Search

Objectives: Identify characteristics distinguishing people who do and do not continue to participate in a longitudinal study and determine whether the longitudinal changes for people who continue are representative of the changes that would have occurred had longitudinal data been available from all of the initial participants., Method: Moderately large samples of returning (N = 2,082) and nonreturning (N = 1,698) participants across a wide age range (i.e., 18-97 years of age) performed a battery of cognitive tests and completed personality and mood questionnaires. Differences between the groups were examined with multiple regression analyses with age, returner status, and their interaction as predictors., Results: Compared with participants who did not return, returning participants at the initial occasion had higher levels of each cognitive ability and of certain personality characteristics (e.g., agreeableness and openness), but many of the differences were only apparent among adults older than 50 years of age. Importantly, there was no evidence that the longitudinal change for nonreturning participants would have been different from that among the participants who did return., Discussion: The phenomenon of selective attrition is more complex than often assumed, and it may not necessarily limit the generalizability of longitudinal comparisons., (© The Author 2013. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.)

Objective: The article describes the strategies adopted by the Brazilian Longitudinal Study for Adult Health (ELSA-Brasil) for participation and retention of subjects. This is key to ensure internal validity of longitudinal studies, and to identify, investigate, and ascertain outcomes of interest., Methods: The follow-up strategies include annual telephone contacts with new assessments and interviews every three to four years this approach aims to identify transient outcomes (reversible or not), permanent outcomes as well as complications related to the progression of major diseases--cardiovascular diseases and diabetes--to be studied., Results: Telephone interviews are designed to monitor subjects' health status and to identify potential health-related events such as hospital admissions, medical visits or pre-selected medical procedures. Subjects are also encouraged to report to the ELSA-Brasil team any new health-related events. When a potential event is identified, a thorough investigation is carried out to collect relevant information about that event from medical records. All data are blinded and reviewed and analyzed by a medical expert committee. Incident outcome ascertainment follows well-established international criteria to ensure data comparability and avoid misclassification. In addition to these strategies, the occurrence of health-related events is also investigated through linkage of secondary databases, such as national mortality and hospital admission databases., Conclusions: Accurate identification of outcomes will allow to estimating their incidence in the study cohort and to investigate the effect of the exposures studied in the ELSA-Brasil at baseline and at its subsequent waves.

In an observational study, one treated subject may be matched for observed covariates to either one or several untreated controls. The common motivation for using several controls rather than one is to increase the power of a test of no effect under the doubtful assumption that matching for observed covariates suffices to remove bias from nonrandom treatment assignment. Does the choice between one or several matched controls affect the sensitivity of conclusions to violations of this doubtful assumption? With continuous responses, it is known that reducing the heterogeneity of matched pair differences reduces sensitivity to unmeasured biases, but increasing the sample size has a highly circumscribed effect on sensitivity to bias. Is the use of several controls rather than one analogous to a reduction in heterogeneity or to an increase in sample size? The issue is examined for Huber's m-statistics, including the t-test, the examination having three components: an example, asymptotic calculations using design sensitivity, and a simulation. Use of multiple controls with continuous responses yields a nontrivial reduction in sensitivity to unmeasured biases. An example looks at lead and cadmium in the blood of smokers from the 2008 National Health and Nutrition Examination Survey. A by-product of the discussion is a new result giving the design sensitivity for the permutation distribution of m-statistics., (Copyright © 2013, The International Biometric Society.)

