Su r g i c a l site infection is a significant complication

Transcription

1 J Neurosurg Spine 13: , 2010 Surgical site infection in spinal surgery: detection and management based on serial C-reactive protein measurements Clinical article By u n g-uk Ka n g, M.D., Sa n g -Ho Le e, M.D., Ph.D., Yo n g Ahn, M.D., Ph.D., Wo n-ch u l Ch o i, M.D., a n d Yo u n g -Ge u n Ch o i, M.D. Departments of Neurosurgery, Wooridul Spine Hospital, Daegu and Seoul, Korea Object. C-reactive protein (CRP) is a well-known sensitive laboratory parameter that shows an increase within 6 hours after the onset of bacterial infection. In relation to surgery, a normal CRP response is a rapid increase followed by a gradual reduction, eventually returning to the normal range. The goal of this study was to determine the diagnostic significance of CRP as a detector for early onset surgical site infection in spinal surgery and to discuss effective medical treatment through clinical interpretation and application of the measured CRP values. Methods. A prospective study was performed in 348 consecutive cases involving patients who underwent spinal surgery under general anesthesia between February and September Blood samples were obtained preoperatively and on postoperative Days 1, 3, and 5 in patients undergoing single-level decompression surgery. An additional blood specimen was obtained at postoperative Day 7 in patients requiring more extensive surgeries. Recorded laboratory results were compared with the patients clinical course to determine the diagnostic significance of CRP. All of the patients received intravenous prophylactic antibiotic therapy. Once an abnormal response of CRP, indicated by a tendency toward continuous elevation, was noted on Day 5 or Day 7, the prophylactic antibiotics were replaced with another regimen and administration was resumed along with careful observation for signs of surgical site infection. Results. Monitoring of CRP revealed a characteristic increase and decrease pattern in 332 of 348 patients (95.4%) showing a normal clinical course with regard to early infectious complications. The mean measured CRP (reference range < 4 mg/l) averaged 14.9 ± 20.3 mg/l on Day 1, 15.4 ± 25.1 mg/l on Day 3, and 7.9 ± 13.3 mg/l on Day 5. In contrast, there were 16 cases (4.6%) of abnormal CRP responses resulting in the resumption of intravenous antibiotic treatment, which included a second rise (in 12 cases) and a steady rise (in 4) in the CRP value. Five (1.4%) of 16 patients experienced infectious complications related to spinal surgery. Three patients (0.9%) received long-term antibiotic therapy for 4 6 weeks; however, all patients recovered with medical treatment alone and did not experience gross wound disruption or subsequent discitis. As a predictor for early wound infection, the sensitivity, specificity, positive predictive value, and negative predictive value for abnormal CRP responses were calculated as 100%, 96.8%, 31.3%, and 100%, respectively. Conclusions. The above results demonstrate that CRP screening is a simple and reliable test for the detection of early infectious complications after spinal surgery. Close observation and appropriate medical management should be performed in a timely fashion when abnormal CRP responses are observed at 5 or 7 days after surgery. (DOI: / SPINE09403) Ke y Wo r d s surgical site infection spinal surgery antibiotic treatment C-reactive protein Su r g i c a l site infection is a significant complication in spinal surgery as it is associated with prolonged morbidity and poor functional outcome. The occurrence is relatively frequent, and the reported incidence varies from 0.7 to 6%. 2,6,10,17,22,26,27 Surgical site infection often requires long-term antibiotic therapy, surgical debridement, and a prolonged Abbreviations used in this paper: CRP = C-reactive protein; ESR = erythrocyte sedimentation rate; NPV = negative predictive value; PPV = positive predictive value; SSI = surgical site infection. hospital stay. Appropriate prevention and early detection of infectious complications can lead to improved outcomes. Of great interest is a simple and reliable screening test for detection of early postoperative infection, which would allow an appropriate therapeutic plan to be made and proper counseling to be given to patients. C-reactive protein is an acute-phase protein synthesized by hepatocytes. It is present only in trace amounts in healthy subjects, but an increase in the CRP value is observed within 6 hours after the onset of bacterial infection. 4,12,20,21,29 In relation to surgery, the normal CRP response is rapid production until the peak level is reached; 158 J Neurosurg: Spine / Volume 13 / August 2010

