The aim of our study was to identify independent factors that predicted a prolongation of the calculated perioperative antibiotic treatment leading in to therapy and to identify predictors for failure of the calculated antibiotic treatment and subsequent escalation to another preparation.

Among others we were able to demonstrate, that SSIs lead to escalation of antibiotic therapy regardless of bacterial culture positivity. Furthermore, mandibular bone reconstruction could be considered a predictor for antibiotic escalation and prolongation over 10 days postoperatively.

The aim of perioperative antibiotic prophylaxis in free flap reconstruction, is to prevent nosocomial infections like SSI and hospital acquired pneumonia which account for up to 20% of postoperative complications and determine microvascular success [4, 5, 11, 17, 23]. Systematic reviews considering perioperative prophylaxis in general head neck surgery suggest a single dose of intravenous antibiotics during surgery, followed by oral antibiotics is suitable [5, 24, 25]. However, these guidelines do not fully address the complexities of surgeries involving microvascular free flap reconstruction and few prospective studies have shown clear protocols for shortening the duration of antibiotic use in these complex surgeries. Consequently, clinicians tend to postoperatively extend antibiotic prophylaxis up to 10 days or longer to prevent nosocomial infections after free flap surgery indistinctly transitioning into postoperative antibiotic therapy [26].

It has been demonstrated that nosocomial infections, particularly surgical site infections can arise despite or due to inappropriate perioperative antibiotic use after head neck surgery [4]. Ultimately, it remains unclear why some patients experience poor outcomes despite all precautions, resulting e.g. in the failure of reconstruction. Synoptically this might be explained either with incorrect antibiotic selection or through the disruption of microbial niches, allowing virulent bacteria, which were previously suppressed by commensals, to postoperatively proliferate, cause SSI or remote site infections like HAP and leading to disintegration of microvascular reconstructions or deterioration of the general condition, with respiratory failure extending to death [12, 16, 17, 27,28,29,30].

With around 25% incidence of cultural positive SSI in our cohort we find the results to be coherent to the literature [4]. Our data indicates that in a total of 25% patient cases, the initially calculated perioperative antibiotic preparation had to be escalated. In 61% of the patients this escalation was necessary due to a clinically suspected SSI with consequent microbiological testing, with around 90% of these tests led to a culture positive SSI with a corresponding antibiogram. Overall, in 27 cases it was not possible to transform the clinical infection into a cultural antibiogram. Early escalation of antibiotics followed by a delayed swab test may result in a “false negative” microbiological culture. This might reduce the bacterial density to a level preventing cultural bacterial growth [31]. According to Rasnake et al., culture-negative SSIs are particularly challenging to treat since treatment options and antibiotic escalation rely on clinical experience rather than on resistance profiling [32]. Therefore, adherence to established infectious disease protocols, including preoperative bacterial screening, is recommended [33]. In reconstructively demanding salvage cases, the authors suggest to supplement preoperative diagnostics with molecular techniques like 16s RNA sequencing [31].

In our multivariate analysis, we found that both suspected surgical site infections, without microbiological confirmation and culture positive infections led to antibiotic escalation. This indicates that clinicians prioritize their clinical examination over microbiological tools, when deciding to escalate antibiotic medication. In our cohort, 43 patients underwent antibiotic escalation for other reasons than clinical surgical site infections. Of these cases, 46% involved resections located in the lower jaw, with 86% being tumor patients and radial forearm and fibula flaps being the most commonly utilized. Additionally, 6 patients from this group developed hospital-acquired pneumonia. This trend may be attributed to SSIs that were not microbiologically confirmed, complications with the anastomosis necessitating revision and subsequent prophylactic enhancement of antibiotic coverage or simply the surgeon’s preference. Hamilton et al. demonstrated that abnormal laboratory results are more likely to contribute to antibiotic escalation than negative cultures are to prompt de-escalation in a hospitalized setting [34]. These finding should encourage improvements in clinicians’ implementation and interpretation of microbiology results.

