Eur J Cardiothorac Surg 1999;15:45-50
© 1999 Elsevier Science NL
A prospective study of wound infection in coronary artery surgery
J. Bellchambers,
J.M. Harris,
P. Cullinan,
H. Gaya,
J.R. Pepper
Department of Surgery, Royal Brompton Hospital, Sydney Street, London SW3 6NP, UK
Received 27 July 1998;
received in revised form 5 October 1998;
accepted 6 October 1998.
Corresponding author. Tel./fax: +44-171-351-8530; e-mail: a.garrick@rbh.nthames.nhs.uk
 |
Abstract
|
|---|
Objective: The incidence of wound infection in coronary artery surgery is low, but there is an appreciable mortality associated with sternal infection. In other fields of surgery there is evidence that an all disposable paper gown and drape system has protective advantages over a fabric system. Using an established wound scoring system (ASEPSIS), we investigated this in a prospective, randomised trial of patients undergoing cardiac surgery. Methods: 505 patients undergoing isolated coronary artery surgery in a single hospital over an 18 month period were randomised to either a disposable, paper drape system or re-usable fabric drapes. Allocation was stratified according to whether or not the patient had had previous coronary artery surgery. Patients were followed for 3 months. Total infection scores for each wound were calculated from daily scores collected in hospital together with the scores at six weeks and three months. Results: The two randomised groups were otherwise well matched. Full follow-up information was available for 464 (92%) patients. There was no evidence of any difference in rates of sternal or leg wound infection between the two groups (P=0.87 and 0.62, respectively). Women were more likely to have infected sternal wounds (P=0.17) and significantly more likely to have infected leg wounds (P=0.04). Patients with sternal wound infections had a significantly higher body mass index (P=0.001). High Parsonnet scores and increased time on ventilation were significantly associated with leg wound but not sternal infections. For both wound sites, patients with infections had spent longer in the operating theatre. Conclusion: In a randomised controlled study of patients undergoing coronary artery surgery we found that the use of paper drapes and gowns conferred no benefit over a reusable fabric in terms of post-operative wound infection. Although females and diabetics are more likely to experience this complication, an important additional risk factor is an extended operating time.
Key Words: Coronary artery surgery • Wound infection • Prospective • Randomised trial • Drape system
|
Introduction
|
|---|
Coronary artery bypass surgery remains the dominant operation in adult cardiac surgery. Although the incidence of wound infection is generally low, the associated mortality rate from sternal infection is high and varies from 14 to 47% in recent reports
[1]
[2]
[3]
[4]
[5]. Even leg wound infection, which rarely represents a threat to life, adds importantly to morbidity; the average cost of maintaining a patient in hospital with a wound infection is three times that of patients with an uncomplicated postoperative course
[3]. Clinically, mediastinitis typically presents between 4 and 30 days after operation, most commonly with a wound discharge. The recent trend towards shorter hospital stays of 4–5 days leads to a situation where the clinical diagnosis of wound infection is often made outside the cardiac unit, by the primary care team or the community hospital. Thus the incidence of postoperative wound infection may be difficult to measure. This study, a randomised controlled trial of two draping systems, was designed also to produce a prospective audit of wound infections in a tertiary referral centre for cardiac surgery, and to evaluate the clinical use of an established wound scoring system (ASEPSIS), developed specifically for cardiac surgery
[6].
|
Patients and methods
|
|---|
Patients undergoing isolated coronary artery surgery in a single hospital between July 1995 and December 1996 were considered for entry into the study. Patients were allocated, using sealed envelopes containing a series of computer-generated random numbers, to surgery using either a disposable, paper drape system including an iodophor impregnated adhesive plastic drape which covered the central thorax and abdomen; or re-usable fabric drapes including an iodophor impregnated adhesive plastic drape covering the anterior thorax
[7]. In each group, the operating surgeon, assistants and scrub nurses wore gowns of the same material as the drapes. Allocation was stratified according to whether or not the patient had had previous coronary artery surgery. All patients received an identical prophylactic antibiotic regime consisting of cephazolin 1G 12 hourly for 48 h with the first dose given at the time of anaesthetic induction. None of these patients had a history of cephazolin sensitivity. A uniform method of wound closure was used: polydioxanone (PDS) monofilament absorbable suture was used for sternum, leg and arm. Single steel wire sutures were used to close the sternum in all patients and metal clips were used to secure branches of the internal mammary artery on the chest wall, branches of the saphenous vein in the leg, and branches of the radial artery in the arm. The study was approved by the Ethics Committee of Royal Brompton Hospital, each patient providing informed consent.
