| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 1996;62:1783-1789
© 1996 The Society of Thoracic Surgeons
Department of Cardiovascular Surgery, Watson Clinic, Lakeland, Florida
Accepted for publication June 25, 1996.
 |
Abstract |
|---|
 TopFootnotes Abstract IntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Methods. Data from careful surveillance for infection have been collected prospectively during a 4.5-year period on 1,717 patients who underwent cardiac operations performed under direct ultraviolet C radiation.
Results. The rate for mediastinitis was 0.23%, and for deep incisional infection without mediastinitis, 0.12%; these rates are significantly lower than those for eight of nine of the most recently published cardiac series. When our infection rates were stratified using the National Nosocomial Infection Surveillance risk index, they were also significantly lower in the most important risk categories than the corresponding stratified rates collected from the participating hospitals of the Centers for Disease Control and Prevention National Nosocomial Infection Surveillance system.
Conclusions. Though we lack the proof that only a large, randomized study might provide, certainly, one possible explanation for our lower wound infection rate was the use of bactericidal ultraviolet C radiation during operation. This is a simple and effective means of minimizing operating room airborne bacteria as one possible source of these infections.
|
Introduction |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Two reports [2, 3] of small series of cardiac operations suggested a preventive effect of "high-efficiency particulate air" filtration during operation. A series of 878 cardiac operations in which a special attempt was made to eliminate operating room airborne bacteria as a possible source of operative wound infection was reported by one of us (I.W.B.) in 1968 with a 0.00% incidence of mediastinitis [4].
The purpose of this study was to determine the operative wound infection rates occurring in a series of 1,717 cardiac operations performed under direct ultraviolet C radiation. We compared these rates with other recently published cardiac operative series, particularly those collected nationwide by the Centers for Disease Control and Prevention (CDC) from the hospitals participating in the CDC National Nosocomial Infection Surveillance (NNIS) system.
|
Material and Methods |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
The hospital has maintained continuous detailed surveillance of nosocomial infections, carefully following the recommendations and guidelines of the CDC and, more recently, the NNIS of the CDC. The hospital Infection Control Office is staffed by a chief nurse trained in epidemiology, two assistant nurse epidemiologists, and a secretary. The office is under the direction of the Chairman of the Infections Committee with an MD infectious disease specialist as consultant.
Infection Surveillance and Data Collection
Each nurse epidemiologist makes daily rounds on assigned wards to check for any suspected or confirmed infections. The hospital bacteriology laboratory automatically forwards a copy of each positive culture report to the Infection Control Office as a further alert for follow-up by a nurse epidemiologist. If an infection is suspected, an "Infection Surveillance Record" is begun on that patient. Daily chart and patient follow-ups are started. Daily notes are made on the surveillance record summarizing any changes in the patient's temperature or condition and pertinent laboratory reports, such as white blood cell count and cultures. If an operative wound is involved, the appearance of the wound and character and amount of any drainage are noted. Antibiotics or other methods of treatment are recorded. Although the nurse epidemiologist confers with the patient's surgeon, particularly regarding the plan of treatment, it is the nurse's wound classification and estimate of severity, not the surgeon's, that is accepted. The patient's course is monitored throughout the remainder of the hospital stay, and arrangements are made for follow-up reports from his or her doctor's office after discharge.
We have used these prospective data from 1,774 cardiac operations performed under direct ultraviolet radiation by three Watson Clinic cardiac surgeons between January 1, 1988, and July 1, 1992. Unfortunately, because of limited storage space, the daily records of the nurse epidemiologists before January 1988 had been discarded. Wishing to base this study only on prospective infection data obtained by independent observers-the nurse epidemiologists-we therefore chose January 1, 1988, as the study starting date.
