記住我
ORIGINAL ARTICLE
Year : 2022 | Volume
: 18
| Issue : 9 | Page : 170-176
Perioperative antimicrobial prophylaxis in clean-contaminated head and neck squamous cell cancer surgeries: Is less better?
Pavneet Kohli1, Prasanth Penumadu1, Naveeth Shukkur1, M Sivasanker2, Arumugam Balasubramanian1, Sachit Ganapathy3
1 Department of Surgical Oncology, JIPMER, Puducherry, India
2 Department of Surgery, Royal Liverpool University Hospitals NHS Trust, Merseyside, England
3 Department of Biostatistics, JIPMER, Puducherry, India
Correspondence Address:
Prasanth Penumadu
Associate Professor, Department of Surgical Oncology, JIPMER, Puducherry
India
Source of Support: None, Conflict of Interest: None
CheckDOI: 10.4103/jcrt.JCRT_1654_20
Introduction: Judicious use of antibiotics and stringent adherence to practice guidelines is the need of the hour as antibiotic resistance is a rampant problem. Despite several reports in the literature describing the optimal duration of antibiotics, there is no consensus. A “one for all” protocol may be impractical and hence the guidelines need to be tweaked to take into consideration local factors. We designed a protocol for prophylactic antibiotics in clean-contaminated head and neck cancer squamous cell carcinoma (HNSCC) surgeries to prevent unchecked abuse and evaluated its feasibility.
Materials and Methods: Two hundred consecutive patients who underwent a clean-contaminated surgery for HNSCC between January 2017 and December 2019 were included. Single-dose intravenous amoxicillin-clavulanate at induction followed by three doses of amoxicillin-clavulanate, metronidazole, and amikacin in the postoperative period was used. Adherence to the antibiotic protocol was assessed from a prospectively maintained database.
Results: The mean age was 55.99 ± 11.71 years. The protocol was effective in 70% of the patients with an acceptable surgical site infection (SSI) rate of 12%. Flap-related complications (9.5%) and oro-cutaneous fistula (5%) were common causes of prolonged antibiotics. On univariate analysis, blood transfusion (P = .014), clinical stage at presentation (P = .028), patients undergoing reconstruction (P = .001), longer operative time (P = .009), and pathological T stage (P = 0.03) were at higher chance of deviating from the protocol. On multivariate analysis, age more than 50 years (OR: 2.14, 95% CI: (1.01, 4.52); P value = 0.047) and reconstruction (OR: 3.36, 95% CI: (1.21, 9.32); P value = 0.020) were found to be significant.
Conclusions: A three-dose perioperative antibiotic prophylaxis in clean-contaminated HNSCC surgeries is feasible. Similar protocols should be developed and validated at other major centers to limit the unnecessary use of antibiotics and prevent the emergence of antibiotic resistance.
Keywords: Antibiotic misuse, antibiotic resistance, clean-contaminated surgery, head and neck surgery, oral cancer, SSI
Prophylactic antibiotic usage was introduced in surgical centers across the world to reduce the incidence of surgical site infection (SSI) and the morbidity associated with it. Despite advances in sepsis preventive protocols, technology, and surgical protocols, SSIs continue to be the most common postoperative complication. They remain one of the most regular nosocomial infections, head and neck surgery being no exclusion.[1],[2],[3] However, liberal use of antibiotics for surgical prophylaxis is a double edge sword and its abuse has increased antimicrobial resistance, a worrisome problem in India.[4],[5],[6] Rational usage of prophylactic antibiotics is pivotal in reducing the probability of developing resistant strains.