A research nurse (J.B.), who was unaware of the patients' drape system, inspected and assessed all wounds on 5 out of the first 7 days following surgery; information at 6 weeks and 3 months post-operatively was collected from patients either over the telephone or in the clinic, according to protocol. Details of antibiotics used were verified by the primary care team or local hospital. Each wound was scored using the ASEPSIS system (6) (Appendix A>). The important feature of this system is that the final score accrues by a simple and objective assessment of each element on the checklist, each element only being scored once during the 3 month period. Such a scoring system provides a more objective assessment. In general, a clinically significant wound infection had a score of 21–30, a deep wound infection, usually associated with a sternal click scored 31–40 and bone infection scored >41.
In this way, 505 (85%) of 593 eligible patients were randomised. Of the 593, consent was not obtained from 72 patients, and a further 16 were not randomised because of individual surgeons' preferences. Fifteen patients died during the follow-up period of the study.
|
Statistical analysis
|
|---|
Total infection scores for each site were calculated from the five daily scores collected in hospital together with the scores at 6 weeks and at 3 months. Full follow-up data were available for 464 (92%) patients. A further 22 patients were only missing one of the five scores collected during their post-operative hospital care. Substituting the missing score with the previous day's score made little difference to the subsequent results and increased the percentage of patients with complete scores to 96%.
Any infection which resulted in an ASEPSIS score of more than 20 was categorised as infected. Comparisons between the two randomisation groups were made using t-tests, Mann–Whitney tests, chi-squared tests and Fisher's exact test. Associations between an infected wound and patient or operative details were also investigated separately using these tests. A stepwise logistic regression model was used to identify which factors were independently related to the risk of a wound becoming infected. All analysis was done using SAS (Cary, NC) and Egret (SERC, Seattle, Washington) statistical software.
With 250 patients in each arm of the trial, this study had an 80% power (alpha=0.05) to detect a 2.4-fold increase in infection rate above the predicted baseline of 5%.
|
Results
|
|---|
The two randomised groups were well matched on demographic and co-morbidity features with the exception of diabetes mellitus which was more prevalent in the group randomised to receive paper drapes (Table 1). Operative and peri-operative details were also similar although patients with fabric drapes were ventilated for longer (P=0.05).
Table 2 shows the total ASEPSIS scores for sternal wounds, and the maximum score for any wound. Twenty five (5%) sternal wounds, 58 (13%) leg wounds and no arm wounds were classified as infected (score 20). There was no evidence of any difference in rates of infection between the two drape groups (P=0.87 and 0.78, respectively). Sternal infections were as likely to occur in hospital (52%) as after discharge (48%). On the other hand, 44 (76%) leg wound infections were detected only after discharge from hospital.
Univariate associations between possible risk factors and the presence of an infected wound are displayed in Table 3. Female patients were more likely to have infected sternal wounds (P=0.17) and significantly more likely to have infected leg wounds (P=0.04). There was a tendency for diabetic patients to develop more infections but the increases were not statistically significant. Patients with infected sternal wounds had on average a significantly higher body mass index (P=0.001). No relationship was observed between the thickness of the subcutaneous fat tissue and wound infection rate High Parsonnet scores
[8] and increased time on ventilation were significantly associated with leg wound but not sternal infections. In this study the use of one or two internal mammary arteries was not associated with an increased incidence of sternal wound infection (Table 3). For both wound sites, patients with infections had spent longer in the operating room but the differences were small. Re-opening of the sternum for excessive blood loss in the early postoperative stage was performed in 22 of 386 (5.6%) uninfected wounds and in eight of 78 (10%) wounds which subsequently became infected; this difference which was not statistically significant. All reopening procedures were performed under sterile conditions in the operating room. The draping method for these second procedures was identical to that determined by randomisation for the first operation.
Sternal wounds tended to be closed by a consultant (28%) or senior trainee surgeon (64%); peripheral wounds, on the other hand, were more frequently closed by a junior surgeon (53%) or surgical assistant (36%). A sternal wound was slightly more likely to become infected if it was closed by a senior trainee but the differences were small and not statistically significant (Table 3). Leg infections were more likely to become infected if the person responsible for closing the wound was a junior surgeon, a difference that also was not significant. Surgical assistants had significantly lower (P=0.042) rates of infections in leg wounds, 15/168 (9%), compared to junior surgical trainees, 39/247 (16%).
