Secondary Logo

Journal Logo

Original Studies

Surgical Site Infections in Pediatric Spine Surgery

Comparative Microbiology of Patients with Idiopathic and Nonidiopathic Etiologies of Spine Deformity

Maesani, Matthieu MD, MSc*†; Doit, Catherine MD, MSc†‡§; Lorrot, Mathie MD, PhD†¶‖; Vitoux, Christine MD; Hilly, Julie MD, MSc*†; Michelet, Daphné MD, MSc*†; Vidal, Christophe MD, MSc†**; Julien-Marsollier, Florence MD, MSc*†; Ilharreborde, Brice MD, PhD†**; Mazda, Keyvan MD, PhD†**; Bonacorsi, Stéphane MD, PhD†‡§; Dahmani, Souhayl MD, PhD*†††

Author Information
The Pediatric Infectious Disease Journal: January 2016 - Volume 35 - Issue 1 - p 66-70
doi: 10.1097/INF.0000000000000925


Although classified as clean surgery, spine surgery is at high risk of surgical site infection (SSI). This risk even increases for pediatric spine surgery with rates ranging from 3.7% to 8.5%.1–6 “Nonidiopathic” scoliosis patients, particularly neuromuscular patients, have the highest rates of infection ranging from 5.4% to 13.3%.7–9 Pediatric patients often require long surgeries, such as scoliosis, with extensive instrumentation and bone graft. SSI leads to significant morbidity, with the need of multiple surgeries, longer hospital stay and increased cost.3,4,10,11

In pediatric populations, malnutrition, obesity, preoperative urinary tract infection, length of surgery, blood loss, use of allograft, number of levels fused, use of sacral implants and inappropriate perioperative antimicrobial prophylaxis have all been highly suspected to be risk factors for SSI.12 Moreover, early onset infections, during the first 3 months, account for the majority of cases.7

It seems that pathogens involved in SSI are related to patients’ characteristics, such as the etiology of the spinal deformity. Although Staphylococcus aureus is the most frequent pathogen,1,4,7 it has been shown that Gram-negative bacilli (GNB) are more frequent in neuromuscular patients.7,12 Anaerobic bacteria are less often described in early onset SSI and involve Propionibacterium acnes in most cases of pediatric SSI.4,7 Only few studies explored the relationship between pathogens and spine deformity etiology.

The purpose of this study was to describe the pathogens found in early onset SSI during instrumented spine surgery, including antibiotics resistances, and the relationship between these pathogens and the etiology of the spinal deformity.


Study Design

We conducted a monocentric observational retrospective study in Robert Debre pediatric teaching hospital from January 2007 to January 2012. We reviewed the records of all cases of SSIs occurring in the early (from the first to the 30th postoperative days) and late (from the first to the 12th postoperative months) postoperative periods according to our national guidelines and Center for Disease Control and Prevention.13,14

The study was approved by a local ethic committee. Because of the retrospective observational nature of the study, the committee waived informed consent. Data files were anonymized according to national regulation.

The primary outcome of this study was to investigate the pathogens involved in the SSI in our population. Secondary outcomes were the resistance profile of the pathogens and the nature of these pathogens in relation with the nature of the spine deformity (idiopathic scoliosis vs. nonidiopathic scoliosis).

Study Population

All patients who had spine surgery with implants were included in this study. Patients undergoing surgeries without implant (ie, herniated disc discectomy) were excluded. Surgical techniques used are provided in the supplementary text (Supplemental Digital Content 1, Anesthesia was standardized in all patients with blood saving strategies, including preoperative recombinant erythropoietin or preoperative blood donation and intraoperative antifibrinolytic therapy. For antibiotic prophylaxis, according with our national recommendations, cefamandole was injected 1 hour before the incision, every 2 hours during the intraoperative period and every 8 hours during the first 48 postoperative hours. In case of penicillin allergy, clindamycin was used.

