Article

lock Open Access lock Peer-Reviewed

7

Views

ORIGINAL ARTICLE

Validation of MagedanzSCORE as a predictor of mediastinitis after coronary artery bypass graft surgery

Michel Pompeu Barros de Oliveira Sá; Evelyn Soares Figueira; Cecília Andrade Santos; Omar Jacobina Figueiredo; Renato Oliveira Albuquerque Lima; Fábio Gonçalves de Rueda; Rodrigo Renda de Escobar; Alexandre Magno Macário Nunes Soares; Ricardo de Carvalho Lima

DOI: 10.5935/1678-9741.20110013

ABSTRACT

OBJECTIVE: The aim of this study is to evaluate the applicability of a new score for predicting mediastinitis - MagedanzSCORE - in patients undergoing coronary artery bypass graft (CABG) surgery in the Division of Cardiovascular Surgery of Pronto Socorro Cardiológico de Pernambuco - PROCAPE.
METHODS: Retrospective study involving 500 patients operated between May/2007 and April/2010. The registers contained all the information used to calculate the MagedanzSCORE. The outcome of interest was mediastinitis. We calculated sensitivity, specificity, positive predictive value, negative predictive value, concordance and accuracy. The accuracy of the model was evaluated by ROC (receiver operating characteristic) curve.
RESULTS: The incidence of mediastinitis was 5.6%, with a lethality rate of 32.1%. In univariate analysis, the five variables of the MagedanzSCORE were predictors of postoperative mediastinitis: chronic obstructive pulmonary disease (OR 6.42; 95.0% CI 2.76-14.96; P<0.001), obesity (OR 3.06; 95.0% CI 1.32-7.09; P=0.009), surgical reintervention (OR 82.40; 95.0% CI 30.40-223.30; P<0.001), multiple transfusion (OR 3.33; 95.0% CI 1.52-7.29; P=0.003) and stable angina class IV or unstable (OR 2.59; 95.0% CI 1.19-7.64; P=0.016) according to Canadian Cardiovascular Society. The score had a sensitivity of 96.4%, specificity of 90.0%, positive predictive value of 36.5%, negative predictive value of 99.8% and 90.4% concordance. The accuracy measured by the area under the ROC curve was 96.2% (95.0% CI 94.5%-97.9%).
CONCLUSIONS: The MagedanzSCORE proved to be a simple and objective index, revealing a satisfactory predictor of development of postoperative mediastinitis in patients undergoing CABG surgery at our institution

RESUMO

OBJETIVO: O objetivo deste estudo é avaliar a aplicabilidade de um novo escore de predição de mediastinite - MagedanzSCORE - em pacientes submetidos à cirurgia de revascularização miocárdica (CRM) na Divisão de Cirurgia Cardiovascular do Pronto Socorro Cardiológico de Pernambuco - PROCAPE.
MÉTODOS: Estudo retrospectivo envolvendo 500 pacientes operados entre maio/2007 e abril/2010. Os registros continham todas as informações utilizadas para calcular o MagedanzSCORE. O desfecho de interesse foi mediastinite. Foram calculados sensibilidade, especificidade, valor preditivo positivo, valor preditivo negativo, concordância e acurácia. A acurácia do modelo foi avaliada pelo ROC (receiver operating characteristic curve).
RESULTADOS: A incidência de mediastinite foi de 5,6%, com uma taxa de letalidade de 32,1%. Na análise univariada, as cinco variáveis do MagedanzSCORE foram preditoras de mediastinite no pós-operatório: doença pulmonar obstrutiva crônica (OR 6,42; IC 95,0% 2,76-14,96; P<0,001), obesidade (OR 3,06; IC 95,0% 1,32-7,09; P=0,009), reintervenção cirúrgica (OR 82,40; IC 95,0% 30,40-223,30; p<0,001), transfusões múltiplas (OR 3,33; IC 95,0% 1,52-7,29; P=0,003) e angina estável classe IV ou instável (OR 2,59; IC 95,0% 1,19-7,64; P=0,016) de acordo com a Sociedade Canadense de Cardiologia. O escore teve uma sensibilidade de 96,4%, especificidade de 90,0%, valor preditivo positivo de 36,5%, valor preditivo negativo de 99,8% e 90,4% de concordância. A acurácia medida pela área sob a curva ROC foi 96,2% (IC 95,0% 94,5%-97,9%).
CONCLUSÕES: O MagedanzSCORE provou ser um índice simples e objetivo, revelando-se um preditor satisfatório de desenvolvimento de mediastinite no pós-operatório de pacientes submetidos à CRM em nossa instituição
INTRODUCTION

