Annals of Surgical Oncology Cite Track
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
10.1245/ASO.2003.07.023
Annals of Surgical Oncology 10:778-782 (2003)
© 2003 Society of Surgical Oncology
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Morris, C. D.
Right arrow Articles by Healey, J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Morris, C. D.
Right arrow Articles by Healey, J.
Related Collections
Right arrow Surgery

ORIGINAL ARTICLES

Prospective Identification of Risk Factors for Wound Infection After Lower Extremity Oncologic Surgery

Carol D. Morris, MD, MS, Kent Sepkowitz, MD, Claudette Fonshell, RN, Neil Margetson, MA, Janet Eagan, RN, MPH, Jeremy Miransky, PhD, Patrick J. Boland, MD and John Healey, MD

From the Division of Orthopaedic Surgery (CDM, PJB, JH), Infectious Disease Service (KS, CF, JE), and Department of Quality Assessment (NM, JM), Memorial Sloan-Kettering Cancer Center, New York, New York. Affiliated with Weill College of Medicine, Cornell University, New York, New York.

Correspondence: Address correspondence and reprint requests to: Carol D. Morris, MD, MS, Memorial Sloan-Kettering Cancer Center, 1275 York Ave., New York, NY 10021; Fax: 212-794-4075; E-mail: morrisc{at}mskcc.org


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Background: Surgical site infections (SSI) are frequent causes of morbidity and mortality after orthopaedic oncologic procedures. This study was conducted to identify the surgical site infection rate following a lower extremity or pelvic procedure and assess the risk factors for acquiring SSI by direct observation of orthopaedic oncology patients’ wounds at a comprehensive cancer center.

Methods: One hundred ten consecutive patients were prospectively studied. The surveillance of surgical site infections was carried out by a surgeon-trained nurse from the Infectious Disease Service. Nineteen variables were analyzed as risk factors.

Results: The overall SSI rate was 13.6% (15 of 110). Excluding those patients with known preoperative infections, the SSI rate was 9.5% (10 of 105). Two statistically significant risk factors for surgical site infection in these patients emerged in the multivariate analysis: blood transfusion (P = .007) and obesity (P = .016). Procedure category was significant in univariate analysis only. Preoperative length of stay, length of procedure, prior adjuvant treatment (chemotherapy or radiotherapy), prior surgery, and use of an implant or allograft were not statistically significant risk factors for wound infection. Antibiotic usage patterns did not influence SSI rate.

Conclusions: Blood transfusion and obesity should be considered individual risk factors for the development of wound infection in patients having orthopaedic oncologic procedures.

Key Words: Surgical site infection • Orthopaedic oncology • Wound infection • Complications


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Surgical site infections (SSI) frequently cause morbidity and mortality in patients having orthopaedic oncologic procedures. According to the National Nosocomial Infectious Surveillance (NNIS) system, more than 500,000 postoperative wound infections occur nationally each year, accounting for 14% to 16% of all nosocomial infections among hospital inpatients.1,2 For surgical patients, SSI are the most common nosocomial infection and they have been shown to be the leading cause of operation-related adverse events.3,4 Furthermore, several studies have demonstrated an increased length of hospitalization and the associated financial implications for patients with SSI compared with noninfected patients having similar surgical procedures.3,5 For example, management of an infected total joint replacement costs an additional $40 million to $80 million per year in the United States, seven times more than the primary operation.6

Surgery represents the cornerstone of therapy in treating patients with malignant and some benign tumors of the bone and soft tissues of the extremities. Surgical infection causes delays in the resumption of adjuvant therapies, which compromises cancer survival, contributes to prosthesis failure, prolongs hospitalization, and increases treatment costs.7,8 Risk factors for developing surgical site infections have been well described for the general surgery population.9 Patients having orthopaedic oncologic procedures represent a unique population compared with that for general surgery, general orthopaedic surgery, and even general surgical oncology. The factors that contribute to infection for orthopaedic oncology surgery have yet to be identified. Although it may seem logical to extrapolate from general surgery and general orthopaedic procedures, no data support these assumptions. This study was conducted to identify the rate of SSI on the Orthopaedic Oncology Service for patients having a lower extremity or pelvic operation at the authors’ institution and to assess the risk factors associated with developing such an infection for this group of patients.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
A single epidemiology nurse (observer) from the Infectious Disease Service was trained by the orthopaedic surgery staff to evaluate surgical wounds. For 3 months before commencement of the study, the observer’s evaluations were validated against the observations of the attending surgeons. From January 1, 1999 to May 1, 1999, 110 consecutive patients who had major lower extremity, pelvic procedures, or both performed by the orthopaedic oncology staff were prospectively followed. A standardized data collection form was used. At the time of surgery, each procedure was classified by the attending surgeon according to the system of the American College of Surgeons (ACS) (Table 1) and entered into an operating room database by the nurse. The SSI database was programmed and customized by the institutions’ Quality Assessment Department. Analysis was performed using SPSS software (version 7.5, SPSS Inc., Chicago, IL).


