*2.3. Definitions*

Periprosthetic shoulder infection was diagnosed according to the last proposed definition criteria of the ICM [10]. According to these criteria, patients were classified into 4 infection subgroups: (1) definitive infection; (2) probable infection; (3) possible infection; (4) infection unlikely. Meeting one of the following criteria was diagnostic of definitive periprosthetic shoulder infection: (1) a sinus tract communicating with the prosthesis; (2) gross intra-articular pus; (3) two positive cultures with phenotypically identical virulent organisms. In the lack of these defining signs, weighted minor criteria (Table 1) are summed and used to distinguish between probable, possible, and unlikely infection.

The three categories in these less distinct scenarios are defined as follows:

	- Fewer than six.





1 Statistical analysis was only undertaken between reimplantation and no reimplantation groups. \* The values are given as the mean and the standard deviation. The values are given as the number with the percentage of the group in parentheses. CCI—Charlson comorbidity index.

> Of the 49 infected patients, 16 met the criteria for definitive infection, 25 for probable infection, and 8 for possible infection. Cutibacterium acnes was the most common infecting microorganism at the time of resection arthroplasty in 18/49 patients (37%), followed by coagulase-negative staphylococci (18/49, 37%), Staphylococcus aureus (8/49, 16%), and other microorganisms (9/49, 18%). In 14 of 49 cases (29%), a polymicrobial infection was evident and eight patients (16%) had no growth in the microbiology. Three of these eight patients with negative microbiology had a definitive infection due to presence of gross intraarticular pus and antibiotic treatment was started before taking samples, as patients were in sepsis. In the remaining five patients, the infection was classified as possible due to presence of minor criteria.

> The definition for successfully treated shoulder PJI, in terms of infection eradication, was based on the Dephi-based international multidisciplinary consensus [11] and was further modified [12,13]. Infection was considered as eradicated if all of the following criteria were fulfilled at the latest follow-up: infection eradication, characterized by a healed wound without fistula and drainage; no recurrence of the infection; no occurrence of periprosthetic joint infection-related mortality; no subsequent surgical intervention for infection after reimplantation surgery; no long-term (>6 months) antimicrobial suppression

therapy. Given that the Delphi criteria do not consider patients who do not undergo the reimplantation stage of the two-stage exchange arthroplasty, in this study successful infection eradication also included no subsequent surgical intervention for infection after explantation and no mortality related to infection in patients who did not undergo the reimplantation stage [8].

#### *2.4. Two-Stage Exchange Arthroplasty Approach*

All patients were treated according to a standardized two-stage exchange protocol. The first stage consisted of removal of all implants, as well as infected tissue, cement, and all other foreign material followed by irrigation and debridement. In most cases an antibiotic-impregnated cement spacer was inserted. Tissue cultures were incubated for 14 days. Antibiotic treatment was started intravenously (IV) after surgery or preoperatively in the case of patients presenting with sepsis after synovial aspiration. A standardized antimicrobial treatment was applied in every case based on a previously published concept under the surveillance of our infectious disease specialists [14]. A revision with irrigation, debridement, and concomitant spacer exchange was performed in case of a persistent infection (discharging wound and/or local sings of infection and/or increasing CRP without any other focus). A reimplantation was performed when the operation site was healed, with soft tissues in a good condition and ready for surgery, and the general health status of the patient was suitable for surgery. The reimplantation was used as another chance to execute another debridement of the surrounding soft tissues and bone before reimplantation of the definitive components. Intravenous antibiotic treatment was given for 2 weeks after reimplantation surgery and changed to oral regime mostly for another 4 weeks to complete a total treatment duration of 6 weeks after reimplantation.
