Emerging Themes in Epidemiology - Latest Articles

Improving epidemiologic data analyses through multivariate regression modelling

Fraser Lewis — 2013-05-17T00:00:00Z

Regression modelling is one of the most widely utilized approaches in epidemiological analyses. It provides a method of identifying statistical associations, from which potential causal associations relevant to disease control may then be investigated. Multivariable regression - a single dependent variable (outcome, usually disease) with multiple independent variables (predictors) - has long been the standard model. Generalizing multivariable regression to multivariate regression - all variables potentially statistically dependent - offers a far richer modelling framework. Through a series of simple illustrative examples we compare and contrast these approaches. The technical methodology used to implement multivariate regression is well established - Bayesian network structure discovery - and while a relative newcomer to the epidemiological literature has a long history in computing science. Applications of multivariate analysis in epidemiological studies can provide a greater understanding of disease processes at the population level, leading to the design of better disease control and prevention programs.

Trends in parameterization, economics and host behaviour in influenza pandemic modelling: a review and reporting protocol

Luis Carrasco — 2013-05-07T00:00:00Z

Background: The volume of influenza pandemic modelling studies has increased dramatically in the last decade. Many models incorporate now sophisticated parameterization and validation techniques, economic analyses and the behaviour of individuals. Methods: We reviewed trends in these aspects in models for influenza pandemic preparedness that aimed to generate policy insights for epidemic management and were published from 2000 to September 2011, i.e. before and after the 2009 pandemic. Results: We find that many influenza pandemics models rely on parameters from previous modelling studies, models are rarely validated using observed data and are seldom applied to low-income countries. Mechanisms for international data sharing would be necessary to facilitate a wider adoption of model validation. The variety of modelling decisions makes it difficult to compare and evaluate models systematically. Conclusions: We propose a model Characteristics, Construction, Parameterization and Validation aspects protocol (CCPV protocol) to contribute to the systematisation of the reporting of models with an emphasis on the incorporation of economic aspects and host behaviour. Model reporting, as already exists in many other fields of modelling, would increase confidence in model results, and transparency in their assessment and comparison.

Estimating HIV prevalence from surveys with low individual consent rates: annealing individual and pooled samples

Lauren Hund — 2013-02-27T00:00:00Z

Many HIV prevalence surveys are plagued by the problem that a sizeable number of surveyed individuals do not consent to contribute blood samples for testing. One can ignore this problem, as is often done, but the resultant bias can be of sufficient magnitude to invalidate the results of the survey, especially if the number of non-responders is high and the reason for refusing to participate is related to the individual’s HIV status. One reason for refusing to participate may be for reasons of privacy. For those individuals, we suggest offering the option of being tested in a pool. This form of testing is less certain than individual testing, but, if it convinces more people to submit to testing, it should reduce the potential for bias and give a cleaner answer to the question of prevalence. This paper explores the logistics of implementing a combined individual and pooled testing approach and evaluates the analytical advantages to such a combined testing strategy. We quantify improvements in a prevalence estimator based on this combined testing strategy, relative to an individual testing only approach and a pooled testing only approach. Minimizing non-response is key for reducing bias, and, if pooled testing assuages privacy concerns, offering a pooled testing strategy has the potential to substantially improve HIV prevalence estimates.

Why is greater medication adherence associated with better outcomes

Arthur Hartz — 2013-02-02T00:00:00Z

Background: Previous studies found an association of greater adherence to placebo medication with better outcomes. The present study tested whether this association was explained by any of the following factors: 1) adherence to other medications, 2) healthcare behaviors, 3) disease risk, or 4) predicted degree of adherence. Data included information on more than 800 risk factors from 27,347 subjects in two randomized controlled trials of hormone therapy in the Women's Health Initiative. Results: Greater adherence to placebo was not associated with colon cancer but was substantially and significantly associated with several diverse outcomes: death, myocardial infarction, stroke, and breast cancer. Adherence to hormone therapy was only weakly associated with outcomes. The WHI risk factors only poorly predicted degree of adherence, R2 < 4%. No underlying factors accounted for the association between placebo adherence and outcome. Conclusion: The results suggest that adherence to placebo is a marker for important risk factors that were not measured by WHI. Once identified these risk factors may be used to increase the validity of observational studies of medical treatment by reducing unmeasured confounding.