View this table:
[in this window]
[in a new window]
  		TABLE 1. Surgical procedure categories according to the system of the American College of Surgeons
 
Following the surgical procedure, surgical sites were examined by the epidemiology nurse on postoperative day 3 and every 3 days thereafter. The observation schedule increased to more frequent intervals at the observer’s discretion or when surgical sites had any sign of infection. Observations were conducted while the patient was in the hospital for a 30-day period. If patients were discharged before conclusion of the observation period or had a hospital stay of <3 days, they were interviewed by telephone and observed in the outpatient clinic. When the telephone interview solicited responses indicative of wound problems or antibiotic usage, a follow-up evaluation was arranged.

Wounds were graded on the following scale: grade 1 = normal healing; grade 2 = suture line erythema <1 cm; grade 3 = suture line erythema >=1 cm; grade 4 = frank, purulent drainage. Gram stains and cultures were obtained on all wounds determined to be infected or as otherwise clinically indicated. Grade 3 or 4 wounds were considered infected. Wounds from which a positive culture was obtained in the setting of physical signs of infection (i.e., fever, inflammation) were also considered infected. Localized suture abscesses without cellulitis or other signs of infection were not considered an SSI.

The incidence of wound infection was compared for (1) the use of blood transfusion products, (2) obesity, (3) preoperative length of stay, (4) length of procedure, (5) gender, (6) prior radiotherapy, (7) prior brachytherapy, (8) prior chemotherapy, (9) prior steroid use, (10) diabetes, (11) prior antibiotics (nonprophylactic), (12) metastatic status, (13) disease recurrence, (14) prior surgery, (15) use of an implant (metal prosthesis or allograft), (16) perioperative antibiotic prophylaxis, (17) Anesthesiology Society of America (ASA) value, (18) ACS procedure category, and (19) age. Obesity was defined as the height-weight distribution that exceeded the largest frame size in the 1985 Metropolitan Life Insurance tables by >=20%. Those risk factors that were univariately significant using {chi}2 analysis at P <= .05 were entered into a stepwise logistic regression equation using SPSS software to evaluate the risk of each factor when adjusted for other factors.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
The overall SSI rate for the 110 consecutive patients identified by the observer was 13.6% (15 of 110). By traditional wound classification, most cases were class I (clean), comprising 93% of all cases performed over the 4-month period. Three wound were class II, none were class III, and five wounds were class IV (infected). When excluding the five class IV cases, the de novo SSI rate was 9.5% (10 of 105).

Seventy-three percent of patients had procedures done for malignant conditions. Thirty patients required blood transfusion intraoperatively, with nine patients receiving more than 6 U of packed red blood cells. External beam radiation therapy was administered to 18 patients before surgery, whereas 2 patients received postoperative brachytherapy. Forty patients received preoperative chemotherapy. These findings are summarized in Table 2. Postoperative external beam radiation was delayed until healing stabilized so as to not contribute to SSI.


View this table:
[in this window]
[in a new window]
  		TABLE 2. Summary of statistical analysis of variables associated with surgical site infection in 110 consecutive patients
 
Of the 15 surgical site infections, 3 (20%) were confined to the space superficial to the deep fascia and 12 (80%) involved tissue below the deep fascia. All patients who were infected preoperatively (n = 5) developed postoperative surgical site infections. Again, among patients not infected preoperatively, 10 of 105 (9.5%) developed a new SSI. Twelve (12 of 15) patients with SSI had a malignant diagnosis.

Six of 15 infected patients required surgery to resolve the infection. Nine patients were successfully treated with antibiotics alone. For two of the patients with class IV wounds, the return to the operating room was planned. The remaining four patients each had from one to three debridement procedures for wound management. Amputation did not have to be performed on any patient. One patient required removal of a structural pelvic allograft.