Background: Household contact with an index case of an infectious disease is a known risk factor for infection transmission. However, such contact may be underestimated due to the dynamic nature of households, particularly in longitudinal studies. Such studies generally begin with contact defined at a single point in time ('snap-shot'), leading to contact misclassification for some individuals who actually experienced contact before and after the snapshot., Objective: To quantify contact misclassification with index cases of disease in households., Methods: Historical data of 112,026 individuals from 17,889 households from an epidemiological study on leprosy in northern Malawi were used. Individuals were interviewed in the early 1980s and followed up over 5 years. It was possible to trace whether individuals died, changed household within the area, or moved out of the area between the two surveys.Using a 10% sample of households as the starting population and parameters for demographic and household changes over 5 years, the extent of contact misclassification was estimated through a simulation model of household dynamics, which traced contact with index cases in households over time. The model thereafter compared initial contact status and 'true' contact status generated from simulations., Results: The starting population had 11,401 individuals, 52% female, and 224 (2%) leprosy index cases. Eleven percent of the households had at least one index case resident and 10% (1, 177) of non-case individuals were initial contacts. Sensitivity of initial contact status ranged from 0.52 to 0.74 and varied by age and sex. Sensitivity was low in those aged 20-29 and under 5 years but high in 5- to 14-year-olds. By gender, there were no differences among those aged under 5; females had lower sensitivity among those aged under 20 and higher for those above 30, respectively. Sensitivity was also low in simulations of long incubation periods., Conclusion: This work demonstrates the implications of changes in households on household contact-associated disease spread, particularly for long durations of follow-up and infections with long incubation periods where earlier unobserved contact is critical.

The number of life events reported by study participants is sensitive to the method of data collection and time intervals under consideration. Individual characteristics also influence reporting; respondents with poor mental health report more life events. Much current research on life events is cross-sectional. Data from a longitudinal study of women's health from 4 waves over a decade suggest that over time additional systematic biases in reporting life events occur. Inconsistency over time is due to both fall-off of reporting and telescoping. Intracategory variability and ambiguity of items, as well as respondent characteristics, also potentially contribute to response biases. Although some factors (e.g., item wording) are controllable, others (e.g., respondents' mental health) are not and must be factored into data analysis and interpretation.

Introduction: In this article, we introduce Marginal Structural Models, which yield unbiased estimates of causal effects of exposures in the presence of time-varying confounding variables that also act as mediators., Objectives: We describe estimation via inverse probability weighting; estimation may also be accomplished by g-computation (Robins in Latent Variable Modeling and Applications to Causality, Springer, New York, pp 69-117, 1997; van der Wal et al. in Stat Med 28:2325-2337, 2009) or targeted maximum likelihood (Rosenblum and van der Laan in Int J Biostat 6, 2010)., Conclusions: When both time-varying confounding and mediation are present in a longitudinal setting data, Marginal Structural Models are a useful tool that provides unbiased estimates.

Objectives: Our objective was to develop and test standardized methods for collection and statistical analysis of longitudinal data on hospital antibacterial use from different countries., Methods: We collected data on monthly supply of antibiotics from pharmacies in one hospital from each of 18 European countries. We applied a standardized method to classify drugs, measure use in defined daily doses and compare the effect of using occupied bed-days (OBDs) or admissions as denominators for longitudinal analysis., Results: Antibiotic use increased in 14 (78%) hospitals and decreased in 4 hospitals. For 16 (89%) hospitals, adjustment of antibiotic use with OBDs resulted in larger changes over time than adjustment with admissions. Inclusion of all hospital clinical activity variables (admissions, length of stay and OBDs) in multivariate time series analysis identified distinct hospital groups. Nine (50%) hospitals had statistically significant changes in antibiotic use (six increasing and three decreasing) that were not explained (n = 3) or only partially explained (n = 6) by change in clinical activity. Three (17%) hospitals had no significant change in antibiotic use. In the remaining six hospitals, apparent changes in antibiotic use were largely explained by changes in clinical activity., Conclusions: This is the first study to use a standardized method for data collection and longitudinal analysis of antibiotic use in different hospitals. These data suggest that determination of changes in antibiotic exposure of hospital patients over a period of time is unreliable if only one clinical activity variable (such as OBDs) is used as the denominator. We recommend inclusion of admissions, OBDs and length of stay in statistical, time series analysis of antibiotic use. This model is also relevant to longitudinal analysis of infections in hospitals.