The R package mice imputes incomplete multivariate data by chained equations. The software mice 1.0 appeared in the year 2000 as an S-PLUS library, and in 2001 as an R package. mice 1.0 introduced predictor selection, passive imputation and automatic pooling. This article documents mice, which extends the functionality of mice 1.0 in several ways. In mice, the analysis of imputed data is made completely general, whereas the range of models under which pooling works is substantially extended. mice adds new functionality for imputing multilevel data, automatic predictor selection, data handling, post-processing imputed values, specialized pooling routines, model selection tools, and diagnostic graphs. Imputation of categorical data is improved in order to bypass problems caused by perfect prediction. Special attention is paid to transformations, sum scores, indices and interactions using passive imputation, and to the proper setup of the predictor matrix. mice can be downloaded from the Comprehensive R Archive Network. This article provides a hands-on, stepwise approach to solve applied incomplete data problems.

/pdf/mice-multivariate-imputation-by-chained-equations-in-r-2cffzza5jj.pdf

mice: Multivariate Imputation by Chained Equations in R

Multiple imputation by chained equations is a flexible and practical approach to handling missing data. We describe the principles of the method and show how to impute categorical and quantitative variables, including skewed variables. We give guidance on how to specify the imputation model and how many imputations are needed. We describe the practical analysis of multiply imputed data, including model building and model checking. We stress the limitations of the method and discuss the possible pitfalls. We illustrate the ideas using a data set in mental health, giving Stata code fragments. Copyright © 2010 John Wiley & Sons, Ltd.

Multiple imputation using chained equations: Issues and guidance for practice

The classic survey design reference, updated for the digital ageFor over two decades, Dillman's classic text on survey design has aided both students and professionals in effectively planning and conducting mail, telephone, and, more recently, Internet surveys. The new edition is thoroughly updated and revised, and covers all aspects of survey research. It features expanded coverage of mobile phones, tablets, and the use of do-it-yourself surveys, and Dillman's unique Tailored Design Method is also thoroughly explained. This invaluable resource is crucial for any researcher seeking to increase response rates and obtain high-quality feedback from survey questions. Consistent with current emphasis on the visual and aural, the new edition is complemented by copious examples within the text and accompanying website.This heavily revised Fourth Edition includes:Strategies and tactics for determining the needs of a given survey, how to design it, and how to effectively administer itHow and when to use mail, telephone, and Internet surveys to maximum advantageProven techniques to increase response ratesGuidance on how to obtain high-quality feedback from mail, electronic, and other self-administered surveysDirection on how to construct effective questionnaires, including considerations of layoutThe effects of sponsorship on the response rates of surveysUse of capabilities provided by newly mass-used media: interactivity, presentation of aural and visual stimuli.The Fourth Edition reintroduces the telephoneincluding coordinating land and mobile.Grounded in the best research, the book offers practical how-to guidelines and detailed examples for practitioners and students alike.

/pdf/internet-mail-and-mixed-mode-surveys-the-tailored-design-3tlkuxua87.pdf

Internet, Mail, and Mixed-Mode Surveys: The Tailored Design Method

Over the last several decades, epidemiological studies have been enormously
successful in identifying risk factors for major diseases However, most of this
research has focused attention on risk factors that are relatively proximal causes of
disease such as diet, cholesterol level, exercise and the like We question the
emphasis on such individually-based risk factors and argue that greater attention
must be paid to basic social conditions if health reform is to have its maximum
effect in the time ahead There are two reasons for this claim First we argue that
individually-based risk factors must be contextualized, by examining what puts
people at risk of risks, if we are to craft effective interventions and improve the
nation's health Second, we argue that social factors such as socioeconomic status
and social support are likely 'fundamental causes" of disease that, because they
embody access to important resources, affect multiple disease outcomes throughmultiple mechanisms, and consequently maintain an association with disease even when intervening mechanisms change Without careful attention to these
possibilities, we run the risk of imposing individually-based intervention strategies
that are ineffective and of missing opportunities to adopt broad-based societal
interventions that could produce substantial health benefits for our citizens

Social Conditions as Fundamental Causes of Disease

Three terms are commonly used interchangeably to identify vulnerable older adults: comorbidity, or multiple chronicconditions,frailty, anddisability. However, in geriatricmedicine,there isagrowingconsensusthatthese are distinct clinical entities that are causally related. Each, individually, occurs frequently and has high import clinically. This article provides a narrative review of current understanding of the definitions and distinguishing characteristics of each of these conditions,including theirclinical relevance and distinct prevention and therapeutic issues, and how they are related. Review of the current state of published knowledge is supplemented by targeted analysesin selectedareas where no current publisheddataexists. Overall,the goalof this articleis to providea basis fordistinguishingbetweenthesethreeimportantclinicalconditionsinolderadultsandshowinghowuseofseparate, distinct definitions of each can improve our understanding of the problems affecting older patients and lead to development of improved strategies for diagnosis, care, research, and medical education in this area.

