Search

Background: In large-scale epidemiology, bloodspot sampling by fingerstick onto filter paper has many advantages, including ease and low costs of collection, processing and transport. We describe the development of an enzyme-linked immunoassay (ELISA) for quantifying insulin from dried blood spots and demonstrate its application in a large trial. Methods: We adapted an existing commercial kit (Mercodia Human Insulin ELISA, 10-1113-01) to quantify insulin from two 3-mm diameter discs (≈6 µL of blood) punched from whole blood standards and from trial samples. Paediatricians collected dried blood spots in a follow-up of 13,879 fasted children aged 11.5 years (interquartile range 11.3–11.8 years) from 31 trial sites across Belarus. We quantified bloodspot insulin levels and examined their distribution by demography and anthropometry. Results: Mean intra-assay (n = 157) coefficients of variation were 15% and 6% for ‘low’ (6.7 mU/L) and ‘high’ (23.1 mU/L) values, respectively; the respective inter-assay values (n = 33) were 23% and 11%. The intraclass correlation coefficient between 50 paired whole bloodspot versus serum samples, collected simultaneously, was 0.90 (95% confidence interval 0.85 to 0.95). Bloodspot insulin was stable for at least 31 months at −80°C, for one week at +30°C and following four freeze-thaw cycles. Paediatricians collected a median of 8 blood spots from 13,487 (97%) children. The geometric mean insulin (log standard deviation) concentrations amongst 12,812 children were 3.0 mU/L (1.1) in boys and 4.0 mU/L (1.0) in girls and were positively associated with pubertal stage, measures of central and peripheral adiposity, height and fasting glucose. Conclusions: Our simple and convenient bloodspot assay is suitable for the measurement of insulin in very small volumes of blood collected on filter paper cards and can be applied to large-scale epidemiology studies of the early-life determinants of circulating insulin.

Background: This paper aims to instigate discussion and publication of methodologies applied to enhance quality management through comprehensive scientific reports. It provides a detailed description of the design, implementation, and results of the quality control program employed in the SMESH study. Methods: Cross-sectional, multicenter, national study designed to assess the prevalence of human papillomavirus in sex workers and in men who have sex with men (MSM). Respondent-driven sampling recruitment was used. An online system was developed for the study and checkpoints were defined for data entry. The system checked the quality of biological samples and performed a retest with part of the sample. Results: A total of 1.598 participants (442 sex workers and 1.156 MSM) were included. Fifty-four health professionals were trained for face-to-face data collection. The retest showed Kappa values ranging between 0.3030 and 0.7663. Conclusion: The retest data were mostly classified as indicating a strong association. The data generated by the checkpoints showed the successful implementation of the quality control program. [ABSTRACT FROM AUTHOR]

Objective: A new method known as 'current-day dietary recall' (current-day recall) is based on an application for mobile phones called 'electronic 12 h dietary recall' (e-12HR). This new method was designed to rank participants into categories of habitual intake regarding a series of key food groups. The present study compared current-day recall against a previously validated short paper FFQ.Design: Participants recorded the consumption of selected food groups using e-12HR during twenty-eight consecutive days and then filled out a short paper FFQ at the end of the study period. To evaluate the association and agreement between both methods, Spearman's correlation coefficients (SCC), cross-classification analysis and weighted kappa statistics (κ w) were used.Setting: Andalusia, Spain, Southern Europe.Subjects: University students and employees over the age of 18 years.Results: One hundred and eighty-seven participants completed the study (64·2 % female, 35·8 % male). For all particpants, for all food group intakes, the mean SCC was 0·70 (SCC≥0·62 were observed for all strata); the mean percentage of participants cross-classified into categories of 'exact agreement+adjacent' was 90·1 % (percentages≥87·8 % were observed for all strata); and the mean κ w was 0·55 (κ w≥0·53 in ten of the twelve strata).Conclusions: For the whole sample and for all strata thereof, the current-day recall has good agreement with the previously validated short paper FFQ for assessing food group intakes, rendering it a useful method for ranking individuals. [ABSTRACT FROM AUTHOR]

Introduction: Respondent-driven sampling (RDS) offers a recruitment strategy for hard-to-reach populations. However, RDS faces logistical and theoretical challenges that threaten efficiency and validity in settings worldwide. We present innovative adaptations to conventional RDS to overcome barriers encountered in recruiting a large, representative sample of men who have sex with men (MSM) who travel internationally. Methods: Novel methodological adaptations for the “International Travel Research to Inform Prevention” or “I-TRIP” study were offering participants a choice between electronic and paper coupons referrals for recruitment and modifying the secondary incentives structure from small cash amounts to raffle entries for periodic large cash prize raffle drawings. Staged referral limit increases from 3 to 10 referrals and progressive addition of 70 seeds were also implemented. Results: There were 501 participants enrolled in up to 13 waves of growth. Among participants with a choice of referral methods, 81% selected electronic referrals. Of participants who were recruited electronically, 90% chose to remain with electronic referrals when it was their turn to recruit. The mean number of enrolled referrals was 0.91 for electronic referrals compared to 0.56 for paper coupons. Median referral lag time, i.e., the time interval between when recruiters were given their referrals and when a referred individual enrolled in the study, was 20 days (IQR 10–40) for electronic referrals, 20 days (IQR 8–58) for paper coupons, 20 days (IQR 10–41) for raffle entries and 33 days (IQR 16–148) for small cash incentives. Conclusions: The recruitment of MSM who travel internationally required maximizing known flexible tools of RDS while at the same time necessitating innovations to increase recruitment efficiency. Electronic referrals emerged as a major advantage in recruiting this hard-to-reach population who are of high socio-economic status, geographically diffuse and highly mobile. These enhancements may improve the performance of RDS in target populations with similar characteristics. [ABSTRACT FROM AUTHOR]

