A retrospective study will be performed to compare the incidence of revision procedures performed for the treatment of postoperative periprosthetic joint infection (PJI) between patients treated with a single preoperative antibiotic prophylaxis and patients treated with multiple shot antibiotic prophylaxis following primary arthroplasty of hip or knee.

This study will be conducted in two phases. In phase 1 the investigators will evaluate which antibiotic prophylaxes (type of antibiotic and duration of prophylaxis) are regularly used in the Netherlands, by contacting all orthopaedic centres implanting total hip and total knee prostheses. In phase 2 the investigators will compare infection rate (defined as revision for infection as registered in the LROI database) between a single-shot and multiple shot antibiotic prophylaxis.

Knee and hip primary arthroplasties will be considered separately.

In phase 1 of this study all Dutch hospitals and clinics performing hip and/or knee arthroplasty will be contacted to supply data concerning type and duration of antibiotic prophylaxis. The investigators will also ask whether debridement with implant retention for early infection is registered in the LROI. The investigators will ask permission to have access to unblended data on institutional level to be able to connect the type of antibiotic prophylaxis to outcome.

In phase 2 the investigators will include all adult patients whom have a primary hip or knee arthroplasty and are registered in the national joint implant registry (LROI) between 2011 and 2016.Those patients will be followed until they are revised (irrespectively of the indication), they died or the end-point of the study, i.e. end of 2016 (to allow at least one year follow-up for those operated late 2015).

Exclusion criteria are patients aged under 18. Patients with no primary procedure recorded in the LROI for the period of interest but records of revision surgeries will be excluded.

The investigators will compare the results found in the LROI database in two groups for both hip and knee arthroplasties. Group one will consist of patients receiving a single shot antibiotic prophylaxis, while group two will consist of patients receiving multiple shot antibiotic prophylaxis.

Analyses will be adjusted for known and available confounders such as age, gender, BMI, and ASA score.

In the LROI debridement without exchange of mobile parts for early infection, is only scarcely reported. The authors realise not all early infection treatments, such as debridement with or without exchange of mobile parts, are registered in the LROI database.

Inclusion Criteria:

- All adult patients whom have a primary hip or knee arthroplasty and are registered in the national joint implant registry (LROI) between 2011 and 2014

Exclusion Criteria:

• Patients aged under 18.
• Patients with no primary procedure recorded in the LROI for the period of interest but records of revision surgeries will be excluded.

All adult patients whom have a primary hip or knee arthroplasty and are registered in the national joint implant registry (LROI) between 2011 and 2014. Those patients will be followed until they are revised (irrespectively of the indication), they died or the end-point of the study, i.e. end of 2015 (to allow at least one year follow-up for those operated late 2014).