| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
Interact CardioVasc Thorac Surg 2005;4:555-560. doi:10.1510/icvts.2005.112714 © 2005 European Association of Cardio-Thoracic Surgery
Poststernotomy mediastinitis: comparison of two treatment modalities
a Department of Cardio-thoracic Surgery, Academic Medical Center, University of Amsterdam, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands Received 6 May 2005; received in revised form 20 July 2005; accepted 21 July 2005
*Corresponding author. Tel.: +31-20-5664323; fax: +31-20-5662289.
Poststernotomy mediastinitis (PM) is a serious and potentially lethal condition with an overall incidence varying from 0.4–5%. There is little consensus on the ideal management of PM. The aim of this retrospective study was to investigate the effectiveness of topical negative pressure (TNP) therapy vs. traditional closed drainage techniques (CDT) as a treatment modality for PM. We reviewed the data of 10,467 patients who underwent median sternotomy between 1 January 1992 and 31 December 2003. During this period 63 patients were treated for PM. Twenty-nine of these patients were treated with TNP and 34 with conventional CDT. Primary points of interest were: treatment modalities, mortality, surgical site infection recurrence and, duration of therapy and hospital stay. In this series, the total incidence of poststernotomy mediastinitis was 0.6% with high morbidity rates. In the TNP group, lower rates of recurring infection, therapeutic failure and fewer defects at discharge were seen (P<0.05). In conclusion, the results of our series add further data to the knowledge that PM is an important cause of morbidity and mortality. TNP is a safe and adequate treatment modality for treating PM.
Key Words: Mediastinitis; Cardiac surgery; Topical; Negative pressure; Retrospective
Median sternotomy allows surgeons excellent access to mediastinal structures without the discomfort associated with muscle-splitting incisions and rib retraction. Surgical site infection (SSI) is not an extinct entity: the Centers for Disease Control and Prevention (CDC) reports that among surgical patients, SSI is the most frequent type of infection (38%) [1]. The incidence of poststernotomy mediastinitis (PM) varies from 0.4–5% [2,3]. Although the incidence is low, the high mortality rate (14–47%) and morbidity rate demand early and aggressive treatment of this potentially devastating complication [3,4]. The associated morbidity, prolonged hospital stay, and need for repeated surgical procedures have important economic consequences [3–5]. There is little consensus on the ideal management of PM. Treatment strategies include closed irrigation and radical debridement with reconstructive procedures using pectoralis major, rectus abdominis or omentum flaps. Although good results have been reported from the latter strategy, it is an extensive surgical procedure [4,7,8]. It also creates an even greater surgical challenge if re-operation is indicated. Sternal refixation using closed mediastinal irrigation has been the therapy for PM in most cardiothoracic surgery centers for many years. Although it is a relatively simple method, a high failure rate (up to 52%) has been reported [3,8]. This has led to some authors stating that closed drainage by means of vacuum-drainage is safe and even superior when compared with irrigation techniques. Also because shorter hospital stays were reported [8,9]. In recent years a new technique, topical negative pressure (TNP) therapy, has been introduced resulting in beneficial effects on blood flow to the wound and the proliferation of granulation tissue [10–12]. TNP is a closed system that applies sub-atmospheric pressure to wound tissues through polyurethane foam. The probable mechanism is the removal of interstitial oedema lowering capillary afterload thereby promoting the microcirculation. By removing excess fluid mitosis inhibitory factors and bacterial loads are reduced [12]. Additionally, the mechanical effect of the vacuum results in approximation of wound edges and appears to have an ‘Ilizarovian’ effect on tissues resulting in a proliferation of granulation tissue [11]. This principle is also known in orthopedic surgery as it is seen in the formation of bony tissue. Pulmonary function and chest wall mechanics are compromised in sternal instability. When TNP is applied, the foam acts as a ‘pseudosternum’ and chest stabilization is restored [13]. This study was designed to evaluate two treatment options for PM: TNP vs. closed drainage techniques (CDT). We also conducted this study to assess the feasibility of a randomized trial.
2.1. Patient population Between 1 January 1992 and 31 December 2003 a retrospective report was made of all patients who were treated for PM after cardiac surgery at the Academic Medical Center in Amsterdam, a tertiary-care University hospital. The diagnosis of mediastinitis was made according to the criteria developed by the CDC [1]. El Oakley proposed a classification of PM based on differences in interval until diagnosis and the presence of risk factors (Table 1) [3]. Patients with a delayed diagnosis usually present without sternal instability as a large portion of the sternum may have healed creating chest stability. Open wound drainage controls the sepsis so there are rarely systemic signs or symptoms. Although lesser signs are present, treatment is a surgical challenge because of more involvement of bony tissue and adhesions.