Objective: To develop and evaluate a pictorial, web-based version of the NCI diet history questionnaire (Web-PDHQ). Design: The Web-PDHQ and paper version of the DHQ (Paper-DHQ) were administered 4 weeks apart with 218 participants randomised to order. Dietary data from the Web-PDHQ and Paper-DHQ were validated using a randomly selected 4 d food record recording period (including a weekend day) and two randomly selected 24 h dietary recalls during the 4 weeks intervening between these two diet history administrations. Setting: Research office in Reston, VA, USA. Participants: Computer-literate men and women recruited through newspaper advertisements. Results: Mean correlation of energy and the twenty-five examined nutrients between the Web-PDHQ and Paper-DHQ was 0∙71 and 0∙51, unadjusted and energy-adjusted by the residual method, respectively. Moderate mean correlations (unadjusted 0∙41 and 0∙38; energy-adjusted 0∙41 and 0∙34) were obtained between both the Web-PDHQ and Paper-DHQ with the 4 d food record on energy and nutrients, but the correlations between the Web-PDHQ and Paper-DHQ with the 24 h recalls were modest (unadjusted 0∙31 and 0∙29; energy-adjusted 0∙37 and 0∙26). A subset of participants (n 48) completing the Web-PDHQ at the initial visit performed a retest on the same questionnaire 1 week later to determine repeatability, and the unadjusted mean correlation was 0∙82. Conclusions: These data indicate that the Web-PDHQ has comparable repeatability and validity to the Paper-DHQ but did not improve the relationship of the DHQ to other food intake measures (e.g. food records, 24 h recall). [ABSTRACT FROM AUTHOR]

Most epidemiological studies examine how risk factors relate to average difference in outcomes (linear regression) or odds of a binary outcome (logistic regression); they do not explicitly examine whether risk factors are associated differentially across the distribution of the health outcome investigated. This paper documents a phenomenon found repeatedly in the minority of epidemiological studies which do this (via quantile regression): associations between a range of established risk factors and body mass index (BMI) are progressively stronger in the upper ends of the BMI distribution. In this paper, we document this finding and provide illustrative evidence of it in the 1958 British birth cohort study. Associations of low childhood socio-economic position, high maternal weight, low childhood general cognition and adult physical inactivity with higher BMI are larger at the upper end of the BMI distribution, on both absolute and relative scales. For example, effect estimates for socio-economic position and childhood cognition were around three times larger at the 90th compared with 10th quantile, while effect estimates for physical inactivity were increasingly larger from the 50th to 90th quantiles, yet null at lower quantiles. We provide potential explanations for these findings and discuss implications. Risk factors may have larger causal effects among those in worse health, and these effects may not be discovered when health is only examined in average terms. In such scenarios, population-based approaches to intervention may have larger benefits than anticipated when assuming equivalent benefit across the population. Further research is needed to understand why effect estimates differ across the BMI outcome distribution and to investigate whether differential effects exist for other physical and mental health outcomes. [ABSTRACT FROM AUTHOR]

An investigation in a randomized population-based Swedish study with 564 participants aged 18-80 years showed that mean physical activity levels obtained using short message service (SMS) by means of cell phones (n = 171) were equal to corresponding levels obtained when sending identical questions by web (n = 182) or paper (n = 211). The response rates were similar for the SMS, web and paper groups. [ABSTRACT FROM AUTHOR]

Objectives: To report some of the most serious omissions and errors which may occur in papers submitted to Occupational and Environmental Medicine, and to give guidelines on the essential components that should be included in papers reporting results from studies of occupational and environmental health.Methods: Since 1994 Occupational and Environmental Medicine has used a panel of medical statisticians to review submitted papers which have a substantial statistical content. Although some studies may have genuine errors in their design, execution, and analysis, many of the problems identified during the reviewing process are due to inadequate and incomplete reporting of essential aspects of a study. This paper outlines some of the most important errors and omissions that may occur. Observational studies are often the preferred choice of design in occupational and environmental medicine. Some of the issues relating to design, execution, and analysis which should be considered when reporting three of the most common observational study designs, cross sectional, case-control, and cohort are described. An illustration of good reporting practice is given for each. Various mathematical modelling techniques are often used in the analysis of these studies, the reporting of which causes a major problem to some authors. Suggestions for the presentation of results from modelling are made.Conclusions: There is increasing interest in the development and application of formal "good epidemiology practices". These not only consider issues of data quality, study design, and study conduct, but through their structured approach to the documentation of the study procedures, provide the potential for more rigorous reporting of the results in the scientific literature. [ABSTRACT FROM AUTHOR]

In a stratified analysis, the results from different strata if homogeneity assumption is met are pooled together to obtain a single summary estimate for the common effect parameter. However, the effect can appear homogeneous across strata using one measure but heterogeneous using another. Consequently, two researchers analyzing the same data can arrive at conflicting conclusions if they use different effect measures. In this paper, the author draws on the sufficient component cause model to develop a stratified-analysis method regarding a particular effect measure, the ‘peril ratio’. When there is no mechanistic interaction between the exposure under study and the stratifying variable (i.e., when they do not work together to complete any sufficient cause), the peril ratio is constant across strata. The author presents formulas for the estimation of such a common peril ratio. Three real data are re-analyzed for illustration. When the data is consistent with peril-ratio homogeneity in a stratified analysis, researchers can use the formulas in this paper to pool the strata. [ABSTRACT FROM AUTHOR]

Background: The SHAMISEN (Nuclear Emergency Situations - Improvement of Medical And Health Surveillance) European project was conducted in 2015-2017 to review the lessons learned from the experience of past nuclear accidents and develop recommendations for preparedness and health surveillance of populations affected by a nuclear accident. Using a toolkit approach, Tsuda et al. recently published a critical review of the article by Cléro et al. derived from the SHAMISEN project on thyroid cancer screening after nuclear accident. Main body: We address the main points of criticism of our publication on the SHAMISEN European project. Conclusion: We disagree with some of the arguments and criticisms mentioned by Tsuda et al. We continue to support the conclusions and recommendations of the SHAMISEN consortium, including the recommendation not to launch a mass thyroid cancer screening after a nuclear accident, but rather to make it available (with appropriate information counselling) to those who request it. [ABSTRACT FROM AUTHOR]