Fifty-seven patients who died less than 5 days postoperatively and who showed no sign of wound infection were eliminated from the study. All other patients who survived at least 5 days postoperatively were included. Thus, this study concerns 1,717 sternum-splitting cardiac operations performed on 1,623 patients, of whom 1,132 (69.83%) were male and 491 (30.17%) were female. The median age was 68 ± 9.95 years (median ± standard deviation). One hundred eleven patients (6.84%) were 80 years of age or older. The various cardiac operative procedures represented in the study are listed in Table 1
.
|
Data from all these sources were combined in a computer database using the Summit Medical Systems version 5 software (Minneapolis, MN).
Preoperative and Postoperative Procedures
Except for emergencies and those totally confined to bed, all patients had showers the evening before operation, usually with antiseptic soap. Shaving was done on the morning of operation on the patient's ward. Patients were given a loading dose of prophylactic antibiotic just before operation, with 1,708 receiving cefazolin and 1 patient receiving cephapirin. Of those sensitive to these two antibiotics, 3 received vancomycin and 5 received tobramycin.
Before moving into the operating room, the patient's eyes were covered to protect them from the ultraviolet light. Skin sterilization was either by Betadine Scrub (Purdue Frederick, Norwalk, CT) followed by painting with Betadine solution, or cleansing with 70% alcohol followed by painting with 70% alcoholic tincture of Zephiran (Winthrop Laboratories, New York, NY). After sterile drapes were applied, disposable adhesive Vidrapes or Ioban (3M Corp, St. Paul, MN) were used to cover all operative fields.
Instruments used in harvesting saphenous veins were kept separate and were removed for resterilization when this portion of the procedure was completed. Gloves also were changed. Before closing the chest wound, we occasionally irrigated the mediastinum with saline, but irrigation with an antibiotic solution was rarely used.
Unless the patient was febrile or showed signs of infection, prophylactic antibiotics were usually discontinued 48 hours postoperatively. All catheters, drains, and vascular access lines were removed as soon as possible. Unless significant drainage continued, mediastinal tubes were removed within 24 hours after operation.
Ultraviolet Bactericidal Radiation
The ultraviolet C light (253.7 nm) from ceiling-mounted ultraviolet lamps was maintained at an intensity of 20 to 25 mW•cm-2•s-1, measured at a level 30 inches above the operating room floor. The intensity was checked monthly with a calibrated ultraviolet photometer and was regulated by a tamper-proof rheostat. The ultraviolet lights remained on continuously from the night before operation until the patient was ready to leave the room after operation.
Definitions and Classification of Operative Wound Infections
We have adopted the recommendations in the consensus paper [5] of the Society for Hospital Epidemiology of America, the Association for Practitioners in Infection Control, Inc, the CDC, and the Surgical Infections Society on the surveillance of operative wound infections. They recommend the following:
).
|
Statistical Methods
Pearson's 
2 test was used to compare our infection rates with those of previously published cardiac operative series. In comparing our infection rates with those collected by the NNIS system, the rates were first stratified according to the NNIS risk index (see legend to Table 6
). Statistical significance (p < 0.05) was determined by the z test, as recommended by the CDC NNIS. Risk-factor variables were first assessed individually by univariate logistic regression analysis. Those variables with p values less than 0.1 were further assessed by multivariate logistic regression at a significance level of p less than 0.05. Statistical analyses were carried out using the SPSS 6.1 Base System and Advanced Statistics software (SPSS Inc, Chicago, IL).
|
Results |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
. There were four (0.23%) organ/space SSIs, all mediastinitis, that occurred in patients undergoing coronary artery bypass grafting (CABG). Two patients (0.12%) having CABG experienced deep incisional SSIs of their chest incisions without mediastinitis. In addition, there were 19 (1.11%) superficial incisional SSIs among the CABG patients. Five (0.29%) were of the chest incisions and 14 (0.82%) occurred at saphenous vein harvest sites in the lower legs.
|
Table 3 lists 15 risk factors for cardiac operative wound infection and their presence or absence in relation to mediastinitis plus deep chest incisional SSI and superficial incisional SSI. Among these, it is of interest that no infection developed in the 61 patients who required reoperation for continued postoperative mediastinal bleeding or in the 11 who underwent reoperation for early noninfected sternal dehiscence. Reoperation for bleeding has been found to be a strong predictor of sternal wound infection in some reported series.
