Clinical Investigations
Received: January 13, 2004
Respiration 2005;72:176–181
DOI: 10.1159/000084049
Blood Gas Analysis and Chest X-Ray
Findings in Infants and Preschool
Children with Acute Airway Obstruction
Massimo Pifferi a Davide Caramella b Angelo Pietrobelli c
Vincenzo Ragazzo a Attilio L. Boner c
Departments of a Pediatrics and b Radiology, University of Pisa, c Department of Pediatrics, University of Verona, and
Istituto Pio XII, Misurina, Italy
Key Words
Blood gas analysis W Chest X-rays W Children W
Pneumonia W Respiratory distress
Abstract
Background: The importance of SaO2 in the assessment
of respiratory distress in bronchial asthma has been
reported. Objectives: To evaluate the correlation be-
tween blood gas analysis and chest X-ray lung opacities
in young children presenting with acute respiratory
symptoms. Methods: Eighty patients (43 males and 37
females aged 0.5–24 months; mean B SD 9.1 B 7.2
months), either with acute wheezing respiratory symp-
toms and/or with crackles were enrolled in our study. In
all children, blood gas analysis and chest X-rays were
performed within 12 h following admission to the emer-
gency department. Results: In 55 children (68.75%) chest
X-rays demonstrated lung opacities. Subjects with nor-
mal X-rays had paO2 and SaO2 higher than subjects with
lung opacities (p ! 0.0001 and p = 0.0001, respectively).
Children with lung opacities almost always presented
paO2 ! 80 mm Hg. Sensitivity and specificity for the
presence of lung opacities of paO2 !80 mm Hg were 81
and 90%, respectively, while sensitivity and specificity
of SaO2 ! 95% were 92 and 40%, respectively. paO2
© 2005 S. Karger AG, Basel
ABC 0025–7931/05/0722–0176$22.00/0
Fax + 41 61 306 12 34
E-Mail karger@karger.ch Accessible online at:
www.karger.com www.karger.com/res
Accepted after revision: August 25, 2004
!80 mm Hg in association with SaO2 !95% had a posi-
tive predictive value for the diagnosis of pneumonia of
90.9%. Conclusions: Our study suggests that blood gas
analysis, particularly paO2, may help in predicting the
presence of lung opacities in patients aged less than 2
years. However, chest X-rays may still be needed to
define the actual extension of opacities as well as the
possible concomitant presence of complications.
Copyright © 2005 S. Karger AG, Basel
Introduction
Several studies demonstrated the importance of arteri-
al oxygen saturation (SaO2) in the assessment of respirato-
ry distress in bronchial asthma [1–3]. Moreover SaO2 has
been used in the diagnosis of pneumonia specifically in
the developing countries, where a good correlation be-
tween low SaO2 levels and pulmonary opacities on chest
X-rays is observed in infants and children [4–7]. In west-
ern countries, better nutritional status and lower viral/
bacterial inoculum are usually associated with less severe
lung involvement [8, 9]. Furthermore discrete lung opaci-
ties may be present in wheezing infants without pneumo-
nia.
Massimo Pifferi, MD
Department of Pediatrics, University of Pisa
Via Roma 67
IT–56126 Pisa (Italy)
Tel. +39 050 992728, Fax +39 050 888622, E-Mail m.pifferi@med.unipi.it
 We have speculated that in these conditions SaO2 eval-
uation may not be sensitive enough in detecting both less
severe cases of pneumonia as well as lung opacities not
related to pneumonia itself. In fact, due to the S-shape of
the oxyhemoglobin dissociation curve, in the flat part of
the curve even significant variations in partial oxygen
pressure (paO2) may cause little variations in SaO2 [10].
Therefore, we have aimed at evaluating whether arteri-
al paO2 may be a better predictor of the presence of lung
opacities in young children presenting with acute respira-
tory symptoms. To verify this hypothesis, we retrospec-
tively analyzed data from children admitted to the emer-
gency unit in whom both blood gas analysis and chest X-
ray were done routinely.
Patients and Methods
Patients
In the period from September to December 1999, 107 infants and
preschool children were admitted to our emergency room with acute
respiratory symptoms. Those who had evidence of crackles and/or
wheezes and were routinely submitted to both blood gas analysis and
chest X-ray within 12 h following admission were therefore included
into this study. Exclusion criteria were cardiovascular, pulmonary or
neurological congenital defects, premature birth or chronic disease or
if parents did not consent to one or both of the proposed diagnostic
procedures.