The Centers for Disease Control and Prevention (CDC) classifies surgery into clean, clean-contaminated, contaminated, and dirty subtypes.[7] In head and neck cancer, surgeries that do not come in contact with the aerodigestive tract (UADT) are considered clean (e.g. thyroidectomy). The incidence of SSI in these clean surgeries is less than 2% and postoperative antibiotic prophylaxis is not necessary.[8],[9] Surgeries involving the UADT are labeled as clean-contaminated and the incidence of SSI can be as high as 50% that may warrant prophylactic antibiotic use.[10],[11] However, irrational use could lead to undesirable adverse effects like acute kidney injury, pseudomembranous colitis, and the more sinister problem of antibiotic resistance.[4],[12]
Perioperative antibiotic prophylaxis in clean-contaminated surgeries is well accepted; however, extended use (up to 5 days) remains controversial with varying practices across different centers.[13] The CDC reports that up to 50% of administered antibiotics are unnecessary, inappropriate, and hence avoidable.[14] Adopting a uniform antibiotic policy is further compounded by multiple variables like malnutrition, diabetes, tobacco usage, preoperative chemotherapy or/and radiotherapy, tumor stage, and type of reconstruction. Hence, there exists a certain dubiety on the optimal antibiotic protocol and its duration in clean-contaminated head and neck cancer surgery.
This study aimed to audit the antibiotic policy of a single center, which specializes in oncological surgery and adopts the middle path of three doses of prophylactic perioperative antibiotics. It also aimed to find out the various risk factors that may require prolonged antibiotic use.
> MethodsThis was a retrospective study of consecutive patients who underwent clean-contaminated surgery for a head and neck squamous cell carcinoma (HNSCC) in a tertiary cancer center between January 2017 and December 2019. The study was approved by the institute ethics committee.
Patients who underwent clean-contaminated head and neck oncological surgeries were included in the study. Patients who were treatment naïve and underwent primary surgery for cancer, as well as patients who underwent salvage surgery for recurrent tumors or second primary cancers, were included. Exclusion criteria included the patients with amoxicillin-clavulanate allergy, any infections in the preceding 2 weeks, and a total leukocyte count of less than 3000 or more than 12,000 cells/mm3. Similarly, patients who underwent emergency or palliative surgeries were excluded. Patients with reduced renal function (creatinine clearance <50 ml/min) were also excluded. Patients whose surgical wounds were considered clean, contaminated, or dirty as per the CDC classification were also excluded as they have a different rate of SSIs.
Patients were advised for a full-body bath with soap and 10% betadine scrub solution a night before surgery. The skin was prepared in the operation theater using an alcohol-based chlorhexidine solution. An antibiotic protocol was formulated after a review of previous audits of antibiotic usage in the department and consultation with the microbiologists keeping in mind the hospital antibiogram, institute infection prophylaxis policies, and local guidelines. This protocol involved a first dose of intravenous co-amoxiclav (amoxicillin + clavulanic acid) 30 mg/kg (max 1.2 g) in the operating room within half-an-hour of induction of anesthesia and a second dose if the surgery was prolonged beyond 4 h or suffered blood loss of more than 500 ml. This was followed by three doses of co-amoxiclav (30 mg/kg (max 1.2 g) every 8 hourly), metronidazole (7.5 mg/kg (max 500 mg) every 8 hourly), and amikacin (7.5 mg/kg every 12 hourly) in the postoperative period. The patients who developed an infection in the postoperative period were treated as per the standard hospital protocol. Euglycemia and normothermia were ensured before surgery.