|
). Univariate analysis of risk factors associated with superficial incisional SSI retained diabetes, duration of operation, and chronic obstructive pulmonary disease as possible predictive factors. After multivariate analysis, only diabetes and duration of operation were significant predictors of superficial incisional SSI (see Table 4).
|
). Our rate of 0.23% is significantly lower by Pearson's 2 test than the rates of mediastinitis in these nine collected series, with the single exception of the low rate of Nishida and associates [7], which they attributed to discriminate use of electrocautery. However, comparison of crude infection rates among institutions has limited applicability. For more meaningful and valid comparison of SSI rates, they must first be stratified as to risk variables [15].
|
), with corresponding stratified rate data collected by the CDC NNIS from their nationwide participating hospitals [16]. Our overall SSI rates for both CABG and non-CABG cardiac operations were significantly lower by z test (p = 0.0002 and p = 0.0096, respectively) than those of the NNIS for risk-index categories 1 (CABG) and 0 to 1 (non-CABG). Risk-index category 1 and combined risk-index categories 0 to 1 are the most important for comparison of infection rates, as 96% of our CABG operations and 79% of the collected NNIS CABG operations are in risk-index category 1, and 95% of our non-CABG cardiac operations and 81% of the NNIS non-CABG cardiac operations are in the combined risk-index categories 0 to 1. Valid rate comparisons cannot be made within categories 2 to 3 because of the small denominator in this combined category for Watson Clinic patients. |
Comment |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
From the time of Lister, airborne bacteria in the operating room have been suspected as one of the sources of operative wound infections. Until recently, however, documentation of infection by this route has been lacking, perhaps accounting for the reluctance of many surgeons to give airborne infection more than passing consideration. In 1963, Walter and associates [17] at Peter Bent Brigham Hospital in Boston documented operative wound infections due to airborne Staphylococcus aureus. More recently, outbreaks of wound infections proved to be due to airborne type A streptococci have been reported [18].
These wound infections from airborne bacteria occurred despite modern air-conditioned operating rooms with air filters, positive room air pressure, and multiple room air changes per hour. As shown first by Hart [19], the air entering the operating room is virtually free of pathogenic bacteria but becomes contaminated by the personnel within it. The amount of air contamination will vary depending on the number of operating room personnel, the presence of disseminating carriers, and the duration of the operation.
Airborne bacteria have a special importance in cardiac operations because environmental air mixes with blood being aspirated through suction and returned to the patient. Blakemore and co-workers [20] took cultures during cardiac operations in the vicinity of the operative chest wounds and from the extracorporeal circuit before and after perfusion. They reported that airborne bacteria aspirated through the coronary suckers contaminated the blood of 75% of the extracorporeal circuits cultured.
In 1969, Charnley and Eftekhar [21] gave new emphasis to the importance of airborne bacteria in the operating room when their introduction of hip replacement was accompanied by a highly discouraging deep wound infection rate of 8.9%. A thorough search as to the source of these infections convinced Charnley and Eftekhar [21] that a substantial number were caused by bacteria sedimenting from the operating room air. To overcome this, Charnley and Eftekhar [21] used laminar airflow enclosure systems along with impervious gowns and closed helmets for the operative team to exclude expired air completely. This method plus high-efficiency particulate air filtration to reduce operating room airborne bacteria have been adopted by orthopedic surgeons throughout the world doing joint replacement. Unfortunately, high-efficiency particulate air filtration with laminar airflow systems is expensive to install and maintain.
Ultraviolet Bactericidal Irradiation
Direct ultraviolet radiation, as used continuously since 1936 by the orthopedic and neurosurgical services at Duke University Hospital, is another means of rapidly and markedly reducing airborne bacteria [4]. This method is simple and inexpensive, requires little maintenance, and, in fact, has been shown to be more effective in reducing airborne bacteria than laminar airflow high-efficiency particulate air filtration systems [22]. Because of the current warnings regarding the dangers of ultraviolet rays in sunlight and the poor general understanding of the biologic effects of various ultraviolet wavelengths, some facts regarding ultraviolet C radiation are briefly reviewed.