It is important that there are models for accurate prediction of risks in the current practice of cardiac surgery. These models allow surgeons and institutions to compare results in a significant way. They are also useful in surgical decision making, development of informed consent preoperatively, quality and management of health care [1]. Few scores assess the postoperative morbidity [2].

Despite technological development and improvement of health professionals, mediastinitis - deep wound infection, with clinical evidence and / or microbiological commitment of the retrosternal space, and may be associated with sternal osteomyelitis with or without its instability - also features a large challenge, since it determines a significant increase in morbidity and mortality rates, hospital costs and significant impact on social life of patients who survive [3].

Magedanz et al. [2] developed a score for use in daily practice to calculate the risk of mediastinitis after coronary artery bypass grafting (CABG). The score includes variables collected routinely and easy to use. We baptized this score in our institution: the MagedanzSCORE.

Risk scoring systems are most applicable when the characteristics of the patient preoperatively and treatment profiles are comparable to those on which the system originated. For this reason, any risk scoring system can be used reliably when its validity was tested in the local patient population [4]. The MagedanzSCORE has not been tested outside the institution where it was created, so that we do not know whether it is useful in predicting mediastinitis after CABG in other populations, in the case, in our local institution.

The aim of this study is to test the MagedanzSCORE on the population that underwent CABG at our institution.

 

METHODS

Source Population

After approval by the ethics committee of institution, we reviewed the records of patients undergoing CABG at the Division of Cardiovascular Surgery of Pronto Socorro Cardiológico de Pernambuco (PROCAPE) from May 2007 to April 2010. We excluded patients whose records did not contain the necessary data concerning the variables to be studied.

Study Design

It was a retrospective study of exposed and nonexposed to certain factors (independent variables) with outcome (dependent variable) followed by validation of a model (the MagedanzSCORE).

The independent variables were: age (years), gender (male or female), surgical reoperation (due to bleeding, cardiac tamponade or other reason other than mediastinitis), chronic obstructive pulmonary disease (COPD - diagnosed clinically and / or radiological examination of the chest and / or spirometry and / or drug treatment - corticosteroids, bronchodilators), obesity (BMI > 30.0 kg / m2), stable angina class IV / unstable by the Canadian Cardiovascular Society, multiple transfusion postoperatively (more than 3 units of any type of blood products).

The dependent variable was mediastinitis after surgical procedure. This variable was categorized into yes or no. Mediastinitis were considered with those who met the criteria according to the Centers for Disease Control and Prevention (CDC) [5]:

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 (38ºC), chest pain, or sternal instability and at least 1 of the following:

a. purulent discharge from mediastinal area

b. organisms cultured from blood or discharge from mediastinal area

c. mediastinal widening on radiography.

Isolated superficial infections from sternal wound with stable and/or sterile sternal dehiscence and/or no macroscopic evidence of deep infection (purulent drainage) were not considered as having mediastinitis.

Each patient was evaluated for the presence or absence of the five risk factors established by MagedanzSCORE, respecting the definition of each of them and giving them the correct score (Table 1). Depending on the final score, each patient was placed in one of the four risk groups (low, medium, high or very high). We recorded the outcomes (development or non-development of mediastinitis).

 

 

 

Statistical Methods

Data were analyzed using percentage and descriptive statistics measures: mean, median and standard deviation. The following tests were used: t-Student with equal or unequal variances (for parametric variables) and chi-square test or Fisher's exact (as appropriate, for non-parametric variables). In the study of univariate association between categorical variables, the values of the Odds Ratio (OR) and a confidence interval (CI) for this parameter with a reliability of 95.0% were obtained.

We calculated sensitivity, specificity, positive predictive value and negative predictive value and established the best cut-off. We used the kappa statistic to assess agreement between the prediction of the outcome by the score and the observed in the study.

