Leukocyte esterase (LE) has shown to be an accurate marker of prosthetic joint infections and has been proposed as an alternative to frozen section (FS) for intra-operative diagnosis. In this study, intra-operative determination of LE was compared with FS for the diagnosis of periprosthetic hip infections.

One hundred and nineteen patients undergoing hip revision surgery due to prosthetic joint failure from June 2015 to December 2017 were considered. Joint fluids were collected before the arthrotomy for determination of LE which was performed by using enzymatic colorimetric strips. Four to six samples were stained with hematoxylin eosin for FS and considered suggestive for infection when at least 5 polymorphonuclear leukocytes in 5 fields at high power fields were found.

Sensitivity and specificity of LE were 100% and 93.8 %, respectively. The positive predictive value was 79.3 %, while the negative predictive value was 100%. Time from collection to response was 20.1 ± 4.4 minutes.

Sensitivity and specificity of FS were 83.3 % and 92.1 %, respectively. The positive and negative predictive values were 84.6 % and 97.1%. Time from sample collection response was 27.2 ± 6.9 minutes.

LE showed a higher sensitivity and a slightly lower specificity and the same diagnostic accuracy of intraoperative FS. The faster turnaround time (about 20 minutes from receiving of sample by the laboratory), its ease of use and the low costs make this test a valuable alternative to frozen sections and is going to replace FS in our clinical practice.

Reliability of microbiological diagnosis of prosthetic joint infection [PJI] strongly depends on the ability to dislodge microorganisms from biofilm and on the rate of contaminating samples during collection in the operating room and processing. The aim of a correct protocol is to avoid false negative and false positive results in order to adapt the correct therapy for each patient.

The object of the present study was to evaluate the impact of a novel closed bag system designed for samples collection and processing based on dithiothreitol (DTT), which is a sulfydryl compound able to remove bacteria from biofilm (MicroDTTect, 4i, Italy), on isolation of contaminant microorganisms in hip prostheses.

Specimens (prostheses, spacers, periprosthetic tissues) were aseptically collected according to a standard protocol into the device, which was transported to the laboratory for culture. Three different models of the system were prospectively evaluated, each being a development of the previous one. The first generation device consisted in an “open” system (DTT eluate was collected with a syringe and dispensed into sterile tubes), the second generation device in a “partially closed” system (DTT eluate collected directly in sterile vacuum tubes) and the third generation device in a “completely closed system” (DTT reservoir directly connected with sealed tubes inside the device). PJI was diagnosed following criteria established by MSIS.

The overall contamination rate, sensitivity and specificity of the first generation “open” system MicroDTTect were respectively 2.6% (1/39), 82.3% and 95.4% in 39 hips. The second generation “partially closed” device was characterized by a contamination rate of 1.96% (1/51), a sensitivity of 84% and a specificity of 96.1% in 51 hips. Contamination rate further decreased in the third generation “closed” system (1.89%, 2/106), while sensitivity (91.3%) and specificity (96.7%) improved in 106 hips. Differences have been also observed in hips (106) when compared to knees (70 cases) prosthetic infections (sensitivity 91.3% vs 89.3% and specificity 96.7% vs 100%).

Our data show as, thanks to its ease of use, low contamination rate and high sensitivity, MicroDTTect can represent a useful tool for improving the microbiological diagnosis of PJIs in hip revisions and has replaced sonication in our practice.