SSIs were defined prospectively when occurring by a local board of specialists, including surgeons, bacteriologists, radiologists and infectious diseases specialists, during the weekly meeting. Cases were defined on the basis of national (French)15 and international (Center for Disease Control and Prevention) guidelines14 according to at least one of the following categories: (1) purulent discharge of the scar or abscess in deep surgical site and (2) clinical signs of infection (fever, swelling or leaking of scar) and positive bacterial culture of superficial and/or deep surgical site sample. Local policy for SSI management included deep surgical debridement of all suspected cases of SSI (either superficial or deep) and deep cultures of the surgical site. Consequently, we defined cases as patients needing surgical debridement, and no classification was performed according to the superficial or deep location of the SSI. Early infection was defined as a SSI occurring between the first and the 30th day after surgery, whereas late SSI was defined as a SSI occurring between the first and the 12th month after surgery. Antibiotherapy was started during the intraoperative period of the debridement. Antibiotics were given intravenously during 15 days and orally for a total duration of 3 months. Follow-up of patients was ensured until at least 1 year after the resolution of the infection episode or as needed by the surgical and medical condition.

Epidemiologic characteristics and data about surgical and anesthetic management were collected from the patients’ files that were available (in 6 cases, patients files had moved to another hospital). Microbiological data included identification of pathogens and antibiotic susceptibility test. Bacteria were classified as S. aureus, coagulase-negative staphylococci (CNS), GNB, anaerobic bacteria and other bacteria.

Samples were inoculated onto sheep blood Columbia and chocolate agar (Bio-Mérieux, Marcy l'Etoile, France) incubated at 37°C in a 5% CO2 atmosphere and in anaerobic atmosphere for 10 days and were also inoculated in BacT/ALERT aerobic and anaerobic blood culture vials (Bio-Mérieux) incubated in BactT/ALERT system at 37°C for 6 days. When standard cultures were sterile, 16S rRNA gene amplification and sequencing were performed to identify pathogens. Antibiotic susceptibility test was performed according to national guidelines (Comité de l'antiobiogramme de la Société Française de Microbiologie).

Finally, pathogens were regarded as responsible (in contrast to contaminant) if present in the deep samples or only in the superficial samples if pathogens were not contaminants (Staphylococcus epidermidis and P. acnes).

Statistical Analysis

Descriptive statistics were displayed as median [range] for continuous variables and n (%) with their 95% confidence interval (according to Wilson methods with continuity correction16) for discrete variables.

A subgroup analysis was made according to the etiology of the spine disease between idiopathic illness and nonidiopathic ones as proposed by Mackenzie.7 The 2 subgroups were compared using the Mann–Whitney test for non-normal variables. Categorical variables were compared using χ2 or Fisher exact test. Significance level was 0.05.

Relation between characteristics of patients and type of pathogens was analyzed through χ2 or Fisher exact test.

All statistical analysis was conducted on SPSS software version 21 (IBM Corporation, Chicago, IL).


Study Population

Between January 2007 and January 2012, there were 496 spine procedures with implants in our institution. There were 51 cases of SSIs involving 49 patients (1 patient experienced 3 different infections following 3 different surgeries). Complete bacteriologic data were obtained for the 51 cases, and complete epidemiological and surgical data were obtained for 43 cases.

Microbiologic Findings

There were 42 cases of early infections and 9 late cases. The incidence rate of SSI was 10.3%. Of the 51 cases of infections, 74 bacteria were isolated. In 88% of cases (95% CI: 75–95) (n = 45), bacteria were found in deep intraoperative specimens (including identification of P. acnes through 16S rRNA amplification in 1 case). Gram-positive anaerobic cocci (GPAC) were not fully identified in this cohort because antibiograms allowing an adequate antibiotic treatment. In 4 other cases, bacteria were isolated on superficial scar swab only (Table 1: S. aureus, GNB and enterococcus). And in 2 cases, all cultures were sterile. In 1 of the latter cases, 16S rRNA identification was performed, but did not find any pathogen.