We explore the different types of causes that are commonly investigated by epidemiologists. We first distinguish between causes which are events (including actions) and causes which are states. Second, we distinguish between modifiable and non-modifiable states. This yields three types of causes: fixed states (non-modifiable), dynamic states (modifiable) and events (including actions). Different causes may have different characteristics: the methods available to study them, the types of possible biases, and therefore the types of evidence needed to infer causality, may differ according to the specific cause-effect relationship under study. Nevertheless, there are also substantial commonalities. This paper is intended to improve understanding of the different types of causes, and the different types of causality, that underpin epidemiological practice. [ABSTRACT FROM AUTHOR]

Abstract Background This paper aims to instigate discussion and publication of methodologies applied to enhance quality management through comprehensive scientific reports. It provides a detailed description of the design, implementation, and results of the quality control program employed in the SMESH study. Methods Cross-sectional, multicenter, national study designed to assess the prevalence of human papillomavirus in sex workers and in men who have sex with men (MSM). Respondent-driven sampling recruitment was used. An online system was developed for the study and checkpoints were defined for data entry. The system checked the quality of biological samples and performed a retest with part of the sample. Results A total of 1.598 participants (442 sex workers and 1.156 MSM) were included. Fifty-four health professionals were trained for face-to-face data collection. The retest showed Kappa values ranging between 0.3030 and 0.7663. Conclusion The retest data were mostly classified as indicating a strong association. The data generated by the checkpoints showed the successful implementation of the quality control program.

Characterizing the relations between exposures and diseases is the central tenet of epidemiology. Researchers may want to evaluate exposure-disease causation by assessing whether the disease under concern is induced by the various exposures - the so-called "attribution". In this paper, the authors propose a method to attribute diseases to multiple pathways based on the causal-pie model. The method can also be used to evaluate the potential impact of an intervention strategy and to allocate responsibility in tort-law liability issues. [ABSTRACT FROM AUTHOR]

Background: More than two-thirds of the world's HIV-positive individuals live in sub-Saharan Africa, where genetic susceptibility to kidney disease is high and resources for kidney disease screening and antiretroviral therapy (ART) toxicity monitoring are limited. Equations to estimate glomerular filtration rate (GFR) from serum creatinine were derived in Western populations and may be less accurate in this population. Methods: We compared results from published GFR estimating equations with a direct measure of GFR by iohexol clearance in 99 HIV-infected, ART-naïve Kenyan adults. Iohexol concentration was measured from dried blood spots on filter paper. The bias ratio (mean of the ratio of estimated to measured GFR) and accuracy (percentage of estimates within 30% of the measured GFR) were calculated. Results: The median age was 35 years, and 60% were women. The majority had asymptomatic HIV, with median CD4+ cell count of 355 cells/mm3. Median measured GFR was 115 mL/min/1.73 m2. Overall accuracy was highest for the Chronic Kidney Disease Epidemiology Consortium (CKD-EPI) equation. Consistent with a prior report, bias and accuracy were improved by eliminating the coefficient for black race (85% of estimates within 30% of measured GFR). Accuracy of all equations was poor in participants with GFR 60–90 mL/min/1.73 m2 (<65% of estimates within 30% of measured GFR), although this subgroup was too small to reach definitive conclusions. Conclusions: Overall accuracy was highest for the CKD-EPI equation. Eliminating the coefficient for race further improved performance. Future studies are needed to determine the most accurate GFR estimate for use in individuals with GFR <90 mL/min/1.73 m2, in whom accurate estimation of kidney function is important to guide drug dosing. Direct measurement of GFR by iohexol clearance using a filter paper based assay is feasible for research purposes in resource-limited settings, and could be used to develop more accurate GFR estimates in African populations. [ABSTRACT FROM AUTHOR]

Background: There are a number of different methods that can be used when estimating the size of the owned cat and dog population in a region, leading to varying population estimates. The aim of this study was to conduct a systematic review to evaluate the methods that have been used for estimating the sizes of owned cat and dog populations and to assess the biases associated with those methods. A comprehensive, systematic search of seven electronic bibliographic databases and the Google search engine was carried out using a range of different search terms for cats, dogs and population. The inclusion criteria were that the studies had involved owned or pet domestic dogs and/or cats, provided an estimate of the size of the owned dog or cat population, collected raw data on dog and cat ownership, and analysed primary data. Data relating to study methodology were extracted and assessed for biases. Results: Seven papers were included in the final analysis. Collection methods used to select participants in the included studies were: mailed surveys using a commercial list of contacts, door to door surveys, random digit dialled telephone surveys, and randomised telephone surveys using a commercial list of numbers. Analytical and statistical methods used to estimate the pet population size were: mean number of dogs/cats per household multiplied by the number of households in an area, human density multiplied by number of dogs per human, and calculations using predictors of pet ownership. Conclusion: The main biases of the studies included selection bias, non-response bias, measurement bias and biases associated with length of sampling time. Careful design and planning of studies is a necessity before executing a study to estimate pet populations. [ABSTRACT FROM AUTHOR]

Sample size calculations are an important part of research to balance the use of resources and to avoid undue harm to participants. Effect sizes are an integral part of these calculations and meaningful values are often unknown to the researcher. General recommendations for effect sizes have been proposed for several commonly used statistical procedures. For the analysis of tables, recommendations have been given for the correlation coefficient for binary data; however, it is well known that suffers from poor statistical properties. The odds ratio is not problematic, although recommendations based on objective reasoning do not exist. This paper proposes odds ratio recommendations that are anchored to for fixed marginal probabilities. It will further be demonstrated that the marginal assumptions can be relaxed resulting in more general results. [ABSTRACT FROM AUTHOR]

Methods for conducting dietary assessment in the United States date back to the early twentieth century. Methods of assessment encompassed dietary records, written and spoken dietary recalls, FFQ using pencil and paper and more recently computer and internet applications. Emerging innovations involve camera and mobile telephone technology to capture food and meal images. This paper describes six projects sponsored by the United States National Institutes of Health that use digital methods to improve food records and two mobile phone applications using crowdsourcing. The techniques under development show promise for improving accuracy of food records. [ABSTRACT FROM PUBLISHER]