The accuracy (discrimination ability of the MagedanzSCORE) was calculated using the area under the ROC curve (receiver operating characteristic curve), built on the sensitivity (correct prediction of mediastinitis) and specificity (correct prediction of non-mediastinitis) calculated for each value of the score.

The level of significance in the decision of the statistical tests was 5%. The program used for data entry and retrieval of statistical calculations was SPSS (Statistical Package for Social Sciences) version 15.0.

 

RESULTS

Five-hundred and forty-two patients underwent CABG during the study period; 42 patients were excluded for lack of information regarding the study in the medical records. So we analyzed 500 patients undergoing CABG with a mean age of 62.1 ± 9.9 years, 59% male. The following clinical features were observed (Table 2): 5.6% underwent reoperation, 14.8% were obese, 9.4% had chronic obstructive pulmonary disease, 21.8% had multiple transfusions in the postoperative period and 36.2% had stable angina class IV or unstable angina. Mediastinitis was diagnosed in 28 patients (5.6%), of which 9 (32.1%) died. According to the MagedanzSCORE, 35.6% of study patients have low risk of mediastinitis, 49.6% medium risk, 11.4% high risk and 3.4% very high risk of mediastinitis.

 

 

 

Analyzing the variables proposed in MagedanzSCORE with the occurrence of mediastinitis, we observed that the requirement for reoperation is the one that was most associated with mediastinitis followed by COPD, multiple transfusion, obesity and angina, with statistically significant increase in the risk of mediastinitis with magnitude of 82.40, 6.42, 3.33, 3.06 and 2.59 times, respectively (Table 3). The odds ratio was calculated after adjusting for age of patients, so the risks are independent of patient age. There was no statistically significant difference in average age of the patients with (63.5 ± 8.0 years) and without mediastinitis (62.1 ± 10.0 years). There was also no statistically significant difference in the frequency of mediastinitis according to gender (Table 3).

 

 

 

The relationship between the risk classification and mediastinitis according to MagedanzSCORE, it was observed that 96.4% of patients diagnosed with mediastinitis in this study are classified as high risk and very high risk (Table 4). This shows a strong association with the proposed score.

 

 

 

According to the measures of sensitivity and specificity, the score 2 has the best probability of identifying the presence and absence of mediastinitis, presenting 96.4% sensitivity and 90.0% specificity (Table 5), being considered the best cut-off for the model. This cut-off separates the low/medium level from the high/very high level. In Table 6, it was considered as cut-off the score 2 for the presence of mediastinitis according to MagedanzSCORE, being patients with low or medium risk being compared between patients with high and very high risk for mediastinitis. Observe that there is a concordance of 90.4% (excellent) with this cutoff. According to the Kappa index, the concordance between the score and observed in the study is moderate (0.488). However, the positive predictive value (probability of mediastinitis in patients classified as high risk and very high) was 36.5%, so it is expected that among 100 patients with high/very high risk score, only 37 are truly mediastinitis. Moreover, the score has a negative predictive value (probability of absence of mediastinitis in patients classified as low/medium risk) 99.8%, thus we can conclude that the use of the score has a great ability to identify the absence of future disease.

 

 

 

 

 

 

According to the results presented in the area under the ROC curve (overall capacity of the measure used to discriminate individuals with and without the disease), measured by 96.2% (95.0% confidence interval 94.5%97.9%), the MagedanzSCORE shown an excellent measure to identify patients with risk of development of postoperative mediastinitis (Figure 1).

 

 

 

DISCUSSION

The incidence of mediastinitis in this study was 5.6%, higher than in the original study Magedanz et al [2] which was 3.3% and slightly higher than the one described in the literature, ranging from 0.4% to 5.0% [6-8]. However, it should be noted that we are analyzing a risk group. A previous study in our institution [9] showed an incidence of mediastinitis of 2.4% among 1038 cardiovascular surgeries (involving all types of cardiac surgeries) and CABG was associated with increased risk compared with other cardiac surgeries (95% CI 3.44-8.30, P = 0.0001), justifying the high incidence on the present study.

Many factors have been associated with development of mediastinitis after cardiac surgery [3]. However, there is no consensus as to which factors are most important and how each is an independent predictor of risk for postoperative mediastinitis [3].

Reoperation was the most associated factor with mediastinitis at our institution. Other studies have also identified this variable as a strong predictor of deep sternal wound infection and mediastinitis [10,11]. It is expected that patients who are reoperated have increased risk of infection, considering that repeated surgical aggression exposes patients to contamination.

COPD was the second most associated factor with mediastinitis at our institution. Other studies have shown the correlation of chronic obstructive pulmonary disease and mediastinitis [12]. Chronic obstructive pulmonary disease causes mechanical problems, leading to sternum instability, which in those cases may suggest the harmful effect of smoking in the pathophysiology of mediastinal infection as well. Apart from that, it is more often possible for patients with chronic obstructive pulmonary disease to be colonized [13]. Patients with COPD are more susceptible to wound infection due to tissue hypoxemia [7]. Many of these patients require steroids in pre-and postoperatively, which could facilitate the emergence of infections [7]. Also possibly as a result of frequent coughing, which might contribute to wound dehiscence and thereby facilitate bacterial migration [13]. Patients with COPD also experience more frequent respiratory infections and prolonged weaning from artificial ventilation [14,15].

We also observed obesity as a risk factor associated to postoperative mediastinitis. Milano et al. [16] discuss some factors that may explain why obesity is a risk factor, such as, for example, the dose of the prophylactic antibiotic, which is not corrected on the basis of the patient's body mass index. They also suggest that skin preparation can be difficult and inadequate due to deep skin folds, and the adipose tissue itself, which may function as a substrate for infection. The etiology also relates to increased postoperative mechanical loads, facilitated bacterial contamination and bradytrophic properties of fatty tissue contribute to poor wound healing [17].

Our study showed that angina was associated with increased risk of mediastinitis in postoperative coronary artery bypass surgery. Only two studies [2,18] were identified citing the presence of angina as a predictor of risk of mediastinitis after CABG, however, none of these make any reference to possible pathogenetic mechanism of this relationship.

Loop et al. [11] showed that blood transfusions had a relative risk of causing sternal wound complications of 1.05 per unit. It is considered to be a risk factor for postoperative infection [19]. This may be explained by a decrease in immune function after homologous blood transfusion [11,20]. It has also been identified that the removal of blood transfusion reduces the risk of infections [21].

Risk stratification allows estimating the surgical operative risk facing the patient and evaluate the results and, ultimately, the quality of care of an institution. However, most current risk scores include little assessment of postoperative morbidity.

The MagedanzSCORE presented an excellent accuracy, since the discrimination power of the model in this study according to the ROC curve was 96.2% (95.0% Confidence Interval 94.5%-97.9%).

 

CONCLUSION

The MagedanzSCORE proved to be a simple and objective index, revealing a satisfactory predictor of development of postoperative mediastinitis in patients undergoing CABG surgery at our institution.

REFERENCES

1. Mejía OAV, Lisboa LAF, Puig LB, Dias RR, Dallan LA, Pomerantzeff PM The 2000 Bernstein-Parsonnet score and EuroSCORE are similar in predicting mortality at the Heart Institute, USP. Rev Bras Cir Cardiovasc. 2011;26(1):1-6. [MedLine] View article

2. Magedanz EH, Bodanese LC, Guaragna JC, Albuquerque LC, Martins V, Minossi SD Risk score elaboration for mediastinitis after coronary artery bypass grafting. Rev Bras Cir Cardiovasc. 2010;25(2):154-9. [MedLine] View article

3. Sá MPBO, Soares EF, Santos CA, Figueiredo OJ, Lima ROA, Escobar RR Risk factors for mediastinitis after coronary artery bypass grafting surgery. Rev Bras Cir Cardiovasc. 2011;26(1):27-35. [MedLine] View article

4. Sá MPBO, Soares EF, Santos CA, Figueredo OJ, Lima ROA, Escobar RR EuroSCORE and mortality in coronary artery bypass graft surgery at Pernambuco Cardiologic Emergency Medical Services [Pronto Socorro Cardiológico de Pernambuco] : Rev Bras Cir. Cardiovasc. 2010;25(4):474-82.

