Ann Thorac Surg 2005;80:2314-2320
© 2005 The Society of Thoracic Surgeons
Original article: Cardiovascular
Pediatric Poststernotomy Mediastinitis
Abdullah A. Al-Sehly, MD
a
,
Joan L. Robinson, MD
a
,
*
,
Bonita E. Lee, MD
a
,
Geoff Taylor, MD
c
,
David B. Ross, MD
b
,
Murray Robertson, MD
a
,
Ivan M. Rebeyka, MD
b
a Department of Pediatrics and Stollery Children's Hospital, University of Alberta, Edmonton, Alberta, Canada
b Department of Surgery, University of Alberta, Edmonton, Alberta, Canada
c Department of Medicine, University of Alberta, Edmonton, Alberta, Canada
Accepted for publication May 12, 2005.
* Address correspondence to Dr Robinson, Stollery Children's Hospital, 2C3 Walter MacKenzie Centre, 8440-112 St, Edmonton, AB, T6G 2B7, Canada (Email: jr3{at}ualberta.ca).
 |
Abstract
|
|---|
BACKGROUND: Mediastinitis results in significant morbidity in pediatric cardiac patients. It is not clear whether delayed sternal closure is a risk factor for these infections. Management of mediastinitis remains controversial.
METHODS: Cases of mediastinitis at the Stollery Children's Hospital from January 1, 1991, to June 30, 2004, were reviewed.
RESULTS: There were 29 cases of mediastinitis in 2,675 open cardiac procedures for an overall incidence of 1.1%. Infection was diagnosed 5 to 27 days after the original surgical procedure (median, 10 days). The odds ratio for infection with delayed sternal closure versus primary sternal closure was 1.88 (95% confidence interval, 0.63 to 5.60). Signs at the onset of infection included fever (86%), incisional erythema (69%), purulent drainage from the incision or pacer wire sites (83%), and wound dehiscence (23%). Debridement was followed by primary sternal closure in all but three cases in which the sternum had not been closed before debridement and rotational muscle flaps were not used. Continuous irrigation systems were used only in the first 7 patients. One patient died of mediastinitis complicated by infective endocarditis, and 2 patients died of multiorgan failure.
CONCLUSIONS: Delayed sternal closure was not a major risk factor for mediastinitis, especially if primary skin closure was used with delayed sternal closure. Excellent results were attained with debridement and primary closure of these infections.
|
Introduction
|
|---|
Mediastinitis in pediatric patients usually occurs in the setting of cardiac surgery. Centers for Disease Control and Prevention guidelines classify surgical site infections (SSI) in cardiac patients as superficial wound infections, deep wound infections, sternal osteomyelitis, or mediastinitis (with the latter two being classified together as organ space SSI 1). Intraoperative introduction of organisms is thought to be the primary etiologic process, but there is some evidence that postoperative factors may also be of importance [2]. Risk factors are assumed to be the same for all types of SSI. Well-established risk factors in adults include obesity [3–5], renal impairment [4], peripheral vascular disease [4], diabetes mellitus [4, 5], prolonged duration of surgery [5], prolonged time on cardiopulmonary bypass [5], reoperation for bleeding [5], and the use of bilateral internal thoracic artery grafts [3]. Many of these factors are not applicable in children, where younger age [6–9], a higher American Society of Anesthesiologists score [6, 7, 9], and a longer duration of surgery [6, 8, 10] appear to be consistent risk factors. Delayed sternal closure (DSC) is commonly used during neonatal cardiac surgery, but it is not clear whether it is a risk factor for SSI.
Mediastinitis generally presents days to weeks after cardiac surgery, and requires surgical drainage or debridement for cure. The use of postoperative continuous mediastinal irrigation [11, 12] or rotational muscle flaps to manage the wound [13, 14] remains controversial. The purpose of this study was to determine whether DSC increases the risk of mediastinitis, and to describe the clinical course and outcome of mediastinitis in a center in which muscle flaps are not used for wound reconstruction.
|
Material and Methods
|
|---|
Approval to conduct this study was obtained from the Health Research Ethics Board of the University of Alberta.
Patients
All patients younger than 18 years of age who had open intracardiac surgery at the University of Alberta Hospital and Stollery Children's Hospital from January 1, 1991, to June 30, 2004, and developed mediastinitis were included. Cases were identified by records maintained by trained and experienced infection control practitioners who prospectively followed patients for SSI from the time of surgery until hospital discharge or for 14 days (whichever occurred first) for the duration of the study. In addition, a retrospective search of the cardiac surgical database from September 1, 1996, to June 30, 2004, was performed to identify cases of mediastinitis that occurred more than 14 days postoperatively or after hospital discharge. Mediastinitis was defined as the presence of pus in the retrosternal space or isolation of organisms from retrosternal cultures obtained at drainage or debridement.
Data Collection
Charts were reviewed on identified patients and data collected on the patient's age, sex, chromosomal abnormalities, presence or absence of a gastrostomy tube, type and duration of surgical procedure, presence and duration of DSC, duration of postoperative ventilation and inotropic agent administration, and the need for emergency reoperation before the onset of infection. For each episode of mediastinitis, data were collected on the timing of the onset with regard to surgery (when an emergency reoperation was done, the time of infection from the most recent surgical procedure was recorded), signs at the onset of infection, white blood cell count between the onset of infection and debridement, presence of bacteremia, therapy administered, and the outcome. The onset of infection was considered to be the first day when there was both fever and signs of a wound infection, or the first day when the diagnosis of a deep surgical site infection was considered if signs of a wound infection were absent.
Statistical Analysis
Fisher's exact test with two-sided significance was used to determine the significance of DSC as a factor for mediastinitis in patients who had cardiac surgery from September 1, 1996, to June 30, 2004 (data on DSCs were incomplete before this time) and whether there was a significant difference in the occurrence of bacteremia in mediastinitis caused by Staphylococcus aureus versus other organisms.
|
Results
|
|---|
Patient Characteristics
There were 29 cases of mediastinitis in 28 patients (13 boys and 15 girls) after 2,675 open cardiac procedures, for an overall incidence of 1.1%. One boy had two episodes of mediastinitis with surgeries 11 months apart, first with S aureus and then with Pseudomonas aeruginosa. The age of the patients ranged from 7 days to 14 years (median, 1.0 years) at the time of the 29 surgical procedures that resulted in infection (Table 1). Six patients had trisomy 21, one had probable Williams' syndrome, and the patient with two episodes of mediastinitis had chromosome 22q11 deletion. Four of the 29 patients had gastrostomy tubes at the time of the surgery. The duration of the surgery ranged from 82 to 411 minutes (median, 202 minutes) and bypass duration ranged from 23 to 215 minutes (median, 111 minutes). Cefazolin (three doses total) was used for perioperative prophylaxis in almost all cases with primary closure of the sternum, and a 48-hour course of cefazolin and gentamicin was standard for DSC. Infection occurred 5 to 27 days after the most recent surgical procedure (median, 10 days). The onset of infection was before discharge from the intensive care unit in 8 patients (1 patient was receiving extracorporeal membrane oxygenation, 4 others were still being ventilated, and 3 of these 4 patients were still receiving inotropic agents). Thirteen cases of mediastinitis occurred while the patients were still in the hospital after the index surgery, and eight cases occurred after hospital discharge.