This paper provides an overview of the baseline data collected in the nationwide Danish Centre for Strategic Research in Type 2 Diabetes (DD2) project. The paper presents descriptive data from the first 580 patients enrolled in the DD2. The DD2 database will contain detailed interview data, clinical examination data, and urine and blood samples from up to 10,000 patients newly diagnosed with type 2 diabetes each year, collected from general practitioners and hospital outpatient clinics in all of Denmark. Of the first DD2 patients enrolled, blood and urine samples have been obtained from 97%. The median age of the first 580 patients was 59 years and 322 (56%) were men. Median weight gain from age 20 to maximum lifetime weight was 29 kg for men and 31 kg for women, and 364 patients (63%) did not currently participate in regular sports activities. Two hundred and ninety two patients (50%) had a known family history of diabetes. Two hundred fifty (43%) of the 580 DD2 patients have also been enrolled in the Danish Diabetes Database for Adults from which additional clinical data can be obtained. Among these 250 patients (154 of whom were men, 96 women), 75 (49%) men were currently obese, and 63 (41%) were overweight, whereas 62 (65%) women were obese, and another 21 (22%) were overweight. Twenty-nine patients (12%) received insulin, 164 patients (66%) received oral antidiabetics only, and 57 (23%) received no antidiabetic treatment. Glycemic regulation was modest (the glycosylated hemoglobin A of 46% was $7.5%). Two thirds of the patients received antihypertensive and hypolipidemic treatment. Self-reported daily tobacco smoking (23%) and alcohol overuse (6%) seemed comparable to occurrence in the general Danish population. One quarter of the patients with newly diagnosed diabetes had a history of hospital-diagnosed comorbidity at baseline as included in the Charlson comorbidity index, in particular prior myocardial infarction (5%), cerebrovascular disease (5%), peripheral vascular disease (4%), chronic pulmonary disease (6%), and previous solid cancer (6%). In the future, the DD2 database represents a valuable source for outcome studies in type 2 diabetes. [ABSTRACT FROM AUTHOR]

Introduction: Asthma and allergies are complex diseases affected by genetic and environmental factors, such as occupational and psychosocial factors, as well as interactions between them. Although childhood is a critical phase in the development of asthma and allergies, few cohort studies on occupational outcomes followed up participants from childhood onwards. We present design, methods, and initial data analysis for the third follow-up of SOLAR (Study on Occupational Allergy Risks), a prospective and population-based German asthma and allergy cohort.Methods: The SOLAR cohort was initially recruited in 1995–1996 for Phase II of the German branch of the International Study of Asthma and Allergies in Childhood (ISAAC II) and followed up three times since, in 2002–2003, 2007–2009, and 2017–2018. During the third follow-up (SOLAR III), participants were between 29 and 34 years old. Since SOLAR focuses on occupational exposures, follow-ups were conducted at important points in time of the development of participants' career. To evaluate the potential of selection bias, responders and non-responders were compared based on variables from earlier study phases. In responders, frequency and pattern of missing values were examined and compared within the subsets of paper and online versions of the used questionnaires.Results: In total, 1,359 participants completed the questionnaire of the third follow-up (47.3% of eligible participants). Initially, the cohort started with 6,399 participants from the ISAAC II questionnaire study. A selection process led to a study population that is more female, higher educated, smokes less and has a higher proportion of certain asthma and allergy symptoms (also in their parents) than the initial cohort. Pattern and frequency of missing values were different for paper and online questionnaires.Discussion: The third follow-up of the SOLAR cohort offers the opportunity to analyze the course of asthma and allergies and their associations to environmental, occupational and psychosocial risk factors over more than 20 years from childhood to adulthood. Selection processes within the cohort might lead to bias that needs to be considered in future analyses.

Risk and rate advancement periods (RAP) measure the impact of an exposure on the relation of age to disease. Specifically, they quantify the time by which the risk or rate of a disease is advanced among exposed subjects conditional on disease-free survival to a certain baseline age. The fact that these measures incorporate timing of disease occurrence makes them appealing for risk communication, and their use has increased over recent years. Unfortunately, their misinterpretations have also increased. In particular, RAP is often misinterpreted as the difference in mean survival time, when in fact it is a distinct concept and a profoundly different quantity. Other misinterpretations include interpreting RAP as the time by which a survival curve is shifted between exposed and unexposed, and equating RAP to a simple function of relative risks. In this paper we review RAP and show why common misinterpretations are flawed. We also show how RAP is profoundly sensitive to the specification of disease dependence on age. We conclude that whereas RAP remains a potentially useful measure, it requires multiple cautions beyond those needed for traditional measures of association, especially when computed for rates. [ABSTRACT FROM AUTHOR]

Background: The Danish National Patient Registry (DNPR) is one of the world's oldest nationwide hospital registries and is used extensively for research. Many studies have validated algorithms for identifying health events in the DNPR, but the reports are fragmented and no overview exists. Objectives: To review the content, data quality, and research potential of the DNPR. Methods: We examined the setting, history, aims, content, and classification systems of the DNPR. We searched PubMed and the Danish Medical Journal to create a bibliography of validation studies. We included also studies that were referenced in retrieved papers or known to us beforehand. Methodological considerations related to DNPR data were reviewed. Results: During 1977-2012, the DNPR registered 8,085,603 persons, accounting for 7,268,857 inpatient, 5,953,405 outpatient, and 5,097,300 emergency department contacts. The DNPR provides nationwide longitudinal registration of detailed administrative and clinical data. It has recorded information on all patients discharged from Danish nonpsychiatric hospitals since 1977 and on psychiatric inpatients and emergency department and outpatient specialty clinic contacts since 1995. For each patient contact, one primary and optional secondary diagnoses are recorded according to the International Classification of Diseases. The DNPR provides a data source to identify diseases, examinations, certain in-hospital medical treatments, and surgical procedures. Long-term temporal trends in hospitalization and treatment rates can be studied. The positive predictive values of diseases and treatments vary widely (,15%-100%). The DNPR data are linkable at the patient level with data from other Danish administrative registries, clinical registries, randomized controlled trials, population surveys, and epidemiologic field studies - enabling researchers to reconstruct individual life and health trajectories for an entire population. Conclusion: The DNPR is a valuable tool for epidemiological research. However, both its strengths and limitations must be considered when interpreting research results, and continuous validation of its clinical data is essential. [ABSTRACT FROM AUTHOR]

