(Chest. 2000;118:1626-1629.)
© 2000
American College of Chest Physicians
Measuring Pleural Fluid pH*
High Correlation of a Handheld Unit to a Traditional Tabletop Blood Gas Analyzer
Gary L. Kohn, MD and
William D. Hardie, MD
*
From the Division of Pulmonary Medicine, Children’s Hospital Medical Center, Cincinnati, OH.
Correspondence to: William D. Hardie, MD, Children’s Hospital Medical Center, Division of Pulmonary Medicine, OSB-5, 3333 Burnet Ave, Cincinnati, OH 45229; e-mail: bill.hardie{at}chmcc.org
 |
Abstract
|
|---|
Study purposes: To survey hospital laboratories in the
United States to determine methods used for measuring pleural fluid pH,
and to compare pleural fluid pH values obtained with a traditional
tabletop blood gas analyzer (BGA) to those obtained with a handheld
analyzer.
Methods: Hospital laboratories nationwide
were contacted by telephone to survey the methods used to measure
pleural fluid pH. In a second phase, pleural fluid was prospectively
collected from 19 pediatric and adult patients with pleural effusions,
and pleural fluid pH was measured simultaneously with a traditional
tabletop BGA and with a handheld unit.
Results: A
total of 220 hospital laboratories were contacted by telephone, and 166
responded (75%). The methods for determining pleural fluid pH for all
hospital laboratories were pH meter (35%; n = 59), BGA (32%;
n = 53), and litmus paper (31%: n = 51); 2% (n = 3) did not
perform the test. University hospitals were more likely to use a BGA,
compared to community hospitals (p < 0.014) or children’s hospitals
(p < 0.001). In the comparison of pleural fluid measurements, the
mean pH for the traditional BGA was 7.358 ± 0.189, and the mean pH
for the handheld unit was 7.382 ± 0.203. The absolute
difference between the two machines was 0.024 U, and the two methods
were correlated (p < 0.01; r = 0.993; degrees of
freedom = 36).
Conclusion: Most hospital
laboratories in the United States do not measure pleural fluid pH using
a traditional BGA and use alternative methods that have previously been
shown to be inaccurate. Pleural fluid pH obtained by a handheld unit
has a high degree of correlation to that of a traditional tabletop BGA,
and it offers a satisfactory alternative for laboratories reluctant to
measure pleural fluid pH with a BGA.
Key Words: blood gas machine • pH • pH meter • pleural effusions
|
Introduction
|
|---|
The
pleural fluid pH is an important laboratory test in the diagnosis and
management of malignant and parapneumonic effusions. The pleural fluid
pH is especially valuable for managing parapneumonic effusions, which
have an estimated incidence of 300,000 cases per year in the United
States.1
2
3
Parapneumonic effusions present with a
spectrum of disease severity, ranging from free-flowing, uncomplicated
effusions that resolve spontaneously following antibiotic therapy, to
complicated effusions that require surgical drainage for resolution.
Because the progression from a free-flowing effusion to the formation
of intrapleural pus and loculations may occur rapidly, most
authors3
4
5
6
recommend prompt pleural drainage when a
complicated course seems likely.
Of the common pleural fluid indexes measured to assess parapneumonic
effusions, pleural fluid pH has been shown to be the most consistent
diagnostic predictor of the need of pleural space drainage, with many
authors citing a pleural fluid pH < 7.2 as an indication for chest
tube drainage.7
8
9
The most accurate method for measuring
pleural fluid pH is the blood gas analyzer (BGA).2
10
11
Studies by Cheng et al12
and Chandler et
al13
have demonstrated that other methods of
measuring pleural fluid pH, including the pH meter and indicator strip,
are far less accurate than the BGA for determining pleural
fluid pH. However, many hospital laboratories are reluctant to measure
pleural fluid pH with the BGA, probably because of concerns of
obstructing the machine with coagulated proteins. A recent analysis of
acute-care hospitals in the southeastern United States demonstrated
that only 32% of institutions used a BGA for measuring pleural fluid
pH.13
The first goal of this study was to survey hospital laboratories
nationwide on the methods used to measure pleural fluid pH, and to
compare differences between university, community, and pediatric
institutions. The second goal was to prospectively compare measurements
of pleural fluid pH between the traditional BGA and a new alternative
method for analyzing pleural fluid pH, a handheld analyzer (i-STAT
Portable Clinical Analyzer; i-STAT Corporation; East Windsor,
NJ). This device is used in ICUs for rapid interpretation of
blood gas samples. Both machines measure pH with an electrode capable
of detecting hydrogen ions by direct potentiometry. In the traditional
BGA, this electrode is part of the tabletop unit, whereas in the
handheld unit, the electrode is contained in the disposable sample
cartridges. The advantage of the handheld device for measuring pleural
fluid pH is that the disposable cartridges allow users to avoid
potential damage to the base unit by precipitation of pleural proteins.
A close correlation in pH measurements between the handheld unit and
the tabletop BGA would thus provide an alternative method to measure
pleural fluid pH.
|
Materials and Methods
|
|---|
Hospital Survey
From August to December 1999, hospital laboratories representing
all 48 continental United States were randomly selected to be surveyed
by phone. The criteria for selection included listing by the 1999
American Hospital Association Handbook and having at least 10,000
admissions per year. The hospitals were categorized by the American
Hospital Association as either community-based, university-based, or
children’s hospital, and the criterion for 10,000 admissions was
waived for the children’s hospitals. At each institution, the
laboratory technician responsible for measuring pleural fluid pH was
interviewed about the method used by their laboratory.
Pleural Fluid pH Measurement
After approval from the Institutional Review Board at
Children’s Hospital Medical Center, Cincinnati, and The University of
Cincinnati Medical Center, samples of pleural fluid were obtained from
19 pediatric and adult patients with pleural effusions requiring
thoracentesis or chest tube placement. Samples were collected in blood
gas syringes with lithium heparin anticoagulant (SIMS; Keene,
NH) and placed on ice. Each sample was measured twice on the BGA
and the handheld unit for a total of four pH readings. All measurements
were performed using a BGA (CIBA-Corning 278 Blood Gas System
[CC-BGA]; CIBA-Corning; Walpole, MA) and the i-STAT analyzer with a
G3+ cartridge at 37°C. Before each measurement, the CC-BGA was
calibrated with a standard solutions, and the i-STAT machine was
calibrated with its electronic simulating module. Both units measure pH
to the third decimal place; the range for pH for the CC-BGA is 6.0 to
8.0, and for the i-STAT analyzer is 6.5 to 8.0.
Statistics
Differences in methods used to measure pleural fluid pH between
categories of hospitals were determined with 
2
analysis and z score. The Pearson product moment correlation
coefficient and z test were used to analyze the pH data between the
traditional BGA unit and the handheld unit. All statistical analyses
were performed using Statistical Analysis System Software (SAS Version
6.0; SAS Institute; Cary, NC).
|
Results
|
|---|
Hospital Survey
Two hundred twenty hospital laboratories were contacted, and 166
laboratories (75%) responded to the telephone survey. Community-based
laboratories represented 55% (n = 89), university-based represented
23% (n = 38), and children’s hospitals represented 22% (n = 36)
of hospitals surveyed. The methods for determining pleural fluid pH for
all hospital laboratories were pH meter (35%; n = 59), BGA (32%;
n = 53), and litmus paper (31%; n = 51); 2% (n = 3) did not
perform the test. Fifty-eight percent of university hospitals used the
BGA for measuring pleural fluid pH, compared to 33% of community
hospitals (p < 0.014) and 11% of children’s hospitals
(p < 0.001; Table 1
).
Pleural Fluid pH Measurement
Pleural fluid was obtained from 19 patients (age range, 1 month to
74 years) with pleural effusions of a variety of etiologies, including
parapneumonic (n = 10), malignancy (n = 7), and postoperative
(n = 2). Each pleural fluid sample was measured twice each for a
total of 37 pH readings (one sample was measured only once due to
limited amount of pleural fluid). The average (± SD) pH for
CC-BGA was 7.358 ± 0.189 (range 6.9 to 7.9), and the average pH for
the i-STAT was 7.382 ± 0.203 (range 6.9 to 8.0; Fig 1
). The absolute difference between the two machines was 0.024 U, which
was statistically significant; however, the two machines were highly
correlated for pH determination (p < 0.01; r = 0.993;
degrees of freedom = 36).
View larger version (23K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1.. Scatter plot with line of regression comparing the
pH in pleural fluid samples as measured by the CC-BGA and the i-STAT
device. Mean pH for the CC-BGA was 7.358, compared to 7.382 for the
i-STAT device. The average difference between the two units is 0.024
(p < 0.01), with a high degree of correlation
(r = 0.993; degrees of freedom = 36).
|
|
|
Discussion
|
|---|
The evaluation of pleural fluid pH is an invaluable test for the
clinician and is useful diagnostically and prognostically in the
management of patients with parapneumonic and malignant pleural
effusions. In parapneumonic effusions, the pleural fluid pH is used to
identify patients with a complicated effusion and to predict the need
for chest tube drainage.2
11
The original studies that
established the clinical guidelines for evaluating and treating
parapneumonic and malignant effusions were performed using a BGA to
measure pleural fluid pH.8
9
14
Hence, the BGA is
considered the "gold standard" for laboratories measuring and
reporting pleural fluid pH values.
Other alternative methods for measuring pleural fluid pH, including the
pH meter and pH indicator strips, have been shown to be inaccurate
compared with the BGA.12
13
The study by Cheng et
al12
comparing pleural fluid pH measured simultaneously by
a BGA and a pH meter found that the pH meter measured pleural fluid pH
0.16 U higher than the BGA and had a correlation of 0.56. When the BGA
was compared with a pH indicator strip, the mean difference was 0.81 U
higher than the BGA, and the correlation was 0.32.12
Chandler et al13
had similar results, with a mean
difference of 0.3 U between BGA and pH meter for measurement of pleural
fluid pH and a difference of 0.16 U between BGA and pH meter. The large
differences in pleural fluid pH obtained from the pH meter and
indicator strips are clearly enough to affect clinical decision
making.12
13
Despite the inaccuracies in measuring pleural fluid pH with the pH
meter and indicator strip, the majority of all hospital laboratories in
our survey did not measure pleural fluid pH with a BGA. While
university hospitals were more likely to use the BGA for measuring
pleural fluid pH than community or children’s hospitals, we were
surprised to find that almost half of all university laboratories used
methods other than BGA for measuring and reporting pleural fluid pH.
Nationwide, our survey demonstrated that only 32% of surveyed
hospitals used the BGA for measuring pleural fluid pH, and this value
is identical to that of a recent survey of critical-care laboratories
in the southeastern United States.13
In our survey, the
most popular method used by laboratories for measuring pH was the pH
meter (35%), while in southeastern United States laboratories, the pH
indicator paper was most popular (56%). Our study did not assess why
hospital laboratories used methods other than the BGA for measuring
pleural fluid pH; however, concerns about coagulation of pleural fluid
proteins resulting in damage to the analyzer are likely explanations.
In this study, we compared the portable handheld clinical analyzer to
the BGA for determining pleural fluid pH. Both machines use the same
technology to measure pH, using an electrode to detect hydrogen ions by
direct potentiometry. In the BGA, the electrode is part of the tabletop
unit; however, in the handheld unit, the electrode is contained in a
disposable cartridge. The benefit of a disposable cartridge is that it
limits potential damage from pleural fluid samples and leaves the main
unit intact. Our analysis of pleural fluid samples demonstrates that
there is a very strong correlation between pleural fluid pH readings
obtained with the handheld device compared with the BGA
(r = 0.993), with a statistically significant difference
between the two units (p < 0.01). This high correlation and
statistical significance indicate that the study was well powered,
despite the relatively few numbers of pleural fluid samples tested.
Although the pleural fluid pH readings were on average 0.02 U higher in
the handheld unit compared with the BGA, this bias is unlikely to
affect clinical management and is a clear improvement in correlation
compared with the alternatives of the pH meter or indicator strip
paper.
A general cost comparison between the handheld device and the BGA
reveals that the handheld device is affordable for hospital
laboratories and is at least comparable in overall cost compared to a
BGA. The i-STAT is more expensive to run per sample, with the cost of
the G3 i-STAT cartridge at $6, while the CC-BGA costs approximately
$0.40 per sample. However, the initial cost of an i-STAT is lower at
$5,000 compared to approximately $28,000 for a CC-BGA.
In summary, our study demonstrates that the majority of hospitals
throughout the United States continue to use methods other than a BGA
for measuring and reporting pleural fluid pH, despite the reported
imprecision of these alternative methods. We present data that
demonstrate a strong correlation in pleural fluid pH measurements
obtained by the handheld device and a tabletop BGA. These data support
use of the handheld device for measuring pleural fluid pH as an
accurate alternative for laboratories who are reluctant to measure
pleural fluid pH with a tabletop BGA.
|
Acknowledgements
|
|---|
Dr. Paul Succop provided assistance in statistical
analysis of our data. Dr. Paul Steele from the Division of Pathology
provided laboratory technical support, and Abbott Diagnostics provided
the G3+ cartridges and temporary use of the i-STAT unit for the study.
|
Footnotes
|
|---|
Abbreviations: BGA = blood gas analyzer;
CC-BGA = CIBA-Corning Blood Gas Analyzer
Financial support provided by the Division of Pulmonary Medicine,
Children’s Hospital Medical Center, Cincinnati, OH.
Received for publication February 24, 2000.
Accepted for publication June 29, 2000.
|
References
|
|---|
-
Niederman, MS, Bass, JBJ, Campbell, GD, et al (1993) Guidelines for the initial management of adults with community-acquired pneumonia: diagnosis, assessment of severity, and initial antimicrobial therapy. Am Rev Respir Dis 148,1418-1426[ISI][Medline]
-
Light, RW, Girard, WM, Jenkinson, SG, et al (1980) Parapneumonic effusions. Am J Med 69,507-512[CrossRef][ISI][Medline]
-
Light, RW, MacGregor, MI, Ball, WCJ, et al (1973) Diagnostic significance of pleural fluid pH and PCO2. Chest 64,591-596[Abstract/Free Full Text]
-
Light, RW (1995) Pleural diseases 3rd ed. ,129-153 Williams & Wilkins Baltimore, MD.
-
Steinbrecher, HA, Najmaldin, AS (1998) Thoracoscopy for empyema in children. J Pediatr Surg 33,708-710[CrossRef][ISI][Medline]
-
Hoff, SJ, Neblett, WW, Edwards, KM, et al (1991) Parapneumonic empyema in children: decortication hastens recovery in patients with severe pleural infections. Pediatr Infect Dis J 10,194-199[ISI][Medline]
-
Ewig, JM (1995) Pleural effusions: diagnostic considerations. Pediatr Rev 16,79[Abstract/Free Full Text]
-
Chavalittamrong, B, Angsusingha, K, Tuchinda, M, et al (1979) Diagnostic significance of pH, lactic acid dehydrogenase, lactate and glucose in pleural fluid. Respiration 38,112-120[Medline]
-
Potts, DE, Levin, DC, Sahn, SA (1976) Pleural fluid pH in parapneumonic effusions. Chest 70,328-331
-
Heffner, JE, Brown, LK, Barbieri, C, et al (1995) Pleural fluid chemical analysis in parapneumonic effusions: a meta-analysis. Am J Respir Crit Care Med 151,1700-1708[Abstract]
-
Light, RW (1981) Management of parapneumonic effusions. Arch Intern Med 141,1339-1341[CrossRef][ISI][Medline]
-
Dong-Sheng Cheng, M, Rodriguez, RM, Rogers, JR, et al (1998) Comparison of pleural fluid pH values obtained using blood gas machine, pH meter, and pH indicator strip. Chest 114,1368-1372[Abstract/Free Full Text]
-
Chandler TM, McCoskey ED, Byrd RB Jr, et al. Comparison of the use and accuracy of methods for determining pleural fluid pH. South Med J 199; 92:214–217
-
Potts, DE, Taryle, DA, Sahn, SA (1978) The glucose-pH relationship in parapneumonic effusions. Arch Intern Med 138,1378-1380[Abstract]
TOP
Abstract
Introduction