2 C-reactive protein for early detection of infection this is followed by reduction and an eventual return to the normal range. 1,3,15,19,28,29 Compared with other laboratory tests such as ESR and leukocyte count, CRP has gained acceptance for the detection of infectious complications due to its predictable kinetics and high speed of response. 7,13,18,19,28,29 To date, however, there has been no consensus regarding the appropriate interpretation or clinical application of serial changes in CRP after spinal surgery. In the current study, the authors describe their experiences with serial measurement of CRP values as an indicator of early wound infection following various types of spinal surgeries. The purpose of this study was to determine the efficacy of CRP as a detector of early wound infection and to discuss the management of infectious complications through clinical application of the measured CRP values. Methods Between February 2008 and September 2008, data were prospectively collected in consecutive patients who met the inclusion criteria for this study and underwent spinal surgery. The inclusion criteria were as follows: elective spinal surgery under general anesthesia, age over 18 years, and preoperative CRP value within the normal range. Exclusion criteria were history of medical conditions affecting the normal response of CRP, including autoimmune disease, immunosuppressive treatment, inflammatory disorder, liver disease, malignancy, and a recent history of other infection. The data collected included characteristics of patients, history of medical disorders, type of surgery, number of spinal levels treated, estimated blood loss, surgical time, and length of hospital stay. Institutional review board approval was obtained for this study. Blood samples were taken preoperatively, and on postoperative Days 1, 3, and 5 to evaluate the CRP and ESR values in patients who underwent single-level decompression surgery. In the cases in which multilevel decompression surgery or surgery with implantation was performed, an additional blood sample was obtained on postoperative Day 7 to determine the above parameters. The reference value for CRP value in adults was < 4.0 mg/l. For ESR, the reference values were < 9 mm/hour for women and 15 mm/hour for men. In cases in which patients were discharged before the completion of scheduled measurements, follow-up samplings were performed in the outpatient department. A positive test was defined as a steady rise or a second rise in either parameter at Day 5 or Day 7. Cases with a minimal increase within the reference range for example, an increase in CRP value from 1.1 mg/l at Day 3 to 1.4 mg/l at Day 5 were not included in the positive results. Laboratory results were compared with the postoperative clinical course of the patients to establish the diagnostic significance of each parameter. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of both parameters were determined based on the results. All of the patients received prophylactic antibiotic therapy (cefmetazole 2 g per day) for 3 days after surgery, starting within 1 hour of the skin incision. If an abnormal J Neurosurg: Spine / Volume 13 / August 2010 response in CRP (a positive CRP test) was observed at 5 or 7 days after surgery, intravenous antibiotic therapy was resumed with ampicillin/sulbactam and isepamicin, after careful examination for evidence of early wound infection. In the presence of an allergic reaction, vancomycin was considered as a secondary choice. Whether to continue with antibiotic therapy was decided on the basis of the follow-up laboratory tests and the clinical presentation of the patient. The antibiotic treatment was discontinued