C.M. Wells, J. Feldman, I. Timmerman, J. Chow, W. Mihalko, M. Neel, J. Jennings, W. Haggard
Decreasing tissue damage and recovery time, while improving quality of life have been the focus of many approaches to total hip arthroplasty (THA). In this study, we compared two approaches, a tissue-sparing superior capsulotomy percutaneously assisted approach (SP) and the traditional posterior approach (TR), to address the question of whether the novel technique reduces tissue damage. The secondary aim of this study focused on the measurement technologies utilized to quantify the damage resulting from either SP or TR. Image J, BioQuant, and cellSens were the image analysis programs employed. Statistical validation and comparisons of results between all platforms were performed.
Both hips of freshly frozen cadaveric specimens (n = 8) were surgically prepared for THA with random procedure performed on left or right hip. All selected specimens had no prior implantation of devices to ensure all observed muscle damage occurred from the surgical technique. Surgeons resected tissue and performed necessary procedural steps up to device implantation. No devices were implanted during the study, as the aim was to quantify the damage caused by the incision and resection. After completion of the surgery, an independent surgeon (IS), who was blinded as to which method was performed on the specimen, excised the muscles and inspected areas of interest Assessment of the tissue damage was executed using a midsubstance cross-sectional area technique, validated by prior studies. High-resolution images of demarcated muscles were used for quantitative analysis. Three blinded independent reviewers quantified damaged tissue. The results were used to detect if statistically significant differences were present between the two methods. Furthermore, an independent reviewer using SPSS statistical software also assessed inter-program and inter-rater reliability.
The SP procedure significantly reduced percent of damaged GM tissue. GS tissue damage reduction with SP procedure was observed but the difference was not statistically significant. Between raters, the intraclass correlation coefficient (ICC) for the tissue damage measures was 0.870 (95% CI: 0.824 – 0.907). Within the individual applications, BioQuant and Image J had ICCs of 0.972 and 0.987, respectively. CellSens, however had an ICC value of 0.671.
For the three image analysis programs chosen, the damaged tissue was quantified within the software application and each user defined areas of interest slightly different, which reduced the interrater reliability. However, variation in the software may affect the degree of difference detected and/or the p-value. There was a statistically significant reduction in percent of damaged tissue, SP vs. TR, in the GM obtained from all programs. A statistically significant reduction in GS damage, SP vs. TR, was not detected in either program, which may be a result of the limited sample size.
New surgical techniques require evaluation to determine if there are objective advantages over other techniques. Preclinical evaluation of these techniques has been limited. Providing quantitative evidentiary support has clinical and scientific relevance to patients, physicians, and healthcare providers.. Independent analysis of these results may depend on the user interface of image analysis programs and should be evaluated by multiple raters to ensure accuracy.