Chest ACCP Career Connection
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Guest Access | Sign In via User Name/Password
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Article Archive
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (2)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Barnes, T. W.
Right arrow Articles by Ryu, J. H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Barnes, T. W.
Right arrow Articles by Ryu, J. H.
(Chest. 2005;127:916-921.)
© 2005 American College of Chest Physicians

Low Yield of Microbiologic Studies on Pleural Fluid Specimens*

Terrance W. Barnes, MD; Eric J. Olson, MD, FCCP; Timothy I. Morgenthaler, MD, FCCP; Randall S. Edson, MD; Paul A. Decker, MS and Jay H. Ryu, MD, FCCP

* From the Divisions of Pulmonary and Critical Care Medicine (Drs. Barnes, Olson, Morgenthaler, and Ryu), Infectious Diseases (Dr. Edson), and Biostatistics (Mr. Decker), Mayo Clinic, Rochester, MN.

Correspondence to: Jay H. Ryu, MD, FCCP, Division of Pulmonary and Critical Care Medicine, Desk East 18, Mayo Clinic, 200 First St SW, Rochester, MN 55905; e-mail: ryu.jay{at}mayo.edu


   Abstract
TOP
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Background: It is generally recommended that pleural fluid samples from pleural effusions of unknown cause be cultured for bacteria, mycobacteria, and fungi. However, the utility of this practice has been not been adequately assessed.

Design: Retrospective review.

Setting: Tertiary care, referral medical center.

Patients: Five hundred twenty-five patients undergoing diagnostic thoracentesis at Mayo Medical Center, Rochester, MN, over a 12-month period from July 1, 2001, to June 30, 2002.

Interventions: None.

Measurements and results: Among 525 patients undergoing diagnostic thoracenteses, 476 patients (91%) had one or more cultures performed on their pleural fluid specimens. Thirty-nine positive results (3.0% of 1,320 cultures) occurred in 35 of these 476 patients (7.4%). After excluding likely contaminants, true pathogens were identified in only 19 of 1,320 pleural fluid cultures (1.4%) belonging to 15 patients (3.2% of those who had cultures performed on their pleural fluid specimen). These positive results included 2.3% of aerobic bacterial, 1.2% of anaerobic bacterial, 1.4% of fungal, and 0% of mycobacterial cultures. Microbiologic smears performed on these pleural fluid samples included 357 Gram stains, 109 fungal smears (potassium hydroxide), and 232 acid-fast smears with positive yields of 2.5%, 0%, and 0%, respectively. These positive findings represented 1.3% of all smears performed. Of the specimens obtained from outpatient thoracenteses, only one had a true-positive result (0.8%). Only 1.1% (four specimens) of the cultures performed on free-flowing effusions demonstrated true pathogens; three of these four specimens grew fungi.

Conclusions: The positive yield of microbiologic smears and cultures on pleural fluid specimens is low, particularly in the outpatient setting and in patients with free-flowing effusions. Microbiologic testing of pleural fluid specimens should be ordered more selectively.

Key Words: bacterial culture • empyema • fungal culture • pleural effusion • thoracentesis


   Introduction
TOP
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Analysis of pleural fluid obtained via thoracentesis is a valuable procedure in determining the cause of pleural effusions. The most common causes of pleural effusions include congestive heart failure (CHF), malignancy, and infection. It is generally recommended that microbiologic cultures and smears be performed on pleural fluid obtained by thoracentesis.1234 Although the yield of pleural fluid cultures in specific pleuropulmonary infections has been studied, very little information exists on the overall yield of these pleural fluid cultures performed at the time of diagnostic thoracenteses. Our aim was to assess the frequency of positive results from pleural fluid cultures and to identify factors associated with these positive results.


