A single site, multi-operator precision study for second-generation HR-pQCT

As HR-pQCT increases in popularity for studying bone dynamics, it becomes increasingly important to quantify the factors that impact rigor and reproducibility in research, particularly for longitudinal studies. Previously reported data for HR-pQCT precision primarily use first-generation HR-pQCT, and focus on single operators, with narrow study populations that reflect the population of a larger research question. This study at a single academic imaging center with standardized, single-analyst post-processing was designed to investigate how operator characteristics in scan acquisition and participant demographics influence measurement precision. In total, 45 adults (58% female; 22-82 yr; BMI 18.4-44.8 kg/m²) underwent same-day repeat HR-pQCT scanning of the distal tibia and radius, with 3 imaging technologists each acquiring 15 participant scan pairs, and registration was applied in post-processing to accurately depict the standard workflow at the center. The root-mean-square coefficient of variation was ≤1.3% for BMD measures and <2% for area measures in both the radius and tibia. Trabecular microarchitecture ranged from 1.22% to 2.24% in the radius and from 1.14% to 2.82% in the tibia, while cortical thickness precision was calculated as 1.99% in the radius and 1.57% in the tibia. Cortical porosity was much higher, at 24.66% in the radius and 20.90% in the tibia. Significant differences were not found when analyzed by technologists, or in statistical modeling of participant demographics as potential covariates. While final image quality scores varied among technologists, it did not influence precision outcomes. These results offer reference values for the least significant change and demonstrate that standardized technologist training can yield consistent scan-rescan precision across operators.