Efficacy of Chest Radiography in a

Respiratory Intensive Care Unit*

A Prospective Study
William B. Bekemeyer, M.D., F.C.C.P; Robert 0. Crapo, M.D., FC.C.P;
Stephen Caihoon, M.D., F.C.C.P.; Clawson Y. Cannon, B.S.; and
Paul D. Clayton, Ph.D.

A prospective study of chest radiographic examinations in a examinations that were prompted by a change in a patients
respiratory intensive care unit was conducted to determine clinical status. Less than 6 percent ofthe radiographic films
the diagnostic and therapeutic efficacy of such examina- taken post-procedure demonstrated abnormalities poten-
tions. Analysis of data from 1,354 x-ray films from 167 tially related to the procedure. We conclude that, in a
patients revealed a 34.5 percent incidence of new (or respiratory intensive care unit: 1) routine morning radio-
increased) abnormalities, or tube or catheter malposition. graphic examination frequently demonstrates unexpected
Changes in diagnostic approach or therapeutic measures, or changing abnormalities, many of which prompt changes
excluding catheter position adjustments, occurred after 28.5 in diagnosis or management; 2) radiographic evaluation of a
percent ofthe examinations. Radiographic yield was higher change in a patients clinical condition has a higher yield
when a change in clinical condition prompted the radio- than routine examinations; and 3) post-procedure radio-
graphic examination than when the examination was a graphic examination uncommonly demonstrates complica-
routine morning study. Changes in the approach to patient tions related to the procedure, but frequently demonstrates
management were also more likely (42.7 percent) following abnormalities of tube or catheter placement.

I n recent years, questions have arisen regarding the physical findings.

efficacy of diagnostic radiographs in varied clinical Despite the 54 percent incidence of serious radio-
situations. Sagel et aP have reported that routine chest graphic abnormalities in patients over the age of 40
radiographic screening may safely be eliminated in with suspected chest diseas& and the 37 percent
patients under 20 years ofage admitted to the hospital incidence of acute radiographic abnormalities in pa-
for noncardiopulmonary complaints, and that lateral tients presenting with acute chest complaints,2 it has
projections of the chest may be eliminated in patients been suggested that radiographs are overutilized, es-
under the age of4O unless chest disease is suspected. pecially in intensive care units.34 To our knowledge,
However, in that same report, over 50 percent of pa- there have been three published studies examining the
tients with a reasonable probability of a chest abnor- utility ofchest radiographic examination in critical care
mality or with suspected chest disease were found to units.7 Greenbaum and Marschall reported that 43
have serious abnormalities on chest radiographs. In a percent of 126 routine chest radiographs performed in
prospective analysis of patients presenting with acute a medical intensive care unit demonstrated abnormali-
chest complaints, Benacerraf et al2 found that chest ties which prompted changes in therapy. These in-
radiographic examination was not useful in detecting cluded malposition of invasive devices, as well as
disease in patients under the age of4O in the absence of changing infiltrates and the development of pneumo-
either hemoptysis or abnormalities apparent on phys- thorax. No information is given regarding the changes
ical examination. These investigators found acute radi- in management that ensued. Janower et al6 examined
ographic abnormalities in almost half of the patients 233 portable chest films obtained in intensive care
over forty years of age with acute chest symptoms units, on medical and surgical floors, and in the
unrelated to the presence or absence of abnormal emergency room and found that 37 percent of follow-
up examinations demonstrated new radiographic find-
*Fmm the Intermountain Respiratory Intensive Care Unit, Depart- ings. No information is available regarding patient
ment of Medical Biophysics and Computing, and Department of management changes based on these radiographs. In
Medicine, LDS Hospital, University of Utah, Salt Lake City; and
the Department of Medicine, University ofTennessee, Memphis. the most extensive study to date, Henschke et al7
Dr. Bekemeyer was a research fellow of the American Lung examined 1,132 chest radiographic films from medical
Association at the time these studies were done.
Presented inpart at the 50th Annual Scientific Assembly, American and surgical intensive care units. Marked cardiopul-
College of Chest Physicians, Dallas, October 8-12, 1984. monary abnormalities, or tube or catheter malposition
Manuscript received December 14; revision accepted May 7.
Reprint requests: Dr Beke,neyer, 956 Court Avenue, 3H29, Mem-
were present in 65 percent. These were stated to
phis 38163 reflect management efficacy of these radiographic