Informed consent from parents and guardians was asked for
obtaining blood gas analysis and chest X-ray. The Hospital Ethical
Committee approved the study.
Physiological Evaluation
On admission, after having measured body temperature and
respiratory rate (observing chest wall movements over 1 min when
the patient was not crying) a blood gas analysis was performed on
blood samples obtained from the radial artery in a syringe prefilled
with heparin sodium (Pulsator, 3 ml; Concord Laboratories, Hythe,
UK) in the supine position [11]. EMLA cream was applied 30 min
previously in order to reduce the pain associated with the proce-
dure.
For the blood gas analysis, the 850 Ciba Corning equipment
(Diagnostics, Medfield, Mass., USA) was used at 37 ° C, regardless of
the patient’s body temperature since substantial controversy exists
regarding the clinical application of the correct temperature of blood
specimens [12].
Chest X-Rays
The chest X-rays were obtained in frontal and lateral projections.
For the purpose of this study, all images were reviewed by the same
radiologist who has extensive experience in pediatric radiology and
who was blinded to the clinical conditions of the patients. He was
asked to focus on the presence of lung opacities having the radiologi-
cal features of pneumonia.
Gas Analysis in Acute Respiratory Distress
Microbiological Evaluation
Respiratory syncytial virus (RSV) infection was evaluated by
indirect immunofluorescence technique to be able to identify the
viral antigen in the nasal secretions (Abbott, Testpack RSV Diagnos-
tic, North Chicago, Ill., USA), enabling a rapid diagnosis. Cell culture
using standard techniques was also performed in order to detect oth-
er viruses.
Statistics
Paired Student’s t test was used to evaluate the differences in pH,
paO2, paCO2, SaO2 and respiratory rate in patients with or without
lung opacities demonstrated by X-ray. The ¯2 test was used to evalu-
ate the correlation between clinical signs (wheezing and crackles) and
the presence of opacities on chest X-ray. Sensitivity and specificity of
paO2 and SaO2 values as indicators of pneumonia were evaluated
using receiver-operating-characteristic curves. Sensitivity and speci-
ficity of paO2 ! 80 mm Hg (which is usually used to define hypox-
emia) [12, 13] and SaO2 ! 95% for the diagnosis of pneumonia were
analyzed.
Subsequently, multiple linear regression analysis was performed
to assess the relationship of X-rays and pH, paO2, paCO2, SaO2, and
respiratory rate in all subjects controlling for age and gender [14].
Interaction with paO2 and SaO2 were tested. Results are expressed in
the text and tables 1 and 2 as group means B SD. All statistical calcu-
lations were performed using the SPSS v9.0 software package for
Windows (SPSS, Chicago, Ill., USA) for personal computers.
Results
Eighty children (43 males and 37 females aged between
0.5 and 24 months, mean B SD 9.1 B 7.2 months) were
enrolled in the study. Their respiratory rate ranged be-
tween 36 and 76 breaths/min (mean B SD 54.9 B 9.1).
Arterial blood gas analysis showed pH values of 7.26–7.48
(mean B SD 7.40 B 0.04), paO2 of 38.6–100 mm Hg
(mean B SD 74.0 B 14.2), paCO2 of 26.8–51.4 mm Hg
(mean B SD 36.5 B 5.4) and SaO2 of 74.7–100% (mean
B SD 95.2 B 4.5). No significant difference was found
between males and females for the variables examined.
Moreover, the distribution of the values obtained for each
considered variable was normal.
In 55 patients (68.75%), chest X-rays showed the pres-
ence of one or more discrete ill-defined pulmonary densi-
ties due to airspace filling. When multiple, the shadows
tended to coalesce. Air bronchogram was often visible
[15]. Table 1 shows values of pH, paO2, paCO2, SaO2, and
respiratory rate in infants with and without lung opacities.
As can be seen, paO2 and SaO2 were significantly higher
in subjects with normal X-rays compared with subjects
with lung opacities (p ! 0.0001 and p = 0.0001, respec-
tively). On the other hand, paCO2 was lower in subjects
with normal X-rays compared to subjects with lung opaci-
ties (p = 0.005), while pH and respiratory rate were not
Respiration 2005;72:176–181 177
 Fig. 1. Sensitivity and specificity of different paO2 and SaO2 values. Receiver-operating curves for paO2 (a) and SaO2
(b) as indicators of lung opacities in our population. The point closest to the upper left corner indicates the value that
would be the best indicator of pneumonia in the study population.

