A national, prospective, open-label, non-interventional, multicentre, controlled study designed to evaluate a diagnostic method in patients undergoing hip, knee or shoulder prosthetic removal regardless of whether or not infection is suspected. Serological test results will have no impact on the therapeutic approach.

Study objective: To evaluate the diagnostic performance of an antibody detection kit for the serological diagnosis of bone and joint prosthetic device infections cause by Staphylococci, Streptococci, Propionibacterium acnes (P. acnes) and Gram-negative bacteria versus the reference method: the bacterial culture of deep intraoperative samples.

Inclusion Criteria:

1. Male or female patients
2. Patients 18 years of age or older
3. Patient with a bone and joint prosthetic device: total hip prosthesis (THP) or total knee prosthesis (TKP) or total shoulder prosthesis (TSP)
4. Patients with a revision of a failed prosthesis, regardless of the cause and whether septic or not
5. Patients who have not expressed their opposition to the use of their personal medical data and blood samples

Exclusion Criteria:

1. Patients already enrolled or clinically reviewed following a relapse
2. Patients with several prostheses requiring at least two revisions during the same surgical procedure
3. Patients for whom the microbiologist and/or doctor responsible for the care refuses that these patients participate in the study
4. HIV+ patients
5. Patients undergoing chemotherapy for a blood disease or solid tumour
6. Patients under guardianship or trusteeship
7. Patients who are expected to be difficult to monitor (life expectancy less than 1 year, or homeless people, foreigner in transit)

The septic patient group will be comprised of patients presenting with bone and joint prosthetic device infection confirmed by intraoperative microbiological culture (at least 2 deep positive samples for the same bacterial strain).

The non-septic patient group will be comprised of prosthetic patients presenting with symptoms of mechanical loosening, and whose deep intraoperative samples have all proved negative.