Surgical site infection in pediatric spinal fusion surgery revisited: outcome and risk factors after preventive bundle implementation

Abstract

Background:

Surgical site infections (SSI) contribute to significant morbidity, mortality, length of stay, and financial burden. We sought to evaluate the incidence and risk factors of surgical site infection following pediatric spinal fusion surgery in patients for whom standard perioperative antibiotic prophylaxis and preventive strategies have been implemented.

Methods:

We conducted a retrospective study of children aged <18 years who underwent spinal fusion surgery from January 2017 to November 2021 at a quaternary academic pediatric medical center. Univariable analysis was used to evaluate associations between potential risk factors and SSI.

Results:

Of 1111 patients, 752 (67.6%) were female; median age was 14.2 years. SSI occurred in 14 patients (1.3%). Infections were superficial incisional (n=2; 14.3%), deep incisional (n=9; 64.3%), and organ/space (n=3; 21.4%). Median time to SSI was 14 days (range, 8 to 45 days). Staphylococcus aureus and Escherichia coli were the most frequently-isolated bacteria. Potential risk factors for SSIs included low body weight (Odds ratio (OR) 0.96, 95% confidence interval (CI) 0.93–0.99, p=0.026), ASA classification of ≥3 (OR 24.53, 95%CI 3.20–188.22, p=0.002), neuromuscular scoliosis (OR 3.83, 95%CI 3.82–78.32, p<0.001), prolonged operative time (OR 1.56, 95%CI 1.28–1.92, p<0.001), prolonged anesthetic time (OR 1.65, 95%CI 1.35–2.00, p<0.001), administration of prophylactic antibiotic ≥60 minutes before skin incision (OR 11.52, 95%CI 2.34–56.60, p=0.003), and use of povidone-iodine alone for skin preparation (OR 5.97, 95%CI 1.27–28.06, p=0.024).

Conclusion:

In the context of a robust bundle for SSI prevention; low body weight, ASA classification of ≥3, neuromuscular scoliosis, prolonged operative and anesthetic times, administration of prophylactic antibiotic ≥60 minutes before skin incision, and use of povidone-iodine alone for skin preparation increased the risk of SSI. Administration of prophylactic antibiotic within 60 minutes of skin incision, strict adherence to high-risk preventive protocol, and use of CHG-alcohol could potentially reduce the rate of SSI.

1. Introduction

Surgical site infection (SSI) is one of the major healthcare-associated infections that contributes to significant morbidity, mortality, length of stay, and financial burden for hospitalized patients in the United States.^{1, 2} Overall SSI incidence in pediatric orthopedic operations is 2.5–2.8%,^{3, 4} and the number is remarkably higher in spinal fusion surgery at 3.6–10.3%.^5–10 Over the past decade, several studies have reported risk factors for SSI in pediatric spinal fusion surgery, including underlying etiology of scoliosis,^11–13 patient comorbidities^{7, 14} such as obesity, suboptimal antibiotic prophylaxis,^{7, 13, 15} and surgical complexity.¹¹

Evidence-based practice guidelines have been developed to address modifiable risk factors to minimize the risk of SSI or the severity of the infection.^16–20 Adherence to a protocol using several strategies to prevent infection resulted in a lower SSI rate.²¹ This study aims to evaluate the incidence and risk factors of SSIs following pediatric spinal fusion surgery in the context of a standardized perioperative antibiotic prophylaxis and implementation of routine preventive strategies into clinical practice.

2. Materials and methods

2.1. Patient cohort and data collection

This retrospective cohort study was approved by the Institutional Review Board at Boston Children’s Hospital, and informed consent was exempted. We included patients younger than 18 years of age at the time of surgery who underwent spinal fusion surgery from January 2017 to November 2021 at our institution.