CHEST I 88 I 5 I NOVEMBER, 1985 691

Downloaded From: http://journal.publications.chestnet.org/ on 12/14/2015

 examinations, though clinical evidence in support of were never informed ofany data examined during the study period.
Personal bias of the individual fellows may have affected this
this is not presented. Routine and nonroutine radio-
investigation, if the fellows believed that radiographs were or were
graphic examinations were not separated in this study. not useful in the IRICU. However, we imposed no bias upon the
The American College of Radiology has divided the fellows by discussing the specific purpose ofthe study. Consultation
issue ofefficacy into three parts:8 1) diagnostic efficacy, with a radiologist was performed when questions arose. The fellow

or the influence of a test result on diagnosis; 2) entered data on the technical adequacy ofthe radiographic film, the
position of the patient at the time of the examination; the level of
therapeutic efficacy, or the effect of a test result on
positive end-expiratory pressure, if applicable; the position of all
clinical management; and 3) outcome efficacy, or the visible tubes and catheters, including thoracostomy tubes, pulmo-
effect that a test result will have on patient outcome. In nary artery catheters, endotracheal tubes, central venous catheters,
this prospective study, we analyzed the occurrence of and nasogastric and small bowel tubes. The fellow also entered

new or increased radiographic abnormalities, as well as information about all pathologic findings and a comment as to
whether the abnormality was new, increased, decreased, or un-
the specific actions taken as a result ofthe information
changed. Finally, the fellow entered information regarding whether
obtained from such radiographic examinations of the the finding elicited a diagnostic or therapeutic response and what
chest, in a respiratory intensive care unit, thus yielding that response was. Diagnostic actions available for selection on the
information about the first two types ofefficacy, but not computer included institution of hemodynamic monitoring, thora-

the third. cocentesis, bronchoscopy, collection of cultures, special radio-

graphs, and other diagnostic tests. Therapeutic actions included
changes in medications, respiratory therapy, ventilator settings,
MATERIALS AND METHODS placement ofthoracostomy tubes, tracheal intubation or extubation,

The Intermountain Respiratory Intensive Care Unit (IRICU) at and other therapy changes. Certain subgroups ofradiographic films,
LDS Hospital is a three-bed primary and referral facility staffed with to be described later, were examined by one of us (WB), and the

onsite residents and pulmonary fellows and oncall pulmonary findings were found to closely match the responses recorded in the
intensivists. The IRICU is designed to treat patients in whom the computerized data entries. The computer data was also examined to
primary threat to life is respiratory or ventilatory failure. The most confirm that findings reported to be new had not been recorded as
frequent clinical problems encountered in this setting include adult present on the two previous radiographic films. Statistical analysis of
respiratory distress syndrome, pneumonia, pulmonary embolism, the potential differences between groups were performed using the
chronic obstructive pulmonary disease with superimposed acute proportionality test.9
respiratory failure, asthma, and aspiration pneumonitis. Prior to any
radiographic exposure in the IRICU, electrocardiographic leads and RESULTS