Table 1. Different respiratory patterns in subjects with and without lung opacities

X-ray pH paO2 paCO2 SaO2 Respiratory rate

Lung opacity 7.40B0.05 68.21B10.93 37.70B5.35 94.09B4.93 57.05B8.60

Normal 7.41B0.02 86.86B12.03 34.17B4.85 97.57B2.13 50.04B8.60
p valuea NS !0.0001 0.005 0.0001 NS

Data are means B SD.

a Student’s t test, X-rays showing lung opacity vs. normal X-rays.

significantly different. The difference in age between sub-
jects showing a lung opacity and those with a normal chest
X-ray was not significant (p = 0.92). Sensitivity and speci-
ficity of different paO2 and SaO2 values are presented in
the receiver-operating characteristic curves in figure 1.
The point closest to the bottom left corner, i.e. 100% sen-
sitivity and specificity, indicate the value that would be
the best indicator of pneumonia in the study population.
Sensitivity and specificity of paO2 !80 mm Hg (which
is usually used to define hypoxemia) [12, 13] were 81 and
90%, respectively, while sensitivity and specificity of
SaO2 !95% were 92 and 40%, respectively. The ¯2 test
showed a significant correlation between respiratory rates
150 and presence of lung opacities in subjects aged !12
months (p = 0.01). On the other hand, a respiratory rate
140 was found in all subjects aged 112 months and was
also associated with wheezing.
Linear regression analysis tested the relationship of
paO2 and SaO2 with chest X-rays. Chest X-ray results
were correlated with paO2 and SaO2 yielding a significant
model (p ! 0.0001, R2 = 0.44) (table 2). Effects of gender
and age were not significant. Adding respiratory rate we