Microbiological Ecology in the Whole Cohort and According to Etiologic Groups

Three different patients had concomitant S. aureus bacteriemia: 2 patients with adolescent idiopathic scoliosis (AIS) and the last 1 with Scheuermann disease. Two other patients developed postoperative non-SSI and then developed SSI with the same germ: 1 patient with cerebral palsy got pyelonephritis and another patient got pneumopathy. Of 31 patients, who got infected with S. aureus, 19 were screened for nasal carriage of S. aureus, of whom 8 were positive. Of the 20 other patients, who did not get infected with S. aureus, 18 were screened and none had nasal carriage of S. aureus. Consequently, all patients who had S. aureus nasal carriage got infected with S. aureus, 100% (95% CI: 60–99, n = 8), whereas 38% (95% CI: 21–58, n = 11) of patients who had negative S. aureus nasal carriage had S. aureus infection.

Pathogens found in the whole population and for idiopathic and nonidiopathic subgroups along with bacterial resistance are shown in Table 1 and Figure 1. Plurimicrobial infection was found in 31% of cases (n = 16 cases), in 38% of idiopathic cases (n = 8 cases) and 26% of nonidiopathic (n = 8 cases). S. aureus was the most frequent microorganism in both groups. The second most frequent pathogens were GNB in nonidiopathic cases and GPAC in idiopathic cases.

Pathogens found according to etiological subgroup (percentage of pathogens per infection cases).

We found 1 methicillin-resistant S. aureus and 1 extended-spectrum β-lactamase–producing enterobacteriaceae. Based on antibiograms, if used as empirical antibiotherapy or antibioprophylaxy, first- or second-generation cephalosporins would have been inefficient on the infecting bacteria in 57% (n = 17 cases) of nonidiopathic infected cases, 13% (n = 4 case) of idiopathic infected cases and only 12% (n = 2 case) of AIS infected cases. Cefotaxime would have been inefficient in 40% (n = 12 cases) of infected nonidiopathic cases, 19% (n = 4 case) of idiopathic cases and only 12% (n = 2 case) of AIS infected cases. Forty-two percent of all bacteria were found clindamycin resistant (n = 31, 13% of S. aureus, 17% of CNS and 67% of GPAC). Consequently, clindamycin would have been inefficient in 55% of infected cases (n = 28 cases).

Relation Between Microbiological Findings and Spine Illness

There were 21 idiopathic cases and 30 nonidiopathic cases (details of etiologies are shown in Table 2). Complete data were obtained for 19 idiopathic and 24 nonidiopathic cases (Table 3). The incidence rates for infection were 6.8% and 15.9% in idiopathic and nonidiopathic cases, respectively. There were significantly more female in the idiopathic group and conversely more male in the nonidiopathic group. Nonidiopathic cases had a higher ASA score, more sacral fusion, less level instrumented, more anterior approach, less intrathecal morphine and more fluid replacement during the first 24 postoperative hours were less often on autologous transfusion program and were more frequently admitted in the intensive care unit for monitoring (Table 2).

Etiologies of Spinal Deformity in Patients (n = 49)
Characteristics of Cases With Surgical Site Infection

Characteristics of cases significantly related to GNB were nonidiopathic etiology relative risk (RR) for GNB: 3.25 (95% CI: 1.06–9.9; P = 0.039), neuromuscular diseases: 2.34 (95% CI: 1.12–4.9; P = 0.039), age <10 years: 2.8 (95% CI: 1.3–5.8; P = 0.025), weight <30 kg: 3.7 (95% CI: 1.1–5.3; P = 0.016) and sacral fusion: 2.7 [1.3–5.4] (P = 0.044). GNB was significantly associated with plurimicrobial infections RR 4.53 [1.9–10.8] (P = 0.001).