any other overlying equipment is removed from the patients chest

Thirteen-hundred and fifty-four technically ade-
and ventilator tubing is held out of the field of intended exposure.
The distance from the x-ray tube to the film cassette, tube output
quate chest radiographic films of 167 patients were
(Ky), and position of the patient are adjusted and recorded on the analyzed from November, 1980 to January, 1983. One-
bed so that all radiographic examinations of a given patient use the hundred and twenty-two (9 percent) were repeated
same technique. because they werejudged to be technically inadequate
A computer program was designed for the entry ofdata regarding
for interpretation due to motion, improper exposure,
two separate sets of information. After consultation with the physi-
cians, the first set of data was entered by the nurses when the
or inadequate patient positioning. All those repeated
radiographic examination was ordered, and included the reason for were judged to be adequate for interpretation. One-
ordering the examination as well as an estimate ofthe likelihood of a hundred and twenty-five others were deleted from the
suspected abnormality. The probability of suspected findings was data base due to incomplete data entry. Upon review-
entered as high (>90 percent), medium (50 percent), low (<10
ing the data that were entered regarding these films,
percent), or very low (<1 percent). The reasons for the radiographic
we were unable to detect any bias in the reasons for
examinations were categorized into six specific subgroups: 1) routine
morning radiographic examination (IRICU policy requires a daily incomplete data being entered. Eighty-two percent of
chest radiograph prior to morning rounds); 2) admission radio- the examinations were anteroposterior in projection;
graphic examination (all patients admitted to the IRICU have a chest 5. 7 percent were lateral decubitus studies. The re-
radiograph taken as part of the initial evaluation, unless a recent
mainder were postero-anterior, crosstable lateral, or
chest radiographic film is available); 3) post-procedure evaluations
other projections. Forty-eight percent were exposed
(all invasive procedures on the central vasculature, airway, and
thoracic cavity included); 4) examinations as a result of a change in with the patient in a semi-erect position; the rest were
the clinical condition ofthe patient (symptoms, physical findings, or equally divided between supine and upright examina-
laboratory data) when deemed necessary by the physician at the tions. Reasons for ordering the examinations are sum-
bedside; 5) evaluation ofa previously known or suspected abnormal-
marized in Table 1.
ity; and 6) routine follow-up examination done other than as a routine
morning examination. In a few instances, two reasons for obtaining
Radiographic Abnormalities
the examination were entered by the nurse (eg, post-procedure and
change in clinical condition). In these instances, the radiographic For effect on diagnosis (efficacy type 1), the number
film was analyzed under only the first reason recorded. ofnew or increased findings were evaluated on routine
The next phase ofdata entry into the computer was performed by a
radiographic examinations, as well as on those re-
fellow in pulmonary medicine after morning rounds with the
quested post-procedure, to evaluate an existing abnor-
attending physician, residents, nurses and respiratory therapists.
Each radiographic film is discussed as part of these rounds. The
mality, and to evaluate a change in clinical condition.
pulmonary fellows knewlittle ofthe general purpose ofthe study and Of 716 routine morning radiographs, 322 (45 percent)

692 Efficacy of Chest RadiOgraphy in RICU (Bekemeyereta!)

Downloaded From: http://journal.publications.chestnet.org/ on 12/14/2015

 Table 1-Number ofICU Radiographic Examinations Table 3-T,pe 1 (diagnostic) Efficacy Related to
by Category Radiographic Category

Category Number (%) Radiographic category Type 1 efficacy4 (%)

Admission 48 (3.6) Admission 72.9

Routine morning 716 (52.8) Routine morning 30.2
Routine followup 18 (1.3) Routine followup 16.7
Post-procedure 305 (22.5) Post-procedure 7.9
Change in clinical condition 124 (9.2) Change in condition 50.0
Evaluate an existing abnormality 143 (10.6) Existing abnormality 16.8
Total 1354 (100) All radiographs 26.9

4Type 1 efficacy is defined as the percentage ofradiographs which are

diagnostic ofnew or increased abnormalities, excluding abnormali-
demonstrated at least one clinically unsuspected new ties of tube or catheter placement.
finding, increased finding, or a tube or catheter
malposition. including pulmonary artery catheters; five of these
Radiographs were performed after 305 procedures. were temporally associated with the development of a
Excluding tube or catheter malposition, 18 (5.9 per- pneumothorax, two of which were strongly suspected
cent) demonstrated new abnormalities that were po- (>90 percent estimated probability) prior to radio-
tentially related to the procedure. Table 2 lists these graphic study and three ofwhich were not suspected.
procedures and the radiographic complications that Pneumothoraces were strongly suspected in five cases
were found. Causality of these findings is not con- after central venous catheterization. The subsequent
firmed by this temporal association, but is inferred. radiographic films demonstrated pneumothoraces on
Eighty-one films (26.6 percent) showed at least one two of these.
tube or catheter that was judged to be in a suboptimal Endotracheal intubation was performed on 57 occa-
position; 42 of these were the catheters that had been sions, and nine endotracheal tubes were thought to
placed immediately prior to the radiographic examina- require adjustment in position after examining the
tion. post-procedure film (eight orotracheal tubes and one
One-hundred and one pulmonary artery catheters nasotracheal tube). Eight of these films were available
were placed. Of the 24 pulmonary artery catheters for review. One demonstrated the endotracheal tube in
noted on post-procedure radiographic examination to the right mainstem bronchus, one was in the larynx,
require a position adjustment, two were proximal, and six required only minor adjustment due to place-
apparently in the right ventricle or right ventricular ment near the carina.
outflow tract; three were looped in the right ventricle, Thoracostomy tubes were placed on 60 occasions.
right atrium or inferior vena cava; and 19 were thought Four of these were thought to be in suboptimal
to be placed distally in segmental or subsegmental position or were ineffective in resolving the abnormal-
pulmonary arteries. ity (three pneumothoraces and one pleural effusion) for
A total of 143 central venous catheters were placed, which they were placed.