once the elevated CRP showed a sharp decline without findings of SSI at follow-up measurements. The results of the microbiological tests were reviewed when available. Magnetic resonance imaging was considered when patients with an abnormal CRP response showed constitutional symptoms or signs of wound infection. Positive cultures of wound drains and/or evidence of wound infection confirmed by MR imaging that resulted in prolonged antibiotic treatment were recorded as infectious complications. In cases in which wound infection was confirmed, antibiotic therapy was continued until the CRP value was normalized. Statistical analysis was performed using the Kruskal- Wallis test to ascertain whether a significant difference in peak CRP values existed among the patients who had undergone single-level decompression, multilevel decompression, or instrumentation surgeries. If a significant difference was found, a post hoc Tamhane T2 test was used to determine the kinds of surgeries for which the values were significantly different. A probability value < 0.05 was considered statistically significant. Results Overall, 348 patients (184 men, 164 women) who met the inclusion criteria were enrolled in this study. Their mean age was 53.4 years. Clinical follow-up was performed at 1, 6, and 12 months after surgery, and the mean duration of follow-up was 11 months (range 6 13 months). Characteristics of the patients and information on surgery and hospitalization are summarized in Table 1. After surgery, the measured values of CRP and ESR increased in all of the patients. The serial monitoring of CRP revealed a characteristic increase-and-decrease pattern, although the peak amplitude varied somewhat between the surgeries. The CRP values in 332 patients (95.4%) reached a maximum on the 1st or the 3rd day after surgery with rapid decrease at serial follow-up. The mean measured CRP was 14.9 ± 20.3 mg/l on Day 1 and 15.4 ± 25.1 mg/l on Day 3 in this group of patients. Normal values were regained in 143 (43%) of the 332 patients within 5 days after surgery, and the mean value for all 332 patients on Day 5 was 7.9 ± 13.3 mg/l (Fig. 1). During hospitalization, all of these patients had uneventful clinical courses in relation to infectious complications. There were 16 cases (4.6%) of abnormal CRP responses resulting in the resumption of intravenous antibiotic treatment, which included a second rise in CRP level in 12 cases and a steady rise pattern in 4 (Table 2). During hospitalization, 8 of these 16 patients experienced infectious complications; these complications were related to surgery in 5 cases and unrelated in 3. Two patients 159

3 B. U. Kang et al. TABLE 1: Summary of demographic and surgical data in 348 patients Variable Value mean age in yrs 53.4 ± 15.5 sex male 184 female 164 region cervical 27 thoracic 2 lumbosacral 319 type of surgical procedure single-level decompression 221 multilevel decompression 44 surgery w/ implantation 83 type of surgical approach anterior 15 posterior 304 combined 29 mean operative time in min 123 ± 67 mean estimated blood loss in ml 214 ± 149 mean length of hospital stay in days 8.7 ± 8.3 experienced acute gastroenteritis accompanied by watery diarrhea, and 1 patient suffered from a symptomatic urinary tract infection. There were 5 patients in whom early wound infection was diagnosed through MR imaging and/or positive cultures, accounting for an overall incidence of 1.4%. Three patients received long-term antibiotic treatment to achieve control of their deep wound infection; however, none of these patients required surgical treatment. In contrast, in the other 8 patients with abnormal CRP responses the CRP values returned to the normal range without significant infectious complications. Remarkably, 4 of these 8 patients suffered from increased axial pain and general ache until their CRP was normalized, but did not have a gross wound problem. Compared with the CRP responses, the serial measurements for ESR showed slow and inconsistent change after surgery. The ESR had a steady rise pattern in most cases, and reached a maximum value on the 5th day after surgery on average (Fig. 2). During the follow-up period, all but 1 of the patients