A stepwise logistic regression model was used to identify which factors were independently associated with a risk of wound infection (Table 4). Increased body mass index, being female and more time spent in the operating room were independently associated with an increased risk of sternal infection. A high Parsonnet score was the only factor which remained significantly and independently associated with risk of leg wound infection. A variable coding for drape system was subsequently added to the regression models in order to estimate its relationship with risk of infection after adjusting for factors known to be associated with infection. For neither sternal nor leg sites was there evidence of increased risk with the fabric drape system (odds ratios=0.86, 95% [CI: 0.37–2.01] and 1.24, [CI: 0.71–2.16], respectively).
Swabs for microbiology were obtained from 23 (5%) of 486 sternal wounds during hospital admission. The micro-organisms isolated in the presence of infection (scored >20) were, in order of frequency (n): Staphylococcus epidermidis (5), Staphylococcus aureus (2), miscellaneous organisms (2). Similarly, swabs were obtained from 27 (6%) of 486 sternal wounds after discharge from hospital. In these cases the micro-organisms isolated in the presence of infection were, in order of frequency (n): Staphylococcus aureus (9), Staphylococcus species (1), Pseudomonas aeruginosa (1), Pseudomonas species (2), miscellaneous organisms (3).
Swabs were also obtained from 13 (3%) of 478 leg wounds. Microorganisms isolated from infected leg wounds in hospital were: Pseudomonas aeruginosa (2) and Pseudomonas species (2), miscellaneous organisms (4). For leg wounds examined out of hospital (54 [11%] of 418 wounds) the frequencies were as follows: Staphylococcus aureus (23), Staphylococcus epidermidis (1), Pseudomonas aeruginosa (2), Pseudomonas species (1), miscellaneous organisms (32).
An assessment of the comparative cost of fabric versus disposable drapes showed that the disposables were 150% more expensive than fabric.
|
Discussion
|
|---|
Elective adult cardiac surgery is `clean' and the number of wound infections should be low. A review of the literature in the last 15 years reveals a range of postoperative sternal wound infection of 1.0–8.6%
[1]
[2]
[3]
[4]. The rate of sternal infection in our own unit based on a routine prospective surgical audit (unpublished) was 5.3 and 5.4% for the years 1993 and 1994, respectively. Unfortunately, the majority of reported studies, whether retrospective or prospective, fail to identify the end points with sufficient clarity to enable comparisons to be made.
Advances in manufacturing techniques have led to the construction of improved quality materials which are resilient
[9]. A 21 month study from Duke University incorporating 2181 general surgical procedures demonstrated the efficiency of a disposable gown and drape system in comparison to a cotton system in decreasing wound infection. Overall the disposable system showed a marked reduction in the number of wound infections (2.8 vs. 6.5%). A similar improvement was shown in `clean' cases (1.8 vs. 3.8%). The barrier effect of the disposable material was independent of all other factors in reducing the wound infection rate.
We did not demonstrate a reduction in wound infection in our study. Although our numbers are smaller than the general surgical study of Moylan
[9], our statistical calculation indicates that with the use of a closely monitored wound score system a significant difference is unlikely to emerge between the two groups with less than 5000 patients in a relatively homogeneous group of coronary artery bypass operations. As in other studies of cardiac operations we found that female gender, diabetes, high body mass and extended operating time were incremental risk factors
[10]
[11]
[12]
[13]
[14]. Unlike Fernandez et al.
[15] we did not find that age had any significant bearing on wound infection. Also in contrast to other reports
[2]
[16]
[17] we did not find that a preoperative hospital stay of two or more days was associated with postoperative wound infection which is reassuring since this is a common event occurring in 32% of our patients, many of whom are transferred directly from community hospitals to our unit. Furthermore, the requirement for re-sternotomy for early post-operative bleeding did not emerge as an incremental risk factor for wound infection. This stands in contrast to other studies where re-operation in this setting was associated with an increased risk of mediastinitis
[1]
[2]
[18]
[19]
[20]
[21]
[22].
Our mean bypass times were less than 90 min so it is not surprising that this factor had no impact on wound infection. Most studies show a higher incidence of sepsis with bypass times over 3 h
[3]. There is evidence that cardiopulmonary bypass affects phagocytic activity. The ability to clear bacteria is diminished, a problem that increases with prolonged bypass time
[23].
The use of one or two internal mammary arteries has been implicated as a risk associated with sternal infections
[24]. Sternal infections may occur with a greater frequency in patients using both internal mammary arteries than among unilateral IMA or vein graft patients (6.9 vs. 1.9 vs. 1.3%; P=0.001; n=1566 patients). A large review
[3] from the Cleveland Clinic assessed relative risk by regression analysis and showed the combination of diabetes and bilateral IMA to have a relative risk of 5.00. However, among conduit groups (SV; 1 IMA or 2 IMA's) there was no difference. Hazelrigg et al.
[25] noted among a series of 2582 patients with an infection rate of 0.81% a significantly higher rate among those having bilateral IMA grafting (1.65%; odds ratio 4.15).
The detection of wound infection after hospital discharge was high in our study (48% of sternal wounds and 76% of leg wounds) similar to the experience of Law et al.
[26] who found that 60% of postoperative wound infections in a general surgery practice were detected after discharge. This supports the need to extend specialist care into the community during the convalescent period, a feature of current `Hospital at Home' programmes. It is our unit policy to maintain contact between the patient and the cardiac surgical nursing staff by telephone for the first postoperative month.
Most wound infections arise from skin organisms. The organisms most commonly involved include Staphylococcus epidermidis, or Staphylococcus aureus
[1]
[26]. Our findings are in line with these authors and like them we found Gram negative bacilli less commonly. We did not find any fungal infections, which is to be expected as they usually appear as superinfections in association with a more complex patient profile.
Our study largely supports the view of Hammermeister et al.
[27] that, with the exception of body mass index and female gender, there are few preoperative factors that predict sternal wound infection; this implies that the most important factors are technical at the time of the primary wound closure.
|
Acknowledgments
|
|---|
We thank the clinical audit committee for the support of J.B., P. Webber for his help with the microbiology and P. Luscombe for help with the database. We are very grateful for the help of the nursing staff in the operating theatre and surgical wards.
|
A. Appendix: ASEPSIS
|
|---|
ASEPSIS is an acronym for Additional treatment, the presence of Serous discharge, Erythema, Purulent discharge and Separation of the deep tissues, the Isolation of bacteria and the duration of inpatient Stay. Points were allotted according to the ASEPSIS system for the extent of wound disturbance showing serous (0–5) or purulent (0–10) exudate, erythema (0–5), and separation of the deep tissues (0–10). Points were also added for criteria up to 3 months following surgery: antibiotic treatment, 10 points; drainage under local anaesthetic, 5 points; debridement under general anaesthetic, 10 points; positive microbiology isolate, 10 points; extended hospital stay beyond 10 days due to wound infection, 5 points. The total score was then used to reflect the severity of infection as shown:- 0–10=satisfactory healing
- 11–20=disturbance of healing
- 21–30=minor wound infection
- 31–40=moderate wound infection
- >41=severe wound infection
|
References
|
|---|
- Grossi E.A., Culliford A.T., Krieger K.H., Kloth D., Press R., Baumann F.G., Spencer F.C. A survey of 77 major infectious complications of median sternotomy: a review of 7949 consecutive operative procedures. Ann Thorac Surg 1985;40:214-223.[Abstract]
- Ottino G., De Paulis R., Pansini S., Rocca G., Tallone M.V., Comoglio C., Costa P., Orzan F., Morea M. Major sternal wound infection after open heart surgery: a multivariate analysis of risk factors in 2579 consecutive operative procedures. Ann Thorac Surg 1987;44:173-179.[Abstract]
- Loop F.D., Lytle B.W., Cosgrove D.M., Mahfood S., McHenry M.C., Goormastic M., Stewart R.W., Golding L.A.R., Taylor P.C. Sternal wound complications after isolated coronary artery bypass grafting: early and late mortality, morbidity and cost of care. Ann Thorac Surg 1990;49:179-187.[Abstract]
- Ivert T., Lindblom D., Sahni J., Eldh J. Management of deep sternal wound infection after cardiac surgery: Hanuman syndrome. Scand J Thorac Cardiovasc Surg 1991;25:111-117.[Medline]
- Serry C., Bleck P.C., Javid H., Hunter J.A., Goldin M.D., DeLaria G.A., Najafi H. Sternal wound complications: management and results. J Thorac Cardiovasc Surg 1980;80:861-867.[Abstract]
- Wilson A.P.R., Treasure T., Sturridge M.F., Gruneberg R.N. A scoring method (ASEPSIS) for post-operative wound infections for use in clinical trials of antibiotic prophylaxis. Lancet 1986;1:311-313.[Medline]
- Johnston D.H., Fairclough J.A., Brown E.M., Morris R. Rate of bacterial colonisation of the skin after preparation: four methods compared. Br J Surg 1987;74:64.[Medline]
- Parsonnet V., Dean D., Bernstein A.D. A method of uniform stratification of risk for evaluating the results of surgery in acquired adult heart disease. Circulation 1989;79(Suppl. 1):13-112.
