Skip to Content
Interactive Textbook on Clinical Symptom Research Logo


Home Button

Clinical Trials in TMD Sections
Author Bio
Introduction
The Biopsychosocial Model
Designing Multicenter RCTs
Players in an RCT
Randomization
Trial Design Quality
TMD Case Definition
Endpoints and Outcome Measures
Blinding & Masking
Study Sample Size
Number and Nature of Interventions
Study Length and Follow up
Currently selected section: Intent-to-treat Analyses and Sample Size
Compliance
Multicenter RCTs
Implementing RCTs: Practical Issues
Analysis of TMD Trials
Conclusions
Acknowledgments
Appendix A
Appendix B

 

Chapter 22: Clinical Trials in Temporomandibular: Intent-to-treat Analyses and Sample Size
        

Gathering Key Data on All Subjects

Intention-to-treat analyses, the stringent requirement that data is analyzed from all patients who have been randomized into an RCT raises -- whether or not they completed the trial -- some special issues in clinical trials of TMD treatment.

Obviously, if patients drop out of an RCT after being randomized but before all follow up data collection is completed, the final sample sizes will be smaller and the possibility arises that statistical analyses will be less valid. Some strategies to reduce loss to follow up or missing data on critical measures include gathering data at least on the primary outcome measures over time, even if the patient will not submit to full batteries of repeated questionnaires.

Thus, it is recommended that investigators obtain permission from TMD patients at the onset of RCTs, before randomization, to gather at least a small amount of data even if the patient doesn't allow full data accumulation. If the primary outcome measure is pain, for instance, then attempts are made to insure that VAS pain levels (and possibly one or two additional, critical variables) can be gathered at each data collection point included in the RCT study design. Such permission is usually more reliably obtained at the onset of the RCT rather than later in the trial when the patient drops out and may become lost, not only to follow up but to any possibility of contact.

 


Page 47 of 81
      Previous Page