Elapsed time between primary surgery and debridement surgery was 21 days [1–273] and was not different for idiopathic and for nonidiopathic cases (22.5 days [6–273] and 21 days [1–180], respectively; P = 0.659). Five patients needed 2 deep debridements to sterilize the surgical site. One patient needed a total of 4 deep debridements with a vacuum-assisted closure and the removal of implants.


The results of our study can be summarized as follows: in pediatric spine surgery, S. aureus was the most frequent bacteria in both idiopathic and nonidiopathic cases, most often found as single pathogen. In contrast, GNB was the second most frequent pathogen in nonidiopathic cases and often found in polymicrobial cultures. Of those GNB, Pseudomonas aeruginosa was the most frequent. Third-generation cephalosporin resistance was frequent among GNB. Strikingly, we found a high proportion of GPAC in idiopathic cases: Finegoldia magna (previously known as Peptostreptococcus magnus)17 and unidentified GPAC. Clindamycin resistance was frequent among these bacteria. As expected by the epidemiology, idiopathic cases were more often female, with significantly higher weight and body mass index. Nonidiopathic cases were more often polypathological with higher ASA score; sacral instrumentation was exclusively done in nonidiopathic patients.

GPAC are commensal bacteria found on the skin and on the mucosae of patients.17,18 Contamination could occur through skin incision. Interestingly, these pathogens were found in healthy idiopathic patients, mainly women. This population is prone to acneic skin lesions (especially in the back), which are known to be colonizated by F. magna and other GPAC.19,20 This could explain why we found such an important proportion of those pathogens in our cohort. Although GPAC have been sometimes described in spine SSI,7,21 to our knowledge, this is the largest cohort of GPAC infections in pediatric spine surgery. However, difficulties to identify these bacteria might explain the low proportion of these pathogens in previous studies. In this study, anaerobic species were systematically looked for through appropriate cultures. F. magna is known as pathogen for primary bone and joint infections and has also been described in SSI in orthopedic surgery, especially during joint replacement surgery.17,22 It is sometimes difficult to isolate this bacterium, some cases being identified only by polymerase chain reaction.22 Although pathogenicity of F. magna is now well characterized during SSI, other GPAC like Anaerococcus prevotii has only been identified in gynecologic, soft tissue primary infections and in abdominal SSIs.17 Interestingly, in our study, we found other GPAC as SSI pathogens in orthopedic surgery.

GNB were described in previous studies and known to be more frequently isolated in SSI in neuromuscular patients.7,23 In our cohort, GNB was associated with all nonidiopathic cases, younger patients and those with lower weight, who frequently are congenital or dystrophic scoliosis. The impairment of sphincter control in these patients might explain the association between SSI and GNB. This hypothesis is supported by the fact that infection often begins at the distal end of the scar in patients who had sacral fusion.24 It has also been shown that the closer to the sacrum the surgical site is the more frequent GNB infections are.25 In addition, sacral fusion was proved to be an independent risk factor for infection.7 In our study, there was a significant association between GNB and sacral fusion. Cognitive impairment might also be involved in this association as previously described.26 However, it is not clear that whether contamination occurs intraoperatively through skin colonization or postoperatively through ascending contamination of the scar from the perineum or the stools.

Our study emphasizes the need for appropriate intraoperative preventive anti-GNB antibiotics in nonidiopathic patients, as a high proportion of GNB was resistant to cefamandole and cefazolin, which are recommended as antibiotic prophylaxis.12 After this study, we shift to a cefazolin–gentamycin antibiotic prophylaxis in non-AIS patients. Clindamycin is commonly used as an alternative prophylaxis in case of allergy.12 Because of the high percentage of bacteria resistant to clindamycin, we replaced clindamycin by vancomycin for prophylaxis in case of penicillin allergy. Some authors advocated systematic vancomycin prophylaxis because of a high number of MRSA and CNS, which was not the case in our cohort.2 As for empiric antibiotherapy of SSI, we also took into account cefotaxime resistance in non-AIS patients for empirical antibiotherapy. In addition, given the presence of S. aureus carriage in all infected cases with this pathogens and the prevalence of GPAC in our cohort, 2 measures were introduced in our daily practice: (1) a systematic screening of S. aureus and its treatment in case of carriage and (2) a careful preoperative skin preparation by patients, nurses and parents, especially for patients with acne lesions. This consists of a preoperative iodine antiseptic shower of all the body (with a focus of the operative site) the day before and the same day surgery.