- Moylan J.A., Fitzpatrick K.T., Davenport K.E. Reducing wound infections. Arch Surg 1987;122(2):150-157.
- Stahle E., Tammelin A., Bergstrom R., Hambreus A., Nystrom S., Hansson H.E. Sternal wound complications: incidence, microbiology and risk factors. Eur J Cardiothorac Surg 1997;11:1146-1153.[Abstract]
- Parisian Mediastinitis Study Group: risk factors for deep sternal wound infection after sternotomy: a prospective multicentre study. J. Thorac. Cardiovasc. Surg. 1996;111:1200–1207.
- Breyer R.H., Mills S.A., Hudspeth A.S., Johnston F.R., Cordell A.R. A prospective study of sternal wound complications. Ann Thorac Surg 1984;37:412-416.[Abstract]
- Demmy T.L., Park S.B., Liebler G.A., Burkholder J.A., Maher T.D., Benckart D.H., Magovern G.J., Jr., Magovern G.J. Recent experience with major sternal wound complications. Ann Thorac Surg 1990;49:458-462.[Abstract]
- Bitkover C.Y., Gardland B. Mediastinitis after cardiovascular operations: a case control study of risk factors. Ann Thorac Surg 1998;65:36-40.[Abstract/Free Full Text]
- Fernandez J., Chen C., Anolik G., Brdlik O.B., Laub G.W., Anderson W.A., McGrath L. Perioperative risk factors affecting hospital stay and hospital costs in open heart surgery for patients >65 years old. Eur J Cardiothorac Surg 1997;11:1133-1140.[Abstract]
- Cruse P., Fuord R. The epidemiology of wound infection: a ten year prospective study of 62939 wounds. Surg Clin N Am 1980;60:27-40.
- Conklin C.M., Gray R.J., Neilson D., Wong P., Tomita D.K., Matloff J.M. Determinants of wound infection incidence after isolated coronary artery bypass surgery in patients randomised to receive prophylactic Cefuroxime or Cefazolin. Ann Thorac Surg 1988;46:172-177.[Abstract]
- Newman L.S., Szczukowski L.C., Bain R.P., Perlino C.A. Supperative mediastinitis after open heart surgery. A case control study of risk factors. Chest 1988;94:546-553.[Abstract/Free Full Text]
- Farifias M.C., Galo Peralta F., Bernal J.M., Rabasa J.M., Revuetta J.M., Gonzales-Macias J. Supperative mediastinitis after open heart surgery: a Case control study covering a seven year period in Santander. Spain. Clin Infect Dis 1995;20:272-279.
- Blanchard A., Hurni M., Ruchat P., Stumpe F., Fisher A., Sadeghi H. Incidence of deep and superficial sternal infections after open heart surgery. A ten year retrospective study from 1981 to 1991. Eur J Cardiothorac Surg 1995;9:153-157.[Abstract]
- Wouters R., Wellens F., Vanermen H., De Geest R., Degrieck I., De Meerleer F. Sternitis and mediastinitis after coronary artery bypass grafting. Analysis of risk factors. Tex. Heart Inst J 1994;21:183-188.[Medline]
- Borger M.A., Rao V., Weisel R.D., Ivanov J., Cohen G., Scully H.E., David T.E. Deep sternal wound infection: risk factors and outcomes. Ann Thorac Surg 1998;65:1050-1056.[Abstract/Free Full Text]
- Sarr M.G., Gott V.L., Townsend T.R. Mediastinal infection after cardiac surgery. Ann Thorac Surg 1984;38:415-423.[Abstract]
- Kouchkoukos N.T., Wareing T.H., Murphey S.F. Risk of bilateral internal mammary artery bypass grafting. Ann Thorac Surg 1990;49:210-219.[Abstract]
- Hazelrigg S.R., Wellons H.A., Jr., Schneider J.A., Kolm P. Wound complications after median sternotomy: relationship to internal mammary grafting. J Thorac Cardiovasc Surg 1989;98:1096-1099.[Abstract]
- Law D.J.W., Mishriki S.F., Jeffrey P.J. The importance of surveillance after discharge from hospital in the diagnosis of post-operative wound infection. Ann Roy Coll Surg Eng 1990;72:207-209.
- Hammermeister K.E., Burchfiel C., Johnson R., Grover F.L. Identification of patients at greatest risk for developing major complications at cardiac surgery. Circulation 1990;82(Suppl. IV):380-389.
Top
Abstract
Introduction