were free from any new infectious complication related to spinal surgery. One male patient who had undergone single-level lumbar decompression developed a retrodiscal abscess 4 weeks after surgery. He had a postoperative course, which was similar to that of the patients with an uneventful course, with normal-appearing wound healing. Chart review revealed a characteristic increase and decrease pattern of CRP that was normalized within 5 days after spinal surgery. The abscess was treated without sequelae after surgical revision followed by 3 weeks of intravenous antibiotic therapy. Negative cultures were reported from specimens obtained during revision surgery. Fig. 1. Mean CRP values for 332 patients without infectious complications. The labels C1, C3, C5, and C7 refer to CRP measurements on Day 1, Day 3, Day 5, and Day 7, respectively. The maximum level of CRP after instrumentation surgery was higher compared with the peak value following single-level decompression surgery (p < 0.001) and multilevel decompression surgery (p < 0.01). The peak CRP value after single-level decompression was significantly lower than the peak value after each of the other types of surgery. Predictor of Infectious Complications When compared with the clinical course, examination of CRP quantification revealed true positive in 5 and false positive in 11 patients. True negative results were recorded in 331 patients, and 1 false negative result was observed. As a predictor for SSI, therefore, the sensitivity, specificity, PPV, and NPV were calculated as 83.3%, 96.8%, 31.3%, and 99.7%, respectively. Except in the case of delayed onset wound infection, CRP had a sensitivity of 100%, specificity of 96.8%, PPV of 31.3%, and NPV of 100% as a predictor for early wound infection. For ESR, the sensitivity, specificity, PPV, and NPV for early wound infection were 100%, 53.6%, 3%, and 100%, respectively. Discussion Based on the time to clinical presentation, spinal SSI can be categorized as early, delayed, or late onset. 3 In the delayed or late-onset types, the infection generally occurs as postoperative discitis or indolent microbial contamination of tissue that presents weeks or months after surgery usually without the preceding breakdown of the overlying skin. 5,8,11,23 The most common type is early onset; in this type of SSI, the signs and symptoms are manifested as a superficial/deep incisional infection within a few days or weeks of surgery. 3 Early detection and prompt treatment are 160 J Neurosurg: Spine / Volume 13 / August 2010

4 C-reactive protein for early detection of infection TABLE 2: Characteristics of cases involving patients with abnormal postoperative CRP responses* C a s e No. Op CRP (mg/l), ESR (mm/hr), WBC (x 10³/mm 3 ) Preop Day 1 Day 3 Day 5 Day 7 Causes for CRP Elevation Appearance of Surgical Site Other Findings Results of Drain Culture 1 MD 0.4, 3, , 5, , 30, , 46, , 60, 9.2 AGE NS watery diarrhea no growth 2 IS 0.5, 16, , 28, , 50, , 58, , 71, 5.6 AGE NS watery diarrhea no growth 3 IS 0.4, 4, , 4, , 11, , 32, , 35, 7.4 UTI NS urinary frequency no growth 4 SD 0.6, 5, , 10, , 24, , 91, , 100, 13.1 deep wound inf inflam signs w/o drainage low-gr fever, incr axial pain S. epidermidis 4 wks 5 SD 0.5, 16, , 17, , 55, , 76, , 95, 7.6 deep wound inf inflam signs w/o drainage low-gr fever, incr axial pain S. epidermidis 6 wks 6 IS 1.3, 35, , 40, , 90, , 107, , 120, 6.7 deep wound inf inflam signs w/o drainage low-gr fever, incr axial pain no growth 4 wks 7 MD 0.9, 28, , 35, , 54, , 65, 7 52, 115, 12.3 superficial inflam signs w/o drainage low-gr fever/incr axial pain gram-pos bacilli 2 wks wound inf 8 MD 0.9, 32, 8 7.9, 39, , 50, , 37, , 69, 17.4 superficial inflam signs w/o drainage low-gr fever, incr axial pain S. epidermidis 2 wks wound inf 9 SD 3, 13, , 15, , 31, , 50, , 80, 11.3 unknown NS incr axial pain no growth 6 days 10 SD 1.6, 27, , 28, , 23, , 75, , 73, 7.3 unknown NS incr axial pain no growth 2 days 11 IS 0.9, 34, , 25, , 59, , 48, , 64, 6.9 unknown NS incr axial, radicular pain no growth 2 days 12 IS 0.7, 24, , 25, , 59, , 45, , 73, 7.8 unknown NS incr axial pain no growth 6 days 13 IS 1.3, 18, , 23, , 29, , 43, , 56, 6.5 unknown NS NS no growth 2 