Introduction

Despite their low incidence, surgical side infection (SSI) remains one of the most feared hospital-acquired infections in neurosurgery due to their potentially serious consequences. Accordingly, SSI may complicate the patient’s post-operative course by increasing morbidity and mortality, prolongation of hospitalization, which results in considerable more costs, a nowadays critical factor for all health care systems [1, 2]. In neuro-oncology patients, SSI may additionally delay adjuvant therapies, which leads to an increased risk of adverse medical and psychosocial outcome [3].

SSI rates range on second to fourth position of hospital-acquired infections with a considerably large range of reported incidences between institutions and countries [4, 5]. In a multicenter analysis by the ACS-NSQIP (American College of Surgeons—National Surgical Quality Improvement Program), the overall neurosurgical post-operative infection rate was 5.3% from 2006 to 2014 in 132,063 neurosurgery adult (i.e., > 18 year of age) cases [6]. In respect to pediatric neurosurgery, there are no comparable large studies available on SSI rates to date.

To prevent SSI, surgical antimicrobial prophylaxis (SAP) was introduced early as a short-term application of an antibiotic just before a surgical procedure to avoid pathogen multiplication in the operative field [7]. Since the availability of prospectively randomized studies on prophylactic antibiotic application, demonstrating the overall potential of SAP to effectively reduce SSI, peri-operative antibiosis was widely acknowledged [8,9,10,11,12,13,14,15]. In regard to “clean” neurosurgical procedures, the prospective surveillance study by Korinek et al. demonstrated that SAP effectively decreased infection rates in low-risk patients (i.e., clean craniotomy cases) [16]. It is of note that no effectiveness in SAP could be shown for high-risk patients (those undergoing emergency, clean contaminated, and dirty procedures; or reoperation or with operative times exceeding 4 h) [16, 17].

Even though SAP is not a substitute for evidence-based hygiene measures, aseptic handling and tissue-conserving surgical techniques (https://www.nice.org.uk/guidance/CG74/chapter/1-Guidance, as of March 30, 2018), it has the potential to essentially reduce complication rates and costs. Thus, SAP is recommended currently for adult and pediatric neurosurgery [18]. However, evidence-based recommendations, specifically for pediatric neurosurgery, remain problematic. Placebo-controlled and randomized studies on SAP in pediatric neurosurgery are missing to date and current guidelines on “clean” neurosurgery for this specific patient group were largely derived from adult (neuro)-surgical databases not taking into account the specific needs of preterm, newborns, and children [18]. In regard to the pediatric population, SAP-specific age-related studies are mainly limited to shunt-related procedures.

This review will summarize current data situation and SAP recommendations in clean surgery and shunt-surgery—as far as available—with special focus on pediatric neurosurgery.

General considerations

Surgical site infections (SSI) are defined as infections of the superficial or deep tissue or organs at the surgical site or related to the site of the surgical procedure. Regardless of age, a national nosocomial infection surveillance (NNIS) risk index was defined for assessing patients at risk for SSI [19]. It ranges from 0 to 3, based on the American Anesthesiologists score, surgical wound classification (Altemeier score), and duration of the surgical procedure [20]. Narotam et al. provided a classification system for neurosurgery in which procedures are divided into five categories: clean, clean with foreign body, clean contaminated, contaminated, and dirty [21]. Risk factors for post-operative infections after neurosurgical procedures are considered an ASA classification of ≥ 2, invasive post-operative intracranial pressure monitoring, ventricular drains for five or more days, cerebrospinal fluid (CSF) leak, procedure duration of more than 2 to 4 h, repeat or additional neurosurgical procedures, concurrent or previous shunt infection, emergency procedures, and metabolic comorbidity (like diabetes). In regard to pediatric neurosurgery, age might also be a relevant factor and comorbidities, which are related to developmental disorders like spina bifida-associated problems [22, 23].

Current recommendations on SAP refer to Bratzler’s et al. (2013) comprehensive analyses of the literature and recommendations. This consortium of different organizations provided analyses of the current literature on meta-analyses and studies, which investigated antimicrobial prophylaxis versus no prophylaxis among others in clean neurosurgical procedures [18]. Assuming less infection-related complications like meningitis, scalp infections, bone flap osteitis, and abscess or empyema with antimicrobial prophylaxis compared to surgery without prophylaxis, they provided recommendations on different aspects of SAP [18]. However, pediatric particularities could not be addressed for neurosurgery (instead of for example cardiac surgeries) due to the current data situation.

Surgical antibiotic prophylaxis (SAP) in clean neurosurgical procedures

Crucial aspects of SAP include the time-point of application, appropriate antibiotic selection, duration, and SAP re-dosing during longer procedures. Establishing antibiotic stewardship programs were shown to increase the compliance to these points, as well as correctly performed team-time outs before starting the surgical procedure [24].