Patient demographic data were collected including gender, age, race, body weight, height, American Society of Anesthesiologists (ASA) physical status classification, and primary etiology of scoliosis presenting for surgery. Intraoperative data included surgical and anesthesia attendings, operative time (duration from skin incision to skin closure), anesthetic time (duration from the start to the end of anesthesia), type and timing of intravenous antibiotic, skin preparation agent, medications, temperature, and blood transfusion. All data were retrieved from electronic medical records and Anesthesia Information Management System^™ (AIMS; Cerner, MO, USA).

2.2. SSI definition and prevention bundles

SSI was diagnosed based on the National Healthcare Safety Network (NHSN) definitions from the US Centers for Disease Control and Prevention and occurred within 90 days after the index procedure.²² Patients who developed SSI after spinal fusion surgery were identified through continuous prospective surveillance by the Infection Prevention and Control department.

At our institution, the general SSI prevention bundle includes 1) preoperative bath with soap or an antiseptic agent; 2) skin preparation with chlorhexidine gluconate (CHG)-alcohol (ChloraPrep^™), povidone-iodine (Betadine^®) alone, povidone-iodine plus CHG-alcohol, povidone-iodine plus alcohol, or all three antiseptics combined; 3) using clippers for hair removal; and 4) perioperative antibiotic prophylaxis. Cefazolin was the first-line prophylactic antibiotic for standard-risk spinal fusion (i.e., idiopathic scoliosis in otherwise healthy patients) and administered within 60 minutes prior to surgical incision and re-dosed every 4 hours intraoperatively. An alternative antibiotic (clindamycin or vancomycin) would be administered if a patient had a history of penicillin or cephalosporin allergy. The choice of antibiotic in those cases was made based on the institutional guideline.

Patients with high-risk spinal fusion (i.e., complex surgery, associated significant comorbidities such as cerebral palsy, neuromuscular scoliosis, etc.) also received cefazolin as a first-line agent with additional gentamicin. Vancomycin would be administered if the patients were labeled with penicillin or cephalosporin allergy or colonized with Methicillin-resistant Staphylococcus aureus (MRSA). Additional measures were exercised including 1) chlorhexidine bath the night before surgery, 2) preoperative patient education sheet, 3) intraoperative wound irrigation, 4) maintaining normothermia, 5) vancomycin powder added to bone graft, and 6) use of impervious dressings.

2.3. Statistical analysis

We presented number and percentage for categorical variables, mean and standard deviation (SD) for continuous variables with normal distribution, or median and interquartile range (IQR) for variables with skewed distribution. Shapiro-Wilk test was used to assess normality. Potential risk factors for SSI were assessed with univariable logistic regression analysis. The results were reported as odds ratio (OR) and 95% confidence interval (CI). P-value of <0.05 was considered statistically significant. Pearson pairwise correlation coefficient (r) between variables was examined to address the potential collinearity of two variables. The statistical analysis was performed using PASW Statistics for Windows, Version 18.0 (SPSS Inc., Chicago, IL, USA).

3. Results

A total of 1,544 patients underwent spine surgery between January 2017 and November 2021 at our institution. Among these, we enrolled 1,111 patients younger than 18 years of age who underwent spinal fusion surgeries regardless of surgical approaches.

In our cohort, the majority of patients were female (n=752, 67.6%) with a median age of 14.2 years (IQR 12.5–15.8), the youngest patient was 8 months old. Demographic data and patient characteristics were presented in Table 1. Adolescent idiopathic scoliosis was the most common indication for surgery (n=430, 38.7%), followed by neuromuscular scoliosis (n=187, 16.8%). Approximately one-third of patients had ASA classification of at least 3 (n=369, 35%). The most common primary surgical antibiotic prophylaxis was cefazolin (n=969, 87.2%), gentamicin was administered in 192 (17.3%) patients. Most of the patients (n=1068, 96.4%) received prophylactic antibiotic within 60 minutes before skin incision, whereas, 20 (1.8%) and 19 (1.7%) patients received antibiotic longer than 60 minutes before skin incision and after skin incision, respectively. Four patients had already received preoperative antibiotic, we did not include these patients in the analysis related to antibiotic timing.

Table 1.

Patient characteristics

	Total (n=1111)
Gender; female	752 (67.7)
Age (yr)	14.2 (12.5,15.8)
Race
White	690 (62.1)
Black/African American	58 (5.2)
Asian	34 (3.1)
American Indian/Alaska native	4 (0.4)
Other	124 (11.2)
Declined/unable to answer	201 (18.0)
Body weight (kg)	52.8 (42.8,64.1)
Height (cm)	160.0 (149.8,167.5)
ASA classification
1	157 (14.2)
2	560 (50.4)
3	360 (32.4)
4	33 (3.0)
Diagnosis
Idiopathic scoliosis	596 (53.6)
Neuromuscular scoliosis	187 (16.8)
Congenital scoliosis	97 (8.7)
Unspecified or unknown	231 (20.8)
Type of prophylactic antibiotic
Cefazolin	969 (87.2)
Clindamycin	128 (11.5)
Vancomycin	12 (1.1)
Ceftriaxone	1 (0.1)
Linezolid	1 (0.1)
Adjunct Gentamicin	192 (17.3)
Adjunct Gentamicin with
Cefazolin	167 (87.0)
Clindamycin	16 (8.3)
Vancomycin	7 (3.6)
Cefazolin and Vancomycin	2 (1.0)
Antibiotic-to-incision time (min)	15.0 (11.0,21.0)
Antibiotic timing
Before skin incision
0–30 minutes	989 (89.3)
31–60 minutes	79 (7.1)
>60 minutes	19 (1.7)
After skin incision	20 (1.8)

Data are presented as number (%), median (IQR)

ASA, American Society of Anesthesiologists

The overall incidence of SSIs following spinal fusion surgery was 1.3% (14 out of 1,111 patients) with an annual infection varying from 0.0 to 2.6%. Table 2 summarized the detailed characteristics of all patients who developed postoperative SSIs. All of the patients underwent posterior spinal fusion surgery with the age ranging from 10.6 to 17.5 years old. Most of the infected patients had neuromuscular scoliosis (n=12, 85.7%). The median onset of SSIs was 14 days after the surgery (range of 8 to 45 days). Deep incisional SSIs were the most frequent which accounted for 64.3% (n=9) of the infections; followed by organ/space and superficial incisional infections at 21.4% (n=3) and 14.3% (n=2), respectively. Polymicrobial infection was common, with the majority of organisms isolated being Gram-negative bacteria (18 out of 26). Staphylococcus aureus (n=4, 15.4%) and Escherichia coli (n=4, 15.4%) were the most common organisms isolated from cultured specimens. Over 80% of patients who developed SSI received cefazolin as the primary prophylactic antibiotic (n=12, 85.6%), the others received clindamycin (n=1, 7.2%) and vancomycin (n=1, 7.2%). Gentamicin was administered to 9 infected patients. Regarding the antibiotic timing, most of the patients received prophylactic antibiotic within 60 minutes prior to skin incision (n=11, 78.6%) except 2 patients and 1 patient whose antibiotics were given longer than 60 minutes before skin incision and after skin incision, respectively. The distribution of the timing of antibiotic administration relative to surgical incision between patients with and without SSI is presented in Figure 1. All patients received intraoperative re-dosing of antibiotic per institutional protocol. There was a total of 8 surgical attendings and 63 anesthesia attendings involved in our cohort. We did not include anesthesia attendings in the statistical analysis.

Table 2.

Detailed characteristics of patients who developed SSIs following spinal fusion surgery.