days 14 MD 3.1, 40, , 31, , 50, , 75, , 72, 6.8 unknown NS NS no growth 2 days 15 IS 0.7, 19, , 23, , 32, , 30, , 53, 8.7 unknown NS NS no growth 2 days 16 SD 0.8, 3, 8.6 1, 2, , 2, , 10, , 19, 11 unknown NS NS unavailable 2 days Duration of ABx Tx * Reference ranges: CRP < 4 mg/l, ESR < 9 mm/hour for male and 15 mm/hour for female, WBC ,000 cells/mm 3. Abbreviations: ABx = antibiotic; AGE = acute gastroenteritis; gr = grade; incr = increased; inf = infection; inflam = inflammatory; IS = surgery with instrumentation; MD = multilevel decompression; NS = nonsignificant; pos = positive; SD = single-level decompression; UTI = urinary tract infection ; WBC = white blood cell. J Neurosurg: Spine / Volume 13 / August

5 B. U. Kang et al. crucial to a successful outcome; however, the detection of impending infection is quite challenging. 1,3,18,26,28 In the early stage of infection, constitutional symptoms such as axial pain, fever, chills, and night sweats are not always present. Clinical examination cannot reliably detect infection in this incubation period. For timely management, an objective and reliable method that enables the early detection of impending infection is required. Many authors have reported methods for the detection of postoperative spinal infection using radionuclide studies, plain radiographs, spinal biopsy, MR imaging, and laboratory tests, including CRP and ESR. 14,24,25,30 Of these methods, CRP screening has been proven to be valuable in the detection of early postoperative infectious complications and is gradually gaining acceptance. 18,19,28 An increase in CRP is a normal response to surgery. In uncomplicated operations, CRP reaches its peak value at 1 3 days after surgery, rapid reduction occurs from 3 to 5 days, and then more gradual reduction follows until normal CRP values are reached. 15,19,28,29 A continuous or renewed elevation of CRP after the initial peak period could be a signal of the onset of an infectious complication. 19,28 In contrast, the ESR peak occurs on approximately the 5th postoperative day and decreases more slowly and with a more irregular pattern than CRP. Sometimes ESR remains elevated even a few weeks or months after surgery, which renders ESR of limited value in the detection of early wound infection. 13,15,29 The results of this study were very similar to those published in previous reports. In uncomplicated cases, CRP values reached a maximum on the 1st or 3rd day after surgery followed by rapid reduction at serial measurements. Normal values were regained within 5 days after surgery in 43% of the patients. Only 4.6% of the patients had abnormal CRP responses that might have been associated with ongoing infectious complications. In spite of the results indicating CRP as a sensitive marker for infectious complications, there is another obstacle to the accurate diagnosis of SSI using CRP measurements. An elevation in CRP and/or ESR is not pathognomonic for a wound infection. The presence of an intercurrent infection or other inflammatory process may be a cause for an elevated CRP level. When an abnormal CRP response is detected, therefore, a question is raised as to whether prompt intervention should be undertaken. No consensus has been reached on a protocol that can be applied to patients in these circumstances. The authors strategy for management was an immediate return to antibiotic therapy when an abnormal CRP response was detected and when there was no clear evidence of intercurrent infection or other inflammatory processes. In the present series, 16 of the patients had a definite elevation of CRP after the initial peak period, which might suggest an ongoing infectious event. Except for the 3 cases involving patients with infectious complications unrelated to surgery, the authors response to the elevated CRP levels included close observation for signs of SSI, resumption of intravenous antibiotic therapy with a different regimen, and serial monitoring of the laboratory parameters. Because the abnormal CRP rise suggests that the prophylactic antibiotics were ineffective, the Fig. 2. Mean ESR for 332 patients without infectious complications. The labels E1, E3, E5, and E7 refer to ESR measurements on