Three risk factors had univariate statistical significance: obesity (P = .005), procedure category (P <= .001), and blood transfusion (P = .007). Blood transfusion (P = .009) and obesity (P = .001) remained independent predictors of surgical site infection in multivariate analysis. When the amount of blood transfused was analyzed, 1 to 2 units emerged as a statistically significant risk factor (P = .0008), whereas >2 units did not, although transfusion of >6 units approached significance (P = .086).

Microbiology cultures were obtained in 13 of 15 infected patients. Five of 13 cultures failed to grow any organisms. From the remaining 8 of 13 cultures, the following organisms were isolated: Pseudomonas aeruginosa(2), coagulase-negative Staphylococcus (2), Streptococcus pneumoniae (1), Citrobacter (1), Haemophilus influenza (1), and multiple organisms (1). No fungal isolates were found.

In this analysis, several variables failed to be statistically significant risk factors for SSI. These included the length of the procedure, prior adjuvant therapy (chemotherapy or radiation), perioperative steroid use, metastatic status, history of diabetes, prior surgery, disease recurrence, antibiotic usage patterns, and ASA status.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
This study was conducted to identify the actual surgical site infection rate at our hospital and to determine the risk factors associated with infection, specifically for patients having orthopaedic oncologic procedures. Such risk factors have most often been described for general surgical patients, although they have also been reported for cancer patients as well as patients having general orthopaedic procedures. Patients requiring surgery for the long bones and pelvis for oncologic reasons represent a unique population compared with those of general surgery, general orthopaedics, and nonextremity surgical oncology.

The NNIS system has reported three risk indices are helpful in predicting SSI risk in the general surgery population: (1) ASA score >=3, (2) operative time >4 hours, and (3) procedure class III or IV.9 These risk factors, which have been validated in hospitals outside of the NNIS study,10 represent excellent predictors of SSI that are easily assessable and which encompass other risk factors such as chronic disease or obesity. In our patient population, only procedure category overlapped with the NNIS findings. We did not have any class III procedures and only 4.5% of all procedures were class IV. All of the class IV procedures developed an SSI. Although it may seem intuitive to predict that infected procedures (class IV) will have postoperative wound infections, these cases were included for completeness in this consecutive patient analysis as the goal of surgery was to clear the infection. It is important to distinguish the differences between these two groups as the associated complications and morbidity of soft tissue wound infections, compared with skeletal infections, are striking. Osseous infections, which are particularly challenging to treat, can cause significant functional impairment and require a prolonged systemic therapy course.

Infection rates and risk factors for general orthopaedic patients have also been described.4,11,12 Joint arthroplasty is believed to have an overall infection rate of 1% over the lifetime of the prosthesis, whereas arthroscopy has far less than 1%.13,14 One population-based study reports on more than 10,000 Dutch patients who had orthopaedic operations for a variety of conditions.3 Four statistically significant risk factors emerged as predictors for SSI: (1) age, (2) additional nosocomial infection, (3) wound contamination class, and (4) number of operations. Again, only wound class (procedure category) was found to be a risk factor in common with our patient population. Orthopaedic oncologic procedures, in general, require substantial exposures and dissection across vascular distributions compared with general orthopaedic procedures. To control hemostasis and safely excise tumors, the affected area is often subject to devascularization, which alone can lead to compromised wound healing. Furthermore, orthopaedic oncology patients frequently receive adjuvant immunomodulating therapeutic modalities, which can complicate their perioperative course. Interestingly, in this investigation, prior chemotherapy, prior radiotherapy, and steroid use were not statistically significant independent predictors of SSI.

Velasco et al. reported on more than 1200 cancer patients over a 24-month period who had surgery for malignant disease.15 The overall SSI rate was 17%. Six statistically significant risk indices for the prediction of SSI following oncology surgery were (1) wound class, (2) operative time >280 minutes, (3) male sex, (4) prior radiotherapy, (5) antibiotic prophylaxis not according to protocol, and (6) ASA >=3. A similar investigation of cancer patients was done at our institution that included the Gastric-mixed Tumor, Colorectal, and Breast surgical services.16 Over a 15-month period, 1226 patients had 1283 surgical procedures, with an overall infection rate of 8%. Statistical analysis revealed five predictors of SSI: (1) obesity, (2) procedure category, (3) operation time >4 hours, (4) ASA >=3, and (5) preoperative length of stay >=3 days. Procedure category continues to emerge as a common independent risk factor in all subsets of surgical patients. Orthopaedic oncology patients are probably closer in their SSI risk status to surgical oncology patients than general orthopaedic surgical patients. Three fourths of all our patients in this study and 12 of 15 patients with an SSI carried a malignant diagnosis.

