Send to:

Choose Destination
See comment in PubMed Commons below
Curr Opin Support Palliat Care. 2012 Dec;6(4):465-70. doi: 10.1097/SPC.0b013e328358441d.

Accommodation of missing data in supportive and palliative care clinical trials.

Author information

  • 1Fred Hutchinson Cancer Research Center, Seattle, Washington 98019, USA.



Clinical trials to evaluate the supportive and palliative care treatments have some different missing data concerns than the other clinical trials. This study reviews the literature on missing data as it may apply to these trials.


Prevention of missing data through study design and conduct is a recent area of focus. Missing data can be minimized by simplifying trial participation for patients, their caregivers, and trialists. Run-in periods with active drug or collecting data from observer (proxy) respondents may complicate a trial but may be used to address some specific concerns. Many analyses can accommodate data missing because of nonresponse by multiple imputation, using carefully chosen imputation models. Analysis of trials evaluating end-of-life care should distinguish between missing data and truncation because of death.


Likely patterns for missing data should be discussed when planning a clinical trial, as modifications to trial design can minimize missing data while still addressing study aims. Many statistical analysis methods are available to accommodate missing data, but robustness of study conclusions to assumptions about mechanisms underlying the missingness should be evaluated by sensitivity analyses.

[PubMed - indexed for MEDLINE]
PubMed Commons home

PubMed Commons

How to join PubMed Commons

    Supplemental Content

    Full text links

    Icon for Lippincott Williams & Wilkins
    Loading ...
    Write to the Help Desk