Spontaneous bacterial empyema in cirrhotic patients: A prospective study

Spontaneous Bacterial Empyema in Cirrhotic Patients: A Prospective Study XAVIER XIOL, JOSEP M. CASTELLVI´, JORDI GUARDIOLA, EVA SESE´, JOSE´ CASTELLOTE, ANTONIA PERELLO´, XAVIER CERVANTES, AND MARIA JESU´S IBORRA

Spontaneous bacterial empyema (SBEM) is an infection of a preexisting hydrothorax in cirrhotic patients and has seldom been reported. To determine its incidence and primary characteristics, all cirrhotic patients with pleural effusion underwent thoracentesis at our hospital either on admission or when an infection was suspected. Pleural fluid (PF) study included biochemical analysis, polymorphonuclear (PMN) leukocyte count, and culture by two methods: conventional and modified (inoculation of 10 mL of PF into a blood culture bottle at the bedside). SBEM was defined according to previously reported criteria: PF culture positive or PMN count greater than 500 cells/mL, and exclusion of parapneumonic effusions. Sixteen of the 120 (13%) cirrhotic patients admitted with hydrothorax had 24 episodes of SBEM. In 10 of the 24 episodes (43%), SBEM was not associated with spontaneous bacterial peritonitis (SBP). PF culture was positive by the conventional method in 8 episodes (33%) and by the modified method (blood culture inoculation) in 18 (75%) (P Å .004, McNemar). The microorganisms identified in PF were Escherichia coli in 8 episodes, Streptococcus species in 4, Enterococcus species in 3, Klebsiella pneumoniae in 2, and Pseudomonas stutzeri in 1. All episodes were treated with antibiotics without inserting a chest tube in any case. Mortality during treatment was 20%. We conclude that SBEM is a common complication of cirrhotic patients with hydrothorax. Almost half of the episodes were not associated with SBP; thus, thoracentesis should be performed in patients with cirrhosis, pleural effusion, and suspected infection. Culture of PF should be performed by inoculating 10 mL into a blood culture bottle at the bedside. (HEPATOLOGY 1996;23:719-723.)

Abbreviations: SBP, spontaneous bacterial peritonitis; SBEM, spontaneous bacterial empyema; PF, pleural fluid; PMN, polymorphonuclear; AF, ascitic fluid; OLT, orthotopic liver transplantation. From the Gastroenterology Service, Hospital de Bellvitge, 08907 L’Hospitalet de Llobregat, Barcelona, Spain. Received March 27, 1995; accepted October 21, 1995. Dr. Guardiola is now at the Hospital de Vilafranca, Sant Pere 4, Vilafranca, Barcelona, Spain. Dr. Castellote is now at the Hospital General de Manresa, Ctra de la Culla s/n, Manresa, Barcelona, Spain. Address reprint requests to: Xavier Xiol, M.D., Servicio Aparato Digestivo, Pta 19, Hospital de Bellvitge, Feixa Llarga s/n, 08907 L’Hospitalet de Llobregat, Barcelona, Spain. Copyright q 1996 by the American Association for the Study of Liver Diseases. 0270-9139/96/2304-0010$3.00/0