Blood product usage was an independent predictor of SSI in this study. The clinical significance of immune suppression secondary to blood replacement in cancer patients is controversial. The adverse effects of blood transfusion have been well documented in patients having colorectal and gastrointestinal surgery as demonstrated by higher infection rates in these patients.17,18 A meta-analysis by McAlister et al.19 that examined perioperative allogenic blood transfusion in surgical oncology patients, however, concluded there was no evidence to support adverse infectious sequelae. Tan and Mankin20 reported that blood transfusion >1000 mL had an adverse effect on infection in osteoarticular allografts. In our study, although the number of transfused units did not provide statistical significance, a trend toward increased SSI rates was observed in patients who received >5 units. Certainly, longer procedures are associated with greater blood loss and perhaps an associated intraoperative "washout" of prophylactic antibiotics. Unlike blood products, which are administered on an as needed basis, antibiotics are usually dispensed according to a schedule at our institution (e.g., cefazolin every 3.5 hours). Hence, as blood is lost and then replaced, the concentration of circulating antibiotics can potentially diminish disproportionately. Using this same argument, however, one might expect a dose-response effect related to the number of units transfused and infection. Scheduled antibiotics likely diminish this effect. The lack of a proportionate relationship between number of blood products transfused and SSI rate questions the clinical significance of this observation.

Regarding antibiotic usage patterns, 94% (103 of 110) of patients received perioperative antibiotics. Perioperative antibiotics were defined as prophylactic antibiotics administered immediately before making the incision or administered intraoperatively after a culture had been obtained. Whereas the vast number of elective procedures performed in orthopaedic surgery are class I type wounds, or clean wounds, the administration of prophylactic antibiotics is common in orthopaedic practice. Cefazolin was our antimicrobial agent of choice for clean wounds, except for patients with known allergic contraindications to cephalosporins. The main indications for prophylactic antibiotics are procedures with a high inherent infection rate and procedures for which the prevalence of infection is low but such an infection would have catastrophic results. Validation of the latter has been extensively documented with total joint arthroplasty. Infected tumors certainly fall into the category of possible catastrophic complications and, as such, antibiotics are administered for most procedures that expose or manipulate bone. It is the routine practice at our institution to withhold antibiotics before obtaining microbiology cultures and diagnostic tissue in lesions of uncertain cause. One of the patients who did not receive perioperative antibiotics was on antibiotics before surgery for a prior wound dehiscence. The other six patients without perioperative antibiotic prophylaxis had incisional or excisional biopsies performed. Three of these six patients received antibiotics in the postanesthesia care unit. The timing of the administration of antibiotics before or during surgery did not have an effect on the SSI rate.

In summary, this prospective study showed that obese patients who have complex orthopaedic oncologic procedures that necessitate transfusion are at high risk for developing an SSI. These patients should be the focus of intervention studies to reduce the rate of wound infections. In addition, previously infected wounds are at particularly high risk for recurrent infection.


   FOOTNOTES
 
This study was conducted to identify the surgical site infection (SSI) rate for orthopaedic oncologic patients and assess the risk factors for acquiring SSI by direct observation at a comprehensive cancer center. Nineteen variables were analyzed as risk factors in 110 consecutive patients having lower extremity, pelvic procedures, or both.

