The presence or absence of prosthetic graft incorporation with surrounding tissue was assessed relative to bacterial culture results, using enhanced microbiologic culture techniques.
University and Veterans Affairs hospital.
Prosthetic samples were removed from 113 aortofemoral, extra-anatomic, infrainguinal, and hemoaccess sites at the time of vascular reoperative surgery. Harvested grafts were sonicated. Density of organisms was determined by quantitative culture.
Main Outcome Measures:
The culture result was predicted from the status of prosthetic incorporation or disincorporation as determined at surgery. For purposes of this study, any bacterial growth represented graft infection.
Cultures positive for bacteria were obtained from 31 sites; cultures with no growth, from 82. Thirty-one of the 113 sites were disincorporated, of which 23 yielded cultures positive for bacteria, and eight, no growth. The remaining 82 sites were well incorporated, of which 74 yielded cultures negative for bacteria, and eight, bacterial growth. Sixteen (14%) incorrect predictions were noted. The concurrence of disincorporation and a culture positive for bacteria relative to all culture-positive grafts (sensitivity) was 74%. The concurrence of incorporation and cultures negative for bacteria relative to all culture-negative grafts (specificity) was 90% in prostheses implanted for longer than 2 weeks; in prostheses implanted for longer than 12 weeks, specificity was 97%.
The surgical finding of incorporation or disincorporation accurately predicted the culture result in 89% of the sites. Disincorporation correlated with presence of bacteria in 71%; incorporation reliably excluded the presence of bacteria in 97%.(Arch Surg. 1995;130:183-187)
Padberg FT, Smith SM, Eng RHK. Accuracy of Disincorporation for Identification of Vascular Graft Infection. Arch Surg. 1995;130(2):183–187. doi:10.1001/archsurg.1995.01430020073013
Customize your JAMA Network experience by selecting one or more topics from the list below.
Create a personal account or sign in to: