Definition for persistent wound drainage (PWD) is still controversial: time (48 h to 1 week), site of drainage (wound or site of removed suction drains), type (clear or not, blood or other fluids), amount of secretion, and eventual microbial content of secretion are variously defined in literature (Butt et al. 2011; Dennis 1997; Hansen et al. 2013; Jaberi et al. 2008; Lonner and Lotke 1999; Saleh et al. 2002; Vince et al. 2007). A recent international expert consensus has defined PWD as “continued drainage from the operative incision site for greater than 72 h” with a strong consensus among participants (80 % agree) (Ghanem et al. 2014). According to the authors, an area greater than 2 × 2 cm of drainage covering the medication is also required to fulfill the definition of PWD. However, it is well accepted that PWD after total joint arthroplasty (TJA) is a predictor of a subsequent onset of a TJA infection (Saleh et al. 2002). Patel et al. (2007) estimated that each day of PWD increases the risk of infection by 29 % after total knee arthroplasty and by 42 % after total hip arthroplasty.

Surgical site infections represent a relevant percentage of hospital-acquired infections (HAIs); according to ECDC 2013 point prevalence survey, of a total of 15,000 reported HAIs, the most frequently reported HAI types were respiratory tract infections (23.5 %), followed by surgical site infections (19.6 %). One third of HAIs present at admission were surgical site infections (ECDC 2013, Fig. 26.3). Wound complications after TJA can lead to deep infection with increase of length of hospitalization, need for additional surgery, and adjunctive costs. Any effort to prevent wound complications must be performed in the preoperative period as some risk factors for the development of wound infections or delayed healing (malnutrition, hypokalemia, other infections, decompensated diabetes mellitus, obesity, smoking, renal failure, hypothyroidism, and alcohol abuse) can be corrected. Identification of nasal carriers of methicillin-resistant Staphylococcus aureus and decontamination with a short course of mupirocin ointment to the nares may be effective in reducing surgical site infections rates (Rao et al. 2008). Other not preventable factors associated with increased wound complications are prior surgical procedures, immunosuppressive therapy, diverticulosis, rheumatic diseases, and multiple previous incisions. The main topics in the postoperative period are thromboprophylaxis and the use of surgical drains, and both are controversial. Anticoagulation has been associated with increased risk of ecchymosis, hemarthrosis, drainage, and need for transfusion as well as significant bleeding in the gastrointestinal tract and central nervous system (Vince et al. 2007). Parvizi et al. (2007) found that a mean international normalized ratio greater than 1.5 was more prevalent in patients who developed postoperative wound complications and subsequent periprosthetic infection. Finally, patients requiring warfarin seem to have a greater incidence of deep infection, superficial infection, and other wound associated complications (McDougall et al. 2013).

The use of surgical drains for elective TJA is a common practice despite limited evidence to support routine use in noninfected cases; if used, the time of permanence of drains is not well established. Strahovnik et al. (2010) found that the duration of wound drainage did not have any influence on the incidence and duration of prolonged serous drainage; absence of drainage was not associated with PWD, but increased swelling and pain of the thigh were present in these patients. In a perspective randomized trial, Dora et al. (2007) found that to omit closed suction drainage in hip arthroplasty allowed a simpler and faster wound management without any disadvantage. Finally, a recent meta-analysis of randomized controlled trials in hip arthroplasty concludes that there is insufficient evidence to support the routine use of closed suction drainage (Chen et al. 2013). The authors found no statistical difference in wound hematomas, occurrence of postoperative deep vein thrombosis, and postoperative wound infection between groups with and without drainage.

An important issue in PWD is to understand the origin of fluid and if it represents a sign of established deep infection or of a wound closure defect; in case of early postoperative drainage with no evidence of pain, erythema, purulence, or other clinical signs of deep infection, PWD should be managed by wound care for the first 5 days.

Superficial wound cultures are not recommended in this setting: specimens obtained from a wound or a sinus tract often generate polymicrobial results and may misguide diagnosis and treatment. In literature concordance rates between sinus tracts and bone cultures range from 38 to 88 % (Cook and Farrar 1978; Esposito and Gleckman 1977; Mackowiak et al. 1978; Mousa 1997; Patzakis et al. 1994; Perry et al. 1991; Ulug et al. 2009). In a recent study, Tetreault et al. (2013) found 47.3 % of agreement between superficial and deep cultures with bacterial growth in superficial cultures in 80 % of cases without deep infection.

