AUTHOR=Abreu Daisy , Ware Jennifer , Georgiou-Karistianis Nellie , Leavitt Blair R. , Fitzer-Attas Cheryl J. , Lobo Raquel , Fernandes Ana Raquel , Handley Olivia , Anderson Karen E. , Stout Julie C. , Sampaio Cristina
TITLE=Utility of Huntington's Disease Assessments by Disease Stage: Floor/Ceiling Effects
JOURNAL=Frontiers in Neurology
VOLUME=12
YEAR=2021
URL=https://www.frontiersin.org/journals/neurology/articles/10.3389/fneur.2021.595679
DOI=10.3389/fneur.2021.595679
ISSN=1664-2295
ABSTRACT=Introduction: An understanding of the clinimetric properties of clinical assessments, including their constraints, is critical to sound clinical study and trial design. Utilizing data from Enroll-HD—a global, prospective HD observational study and clinical research platform—we examined several well-established HD clinical assessments across all stages of disease for evidence of instrument constraints, specifically floor/ceiling effects, to inform selection of appropriate instruments for use in future studies/trials and identify gaps in instrument utility over the life-course of the disease.
Material and Methods: Analyzing publicly available data from 6,614 HD gene-expansion carriers (HDGECs), we grouped participants into deciles based on baseline CAP score, which ranged from 26 to 229. We used descriptive statistics to characterize data distribution for 25 outcome measures (encompassing motor, function, cognition, and psychiatric/behavioral domains) in each CAP decile. A skewness statistic threshold of ±2 was defined a priori to indicate floor/ceiling effects.
Results: We found evidence of floor/ceiling effects in the early premanifest stages of disease for most motor and function assessments (e.g., TMS, TFC) and select cognitive tasks (MMSE, Trail Making tests). Other cognitive assessments, and the HADS-SIS scales, performed well ubiquitously, with no evidence of floor/ceiling effects at any disease stage. Floor/ceiling effects were evident at every disease stage for certain assessments, including PBA-s measures. Ceiling effects were apparent for DCL from onset stages onwards, as expected.
Discussion: Developing instruments sensitive to subtle differences in performance at the earlier stages of the disease spectrum, particularly in motor and function domains, is warranted.