178 Respiration 2005;72:176–181 Pifferi/Caramella/Pietrobelli/Ragazzo/

Boner
Table 2. X-ray regression model
Variable
paO2
SaO2
Constant
a ß refers to the unstandardized regression coefficient.
increased the significance of the model only by 3%, while
paCO2 did not add any significance. In addition, interac-
tions between paO2 and SaO2 were not significant.
Finally, using paO2 !80 mm Hg in association with
SaO2 !95% for the diagnosis of pneumonia, the blood gas
analysis showed positive and negative predictive values of
90.9 and 88.0% for lung opacities, respectively.
Wheezing was detected in 98.2% of children with lung
opacities and in 100% of patients without lung opacities
(p = 0.68). Crackles were detected in 81.8% of children
with lung opacities and in 76% of patients without lung
opacities (p = 0.76).
RSV was found in 41.8% of children with lung opaci-
ties (in 1 case associated with cytomegalovirus, CMV) and
in 44% of patients without lung opacities. Other viruses
detected in patients with pneumonia were CMV (2 pa-
tients) and adenovirus (1 patient). In one of the patients
without lung opacities, CMV was cultured. No positive
culture was obtained in 52.7% of patients with lung opaci-
ties and in 52% of patients with normal chest X-ray.
Discussion
Our study suggests that arterial paO2 analysis may play
a role in the diagnosis of lung opacities in pediatric
patients less than 2 years of age. In these conditions, lung
opacities may indicate pneumonia as well as asthma/
bronchiolitis. While hypoxemia in asthma is mainly due
to a disproportionate ventilation/perfusion (V̇A/Q̇) ratio
[16–18], in pneumonia it is mainly due to an intrapul-
monary shunt (V̇A/Q̇ = 0) secondary to the perfusion of
airspaces, which may be completely flooded by exudates
and cell debris [19–22]. The perfusion of airspaces not
involved in gas exchange is possible because there is an
insufficient capacity of hypoxic pulmonary vasoconstric-
tion [23] during acute pneumonia [20–22, 24, 25]. This is
due to the presence of endogenous vasodilator prostaglan-
dins and other as yet undefined mechanisms [22]. More-
Gas Analysis in Acute Respiratory Distress
ßa SEß p Overall R2 SE p model
–0.03 0.005 !0.0001
0.04 0.01 0.004
–1.2 1.1 0.3 0.44 0.35 !0.0001
over, the effect of hypoxic pulmonary vasoconstriction on
the reduction in the venous admixture is more pro-
nounced when the hypoxic segments are small and scat-
tered, as in our patients, than when an entire lobe is hyp-
oxic [26]. Furthermore, in pneumonia, there are other
mechanisms influencing the levels of arterial hypoxemia,
such as the disequilibrium of alveolar to endocapillary
oxygen diffusion or increased intrapulmonary parenchy-
mal oxygen consumption [20, 22]. It could be argued that
errors may arise when there is airway obstruction and sub-
sequently reduced peripheral ventilation. However, in the
microvascular gas exchange unit there is an excess capaci-
ty for gas exchange. This means that the capillary blood
reaches equilibrium with alveolar air long before it leaves
the capillary bed [27]. In diseases with airflow limitation
there is a reduction in gas exchange but the effect on paO2
is less evident than space-occupying processes where a
number of gas exchange units are completely excluded. In
our population paO2 was significantly lower (p ! 0.0001)
in patients with lung opacities as a result of pneumonia or
asthma/bronchiolitis. In infants with a more severe reduc-
tion in airflow, reduced collateral ventilation is associated
with the appearance of areas of atelectasis [28, 29]. When
the ventilation-perfusion ratio is reduced to a ‘critical val-
ue’, in the affected gas exchange units more gas is
absorbed by the blood than it is delivered during inspira-
tion. Such units are inherently unstable and may collapse.
On the other hand, in the completely developed lung,
there will be little effect on paO2 when the shunt is small,
but decreases in paO2 are dramatic when shunts are
630% [30].
In wheezing patients, ventilation-perfusion mismatch
is often typically reflected by a pattern in which a distinct
population of low ventilation-perfusion units exists inde-
pendent of units with a normal ratio. Shunting (V̇A/Q̇ = 0)
is notably absent until the patients develop the most
severe level of obstruction, e.g. status asthmaticus [16]. In
the majority of the cases the correlation between airflow
rates and V̇A/Q̇ mismatch is almost nonexistent, suggest-
Respiration 2005;72:176–181 179
ing that bronchoconstriction does not significantly affect
the V̇A/Q̇ ratio and that gas exchange is determined by
events in the most peripheral airways which are poorly
reflected by flow rate data [31]. As a corollary, there is
only a weak association between severity of V̇A/Q̇ mis-
match and clinical severity of asthma.
In our study, children with lung opacities presented
significantly lower SaO2 (p = 0.0001) and paO2 (p !
0.0001) than patients with normal X-ray. However, while
a paO2 !80 mm Hg was sensitive (81%) and specific
(90%), the SaO2 !95% was sensitive (92%) but not spe-
cific (40%). Due to the S-shape of the oxyhemoglobin dis-
sociation curve, at levels 160 mm Hg, paO2 is a sensitive
measure of blood oxygenation since neither percentage
saturation nor oxygen content change as much as paO2 in
this range [10, 12]. However, at paO2 !60 mm Hg rela-
tively small changes in paO2 produce large changes in sat-
uration, and in this range the measurement of SaO2 is
more reliable than paO2 evaluation [32].
Tachypnea and a history of rapid breathing have been
identified as two suitable criteria for the diagnosis of low-
er respiratory tract infections [33, 34]. Specifically, respi-
ratory rates over 50/min in infants and over 40/min in
children 12–35 months of age were found to be sensitive
and specific indicators of lower respiratory tract infec-
tions [35]. A respiratory rate of 40/min was found to be a
useful predictor of lower respiratory tract infections with
a sensitivity of 79%, according to auscultatory signs and/
or radiological abnormalities, even in children over 3
years old [36]. In another study performed in children
aged 1–4 years, fever higher than 38.5 ° C and a respirato-
ry rate above 60/min were the best predictors of lobar con-
solidation [37]. In our series only 1 infant had fever
(38.5 ° C).
Moreover, in our study a respiratory rate 150/min was
significantly associated (p = 0.01) with the presence of
lung opacities only in subjects aged !12 months. How-
ever, a respiratory rate 140/min was found in all subjects
112 months. Considering all the patients together, no sig-
nificant difference in respiratory rate was observed be-
tween the children with and without lung opacities (ta-
ble 1) since tachypnea is found in both radiologically doc-
umented lung opacities and in lower airway wheezing dis-
orders without lung opacities.
In conclusion, although it has been observed that in
infants there is no single sign that could be used to exclude
pneumonia definitively [38], our results suggest that if
paO2 is 180 mm Hg the chest X-ray is unlikely to present
lung opacities either due to pneumonia or severe asthma/
bronchiolitis.
Blood gas analysis in general reflects complications of
acute wheezing and respiratory distress, and our data sug-
gest that chest X-ray is not needed if paO2 is 180 mm Hg.
Routine chest X-ray is not suggested in asthmatic children
[39, 40].
The utilization of minimally invasive techniques such
as the analysis of an arterialized capillary blood sample
will provide reliable and accurate values of gas tension
with minimal discomfort for the children [41] and mini-
mal expenses, and will present important clinical infor-
mation, which may reduce hospitalization in these chil-
dren. Very likely the presence of a paO2 180 mm Hg con-
comitant with the absence of other biomarkers suggestive
of an individual infection will also enable to reduce anti-
biotic treatment in these patients.