   Materials and Methods
TOP
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
A computer-assisted search of medical records was performed to identify all subjects listed as having undergone thoracentesis. In addition, a search was performed through the Laboratory Medicine computer database in identifying those subjects whose body fluid was assayed for protein or lactate dehydrogenase (LDH [normal serum value, 122 to 222 U/L]), two common tests performed on pleural fluid specimens. We identified 538 subjects undergoing their initial diagnostic or combined diagnostic and therapeutic thoracentesis at Mayo Medical Center over a 12-month period from July 1, 2001, to June 30, 2002. Thirteen patients who did not consent for their records to be reviewed were excluded from the analysis. The medical records from the remaining 525 patients were examined to ensure the diagnostic nature of the thoracentesis procedure. The following clinical data were abstracted: age, sex, clinical findings, laboratory test results, radiologic findings, final diagnosis, treatment, and outcome. We specifically identified those patients that had received antibiotics at any time within the 7 days prior to their thoracentesis. A pleural effusion was considered "loculated" if any imaging study (radiograph, CT, or ultrasound) clearly demonstrated septations or illustrated that the effusion was not free flowing. Fever was defined as any body temperature ≥ 38.5°C that was documented in the medical record within 24 h of the time the thoracentesis was performed.

Microbiologic samples were collected and processed according to well-established, published guidelines.5 Microbiologic cultures and smears from pleural fluid from each patient were recorded along with respective results. We analyzed pleural fluid data pertaining only to the initial diagnostic thoracentesis and excluded those related to repeat procedures. Isolates were considered contaminants if they are typically not associated with empyema (ie, coagulase-negative staphylococcus) and/or recovered from patients with no clinical evidence of infection or with alternative diagnoses that explained their illness. The underlying cause of the pleural effusion was based on the clinician’s diagnosis documented in the medical records, unless there was strong evidence suggesting another diagnosis.

Data are summarized using median, range for continuous variables, and frequency, percentage for categorical variables. Comparisons of true-positive vs negative and false-positive results in regard to associated demographic features were performed using an exact test for categorical variables and the two-sample rank-sum test for continuous variables. In all cases, two-tailed p values ≤ 0.05 were considered statistically significant.


   Results
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Among 525 patients undergoing diagnostic thoracenteses, 476 patients (91%) had one or more cultures performed on their pleural fluid specimens: aerobic bacterial (n = 469), anaerobic bacterial (n = 335), fungal (n = 277), mycobacterial (n = 227), viral (n = 5), actinomyces (n = 3), Legionella (n = 3), and mycoplasma (n = 1) cultures (Table 1 ).


View this table:
[in this window]
[in a new window]

  		Table 1. Cultures Performed on 476 Pleural Fluid Specimens

 
Thirty-nine positive results occurred in 35 of these 476 patients (7.4%). These 39 positive results represented 3.0% of all 1,320 cultures that were performed (Table 2 ); however, 20 of these 39 positive results (51%) were contaminants. Thus, true pathogens were identified in only 19 of 1,320 pleural fluid cultures (1.4%), including 11 of 469 aerobic bacterial (2.3%), 4 of 335 anaerobic bacterial (1.2%), 4 of 277 fungal (1.4%), and 0 of 227 mycobacterial cultures (0%). These 19 positive results for pathogens belonged to 15 patients or 3.2% of those who had cultures performed on their pleural fluid specimens. Three patients had more than one pathogen in their culture results.


View this table:
[in this window]
[in a new window]

  		Table 2. List of Positive Pleural Fluid Culture Results

 
Microbiologic smears performed on these pleural fluid samples included 357 Gram stains, 109 fungal smears (potassium hydroxide), and 232 smears for acid-fast bacilli (AFB) [Table 3 ]. Positive results were noted in 2.5%, 0%, and 0% of Gram stain, fungal, and AFB smears, respectively. These positive findings represented 1.3% of all microbiologic smears performed. Among the nine positive Gram stain results, seven also had true-positive pleural fluid findings. Of the remaining two smears, one likely represented a treated empyema and the other was a contaminant.


View this table:
[in this window]
[in a new window]

  		Table 3. List of Positive Smear Results

 
For 13 of 15 pleural fluid cultures that were positive for pathogens, the gross appearance was described as purulent (n = 5), yellow and cloudy (n = 4), and bloody or blood tinged (n = 4). The remaining two fluid samples were described as straw colored with no turbidity or discoloration.