Evidence for Infection
At the onset of the 29 episodes of mediastinitis, 25 patients (86%) had a temperature of 38.5°C or greater (with fever sometimes resolving within 24 hours of starting antimicrobial therapy even if debridement had not yet been performed), 20 patients (69%) had erythema of the incision, 24 patients (83%) had purulent drainage from the incision (n = 22) or from pacer wire sites (n = 2), and 6 patients (23%) of the 26 patients with a closed sternum had wound dehiscence. Excluding the 3 patients with sternums that had not been closed by the time of onset of infection, all but 1 patient had either incisional erythema or drainage. Total white blood cell count at the onset of SSI was greater than 15 x 109/L in 14 of the 27 patients in whom it was recorded. Five patients had white blood cell count less than 15 x 109/L with greater than 80% polymorphonuclear cells (n = 2) or greater than 15% bands (n = 3) in the differential count. The other eight patients had a white blood cell count less than 15 x 109/L with 80% polymorphonuclear cells or less and 15% bands or less. Computerized tomography was performed in 7 patients, and all showed a possible or definite mediastinitis. Four of the 29 cases of mediastinitis occurred in patients with DSC after the original surgery. One patient had DSC with skin closure for 3 days after open intracardiac surgery and developed mediastinitis 6 days after placement of an epicardial pacemaker 25 days after sternal closure. The other three cases occurred while the sternum remained open on postoperative days 6, 8, and 19 (the third patient received an intervening cardiac transplant 10 days after a failed Fontan, and 9 days before mediastinitis occurred). Two of these patients had their open sternum covered with a sheet of rubber latex (Esmarch bandage), and the third patient was on extracorporeal membrane oxygenation when the mediastinitis developed. Four of the 218 patients with DSC and 18 of 1,824 patients with primary wound closure had mediastinitis (p = 0.28; odds ratio, 1.88; 95% confidence interval, 0.63 to 5.60). For children with DSC, mediastinitis occurred in 3 of 151 patients younger than 6 months of age and 1 of 67 patients 6 months of age or older at the time of surgery. Importantly, there have been no cases of mediastinitis in 178 cases with DSC (except the one mentioned above in which the mediastinitis was likely related to a subsequent pacemaker placement) since local practice changed to coverage of all open sternums with sutured skin rather than with an Esmarch bandage.
Microbiology
Results of cultures are shown in Table 1. Mediastinal infections were caused by S aureus (n = 21), Pseudomonas aeruginosa (n = 1), Enterobacter cloacae (n = 1), Enterococcus faecalis (n = 1), Candida albicans (n = 1), Mycoplasma hominis (n = 1), and mixed organisms (n = 3). The rate of bacteremia was higher with S aureus (17 of 21; 81%) than with other causative organisms (1 of 8; 13%; p = 0.001; odds ratio, 31.5; 95% confidence interval, 3.5 to 287). Bacteremia persisted for a median of 2.5 days, with the longest duration being 10 days in a patient with infective endocarditis.
Therapy
All patients underwent semiurgent operative debridement after the diagnosis of mediastinitis (usually within 24 hours of onset of infection). Postoperative continuous mediastinal irrigation was used in seven cases (all before 1997), and a mediastinal drain was used until drainage was minimal in the other 22 cases. Sternum was closed after debridement in all but the three cases in which the sternum remained open before debridement. Rotational muscle flaps were not used. Four patients were still being ventilated and receiving inotropic agents at the time of debridement, and data could not be obtained in 2 other patients as they underwent debridement in other centers. Fifteen of the remaining 23 patients required a new course of ventilation, and 8 required a new course of inotropic agents administration after debridement, usually for less than 24 hours. Excluding a patient who died 3 days after the diagnosis of mediastinitis, intravenous antimicrobial therapy was administered for 4 to 8 weeks through a tunneled central venous catheter (n = 17), a peripherally inserted central catheter (n = 4), a nontunneled central venous catheter (n = 4), a peripheral venous catheter (n = 3), or an unknown route (n = 1). Oral antibiotics were continued at discharge in 1 patient with infective endocarditis and prosthetic cardiac material.
Outcome
Sequelae of infection are as noted in Table 1. One patient died of an intracerebral aneurysm secondary to endocarditis 3 days after diagnosis of mediastinitis, and 2 patients died of multiorgan failure that may have been exacerbated by mediastinitis, although they had no evidence of ongoing infection at the time of death. There were no other permanent sequelae of mediastinitis.
|
Comment
|
|---|
This study describes 29 cases of mediastinitis after pediatric cardiac surgery during a 13.5-year period. Patients presented 5 to 27 days postoperatively. The incidence of 1.1% mediastinitis from January 1, 1991, to June 30, 2004, could be an underestimate as cases presenting more than 14 days postoperatively would have been missed from 1991 to 1995. However, complete data were available from September 1, 1996, to June 30, 2004, and the rate of 1.1% mediastinitis (23 of 2,042 cases) during this period is identical to the earlier period and is comparable to the rates of 1.0% [15] and 0.85% [13] in previous pediatric studies that used a similar definition of mediastinitis, and 2.1% [16] in a study that included patients with mediastinitis or isolated sternal infection. Because the majority of cases of mediastinitis develop within 2 weeks of sternotomy, active surveillance limited to that period maximizes surveillance efficiency.