Comparing to other cohorts, we had relatively high prevalence of SSI especially in idiopathic patients.1–9 Interestingly, most germs found in this population are commensal of the skin (Staphylococcus, GPAC). Contamination might occur perioperatively from skin incision and might be preventable by adequate skin preparation.

We did not find any bacteria in deep surgical site in one eighth of our cases. Some studies highlight the usefulness of systematic broad-range polymerase chain reaction (16S ribosomal DNA polymerase chain reaction) for the diagnosis of osteoarticular infections in culture-negative samples.27 As we treated all suspected SSI by surgical debridement through every layer of the scar, we did not differentiate superficial and deep wound infection, thus we consider all SSI as deep. This may have overestimated the number of deep SSI.

After this work, it could be of interest to investigate the mode of contamination, indeed if contamination occurs mostly through skin incision, care should be given to skin preparation. Given the high incidence of GPAC, we should evaluate treatment of acne before scoliosis surgery in an adolescent. In nonidiopathic scoliosis, it should be investigated whether stool cultures might help aiming the right bacteria and whether stool decontamination might be of any help.

Our study suffers from some limitations. First, this is a monocentric study and bacterial ecology may only reflect our center ecology, as bacterial epidemiology is related both to patients and to local ecology as shown by the variability of the bacteria found among the studies.4 We only had 1 case of MRSA, whereas some North American series exhibit higher percentage.3,7,21,28 Nevertheless, these results are consistent with previous studies, and bacterial epidemiology seems to be more related to etiologic groups than specific centers.4,7,12 Because of the number of patients included, we did not conduct any multivariate analysis of the relationship between pathogens and patients characteristics.


