Radiostereometric analysis (RSA) is an analytical technique in which biplanar radiographs are used to quantify the migration of implants in vivo. RSA is now considered the benchmark for quantifying implant migration in clinical studies of total joint replacement, yet its use in preclinical animal models has not been widely reported. The same attributes that make RSA appealing as an analytical method in humans (i.e" high precision and accuracy, noninvasiveness, objectivity) also make it a promising option for animal studies. The specific aims of this study were to determine the technical feasibility and analytical precision of RSA in a canine model of cemented total hip replacement. The precision of RSA was assessed in (1) a Plexiglas phantom, (2) a canine Sawbone model, and (3) a pilot series of dogs implanted with cemented canine THR implants. In vitro precision values, calculated as the 95% confidence limits for the error between duplicate RSA examinations, ranged from 4.3 to 17.9 mum for translation and from 0.01degrees to 0.22degrees for rotation. In vivo precision values ranged from 16.2 to 41.1 mum for translation and 0.17degrees to 0.44degrees for rotation. As is the case in humans, RSA appears to provide an order-of-magnitude improvement in technical precision as compared to plain film X-ray. RSA can therefore be considered a practical and potentially valuable noninvasive outcome measure for assessing implant function in canine cemented THR.