It is possible the true rate of mediastinitis in patients with DSC in our center is even lower than the 1.1% in our study as the diagnosis of mediastinitis in 2 of the 4 patients who had DSC was based on polymicrobial growth from mediastinal cultures taken at the time of debridement for possible mediastinitis rather than visible purulent material in the mediastinum, and it is possible these cultures were contaminated. One of the other cases occurred 6 days after placement of an epicardial pacemaker 25 days after sternal closure, and it seems most likely the mediastinitis was related to the pacemaker placement rather than the original surgery. However, even when these 3 patients are included, the rate of mediastinitis was minimally increased in patients with DSC (odds ratio, 1.88, but with very wide confidence intervals). With the observed difference in the rate between the two groups being very small, the power of the current study is very low (0.26), and it is possible that a significant difference in risk was not observed because of the small number of cases.
The risk of mediastinitis with DSC is not clear from the literature. Multiple small studies reported only one superficial SSI with no deep infections in a total of 48 patients with DSC [17]. Table 2
shows the incidence of SSI in studies of 20 or more children with DSC, with a 4.1% rate of superficial SSI in 581 cases and a 1.9 % rate of mediastinitis in 856 cases. These rates are clearly sufficiently low that concerns about SSI and sepsis should not preclude use of DSC. It is not clear whether there is an absolute increase in the rate of SSI with DSC, as only one of these studies included a comparative group, with 1 of 55 patients with DSC and 0 of 29 patients with primary sternal closure having an SSI [17]. However, a recent abstract reported an odds ratio of 4.9 (95% confidence interval, 2.0 to 24.8; p = 0.05) for DSC in 17 children with SSI, but did not report how many patients with DSC were included in the study [26]. This relative risk is higher than in the current study, but perhaps the findings would have been different had all SSI (including superficial wound infections) been included in the current study. The low rate of infection in patients with DSC in the current study may be because skin closure is attained over the open sternum whenever possible, although similar low rates were attained in previous studies that used mainly surgical membranes [21, 24] or bovine pericardium [17] to cover the incision.
View this table:
[in this window]
[in a new window]
|
Table 2. Incidence of Surgical Site Infections in Case Series of 20 Children or More With Delayed Sternal Closure After Cardiac Surgery
|
|
Our current practice for DSC is to achieve primary skin closure while stenting the sternum open. The skin and subcutaneous layers are mobilized using cautery dissection in the plane immediately superficial to the pectoral fascia while retracting the skin edges with sharp hooks. Mobilization for 2 to 3 cm usually allows skin closure without undue tension on the skin edges (Fig 1). The sternum is stented open using a sterile plastic strut approximately 2 cm in length trimmed from a plastic chest tube (28F). The length of the tube is estimated on the basis of the prominence of the heart within the anterior mediastinum but is usually 2.5 to 3.5 cm in an infant. The ends of the tube are trimmed leaving short tabs, which are used to secure the tube to the sternal margins using 3-0 polypropylene suture (Fig 2). The skin layer is then brought together with continuous 4-0 or 5-0 Prolene (Ethicon, Somerville, NJ). When the infant's condition has improved, under sterile conditions the incision is reopened, the strut is removed, and the sternum is closed with wire in the usual fashion, usually in the intensive care unit.
View larger version (124K):
[in this window]
[in a new window]
|
Fig 2. The skin and subcutaneous layers are mobilized away from the pectoral fascia to allow tension-free skin closure over the stented sternum.
|
|
This study suggests that if primary sternal closure was attained, mediastinitis is unlikely in the absence of incisional erythema or drainage. The white blood cell count and differential were suggestive of bacterial infection in only 70% of cases. As in a previous study [27], the risk of bacteremia was much higher with S aureus SSI than with other organisms (p = 0.001). It seems likely that this occurs because of the virulent nature of this organism. As reported in the literature, computed tomography was sensitive for diagnosis of mediastinitis, but our methodology did not allow us to confirm the previous finding that specificity is low in the first 14 postoperative days [28].