5. Horan TC, Andrus M, Dudeck MA CDC/NHSN surveillance definition of health care : associated infection and criteria for specific types of infections in the acute care setting. Am J Infect Control. 2008;36(5):309-32. [MedLine]

6. Sá MPBO, Soares EF, Santos CA, Figueiredo OJ, Lima ROA, Escobar RR Skeletonized left internal thoracic artery is associated with lower rates of mediastinitis in diabetic patients. Rev Bras Cir Cardiovasc. 2011;26(2):183-9. [MedLine] View article

7. Guaragna JC, Facchi LM, Baião CG, Cruz IBM, Bodanese LC, Albuquerque L Preditores de mediastinite em cirurgia cardíaca. Rev Bras Cir Cardiovasc. 2004;19(2):165-70. View article

8. Ledur P, Almeida L, Pellanda LC, Schaan BD Predictors of infection in post-coronary artery bypass graft surgery. Rev Bras Cir Cardiovasc. 2011;26(2):190-6. View article

9. Sá MP, Silva DO, Lima EN, Lima RC, Silva FP, Rueda FG Postoperative mediastinitis in cardiovascular surgery postoperation : Analysis of 1038 consecutive surgeries. Rev Bras Cir Cardiovasc. 2010;25(1):19-24. [MedLine]

10. Ridderstolpe L, Gill H, Granfeldt H, Ahlfeldt H, Rutberg H Superficial and deep sternal wound complications : incidence, risk factors and mortality. Eur J Cardiothorac Surg. 2001;20(6):1168-75. [MedLine]

1. Loop FD, Lytle BW, Cosgrove DM, Mahfood S, McHenry MC, Goormastic M J. Maxwell Chamberlain memorial paper : Sternal wound complications after isolated coronary artery bypass grafting: early and late mortality, morbidity, and cost of care. Ann Thorac Surg. 1990;49(2):179-86.

12. Bitkover CY, Gardlund B Mediastinitis after cardiovascular operations : a case-control study of risk factors. Ann Thorac Surg. 1998;65(1):36-40. [MedLine]

13. 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(3):183-8. [MedLine]

14. Baskett RJ, MacDougall CE, Ross DB Is mediastinitis a preventable complication : A 10-year review. Ann Thorac Surg. 1999;67(2):462-5. [MedLine]

15. Marggraf G, Splittgerber FH, Knox M, Reidemeister JC Mediastinitis after cardiac surgery : epidemiology and current treatment. Eur J Surg Suppl. 1999;(584):12-6. [MedLine]

16. Milano CA, Kesler K, Archibald N, Sexton DJ, Jones RH Mediastinitis after coronary artery bypass graft surgery : Risk factors and long-term survival. Circulation. 1995;92(8):2245-51. [MedLine]

17. Diez C, Koch D, Kuss O, Silber RE, Friedrich I, Boergermann J Risk factors for mediastinitis after cardiac surgery : a retrospective analysis of 1700 patients. J Cardiothorac Surg. 2007;2:23. [MedLine]

18. Abboud CS, Wey SB, Baltar VT Risk factors for mediastinitis after cardiac surgery. Ann Thorac Surg. 2004;77(2):676-83. [MedLine]

19. Dorneles CC, Bodanese LC, Guaragna JCVC, Macagnan FE, Coelho JC, Borges AP The impact of blood transfusion on morbidity and mortality after cardiac surgery. Rev Bras Cir Cardiovasc. 2011;26(2):222-9. View article

20. Murphy PJ, Connery C, Hicks Jr GL, Blumberg N Homologous blood transfusion as a risk factor for postoperative infection after coronary artery bypass graft operations. J Thorac Cardiovasc Surg. 1992;104(4):1092-9. [MedLine]

21. Cartier R, Brann S, Dagenais F, Martineau R, Couturier A Systematic off-pump coronary artery revascularization in multivessel disease : experience of three hundred cases. J Thorac Cardiovasc Surg. 2000;119(2):221-9. [MedLine]

Article receive on Wednesday, September 15, 2010

CCBY All scientific articles published at rbccv.org.br are licensed under a Creative Commons license

Indexes

All rights reserved 2017 / © 2024 Brazilian Society of Cardiovascular Surgery DEVELOPMENT BY