Whereas spontaneous bacterial peritonitis (SBP) is a well-known entity with a reported incidence between 15% and 20% in hospitalized cirrhotic patients with ascites,1-3 spontaneous bacterial empyema (SBEM)— the infection of a preexisting hydrothorax—has seldom been reported.4 Because 5% to 10% of cirrhotic patients with ascites have an associated hydrothorax,5,6 SBEM could be expected to appear in 1% to 2% of hospitalized cirrhotic patients with ascites. Apart from case reports, only two retrospective series including a total of 15 episodes have been published.4,7 The aim of this study was to investigate incidence, bacteriology, and clinical characteristics of SBEM and to confirm the data obtained in our previous retrospective study.4 PATIENTS AND METHODS In a university-based reference hospital, from September 1988 to December 1992, a thoracentesis was performed on all cirrhotic patients with pleural effusion on admission (or when the effusion was detected for the first time during hospitalization) or when an infection was suspected during admission because of fever, abdominal or chest pain, hepatic encephalopathy, or shock. If ascites was present, a paracentesis also was performed at the same time. Pleural fluid (PF) study included bacteriologic study, cytology, polymorphonuclear (PMN) leukocyte count, and glucose, protein, amylase, lactic dehydrogenase, and adenosine deaminase determinations. pH also was performed if an infection was suspected. The bacteriologic study was performed using two different methods: conventional and modified.8 In the conventional method, a sample of PF was collected in an empty sterile container and sent to the Clinical Microbiology Laboratory. Ten milliliters was centrifuged at 3,000 rpm for 20 minutes. The sediment was cultured on enriched chocolate agar, blood agar, MacConkey’s agar, and thioglycolate broth. The cultures were incubated at 357C in a CO2 (10%-15%) incubator for 48 hours. Both plates and broth were examined at 24 and 48 hours for visible growth. In the modified method, 10 mL of PF was inoculated into a 70-mL ‘‘Liquoid’’ Blood Culture TSB Roche (Hoffman-La Roche, Basel, Switzerland) at the patient’s bedside. It was incubated at 357C for 24 hours. Then, the BCB Roche slide (Hoffman-La Roche) with three solid culture mediums (chocolate agar, MacConkey’s agar, and malt agar) was screwed to the culture bottle, and the agar surfaces of the culture mediums were flooded with the broth of the culture bottle. The assembly was incubated at 357C for a further 24 hours. In both cases, the organism was identified with routine laboratory methods if growth occurred.9 Antibiotic susceptibility was studied using standard agar dilution

719

5p0c$$0014

03-20-96 08:27:42

hepa

WBS: Hepatology

720 XIOL ET AL.

HEPATOLOGY April 1996

methods.10 If there was no growth by 48 hours, the culture was considered negative. In patients with ascitic fluid (AF), AF culture was performed only by the modified method. The diagnosis of SBEM was established according to previously reported criteria4: 1. Positive PF culture and a PMN count greater than 250 cells/mL. Patients with negative culture, compatible clinical course, and a PF PMN count ú500 cells/mL also were included as culture-negative SBEM; 2. Exclusion of parapneumonic infections11; a. no image of pneumonia on a chest radiograph or computed tomography scan; b. evidence of pleural effusion before the infectious episode or PF transudate characteristics during infection12; 3. Patients treated with esophageal variceal sclerotherapy in the previous week or who tested seropositive for the human immunodeficiency virus were excluded from the study. SBP was defined13 as a positive AF culture plus an AF PMN count greater than 250 cells/mL. Culture-negative SBP was diagnosed by an AF PMN count greater than 500 cells/ mL with neither bacterial growth in culture methods, nor carcinomatosis, tuberculosis, or pancreatic ascites. The absence of an intraabdominal source of infection was required in both cases. When an infection was suspected, patients were treated empirically with a third-generation cephalosporin (ceftriaxone 1 g/24 hr) after drawing blood cultures, PF, and AF. Once the results of cultures and antibiotic sensitivities were known, appropriate changes in treatment were made. In patients with confirmed SBEM, a control thoracentesis was performed after 7 to 10 days of antibiotic treatment. SBEM was considered to be healed when PF culture became negative and the PF PMN count was less than 250 cells/mL. Criteria used in our hospital for chest tube insertion are frank pus or pH õ7.1 plus glucose levels õ40 mg/100 mL.11 McNemar’s test, Fisher’s Exact Test, and ANOVA were used for statistical analysis when necessary. Results are expressed as mean { SEM and were considered significant at P õ .05. RESULTS

During the study period, 120 cirrhotic patients with hydrothorax were admitted to our unit; 95 (79%) had detectable ascites in addition to pleural effusion. Sixteen of the 120 (13%) had 24 episodes of SBEM. One patient had five episodes, 1 had three episodes, 2 patients had two episodes, and 12 had one episode. Clinical data of patients are presented in Table 1. All patients had advanced cirrhosis and most had been hospitalized on previous occasions with clinical signs of progressive liver dysfunction. Abdominal pain, one of the most common symptoms, was caused by an associated SBP in all cases. Ascites was absent in 6 of 24 episodes of SBEM (25%). In 18 episodes, ascites was present; AF was infected in 14 and noninfected in 4. In 2 of the 4 patients with noninfected ascites, a PMN count was not performed, because the paracentesis yielded a few drops of AF, which was only enough to practice a culture that was negative (Table 1, episodes 1a and 1d). Therefore, in 10 of 24 episodes (43%), SBEM was not associated with SBP (6 without ascites and 4 with noninfected ascites). Blood cultures were positive

5p0c$$0014

03-20-96 08:27:42

hepa

in 11 of 24 episodes, 4 of 10 patients (40%) without SBP and in 7 of 14 patients with SBP being positive (50%) (not significant). PF culture was positive by the conventional method in 8 episodes (33%), whereas it was positive by the modified method in 18 cases (75%) (P Å .004) (Table 2). All cases with positive conventional culture also had positive inoculated culture. The microorganisms identified in PF were Escherichia coli in 8 episodes, Streptococcus species in 4, Enterococcus species in 3, Klebsiella pneumoniae in 2, and Pseudomonas stutzeri in 1. Despite improving culture technique, 6 patients had a culture-negative SBEM. The etiologic diagnosis was confirmed in 4 of 6 culture-negative SBEM, 2 by AF culture, 1 by blood culture, and 1 by both ascitic and blood cultures (Table 1). PF characteristics before, during, and after infection appear in Table 3. PF LDH during infection (4.70 { 0.77 mkat/L) was significantly higher than LDH before infection (2.46 { 0.37 mkat/L) and after infection (2.57 { 0.37 mkat/L). However, there were no changes in glucose and protein levels before, during, or after infection. A chest tube was not placed in any case, because no PF fulfilled the biochemical criteria required for its insertion. In one patient, a pneumothorax appeared after a control thoracentesis, when the infection was already cured. A chest tube was inserted for 7 days, and the patient was discharged a few days later. The mortality rate during treatment was 20% (5 of 24). The cause of death was septic shock within 48 hours after starting treatment in 2 cases, esophageal variceal hemorrhage in 2, and hepatic insufficiency in 1. In the remainder, the infection was cured after 7 to 10 days of antibiotic treatment. Two other patients died during admission because of hepatic insufficiency, in both cases several days after having finished antibiotic treatment. Four of the 10 patients who were discharged died between 1 and 19 months after an SBEM episode. In 5 patients, an orthotopic liver transplantation (OLT) was performed between 2 and 24 months after SBEM (Fig. 1). All 5 patients who underwent transplantation are presently alive (2-5 years after OLT). DISCUSSION

This study confirms that SBEM is a frequent complication of cirrhotic patients with hydrothorax; its 13% incidence is similar to the reported SBP incidence in cirrhotic patients with ascites.1-3,13 We believe that SBEM is rarely diagnosed, not only because patients with hydrothorax are unusual, but because thoracenteses are not performed routinely in cirrhotic patients with hydrothorax. In fact, pleural effusion is not as obvious as ascites, and thoracentesis is difficult to perform in comatose patients. We had few complications related to the practice of diagnostic thoracentesis; thus, we think that this procedure may be as safe and efficient as paracentesis.14 As shown in Table 1, there are patients having a culture-positive SBEM associated with a culture-negative SBP, and vice versa, underlin-