References

1 Kerem E, Canny G, Tibshirani R, Reisman J,
Bentur L, Schuh S, Levison H: Clinical-physio-
logic correlations in acute asthma of childhood.
Pediatrics 1991;87:481–486.
2 Sole D, Komatsu MK, Carvalho KV, Naspitz
CK: Pulse oximetry in the evaluation of the
severity of acute asthma and/or wheezing in
children. J Asthma 1999;36:327–333.
3 Mayefsky JH, el-Shinaway Y: The usefulness of
pulse oximetry in evaluating acutely ill asth-
matics. Pediatr Emerg Care 1992;8:262–264.
4 Lozano JM, Steinhoff M, Ruiz JG, Mesa ML,
Martinez N, Dussan B: Clinical predictors of
acute radiological pneumonia and hypoxaemia
at high altitude. Arch Dis Child 1994;71:323–
327.
5 Reuland DS, Steinhoff MC, Gilman RH, Bara
M, Olivares EG, Jabra A, Finkelstein D: Preva-
lence and prediction of hypoxemia in children
with respiratory infections in the Peruvian
Andes. J Pediatr 1991;119:900–906.
6 Madico G, Gilman RH, Jabra A, Rojas L, Her-
nandez H, Fukuda J, Bern C, Steinhoff M: The
role of pulse oximetry. Its use as an indicator of
severe respiratory disease in Peruvian children
living at sea level. Respiratory group in Peru.
Arch Pediatr Adolesc Med 1995;149:1259–
1263.
7 Onyango FE, Steinhoff MC, Wafula EM, War-
iua S, Musia J, Kitonyi J: Hypoxaemia in
young Kenyan children with acute lower respi-
ratory infection. BMJ 1993;306:612–615.
8 Selwyn B: The epidemiology of acute respirato-
ry tract infections in young children: Compari-
son of findings from several developing coun-
tries. Rev Infect Dis 1990;12(suppl):S870–
S888.
9 Pio AJ, Leowski J, Ten Dam HG: The magni-
tude of the problem of acute respiratory infec-
tions. Proceedings of an International Work-
shop on Acute Respiratory Infections in Chil-
dren. Sydney, University of Adelaide, 1985, pp
3–16.
10 Foster RE II, Dubois AB, Briscoe WA, Fisher
AB: The lung: Physiologic basis of pulmonary
function tests, ed 3. Chicago, Year Book Medi-
cal Publishers, 1986.