There is agreement in the literature that all patients with mediastinitis require surgical drainage or debridement [3–5, 7, 9], but the details of surgical management remain controversial. There are no randomized trials evaluating the efficacy of postoperative continuous irrigation systems for mediastinitis in adults or children. A study in adults advocated routine use [12], but our results in conjunction with a more recent adult study showing improved outcome with vacuum drainage versus a continuous irrigation system [11] and a pediatric study showing excellent results with a brief period of closed drainage [13] suggest continuous irrigation is not necessary. Our study methodology did not allow us to assess the long-term outcome as many patients receive follow-up in other centers, but all patients had a satisfactory cosmetic result at the time of discharge despite the fact no rotational muscle flaps were used, and we are aware of only 1 patient who required further surgery as a result of sequelae of mediastinitis (malunion of the sternum). Previous studies advocated use of muscle flaps in adult cases that occur more than 14 days postoperatively [15], in neonates [29], and in almost all pediatric cases [14, 15], but a recent pediatric study described excellent outcome with primary sternal closure after debridement and minimal use of muscle flaps [13].
In conclusion, it appears that the increased risk of mediastinitis with DSC is minimal, especially with primary skin closure. Debridement followed by primary closure yields excellent results in the vast majority of cases of mediastinitis.
|
Acknowledgments
|
|---|
The authors would like to acknowledge the efforts of hospital infection control practitioners who performed the surveillance necessary for this study.
|
References
|
|---|
- Garner JS, Jarvis WR, Emori TG, et al. CDC definitions for nosocomial infections, 1988 Am J Infect Control 1988;16:128-140.[Medline]
- Ehrenkranz NJ, Pfaff SJ. Mediastinitis complicating cardiac operationsevidence of postoperative causation. Rev Infect Dis 1991;13:803-814.[Medline]
- The Parisian mediastinitis study group Risk factors for deep sternal wound infection after sternotomya prospective, multicenter study. J Thorac Cardiovasc Surg 1996;111:1201-1207.
- Ridderstrope L, Gill H, Granfeldt H, et al. Superficial and deep sternal wound complicationsincidence, risk factors and mortality. Eur J Cardiothoac Surg 2001;20:1168-1175.
- Bitkover C, Gardlund B. Mediastinitis after cardiovascular operationsa case-control study of risk factors. Ann Thorac Surg 1998;65:36-40.[Abstract/Free Full Text]
- Brady M, McGlone C, Vinsel J, Hayes J. Sternal wound infections following cardiac surgery in infants [Abstract J158] Presented at the 36th Interscience Conference on Antimicrobial Agents and Chemotherapy. 1998.
- Macanally H, Cutter G, Rutterben A, et al. Hypothermia as a risk factor for pediatric cardiothoracic surgical site infection Pediatr Infect Dis J 2001;20:459-462.[Medline]
- Allpress AL, Rosenthal GL, Goodrich KM, Zerr DM. Risk factors for surgical site infections following pediatric cardiovascular surgery [Abstract 407]Chicago: Presented at the 40th Infectious Diseases Society of American Annual Meeting; 2002Oct 24–27.
- Mehta P, Cunningham C, Colella C, et al. Risk factors for sternal wound and other infections in pediatric cardiac surgery patients Pediatr Infect Dis J 2000;19:1000-1004.[Medline]
- Nateghian A, Taylor G, Robinson J. Risk factors for surgical site infections following open heart surgery in a Canadian pediatric population Am J Infect Control 2004;32:397-401.[Medline]
- Berg HG, Brands WG, van Geldopr TR, Kluytmans-VandenBergh FQ, Kluytmans JA. Comparison between closed drainage techniques for the treatment of postoperative mediastinitis Ann Thorac Surg 2000;70:924-929.[Abstract/Free Full Text]
- Verkkala K, Jarvinen A. Mediastinal infection following open-heart surgeryTreatment with retrosternal irrigation. Scand J Thorac Cardiovasc Surg 1986;20:203-207.[Medline]