Deep sepsis can be ruled out with a joint aspiration and cell count, differential, and negative culture and sensitivity. There isn’t a well-delineated cutoff for the levels of synovial cell count and PMN% in the infected hip arthroplasty (Parvizi et al. 2011). Schinsky et al. (2008), in a chronic setting, propose a threshold of 3,000 cells/μL for leukocytes and 80 % for PMN% for the infected hip arthroplasty. Blood cultures should be obtained if fever is present. Serum tests (erythrocyte sedimentation rate and C-reactive protein) are not useful in this setting as these markers can be elevated up to 60 days in the postoperative period, and they are affected by co-medications, comorbidities, age, and sex (Bilgen et al. 2001; Larsson et al. 1992). C-reactive protein seems to be a promising marker of prosthetic joint infection when measured in synovial fluid rather than in serum: in a prospective study by Parvizi et al. (2012), a threshold of 0.06 mg/L for individual ELISA and of 3.7 mg/L for multiplex ELISA provided a sensitivity of 70 and 84 % and a specificity of 100.0 and 97.1 %, respectively. Finally promising methods for diagnosing infection are polymerase chain reaction techniques for detection of microorganisms and analysis of circulating cytokines levels (Bergin et al. 2010; Kobayashi et al. 2008).

Various interventions have been recommended in order to reduce the amount of wound drainage. Negative-pressure wound therapy (NPWT) after total hip arthroplasty was evaluated in a randomized prospective study and compared with standard dressing, and the authors demonstrated a decreased development of postoperative seromas with NPWT (Pachowsky et al. 2012). Hansen et al. (2013) reported cessation of wound drainage after total hip arthroplasty with NPWT in 76 % of cases, while the remaining 24 % required subsequent surgery. However, a previous Cochrane meta-analysis found no evidence for the effectiveness of NPWT in PWD in trauma surgery, arthroplasty, and skin graft patients (Webster et al. 2012a, b).

Antibiotic therapy is not recommended in patients with early PWD for several reasons. First of all, in the absence of signs of deep infection, an antibiotic therapy is unnecessary, and prevention of ingress of germs through draining wound should be warranted by a correct wound care. The emergence of antibiotic resistance actually represents a dramatic challenge in HAIs and any effort should be made in order to prevent this event; to avoid unnecessary and/or inadequate antibiotic prescription is mandatory. In addition antibiotic therapy can cause adverse effects. Finally administration of antibiotics could mask an underlying infection and reduce the sensitivity of microbiological cultures.

Surgical revision is recommended after 5 days of PWD to reduce the likelihood of a subsequent prosthetic joint infection. During the procedure (exploration, adequate irrigation and debridement, exchange of modular components, and meticulous fascia and layer wound reclose), deep cultures are strongly recommended to drive antibiotic therapy (Ghanem et al. 2014).The number of samples is also important; the isolation of a microorganism from three or more independent samples is highly predictive of infection with a specificity of 99.6 % (Atkins et al. 1998). In case of negative culture (due to low bacterial burden or previous antibiotic therapy) with the presence of clinical signs of infection, an empiric antibiotic therapy covering oxacillin-resistant Staphylococcus aureus should be administered until further workup rules out deep infection. In this setting knowledge of local epidemiology and evaluation of patient risk factors for multidrug-resistant microorganism is fundamental to tailor antimicrobial therapy (Table 26.1). In case of positive culture, most common microorganisms and their treatment are summarized in Table 26.2. Staphylococci are the most common agents causing prosthetic joint infection (Kuiper et al. 2013; Saleh et al. 2002). For staphylococcal infections, a pathogen-specific intravenous or highly bioavailable oral antimicrobial therapy (Table 26.2) in combination with rifampin 600 mg orally daily should be administered for 2–6 weeks, followed by rifampin plus a companion oral drug for a total of 3 months for a total hip arthroplasty infection (Osmon et al. 2013). Rifampin may decrease serum linezolid concentrations in combination therapy (Gebhart et al. 2007; Hoyo et al. 2012).