Patients were selected with the aid of a surgical site infection database, registrations of surgical procedures and wound cultures. Information on the latter was provided by our Department of Medical Microbiology. Data were collected from surgical and cardiology medical and nursing records. Primary points of interest were: SSI treatment techniques, mortality, SSI recurrence and, duration of therapy and hospital stay. To predict the morbidity and mortality risk perioperatively, we used the EuroSCORE [14]. 2.2. Surgical technique After diagnosis, aggressive debridement of all foreign material (steel wires, sutures), necrotic and infectious tissues was done. Sternal edges were revised with a curette or oscillating saw. If TNP was to be applied, particular care was taken to leave no sharp sternal edges especially in the area overlying the right ventricle. Before closing with internal fixation material or TNP, the wound was irrigated and thorough hemostasis was performed.If the sternum was closed, this was done in the usual manner using interrupted steel wires and further closure in layers. If tissue cultures were repeatedly negative, drains were removed. If TNP was chosen (KCI Medical BV, Houten, the Netherlands), we used a hydrocolloid adhesive (Allevyn® Thin, Smith&Nephew BV, Hoofddorp, the Netherlands) as protection for skin edges. Delicate anatomical structures (aorta, ventricles and bypass grafts) were covered with gauze dressings (Adaptic®, Johnson&Johnson Wound Management, Amersfoort, the Netherlands). Several layers of polyurethane foam dressing were trimmed to fit between the wound edges. When adhesions allowed, one layer of foam was positioned underneath the sternum. Subsequently, a strip of foam was placed into the sternal defect itself. Depending on the patient's body mass, the most superficial layer of foam should rise to 4 cm above skin level when a vacuum is not in use, to allow meticulous contact of the foam to the wound when the system collapses on application of the vacuum (Video 1). This is particularly important if a patient is likely to cough e.g. obese patients or those with COPD.
The pump unit was programmed at 125 mmHg of continuous negative pressure. We did not use intermittent pressure therapy because frequent cessation of vacuum was experienced as being painful by most patients. The dressing was initially changed after 48 h and thereafter every four to five days. Whenever possible this took place on the surgical ward after reassuring the patient. If clinical inspection demonstrated a well-vascularized wound completely covered by granulation tissue, C-reactive protein levels were <50 mg/l and cultures did not show pathogenic bacteria, sternal closure was performed. This involved closing the sternum with interrupted steel-wire sutures. The pectoral fascia was mobilized as far as possible to cover the fixation material. Further closure in layers was done using interrupted stitches. Drains were positioned substernally and subcutaneously to avoid seroma. If sternal closure was not possible or the skin edges were subject to too great a force, we used partial or total mobilization of the pectoral muscle either as a myocutaneous flap or as a muscle flap to close the defect with leaving behind drains subpectoral for five days at a minimum. 2.3. Statistical analysis Individual variables were compared using the two-independent-samples test (Mann–Whitney) with significance for P-values < 0.05. Independent factors determining survival, therapeutic failure and wound defect at discharge were analyzed by a Cox proportional hazard model or logistic regression analysis
Between 1 January 1992 and 31 December 2003 approximately 10,467 sternotomy procedures were performed at our hospital. The incidence of PM at our hospital was 0.6% (N=67). Of 63 included PM patients, 34 underwent CDT. Twenty-nine patients with PM were treated by TNP. Four patients were excluded because of missing data. 3.1. Patient characteristics Patient characteristics are presented in Table 2.
In our study population we have seen older patients (66.7 vs. 65.9 years), higher EuroSCOREs (7.3 vs. 6.4) and more combined CABG and valve surgery, in recent years (Table 2). The more complex surgery was also translated into longer surgical procedure times (P=0.04). It is striking that more females have been treated for PM in recent years (12 vs. 4, P=0.008). Other preoperative, intraoperative and postoperative risk factors showed no major differences. 3.2. Diagnosis, morbidity and hospital stay Table 3 shows variables related to SSI.