WBS: Hepatology

5p0c$$0014

03-20-96 08:27:42

hepa

WBS: Hepatology

39

38

39 38 37.5

37.7

37 39 38.8 37.5

10

11a

11b 11c 12

13

14a 14b 15 16

IH IH IH IH

EH

EH IH IH

EH

EH

IH

IH

IH

EH EH

EH

EH

EH

IH

IH

EH IH IH IH

EH/IH

Abdominal pain Septic shock

Thoracic pain

Thoracic pain, cough Thoracic pain, dyspnea Thoracic pain Dyspnea Thoracic pain, dyspnea Abdominal pain

Encephalopathy, dyspnea Abdominal pain

Encephalopathy Abdominal pain, thoracic pain, dyspnea Abdominal pain

Chills Abdominal pain Abdominal pain Thoracic pain, cough, dyspnea Thoracic pain, dyspnea Abdominal pain, cough Abdominal pain, cough, dyspnea Abdominal pain, thoracic pain, dyspnea Encephalopathy

Clinical Findings

9 9 12 11

12

10 11 11

10

8

12

11

12

13 12

11

10

11

10

9

10 11 10 11

Pugh

No Yes No No

No

No No No

No

No

No

No

No

Yes No

Yes

No

Yes

Yes

No

No No No No

PVS

Negative Negative Negative E. faecium

Negative

E. coli Negative Negative

E. coli

Negative

Negative

Negative

Negative

Negative Negative

E. coli

Negative

S. sanguis

E. coli

S. bovis

Negative Negative S. bovis Negative

Modified Culture

Negative E. faecalis K. pneumoniae E. faecium

Negative

E. coli P. stutzeri Negative

E. coli

S. mitis

E. coli

Negative

S. bovis

E. coli Negative

E. coli

E. coli

S. sanguis

E. coli

S. bovis

Negative E. coli S. bovis K. pneumoniae

PF

Abbreviations: EH, extrahospitalary; IH, intrahospitalary; PVS, peritoneovenous shunt. * No AF present.

36.5

37.5 39

5 6

9

38.5

4

36.8

39.4

3

8

37.8

2b

38

39

2a

7

38

37.5 38.1 37 38

1a 1b 1c 1d

1e

T

Patient

Conventional Culture

3,404 4,680 1,856 2,890

1,600

6,674 1,600 2,574

7,480

13,366

560

826

3,026

1,500 530

2,250

11,440

3,404

7,905

2,950

3,000 3,895 8,500 1,700

PMN

*

*

E. faecalis E. faecalis K. pneumoniae *

Negative

* * Negative

*

Negative

Citrobacter freundii E. coli

S. bovis

E. coli

E. coli

S. sanguis

E. coli

Negative

Negative Negative S. bovis Negative

Culture

AF

3,540 528 1,240 *

1,850

* * 2,180

*

206

4,900

2,760

3,450

* 7,860

*

21,712

4,185

10,545

ND

ND 1,575 34,265 128

PMN

Yes Yes Yes No

Yes

No No Yes

No

No

Yes

Yes

Yes

No Yes

No

Yes

Yes

Yes

No

No Yes Yes No

SBP

Negative Negative Negative E. faecium

E. coli

Negative Negative Negative

Negative

Negative

Negative

Negative

Negative

E. coli E. coli

E. coli

E. coli

S. sanguis

E. coli

Negative

Negative E. coli S. bovis K. pneumoniae

Blood Cultures

TABLE 1. Clinical Characteristics in the 24 Episodes of SBEM in 16 Patients

Ampicillin Ampicillin Ceftriaxone Ceftriaxone

Ampicillin

Ceftriaxone Ciprofloxacin Ceftriaxone

Ceftriaxone

Ceftriaxone

Ceftriaxone

Ceftriaxone

Ceftriaxone

Ceftriaxone Ceftriaxone

Ceftriaxone

Ceftriaxone

Ceftriaxone

Ceftriaxone

Ceftriaxone

Ceftriaxone Ceftriaxone Ceftriaxone Ceftriaxone

Antibiotic

Reinfected Reinfected Reinfected Reinfected

5 2 1 2

mo mo mo mo

later later later later

Cured. Died of hepatic failure 2 mo later Cured. Reinfected 1.5 mo later Cured. OLT 16 mo later Cured. Died of SBP 3 mo later Died of sepsis on day 1