Ultraviolet C radiation used in the operating room has the shortest wavelength of ultraviolet light (100 to 260 nm) and, unlike the harmful ultraviolet A and B radiation of longer wavelengths, has the least penetration. In fact, very little, if any, ultraviolet C from sunlight reaches the earth because it is absorbed in the upper atmosphere. However, it can penetrate to and denude the outermost cell layer of the cornea, resulting in a temporary but uncomfortable superficial keratitis unless an eyeshade and glasses are worn. Ultraviolet C has no permanent effect on the eye [22]. Because it cannot penetrate to the lens of the eye, it cannot cause cataracts.
Ultraviolet C can also penetrate the epidermis of the skin which, with prolonged exposure, can cause a mild erythema if the skin is uncovered or unprotected by sunscreens. There is no evidence that this poorly penetrating wavelength of ultraviolet light can cause skin malignancies. Experimentally, this wavelength (253.7 nm) failed to produce skin malignancies in hairless mice [22]. At Duke University Hospital, where ultraviolet C has been used in the operating rooms for 60 years, there have been no ultraviolet light-induced malignancies in personnel-surgeons, nurses, or assistants [23].
The ultraviolet intensity of 20 to 25 mW•cm-2•s-1 measured at 30 inches above the operating room floor is sufficient to kill almost all bacteria in the air in seconds [22]. With constant exposure at this intensity, it will also sterilize exposed metal countertops and dry smooth surfaces. Because of its poor penetration and absorption by the thinnest layer of fluid, ultraviolet C will not harm the tissues exposed in an operative wound [22]. It does require a certain amount of eye and skin protection, but this is minimal and not uncomfortable. For details concerning ultraviolet radiation in the operating room, the reader is referred to the excellent monograph by Berg-Perier [22].
Conclusion
A serious SSI such as mediastinitis, though infrequent, can be a devastating, costly, and life-shortening complication of cardiac operations. In 1,717 cardiac operations performed during a 4.5-year period under bactericidal ultraviolet radiation, the rate of mediastinitis was significantly lower than the rates in eight of nine series of cardiac operations published in the last 5 years.
More important and meaningful, when our cardiac operations were stratified for risk variables by the CDC NNIS risk index, the overall infection rates were significantly lower in the most important risk categories than the corresponding stratified rates collected nationally by the CDC NNIS system. As noted by others [1, 7, 10], duration of operation was found to be the strongest predictor of serious chest SSI. This is not in disagreement with the hypothesis that some SSIs result from airborne bacteria.
Ultraviolet radiation has been used in the cardiac operating rooms at Lakeland Regional Medical Center from the beginning of our heart surgery program in 1972. Consequently, for comparison, we have no historic data of our own without ultraviolet radiation that might support our hypothesis that ultraviolet C radiation in the operating room helps reduce SSI.
Though we lack the proof that only a very large randomized study might provide, certainly, one possible explanation for our lower wound infection rate was the use of bactericidal ultraviolet radiation during operations. This is a simple and effective means of minimizing operating room airborne bacteria as a possible source of these infections. We recommend its further evaluation by other cardiac surgery centers.
|
Acknowledgments |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
|
Footnotes |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. Dial, D. Nguyen, and D. Menzies Autotransfusion of Shed Mediastinal Blood: A Risk Factor for Mediastinitis After Cardiac Surgery? Results of a Cluster Investigation Chest, November 1, 2003; 124(5): 1847 - 1851. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Jonkers, T. Elenbaas, P. Terporten, F. Nieman, and E. Stobberingh Prevalence of 90-days postoperative wound infections after cardiac surgery Eur. J. Cardiothorac. Surg., January 1, 2003; 23(1): 97 - 102. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
