Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Estimating treatment effects from longitudinal clinical trial data with missing values: comparative analyses using

Patricia R Houck1, Sati Mazumdar, Tulay Koru-Sengul

  • 1Department of Psychiatry, University of Pittsburgh, Western Psychiatric Institute and Clinic, UPMC Health System, Thomas Detre Hall, 3811 O'Hara Street, Pittsburgh, PA 15213-2593, USA. houckpr@upmc.edu

Psychiatry Research
|December 14, 2004
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Retreatment with theta burst stimulation (TBS) for late life depression (LLD): A retrospective chart review.

Journal of psychiatric research·2023
Same author

Change in patient-centered outcomes of psychological well-being, sleep, and suicidality following treatment with intravenous ketamine for late-life treatment-resistant depression.

International journal of geriatric psychiatry·2023
Same author

Exploratory genome-wide analyses of cortical inhibition, facilitation, and plasticity in late-life depression.

Translational psychiatry·2023
Same author

Effect of prior pharmacotherapy on remission with sequential bilateral theta-burst versus standard bilateral repetitive transcranial magnetic stimulation in treatment-resistant late-life depression.

The British journal of psychiatry : the journal of mental science·2023
Same author

Brain-Cognition Associations in Older Patients With Remitted Major Depressive Disorder or Mild Cognitive Impairment: A Multivariate Analysis of Gray and White Matter Integrity.

Biological psychiatry·2023
Same author

Effects of antipsychotic medication on functional connectivity in major depressive disorder with psychotic features.

Molecular psychiatry·2023
Same journal

Association of TyG-RFM with depressive symptoms: a nationally representative study enhanced by interpretable machine learning.

Psychiatry research·2026
Same journal

Cross-national patterns of probable complex PTSD in post-communist Europe: A network analysis of five countries.

Psychiatry research·2026
Same journal

The overlooked majority of hikikomori: Latent profiles reveal a discrepancy between social anxiety and life functioning.

Psychiatry research·2026
Same journal

Promoting self-compassion meditation with intranasal oxytocin in borderline personality disorder: a pilot study.

Psychiatry research·2026
Same journal

Low-dose lithium vs valproate and incident dementia diagnoses in older adults: A propensity-matched cohort study using electronic health record data.

Psychiatry research·2026
Same journal

Association between depressive symptoms and hepatic steatosis in young adults: a nationwide study.

Psychiatry research·2026
See all related articles

Choosing the right method for analyzing clinical trial data with missing values is crucial. Flexible methods like Ignorable Maximum Likelihood (IML) and Multiple Imputation (MI) are favored over simpler approaches when data are not completely random.

Area of Science:

  • Clinical Trials
  • Biostatistics
  • Psychiatry

Background:

  • Estimating treatment effects in clinical trials with missing data is complex.
  • Common methods like Last Observation Carried Forward (LOCF) and Completers Only have unrealistic assumptions.
  • Ignorable Maximum Likelihood (IML) and Multiple Imputation (MI) require less stringent missing data assumptions (Missing At Random - MAR).

Purpose of the Study:

  • To compare the performance of four different methods for handling missing data in an intent-to-treat analysis.
  • To evaluate these methods in a randomized clinical trial assessing antidepressant effects in elderly patients.
  • To determine the impact of missing data mechanisms on treatment effect estimation.

Main Methods:

  • Applied four methods: Last Observation Carried Forward (LOCF), Completers Only, Ignorable Maximum Likelihood (IML), and Multiple Imputation (MI).

Related Experiment Videos

  • Utilized a mixed-effects model to analyze antidepressant treatment effects over time.
  • Compared explanatory and pragmatic analysis approaches, including piecewise splines and rescue medication.

    • Main Results:

    • Explanatory analysis showed significant treatment differences with LOCF and IML.
    • Statistical tests indicated a violation of the Missing Completely At Random (MCAR) assumption.
    • Pragmatic analysis using IML and MI, alongside other methods, revealed no significant treatment differences.

    Conclusions:

    • The choice of missing data handling method significantly impacts clinical trial results.
    • Flexible methods (IML, MI) are preferable when data are not Missing Completely At Random (MCAR).
    • Careful consideration of the missing data mechanism is essential for robust clinical trial analysis.