase in TVRs was associated with a significant increase in recall and a reduction in mean cigarette consumption in the following month and smoking prevalence 2 months later (appendix). 45% of advertisements screened between 2004 and 2010 were coded as having positive emotional content. Negative, but not positive, emotive campaigns resulted in significantly increased and longer durations of recall (appendix). There was a significant decrease in text requests for quit support packs, quitline calls, and web hits after April, 2010, but no effect on the use of the NHS SSS (appendix). We studied recall and tobacco use at the individual level; however, TVRs measure mean potential exposure, and individual-level exposure can vary dependent on frequency of television viewing and attention to advertisements. We cannot completely rule out that other unmeasured variables confound the association with TVRs or the timing of the freeze. Interpretation Televised campaigns are effective in changing smoking behaviours in adults. We are now exploring whether the stronger effect of negative emotive campaigns on recall translates to effects on quitting behaviour. Funding Universities of Bath, Nottingham, and Kings College London, received funding from the National Prevention Research Initiative (www.mrc.ac.uk/npr). Funding partners relevant to this award are: Alzheimers Research Trust; Alzheimers Society; Biotechnology and Biological Sciences Research Council; British Heart Foundation (BHF); Cancer Research UK (CRUK); Chief Scientist Office, Scottish Government Health Directorate; Department of Health; Diabetes UK; Economic and Social Research Council (ESRC); Health and Social Care Research and Development Division of the Public Health Agency; MRC; The Stroke Association; Wellcome Trust; and Welsh Assembly Government. All authors are members of the UK Centre for TC Studies. Funding from the BHF, CRUK, ESRC, MRC, and the NIHR, under the auspices of the UK Clinical Research Collaboration, is gratefully acknowledged. TC-4 was supported by the US National Cancer Institute; Canadian Institutes of Health Research; National Health and Medical Research Council of Australia; CRUK; Ontario Institute for Cancer Research; and Canadian Cancer Society Research Institute through awards to Geoffrey T Fong.

The freeze on mass media campaigns in England: a natural experiment of the impact of tobacco control campaigns on quitting behaviour: Freeze on mass media campaigns

AIMS To measure the impact of the suspension of tobacco control mass media campaigns in England in April 2010 on measures of smoking cessation behaviour. DESIGN Interrupted time series design using routinely collected population-level data. Analysis of use of a range of types of smoking cessation support using segmented negative binomial regression. SETTING England. MEASUREMENTS Use of non-intensive support: monthly calls to the National Health Service (NHS) quitline (April 2005-September 2011), text requests for quit support packs (December 2007-10) and web hits on the national smoking cessation website (January 2009-March 2011). Use of intensive cessation support: quarterly data on the number of people setting a quit date and 4-week quitters at the NHS Stop Smoking Services (SSS) (quarter 1, 2001 and quarter 3, 2011). FINDINGS During the suspension of tobacco control mass media spending, literature requests fell by 98% [95% confidence interval (CI) = 96-99], and quitline calls and web hits fell by 65% (95% CI = 43-79) and 34% (95% CI: 11-50), respectively. The number of people setting a quit date and 4-week quitters at the SSS increased throughout the study period. CONCLUSIONS The suspension of tobacco control mass media campaigns in England in 2012 appeared to markedly reduce the use of smoking cessation literature, quitline calls and hits on the national smoking cessation website, but did not affect attendance at the Stop Smoking Services. Within a comprehensive tobacco control programme, mass media campaigns can play an important role in maximizing quitting activity.

The freeze on mass media campaigns in England

AIMS To measure the impact of the suspension of tobacco control mass media campaigns in England in April 2010 on measures of smoking cessation behaviour. DESIGN Interrupted time series design using routinely collected population-level data. Analysis of use of a range of types of smoking cessation support using segmented negative binomial regression. SETTING England. MEASUREMENTS Use of non-intensive support: monthly calls to the National Health Service (NHS) quitline (April 2005-September 2011), text requests for quit support packs (December 2007-10) and web hits on the national smoking cessation website (January 2009-March 2011). Use of intensive cessation support: quarterly data on the number of people setting a quit date and 4-week quitters at the NHS Stop Smoking Services (SSS) (quarter 1, 2001 and quarter 3, 2011). FINDINGS During the suspension of tobacco control mass media spending, literature requests fell by 98% [95% confidence interval (CI) = 96-99], and quitline calls and web hits fell by 65% (95% CI = 43-79) and 34% (95% CI: 11-50), respectively. The number of people setting a quit date and 4-week quitters at the SSS increased throughout the study period. CONCLUSIONS The suspension of tobacco control mass media campaigns in England in 2012 appeared to markedly reduce the use of smoking cessation literature, quitline calls and hits on the national smoking cessation website, but did not affect attendance at the Stop Smoking Services. Within a comprehensive tobacco control programme, mass media campaigns can play an important role in maximizing quitting activity.