Recent methodological innovation is giving rise to an increasing number of applied papers in medical and epidemiological journals in which natural direct and indirect effects are estimated. However, there is a longstanding debate on whether such effects are relevant targets of inference in population health. In light of the repeated calls for a more pragmatic and consequential epidemiology, we review three issues often raised in this debate: (i) the use of composite cross-world counterfactuals and the need for cross-world independence assumptions; (ii) interventional vs non-interventional identifiability; and (iii) the interpretational ambiguity of natural direct and indirect effect estimates. We use potential outcomes notation and directed acyclic graphs to explain ‘cross-world’ assumptions, illustrate implications of this assumption via regression models and discuss ensuing issues of interpretation. We argue that the debate on the relevance of natural direct and indirect effects rests on whether one takes as a target of inference the mathematical object per se, or the change in the world that the mathematical object represents. We further note that public health questions may be better served by estimating controlled direct effects. [ABSTRACT FROM AUTHOR]

As a performance measure for a prediction model, the area under the receiver operating characteristic curve (AUC) is insensitive to the addition of strong markers. A number of measures sensitive to performance change have recently been proposed; however, these relative-performance measures may lead to self-contradictory conclusions. This paper examines alternative performance measures for prediction models: the Lorenz curve-based Gini and Pietra indices, and a standardized version of the Brier score, the scaled Brier. Computer simulations are performed in order to study the sensitivity of these measures to performance change when a new marker is added to a baseline model. When the discrimination power of the added marker is concentrated in the gray zone of the baseline model, the AUC and the Gini show minimal performance improvements. The Pietra and the scaled Brier show more significant improvements in the same situation, comparatively. The Pietra and the scaled Brier indices are therefore recommended for prediction model performance measurement, in light of their ease of interpretation, clinical relevance and sensitivity to gray-zone resolving markers. [ABSTRACT FROM AUTHOR]

Background: It is known that experience of a previous crash is related to incidence of future crashes in a cohort of New Zealand cyclists. This paper investigated if the strength of such association differed by crash involvement propensity and by the need for medical care in the previous crash. Methods: The Taupo Bicycle Study involved 2590 adult cyclists recruited in 2006 and followed over a median period of 4.6 years through linkage to four national databases. The crash involvement propensity was estimated using propensity scores based on the participants' demographic, cycling and residential characteristics. Cox regression modelling for repeated events was performed with multivariate and propensity score adjustments. Analyses were then stratified by quintiles of the propensity score. Results: A total of 801 (31.0%) participants reported having experienced at least one bicycle crash in the twelve months prior to the baseline survey. They had a higher risk of experiencing crash events during follow-up (hazard ratio (HR): 1.43; 95% CI: 1.28, 1.60) but in the stratified analysis, this association was significant only in the highest two quintiles of the propensity score where the likelihood of having experienced a crash was more than 33%. The association was stronger for previous crashes that had received medical care (HR 1.63; 95% CI: 1.41, 1.88) compared to those that had not (HR 1.30; 95% CI: 1.14, 1.49). Conclusions: Previous crash experience increased the risk of future crash involvement in high-risk cyclists and the association was stronger for previous crashes attended medically. What distinguishes the high risk group warrants closer investigation, and the findings indicate also that health service providers could play an important role in prevention of bicycle crash injuries. [ABSTRACT FROM AUTHOR]

With the proliferation of social networks and blogs, the Internet is increasingly being used to disseminate personal health information rather than just as a source of information. In this paper we exploit the wealth of user-generated data, available through the micro-blogging service Twitter, to estimate and track the incidence of health conditions in society. The method is based on two stages: we start by extracting possibly relevant tweets using a set of specially crafted regular expressions, and then classify these initial messages using machine learning methods. Furthermore, we selected relevant features to improve the results and the execution times. To test the method, we considered four health states or conditions, namely flu, depression, pregnancy and eating disorders, and two locations, Portugal and Spain. We present the results obtained and demonstrate that the detection results and the performance of the method are improved after feature selection. The results are promising, with areas under the receiver operating characteristic curve between 0.7 and 0.9, and f-measure values around 0.8 and 0.9. This fact indicates that such approach provides a feasible solution for measuring and tracking the evolution of health states within the society. [ABSTRACT FROM AUTHOR]

Individual behavioral response to the spreading of an epidemic plays a crucial role in the progression of the epidemic itself. The risk perception induces individuals to adopt a protective behavior, as for instance reducing their social contacts, adopting more restrictive hygienic measures or undergoing prophylaxis procedures. In this paper, starting with a previously developed lattice-gas SIR model, we construct a coupled behavior-disease model for influenza spreading with spontaneous behavioral changes. The focus is on self-initiated behavioral changes that alter the susceptibility to the disease, without altering the contact patterns among individuals. Three different mechanisms of awareness spreading are analyzed: the local spreading due to the presence in the neighborhood of infective individuals; the global spreading due to the news published by the mass media and to educational campaigns implemented at institutional level; the local spreading occurring through the “thought contagion” among aware and unaware individuals. The peculiarity of the present approach is that the awareness spreading model is calibrated on available data on awareness and concern of the population about the risk of contagion. In particular, the model is validated against the A(H1N1) epidemic outbreak in Italy during the season, by making use of the awareness data gathered by the behavioral risk factor surveillance system (PASSI). We find that, increasing the accordance between the simulated awareness spreading and the PASSI data on risk perception, the agreement between simulated and experimental epidemiological data improves as well. Furthermore, we show that, within our model, the primary mechanism to reproduce a realistic evolution of the awareness during an epidemic, is the one due to globally available information. This result highlights how crucial is the role of mass media and educational campaigns in influencing the epidemic spreading of infectious diseases. [ABSTRACT FROM AUTHOR]

