Systematic handling of missing data in complex study designs – experiences from the Health 2000 and 2011 Surveys

Härkänen, Tommi; Karvanen, Juha; Tolonen, Hanna; Lehtonen, Risto; Djerf, Kari; Juntunen, Teppo; Koskinen, Seppo

doi:10.1080/02664763.2016.1144725

Cited by 55 publications

(64 citation statements)

References 54 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The possible bias of Rubin's MI variance estimator for data that are collected with a complex sample design has been discussed by many authors [25][26][27][28]. Although unbiased estimation cannot be guaranteed theoretically, MI methods seem to work well in the practice [25,29]. In our analysis, the imputation model includes the stratification variables age, sex and region as explanatory variables.…”

Section: Statistical Methods For Estimation Of Health Indicatorsmentioning

confidence: 98%

“…As these are available for the full cohort, the imputation based estimates can be benchmarked against the real data [29,31]. In the imputation, it is assumed that the deaths and hospital visits are available for participation groups 1 and 2 but missing for participation group 3 and the imputation model is similar to the model used for the questionnaire variables.…”

Section: Alternative Statistical Methods For Estimation Of Health Indmentioning

confidence: 99%

See 1 more Smart Citation

Selection bias was reduced by recontacting nonparticipants

Karvanen

Tolonen

Härkänen

et al. 2016

Journal of Clinical Epidemiology

Self Cite

View full text Add to dashboard Cite

Objective: One of the main goals of health examination surveys (HES) is to provide unbiased estimates of health indicators at the population level. We demonstrate how multiple imputation methods may help to reduce the selection bias if partial data on some non-participants are collected. Study Design and Setting: In the FINRISK 2007 study, a populationbased health study conducted in Finland, a random sample of 10 000 men and women aged 25-74 were invited to participate. The study included a questionnaire data collection and a health examination. 6255 individuals participated in the study. Out of 3745 non-participants, 473 returned a simplified questionnaire after a recontact. Both the participants and the nonparticipants were followed up for death and hospitalizations. The follow-up data allowed to check the assumptions on the missing data mechanism and tailored multiple imputation methods were used to handle the missing data. Results: Non-participation is a strong predictor for mortality in the fiveyear follow-up. However, the recontact response does not predict mortality or morbidity among the non-participants when adjusted for age and sex. The M A N U S C R I P T A C C E P T E D ACCEPTED MANUSCRIPT 2 result suggests that the recontact respondents can be used as proxy for all non-participants. A comparison of raw estimates and estimates adjusted for selection bias reveals clear differences in the estimated population prevalences of smoking and heavy alcohol usage. Conclusion: All efforts to collect data on non-participants are likely to be useful even if the response rate for the recontact remains low. Statistical analysis of the recontact respondents provides an indication of the extent of the selection bias, even in studies where follow-up data are not available to check the assumptions.Keywords: survey; non-response; bias; missing data; multiple imputation What is new?:• Register data on cause specific mortality and morbidity can be used to check the assumptions on the missing data mechanism.• In the FINRISK 2007 study, we found that while the participants and the non-participants clearly differ, the non-participants with and without a recontact response have similar mortality and morbidity when adjusted for age and sex.• We propose a multiple imputation approach to handle data missing not at random. The approach is applicable when data on a non-response questionnaire are available.

show abstract

Section: Statistical Methods For Estimation Of Health Indicatorsmentioning

confidence: 98%

Section: Alternative Statistical Methods For Estimation Of Health Indmentioning

confidence: 99%

Selection bias was reduced by recontacting nonparticipants

Karvanen

Tolonen

Härkänen

et al. 2016

Journal of Clinical Epidemiology

Self Cite

View full text Add to dashboard Cite

show abstract

“…In addition to the previously mentioned factors, there might have been some other residual confounding that we have not been able to define. In addition, one limitation of our study is that the participation rates were lower in H2011 than in H2000, especially among men, those in younger age groups, and those with a lower educational status (36). To handle the missing data, we used both baseline and updated H2011 weights, which take into account the increased nonparticipation (36).…”

Section: Methodologic Issuesmentioning

confidence: 99%

“…The baseline and updated weights for H2011 were used for the corresponding survey data sets (36). Weighted model-adjusted mean and prevalence estimates were based on predictive margins (37).…”

Section: Methodsmentioning

confidence: 99%

The positive impact of general vitamin D food fortification policy on vitamin D status in a representative adult Finnish population: evidence from an 11-y follow-up based on standardized 25-hydroxyvitamin D data

Jääskeläinen

Itkonen

Lundqvist

et al. 2017

The American Journal of Clinical Nutrition

Self Cite

189

203

View full text Add to dashboard Cite

“…The moderate response rate can cause selection bias. However, the response rate at baseline was high (92%), allowing for a quite accurate imputation of the missing outcome and risk factor values in 2011 31. As our results were based on an observational study, the projections based on different scenarios do not necessarily reflect the causal effects of risk factor changes, which would require incorporation of randomised controlled trials and evidence synthesis 43 44…”

Section: Discussionmentioning

confidence: 97%

Projecting long-term trends in mobility limitations: impact of excess weight, smoking and physical inactivity

Härkänen

Sainio

Stenholm

et al. 2019

J Epidemiol Community Health

Self Cite

View full text Add to dashboard Cite

BackgroundPolicy makers need disability projections for planning adequate services and measures for health promotion. The aim of this study is to provide projections on severe mobility limitations up to year 2044 and illustrate how the projected prevalence and the number of persons with severe mobility limitations are affected by potential changes in the modifiable risk factors, namely excess weight, physical inactivity and smoking.MethodsWe analysed the nationally representative, repeated measures Health 2000 and 2011 Surveys (BRIF8901) with 8615 and 6740 participants, respectively, aged 18 years and older. Severe mobility limitations were defined as major difficulties or unable to walk about half a kilometre. We applied a multistate model on repeated measures to account for both individual risk factors and their changes over time.ResultsThe number of people with severe mobility limitations was projected to double by the year 2044 in Finland, due to the rapid ageing of the population. Eliminating half of the excess weight would reduce their number by one-fifth, while reductions in the prevalence of smoking and physical inactivity would have a minor impact. Even if excess weight, smoking and physical inactivity were completely eliminated, the number of persons with severe mobility limitations is projected to increase.ConclusionsDesigning and implementing strategies to promote healthy weight are important to slow down the rapid increase in mobility limitations due to population ageing. Providing adequate health and social services for the increasing population with disabilities will nevertheless be an increasing national challenge.

show abstract

Systematic handling of missing data in complex study designs – experiences from the Health 2000 and 2011 Surveys

Cited by 55 publications

References 54 publications

Selection bias was reduced by recontacting nonparticipants

Selection bias was reduced by recontacting nonparticipants

The positive impact of general vitamin D food fortification policy on vitamin D status in a representative adult Finnish population: evidence from an 11-y follow-up based on standardized 25-hydroxyvitamin D data

Projecting long-term trends in mobility limitations: impact of excess weight, smoking and physical inactivity

Contact Info

Product

Resources

About