Pre-operative dosing time

Applying the antibiotic shortly before skin incision is considered effective SSI prevention as the applied antibiotic lowers bacterial inoculation before incision and prevents proliferation during surgery [25, 26]. In the current recommendations, the optimal time-point for SAP administration is given when applied within 60 min before surgical incision [18, 27]. This point seems to be often ignored during daily clinical practice and SAP application “at induction of anesthesia” is not an adequate approach [28]. Here, a correctly performed team-time out may be a helpful tool for a constructive interdisciplinary interaction and adherence to standards and recommendations [29, 30].

Choice of antibiotic for SAP

SSI in neurosurgical procedures is mainly due to gram-positive organisms, Staphylococcus aureus, and coagulase-negative staphylococci, and other skin organisms (for review, see [18]. One major problem is the current high rate of infection with MRSA and coagulase-negative staphylococci, with a reported rate of 75–80% isolated patients (among others [16, 17, 31,32,33,34]).

In regard to the antibiotic choice, different antimicrobial regimens were reported to be effective in SSI prevention including clindamycin, vancomycin, cefotiam, piperacillin, cloxacillin, oxacillin, cefuroxime, cefotaxime, sulfamethoxazole–trimethoprim, cefazolin, penicillin G, and amoxicillin–clavulanate [9, 16, 17, 33,34,35,36,37,38].

Currently, a single dose of cefazolin (not followed by further antimicrobiotic administration) is recommended by the AAHP/IDSA/SIS/SHEA (American Society of Health-System Pharmacists, the Infection Disease Society if America, Surgical Infection Society, the Society for Healthcare Epidemiology of America) for patients undergoing clean neurosurgical procedures, i.e., “elective craniotomy and cerebrospinal fluid-shunting procedures” (strength of evidence A) and “implantation of intrathecal pumps” (strength of evidence C) [18], as this first-generation cephalosporin is active against a variety of staphylococci and streptococci [39]. These recommendations were followed in studies evaluating SAP for different neurosurgical procedures—in regard to pediatric neurosurgery—including vagus nerve stimulator implantation and external ventricular drains, and found to be effective in reducing or preventing SSI [40,41,42]. Thereby, clindamycin or vancomycin should be reserved as an alternative agent for patients with a documented β-lactam allergy or for MRSA-colonized patients (vancomycin) [18]. In a further meta-analysis, third-generation cephalosporins did not show superiority over first-generation cephalosporins in regard to incisional- and organ-related SSI [43].

To ensure sufficient serum and tissue antibiotic concentration level, SAPs have to be given in a weight-based manner. Further, there is the need of re-dosing during longer procedures if the duration exceeds two half-lives of the drug or there is excessive blood loss during the procedure. For cefazolin, current recommendations are 30 mg/kg children with a re-dosing interval of 4 h (after initiation of the pre-operative dose) [18].

Duration of SAP

Further care has to be taken regarding duration and choice as arbitrary usage can lead to resistant microorganisms (European Center for Disease Prevention and Control; https://ecdc.europa.eu/sites/portal/files/media/en/publications/Publications/Perioperative%20antibiotic%20prophylaxis%20-%20June%202013.pdf; accessed May 16, 2018). There is a great variety in clinical practice regarding this point. In the recent literature, SAP is recommended for a shortened post-operative course, i.e., a single dose or continuation for less than 24 h. There is no evidence that prolonged AB application bears any benefit for the patient (National Institute of Health and Clinical Excellence. Surgical site infection is www.nice.org.uk/CG74, [17].

Antibiotic prophylaxis for shunt-related procedures

Deleterious sequelae resulting from shunt infections are well known and feared complications. The role of SAP in such shunt-related procedures has been studied in various early trials (among others [44, 45]). Despite ongoing efforts to prevent these infections, there is still no consensus met about application and duration of SAP in pediatric shunt surgeries. However, SAP application in shunt surgery for children is recommended and integrated as one crucial step in current shunt infection prevention protocols (among others [46,47,48,49]. Thereby, reported shunt infection rates themselves vary widely between different institutions and countries and depend on individual medical preconditions like prematurity, age, previous shunt infection and others. The major source of shunt infections are bacteria of the patient’s own skin. Hence, infection often occurs during the first post-operative month and is most often due to infections with Staphylococcus species (90%) [50].

Due to small patient numbers of single institutional studies, the SAP problematic for shunt surgeries has been approached by meta-analyses of multiple individual studies. By doing so, Langeley et al. [45], found that SAP significantly (up to 50%) reduced shunt related infection rates similar to Haines and Walter 1994 who demonstrated relative effectiveness of SAP depending on the initial infection rate [44]. Ratilal et al. performed meta-analyses of 17 trials with 2134 participants to evaluate the use of SAP in shunt procedures. They came to the conclusion that systemic SAP within the first 24 h post-operatively was effective to prevent shunt infection regardless of patients’ characteristics (age, type of shunt) [51]. Xu et al. recently performed a meta-analysis for effectiveness of SAP in pediatric shunt surgery; they concluded that shunt prophylaxis is effective, again pointing out that more evidence from advanced multi-center studies are necessary [52].