The number of travel-acquired dengue infections has seen a consistent global rise over the past decade. An increased volume of international passenger air traffic originating from regions with endemic dengue has contributed to a rise in the number of dengue cases in both areas of endemicity and elsewhere. This paper reports results from a network-based risk assessment model which uses international passenger travel volumes, travel routes, travel distances, regional populations, and predictive species distribution models (for the two vector species, Aedes aegypti and Aedes albopictus) to quantify the relative risk posed by each airport in importing passengers with travel-acquired dengue infections. Two risk attributes are evaluated: (i) the risk posed by through traffic at each stopover airport and (ii) the risk posed by incoming travelers to each destination airport. The model results prioritize optimal locations (i.e., airports) for targeted dengue surveillance. The model is easily extendible to other vector-borne diseases. [ABSTRACT FROM AUTHOR]

Background: Up to 2% of suicides in young people may occur in clusters i.e., close together in time and space. In early 2008 unprecedented attention was given by national and international news media to a suspected suicide cluster among young people living in Bridgend, Wales. This paper investigates the strength of statistical evidence for this apparent cluster, its size, and temporal and geographical limits. Methods and findings: The analysis is based on official mortality statistics for Wales for 2000–2009 provided by the UK's Office for National Statistics (ONS). Temporo-spatial analysis was performed using Space Time Permutation Scan Statistics with SaTScan v9.1 for suicide deaths aged 15 and over, with a sub-group analysis focussing on cases aged 15–34 years. These analyses were conducted for deaths coded by ONS as: (i) suicide or of undetermined intent (probable suicides) and (ii) for a combination of suicide, undetermined, and accidental poisoning and hanging (possible suicides). The temporo-spatial analysis did not identify any clusters of suicide or undetermined intent deaths (probable suicides). However, analysis of all deaths by suicide, undetermined intent, accidental poisoning and accidental hanging (possible suicides) identified a temporo-spatial cluster (p = 0.029) involving 10 deaths amongst 15–34 year olds centred on the County Borough of Bridgend for the period 27th December 2007 to 19th February 2008. Less than 1% of possible suicides in younger people in Wales in the ten year period were identified as being cluster-related. Conclusions: There was a possible suicide cluster in young people in Bridgend between December 2007 and February 2008. This cluster was smaller, shorter in duration, and predominantly later than the phenomenon that was reported in national and international print media. Further investigation of factors leading to the onset and termination of this series of deaths, in particular the role of the media, is required. [ABSTRACT FROM AUTHOR]

Advances in next-generation sequencing technologies have enabled the identification of multiple rare single nucleotide polymorphisms involved in diseases or traits. Several strategies for identifying rare variants that contribute to disease susceptibility have recently been proposed. An important feature of many of these statistical methods is the pooling or collapsing of multiple rare single nucleotide variants to achieve a reasonably high frequency and effect. However, if the pooled rare variants are associated with the trait in different directions, then the pooling may weaken the signal, thereby reducing its statistical power. In the present paper, we propose a backward support vector machine (BSVM)-based variant selection procedure to identify informative disease-associated rare variants. In the selection procedure, the rare variants are weighted and collapsed according to their positive or negative associations with the disease, which may be associated with common variants and rare variants with protective, deleterious, or neutral effects. This nonparametric variant selection procedure is able to account for confounding factors and can also be adopted in other regression frameworks. The results of a simulation study and a data example show that the proposed BSVM approach is more powerful than four other approaches under the considered scenarios, while maintaining valid type I errors. [ABSTRACT FROM AUTHOR]

We consider the recently introduced edge-based compartmental models (EBCM) for the spread of susceptible-infected-recovered (SIR) diseases in networks. These models differ from standard infectious disease models by focusing on the status of a random partner in the population, rather than a random individual. This change in focus leads to simple analytic models for the spread of SIR diseases in random networks with heterogeneous degree. In this paper we extend this approach to handle deviations of the disease or population from the simplistic assumptions of earlier work. We allow the population to have structure due to effects such as demographic features or multiple types of risk behavior. We allow the disease to have more complicated natural history. Although we introduce these modifications in the static network context, it is straightforward to incorporate them into dynamic network models. We also consider serosorting, which requires using dynamic network models. The basic methods we use to derive these generalizations are widely applicable, and so it is straightforward to introduce many other generalizations not considered here. Our goal is twofold: to provide a number of examples generalizing the EBCM method for various different population or disease structures and to provide insight into how to derive such a model under new sets of assumptions. [ABSTRACT FROM AUTHOR]

Background: Physical activity is inversely correlated to morbidity and mortality risk. Large cohort studies use wearable accelerometer devices to measure physical activity objectively, providing data potentially relevant to identify different activity patterns and to correlate these to health-related outcome measures. A method to compute relevant characteristics of such data not only with regard to duration and intensity, but also to regularity of activity events, is necessary. The aims of this paper are to propose a new method – the ATLAS index (Activity Types from Long-term Accelerometric Sensor data) – to derive generic measures for distinguishing different characteristic activity phenotypes from accelerometer data, to propose a comprehensive graphical representation, and to conduct a proof-of-concept with long-term measurements from different devices and cohorts. Methods: The ATLAS index consists of the three dimensions regularity (reg), duration (dur) and intensity (int) of relevant activity events identified in long-term accelerometer data. It can be regarded as a 3D vector and represented in a 3D cube graph. 12 exemplary data sets of three different cohort studies with 99,467 minutes of data were chosen for concept validation. Results: Five archetypical activity types are proposed along with their dimensional characteristics (insufficiently active: low reg, int and dur; busy bee: low dur and int, high reg; cardio-active: medium reg, int and dur, endurance athlete: high reg, int and dur; and weekend warrior: high int and dur, low reg). The data sets are displayed in one common graph, indicating characteristic differences in activity patterns. Conclusion: The ATLAS index incorporates the relevant regularity dimension apart from the widely-used measures of duration and intensity. Along with the 3D representation, it allows to compare different activity types in cohort study populations, both visually and computationally using vector distance measures. Further research is necessary to validate the ATLAS index in order to find normative values and group centroids. [ABSTRACT FROM AUTHOR]