No	Age (yr)	Gender	Type of scoliosis	Levels of spinal fusion	SSI onset (POD)	SSI type	Operative time (min)	Organism(s)	Prophylactic antibiotic	Gentamycin	Antibiotic-to-incision time (min)	Skin prep solution
1	14.9	F	Neuromuscular	T7 to pelvis	14	Deep incisional	301	Klebsiella oxytoca Raoultella ornitholytica	Cefazolin	Y	13	Chloraprep™
2	17.0	M	Neuromuscular	T2 to S1	12	Deep incisional	664	Escherichia sp. Proteus	Cefazolin	N	18	Chloraprep™
3	13.4	M	Neuromuscular	T3 to pelvis	40	Organ/Space	532	Staphylococcus aureus	Cefazolin	N	28	Betadine®
4	16.5	F	Neuromuscular	thoracic to lumbar (unspecified)	45	Deep incisional	666	Escherichia coli Klebsiella pneumoniae Bacteroides fragilis	Cefazolin	Y	118	Betadine®
5	15.8	M	Neuromuscular	T2 to L2	8	Deep incisional	428	Cutibacterium granulosum Cutibacterium acnes Enterobacter cloacae complex sp. SCP12–74	Clindamycin	N	14	Chloraprep™
6	14.6	F	Neuromuscular	T2 to pelvis	14	Superficial incisional	523	Proteus mirabilis	Cefazolin	Y	20	Chloraprep™
7	16.8	F	Idiopathic	T3 to iliac	22	Deep incisional	566	Staphylococcus aureus	Cefazolin	Y	17	Chloraprep™
8	11.3	F	Neuromuscular	T2 to S1	12	Organ/Space	592	Staphylococcus aureus Klebsiella pneumoniae	Vancomycin	Y	73	Chloraprep™
9	12.0	F	Neuromuscular	T4 to pelvis	9	Superficial incisional	649	Bacteroides caccae	Cefazolin	N	25	Chloraprep™
10	17.5	F	Idiopathic	T11 to L4	23	Deep incisional	222	Streptococcus pyogenes Staphylococcus aureus Serratia marcescens	Cefazolin	N	0	Chloraprep™
11	10.6	M	Neuromuscular	T3 to pelvis	9	Deep incisional	575	Bacteroides caccae Veillonella parvula Escherichia coli	Cefazolin	Y	43	Chloraprep™
12	12.3	M	Neuromuscular	T3 to pelvis	12	Deep incisional	402	Escherichia coli	Cefazolin	Y	2	Chloraprep™
13	16.4	M	Neuromuscular	T2 to S5	15	Organ/Space	721	Pseudomonas aeruginosa Propionibacterium sp.	Cefazolin	Y	−55	Chloraprep™
14	11.8	F	Neuromuscular	T1 to pelvis	21	Deep incisional	322	Escherichia coli	Cefazolin	Y	30	Chloraprep™

POD, postoperative day number; SSI, surgical site incision

Figure 1.

The distribution of timing interval of antibiotic administration relative to skin incision between the SSI and non-SSI groups. The horizontal lines indicate skin incision time (0-minute) and 60-minute before skin incision, the recommended interval for prophylactic antibiotic administration and surgical incision.

In the univariable analysis (Table 3), potential risk factors for SSIs included low body weight (OR 0.96, 95%CI 0.93–0.99, p=0.026), ASA classification of ≥3 (OR 24.53, 95%CI 3.20–188.22, p=0.002), neuromuscular scoliosis (OR 3.83, 95%CI 3.82–78.32, p<0.001), prolonged operative time (OR 1.56, 95%CI 1.28–1.92, p<0.001), prolonged anesthetic time (OR 1.65, 95%CI 1.35–2.00, p<0.001), administration of prophylactic antibiotic ≥60 minutes before skin incision (OR 11.52, 95%CI 2.34–56.60, p=0.003), and use of povidone-iodine alone for skin preparation (OR 5.97, 95%CI 1.27–28.06, p=0.024). There was a very high correlation between operative time and anesthetic time (r=0.96, p<0.001).

Table 3.

Univariable analysis of risk factors for SSIs following spine fusion surgery.

	Total	SSI		Unadjusted OR	p-value
	(n=1111)	Yes (n=14)	No (n=1097)	(95%CI)
Gender; female	752 (67.7)	7 (50.0)	745 (67.9)	0.47 (0.16,1.36)	0.16
Age (yr)	14.2 (12.5,15.8)	12.3 (11.5,16.7)	14.1 (12.5,15.6)	1.09 (0.88,1.35)	0.45
Body weight (kg)	52.8 (42.8,64.1)	38.1 (34.9,41.3)	53.1 (44.5,64.3)	0.96 (0.93,0.99)	0.026*
ASA ≥3	393 (35.4)	13 (92.9)	380 (34.6)	24.53 (3.20,188.22)	0.002*
Diagnosis
Idiopathic scoliosis	628 (56.5)	2 (14.3)	626 (57.1)	1.00 (reference)
Neuromuscular scoliosis	228 (20.5)	12 (85.7)	216 (19.7)	17.39 (3.86,78.32)	<0.001*
Congenital scoliosis	105 (9.5)	0 (0.0)	105 (9.6)	n/a	n/a
Revision surgery	30 (2.7)	0 (0.0)	30 (2.7)	n/a	n/a
Surgeon				n/a	0.72
1	126 (11.4)	1 (7.1)	125 (11.4)
2	189 (17.0)	2 (14.3)	187 (17.1)
3	54 (4.9)	5 (35.7)	49 (4.5)
4	92 (8.3)	1 (7.1)	91 (8.3)
5	347 (31.3)	1 (7.1)	346 (34.6)
6	150 (13.5)	1 (7.1)	149 (13.6)
7	65 (5.9)	1 (7.1)	64 (5.8)
8	86 (7.8)	2 (14.3)	84 (7.7)
Operative time (hr)	4.9 (3.6,6.7)	8.89 (5.9,10.7)	4.5 (3.4,6.3)	1.56 (1.28,1.92)	<0.001*
Anesthetic time (hr)	6.1 (4.7,8.2)	12.0 (8.3,12.9)	6.1 (4.7,8.2)	1.65 (1.35,2.00)	<0.001*
Type of antibiotic
Cefazolin	969 (87.2)	12 (85.8)	957 (87.2)	1.00 (reference)
Clindamycin	128 (11.5)	1 (7.1)	127 (11.6)	0.63 (0.08,4.89)	0.66
Vancomycin	12 (1.1)	1 (7.1)	11 (1.0)	7.25 (0.87,60.7)	0.07
Ceftriaxone	1 (0.1)	0 (0.0)	1 (0.1)	n/a	n/a
Linezolid	1 (0.1)	‘ (0.0)	1 (0.1)	n/a	n/a
Non-cefazolin antibiotic	142 (12.8)	2 (14.3)	140 (12.8)	1.14 (0.25,5.14)	0.87
Antibiotic timing
After skin incision	20 (1.8)	1 (7.1)	19 (1.7)	5.15 (0.63,42.30)	0.13
0–30 min	98£ (89.3)	10 (71.4)	979 (89.6)	1.00 (reference)
31–60 min	79 (7.1)	1 (7.1)	78 (7.1)	1.26 (0.16,9.93)	0.83
>60 min	19 (1.7)	2 (14.3)	17 (1.6)	11.52 (2.34,56.60)	0.003*
Skin preparation agent
ChloraPrep™	944 (95.9)	12 (85.7)	932 (96.1)	1.00 (reference)
Betadine	28 (2.8)	2 (14.3)	26 (2.7)	5.97 (1.27,28.06)	0.024*
Others	12 (1.2)	0 (0.0)	12 (1.2)	n/a	n/a
Average temperature (°C)	36.2 (35.8,36.6)	36.4 (35.7,36.6)	36.2 (35.8,36.6)	1.53 (0.62,3.80)	0.36
Dexamethasone use	725 (65.3)	10 (71.4)	715 (65.2)	1.34 (0.42,4.29)	0.63
Inotrope/vasopressor infusion	577 (81.4)	8 (80.6)	569 (81.4)	0.91 (0.19,4.35)	0.91
PRBC transfusion	144 (13.0)	2 (14.3)	142 (12.9)	1.12 (0.25,5.06)	0.88
Cell saver transfusion	659 (59.3)	8 (57.1)	651 (59.3)	0.91 (0.32,2.65)	0.87