Day 1, Day 3, Day 5, and Day 7, respectively. change of antibiotic agents constituted the most crucial part of the management. Within 7 days of antibiotic therapy with the changed regimen, all but 5 patients showed a gradual decrease in CRP values at follow-up measurements. These patients had a normal postoperative course without further infectious complications after discharge. In the remaining 5 patients, however, the constitutional symptoms suggesting SSI continued 6 10 days after surgery. Remarkably, the elevation of CRP always preceded the presentation of constitutional symptoms. Of these 5 patients, 3 had a CRP value on Day 5 that was several times higher than the value of the other patients with elevated CRP levels, and these 3 patients required antibiotic treatment for more than 4 weeks. Magnetic resonance imaging revealed deep wound infection with abscess formation in these patients. Successful control of SSI was achieved with medical treatment alone, and no patients experienced gross wound disruption requiring surgical intervention or subsequent discitis. The above results demonstrate that early awareness of a potential deep space infection is a crucial factor in the successful outcome of treatment. The results also indicate that appropriate medical treatment should be considered before purulent discharge or gross wound dehiscence develops. Although the administration of antibiotics should be delayed in principle until reports on positive cultures are obtained, delays of only a few days may cause the infection to be medically uncontrollable, eventually resulting in poor functional outcomes. During the study period, intravenous antibiotic treatment was resumed before the results of the surgical drain culture were available. A combination therapy of ampicillin/sulbactam and isepamicin was empirically chosen because this regimen is 162 J Neurosurg: Spine / Volume 13 / August 2010

6 C-reactive protein for early detection of infection effective against most gram-positive and gram-negative organisms except for methicillin-resistant Staphylococcus aureus. Fortunately, in most cases with positive cultures, the cultures revealed the growth of S. epidermidis, which was sensitive to the selected antibiotic agents. In terms of effective medical treatment, therefore, the authors recommendation is the implementation of a different regimen before a gross wound problem develops. The few days of the latent period between CRP elevation and a gross wound problem is believed to play a decisive role in the progression of infection and the resulting clinical outcomes. Another advantage of serial CRP quantification is the accurate demonstration of absence of postoperative infectious complication. A negative test on Day 5 or Day 7 rules out an early infectious complication, which allows for earlier institution of a routine treatment plan. In a study involving 400 patients who underwent singlelevel microdiscectomy, Meyer et al. 18 reported that the sensitivity and NPV for serial CRP testing was 100%. The calculated specificity reached 95.8%, and PPV was 48.4%. They concluded that normal results of a CRP test performed on Day 5 ruled out postoperative infection, and thus no further examination was required. In relation to instrumentation surgery, Takahashi et al. 28 suggested that postoperative infection should be suspected and that the antibiotic regimen should be changed or added to when the CRP level is higher on Day 7 than on Day 4. In another study involving 149 patients who underwent decompression alone or decompression with placement of instrumentation, Mok et al. 19 endeavored to derive the normal kinetics of CRP levels associated with surgery. They reported that deviation from normal CRP kinetics, as represented by a second rise or failure to decrease, had good sensitivity and a good predictive value for postoperative infection. Another remarkable finding in their study was that a greater CRP peak value was significantly related to number of surgical levels. The present results are generally consistent with previous results. As a predictor for SSI, a positive CRP test had a sensitivity of 83.3%, specificity of 96.8%, PPV of 31.3%, and NPV of 99.7%. When