Optimizing Results From pQCT Reliability of Operator-Dependent pQCT Variables in Cadavers and Humans With Low Bone Mass Maureen C. Ashe, 1,2 Teresa Liu-Ambrose, 1,2 Karim M. Khan, 1,2,3 Neil White, 4 and Heather A. McKay *,1,2,3 1 University of British Columbia, Vancouver, BC, Canada; 2 British Columbia Women’s and Children’s Hospital; 3 Vancouver Coastal Health Research Institute; and 4 University of Calgary Abstract Peripheral quantitative computed tomography (pQCT) can assess bone geometric properties and separate cortical from trabecular bone. Despite pQCT’s potential benefits for research, most reliability and accuracy studies have used a constant acquisition and analysis protocol. There are, however, numerous steps in the pQCT scan acquisition and analysis that are operator dependent. Whether or not these influence the quality of the pQCT scans and, potentially, the precision and validity of the data collected has been little explored. We investigated how pQCT outputs changed when operator-dependent parameters were varied, particularly when the bone of interest was of low mineral density. We found that bone parameters and scan failure rate varied significantly depending on the acquisition resolution; only one scan slice at the 10 and 30% radius is required to maintain adequate precision, and reference lines for sites should use a reproducible landmark. These results provide a foundation for recommending scan acquisition and analysis options for patients with low bone mass. Key Words: Peripheral quantitative computed tomography; reliability; radius. Received 09/07/04; Revised 01/21/05; Accepted 01/21/05. *Address Correspondence to: Heather A. McKay; University of British Columbia, Department of Orthopaedics and Department of Family Practice, Room 588, 828 West 10th Avenue,Vancouver, BC, Canada. V5Z 1L8. E-mail: mckayh@interchange.ubc.ca 335 Original Article Introduction Peripheral quantitative computed tomography (pQCT) per- mits researchers to investigate bone geometric properties and the relative contribution of cortical and trabecular bone in humans. Although there have been a number of accuracy and precision studies of pQCT in animals (1,2) and humans (3–5), there are few data on the accuracy and precision of the most commonly used pQCT, the Norland/Stratec XCT 2000/3000 (Stratec Medizintechnic GmbH, Pforzheim, Germany) (6–9). Of great importance, most precision studies have used a constant analysis and acquisition protocol, whereas a number of steps in the Norland Stratec pQCT scan acquisition are operator dependent. These include (1) resolution/ voxel size, (2) whether to obtain a single slice or multiple slices at each region of interest, (3) locat- ing the anatomical reference line, (4) participant positioning, and (5) the approach to including the analysis of scans that contain image artifacts. At the time of commencing this study, we could find no studies that informed us as to which options among these protocols were most reliable. Furthermore, precision studies should ideally be undertaken in the type of bone that will be investigated in future studies. There have been very few reports of pQCT reliability in populations with bone compromised by disuse or osteoporosis. These patients might provide the great- est challenges to those seeking high-quality reproducible scans. Therefore, the purpose of this methodological investigation was to investigate pQCT outputs when several technical aspects of the scan were altered. Specifically, we report the variability that resulted from (1) changing resolution, (2) using single or multiple slices for analysis, and (3) changing the scout view reference line. Also, we (4) quantify failure of scan Journal of Clinical Densitometry, vol. 8, no. 3, 335–340, 2005 © Copyright 2005 by Humana Press Inc. All rights of any nature whatsoever reserved. 1094-6950/05/8:335–340/$30.00 doi: 10.1220/1094-6950