Data are presented as number (%), median (IQR)

ASA, American Society of Anesthesiologists; PRBC, packed red blood cell

4. Discussion

In the context of an implementation of SSI prevention bundles and standardized perioperative antibiotic prophylaxis, the incidence of SSIs following pediatric spinal fusion surgery in our study was 1.3%. We identified lower body weight, ASA classification of ≥3, neuromuscular scoliosis, prolonged operative time, administration of prophylactic antibiotic ≥60 minutes before skin incision, and use of povidone-iodine alone for skin preparation as potential risk factors for postoperative SSIs. Although there was no statistical significance, one surgeon, with the least number of cases performed, demonstrated a distinctly high proportion of SSI incidence suggesting surgeon experience might affect the rate of SSI. To test variability, we need to have a much higher case number in a future study.

Although we did not compare the incidence of SSI before and after the implementation of SSI prevention bundles at our institution, the SSI rate for pediatric spine fusion is remarkably lower than previously reported by others.^5–10 Nevertheless, considering the potential risk factors we have identified, there are still rooms for improvement to maintain as low SSI rate as possible.

Most current guidelines recommend the administration of prophylactic antibiotic within 60 minutes prior to the surgical incision for most antibiotic classes, and within 120 minutes for vancomycin and fluoroquinolones.^17–20 Our institutional guideline mandates the administration of prophylactic antibiotics (other than vancomycin and fluoroquinolones) within 60 minutes of the skin incision. However, the protocol of 60-minute interval was violated in some patients (n=39, 3.5%), and three of them belonged to the SSI group. Among the patients with SSI, three patients received prophylactic antibiotics inappropriately, either greater than 60 minutes (two patients) or after skin incision (one patient). Additionally, the antibiotic was presumably administered at the time of skin incision (time zero) in one SSI patient. This finding strongly emphasized the importance of appropriate antibiotic timing to prevent SSI.

Positioning for spine surgery is one of the factors that likely hinders the proper antibiotic timing. We generally administer prophylactic antibiotic after general anesthetic induction is completed. When antibiotic was given ≥ 60 min before skin incision, the patient was still in a supine position. As it usually takes some time to position the patient to prone, it is reasonable to give the antibiotic after the patient is in a prone position to minimize the antibiotic-to-skin incision window.

Cefazolin has a short half-life and exhibits time-dependent killing, in which efficacy is optimized when drug concentration remains above the minimum inhibitory concentration (MIC) of a bacterium for at least 40–50% of the dosing interval.²³ Previous studies demonstrated approximately 30% reduction of plasma cefazolin concentration by 60 minutes after the administration.^{24, 25} These pharmacokinetics studies are consistent with a clinical study of antimicrobial prophylaxis and SSI risk; the authors reported that SSI risk was elevated as the interval between antibiotic infusion and skin incision increased.²⁶ Given a large reduction in plasma cefazolin levels over time, the effective tissue drug concentration may fall below a target level for common Gram-positive, and particularly, Gram-negative pathogens that were prevalent among our infected patients.²⁷

Our study showed a higher proportion of polymicrobial and Gram-negative infections compared with previously published literature.^{5, 28} This finding likely represents the patient population of our cohort, with the majority of the infected patients having neuromuscular scoliosis, a group in whom polymicrobial and Gram-negative infections are more frequent.^{8, 29, 30} The higher incidence of isolated Gram-negative bacteria in these patients is the reason behind the recommendation for additional antibiotic prophylaxis in high-risk pediatric populations. Although we were unable to extract high-risk cases due to the inconsistency of high-risk case database collection during the pandemic, patients with ASA classification of ≥3 and/or neuromuscular scoliosis presumably represented high-risk candidates. Given that ASA classification of ≥3 and neuromuscular scoliosis are risk factors for SSI, it is noticeable that the rate of gentamicin administration was lower than the number of these presumed high-risk patients. Additionally, gentamicin was not administered in some of the infected patients who had neuromuscular scoliosis. This possibly highlights the necessity of Gram-negative bacterial coverage in high-risk populations.

Prolonged operative time usually involves surgical complexity and significant blood loss. The longer the skin is exposed, the higher the risk of developing wound infection as skin flora gain access during surgery while the incision is open.^{31, 32} Even a small number of bacteria can cause infection in the presence of a prosthesis.³³ Appropriate choice, timing, and re-dosing of prophylactic antibiotic is essential to prevent SSI in this context. However, a standard re-dosing interval may not be effective in the presence of excessive blood loss. In adults, it is recommended that an additional dose of cefazolin should be administered when the blood loss is greater than 1,500 mL,^{19, 34} despite that, the threshold for pediatric patients has not yet been established. Significant intraoperative blood loss and fluid replacement also result in plasma and tissue gentamicin concentrations lower than MIC for Gram-negative bacteria.³⁵ We also identified prolonged anesthetic time as another risk factor for SSI. Given that the duration of anesthesia is substantially correlated with the duration of surgery, it is difficult to determine statistically which factors are more likely to be associated with SSI risk. Besides operative time, anesthetic time typically involves patient preparation, line access, positioning from supine to prone, and radiographic imaging. The prolonged duration of anesthesia indicates longer anesthetics exposure; most of our patients were maintained with volatile anesthetics that may predispose them to infections due to the immunosuppressive effects of volatile anesthetics as previously described.^{36, 37}

Decontamination of the skin with antiseptic prior to skin incision is one of the standard measures to prevent SSI. We found that the use of povidone-iodine alone was a potential risk for SSI. Povidone-iodine may not be suitable for spine surgery as it has only two-hour duration of action, in contrary to chlorhexidine-based solutions that could be effective for as long as 48 hours.³⁸ Another disadvantage for povidone-iodine is that transcutaneous iodine absorption may result in iatrogenic hyperthyroid or hypothyroidism in pediatric population.³⁹ Our hospital policy has recently changed to omit the use of povidone-iodine alone and require an alcohol-based skin preparation agent unless the patient is allergic to alcohol or has abundant hair.

Overweight or obesity, not low body weight, is more likely a risk factor for SSI according to the previous study.⁷ The caveats are possibly owing to that we enrolled all pediatric patients regardless of age, and that we did not standardize body weight with age-related percentile which might interfere with the validity of the result interpretation. Nevertheless, low body weight may indicate poor nutritional status and/or significant comorbidities, suggesting preoperative nutritional supplementation should be important.

Despite our institution being a large referral center for pediatric spine surgery with a large number of high-risk cases, the SSI rate is still relatively low, likely demonstrating the effectiveness of our SSI prevention bundle in place during the study period. The small number of SSI in our study limits the power to perform a multivariable analysis. Additionally, our patient cohort may not be representative of other institutions. Another potential limitation is overall SSI bundle reliability for each surgery. During this time overall bundle reliability (that is, compliance with all elements of the bundle) has ranged from 64% to 87%. Gaps in compliance with other elements of the bundle (such as lack of preoperative bathing or utilizing povidone-iodine alone as a skin preparation prior to incision) could have increased the risk of SSI for individual patients, and we did not account for these factors in the analysis. This study is also limited by its retrospective nature.

5. Conclusions

In the context of a standard SSI prevention bundle including antibiotic prophylaxis, we demonstrated a relatively low incidence of SSI following pediatric spinal fusion surgery. Administration of prophylactic antibiotic within 60 minutes of skin incision, strict adherence to high-risk protocol, and use of CHG-alcohol could potentially reduce the rate of SSI. However, larger prospective studies are needed to validate this finding before existing SSI prevention guidelines are modified.