A tutorial in estimating the prevalence of disease in humans and animals in the absence of a gold standard diagnostic

Fraser Lewis — 2012-12-28T00:00:00Z

Epidemiological methods for estimating disease prevalence in humans and other animals in the absence of a gold standard diagnostic test are well established. Despite this, reporting apparent prevalence is still standard practice in public health studies and disease control programmes, even though apparent prevalence may differ greatly from the true prevalence of disease. Methods for estimating true prevalence are summarized and reviewed. A computing appendix is also provided which contains a brief guide in how to easily implement some of the methods presented using freely available software.

The validation of a three-stage screening methodology for detecting active convulsive epilepsy in population-based studies in health and demographic surveillance systems

Anthony Ngugi — 2012-11-21T00:00:00Z

Background: There are few studies on the epidemiology of epilepsy in large populations in Low and Middle Income Countries (LMIC). Most studies in these regions use two-stage population-based screening surveys, which are time-consuming and costly to implement in large populations required to generate accurate estimates. We examined the sensitivity and specificity of a three-stage cross-sectional screening methodology in detecting active convulsive epilepsy (ACE), which can be embedded within on-going census of demographic surveillance systems.We validated a three-stage cross-sectional screening methodology on a randomly selected sample of participants of a three-stage prevalence survey of epilepsy. Diagnosis of ACE by an experienced clinician was used as ‘gold standard’. We further compared the expenditure of this method with the standard two-stage methodology. Results: We screened 4442 subjects in the validation and identified 35 cases of ACE. Of these, 18 were identified as false negatives, most of whom (15/18) were missed in the first stage and a few (3/18) in the second stage of the three-stage screening. Overall, this methodology had a sensitivity of 48.6% and a specificity of 100%. It was 37% cheaper than a two-stage survey. Conclusion: This was the first study to evaluate the performance of a multi-stage screening methodology used to detect epilepsy in demographic surveillance sites. This method had poor sensitivity attributed mainly to stigma-related non-response in the first stage. This method needs to take into consideration the poor sensitivity and the savings in expenditure and time as well as validation in target populations. Our findings suggest the need for continued efforts to develop and improve case-ascertainment methods in population-based epidemiological studies of epilepsy in LMIC.

The effect of improved rural sanitation on diarrhoea and helminth infection: design of a cluster-randomized trial in Orissa, India

Thomas Clasen — 2012-11-13T00:00:00Z

Background: Infectious diseases associated with poor sanitation such as diarrhoea, intestinal worms, trachoma and lymphatic filariasis continue to cause a large disease burden in low income settings and contribute substantially to child mortality and morbidity. Obtaining health impact data for rural sanitation campaigns poses a number of methodological challenges. Here we describe the design of a village-level cluster-randomised trial in the state of Orissa, India to evaluate the impact of an ongoing rural sanitation campaign conducted under the umbrella of India’s Total Sanitation Campaign (TSC).We randomised 50 villages to the intervention and 50 villages to control. In the intervention villages the implementing non-governmental organisations conducted community mobilisation and latrine construction with subsidies given to poor families. Control villages receive no intervention. Outcome measures include (1) diarrhoea in children under 5 and in all ages, (2) soil-transmitted helminth infections, (3) anthropometric measures, (4) water quality, (5) number of insect vectors (flies, mosquitoes), (6) exposure to faecal pathogens in the environment. In addition we are conducting process documentation (latrine construction and use, intervention reach), cost and cost-effectiveness analyses, spatial analyses and qualitative research on gender and water use for sanitation. Results: Randomisation resulted in an acceptable balance between trial arms. The sample size requirements appear to be met for the main study outcomes. Delays in intervention roll-out caused logistical problems especially for the planning of health outcome follow-up surveys. Latrine coverage at the end of the construction period (55%) remained below the target of 70%, a result that may, however, be in line with many other TSC intervention areas in India. Conclusion: We discuss a number of methodological problems encountered thus far in this study that may be typical for sanitation trials. Nevertheless, it is expected that the trial procedures will allow measuring the effectiveness of a typical rural sanitation campaign, with sufficient accuracy and validity.

