Main Menu Browse by Day or Program Author Keywords

48 Multicenter Study of Post-operative Surgical Site Infections following Scoliosis Repair in Children

Saturday, April 2, 2011: 12:00 PM
Coronado A (Hilton Anatole)
Lisa Covington, RN, MPH , New York- Presbyterian Hospital, New York, NY
W. Stuart G. MacKenzie, BS , Columbia University Medical Center, New York, NY
Hiroko Matsumoto, MA , Columbia University Medical Center, New York, NY
Jacqueline Corona, MD , Columbia University Medical Center, New York, NY
Christopher Lee, BS , Children's Hospital of Los Angeles, Los Angeles, CA
Stephanie R. Cody, BS , Children's Hospital of Philadelphia, Philadelphia, PA
John M. Flynn, MD , Children's Hospital of Philadelphia, Philadelphia, PA
David L. Skaggs, MD , Children's Hospital of Los Angeles, Los Angeles, CA
Michael G. Vitale MD MPH , Columbia University Medical Center, New York, NY
Lisa Saiman MD MPH , Columbia University Medical Center, New York, NY
Background:  

Recent single center reports have suggested the increasing importance of gram negative bacilli (GNB) causing surgical site infections (SSIs) following scoliosis surgery in children. However, risk factors are not well understood including the risk of SSIs following vertical insertion and lengthening of expandable rods (VERs). We hypothesized that SSIs among patients with non-idiopathic scoliosis would be more likely to be caused by GNBs.

Objective:  

We sought to assess the present-day epidemiology and microbiology of SSIs following posterior spinal instrumentation in 3 pediatric referral hospitals.

Methods:  

We performed a retrospective study of SSIs at 3 centers that perform a large volume of spinal surgery, including VER procedures. We used the Center for Disease Control and Prevention’s National Healthcare Safety Network case definition for SSIs following surgery for scoliosis and patients were followed for one year after surgery. Multivariable regression analysis was used to assess risk factors associated with SSIs.

Results:   

From Jan 2006 – Dec 2008, 905 patients (38.7% with idiopathic and 61.3% with non-idiopathic scoliosis) underwent 1352 spine procedures. In all, 78 SSIs were identified for an SSI rate of 5.8% (site range: 3.0-7.7%). Scoliosis etiology was associated with risk of SSI; the SSI rate in idiopathic vs. non-idiopathic scoliosis was 2.9% vs. 7.6%, respectively (p<0.001). The SSI rate following VER insertion or lengthening/ revision was 10.8% (4/37) vs. 3.0% (4/135), respectively. GNB were isolated from 45.9% of SSIs and 41.0% of SSIs were polymicrobial infections. Multivariate regression analysis revealed that patients with SSIs caused by GNB were more likely to have non-idiopathic scoliosis than idiopathic scoliosis (31/61 [50.8%] vs. 1/13 [7.7%], p=0.009) and were older than those without GNB SSIs (12.8 + 4.5 vs. 9.3 + 4.8 years, p=0.003).

Conclusions:  

Patients with non-idiopathic scoliosis had higher rates of SSIs and were more likely to have GNB isolated from wound infections than patients with idiopathic scoliosis. In fact, GNB accounted for nearly half of SSIs. Preventive strategies should consider the local epidemiology of SSIs, including the pathogens causing SSIs; the risk associated with new technologies such as VERs; and the need to potentially modify antimicrobial prophylaxis.