Cured. Reinfected 2 mo later Cured. OLT 2 mo later Cured. OLT 2 mo later

Cured. Reinfected 1.5 mo later

Cured. Died of hepatic failure 7 mo later Cured. OLT 10 mo later

Died of bleeding esophageal varices on day 2 Died of hepatic failure on day 8

Cured. Died of hepatic failure 3 mo later (during admission) Died of sepsis on day 2 Cured. Died of hepatic failure 10 mo later

Cured. Died of hepatic failure 1 mo later (during admission) Died of bleeding esophageal varices on day 5

Cured. Reinfected 7 mo later

Cured. OLT 1 mo later

Cured. Cured. Cured. Cured.

Outcome

HEPATOLOGY Vol. 23, No. 4, 1996 XIOL ET AL.

721

722 XIOL ET AL.

HEPATOLOGY April 1996

TABLE 2. Comparison of Two PF Culture Methods in 24 Episodes of SBEM No. of Isolates Microorganisms

Conventional

Modified

E. coli Streptococcus species Enterococcus species K. pneumoniae P. stutzeri Negative* Total

4 3 1 0 0 16 24

8 5 2 2 1 6 24

* All microorganisms isolated by the conventional method were also isolated by the modified method.

ing the importance of performing a thoracentesis in cirrhotic patients with pleural effusion and suspected infection. Because of our strategy of practicing thoracentesis in addition to paracentesis and blood cultures, an etiologic diagnosis was obtained in 22 of the 24 infections of the present series. In contrast with our previous report,4 40% of SBEM episodes were not associated with SBP, although the criteria used for SBEM diagnosis were the same. The difference could be because the previous study was retrospective. Six of 24 patients in the present series had hydrothorax without ascites, indicating that ascites is

not a prerequisite for SBEM, and supporting the hypothesis that enteric microorganisms reach the PF through a bacteremia, as has been reported in SBP.13,15 This is also suggested by the fact that blood cultures were positive in 4 of 10 cases without SBP. The inoculation of 10 mL of PF into a TSB blood bottle at the patient’s bedside had a significantly higher sensitivity for the diagnosis of SBEM than the conventional method (77% vs. 33%, P Å .004). This is in accordance with previous SBP reports.16,17 Analogous results in AF have been described in our hospital8 (77% vs. 57%, P Å .0001). The reason for improving sensitivity should be immediate inoculation,18 since a TSB blood culture bottle contains an anticoagulant and opsonin inhibitor that protects bacteria from further complement- or phagocyte-mediate killing,19 as well as protecting the cultured volume.20 The modified method allows the culture of a large volume of fluid, and SBEM is probably an infection that involves a low concentration of bacteria, as is SBP.20 None of the patients in our series were treated with a chest tube because PF did not meet the biochemical criteria required for its insertion (frank pus, or pH õ7.1 and glucose õ2.22 mmol/L [40 mg/L]). Because most of our patients were cured without inserting a chest tube, and because its insertion in cirrhotic patients with hydrothorax can be harmful,21 a chest tube should not be used in the treatment of SBEM. The biochemical

TABLE 3. PF Characteristics Before, During, and After SBEM Before SBEM Patient

1a 1b 1c 1d 1e 2a 2b 3 4 5 6 7 8 9 10 11a 11b 11c 12 13 14a 14b 15 16

During SBEM

Previous Effusion/ Thoracentesis

LDH ( mkat/L)

Protein (g/L)

PMN

LDH

Protein

Glucose (mmol/L)

Yes/Yes

2.0 1.2 0.7 0.6 1.3

23 4 2 4 4

4 50 100 100 32

2.2 1 3

13 8 18

12 76 24

1.6 1.3 1.4 1.1 1.0 5.5 5.8 3.9 11.4 4.0 4.2 4.5 1.0 1.9 18.5 4.9 6.8 4.3 4.4 5.4 2.8 5.7 5.9 5.4

