Imputation strategies for missing continuous outcomes in cluster randomized trials

Imputation strategies for missing binary outcomes in cluster randomized trials

Background

Attrition, which leads to missing data, is a common problem in cluster randomized trials (CRTs), where groups of patients rather than individuals are randomized. Standard multiple imputation (MI) strategies may not be appropriate to impute missing data from CRTs since they assume independent data. In this paper, under the assumption of missing completely at random and covariate dependent missing, we compared six MI strategies which account for the intra-cluster correlation for missing binary outcomes in CRTs with the standard imputation strategies and complete case analysis approach using a simulation study.

Method

We considered three within-cluster and three across-cluster MI strategies for missing binary outcomes in CRTs. The three within-cluster MI strategies are logistic regression method, propensity score method, and Markov chain Monte Carlo (MCMC) method, which apply standard MI strategies within each cluster. The three across-cluster MI strategies are propensity score method, random-effects (RE) logistic regression approach, and logistic regression with cluster as a fixed effect. Based on the community hypertension assessment trial (CHAT) which has complete data, we designed a simulation study to investigate the performance of above MI strategies.

Results

The estimated treatment effect and its 95% confidence interval (CI) from generalized estimating equations (GEE) model based on the CHAT complete dataset are 1.14 (0.76 1.70). When 30% of binary outcome are missing completely at random, a simulation study shows that the estimated treatment effects and the corresponding 95% CIs from GEE model are 1.15 (0.76 1.75) if complete case analysis is used, 1.12 (0.72 1.73) if within-cluster MCMC method is used, 1.21 (0.80 1.81) if across-cluster RE logistic regression is used, and 1.16 (0.82 1.64) if standard logistic regression which does not account for clustering is used.

Conclusion

When the percentage of missing data is low or intra-cluster correlation coefficient is small, different approaches for handling missing binary outcome data generate quite similar results. When the percentage of missing data is large, standard MI strategies, which do not take into account the intra-cluster correlation, underestimate the variance of the treatment effect. Within-cluster and across-cluster MI strategies (except for random-effects logistic regression MI strategy), which take the intra-cluster correlation into account, seem to be more appropriate to handle the missing outcome from CRTs. Under the same imputation strategy and percentage of missingness, the estimates of the treatment effect from GEE and RE logistic regression models are similar.

Quantifying the impact of fixed effects modeling of clusters in multiple imputation for cluster randomized trials

In cluster randomized trials (CRTs), identifiable clusters rather than individuals are randomized to study groups. Resulting data often consist of a small number of clusters with correlated observations within a treatment group. Missing data often present a problem in the analysis of such trials, and multiple imputation (MI) has been used to create complete data sets, enabling subsequent analysis with well-established analysis methods for CRTs. We discuss strategies for accounting for clustering when multiply imputing a missing continuous outcome, focusing on estimation of the variance of group means as used in an adjusted t-test or ANOVA. These analysis procedures are congenial to (can be derived from) a mixed effects imputation model; however, this imputation procedure is not yet available in commercial statistical software. An alternative approach that is readily available and has been used in recent studies is to include fixed effects for cluster, but the impact of using this convenient method has not been studied. We show that under this imputation model the MI variance estimator is positively biased and that smaller intraclass correlations (ICCs) lead to larger overestimation of the MI variance. Analytical expressions for the bias of the variance estimator are derived in the case of data missing completely at random, and cases in which data are missing at random are illustrated through simulation. Finally, various imputation methods are applied to data from the Detroit Middle School Asthma Project, a recent school-based CRT, and differences in inference are compared.

Improving the care for people with acute low-back pain by allied health professionals (the ALIGN trial): A cluster randomised trial protocol

BACKGROUND

Variability between clinical practice guideline recommendations and actual clinical practice exists in many areas of health care. A 2004 systematic review examining the effectiveness of guideline implementation interventions concluded there was a lack of evidence to support decisions about effective interventions to promote the uptake of guidelines. Further, the review recommended the use of theory in the development of implementation interventions. A clinical practice guideline for the management of acute low-back pain has been developed in Australia (2003). Acute low-back pain is a common condition, has a high burden, and there is some indication of an evidence-practice gap in the allied health setting. This provides an opportunity to develop and test a theory-based implementation intervention which, if effective, may provide benefits for patients with this condition.

AIMS

This study aims to estimate the effectiveness of a theory-based intervention to increase allied health practitioners' (physiotherapists and chiropractors in Victoria, Australia) compliance with a clinical practice guideline for acute non-specific low back pain (LBP), compared with providing practitioners with a printed copy of the guideline. Specifically, our primary objectives are to establish if the intervention is effective in reducing the percentage of acute non-specific LBP patients who are either referred for or receive an x-ray, and improving mean level of disability for patients three months post-onset of acute LBP.

METHODS

The design of the study is a cluster randomised trial. Restricted randomisation was used to randomise 210 practices (clusters) to an intervention or control group. Practitioners in the control group received a printed copy of the guideline. Practitioners in the intervention group received a theory-based intervention developed to address prospectively identified barriers to practitioner compliance with the guideline. The intervention primarily consisted of an educational symposium. Patients aged 18 years or older who visit a participating practitioner for acute non-specific LBP of less than three months duration over a two-week data collection period, three months post the intervention symposia, are eligible for inclusion. Sample size calculations are based on recruiting between 15 to 40 patients per practice. Outcome assessors will be blinded to group allocation.

TRIAL REGISTRATION

Australian New Zealand Clinical Trials Registry ACTRN12609001022257 (date registered 25th November 2009).

Community based lifestyle intervention for blood pressure reduction in children and young adults in developing country: cluster randomised controlled trial

OBJECTIVE

To assess the effectiveness of a community based lifestyle intervention on blood pressure in children and young adults in a developing country setting.

DESIGN

Cluster randomised controlled trial.

SETTING

12 randomly selected geographical census based clusters in Karachi, Pakistan.

PARTICIPANTS

4023 people aged 5-39 years.

INTERVENTION

Three monthly family based home health education delivered by lay health workers.

MAIN OUTCOME MEASURE

Change in blood pressure from randomisation to end of follow-up at 2 years.

RESULTS

Analysed using the intention to treat principle, the change in systolic blood pressure (adjusted for age, sex, and baseline blood pressure) was significant; it increased by 1.5 (95% confidence interval 1.1 to 1.9) mm Hg in the control group and by 0.1 (-0.3 to 0.5) mm Hg in the home health education group (P for difference between groups=0.02). Findings for diastolic blood pressure were similar; the change was 1.5 mm Hg greater in the control group than in the intervention group (P=0.002).

CONCLUSIONS

Simple, family based home health education delivered by trained lay health workers significantly ameliorated the usual increase in blood pressure with age in children and young adults in the general population of Pakistan, a low income developing country. This strategy is potentially feasible for up-scaling within the existing healthcare systems of Indo-Asia.

TRIAL REGISTRATION

Clinical trials NCT00327574.

Preventing bias in cluster randomised trials

Bruno Giraudeau and Philippe Ravaud discuss the difficulties in preventing selection bias and applying intention-to-treat analysis in cluster randomized trials, and propose some solutions.