The correlation of demographics and pleural fluid characteristics with culture results is shown in Table 4 . Since pleural fluid characteristics, eg, transudate vs exudate, would not be known until the fluid is analyzed, we attempted to identify factors that could be associated with true-positive results and are ascertainable at the time of fluid sampling. Eighty percent of the patients with true-positive results (95% confidence interval [CI], 52 to 96%) received antibiotics within 7 days prior to thoracentesis vs 47% (95% CI, 42 to 51%) of those with negative or false-positive results (p = 0.016). Loculated pleural effusions were present in 11 of 15 patients (73%) with true-positive results (95% CI, 45 to 92%), while only 23% (95% CI, 19 to 27%) of the effusions from patients with negative or false-positive results were loculated (p < 0.001). Among the 359 patients with free-flowing effusions, only four of the cultures (1.1%) demonstrated true pathogens. Of these four positive findings, three were fungal (one each of C albicans, C immitis, and C neoformans). The remaining true-positive finding from a free-flowing effusion grew only ß-hemolytic group A streptococcus. Data on body temperature within 24 h of thoracentesis were available for 421 of the 461 patients. Fever tended to be more common among patients with true-positive results (33%; 95% CI, 12 to 62%) than for those with negative or false-positive findings (12%; 95% CI, 9.1 to 15.2%), although the difference did not meet statistical significance (p = 0.062).


View this table:
[in this window]
[in a new window]

  		Table 4. Demographic and Pleural Fluid Characteristics and Culture Results*

 
The median value for pleural fluid protein was not significantly different in patients with true-positive results (p = 0.589), but the median LDH values was significantly higher in this group of 15 patients (p < 0.001) compared to those without pathogens. Fourteen of these 15 patients with true-positive results had their diagnostic thoracentesis performed while hospitalized, with the one remaining patient being hospitalized on the day following his thoracentesis (Table 4).

The underlying causes of the pleural effusions, as determined by the clinicians (unless strong evidence suggested another diagnosis), are summarized in Table 5 . Sixteen percent of all effusions were believed to be of infectious origin (empyema and parapneumonic effusions) and 60% were of noninfectious causes; the cause was not known for the remaining 24%. Not surprisingly, true-positive results were more likely to be found in patients with pleural effusions associated with infections (p < 0.001). The final diagnoses for 20 patients with false-positive results included CHF/fluid overload in 5 patients, inflammatory/noninfectious in 2 patients, malignant in 4 patients, unknown in 6 patients, and parapneumonic in 3 patients. The organisms noted in these parapneumonic effusions were believed to be contaminants.


View this table:
[in this window]
[in a new window]

  		Table 5. Diagnoses of Pleural Effusions*

 
We calculated the total cost of all the cultures performed in the patients in this study to assess the economic ramifications of these tests. The total cost of the 1,320 cultures was $72,908. This translated into a cost of $3732 for every true-positive pleural fluid finding.


   Discussion
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
The results of our study demonstrate that pleural fluid cultures are commonly ordered on pleural fluid samples, but the frequency of true-positive results is very low. Without a "gold standard" with which one can compare these results, it is difficult to calculate the sensitivity of the cultures in our analysis; however, even among the cultures that demonstrated microbial growth, the organisms were more likely to represent contaminants than true pathogens. In our study, only 49% (19 of 39) of the positive results represented true pathogens.