Background: Recent focus on earlier detection of pathogen introduction in human and animal populations has led to the development of surveillance systems based on automated monitoring of health data. Real- or near real-time monitoring of pre-diagnostic data requires automated classification of records into syndromes–syndromic surveillance–using algorithms that incorporate medical knowledge in a reliable and efficient way, while remaining comprehensible to end users. Methods: This paper describes the application of two of machine learning (Naïve Bayes and Decision Trees) and rule-based methods to extract syndromic information from laboratory test requests submitted to a veterinary diagnostic laboratory. Results: High performance (F1-macro = 0.9995) was achieved through the use of a rule-based syndrome classifier, based on rule induction followed by manual modification during the construction phase, which also resulted in clear interpretability of the resulting classification process. An unmodified rule induction algorithm achieved an F1-micro score of 0.979 though this fell to 0.677 when performance for individual classes was averaged in an unweighted manner (F1-macro), due to the fact that the algorithm failed to learn 3 of the 16 classes from the training set. Decision Trees showed equal interpretability to the rule-based approaches, but achieved an F1-micro score of 0.923 (falling to 0.311 when classes are given equal weight). A Naïve Bayes classifier learned all classes and achieved high performance (F1-micro = 0.994 and F1-macro = .955), however the classification process is not transparent to the domain experts. Conclusion: The use of a manually customised rule set allowed for the development of a system for classification of laboratory tests into syndromic groups with very high performance, and high interpretability by the domain experts. Further research is required to develop internal validation rules in order to establish automated methods to update model rules without user input. [ABSTRACT FROM AUTHOR]

Measurement of malaria burden is fraught with complexity, due to the natural history of the disease, delays in seeking treatment or failure of case management. Attempts to establish an appropriate case definition for a malaria episode has often resulted in ambiguities and challenges because of poor information about treatment seeking, patterns of infection, recurrence of fever and asymptomatic infection. While the primary reason for treating malaria is to reduce disease burden, the effects of treatment are generally ignored in estimates of the burden of malaria morbidity, which are usually presented in terms of numbers of clinical cases or episodes, with the main data sources being reports from health facilities and parasite prevalence surveys. The use of burden estimates that do not consider effects of treatment, leads to under-estimation of the impact of improvements in case management. Official estimates of burden very likely massively underestimate the impact of the roll-out of ACT as first-line therapy across Africa. This paper proposes a novel approach for estimating burden of disease based on the point prevalence of malaria attributable disease, or equivalently, the days with malaria fever in unit time. The technique makes use of data available from standard community surveys, analyses of fever patterns in malaria therapy patients, and data on recall bias. Application of this approach to data from Zambia for 2009–2010 gave an estimate of 2.6 (95% credible interval: 1.5–3.7) malaria attributable fever days per child-year. The estimates of recall bias, and of the numbers of days with illness contributing to single illness recalls, could be applied more generally. To obtain valid estimates of the overall malaria burden using these methods, there remains a need for surveys to include the whole range of ages of hosts in the population and for data on seasonality patterns in confirmed case series. [ABSTRACT FROM AUTHOR]

Background: Research using electronic health records (EHRs) relies heavily on coded clinical data. Due to variation in coding practices, it can be difficult to aggregate the codes for a condition in order to define cases. This paper describes a methodology to develop ‘indicator markers’ found in patients with early rheumatoid arthritis (RA); these are a broader range of codes which may allow a probabilistic case definition to use in cases where no diagnostic code is yet recorded. Methods: We examined EHRs of 5,843 patients in the General Practice Research Database, aged ≥30y, with a first coded diagnosis of RA between 2005 and 2008. Lists of indicator markers for RA were developed initially by panels of clinicians drawing up code-lists and then modified based on scrutiny of available data. The prevalence of indicator markers, and their temporal relationship to RA codes, was examined in patients from 3y before to 14d after recorded RA diagnosis. Findings: Indicator markers were common throughout EHRs of RA patients, with 83.5% having 2 or more markers. 34% of patients received a disease-specific prescription before RA was coded; 42% had a referral to rheumatology, and 63% had a test for rheumatoid factor. 65% had at least one joint symptom or sign recorded and in 44% this was at least 6-months before recorded RA diagnosis. Conclusion: Indicator markers of RA may be valuable for case definition in cases which do not yet have a diagnostic code. The clinical diagnosis of RA is likely to occur some months before it is coded, shown by markers frequently occurring ≥6 months before recorded diagnosis. It is difficult to differentiate delay in diagnosis from delay in recording. Information concealed in free text may be required for the accurate identification of patients and to assess the quality of care in general practice. [ABSTRACT FROM AUTHOR]

The overall aim of the Danish Centre for Strategic Research in Type 2 Diabetes (DD2) is to near-normalize metabolic control in newly diagnosed patients with type 2 diabetes (T2D) using an individualized treatment approach. We hypothesize that this will not only prevent complications and improve quality of life for T2D patients but also result in increased cost efficiency compared with current treatment modalities. This paper provides an overview of the expected outcomes from DD2, focusing on the two main intervention studies. The main data for the DD2 project are collected during patient enrollment and stored using the individual civil registration number. This enables subsequent linking to other national databases where supplemental data can be obtained. All data will be used for designing treatment guidelines and continuously monitoring the development of diabetic complications, thereby obtaining knowledge about predictors for the long-term outcome and identifying targets for new interventions. Further data are being collected from two intervention studies. The aim of the first intervention study is to improve T2D treatment using an individualized treatment modality optimizing medication according to individual metabolic responses and phenotypic characteristics. The aim of the second intervention study is to develop an evidence-based training protocol to be implemented as a treatment modality for T2D and used for initiating lifelong changes in physical activity levels in patients with T2D. An initial pilot study evaluating an interval-based walking protocol is ongoing, and preliminary results indicate that this protocol is an optimal "free-living" training intervention. An initial health-economic analysis will also be performed as a basis for analysis of the data collected during the project. A cost-benefit analysis of the two intervention studies will be conducted. The DD2 project is expected to lead to improved treatment modalities and increased knowledge about existing treatment guidelines, and will also provide a solid base for health-economic decision-making. [ABSTRACT FROM AUTHOR]