Basic demographic data, laboratory studies, and treatment procedures including gender, age, body mass index (BMI), tumor location, tumor stage (American Joint Committee on Cancer, AJCC 8th edition), preoperative radiotherapy or chemotherapy, preoperative serum albumin (<3.5 g/dl, ≥3.5 g/dl) and hemoglobin level (<10 or >10 g %), reconstruction type (local, regional, or free flaps), operative time (<4 or ≥4 h), and perioperative blood loss and blood transfusion were obtained and recorded from a prospectively maintained database as well as from the electronic medical records. Details of surgery, intraoperative and postoperative complications, length of hospital stay, and the complications were documented. Complications were graded as per the Clavien–Dindo classification.[15]
Statistics
Statistical analysis was performed using SPSS version 19 (IBM Corp, Armonk, New York). The distributions of categorical variables such as gender, co-morbidity, clinical profile, and pathological profile were expressed as frequency and percentages. The continuous variables such as age, albumin level, and hemoglobin levels were expressed as mean with standard deviation. Continuous variables like the length of hospital stay (LOHS), blood loss, and operative time were depicted as median and interquartile range (IQR) as did not follow the normal distribution assessed by the Kolmogorov–Smirnov test. The association of deviation from antibiotic protocol (PD) with categorical variables was carried out by using the Chi-square test/Fisher's exact test. The association of continuous variables with PD was carried out using the independent student t-test/Mann–Whitney U test depending on the distribution of the variable. Univariate logistic regression was performed to assess the potential factors predisposing the patients to deviate from antibiotic protocol. The variables, which were statistically significant with a relaxed P value <0.1, were considered for multiple logistic regression. Adjusted odds ratio along with 95% confidence was reported. The statistical analysis was carried out at a 5% level of significance and P < 0.05 was considered as significant.
> ResultsA total of 200 patients who met the inclusion criteria during the study period were included in the study. There were 127 (63.5%) males and 73 (36.5%) females. The mean age of the study cohort was 55.99 ± 11.71 (range 34–73 years) years. One hundred and forty-six patients (73%) were treatment naïve. Twenty-six (13%) patients received preoperative chemotherapy and 28 (14%) patients received preoperative radiotherapy. One hundred and seventy-seven (88.5%) patients had the primary tumor located in one of the subsites of the oral cavity as defined by the AJCC 8th edition. Twenty-three (11.5%) patients had tumors located in the larynx or the pharynx. The demographic details are mentioned in [Table 1].
Thirty-seven (18.5%) patients were diabetic and 42 (21%) patients had albumin lower than 3.5 g/dl. Twenty-eight (14%) patients had a hemoglobin level of <10 g%. The median albumin of the cohort was 4 g/dl and the median hemoglobin was 12.5 g%.
Twenty-three (11.5%) patients were diagnosed as Stage I after clinical examination and imaging. Twenty-nine (14.5%) patients were diagnosed as Stage II, 43 (21.5%) patients Stage III, and 104 (52%) patients Stage IV (99 Stage IVA and 5 Stage IVB). One patient underwent a salvage laryngectomy post concurrent chemo-radiotherapy (CCRT) for a dysfunctional larynx. One hundred and thirty-three (66.5%) patients underwent some form of reconstruction. Eight (4%) patients underwent a local flap, 92 (46%) patients a regional flap, and 33 (16.5%) patients underwent a distant “free” flap reconstruction.
The median duration of hospital stay was 13 days (IQR 8-22) and the median operative time was 300 min. The median intraoperative blood loss in the study cohort was 350 ml. Thirty-three (16.5%) patients received a perioperative blood transfusion. There was no mortality in the study group. Eighty-one (40.5%) patients had no morbidity post-surgery, while 73 (36.5%) patients had Clavien–Dindo Grade III/IV morbidity.
On histopathological analysis, 63 (31.5%) patients were node-positive while 137 (68.5%) were negative. Twelve patients (6%) were T1, 47 (23.5%) T2, 70 (35%) T3, and 24 (12%) patients were T4 on histopathological examination.
Deviation from antibiotic policy occurred in 60 (30%) patients in our study. In 19 (9.5%) patients, flap-related complications accounted for deviation. SSI accounted for deviation in 24 patients (12%). Antibiotics were continued or restarted due to oro-cutaneous fistula in 10 (5%) patients. Two patients underwent a re-exploration due to neck hematoma and were continued on antibiotics. Five patients deviated due to other reasons (lower respiratory tract infection (LRTI), gastroenteritis, etc.).
