Description
Data was collected between December 1st 2012 through December 31st 2014 by Tissue Bank A for the purpose of evaluating the risk factors for musculoskeletal (MS) tissue allograft microbial contamination. Two data sets were created for this study. One data set had any microbial growth as the outcome variable and the second data set included only those allografts with microbial contamination which was divided into low and high pathogenicity. A total of 12,780 allografts were evaluated, which were procured from 805 tissue donors. The study subject was the allograft tissue. A generalized linear mixed model, with the donor as the mixed effects term, was used to analyze the relationship between autopsy status and musculoskeletal allograft contamination while controlling for BMI, number of technicians present on a procurement, skin recovery prior to MS tissue procurement, prior organ recovery, trauma present at the time of death, and recovery environment. The risk of any microbial contamination in allografts recovered from donors who had an autopsy prior to MS tissue recovery was increased by 280% relative to those allografts recovered from donors without an autopsy prior, after controlling for skin procurement in the model (p=0.030; RR: 2.8; CI:1.10, 7.12). The risk of any microbial contamination in allografts recovered from donors following a skin recovery was increased by 192% relative to those allografts recovered from donors without a skin procurement prior after adjusting for autopsy status in the model (p=0.011; RR: 1.92; CI:1.16, 3.15). Regarding the dataset with low and high virulence organisms, the risk of high virulence contamination in MS allografts recovered from donors post autopsy was reduced by 75% relative to allografts recovered from donors without an autopsy prior, after controlling for trauma present at the time of death (p=0.583; RR:0.75; CI:27, 2.10). This result was not statistically significant. The risk of high virulence contamination in allografts recovered from donors with trauma present at the time of death occurred 26.6% less relative to those allografts recovered from donors that did not have trauma, after controlling for autopsy status (p=0.024; RR:0.27; CI:0.08, 0.84). This result was statistically significant.