Like in other clean procedures benefit of SAP beyond 24 h after surgery in shunt-related procedures is nowadays discussed controversial and there are no general recommendations available on SAP duration after surgery. Some institutions follow Kestle et al.’s protocol and apply a further dose of cafazolin 6 h after surgery (first author’s experience at two different hospitals), in others antibiotics are applied during 3 following days after surgery (authors’ institution) or are limited to the single dose before surgery. In their meta-analyzes of trials with 2134 patients, Ratilal et al. demonstrated a beneficial effect of SAP for the first 24 h post-operatively to prevent shunt infections; they stated that the benefit of its use after this period remains uncertain [51].

Due to the widespread inconsistency regarding SAP choice, SAP application time and duration and recommending specific antibiotics remains problematic [53]. Antibiotics used for SAP in shunt surgery included first-generation cephalosporins, second-generation cephalosporin, and vancomycin (in different forms from i.v. or local powder administration). Thereby, none of these antibiotics showed any better effect compared to the other. Another aspect not cleared to date is if the antimicrobiotic’s ability to penetrate the blood-brain-barrier has an impact on the prevention of CNS infections after shunting procedures [18]. In their latest shunt infection prevention protocol, Kestle et al. recommended cefazolin (30 mg/kg) application before surgery (“before incision”) and one further application of this dose after surgery [48].

However, large cohort studies on SAP in pediatric shunt surgeries are not available and evidence-based protocols for SAP choice and application times are lacking. Accordingly, we are not past the stage of often arbitrary application of prophylactic antibiotics for shunt-related procedures in children. Same applies to the use of antimicrobial-impregnated shunts. There are hints from published case series and few randomized controlled studies that the use of these shunts along with SAP decreases the rate of shunt infections [50, 54,55,56]. However, data from further well-designed studies are needed to define the role of antimicrobial-impregnated shunts in SAP in pediatric neurosurgery.

In regard to SAP for ventricular drains or baclofen pumps or other devices, the factual situation is even unclear. There are minimal published trial data regarding appropriate prophylaxis for these procedures. Thus, there are no recommendations available for children, which can be deduced from the literature right now [18]. However, considering publications of the current literature, SAP is mostly performed like in clean or shunt-procedures (i.e., application of an antibiotic before incision).

Perspective

In addition to the lack of evidence, authors increasingly raised concerns about clinicians’ compliance in adhering to current available SAP recommendations. In Ciofi degli Atti et al.’s prospective study on adherence to SAP-indication, SAP-choice, SAP-timing, and SAP-duration in pediatric surgical procedures, there was an underuse of SAP when it was indicated and an overuse was it was not indicated [28]. Also in case of application, there was a quite arbitrary manner in the choice of antibiotics, timing, and duration.

To address this point, the European Center for Disease Prevention and Control proposed five key SAP modalities that were shown to improve compliance of healthcare professionals in regard to appropriate administration, timing, dosage, and duration of SAP. These modalities included (1) forming a multidisciplinary team to develop, implement and update a SAP protocol, conduct an audit of compliance, and provide feedback; (2) ensuring administration of SAP within 60 min prior to incision; (3) assigning responsibility for timely administration of SAP to the anesthesiologist; (4) administering only a single dose of SAP; and (5) to discontinue SAP at the end of surgery (https://ecdc.europa.eu/sites/portal/files/media/en/publications/Publications/Perioperative%20antibiotic%20prophylaxis%20-%20June%202013.pdf, accessed May 16, 2018; summarized in Table 1).

Table 1 Suggested standard operation procedures (SOP) for peri-operative antibiotic prophylaxis by the European Center for Disease Prevention and Control (details are provided in the text)
Full size table

However, despite these efforts, Weiss et al. demonstrated a lacking compliance and fostered individualism in regard to adherence to internal institutional standard operation procedures (SOPs). Despite a formulated SOP which was developed by a multidisciplinary team (neurosurgeons, neurooncologists, microbiologists) based on a medline search, different antibiotic agents were found to be used other than recommended or different time-periods and different doses [57].

These observations again reflected the dilemma of missing evidence-based and thus accepted best practice workflows. Thereby, critical points that challenged randomized controlled studies on antimicrobial prophylaxis in children is the low SSI rates for neurosurgical procedures resulting in low numbers with an increase in the likelihood of type II errors and difficulty in establishing controls (including placebo, no treatment, or other antimicrobial agents) due to study design.

Conclusion

Controlled randomized multicenter studies are urgently needed for implementing SAP guidelines in pediatric neurosurgical procedures. In regard to shunt-related procedures, which have been addressed more frequently compared to other neurosurgical interventions, adherence to existing protocols is desirable. This emphasizes the need for a robust surveillance programs to provide valid data for SAP. Other procedures like clean craniotomies have to be addressed at all for the pediatric population in order to meet the needs of this age group and prevent inappropriate use of SAP, which promotes selection of multi-resistant microorganisms. Due to statistical needs, multi-center surveillance studies are needed for implementing SAP evidence-based recommendations in pediatric neurosurgery, which is the prerequisite to overcome traditions and arbitrary individual decisions that may endanger the patient’s safety.