Table 2-Radiographically Evident Complications of Procedures

Number (%) with

abnormal tube or
Procedure Number catheter placement Other complications (number)

Pulmonary artery 101 24 (23.8) Vascular abnormality (1)

catheterization Pleural effusion (1)
Mediastinal abnormality (1)
Pneumothorax (3)
Tube thoracostomy 60 4 (6.7) Mediastinal abnormality (1)
Volume loss (1)
Infiltrate (1)
Subcutaneous emphysema (3)
Endotracheal 57 9 (15.8) None
intubation
Central venous 42 5 (11.9) Extrathoracic abnormality (1)
catheterization Pneumothorax (2)
Thoracocentesis 36 - Pneumothorax (1)
Bronchoscopy 5 - None
Surgery/other 4 - Infiltrate (1)
Pleural effusion (1)

CHEST I 88 I 5 I NOVEMBER, 1985 693

Downloaded From: http://journal.publications.chestnet.org/ on 12/14/2015

 Table 4-Percentage ofRadiographic Films of Different change in clinical condition, those ordered post-proce-
Categories Evoking an Action (Type 2 or Therapeutic dure, and those ordered to evaluate a known or sus-
Efficacy)
pected abnormality. When the 716 routine films were
Dx or compared to the 572 nonroutine radiographic examina-
Category Dx Rx Adj Dx or Rx Rx or Adj tions (Table 5), significant differences (p<0.05) were
found between the two groups in the incidence of Dx
Admission 33.3 33.3 4.2 50.0 50.0
Routine morning 9.9 18.3 20.7 23.7 38.7 and Rx actions. Significant differences between the
Routine followup 11.1 16.7 0.0 16.7 16.7 groups were also found for the combined groupings of
Post-procedure 9.2 20.0 26.6 24.6 45.9 Dx or Rx, and Dx, Rx or Adj. There was no significant
Change in condition 12.1 34.7 19.4 42.7 57.3
difference between routine and nonroutine radio-
Existing abnormality 29.4 18.2 12.6 42.6 51.1
graphic films with respect to the incidence of tube
All radiographs 12.9 20.7 20.2 28.5 43.4
or catheter malposition. The Dx or Rx or Adj column
Abbreviations used: Dx radiograph induced institution of diag-
of Table 4 represents the type 2 (therapeutic) efficacy
nostic procedures; Rx radiograph induced change in therapy; Adj
of radiographic examinations of the different cate-
= radiograph induced a change in the position ofa tube or catheter.
gories.
Analysis ofall radiographic films in the study period
all radiographic
Of examinations performed, 364
revealed that new or increased infiltrates were docu-
(26. 9 percent) revealed new or increased abnormalities
mented 152 times, and that these were associated with
and 273 (20.2 percent) revealed inadequately posi-
diagnostic or therapeutic changes 75 times (49 per-
tioned catheters. Four hundred sixty-seven (34.5 per-
cent). The changes included changes in respiratory
cent) demonstrated one or more of these abnormali-
therapy, changes in ventilator settings, special radio-
ties. Table 3 summarizes type 1 (diagnostic) efficacy of
graphic examinations, changes in medications, and
radiographic examinations in the different categories.
collection of cultures.
The incidence of new or increased abnormality was
Pleural fluid was detected as a new or increased
significantly greater (p<O. 001) on radiographic films
finding on 71 occasions and diagnostic or therapeutic
taken to evaluate a change in clinical status than for
maneuvers were initiated in 37 (52 percent) of these,
routine examination.
including lateral decubitus radiographic studies 19
Diagnostic or Therapeutic Interventions times, prior to thoracocentesis in 15 patients.
Pneumothorax was a new or increased finding 58
In order to evaluate the effect of radiographic times, 21 ofwhich were immediately treated with tube
abnormalities on patient management (efficacy type 2),
thoracostomy, and ten of which were immediately
we evaluated physician responses to the radiographic
addressed with other specific therapeutic measures
abnormalities. Ofthe 716 routine radiographic studies,
such as thoracostomy tube adjustment or ventilator
71 (9.9 percent) led to further diagnostic procedures
changes. Others underwent special radiographic ex-
(Dx), 131 (18.3 percent) led to changes in therapy (Rx),
aminations, or timed follow-up radiographic examina-
and 148 (20. 7 percent) led to the adjustment of a tube
tions were scheduled.
or catheter (Adj). One hundred seventy (23.7 percent)
Atelectasis (40) and reduced lung volume (35) were
led to Dx or Rx, and 277 (38.7 percent) were followed
associated with diagnostic and/or therapeutic interven-
by Dx, Rx, or Adj. Table 4 compares these figures to
tion in 60 and 57 percent of the occurrences, respec-
those for the other categories of films.
tively. The majority of interventions in patients with
The categories of radiographic examination were
these abnormalities were changes in respiratory ther-
then grouped into: 1) routine radiographic examina-
apy or ventilator settings, or both.
tions, including only those examinations performed as
scheduled morning studies; and 2) nonroutine ex- DISCUSSION
aminations, including those ordered to evaluate a
The use of routine screening chest radiography for
detection of lung cancer,#{176} tuberculosis,11 or other
Table 5-Comparison ofRoutine and Nonroutine
cardiopulmonary diseasess has been demonstrated to
Radiographs Based on Actions Taken
be oflittle use in asymptomatic individuals. Similarly,
Routine Nonroutine p value (proportionality routine radiographic chest examination was found to
Action (N = 716) (N = 572) test) (9) have a low yield of abnormalities in patients without
Dx 9.9% 14.9% 0.003
suspected chest diseases, as well as in young patients
Ri 18.3% 22.7% 0.025 with acute chest complaints but no hemoptysis or
Adj 20.7% 21.5% 0.358 abnormality on physical examination.2 The increasing
Dx or Rx 23.7% 33.0% <0.001 cost of medical care and the potential hazard of
Dx, Rx, or Adj 38.7% 49.7% <0.001
radiation exposure have led to an examination of tests
For definitions of abbreviati ons see Table 4. that may be omitted from patient care in general, or to