This paper aims to assess the yielding of the sampling procedures and the representativeness of a sample evaluated in a national telephone survey in Portugal. The Portuguese telephone book (2007/2008) was the sampling frame for subjects >39 years ( n = 1,934). The time spent with interviews and unsuccessful contacts was quantified and the yielding of the contact attempts compared according different schedules. Demographic and health characteristics of the participants were compared with those of the Portuguese population. Among eligible individuals, the percentage of refusals was 40.4%, higher on Saturdays (45.2% vs. 30.7%) and evenings (45.2% vs. 36.7%). Evening interviews yielded a higher percentage of men (26.2% vs. 21.0%) and active workers (31.8% vs. 21.8%). Compared to the Portuguese population, our sample included a higher proportion of women (76.7% vs. 54.1%), elderly (>59 years: 61.4% vs. 46.2%) and more educated participants (>4 years: 38.3% vs. 26.3%). Sex- and age-specific estimates of smoking and obesity were similar to those in the Portuguese population. Despite the difficulties in reaching participants, this strategy may produce unbiased estimates for important health indicators. [ABSTRACT FROM AUTHOR]

Instrumental-variable (IV) methods were invented over 70 years ago, but remain uncommon in epidemiology. Over the past decade or so, non-parametric versions of IV methods have appeared that connect IV methods to causal and measurement-error models important in epidemiological applications. This paper provides an introduction to those developments, illustrated by an application of IV methods to non-parametric adjustment for non-compliance in randomized trials. [ABSTRACT FROM AUTHOR]

Preliminary evidence suggests that neighborhood environments, such as socioeconomic disadvantage, pedestrian and physical activity infrastructure, and availability of neighborhood destinations (e.g., parks), may be associated with late-life cognitive functioning and risk of Alzheimer's disease and related disorders (ADRD). The supposition is that these neighborhood characteristics are associated with factors such as mental health, environmental exposures, health behaviors, and social determinants of health that in turn promote or diminish cognitive reserve and resilience in later life. However, observed associations may be biased by self-selection or reverse causation, such as when individuals with better cognition move to denser neighborhoods because they prefer many destinations within walking distance of home, or when individuals with deteriorating health choose residences offering health services in neighborhoods in rural or suburban areas (e.g., assisted living). Research on neighborhood environments and ADRD has typically focused on late-life brain health outcomes, which makes it difficult to disentangle true associations from associations that result from reverse causality. In this paper, we review study designs and methods to help reduce bias due to reverse causality and self-selection, while drawing attention to the unique aspects of these approaches when conducting research on neighborhoods and brain aging.

Bronchial provocation tests using pharmacological agents such as methacholine or histamine are used in epidemiological studies to identify asthma despite recognition of limitations in specificity, positive predictive value and availability of reagents. Hypertonic saline (4.5%) bronchial challenge (HSBC), although less sensitive than pharmacological challenges, is reportedly highly specific in diagnosing current asthma. Added advantages are that reagents are cheap, stable and recognized by participants. Thus, HSBC may offer benefits over pharmacological tests in epidemiological surveys. This paper reports on the second field survey using the test, a study of 99 adults from the timber industry in Western Australia. The test is described and critically appraised as a practical epidemiological tool for assessing asthma prevalence. At a cutoff point of 20% FEV1 fall, HSBC was positive in 8% of subjects, appeared specific for asthma, was safe, well-accepted and easy to use in the field. [ABSTRACT FROM PUBLISHER]

Background: During the last decades, sex and gender biases have been identified in various areas of biomedical and public health research, leading to compromised validity of research findings. As a response, methodological requirements were developed but these are rarely translated into research practice. The aim of this study is to provide good practice examples of sex/gender sensitive health research.Methods: We conducted a systematic search of research articles published in JECH between 2006 and 2014. An instrument was constructed to evaluate sex/gender sensitivity in four stages of the research process (background, study design, statistical analysis, discussion).Results: In total, 37 articles covering diverse topics were included. Thereof, 22 were evaluated as good practice example in at least one stage; two articles achieved highest ratings across all stages. Good examples of the background referred to available knowledge on sex/gender differences and sex/gender informed theoretical frameworks. Related to the study design, good examples calculated sample sizes to be able to detect sex/gender differences, selected sex/gender sensitive outcome/exposure indicators, or chose different cut-off values for male and female participants. Good examples of statistical analyses used interaction terms with sex/gender or different shapes of the estimated relationship for men and women. Examples of good discussions interpreted their findings related to social and biological explanatory models or questioned the statistical methods used to detect sex/gender differences.Conclusions: The identified good practice examples may inspire researchers to critically reflect on the relevance of sex/gender issues of their studies and help them to translate methodological recommendations of sex/gender sensitivity into research practice. [ABSTRACT FROM AUTHOR]

Contact tracing, where exposed individuals are followed up to break ongoing transmission chains, is a key pillar of outbreak response for infectious disease outbreaks. Unfortunately, these systems are not fully effective, and infections can still go undetected as people may not remember all their contacts or contacts may not be traced successfully. A large proportion of undetected infections suggests poor contact tracing and surveillance systems, which could be a potential area of improvement for a disease response. In this paper, we present a method for estimating the proportion of infections that are not detected during an outbreak. Our method uses next generation matrices that are parameterized by linked contact tracing data and case line-lists. We validate the method using simulated data from an individual-based model and then investigate two case studies: the proportion of undetected infections in the SARS-CoV-2 outbreak in New Zealand during 2020 and the Ebola epidemic in Guinea during 2014. We estimate that only 5.26% of SARS-CoV-2 infections were not detected in New Zealand during 2020 (95% credible interval: 0.243 – 16.0%) if 80% of contacts were under active surveillance but depending on assumptions about the ratio of contacts not under active surveillance versus contacts under active surveillance 39.0% or 37.7% of Ebola infections were not detected in Guinea (95% credible intervals: 1.69 – 87.0% or 1.70 – 80.9%). • Next generation matrices can be used to estimate the proportion of infections that are not detected during an outbreak. • Our method is parameterised by linking case and contact tracing data, which should be routine in responding to outbreaks. • Our method is parameterised by less data than traditional methods. [ABSTRACT FROM AUTHOR]