In assessing the test of association between the factors with prolonged antibiotic use, blood transfusion (P = .014), clinical stage at presentation (P = .028), undergoing reconstruction (P = .001), type of reconstruction (P = 0.01), operative time more than 4 h (P = .009), and pathological T stage (P = 0.03) were found to be statistically significant as shown in [Table 2]. The PD was significantly more in patients undergoing free flap reconstruction as compared to another form of reconstruction (P = .001). Age, preoperative hemoglobin or albumin, diabetes mellitus, tumor subsites, treatment naïve or post-chemo/radiotherapy, recurrent tumors, and node positivity on neck dissection were not statistically significant [Table 2].
Table 2: Univariate analysis to identify the factors affecting deviation from predefined perioperative prophylactic antibiotic policyUnivariate logistic regression suggested that age, clinical stage, reconstruction, and operative duration were found to be potentially significant factors. These factors were considered for multiple logistic regression. It was found that age of more than 50 years (OR: 2.14, 95% CI: (1.01, 4.52); P value = 0.047) and reconstruction (OR: 3.36, 95% CI: (1.21, 9.32); P value = 0.020) were significant factors in deviation in antibiotic intake after adjusting for the effect of other covariates in the model [Table 3].
Table 3: Effect estimates of potential risk factors associated with antibiotic deviation from univariate and multiple logistic regressionThe receiver operator curve (ROC) along with the area under curve (AUC) was used to assess and compare the association of blood loss with deviation from antibiotic policy. 375 ml was used as the cut-off with 66.7% sensitivity and 60.7% specificity (AUC value 0.682) [Figure 1].
Figure 1: ROC for assessing the predictive accuracy of the amount of blood loss leading to deviation from antibiotic policy > Discussion In this retrospective study, we audited a shorter duration of the perioperative prophylactic antibiotic protocol in clean-contaminated head and neck cancer surgery and assessed its efficacy and impact on SSI. In the study cohort, 70% of the patients did not require antibiotics beyond the prophylactic schedule. Factors like blood transfusion, clinical stage at presentation, longer operating time, and higher T stage were significant in univariate analysis but failed to show any significant association in multivariate analysis.
The optimal duration of antibiotic prophylaxis in clean-contaminated head and neck surgery has been a bone of contention and despite several randomized controlled trials (RCTs) and prospective studies, there is no consensus.[10],[16],[17],[18],[19] There is a wide difference between the protocols followed in a clinical trial setting and the real-world experience. There is also a paucity of prevalent antibiotic practices in an Indian setting, where generally antibiotic use and its prolonged use are discretionary, rampant, and often unwarranted. This study highlights the experiences of a tertiary health center and its adherence to a predefined antibiotic protocol.
The SSI rate in this study cohort was 12%. There is a wide variation in the incidence of SSI in clean-contaminated head and neck surgery with rates ranging from 16% to 48%.[11],[20],[21],[22] Our cohort was well within these limits with a shorter duration perioperative antibiotic course which highlights the efficacy of this “middle path.” Several patient-related factors like impaired sugar control,[23] increased BMI,[20] malnutrition,[24] anemia,[23] and higher American Society of Anesthesiologists (ASA) score[25] have been associated with increasing the risk of SSI. Similarly, tumor-related factors like advanced disease and positive lymph nodes can increase SSIs.[10],[19] Use of preoperative radiotherapy/chemotherapy or free flap reconstruction has also been documented in some studies to increase SSI.[19],[26] We used a universal antibiotic protocol for our patients and assessed the feasibility of it.
Several trials have compared the efficacy of first/second generation cephalosporins with amoxicillin-clavulanate and have not reported significant differences in the incidence of SSIs.[27],[28],[29] Murphy et al.[30] evaluated the rate of SSIs in patients undergoing osteomyocutaneous free flap reconstruction and concluded that the incidence of SSI in patients receiving cefazolin (OR = 1.2,95% CI 0.3–4.8) was a little higher than the ampicillin–sulbactam (OR = 1) group though the difference was not statistically significant (P = 0.73). Similarly, Robbins et al.[31] compared cefazolin alone to cefazolin–metronidazole to study the need for anaerobic bacteria coverage and found a significant SSI risk reduction with the addition of metronidazole (P < 0.05).