694 Efficacy of Chest Radiography in RICU (Bekemeyeret a!)

Downloaded From: http://journal.publications.chestnet.org/ on 12/14/2015

 the definition of circumstances in which tests may be obtained from malpositioned catheters may be incor-
omitted. Respiratory intensive care units and other rect or misleading, but careful attention to pressure
critical care units are high cost, high utilization areas waveforms obtained from pulmonary artery catheters,
that should not be exempt from scrutiny. The mci- as well as observation for spontaneous wedge pressure
dence ofabnormal radiographic findings is reported to waveforms and knowledge of the volume of air re-
be between 43 and 70 percent in intensive care units.7 quired for balloon inflation to obtain a wedge pressure
These studies differ in that daily radiographic examina- may provide insight to peripheral catheter tip place-
tions were not routine and there were no comparisons 13 In addition, pulmonary artery catheters are
of routine daily examinations to those performed to reported to migrate peripherally during the 12 hours
study a change in clinical condition.7 Additionally, the following initial placement.13 If this is true, then
abnormalities on the films were not stated to be new or delayed radiographic examination of the chest (eg, the
increased.6 Thus, these studies may not actually reflect next routine radiographic examination) would likely
type 1 efficacy. Two of these studies7 noted that give the most reliable information regarding periph-
abnormal films were or should be associated with eral location of the catheter tip. Central venous
changes in management, but no information is for- catheters that are used for pressure measurements may
warded as to the type ofchanges or actual incidence of fluctuate with ventilations and still be abnormally
changes in management. The present study sought to placed outside the central 14 These
analyze the contribution of chest radiography to the catheters tend to give false information regarding
care of patients in a respiratory intensive care unit. central venous pressure. Thus, catheters used for this
Greater than one-fourth (26.9 percent) of all chest purpose, especially if therapy is to be based on these
radiographic examinations performed in the IRICU pressures, should have their position documented at
demonstrate new or increased abnormalities (type 1 least once prior to use.4 Central venous catheters
efficacy) other than tube or catheter malposition. The placed abnormally may still be used for venous access
incidence of these findings was higher on admission without confirmation of position.
examinations and on examinations done to evaluate a We found that tube or catheter malposition was
change in clinical condition than it was on routine equally likely to be found on routine and nonroutine
examinations or on those done to evaluate an existing radiographic examinations. We also noted that it was
abnormality. uncommon (9. 0 percent) for radiographic examinations
Our results also demonstrate that routine respira- to be repeated due to technical inadequacy, and that
tory ICU radiographic examinations are associated repeat examinations were uniformly adequate for eval-
with at least one physician action (type 2 efficacy) 38.7 uation.
percent ofthe time, a figure statistically lower than the In summary, both routine and nonroutine radio-
49. 7 percent for nonroutine radiographic examina- graphic examinations done in a regional respiratory
tions. Even if tube or catheter positional adjustments intensive care unit have a high incidence of revealing
are excluded, other diagnostic and therapeutic actions abnormalities that were either clinically unsuspected
were undertaken, at least partially due to radiographic or unconfirmed. Of course, demonstration of radio-
findings, after 23. 7 percent ofroutine and 33. 0 percent graphic abnormalities does not necessarily justify the
of nonroutine examinations. time, expense, and risk of the examination. However,
The incidence of radiographically demonstrable changes in patient management, especially if followed
procedural complications, other than tube or catheter by improved outcome, would seem to justify the pro-
malposition, was low (5.6 percent). Procedures with a cedures. We have no data regarding the question of
moderate-to-high probability of being complicated by radiographic findings altering patient outcome (type 3
pneumothorax (pulmonary artery catheterization, efficacy), but our results suggest that alterations in
tube thoracostomy, central venous catheterization, diagnosis and alterations in patient management are
and thoracocentesis) were performed on 239 occasions. commonly based, at least partially, on radiographic
These procedures were associated with the develop- findings in respiratory intensive care units.
ment of four (1.7 percent) unsuspected pneumo-
ACKNOWLEDGMENTS: The authors thank the nursing staff of the
thoraces. Had these unsuspected pneumothoraces Intermountain Respiratory Intensive Care Unit and the pulmonar
caused life-threatening hemodynamic or ventilatory fellows of the University of Utah. We also thank Jerri Duncan-Go
for assistance in preparation of the manuscript.
complications, the availability of on-site physicians
may have prevented untoward consequences. Other
REFERENCES
noted complications (Table 2) were not life-threatening
1 Sagel SS, Evens RG, Forrest JV, Bramson RT Efficacy of routine
and were apparent on subsequent routine films.
screening and lateral chest radiographs in a hospital-based
Similarly, suboptimal tube or catheter position is
population. N EngI J Med 1974; 291:1001-04
unlikely to be acutely life-threatening, with the possi- 2 Benacerraf BR, McLoud TC, Rhea JT, Tritschler V, Libby P. An
ble exception ofbronchial intubation. The information assessment of the contribution of chest radiography in outpa-