180 Respiration 2005;72:176–181 Pifferi/Caramella/Pietrobelli/Ragazzo/

Boner
11 Hardie JA, Mørkve O, Ellingsen I: Effect of
body position on arterial oxygen tension in the
elderly. Respiration 2002;69:123–128.
12 Yorgin PD, Rhee KH: Gas exchange and acid-
base physiology; in Taussig LM, Landau LI, Le
Souëf PN, Martinez FD, Morgan WJ, Sly PD
(ed): Pediatric respiratory medicine. St. Louis,
Mosby, 1999, pp 212–243.
13 Lane EE, Walker JF: Clinical arterial blood gas
analysis. St. Louis, Mosby, 1987.
14 Hays WI: Statistics. New York, CBS College
Publishing, 1981.
15 Armstrong P, Lee P: Infections of the lungs and
pleura; in Armstrong P, Wilson AG, Dee P,
Hansell DM (eds): Imaging of Diseases of the
Chest, ed 2. St. Louis, Mosby, 1995, pp 145–
228.
16 Wagner PD, Dantzker DR, Iacovoni VE, Tom-
lin WC, West JB: Ventilation-perfusion in-
equality in asymptomatic asthma. Am Rev
Respir Dis 1978;118:511–524.
17 Rodriguez-Roisin R, Ballester E, Roca J,
Torres A, Wagner PD: Mechanisms of hypox-
emia in patients with status asthmaticus re-
quiring mechanical ventilation. Am Rev Res-
pir Dis 1989;139:732–739.
18 Ballester E, Reyes A, Roca J, Guitart R, Wag-
ner PD, Rodriguez-Roisin R: V/Q mismatch-
ing in acute severe asthma: Effects of salbuta-
mol and 100% oxygen. Thorax 1989;44:258–
267.
19 Henning R, South M: Respiratory failure; in
Taussig LM, Landau LI, Le Souëf PN, Marti-
nez FD, Morgan WJ, Sly PD (eds): Pediatric
Respiratory Medicine. St Louis, Mosby, 1999,
pp 404–430.
20 Rodriguez-Roisin R, Roca J: Update ‘96 on
pulmonary gas exchange pathophysiology in
pneumonia. Semin Respir Infect 1996;11:3–
12.
Gas Analysis in Acute Respiratory Distress
21 Light RB, Mink SN, Wood LDH: Pathophysi-
ology of gas exchange and pulmonary perfusion
in pneumococcal lobar pneumonia in dogs. J
Appl Physiol 1981;50:524–530.
22 Light RB: Pulmonary pathophysiology of
pneumococcal pneumonia. Semin Respir In-
fect 1999;14:218–226.
23 Tagaya E, Tamaoki J, Kawatani K, Nagai A:
Role of Na+-K+-ATPase in sodium nitroprus-
side-induced relaxation of pulmonary artery
under hypoxia. Respiration 2001;68:186–191.
24 McCormack DG, Paterson NA: Loss of hypox-
ic pulmonary vasoconstriction in chronic pneu-
monia is not mediated by nitric oxide. Am J
Physiol 1993;265:H1523–H1528.
25 Hampl V, Herget J: Acute pneumonia reversi-
bly inhibits hypoxic vasoconstriction in iso-
lated rat lungs. Physiol Res 1992;41:147–150.
26 Marshall BE, Marshall C, Benumoff J, Said-
man LJ: Hypoxic pulmonary vasoconstriction
in dogs: Effects of lung segment size and oxygen
tension. J Appl Physiol 1981;51:1543–1551
27 Hill EP, Power GG, Longer LD: Mathematical
simulation of pulmonary O2 and CO2 ex-
change. Am J Physiol 1973;224:904–917.
28 Murray JF: The Normal Lung. Philadelphia,
Saunders, 1986.
29 Thurlbeck WM: Postnatal growth and develop-
ment of the lung. Am Rev Respir Dis 1975;
111:803–844.
30 Dantzker RM: Gas exchange in adult respirato-
ry distress syndrome. Clin Chest Med 1982;3:
57–67.
31 Roca J, Ramis L, Rodriguez-Roisin R, Balles-
ter E, Montserrat JM, Wagner PD: Serial rela-
tionships between ventilation-perfusion in-
equality and spirometry in acute severe asthma
requiring hospitalization. Am Rev Respir Dis
1988;137:1055–1061.
Respiration 2005;72:176–181
32 Kohler D: pO2 – Do we treat the right thing?
Respiration 2003;70:455–456.
33 Shann FA, Hart K, Thomas D: Acute lower
respiratory tract infections in children: Possi-
ble criteria for selection of patients for antibiot-
ic therapy and hospital admission. Bull WHO
1984;62:749–753.
34 Leventhal JM: Clinical predictors of pneumo-
nia as a guide to ordering chest roentgeno-
grams. Clin Pediatr 1982;21:730–734.
35 Cherian T, John TJ, Simoes E, Steinhoff MC,
John M: Evaluation of simple clinical signs for
the diagnosis of acute lower respiratory tract
infection. Lancet 1988;ii:125–128.
36 Campbell H, Byass P, Greenwood BM: Simple
clinical signs for diagnosis of acute lower respi-
ratory infections. Lancet 1988;ii:742–743.
37 Campbell H, Byass P, Lamont AC, Forgie IM,
O’Neill KP, Lloyd-Evans N, Greenwood BM:
Assessment of clinical criteria for identification
of severe acute lower respiratory tract infec-
tions in children. Lancet 1989;i:297–299.
38 Margolis P, Gadomski A: Does this infant have
pneumonia? JAMA 1998;279:308–313.
39 Brooks LJ, Cloutier MM, Afshani E: Signifi-
cance of roentgenographic abnormalities in
children hospitalized for asthma. Chest 1982;
82:315–318.
40 Gershel JC, Golman HS, Stein RE, Shelov SP,
Ziprkowski M: The usefulness of chest radio-
graphs in first asthma attacks. N Engl J Med
1983;309:336–339.
41 Pitkin AD, Roberts CM, Wedziche JA: Arteri-
alized earlobe blood gas analysis: An under use
technique. Thorax 1994;49:364–366.
181