Received for publication July 23, 2002. Accepted for publication April 22, 2003.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Haley RW, Culver DH, White JW, et al. The nationwide nosocomial infection rate. A new need for vital statistics. Am J Epidemiol 1985; 121: 159–67.[Abstract/Free Full Text]
  2. Smyth ET, Emmerson AM. Surgical site infection surveillance. J Hosp Infect 2000; 45: 173–84.[CrossRef][Medline]
  3. De Boer As, Mintjes-de Groot AJ, Severijnen AJ, et al. Risk assessment for surgical-site infections in orthopedic patients. Infect Control Hosp Epidemiol 1999; 6: 402–7.
  4. Mangram AJ, Horan TC, Pearson ML, et al. Guideline for the prevention of surgical site infection, 1999. Infect Control Hosp Epidemiol 20; 247–80: 1999.
  5. Wenzel RP. The economics of nosocomial infections. J Hosp Infect 1995; 31: 79–87.[CrossRef][Medline]
  6. Lew D, Pittet D, Waldvogel F. Infections that complicate the insertion of prosthetic devices. In: Mayhall CG, ed. Hospital Epidemiology and Infection Control. Baltimore: Williams & Wilkins, 1996; 731–48.
  7. Kawai A, Boland P, Lin P, et al. Relationship between magnitude of resection, complication, and prosthetic survival after prosthetic knee reconstructions for distal femoral tumors. J Surg Oncol 1999; 70: 109–15.[CrossRef][Medline]
  8. Meyers PA, Heller J, Healey JH, et al. Chemotherapy for nonmetastatic osteogenic sarcoma: the Memorial Sloan-Kettering experience. J Clin Oncol 1992; 10: 5–15.[Abstract]
  9. Horan TC, Culver DH, Gaynes RP, et al. Nosocomial infections in surgical patients in the United States, January 1986–June 1992. National Nosocomial Infections Surveillance (NNIS) System. Infect Control Hosp Epidemiol 1993; 2: 73–80.
  10. Freitas PF, Campos ML, Cipriano ZM. Suitability of the NNISS risk index to predict the incidence of surgical site infection. Rev Assoc Med Bras 2000; 46: 359–62.
  11. Bengston S. Prosthetic osteomyelitis with special reference to the knee: risks, treatment and costs. Ann Med 1993; 25: 523–9.[Medline]
  12. Miller WE, Counts GW. Orthopedic infections: a prospective study of 378 clean procedures. South Med J 1975; 4: 386–91.
  13. Complications of arthroscopy and arthroscopic surgery: results of a national survey. Committee on Complications of Arthroscopy Association of North America. Arthroscopy 1985; 1: 214–20.[Medline]
  14. Lew D, Waldvogel F. Infections in skeletal prosthesis. In: Bennett JV, Brachman PS, eds. Hospital Infections. 4th ed. Philadelphia: Lippincott-Raven, 1998; 613–20.
  15. Velasco E, Thuler LC, Martines CA, et al. Risk index for prediction of surgical site infection after oncology operations. Am J Infect Control 1998; 3: 217–23.
  16. Barber GR, Miransky J, Brown AE, et al. Direct observations of surgical wound infections at a comprehensive cancer center. Arch Surg 1995; 130: 1042–7.[Abstract]
  17. Braga M, Vignali A, Radaelli G, Gianotti L, Di Carlo V. Association between perioperative blood transfusion and postoperative infection in patients having elective operations for gastrointestinal cancer. Eur J Surg 1992; 158: 531–6.[Medline]
  18. Jahnson S, Andersson M. Adverse effects of perioperative blood transfusion in patients with colorectal cancer. Eur J Surg 1992; 158: 419–25.[Medline]
  19. McAlister FA, Clark HD, Wells PS, et al. Perioperative allogeneic blood transfusion does not cause adverse sequelae in patients with cancer: a meta-analysis of unconfounded studies. Br J Surg 1998; 85: 1162–3.
  20. Tan MH, Mankin HJ. Blood transfusion and bone allografts. Effect on infection and outcome. Clin Orthop 1997; 340: 207–14.



This article has been cited by other articles:


Home page
 ChestHome page
A. E. Morris, R. D. Stapleton, G. D. Rubenfeld, L. D. Hudson, E. Caldwell, and K. P. Steinberg
The Association Between Body Mass Index and Clinical Outcomes in Acute Lung Injury
Chest, February 1, 2007; 131(2): 342 - 348.
[Abstract] [Full Text] [PDF]

Home page
 PediatricsHome page
E. D. Taylor, K. R. Theim, M. C. Mirch, S. Ghorbani, M. Tanofsky-Kraff, D. C. Adler-Wailes, S. Brady, J. C. Reynolds, K. A. Calis, and J. A. Yanovski
Orthopedic complications of overweight in children and adolescents.
Pediatrics, June 1, 2006; 117(6): 2167 - 2174.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Morris, C. D.
Right arrow Articles by Healey, J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Morris, C. D.
Right arrow Articles by Healey, J.
Related Collections
Right arrow Surgery

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS