Perioperative Infection in the Patient with Rheumatic Disease
The risk of infection accompanies the benefits of surgery. Immunomodulatory chronic illnesses may increase the risk of surgical infections. Surgical patients with rheumatologic illness need close preoperative assessment regarding their infection risks (fixed and modifiable), which vary on the basis of the proposed procedure, specific rheumatologic illness, and underlying comorbidities. Modification of the medication regimens in the preoperative period may decrease risk and enhance healing. Intraoperative antisepsis and antibiotic prophylaxis remain critical in this patient population. Postoperative fevers within 3 days of surgery are usually noninfectious but require vigilance and attention. The principles of surgical infection reduction are not different in the rheumatologic and general patient populations, but best practice depends on expertise in caring for patients with these illnesses.
KeywordsInfection Perioperative Rheumatology Rheumatoid arthritis Orthopedics Wound Healing Prophylaxis Antibiotic Rheumatic disease
Perioperative infection remains among the most common, and frequently preventable, adverse surgical outcomes. Surgical infection rates differ broadly by institution, case mix index, and procedure type. Very high surgical volumes (500,000 arthroplasties yearly in the United States) yield large numbers of infections, even if infection rates are kept low . Beyond the devastation that surgical infection can have on the health of an individual patient, they are costly to caregivers, surgeons and others on the therapeutic team, hospitals, payors, and the public’s confidence in modern medicine . Increasing governmental and regulatory oversight of surgical infections – for they are easily quantified quality indicators – has further increased incentives on healthcare organizations to minimize infection rates. It is unclear whether there can be a zero rate of surgical infections, but it is unlikely that national rates are at their theoretical minimum.
Preoperative Reduction of Infection Risk
The preoperative medical examination, typically used to assess pulmonary and cardiac issues related to surgery, can be a useful way to assess, and decrease, the risk of perioperative infections.
It is helpful to consider perioperative infectious risk factors in two categories, broadly defined: those risk factors that are modifiable and those that are fixed. Common modifiable risk factors include the presence of active infection (at the surgical site or remote from it), tobacco use, diabetes mellitus, and dermatitis. Fixed risks include previous surgery and infection history at the same site, the presence of preexisting foreign bodies at the operative site, and age. There is a gray zone between modifiable and fixed risks. For instance, although Staphylococcus aureus nasal and skin carriage confers increased operative risk of infection, it is not clear that perioperative decolonization treatment modifies this risk substantially such that universal screening for and/or treatment of the colonized state is needed [3, 4, 5, 6]. Rheumatologic disease in general has not been found to confer increased risk of S. aureus colonization in the very scant extant literature . Wegener's granulomatosis is one rheumatologic disease that has been found to be highly associated with intranasal carriage of S. aureus .
Rheumatic diseases, and the effects of their treatments, represent risks that are not always modifiable. These diseases have been found to raise the risk of surgical infections in limited studies , and the disease activity may be an important correlate of risk . The modifiability of this risk may be achieved only through perioperative changes to the immunomodulating regimen. For instance, among rheumatoid arthritis patients undergoing arthroplasty, perioperative cessation of low-dose methotrexate may paradoxically increase infection rate , but cessation of tumor necrosis factor alpha blockers appears to lower the infection rate [12, 13].
Patients with and without rheumatic illness are all faced with a balance between risks (only some of which are modifiable) and benefits as they face surgical procedures. Surgery is often categorized into elective and urgent cases, but not many are truly elective, when quality of life without the proposed surgery is brought into the equation. Deferring elective surgery can carry risk. Decisions to proceed or defer elective surgery quickly become individualized.
Dermatitis at the site of surgical incision may increase the risk of surgical infection; this relationship is best documented for atopic dermatitides [14, 15]. Control of dermatitis, at the incisional site and beyond, may decrease the risk of incisional and deep infection. Catheterization of the bladder, a common procedure in patients requiring anesthesia, is associated with transient bacteremia , although the clinical significance of this bacteremia is unclear. Whether or not asymptomatic bacteriuria should be treated preoperatively for any or all surgical candidates is not known, but we routinely treat preoperative patients with bacteriuria and pyuria who are likely to undergo Foley catheterization periprocedurally. In addition—particularly in elective surgery—maximizing reduction of risk of genitourinary infection via urologic procedures (prostatectomy, bladder resuspension, and the like) may substantially decrease the risk of infection, although risks and benefits need to be considered carefully; available evidence shows that urinary tract pathogens present at the time of surgery are uncommon causes of surgical site infections . In patients undergoing implantation of orthopedic or cardiac hardware, preoperative consideration of intraoral infection is important. The prevalence of untreated dental infection in certain surgical populations is high  and can be a risk factor for device-related infection by bacteremic seeding.
Although comparative data are lacking, there is nothing to suggest that patients with rheumatologic disease are immune from the increased surgical infection risks conferred by tobacco smoke [19, 20•] or the decreased risk when tobacco is stopped . Likewise, obesity increases surgical infection risk (see, e.g., [22, 23]). It is not clear whether subsequent weight loss—and by which intervention such weight loss occurs—decreases it.
Intraoperative Reduction in Infection Risk
Antimicrobial prophylaxis (AMP), when properly administered, is effective at decreasing surgical infections in multiple surgical subspecialties. AMP is indicated “for all operations or classes of operations in which its use has been shown to reduce SSI rates based on evidence from clinical trials or for those operations after which incisional or organ/space SSI would represent a catastrophe” . In general, these include neurosurgical, intrathoracic, intraabdominal, and major orthopedic surgeries such as arthroplasties and spine surgery. Although commonly provided, antibiotic prophylaxis is not clearly indicated for lower risk surgery . It is important to note that preincisional antibiotics are relatively contraindicated when the indication for surgery is to obtain specimens for microbiologic culture—for example, suspected infection.
Perioperative prophylactic antibiotics prolong the sterility of operative sites , especially when timed appropriately [27, 28]. Decisions regarding selection and dosing of AMP regimens are made on the basis of local microbial health-care ecology, patient factors (allergies, renal disease, weight), type of surgery, and the preoperative concern for the presence of infection. There are no data suggesting the generalized superiority of any AMP regimen for patients with specific rheumatologic diseases. First-generation cephlosporins (i.e., cefazolin) are preferred because of their longer half-life (as compared with antistaphylococcal penicillins), low cost, high efficacy, and safety. In general, first-generation cephalosporins are indicated for AMP unless there is (1) a high local incidence of beta-lactam-resistant Gram positive cocci (in which case vancomycin is added), (2) bowel surgery (in which case coverage is broadened), or (c) an allergy (in which case vancomycin or clindamycin is typically substituted). The incidence of allergy in the rheumatologic population is increased, so a careful antibiotic allergy history is important.
Dosing of AMP is dependent on patient factors (weight, creatinine clearance) and procedure duration. In longer procedures, additional intraoperative doses must be given when there is extensive blood loss and after two half-lives of the antibiotic (e.g., every 4 h for cefazolin and every 10 h for vancomycin in patients with normal creatinine clearance).
Preparation of skin just prior to surgery is an important aspect of infection prevention. Clippers used immediately prior to surgery, not razors or chemical depilatories, are best practice for removal of periincisional hair without local abrasion. There is controversy regarding skin antisepsis in the operating room. Care must be taken that allergies to potent skin antiseptics—typically, Betadine or chlorhexidine, in preparations that increasingly include alcohol—be taken into account prior to surgery and that betadine, in particular, be removed as much as possible postoperatively to prevent contact dermatitides.
Surgical scrub solutions now frequently come as mixtures of iodine-based compounds or chlorhexidine gluconate mixed with alcohol. These mixtures may have superior efficacy [29••]. However, the risk of operating room fire from electrocautery-induced ignition of pooled prep solution remains a concern .
Risks of Infection Remote from Surgical Site
Postoperative fever policy at the Hospital for Special Surgery
Management of Postoperative Fever in Total Joint Arthroplasty Patients
Linda A. Russell MD, Charles C. Cornell MD, Mary McDermott RN, and Barry Brause MD
(1) Evaluation of postoperative fever in total arthroplasty patients can be costly and is often fruitless.
(2) Ward and colleagues  demonstrated that fever (temperature >38.5 °C) is a common occurrence after hip and knee arthroplasty. They looked at 1,100 patients, at a single institution, undergoing hip (THR) or knee arthroplasty (TKR) over a 2-year period. The rate of positive tests was as follows: chest radiograph 2 %, blood culture 6 %, urine culture 22 %, and urinalysis 23.7o/o. Fever occurring after POD #3 (OR 23.3; p < .001), multiple days febrile (OR 8.6; p = .003) and maximum temp > 39 °C (25.4 % vs. 6.9 %) had a significantly higher rate of positive fever evaluations. The total direct cost of the fever evaluations was $73,878.
(3) An earlier study by Shaw and Chung  looked at 100 patients undergoing TKR and 100 patients undergoing THR. The postoperative fever curve for each patient was reviewed, the maximum temperature was reviewed, and the maximum temperature for previously defined intervals was recorded. The maximum daily temperature occurred on POD #1 and gradually leveled off by POD #5. Only one patient had a maximum temperature on POD #4 that was greater than POD #3. Patients undergoing revision procedures tended to have a more pronounced febrile response, but it was not statistically significant. There were no abnormal findings on the 17 chest radiographs ordered. The presence of a positive urine culture had no effect on the fever response, with most positive results being identified after the fever had returned toward normal. Authors suggested that early discharge is appropriate if the febrile response is decreasing appropriately.
Recommendations for Fever Evaluation in Postoperative Arthroplasty Patients
(1) When evaluating a postoperative patient with fever, an appropriate history and physical should be done.
(2) If there are no localizing signs or symptoms, the patient is nontoxic appearing, and the temp is ≤ 39 °C, no additional workup is indicated until POD #4.
(3) Hospital discharge is not contraindicated if the patient has ALL five of these criteria:
a. appears nontoxic and has no localizing signs or symptoms of a fever, AND
b. temperature elevation which is trending downward and is below 39 °C AND
c. no comorbidities that would make hospital discharge inappropriate AND
d. adequate access to medical and surgical surveillance and follow-up care to allow for a safe discharge, including ongoing evaluation of temperature evaluation
e. the medical attending has cleared the patient for discharge.
(4) If the patient has temperature >39°, fever greater than 3 days, or progressively elevated temperature, a directed fever evaluation is indicated.
The rheumatologic patient may at times be at lower risk of postoperative fever (e.g., on stress-dose perioperative steroids or other immunomodulators [33, 34]) versus a higher risk (nonspecifically in patients with systemic lupus). However, there is little to suggest that fever is any more of a reliable indicator of perioperative infection in these patients than in those without rheumatic disease. Postoperative fever in patients with systemic lupus erythematosus (SLE), rheumatoid arthritis (RA), and other rheumatologic illness is not obviously associated with greater predictive value for infection. For patients on prednisone or with altered corticosteroid physiology, postoperative fever can accompany adrenal insufficiency.
Postoperative fevers after the initial 72 h need to be evaluated on an individual basis with consideration of a gamut of possibilities, including the wound, intravenous access sites, catheters, lungs, genitourinary and aerodigestive tracts (including the biliary tree), and noninfectious issues such as drug fevers, chemical hepatitides, venous thromboses and emboli, and their underlying rheumatological disease.
Rheumatologic Medications in the Perioperative Period: Impact on Infection Risk
Perioperative oral steroid use decreases neutrophil diapedesis, impairs wound healing, is broadly suppressive to both the innate and adaptive immune response, and increases the rate of wound dehiscence and surgical site infection . Steroids injected into orthopedic tissues to manage inflammation and pain prior to surgery also may increase infectious risk, although this remains controversial [36, 37]. Minimizing exposure to the lowest possible dose to maintain hemodynamic stability is recommended.
Steroids, particularly when given in “stress doses” above the patient’s baseline to prevent acute adrenal insufficiency, frequently induce an acute neutrophilia and suppress the fever response. We are unaware of evidence that these acute effects interfere with postoperative fever workups and infection rates per se, but the combined effects of general immunosuppression and obfuscation of normal inflammatory signals make perioperative steroid use a particular challenge in the patient with a potential perioperative infection. Evidence that stress dose steroids are not routinely needed in surgery (see  as an example) will hopefully limit perioperative increases in steroid dosage in a way that may decrease rates of infection and allow for, when needed, meaningful postoperative fever evaluations.
There is incomplete understanding of the perioperative safety of many DMARDs, particularly leflunomide, sulfasalazine, azathioprine, and hydroxychloroquine; even less known is the role of these medications in surgical infection risk.
Methotrexate (at doses low enough to avoid neutropenia) does not increase surgical infection risk in orthopedic surgery and is typically continued throughout the postoperative period . We recommend maintaining usual doses of less than 20 mg weekly and decreasing doses, if possible, prior to surgery in other patients.
Leflunomide, a pyrimidine synthesis inhibitor with a half-life of more than 2 weeks, may have effects on wound healing [40, 41]. In patients with high presurgical risk, we hold it for 2 weeks preoperatively. Sulfasalazine, azathioprine, and hydroxychloroquine are typically continued until surgery and resumed when the patient is eating with normal renal function—typically, soon after surgery. Hydroxychloroquine and methotrexate are continued.
Each of the immunomodulatory biologic agents used in rheumatologic therapy is associated with particular nonsurgical infectious risks . For instance, the link between TNF-α inhibitors (particularly infliximab) and tuberculosis is well-documented , and the risk of reactivation of the hepatitis B virus in the setting of biologic and steroid therapy is documented in both the oncology and rheumatology literature [44, 45]. In addition, biologic immunomodulators as a class may raise the risk of perioperative infection. TNF-alpha inhibitors have been the best studies , although we are not convinced that the associations are independent of the severity of the underlying inflammatory condition. Not all studies have supported the association between receipt of drug and surgical infection rate , particularly in nonorthopedic surgery. However, because of the published and perceived risks, we advocate holding most TNF inhibitors for twice their half-life prior to surgery to offset any theoretical risk.
Little is known about the change in infection risk in surgical patients receiving rituximab, whose long half-life makes perioperative management impractical. Immunoglobulin levels can be monitored in higher-risk preoperative patients, although perioperative receipt of IVIG has not been rigorously tested as an infection prevention measure in immunoglobulin-deficient patients.
Abatacept, the T-cell activation inhibitor, and tocilizumab, the IL-6 receptor blocker, may have poorly defined effects on surgical wound healing; we typically hold them for a month prior to surgery. They may delay wound healing in RA patients undergoing orthopedic surgery, although the incidence of actual infection is not clearly increased .
Infections are part of orthopedic surgical and perioperative practice of the patient with rheumatologic disease. Prevention, prompt diagnosis, and appropriate therapy are key. Multiple interests—patients, physicians, institutions, and systems—are negatively affected by surgical infection and share a strong interest in minimizing the impact of these devastating complications.
Risk factors for infections can be identified, and often decreased, preoperatively. In addition to optimal management of a patient’s rheumatologic health and medications, attention to the details of their dental, skin, and genitourinary diseases, tobacco use, and underlying cardiovascular and endocrine conditions can have profound effects. S. aureus colonization, as discussed above, may be a modifiable risk factor. Nonmodifiable risk factors, such as age, and prior procedures and infections at the site of proposed surgery still are important factors in the complex preoperative risk–benefit equation.
Antimicrobial prophylaxis is of known benefit in general, cardiothoracic, neurosurgical, and major orthopedic procedures and is commonplace in others.
Fever during the initial 3 days following surgery is rarely from an infection, and even more rarely from a surgical site infection. Suppression of the postoperative fever in patients with rheumatologic disease is not well described. The typical patient does not need more than a close examination for fevers less than 39.0° in the 72 h after orthopedic surgery, and little suggests that rheumatology patients differ in this regard. Nevertheless, focal complaints and other clinical considerations must be taken into account. Clinical vigilance is always appropriate.
Perioperative patients with rheumatologic illness are a complex and heterogenous group, broadly at higher risk of infection. Close attention to detail and careful evaluation perioperatively are critical to the prevention, detection, and treatment of perioperative infection.
Compliance with Ethics Guidelines
Conflict of Interest
Andy O. Miller and Barry D. Brause declare that they have no conflict of interest.
Human and Animal Rights and Informed Consent
This article does not contain any studies with human or animal subjects performed by any of the authors.
Papers of particular interest, published recently, have been highlighted as: • Of importance •• Of major importance
- 2.Whitehouse JD, Friedman ND, Kirkland KB, et al. The impact of surgical-site infections following orthopedic surgery at a community hospital and a university hospital: adverse quality of life, excess length of stay, and extra cost. Infect Control Hosp Epidemiol. 2002;23(4):183–9.PubMedCrossRefGoogle Scholar
- 14.Lim CT, Tan KJ, Kagda F, et al. Implant infection caused by dermatitis: a report of two cases. J Orthop Surg (Hong Kong). 2007;15(3):365–7.Google Scholar
- 20.• Turan A, Mascha EJ, Roberman D, et al. Smoking and perioperative outcomes. Anesthesiology. 2011;114(4):837–46. Another good reason smoking is not recommended.Google Scholar
- 29.•• Darouiche RO, Wall Jr MJ, Itani KM, et al. Chlorhexidine-alcohol versus povidone-iodine for surgical-site antisepsis. N Engl J Med. 2010;362(1):18–26. Large randomized trial of alcohol-based versus conventional skin antisepsis.Google Scholar
- 32.• Athanassious C, Samad A, Avery A, et al. Evaluation of fever in the immediate postoperative period in patients who underwent total joint arthroplasty. J Arthroplasty. 2011;26(8):1404–8. Perioperative fever is poorly associated with subsequent infection.Google Scholar
- 34.Momohara S, Hashimoto J, Tsuboi H, et. al. Analysis of perioperative clinical features and complications after orthopaedic surgery in rheumatoid arthritis patients treated with tocilizumab in a real-world setting: results from the multicentre TOcilizumab in Perioperative Period (TOPP) study. Mod Rheumatol. 2013;23(3):440–9.Google Scholar
- 42.Singh JA, Wells GA, Christensen R, et. al. Adverse effects of biologics: a network meta-analysis and Cochrane overview. Cochrane Database Syst Rev. 2011;2Google Scholar
- 45.Oshima Y, Tsukamoto H, Tojo A. Association of hepatitis B with antirheumatic drugs: a case-control study. Mod Rheumatol. 2013;23(4):694–704.Google Scholar