Both can be difficult to treat due to bacterial biofilm formation on the implant/fixation surface. The main difference is that with fixation antibiotic suppression whilst waiting for fracture union could be possible, or alternatively a change of type of fracture fixation e.g. to external fixation. With a joint prosthesis if a DAIR is unsuccessful then to avoid long term suppression, removal of the implant is required.
Infection within joints may be catastrophic event that often requires multiple or specialized care. Prosthetic joint infection (PJI) may soon becomes sure speciality in itself owing to high number of prosthesis done worldover. Sppecific classification and investigation are advocated but no universal guidelines yet. Fracture infections are now more salvageable by better reaming tecques, local antibiotic delivery or limb reconstruction system etc.
Based on your given affiliation I'm a bit surprised about the question to the audience here?? Is it only for collecting different expert opinions to write a review? If yes, than you are invited to contact me, because I'm working on this.
From my perspective and experience it makes no sense to give a short answer. Why? Simply, it is not possible. Basically, there are different aspects to be recognized, e.g. disease entity, different pathomechanisms, differently triggered immune systems of the hosts, different biomaterials with different biointerface characteristics etc. etc.
A fracture is neither biologically nor pathologically with an OA situation comparable, except in case of a fracture endoprosthesis. Sometimes you will face a bit of both in case of Vancouver fractures etc.
Another critical point is the different design of e.g. fixation plates and intramedullary nails without mobile parts versus an endoprosthesis with mobile parts etc.