Performance of small cluster surveys and the clustered LQAS design to estimate local-level vaccination coverage in Mali

Andrea Minetti — 2012-10-12T00:00:00Z

Background: Estimation of vaccination coverage at the local level is essential to identify communities that may require additional support. Cluster surveys can be used in resource-poor settings, when population figures are inaccurate. To be feasible, cluster samples need to be small, without losing robustness of results. The clustered LQAS (CLQAS) approach has been proposed as an alternative, as smaller sample sizes are required. Methods: We explored (i) the efficiency of cluster surveys of decreasing sample size through bootstrapping analysis and (ii) the performance of CLQAS under three alternative sampling plans to classify local VC, using data from a survey carried out in Mali after mass vaccination against meningococcal meningitis group A. Results: VC estimates provided by a 10 × 15 cluster survey design were reasonably robust. We used them to classify health areas in three categories and guide mop-up activities: i) health areas not requiring supplemental activities; ii) health areas requiring additional vaccination; iii) health areas requiring further evaluation. As sample size decreased (from 10 × 15 to 10 × 3), standard error of VC and ICC estimates were increasingly unstable. Results of CLQAS simulations were not accurate for most health areas, with an overall risk of misclassification greater than 0.25 in one health area out of three. It was greater than 0.50 in one health area out of two under two of the three sampling plans. Conclusions: Small sample cluster surveys (10 × 15) are acceptably robust for classification of VC at local level. We do not recommend the CLQAS method as currently formulated for evaluating vaccination programmes.

Choosing a survey sample when data on the population are limited: a method using Global Positioning Systems and aerial and satellite photographs

Harry Shannon — 2012-09-11T00:00:00Z

Background: Various methods have been proposed for sampling when data on the population are limited. However, these methods are often biased. We propose a new method to draw a population sample using Global Positioning Systems and aerial or satellite photographs. Results: We randomly sampled Global Positioning System locations in designated areas. A circle was drawn around each location with radius representing 20 m. Buildings in the circle were identified from satellite photographs; one was randomly chosen. Interviewers selected one household from the building, and interviews were conducted with eligible household members. Conclusions: Participants had known selection probabilities, allowing proper estimation of parameters of interest and their variances. The approach was made possible by recent technological developments and access to satellite photographs.

Event-based Internet Biosurveillance: Relation to Epidemiological Observation

Noele Nelson — 2012-06-18T00:00:00Z

Background: The World Health Organization (WHO) collects and publishes surveillance data and statistics for select diseases, but traditional methods of gathering such data are time and labor intensive. Event-based biosurveillance, which utilizes a variety of Internet sources, complements traditional surveillance. In this study we assess the reliability of Internet biosurveillance and evaluate disease-specific alert criteria against epidemiological data. Methods: We reviewed and compared WHO epidemiological data and Argus biosurveillance system data for pandemic (H1N1) 2009 (April 2009 – January 2010) from 8 regions and 122 countries to: identify reliable alert criteria among 15 Argus-defined categories; determine the degree of data correlation for disease progression; and assess timeliness of Internet information. Results: Argus generated a total of 1,580 unique alerts; 5 alert categories generated statistically significant (p < 0.05) correlations with WHO case count data; the sum of these 5 categories was highly correlated with WHO case data (r = 0.81, p < 0.0001), with expected differences observed among the 8 regions. Argus reported first confirmed cases on the same day as WHO for 21 of the first 64 countries reporting cases, and 1 to 16 days (average 1.5 days) ahead of WHO for 42 of those countries. Conclusion: Confirmed pandemic (H1N1) 2009 cases collected by Argus and WHO methods returned consistent results and confirmed the reliability and timeliness of Internet information. Disease-specific alert criteria provide situational awareness and may serve as proxy indicators to event progression and escalation in lieu of traditional surveillance data; alerts may identify early-warning indicators to another pandemic, preparing the public health community for disease events.