1. Bachy M, Bouyer B, Vialle R.. Infections after spinal correction and fusion for spinal deformities in childhood and adolescence. Int Orthop. 2012;36:465–469
2. Ho C, Skaggs DL, Weiss JM, et al. Management of infection after instrumented posterior spine fusion in pediatric scoliosis. Spine (Phila Pa 1976). 2007;32:2739–2744
3. Cahill PJ, Warnick DE, Lee MJ, et al. Infection after spinal fusion for pediatric spinal deformity: thirty years of experience at a single institution. Spine (Phila Pa 1976). 2010;35:1211–1217
4. Gerometta A, Rodriguez Olaverri JC, Bitan F.. Infections in spinal instrumentation. Int Orthop. 2012;36:457–464
5. Labbé AC, Demers AM, Rodrigues R, et al. Surgical-site infection following spinal fusion: a case-control study in a children’s hospital. Infect Control Hosp Epidemiol. 2003;24:591–595
6. Linam WM, Margolis PA, Staat MA, et al. Risk factors associated with surgical site infection after pediatric posterior spinal fusion procedure. Infect Control Hosp Epidemiol. 2009;30:109–116
7. Mackenzie WG, Matsumoto H, Williams BA, et al. Surgical site infection following spinal instrumentation for scoliosis: a multicenter analysis of rates, risk factors, and pathogens. J Bone Joint Surg Am. 2013;95:800–806
8. Sponseller PD.. Pediatric revision spinal deformity surgery: issues and complications. Spine (Phila Pa 1976). 2010;35:2205–2210
9. Master DL, Poe-Kochert C, Son-Hing J, et al. Wound infections after surgery for neuromuscular scoliosis: risk factors and treatment outcomes. Spine (Phila Pa 1976). 2011;36:E179–E185
10. Calderone RR, Garland DE, Capen DA, et al. Cost of medical care for postoperative spinal infections. Orthop Clin North Am. 1996;27:171–182
11. Collins I, Wilson-MacDonald J, Chami G, et al. The diagnosis and management of infection following instrumented spinal fusion. Eur Spine J. 2008;17:445–450
12. Glotzbecker MP, Riedel MD, Vitale MG, et al. What’s the evidence? Systematic literature review of risk factors and preventive strategies for surgical site infection following pediatric spine surgery. J Pediatr Orthop. 2013;33:479–487
13. La Societe de Pathologie Infectieuse de Langue Francaise, College des Universitaires de Maladies Infectieuses et Tropicales, Groupe de Pathologie Infectieuse Pédiatrique. . [Clinical practice recommendations. Osteoarticular infections on materials (prosthesis, implant, osteosynthesis]. Medecine et maladies infectieuses. 2009;39:815–863
14. Horan TC, Andrus M, Dudeck MA.. CDC/NHSN surveillance definition of health care-associated infection and criteria for specific types of infections in the acute care setting. Am J Infect Control. 2008;36:309–332
15. Comité technique des infections nosocomiales et des infections liées aux soins. Actualisation de la définition des infections nosocomiales. 2007 Paris Ministère de la santé, de la jeunesse et des sports
16. Gosh B.. A comparison of some approximate confidence intervals for binomial parameter. J Am Statis Assoc. 1979;74:894–900
17. Murphy EC, Frick IM.. Gram-positive anaerobic cocci–commensals and opportunistic pathogens. FEMS Microbiol Rev. 2013;37:520–553
18. Veloo AC.Dongyou L. Finegoldia magna. Molecular Detection of Human Bacterial Pathogens. 2011 New South Wales, Australia CRC Press
19. Kwon HH, Yoon JY, Park SY, et al. Analysis of distribution patterns of Propionibacterium acnes phylotypes and Peptostreptococcus species from acne lesions. Br J Dermatol. 2013;169:1152–1155
20. Brook I, Frazier EH, Cox ME, et al. The aerobic and anaerobic microbiology of pustular acne lesions. Anaerobe. 1995;1:305–307
21. Rihn JA, Lee JY, Ward WT.. Infection after the surgical treatment of adolescent idiopathic scoliosis: evaluation of the diagnosis, treatment, and impact on clinical outcomes. Spine (Phila Pa 1976). 2008;33:289–294
22. Levy PY, Fenollar F, Stein A, et al. Finegoldia magna: a forgotten pathogen in prosthetic joint infection rediscovered by molecular biology. Clin Infect Dis. 2009;49:1244–1247
23. Sponseller PD, LaPorte DM, Hungerford MW, et al. Deep wound infections after neuromuscular scoliosis surgery: a multicenter study of risk factors and treatment outcomes. Spine (Phila Pa 1976). 2000;25:2461–2466
24. Szöke G, Lipton G, Miller F, et al. Wound infection after spinal fusion in children with cerebral palsy. J Pediatr Orthop. 1998;18:727–733
25. Abdul-Jabbar A, Berven SH, Hu SS, et al. Surgical site infections in spine surgery: identification of microbiologic and surgical characteristics in 239 cases. Spine (Phila Pa 1976). 2013;38:E1425–E1431
26. Lipton GE, Miller F, Dabney KW, et al. Factors predicting postoperative complications following spinal fusions in children with cerebral palsy. J Spinal Disord. 1999;12:197–205
27. Fenollar F, Roux V, Stein A, et al. Analysis of 525 samples to determine the usefulness of PCR amplification and sequencing of the 16S rRNA gene for diagnosis of bone and joint infections. J Clin Microbiol. 2006;44:1018–1028
28. Pull ter Gunne AF, Mohamed AS, Skolasky RL, et al. The presentation, incidence, etiology, and treatment of surgical site infections after spinal surgery. Spine (Phila Pa 1976). 2010;35:1323–1328

surgical site infection; spine surgery; pediatric

Supplemental Digital Content

Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved.