Article Text

Download PDFPDF

SAT0666 Sample size for RA clinical trials using ultrasound outcome measures may be reduced by novel joint selection methods: a pilot study
  1. JC Allen Jr1,
  2. J Thumboo2,3,4,
  3. WK Lye1,
  4. PG Conaghan5,6,
  5. LC Chew2,3,4,
  6. YK Tan2,3,4
  1. 1Office of Clinical Sciences, Centre for Quantitative Medicine, Duke-NUS Medical School
  2. 2Yong Loo Lin School of Medicine, National University of Singapore
  3. 3Duke-NUS Medical School
  4. 4Department of Rheumatology and Immunology, Singapore General Hospital, Singapore, Singapore
  5. 5NIHR Leeds Musculoskeletal Biomedical Research Unit
  6. 6Leeds Institute of Rheumatic and Musculoskeletal Medicine, University of Leeds, Leeds, United Kingdom


Background Novel outcome measures selecting a reduced joint count for ultrasonography can be highly responsive in demonstrating the improvement in joint inflammation seen in rheumatoid arthritis (RA) patients on treatment [1].

Objectives To determine whether the use of the novel methods can translate into smaller sample sizes for subject recruitment into RA clinical trials. Results from the existing methods are used for comparison.

Methods 24 RA patients with treatment starts or escalation had clinical and ultrasound joint assessment at baseline and 3 months. The novel methods select joints based on (A) ultrasound joint findings (i.e. Individualized Ultrasound (IUS) method) or (B) a composite of ultrasound and clinical joint findings (i.e. Individualized Composite Ultrasound (ICUS) method). In contrast, the existing methods utilize pre-determined joint sites for ultrasonography. Scores at the relevant joints per patient are summed up to obtain the total inflammatory score (TIS). The effect size (ES) was measured as the mean change of the TIS divided by the standard deviation of the change in the TIS. Sample sizes were calculated from confidence intervals (CIs) on ES that reflect uncertainty in estimating ES. For a given CI on ES, sample sizes are computed as the minimum number of patients required to provide ≥80% power at α =0.05 for rejecting the null hypothesis (defined as no difference in the 3-month mean change in TIS comparing novel versus existing methods).

Results Based on the 95% CI analysis, sample sizes using existing joint assessment methods in conjunction with the 12-joint approach ranged from 10 to 234. The corresponding sample sizes using the ICUS method with the 12-joint approach ranged from 7 to 39, and using the IUS method with the 12-joint approach ranged from 6 to 37. The corresponding sample sizes using the ICUS method with the 7-joint approach ranged from 6 to 24, and using the IUS method with the 7-joint approach ranged from 6 to 35.

Table 1.

Summary statistics for novel versus existing methods on 3-month change in scores

Conclusions Our findings strongly suggest that novel ultrasound joint selection methods result in smaller sample size requirements compared to existing methods, and provide justification for larger studies to confirm these observations.


  1. Tan YK et al. Novel Ultrasound Joint Selection Methods Using a Reduced Joint Number Demonstrate Inflammatory Improvement when Compared to Existing Methods and Disease Activity Score at 28 Joints. J Rheumatol. 2016;43:34–7.


Disclosure of Interest None declared

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.