The efficacy of short- versus long-term prophylactic antibiotics has been studied comprehensively. A systematic review found no significant difference between 1 day or 5 days of prophylaxis using various types of antibiotics including penicillins, cephalosporins, and clindamycin.[32] Sawyer et al.[33] concluded that 2 days of administration of cefazolin–metronidazole had less incidence of wound infection as compared to patients who received it for more than 7 days (P = 0.04). Busch et al.[10] compared the rate of infection between short- and long-term prophylaxis in patients with a tracheotomy in patients with or an increased BMI and found no difference in the incidence of SSI or fistula formation. Another recent meta-analysis concluded that clean-contaminated head and neck surgery, cefazolin, amoxicillin-clavulanate, and ampicillin-sulbactam for 24–48 after surgery were associated with the lowest rate of SSI.[26] Hence, we developed a protocol of a single dose of preoperative amoxicillin-clavulanate and three doses of postoperative amoxicillin-clavulanate, amikacin, and metronidazole injection under the guidance of the Institute of Antibiotic Stewardship committee. An important factor that led us to formulate this three-drug protocol was the lower socioeconomic status and advanced nature of malignancies, which comprised the majority of our patients.
On further analyzing the protocol failure cases, we found a higher failure rate in patients who received an intraoperative blood transfusion, with longer operative time, with a more advanced clinical stage at presentation, and those undergoing reconstruction. There was however no difference in patients who had undergone non-surgical treatment before definite surgery. Similarly, the rate of protocol failure was similar in patients undergoing laryngectomy or oral cavity surgeries.
The benefits of antibiotic stewardship are far-reaching. In addition to simplifying postoperative nursing care, reducing antibiotic doses presents a potential for cost savings to the health care system. A longer duration of antibiotic prophylaxis has also been associated with emerging microbial antibiotic resistance. Decreasing the duration of prophylaxis has been shown to reduce methicillin resistant Staphylococcus aureus (MRSA) isolation levels. Fonseca et al.[34] stated that single-dose antibiotic prophylaxis did not lead to an increase in the rates of SSI and brought a monthly savings of $1980.
To the best of our knowledge, this is the first Indian study, which performs an audit of a predefined antibiotic protocol in clean-contaminated head and neck surgery. Though our protocol failure rate is slightly high at 30%, this can be attributed to the advanced nature of malignancy at presentation, poor hygiene, and low socioeconomic factors. This study is limited by its retrospective and single-center nature. However, it provides an engaging argument for decreasing antibiotic usage, increasing awareness, and providing hospital administrative bodies with an educational intervention to reduce extra costs. This is extremely important in developing countries like ours.
> ConclusionWith multidrug resistance being an emerging problem, protocol-based antibiotic prophylaxis is the need of the hour. Our study shows the feasibility of developing and successfully implementing such protocols. These protocols are likely to decrease antibiotic resistance and the costs of treatment. A multicentric population is required to validate the protocol-based prophylaxis in tertiary care setups in the developing world.
Source of the study
Jawaharlal Institute of Postgraduate Medical Education and Research, Dhanvantari Nagar, Puducherry – 605006.
Ethical statement
The authors confirm that all accepted principles of ethical and professional conduct have been met. The manuscript has been read and approved by all the authors and each author believes that the manuscript represents honest work.
Author contribution
Prasanth Penumadu (PP) was responsible for study concepts. Study was designed by Pavneet Kohli (PSK) and PP. Data Acquisition was done by Naveeth Shukkur (NS). Quality control of data was done by A Balasubramanian (AB). Sachit Ganapathy (SG) and PSK performed data and statistical analysis and its interpretation. PSK, PP and Sivasanker M (SM) prepared the manuscript and edited it. PP, PSK and AB reviewed the manuscript. All authors approved the final manuscript for submission.
Financial support and sponsorship
Nil.
Conflicts of interest
There are no conflicts of interest.
> References
留言 (0)