Recommendations regarding the use of pleural fluid cultures are not well established and are somewhat contradictory. In 1988, Sahn6 recommended that "Gram, KOH [potassium hydroxide], and AFB stain and pleural fluid cultures should be done only when an infection is suspected with an exudative effusion." However, Light in 1999 stated, "Pleural fluid on patients with undiagnosed exudative pleural effusions should be cultured for bacteria (both aerobically and anaerobically), mycobacteria, and fungi. Routine smears for mycobacteria are not indicated because of its low yield."7 A recent statement from the British Thoracic Society recommended that Gram stain, AFB stain, and microbiologic culture be obtained on all pleural fluid samples.1 A review of the literature finds very few published studies assessing the yield of microbiologic cultures in pleural fluid. One exception to this lack of data is noted regarding the utility of pleural fluid smear and culture for tuberculosis.89 Valdes et al8 reviewed the records of 254 patients with proven tuberculous pleurisy and documented a sensitivity of pleural fluid AFB smear and mycobacterial culture of 5.5% and 36.6%, respectively.

Despite the fact that bacterial and fungal smears and cultures are frequently ordered, the yield of these studies is not well established.49 Walshe and colleagues10 reviewed 150 thoracenteses and found that 137 had fluid sent for microbiological investigation (91%). Of these 137 patients, only 5 patients (3.5%) had positive results (four bacteria, one tuberculosis). In a prospective study intended to compare different methods of culture and transport, Ferrer et al11 found 15.5% of their pleural fluid samples were positive for microorganisms. According to the definitions they established, results were positive in 69.4% of the patients with empyema and in 33% of those with an "infectious pathology." The sensitivity of the bacterial and fungal culture findings in these studies and in our own cannot be definitively established because there is not a diagnostic "gold standard" with which to compare. Few patients undergo pleural biopsy in an attempt to identify bacterial or fungal pathogens in pleural fluid. The surgical evacuation of empyemas and parapneumonic effusions is performed for therapeutic, rather than diagnostic purposes.12 Our study demonstrated a lower frequency of true-positive findings than did the studies of Walshe et al10 or Ferrer et al.11 It is unclear if this reflects differences in technique, antibiotic use, or the prevalence of effusions caused by infectious processes. Some of the variations are likely explained by differences in the study populations.

Given the low positive yield of pleural fluid cultures, we tried to determine factors associated with a higher likelihood of true-positive results. Although one could argue for microbiologic studies to be performed only on exudative effusions, this is difficult to accomplish in actual practice since the exudative nature of pleural fluid cannot be accurately anticipated prior to fluid results being available.

In our analysis, the strongest predictor of a true-positive pleural fluid result was the presence of loculations in the pleural fluid (Table 4). Loculated pleural effusion was noted in < 25% of the patients in our study. Among the remaining 359 free-flowing effusions, only 4 demonstrated growth of true pathogens, 3 of these being fungal. This observation may be useful in guiding the appropriate utilization of pleural fluid cultures. Despite being the strongest predictor in our study, loculated fluid was associated with a true-positive result in only 10% of these patients. Perhaps more useful information for clinicians is that true-positive results were found in only 1.1% of the effusions that were free flowing.

Not surprisingly, pleural fluid results were more likely to be positive among patients with effusions associated with infections (parapneumonic effusion and empyema) than for noninfectious effusions. When only the patients with effusions determined to be related to infection are considered, the frequency of true-positive results is still only 18%, lower than the rate of 33% published by Ferrer et al.11 This suggests that the sensitivity of pleural fluid cultures is low, even for patients expected to have the highest likelihood of having an infectious cause for their pleural effusion.

The correlation of pleural fluid characteristics with culture results did not yield useful guides. Pleural fluid LDH level was significantly different between true-positive culture results and those that were negative or false-positive (p < 0.001); fluid protein, however, was not significantly different (p = 0.589). The clinical significance of these data is not entirely clear. The broad range for both of these variables makes the application in individual cases problematic. In addition, these values would not be known at the time of pleural fluid sampling.

Our study is limited by several factors. First of all, being a retrospective survey, we were unable to control for several potentially confounding factors. Because of the variability in ordering practices, not all pleural effusions could be categorized as transudates or exudates. Second, many of the patients had received antibiotics in the seven days before their thoracentesis, likely decreasing the yield of the cultures.13 Of note is the fact that a significantly higher percentage of patients in the true-positive group received antibiotics, likely reflecting an increased clinical suspicion for an infectious etiology in this group of patients. Finally, it should be noted that the patient population seen at our medical center has a low prevalence of tuberculosis.