the criteria of infectious complications are limited to the cases with early onset infections, the sensitivity, specificity, PPV, and NPV were increased to 100%, 96.8%, 31.3%, and 100%, respectively. Since SSIs occurring after a prolonged period of time may have a hematogenous origin or result from the intraoperative seeding of microbes, with infections remaining subclinical for an extended period of time, 9,31 CRP monitoring in the early period may have no predictive value with regard to late infection. The patient with a delayed retrodiscal abscess in this study demonstrates that CRP monitoring has limited value in detecting infections with delayed onset, in addition to its known limitation in predicting late-onset infections. 18 There has been no consensus regarding the period of measurements required to rule out early postoperative infection. The authors recommend an additional measurement at Day 7 in cases of extensive surgery. In extensive surgeries, which produce a large amount of CRP, it would take more time to degrade the high plasma level of CRP J Neurosurg: Spine / Volume 13 / August 2010 to the normal range. It seems logical that additional measurements should be obtained in those cases. On the basis of the above results, therefore, the authors protocol for the use of CRP measurements is as follows. A negative CRP test on Day 5 is considered to rule out a postoperative infection following short-segment decompression surgeries. Additional measurement may be required in more extensive surgeries. A positive CRP test requires close examination for ongoing infectious complications and justifies the resumption of antibiotic therapy with different agents after ruling out intercurrent infections or other inflammatory responses. Despite advances in infection control practices, SSI remains a substantial cause of morbidity and mortality among hospitalized patients. 16 Prophylactic antibiotic treatment can help to decrease the incidence of infection; however, prophylaxis alone cannot eradicate SSI. 32 Concerning the optimal prophylaxis, review of the current literature does not specify whether a particular antibiotic protocol or dosing regimen is superior to another in the prevention of SSI. 32 Those uncertainties regarding the prevention and management of SSI underscore the necessity for a reliable detector and protocol for patients with the risk of SSI. Overall, it is expected that these results will help in the perioperative care of patients who undergo surgical procedures. Conclusions C-reactive protein screening is a simple and reliable test for the detection of early infectious complications after spinal surgery. Normal CRP responses can rule out an early SSI. Close observation for ongoing infectious complications is required when abnormal CRP responses are observed at 5 or 7 days after surgery. The resumption of different antibiotic agents should be considered in a timely fashion after ruling out intercurrent infections or other inflammatory responses. Disclosure The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper. Author contributions to the study and manuscript preparation include the following. Conception and design: Kang, Lee, WC Choi. Acquisition of data: Kang. Analysis and interpretation of data: Kang. Drafting the article: Kang. Critically revising the article: Kang. Reviewed final version of the manuscript and approved it for submission: all authors. Statistical analysis: Kang. Administrative/technical/ material support: Kang. Study supervision: Kang, Lee, Ahn, YG Choi. Acknowledgments The authors thank Mi-Sung Oh, Young-Jin Son, and Bu Hee Kim for their assistance in preparing this manuscript. References 1. An HS, Seldomridge JA: Spinal infections: diagnostic tests and imaging studies. Clin Orthop Relat Res 444:27 33, Bassewitz HL, Fischgrund JS, Herkowitz HN: Postoperative spine infections. Semin Spine Surg 12: ,