- Ohye RG, Maniker RB, Graves HL, Devaney EF, Bove EL. Primary closure for postoperative mediastinitis in children J Thorac Cardiovasc Surg 2004;128:480-486.[Abstract/Free Full Text]
- Tortoreillo TA, Friedman JD, McKenzie ED, et al. Mediastinitis after pediatric cardiac surgerya 15-year experience at a single institution. Ann Thorac Surg 2003;76:1655-1660.[Abstract/Free Full Text]
- Sung K, Jun T, Park PW, Park K, Lee YT, Yang J. Management of deep sternal infections in infants and children with abdominal pectoralis major muscle flaps Ann Thorac Surg 2004;77:1371-1375.[Abstract/Free Full Text]
- Levy I, Ovadia B, Erez E, et al. Nosocomial infections after cardiac surgery in infants and childrenincidence and risk factors. J Hosp Infect 2003;53:111-116.[Medline]
- Hakimi M, Walters 3rd HL, Pinsky WW, Gallagher MJ, Lyons JM. Delayed sternal closure after neonatal cardiac operations J Thorac Cardiovasc Surg 1994;107:925-933.[Abstract/Free Full Text]
- Ziemer G, Karck M, Muller H, Luhmer I. Staged chest closure in pediatric cardiac surgery preventing typical and atypical cardiac tamponade Eur J Cardiothorac Surg 1992;6:91-95.[Abstract]
- Korbmacher B, Rammos S. Two-stage sternal closure after repair of congenital heart disease in neonates Cardiovasc Surg 1993;1:660-663.[Medline]
- Elami A, Permut LC, Laks H, Drinkwater Jr DC, Sebastian JL. Cardiac decompression after operation for congenital heart disease in infancy Ann Thorac Surg 1994;58:1392-1396.[Abstract]
- Alexi-Meskishvili V, Weng Y, Uhlemann F, Lange PE, Hetzer R. Prolonged open sternotomy after pediatric open heart operationexperience with 113 patients. Ann Thorac Surg 1995;59:379-383.[Abstract/Free Full Text]
- Tabbutt S, Duncan BW, McLaughlin D, Wessel DL, Jonas RA, Laussen PC. Delayed sternal closure after cardiac operations in a pediatric population J Thorac Cardiovasc Surg 1997;113:886-893.[Abstract/Free Full Text]
- Iyer RS, Jacobs JP, de Leval MR, Stark J, Elliott MJ. Outcomes after delayed sternal closure in pediatric heart operationsa 10-year experience. Ann Thorac Surg 1997;63:489-491.[Abstract/Free Full Text]
- McElhinney DB, Reddy VM, Parry AJ, Johnson L, Fineman JR, Hanley FL. Management and outcomes of delayed sternal closure after cardiac surgery in neonates and infants Crit Care Med 2000;28:1180-1184.[Medline]
- Owens WA, Vitale N, Hasan A, Hamilton JR. A policy of elective delayed sternal closure does not improve the outcome after arterial switch Ann Thorac Surg 2001;71:1553-1555.[Abstract/Free Full Text]
- Holzmann-Pazgal G, Huddleston C, Grim A, et al. Pediatric cardiothoracic surgical site infectionsa case control study [Abstract 270]. Boston: Presented at the 42nd Annual Meeting of Infectious Diseases Society of American Annual Meeting; 2004.
- Shah SS, Lautenbach E, Long CB, et al. Risk factors for bloodstream infections in children with post-operative mediastinitisrole of S. aureus [Abstract 255]. Boston: Presented at the 42nd Annual Meeting of Infectious Diseases Society of American Annual Meeting; 2004.
- Jolles H, Henry DA, Roberson JP, Cole TJ, Spratt JA. Mediastinitis following median sternotomyCT findings. Radiology 1996;201:463-466.[Abstract/Free Full Text]
- Erez E, Katz M, Sharoni E, et al. Pectoral major muscle flap for deep sternal wound infection in neonates Ann Thorac Surg 2000;69:572-577.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
A. M. Gaca, J. J. Jaggers, L. T. Dudley, and G. S. Bisset III
Repair of Congenital Heart Disease: A Primer-Part 1
Radiology,
June 1, 2008;
247(3):
617 - 631.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
T. Kadohama, N. Akasaka, A. Nagamine, K. Nakanishi, K. Kiyokawa, K. Goh, and T. Sasajima
Vacuum-Assisted Closure for Pediatric Post-Sternotomy Mediastinitis: Are Low Negative Pressures Sufficient?
Ann. Thorac. Surg.,
March 1, 2008;
85(3):
1094 - 1096.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
C. Anslot, S. Hulin, and Y. Durandy
Postoperative Mediastinitis in Children: Improvement of Simple Primary Closed Drainage
Ann. Thorac. Surg.,
August 1, 2007;
84(2):
423 - 428.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