7 3 2 6 5 12 13 15 29 19 17 6 3 15 17 22 37 25 5 17 23 43 21 15

1.0 7.0 3.7 8.9 8.2 10.7 10.2 22.3 6.3 4.5 5.8 19.1 12.1 9.1 3.8 4.9 4.7 5.3 8.8 12.3 5.3 3.7 7.5 5.4

Yes/No Yes/Yes Yes/Yes Yes/No No/No Yes/No Yes/Yes Yes/No Yes/Yes Yes/No

Yes/Yes No/No Yes/Yes Yes/Yes Yes/Yes

0.5

2

40

1.7

12

10

4.4

19

8

5.2

9

180

2.2 2.1 2.6 3.6

19 16 19 20

24 70 66 170

Abbreviation: LDH, lactic dehydrogenase. * Died before completing treatment. † No PF available at the end of treatment.

5p0c$$0014

03-20-96 08:27:42

hepa

WBS: Hepatology

After SBEM pH

PMN

LDH

Protein

PMN

7.30 7.35 7.04 7.26 7.40 7.30 7.10 7.26 7.32 7.50 7.42 7.40 7.34 7.42 7.19 7.26 7.24 7.24 7.32 7.30

3,800 3,895 8,500 1,700 2,950 7,905 3,404 11,440 2,250 1,500 530 3,026 826 560 13,366 7,480 6,674 1,600 2,574 1,600 3,404 4,680 1,856 2,890

1.2 1.3 0.6 0.6 1.3 2.2 1.9 * 5.2 * 1.1 * 0.6 4.9 2.9 4.3 5.6 4.2 2.4 3.2 2.1 † 3.2 *

5 7 4 2 5 13 12

144 20 100 17 116 16 225

20

3

15

4

3 24 13 19 30 16 3 10 16 † 18

240 10 10 8 48 10 24 0 70 † 144

7.30 7.45

HEPATOLOGY Vol. 23, No. 4, 1996

XIOL ET AL.

FIG. 1. Long-term evolution of the 16 patients who had an episode of SBEM.

characteristics of PF found in the present series were similar to those reported previously4 and consist of an increase in pleural LDH during infection. SBP is a recognized indication for OLT.22 Because of SBP’s similarities between and frequent association with SBP, we have indicated OLT in five patients with SBEM. Three of these patients experienced SBP (in addition to SBEM), and two of them had suffered SBEM without SBP. In the other patients who survived the infection, this procedure was contraindicated because of age, active enolism, or portal thrombosis. In conclusion, SBEM is a frequent complication in cirrhotic patients with hydrothorax. Our study shows that almost half of the cases of SBEM are not associated with SBP, indicating the importance of performing thoracentesis in patients with cirrhosis, pleural effusion, and signs or symptoms of infection, independently of their having ascites. PF culture should be performed by inoculating 10 mL into a blood culture bottle at the bedside, because this method is more sensitive than the conventional method. A chest tube is not even necessary in patients with positive PF culture. SBEM should be considered an indication for liver transplantation, independently of SBP. Acknowledgment: We thank Catalina Perello´ for useful linguistic correction and Dr. Josep Sol for statistical assistance. REFERENCES 1. Conn HO, Fessel JM. Spontaneous bacterial peritonitis in cirrhosis: variations on a theme. Medicine 1971;50:161-197.

5p0c$$0014

03-20-96 08:27:42

hepa

723

2. Almdal TP, Skinhoj P. Spontaneous bacterial peritonitis in cirrhosis: incidence, diagnosis and prognosis. Scand J Gastroenterol 1987;22:295-300. 3. Albillos A, Cuervas-Mons V, Millan I, Canton T, Montes J, Barrios C, Garrido A, et al. Ascitic fluid polymorphonuclear cell count and serum to ascites albumin gradient in the diagnosis of bacterial peritonitis. Gastroenterology 1990;98:134-140. 4. Xiol X, Castellote J, Baliellas C, Ariza J, Gimenez A, Guardiola J, Casais L. Spontaneous bacterial empyema in cirrhotic patients: Analysis of eleven cases. HEPATOLOGY 1990;11:365-370. 5. Esteve M, Xiol X, Fernandez F, Gonzalez-Huix F, Baliellas C, Casais L. Treatment and outcome of hepatic hydrothorax in liver cirrhosis. J Clin Nutr Gastroenterol 1986;1:139-144. 6. Alberts WM, Salomon DA, Boyce G. Hepatic hydrothorax. Cause and management. Arch Intern Med 1991;151:2383-2388. 7. Chesta J, Ponichik J, Brahm J, Gil R, Gil LC, Ruiz M, Latorre R, et al. Spontaneous bacterial pleuritis in patients with liver cirrhosis. Rev Med Chil 1991;119:295-298. 8. Castellote J, Xiol X, Verdaguer J, Ribes J, Guardiola J, Gimenez A, Casais L. Comparison of two ascitic fluid culture methods in cirrhotic patients with spontaneous bacterial peritonitis. Am J Gastroenterol 1990;85:1605-1608. 9. Martin WJ, Washington JA II. Enterobacteriaceae. In: Lenette EH, Ballows A, Hausler VJ, eds. Manual of clinical microbiology. 3rd ed. Washington, DC: American Society for Microbiology; 1980:195-219. 10. Washington JA II, Sutter VL. Dilution susceptibility test: Agar and macrobroth dilution procedures. In: Lenette EH, Ballows A, Hausler VJ, eds. Manual of clinical microbiology. 3rd ed. Washington, DC: American Society for Microbiology; 1980:453-458. 11. Sahn SA. The pleura. Am Rev Respir Dis 1988;138:184-234. 12. Light RW, McGregor MI, Luchsinger PV, Ball WC. Pleural effusions: the diagnostic separation of transudates and exudates. Ann Intern Med 1972;77:507-513. 13. Garcia Tsao G. Spontaneous bacterial peritonitis. Gastroenterol Clin North Am 1992;21:257-275. 14. Runyon BA. Paracentesis of ascitic fluid. A safe procedure. Arch Intern Med 1986;146:2259-2261. 15. Wyke RJ. Bacterial infections complicating liver disease. Baillieres Clin Gastroenterol 1989;3:187-210. 16. Runyon BA, Umland ET, Merlin T. Inoculation of blood cultures with ascitic fluid. Improved detection of spontaneous bacterial peritonitis. Arch Intern Med 1987;147:73-75. 17. Bobadilla M, Sifuentes J, Garcia Tsao G. Improved method for bacteriological diagnosis of spontaneous bacterial peritonitis. J Clin Microbiol 1989;27:2145-2147. 18. Runyon BA, Antillon MR, Akriviadis EA, McHutchison JG. Bedside inoculation of blood culture bottles with ascitic fluid is superior to delayed inoculation in the detection of spontaneous bacterial peritonitis. J Clin Microbiol 1990;28:2811-2812. 19. Washington JA II. Cultures of normally sterile body fluids, tissue wounds and abscesses. In: Washington JA II, Brewer NS, eds. Laboratory procedures in clinical microbiology. New York: Springer-Verlag; 1985:97-103. 20. Runyon BA, Canawati HN, Akriviadis EA. Optimization of ascitic fluid culture technique. Gastroenterology 1988;95:13511355. 21. Runyon BA, Greenblatt M, Ming HC. Hepatic hydrothorax is a relative contraindication to chest tube insertion. Am J Gastroenterol 1986;81:566-567. 22. Tito L, Rimola A, Gines P, Llach J, Arroyo V, Rodes J. Recurrence of spontaneous bacterial peritonitis in cirrhosis: Frequency and predictive factors. HEPATOLOGY 1988;8:27-31.

WBS: Hepatology