7 B. U. Kang et al. 3. Boden SD, Bohlman HH: The Failed Spine. Philadelphia: Lippincott Williams & Wilkins, 2003, pp Carr WP: The role of the laboratory in rheumatology. Acutephase proteins. Clin Rheum Dis 9: , Clark CE, Shufflebarger HL: Late-developing infection in instrumented idiopathic scoliosis. Spine 24: , Esses SI, Sachs BL, Dreyzin V: Complications associated with the technique of pedicle screw fixation. A selected survey of ABS members. Spine 18: , Foglar C, Lindsey RW: C-reactive protein in orthopedics. Orthopedics 21: , Hatch RS, Sturm PF, Wellborn CC: Late complication after single-rod instrumentation. Spine 23: , Heggeness MH, Esses SI, Errico T, Yuan HA: Late infection of spinal instrumentation by hematogenous seeding. Spine 18: , Hodges SD, Humphreys SC, Eck JC, Covington LA, Kurzynske NG: Low postoperative infection rates with instrumented lumbar fusion. South Med J 91: , Jiménez-Mejías ME, de Dios Colmenero J, Sánchez-Lora FJ, Palomino-Nicás J, Reguera JM, García de la Heras J, et al: Postoperative spondylodiskitis: etiology, clinical findings, prognosis, and comparison with nonoperative pyogenic spondylodiskitis. Clin Infect Dis 29: , Kindmark CO: The concentration of C-reactive protein in sera from healthy individuals. Scand J Clin Lab Invest 29: , Kock-Jensen C, Brandslund I, Søgaard I: Lumbar disc surgery and variations in C-reactive protein, erythrocyte sedimentation rate and the complement split product C 3 d. Acta Neurochir (Wien) 90:42 44, Kornblum MB, Wesolowski DP, Fischgrund JS, Herkowitz HN: Computed tomography-guided biopsy of the spine. A review of 103 patients. Spine 23:81 85, Larsson S, Thelander U, Friberg S: C-reactive protein (CRP) levels after elective orthopedic surgery. Clin Orthop Relat Res 275: , Mangram AJ, Horan TC, Pearson ML, Silver LC, Jarvis WR: Guideline for prevention of surgical site infection, Centers for Disease Control and Prevention (CDC) hospital infection control practices advisory committee. Am J Infect Control 27:97 132, Massie JB, Heller JG, Abitbol JJ, McPherson D, Garfin SR: Postoperative posterior spinal wound infections. Clin Orthop Relat Res 284:99 108, Meyer B, Schaller K, Rohde V, Hassler W: The C-reactive protein for detection of early infections after lumbar microdiscectomy. Acta Neurochir (Wien) 136: , Mok JM, Pekmezci M, Piper SL, Boyd E, Berven SH, Burch S, et al: Use of C-reactive protein after spinal surgery: comparison with erythrocyte sedimentation rate as predictor of early postoperative infectious complications. Spine 33: , Palosuo T, Husman T, Koistinen J, Aho K: C-reactive protein in population samples. Acta Med Scand 220: , Pepys MB: C-reactive protein fifty years on. Lancet 1: , Rechtine GR, Bono PL, Cahill D, Bolesta MJ, Chrin AM: Postoperative wound infection after instrumentation of thoracic and lumbar fractures. J Orthop Trauma 15: , Richards BS: Delayed infections following posterior spinal instrumentation for the treatment of idiopathic scoliosis. J Bone Joint Surg Am 77: , Rodiek SO: [Diagnostic methods in spinal infections.] Radiologe 41: , 2001 (Ger) 25. Rothman SL: The diagnosis of infections ot the spine by modern imaging techniques. Orthop Clin North Am 27:15 31, Sasso RC, Garrido BJ: Postoperative spinal wound infections. J Am Acad Orthop Surg 16: , Sponseller PD, LaPorte DM, Hungerford MW, Eck K, Bridwell KH, Lenke LG: Deep wound infections after neuromuscular scoliosis surgery: a multicenter study of risk factors and treatment outcomes. Spine 25: , Takahashi J, Ebara S, Kamimura M, Kinoshita T, Itoh H, Yuzawa Y, et al: Early-phase enhanced inflammatory reaction after spinal instrumentation surgery. Spine 26: , Thelander U, Larsson S: Quantitation of C-reactive protein levels and erythrocyte sedimentation rate after spinal surgery. Spine 17: , Varma R, Lander P, Assaf A: Imaging of pyogenic infectious spondylodiskitis. Radiol Clin North Am 39: , Viola RW, King HA, Adler SM, Wilson CB: Delayed infection after elective spinal instrumentation and fusion. A retrospective analysis of eight cases. Spine 22: , Watters WC III, Baisden J, Bono CM, Heggeness MH, Resnick DK, Shaffer WO, et al: Antibiotic prophylaxis in spine surgery: an evidence-based clinical guideline for the use of prophylactic antibiotics in spine surgery. Spine J 9: , 2009 Manuscript submitted May 9, Accepted March 24, Address correspondence to: Byung-Uk Kang, M.D., 50-3, Dongin- Dong, Jung-Gu, Wooridul Spine Hospital, Daegu, Korea grussin@dreamwiz.com. 164 J Neurosurg: Spine / Volume 13 / August 2010