CHEST I 88 I 5 I NOVEMBER, 1985 695

Downloaded From: http://journal.publications.chestnet.org/ on 12/14/2015

 tients with acute chest complaints: a prospective study. Radi- 9 Armitage P Statistical methods in medical research. New York:
ology 1981; 139:293-99 John Wiley and Sons Inc, 1971
3 Griner PF, Glaser RJ. Misuse oflaboratory tests and diagnostic 10 Boucot KR, Weiss W. Is curable lung cancer detected by semi-
procedures. N EngI J Med 1982; 307:1336-39 annual screening? JAMA 1973; 224:1361-65
4 Hall FM. Overutilization ofradiological examinations. Radiology 11 Feingold AO. Routine chest roentgenograms on hospital admis-
1976; 120:443-48 sion do not discover tuberculosis. South Med J 1977; 70:579-80
5 Greenbaum DM, Marschall KE. The value ofroutine daily chest 12 Mackenzie CJG. A two year follow-up ofpersons with nontuber-
x-rays in intubated patients in the medical intensive care unit. culous chest disease found at Operation Doorstep, Vancouver
Crit Care Med 1982; 10:29-30 1964. Can Med Assoc J 1970; 103:1019-25
6 Janower ML, Jennas-Nocera Z, Mukai J. Utility and efficacy of 13 Foote GA, Schabel SI, Hodges M. Pulmonary complications of
portable chest radiographs. AJR 1984; 142:265-67 the flow-directed balloon-tipped catheter. N Engl J Med 1974;
7 Henschke CI, Pasternack GS, Schroeder 5, Hart KK, Herman 290:927-31
PG. Bedside chest radiography: diagnostic efficacy. Radiology 14 Keilner GA, Smart JR Percutaneous placement of catheters to
1983; 149:23-26 monitor central venous pressure. Anesthesiology 1972; 36:
8 Loop JW, Lusted LB. American College ofRadiology diagnostic 515-16
efficacy studies. AJR 1978; 131:173-79

696 Efficacy of Chest Radiography in RICU (Bekemeyer et a!)

Downloaded From: http://journal.publications.chestnet.org/ on 12/14/2015