According to El Oakley's classification [3], no significant differences were noted. However, there seems to be a shift towards a more delayed diagnosis (class IVa 31% vs. 17.6%) with a prolonged mean surgery-to-infection interval (20.3 vs. 16 days). Morbidity was high e.g. blood stream infection (63.5%), renal failure (38.1%), bleeding (17.5%), multiple organ failure (15.9%) and other infections (14.3%). No significant differences were seen between TNP and CDT. Few valve related infections were seen (N=2). Both patients were treated by closed irrigation techniques. Procedure-related complications after TNP were rare (N=1). One patient had a minor substernal right ventricular bleed. A non-significant increase in duration of therapy was seen in the TNP group (22.8 vs. 16.5 days). This did not lead to a significant increase in hospital stay. A comparable average duration of TNP is seen in literature [10,12]. Most patients were extubated immediately after TNP application and only when indicated by sepsis or other reasons they were taken to the intensive care unit (N=7). During TNP treatment, dressings were changed on average five times with a range of 2–14 times. 3.3. Mortality During the study period 18 patients died. In 11 patients SSI was the cause of death (17.5% vs. 20.6%, NS). The other 7 patients died after discharge, due to heart failure or other non-surgical causes.Significant factors determining survival proved to be: age 70 (P=0.03), EuroSCORE (P=0.01), combined surgery (P=0.03), duration of surgical procedure (P=0.03), left ventricular function (P=0.04), prolonged postoperative inotropic therapy (P=0.008), renal failure (P=0.02) and multiple organ failure (P<0.0001). 3.4. Therapeutic failure Therapeutic failure was defined as recurrence of wound infection, a change to other treatment techniques and the need for multiple surgical interventions to control infection or mortality caused primarily by the surgical site infection. Therapeutic failure occurred in 27.6% of TNP patients and in 58.9% of CDT patients (P=0.01).A total of 26 patients (41.3%) developed recurrent wound infection (27.6% vs. 52.9%, P=0.04). In the statistical analysis, treatment modality (P=0.04), age 75 (P=0.03) and left ventricular function (P=0.04) were seen as significant factors for recurrent disease. Although shorter duration of therapy was seen in CDT (NS), this was not a factor in determining recurrent SSI after statistical calculations. In total, 28 patients (44.4%) complete healing was not achieved at discharge (27.6% vs. 58.9%, P=0.01) and prolonged treatment as an outpatient was necessary. Statistical analysis revealed only treatment modality (P=0.01), age 75 (P=0.05) and prolonged postoperative inotropic therapy (P=0.03), as significant factors associated with a surgical site defect at discharge. Duration of therapy or hospital stay was not associated. 3.5. Surgical closure techniques after topical negative pressure therapy Surgical closure techniques after TNP are reported in Table 4.
At our institution secondary closure including sternal refixation is the treatment of choice. However, closure without sternal refixation (N=4) resulted in a clinically satisfactory situation without loss of respiratory function or patient discomfort. Reconstructive techniques using a pectoral muscle were successfully applied in 9 patients.
The overall incidence of PM is low but it has devastating and potentially lethal consequences. In this series, the incidence of PM was 0.6%. Total mortality resulting from PM remained high (17.5%). Also as described in Table 3, high morbidity rates were seen. With the current trends of treating older people and those with complex conditions, and favoring shortened hospital stay [15], SSI remains an important complication of cardiac surgery although great efforts are being made in identifying risk factors and developing preventive strategies. Trends towards higher EuroSCOREs (NS), older patients (NS), longer surgical procedure times (P=0.04) and delay in diagnosis (longer surgery-to-infection interval, NS) were also observed in this study, but this did not result in a significant increase in hospital stay after SSI in patients treated with TNP. In the CDT group, however, higher rates of recurring infection (P=0.04), therapeutic failure (P=0.01) and more defects at discharge were seen (P=0.01). TNP shows good clinical results of beneficial effects on blood flow to the wound and proliferation of healing granulation tissue. Progression can be easily monitored as it allows immediate wound inspection. It gives good patient and nurse satisfaction as it enables a more normal mobilization than traditional CDT. In addition, dressings do not need to be changed several times a day like in traditional open wound management. Furthermore, when the vacuum is applied, the polyurethane foam dressing acts as a ‘pseudosternum’ creating chest stability. In conclusion, the results of our series add further data to the knowledge that PM after cardiac surgery is an important cause of morbidity and mortality. Based on its clinical success and supported by results of this study, we believe that TNP is the most effective treatment for PM. 4.1. Limitations of the study An evidence-based introduction of a new wound treatment modality like TNP appears to be very difficult. Promising clinical results and advances in TNP in other areas of surgery, make is very difficult to withhold this therapy from patients suffering from SSI after cardiac surgery. And, only small numbers are available for research purposes. These factors preclude well-designed randomized trials. We also realise the limitations of this retrospective study as our patients' sample-size is small and drawn over a long period of time. Statistical differences need to be interpreted with care because confounding factors may be present.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
Top
Abstract
1. Introduction
