Introduction
Poststernotomy mediastinitis (PSM) is still one of the most complex and costly infectious processes to treat. Changes in cardiac surgical patient population and contributing pathogens, amongst others, have ensured that the incidence of PSM, despite many advances in prevention, remains significant.
Data recently presented by the Netherlands Association for Cardio-Thoracic Surgery, which includes all 16 cardiac surgical centers, showed an increase in open cardiac procedures between 2007 and 2011 from 15 500 to 16 500 per year [
1]. Data procured from 8 of the 16 centers from 2002 to 2007 revealed a cumulative incidence for surgical site sternal wound infection of 2.4% (95% confidence interval [CI], 1.9 to 3.1) following coronary artery bypass grafting (CABG) [
2]. This figure rose to 3.2% (95% CI, 2.0 to 5.1) where CABG was combined with concomitant valve surgery. This roughly translates to 396 to 530 new cases a year. Between hospitals, the adjusted rates varied from 0.0% to 9.7% [
2]. By comparison, in the United States approximately 700,000 median sternotomies are performed each year, leading to nearly 8,300 cases of deep sternal wound infections [
3].
The acute mortality rates in some large centers still approach 40% internationally [
4]. Even if the patient survives, the long-term mortality rate is significantly higher. In a 10-year follow-up study after CABG [
5], the adjusted survival rate was 39% for patients who had suffered from PSM compared with 70% who did not. The adjusted hazard ratio for mortality during that period was 2.12 (95% CI, 1.80 to 2.49; P < 0.001). The study by Graf and colleagues [
4] confirmed the findings of an earlier study from Uppsala, Sweden [
6], with a similar length of follow-up.
PSM has a significant impact on both healthcare and hospital budgets. In a recent case-control study performed by Graf and colleagues [
4] from Hannover, Germany, the median costs for treatment were almost three times higher (P < 0.0001) and the median length of stay more than doubled (P = 0.0006). Ennker and colleagues [
7] demonstrated an opportunity cost, or turnover increase, in excess of 10 600 euro through reducing the average hospitalization of their PSM patients from 48,43 to 36,73 days by improving treatment efficiency.
The enormous amount of intellectual investment required from a multidisciplinary team including, among others, microbiologists, intensivists, radiologists, cardiologists, cardiothoracic and plastic surgeons in effectively dealing with these cases has yet to be calculated.
Recent developments such as the introduction of topical negative pressure therapy (TNP), new insights into the timing of flap surgery and flap choice, design and methods of harvesting have all become significant in recent years. Further improvement in the treatment and research of PSM requires a targeted approach. A new, evidence-based classification may allow for better comparison between different treatment protocols and may further both research and refinement in the management of PSM.
The aim of this paper is to categorize and assess the effectiveness of a range of primary and secondary reconstructive management options for PSM, based on currently available evidence and present an evidence-based classification. See the “Clinical definition of PSM”.
Clinical definition of poststernotomy mediastinitis (van Wingerden JJ, de Mol BAJM, van der Horst CMAM: Poststernotomy mediastinitis: definition and terminology, Submitted)
Poststernotomy mediastinitis (in adults) must meet the following criteria:
Infection occurs within 1 year, regardless of whether an implant is in place or not
AND infection appears related to the operative procedure
AND, at least 1 of the following criteria:
1.
Patient has organisms cultured from mediastinal tissue or fluid obtained during a surgical operation or needle aspiration.
2.
Patient has evidence of mediastinitis seen during a surgical operation or histopathologic examination.
3.
Patient has at least 1 of the following signs or symptoms with no other recognized cause: fever (>38 0C), chest pain, or sternal instability
AND at least 1 of the following:
a.
Purulent discharge from mediastinal area
b.
Organisms cultured from blood or spontaneous discharge from mediastinal area
c.
Radiological evidence of an infective process in the mediastinum.
Methods
Literature search strategy
A search of the literature from 1990 to March 2014 was conducted without language restrictions using the Cochrane Central Register of Controlled Trials, Ovid Medline, and PubMed and Web of Science databases. Key words and MeSH terms were used to identify and a manual search of the reference lists was done regarding all possible factors (excluding pre-, or intra-operative risk factors) that could influence the current treatment decisions of PSM. Only procedures for which either usefulness or efficacy was claimed were considered. The quality and strength of the evidence was weighed according to the Rating Scheme used by the Society of Thoracic Surgeons (STS) Workforce on Evidence-based Medicine (EBM) for Classification of Recommendation (see below) [
8]. Two hundred thirty three records were identified. Two members (JJvW & DU) reviewed titles and abstracts to exclude records that were of interest. In addition, a manual search of the reference lists was done.
STS workforce on EBM rating scheme to assess the quality and strength of the evidence
Rating Scheme for the Strength of the Evidence
Levels of Evidence
Level A. Data derived from multiple randomized clinical trials.
Level B. Data derived from a single randomized trial or from nonrandomized trials.
Level C. Consensus expert opinion.
Rating Scheme for the Strength of the Recommendations
Classification of Recommendations
Class I. Conditions for which there is evidence or general agreement, or both, that a given procedure is useful and effective.
Class II. Conditions for which there is conflicting evidence or a divergence of opinion, or both, about the usefulness/efficacy of a procedure.
Class IIa. Weight of evidence favors usefulness/efficacy.
Class IIb. Usefulness/efficacy is less well established by evidence.
Class III. Conditions for which there is evidence or general agreement, or both, that the procedure is not useful/effective.
The Level of Evidence was independently established by two authors (JJvW & DU) and a Classification of Recommendation was set by mutual consent. In case of disagreement, a third reviewer (either BdM or CvdH) was consulted.
Discussion
This study shows that, by careful application of the rating scheme recommended by the respected STS workforce on EBM, none of the current reconstructive options, with a few exceptions (case reports), available to the surgical team reaches a level of evidence higher than B. This means that data were derived from either a single randomized trial (RT) or from non-randomized trials, but not from multiple RTs nor through consensus expert opinion. The absence of variation in the level of evidence for each reconstructive option also implies that there is insufficient evidence available at this stage to confirm superiority of one reconstructive method above another. This is the reason why comparison between the various modalities of treatment available to the surgical team is so difficult. Yet, it also suggests that some treatment modalities can be recommended more strongly than others (Class I recommendation) under comparable circumstances. For example both TNP (only) and a delayed flap come with a Class I recommendation in cases where the sternum is found to be fairly stable and viable following an assiduous debridement. This too may complicate determination of the best treatment option. Both difficulties caused by similarities in the level of evidence and classification of recommendation underscore the need for an EB classification as a starting point from which the best decision can be derived. This is the ultimate aim of the proposed Classification. Consistent cognizance of the variables should allow comparison between treatment that are now regarded to be either “apples or pears” (for example TNP (only) vs. flap reconstruction). For example, in our institution our first choice, in cases with a fairly stable and reasonable viable bone, would be the application of TNP following a thorough debridement [
16]. In other centers, in particular in the Eastern hemisphere, where TNP is either not available or popular a flap reconstruction, either with muscle or omentum would be done [
78]. In others again in others [
35],[
36] muscle flap reconstruction, as a one-step procedure, is still routinely preferred. Hitherto, comparison of the outcomes was difficult.
At the other end of the spectrum, the surgical team is left with few options when the sternum is very necrotic and found to be insufficient for closure following debridement. But then, which flap would be preferable and at which stage (primary or delayed) should it be applied? The most difficult cases to decide on (and to compare) are the scenarios in between the two Types I and 4, because of the relatively small number of cases per institution. Comparison could be made easier if some of the variables are standardized. Eventually it would also open a way for meta-analysis, which is necessary in small case series of each institution. The proposed Classification could be a first step in this direction. What this Classification is not primarily intended for is to encourage institutions to alter their approach to each case of PSM. The preferred approach to the further management of PSM is decided on during the first, vigorous debridement that follows immediately upon diagnosing PSM. Patient-related factors (hemodynamic stability, sternal bone viability, etc.) and local circumstances (e.g. availability of material and expertise) will influence this first decision and determine the type of management is planned accordingly. It is this first decision that determines the type of management to which the patient is allocated. Further treatment may change according to altering circumstances (e.g. a flap may be required where initially it was thought unnecessary) but the Type will not change - regardless of what treatment was ultimately given. In this way a prognostic balance generated from the original treatment allocation is ensured. Management approach (type) allocation allows for outcome analysis, which in turn reflects the real-world clinical scenario because it admits treatment-decision alterations and deviations. Insight into the reasons for, and timing of, these decisions will become clearer. A potential weakness of the proposed Classification, some may argue, could be that it is over simplified. This will have to be weighed against a Classification that is so cumbersome that it becomes impractical. Time will tell - at least it is a start.
Conclusion
The development of further, evidence-based guidelines and research would benefit from universally adopted classification of the severe infectious process in the wound that may follow a sternotomy.
In this paper a classification system is introduced based on Evidence-based principles with the aim of providing the highest level of evidence currently available. This system, due to its simplicity, could easily be adopted provided that there is an early, collaborative approach between cardiothoracic-, and reconstructive plastic surgeons. As more evidence becomes available, undoubtedly the classification proposed here will be refined. For more evidence, larger numbers, from numerous centers internationally, are required and for this, a classification as presented here, may serve as a starting as a starting point.
Authors’ contributions
JJvW: Literature search, preparation of the manuscript, revising the manuscript, submission, corresponding author; DU: Literature search, preparation of the manuscript; CvdH: Literature assessment, preparation of the manuscript, revising the manuscript; BdM: Literature evaluation, preparation of the manuscript, revising the manuscript. All authors read and approved the final manuscript.
Competing interests
The authors declare that they have no competing interests.