Statistically antibiotic treatment was mostly escalated to Piperacillin/Tazobactam (53.1%) and in some cases to Ciprofloxacin (18.9%). Literature indicates that SSIs in clean-contaminated surgery frequently involve polymicrobial bacterial spectra [4, 35]. This explains the calculated escalation to a Piperacillin (53%) formulation in our cohort, which typically provides broad-spectrum coverage, especially for Gram-negative bacteria [36, 37]. Ciprofloxacin is frequently used when Pseudomonas is detected, even though the relative prevalence in our study was rather low with 4% [38].

Under multivariate analysis, culture-positive SSIs significantly prolonged the total duration of antibiotic application, with an odds ratio of 3.5 (12.3 vs. 17.3 days) (Table 3). Numerous studies have demonstrated that the use of antibiotics, particularly broad-spectrum antibiotics, is associated with significant adverse effects. Gastrointestinal complications, including diarrhoea and C. difficile infections, as well as Candida overgrowth, particularly in patients with diabetes, are.

well-documented [39]. Additionally, evidence suggests an elevated risk of cardiac mortality in women who consume antibiotics for even less than 15 days during adulthood [40]. We know from literature reviews in other medical fields that it is not uncommon to see delayed de-escalation and hesitant transition to oral therapy resulting in antibiotic mis and overuse [41, 42]. Furthermore, studies have shown that prolonged antibiotic use offers no benefit to tumor patients at all. Instead, those needing adjuvant therapy after ablative tumor surgery benefit from the shortest possible perioperative antibiotic duration, as extended use can reduce response rates to adjuvant therapy and significantly shorten overall survival [13, 14]. The dogma of perioperative antibiotic prophylaxis “one-size-fits-all” should be tailored to the specifics of the surgical procedure, the patient and the unique aspects of the reconstruction, allowing for safe discontinuation of antibiotics after a short period.

Additionally in univariate analysis, haemoglobin levels and postoperative transfusion were found to correlate with antibiotic escalation (Table 1). However, this association was not confirmed in the multivariate model. It is recognized that low haemoglobin levels and subsequent transfusions are risk factors for nosocomial infections [43, 44]. The immune system may be directly compromised by a reduced oxygen-binding capacity otherwise low haemoglobin levels often co-occur with other comorbidities. Direct interactions between erythrocyte concentrates and the host immune system after transfusion is also a subject of ongoing discussion [45,46,47,48]. Nevertheless, it is evident that patient blood management, particularly in oncology patients, can enhance patient outcome, shorten overall hospital stay, and simultaneously reduce the risk of nosocomial infections [49].

Examining multivariate factors in our cohort that lead to an escalation or prolongation of calculated perioperative antibiotics, the most notable factor is microvascular bone reconstruction (Tables 2 and 3). Previous studies on this patient population have highlighted that managing patients with mandibulectomy and microvascular bone reconstructions poses significant challenges and is associated with increased all over complication rates [4, 50, 51]. This may be attributed to the prolonged duration of surgery, extensive incisions and the complex three-dimensional nature of the defect. Our data indicate that segmental mandibulectomy (p = 0.02) necessitates significantly longer durations of antibiotic therapy compared to partial mandibulectomy (Fig. 2). While the causality between SSIs and the use of small titanium plates remains unclear, a significantly higher complication rate is reported with the use of solitary reconstruction plates following segmental mandibulectomy suggesting that immediate microvascular reconstruction may yield superior outcomes [52]. Regarding preoperative radiotherapy, length of stay in the intensive care unit and nosocomial pneumonia, these factors are well-recognized risk factors for more complicated hospitalizations and impairment of microvascular reconstructions. Preoperative irradiation is a risk factor for worse outcomes following head and neck surgery, including flap loss and nosocomial infections [4]. In addition to impaired wound healing, there is a significantly increased likelihood of surgical site infections, likely attributable to alterations in the oral microbiome induced by radiation [8, 23]. For management targeted interventions, such as elective perioperative tracheotomy to shorten postoperative ventilation and preoperative dental rehabilitation with tooth extraction is controversy discussed [53,54,55].

In our retrospective analysis of antibiotic application time, the median duration was 10 days, with an average exceeding 12 days (Table 1). These findings require critical discussion, particularly regarding the transition from prophylaxis to therapy. The literature struggles to distinct clearly between prolonged perioperative antibiotic prophylaxis and the initiation of actual postoperative antibiotic treatment. As demonstrated by Mitchell et al. in a large cohort study of 427 patients undergoing microvascular head and neck reconstruction, the duration of postoperative antibiotic administration seems highly variable. In their study, only 23% of patients received antibiotics for 24 h or less, while the majority were treated for > 7days, with some exceeding 20 days [56]. The authors referred to this as prolonged perioperative antibiotic use and reported that patients receiving extended antibiotic prophylaxis had a significantly reduced risk of SSI [56]. However, the point at which prophylactic antibiotic use transitions to therapy remains poorly defined.

Our findings suggest that perioperative antibiotic prophylaxis, followed by postoperative therapy for over 10 days in at least 27% of our cohort, may be unnecessary in the absence of signs for clinical infection since there is no correlation avoiding SSI or HAP. Current systematic studies, such as those by Haidar et al., recommend that adequate perioperative antibiotic prophylaxis following microvascular head and neck reconstruction should not exceed 24 h [24]. However, there is evidence suggesting that patients with altered microbiomes—such as those who have undergone radiation therapy or mandibular reconstruction after osteonecrosis —might benefit from longer prophylaxis but remains vague due to the lack of prospective studies on this topic [4]. This may explain why clinicians in our study, as well as in other referenced studies, hesitated to discontinue perioperative antibiotic administration despite the absence of clinical evidence for surgical site infections.

This “dilemma” of extended postoperative antibiotic administration is likely attributable to the complex nature of the reconstructive cases within our patient population and missing antibiotic stewardship programmes in the past [57]. Studies on antibiotic therapy indicate a consensus towards minimizing the duration of antibiotic therapy, despite the considerable heterogeneity observed in some of the patient groups studied [58]. Prolonged antibiotic use is well-documented to promote the selection of resistant pathogens, inducing highly resistant mutans which can subsequently lead to surgical site infections and necessitate further escalation of antibiotic treatment [59,60,61]. In irradiated patients literature lacks systematic knowledge of patient’s oral microbiome, missing clear protocols for perioperative or neoadjuvant antibiotic therapy [62]. This underscores a significant gap, as it is well-established that the intraoral microbiome of patients with tumours or osteonecrosis, who constitute the core of the reconstructive patient population, can substantially differ from that of healthy individuals [7, 10, 63]. Such patients may harbour bacteria that, due to localized selection pressures, prolonged antibiotic use or radiotherapy do not conform to standard perioperative therapy protocols, leading to selective enrichment of these resistant strains after perioperative antibiotic prophylaxis. This necessitates further molecular biological studies to enable a systematic analysis of this reconstructive cohort and to derive evidence-based recommendations for pre-emptive antibiotic treatment.

However, the implementation of antibiotic stewardship guidelines in surgical disciplines remains yet limited [64]. To address this issue and reduce antibiotic duration, several key measures are necessary, particularly in complex reconstructive cases with a history of radiation or bone reconstruction. These include establishing local treatment protocols, monitoring antibiotic duration, optimizing dosage, and ensuring timely transitions to oral therapy. Additionally, collaboration with infectious disease specialists and clinical microbiologists could enable preoperative microbiome analysis to identify potential resistant pathogens, thereby allowing for more targeted antibiotic therapy adjustments[65, 66].

In summary, the authors conclude that the study’s results must have direct clinical implications. Firstly, a structured Antimicrobial Stewardship program needs to be established, which defines clear and reproducible protocols for perioperative antibiotic prophylaxis to prevent the increase in multidrug-resistant organisms and antibiotic-associated side effects. Patients with specific risk factors, such as preoperative radiation therapy or colonization with resistant organisms, should be identified preoperatively to determine those who may benefit from extended postoperative antibiotic coverage. Additionally, patients who are more likely to experience harm from prolonged therapy should be identified, as a short perioperative prophylaxis would suffice for them. A stricter Surgical Site Infection surveillance system will help identify microbial spectra, ensuring that even after multiple reconstructions, patients receive appropriately tailored perioperative antibiotic coverage.