/pdf/untangling-the-concepts-of-disability-frailty-and-3jhe15an05.pdf

Untangling the Concepts of Disability, Frailty, and Comorbidity: Implications for Improved Targeting and Care

This article describes and evaluates a procedure for imputing missing values for a relatively complex data structure when the data are missing at random. The imputations are obtained by fitting a sequence of regression models and drawing values from the corresponding predictive distributions. The types of regression models used are linear, logistic, Poisson, generalized logit or a mixture of these depending on the type of variable being imputed. Two additional common features in the imputation process are incorporated: restriction to a relevant subpopulation for some variables and logical bounds or constraints for the imputed values. The restrictions involve subsetting the sample individuals that satisfy certain criteria while fitting the regression models. The bounds involve drawing values from a truncated predictive distribution. The development of this method was partly motivated by the analysis of two data sets which are used as illustrations. The sequential regression procedure is applied to perform multiple imputation analysis for the two applied problems. The sampling properties of inferences from multiply imputed data sets created using the sequential regression method are evaluated through simulated data sets.

A multivariate technique for multiply imputing missing values using a sequence of regression models

Context.— A prominent hypothesis regarding social inequalities in mortality is that the elevated risk among the socioeconomically disadvantaged is largely due to the higher prevalence of health risk behaviors among those with lower levels of education and income. Objective.— To investigate the degree to which 4 behavioral risk factors (cigarette smoking, alcohol drinking, sedentary lifestyle, and relative body weight) explain the observed association between socioeconomic characteristics and allcause mortality. Design.— Longitudinal survey study investigating the impact of education, income, and health behaviors on the risk of dying within the next 7.5 years. Participants.— A nationally representative sample of 3617 adult women and men participating in the Americans’ Changing Lives survey. Main Outcome Measure.— All-cause mortality verified through the National Death Index and death certificate reviews. Results.— Educational differences in mortality were explained in full by the strong association between education and income. Controlling for age, sex, race, urbanicity, and education, the hazard rate ratio of mortality was 3.22 (95% confidence interval [CI], 2.01-5.16) for those in the lowest-income group and 2.34 (95% CI, 1.49-3.67) for those in the middle-income group. When health risk behaviors were considered, the risk of dying was still significantly elevated for the lowestincome group (hazard rate ratio, 2.77; 95% CI, 1.74-4.42) and the middle-income group (hazard rate ratio, 2.14; 95% CI, 1.38-3.25). Conclusion.— Although reducing the prevalence of health risk behaviors in lowincome populations is an important public health goal, socioeconomic differences in mortality are due to a wider array of factors and, therefore, would persist even with improved health behaviors among the disadvantaged.

https://jamanetwork.com/journals/jama/articlepdf/187597/joc72105.pdf

Socioeconomic factors, health behaviors, and mortality: Results from a nationally representative prospective study of US adults

The way health varies with age is importantly stratified by socioeconomic status (SES)--specifically, education and income. Prior theory and cross-sectional data suggest that among higher SES persons the onset of health problems is usually postponed until rather late in life, while health declines are prevalent in lower SES groups by middle age. Thus, SES differences in health are small in early adulthood, but increase with age until relatively late in life, when they diminish due to selection or greater equalization of health risks and protections. The present paper strengthens our causal and interpretive understanding of these phenomena by showing: (1) that results previously reported for indices of SES hold separately for education and income; (2) that the interaction between age and SES (i.e., education or income) in predicting health can be substantially explained by the greater exposure of lower SES persons to a wide range of psychosocial risk factors to health, especially in middle and early old age, and, to a lesser degree, the greater impact of these risk factors on health with age; and (3) that results (1) and (2) generally hold in short-term longitudinal as well as in cross-sectional data. Implications for science and policy in the areas of aging, health, and social stratification are discussed.

The social stratification of aging and health

http://www.isr.umich.edu/williams/All%20Publications/DRW%20pubs%202001/socioeconomic%20disparities%20in%20health%20change.pdf

Socioeconomic disparities in health change in a longitudinal study of US adults: the role of health-risk behaviors.

Older people often suffer from comorbidity, or several chronic conditions simultaneously. Disability rises rapidly as the number of chronic conditions grows, although very ill people who acquire another condition experience attenuated increases. High prevalence conditions such as arthritis tend to have a low or occasionally moderate impact for community residents, while low prevalence ones such as osteoporosis have a high impact; paired conditions sometimes give extra propulsion to disability, as when cerebrovascular disease and hip fracture co-occur. Further research is needed to pin-point combinations of conditions posing great risks and to identify demographic segments in which comorbidity has elevated effects.

https://www.jstor.org/stable/pdfplus/3350223.pdf

James M. Lepkowski

Papers

A multivariate technique for multiply imputing missing values using a sequence of regression models

Socioeconomic factors, health behaviors, and mortality: Results from a nationally representative prospective study of US adults

The social stratification of aging and health

Socioeconomic disparities in health change in a longitudinal study of US adults: the role of health-risk behaviors.

Comorbidity and Its Impact on Disability