Pleural effusions remain a common problem in medicine, and any diagnostic algorithm should be based on knowledge of the limitations of the studies involved. Based on the results of our study, we recommend that the inherent limitations of pleural fluid cultures be taken into account. While true-positive results are clinically valuable, the limitations of their sensitivity and specificity should be taken into account when interpreting positive and negative results. This is particularly true for certain patient groups. Because of the low frequency of true-positive results from patients with free-flowing effusions, obtaining pleural fluid culture specimens from these patients appears to be of limited value, unless there is a strong clinical suspicion that the effusion represents an early empyema or a fungal process. Also, because only 1 of the 119 pleural fluid cultures obtained in the outpatient setting represented a true pathogen, we would recommend that cultures not be performed in these patients unless infection is strongly suspected.


   Footnotes
 
Abbreviations: AFB = acid-fast bacilli; CHF = congestive heart failure; CI = confidence interval; LDH = lactate dehydrogenase

Funding was provided by Mayo Institutional funds.

Received for publication February 4, 2004. Accepted for publication October 14, 2004.


   References
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Maskell, NA, Butland, RJ, . Pleural Diseases Group Standards of Care Committee, British Thoracic Society. (2003) BTS guidelines for the investigation of a unilateral pleural effusion in adults. Thorax 58(suppl 2),ii8-ii17[Medline]
  2. Vikram, HR, Quagliarello, VJ Diagnosis and management of empyema. Curr Clin Top Infect Dis 2002;22,196-213[Medline]
  3. Bryant, RE, Salmon, CJ Pleural empyema. Clin Infect Dis 1996;22,747-762[ISI][Medline]
  4. Everts, RJ, Reller, LB Pleural space infections: microbiology and antimicrobial therapy. Sem Respir Infect 1999;14,18-30[Medline]
  5. Miller, J Bacteriology. Murray, PR Baron, EJ Pfaller, MAet al eds. Manual of clinical microbiology 7th ed. 1999,33-104 ASM Press. Washington, DC:
  6. Sahn, SA State of the art: the pleura. Am Rev Respir Dis 1988;138,184-234[ISI][Medline]
  7. Light, RW Useful tests on the pleural fluid in the management of patients with pleural effusions. Curr Opin Pulm Med 1999;5,245-249[CrossRef][Medline]
  8. Valdes, L, Alvarez, D, San Jose, E, et al Tuberculous pleurisy: a study of 254 patients. Arch Intern Med 1998;158,2017-2021[Abstract/Free Full Text]
  9. Hasaneen, NA, Zaki, ME, Shalaby, HM, et al Polymerase chain reaction of pleural biopsy is a rapid and sensitive method for the diagnosis of tuberculous pleural effusion. Chest 2003;124,2105-2111[Abstract/Free Full Text]
  10. Walshe, AD, Douglas, JG, Kerr, KM, et al An audit of the clinical investigation of pleural effusion. Thorax 1992;47,734-737[Abstract]
  11. Ferrer, A, Osset, J, Alegre, J, et al Prospective clinical and microbiological study of pleural effusions. Eur J Clin Microbiol Infect Dis 1999;18,237-241[CrossRef][Medline]
  12. Light, RW A new classification of parapneumonic effusions and empyema. Chest 1995;108,299-301[Free Full Text]
  13. Strange, C, Sahn, SA Management of parapneumonic pleural effusions and empyema. Infect Dis Clin North Am 1991;5,539-559[Medline]



This article has been cited by other articles:


Home page
 JWatch Infect. DiseasesHome page
Cultures of Pleural Effusion Specimens: Low Diagnostic Yield
Journal Watch Infectious Diseases, April 22, 2005; 2005(422): 5 - 5.
[Full Text]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Article Archive
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (2)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Barnes, T. W.
Right arrow Articles by Ryu, J. H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Barnes, T. W.
Right arrow Articles by Ryu, J. H.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS