Paediatrica Indonesiana

November 

VOLUME 48

NUMBER 6

Original Article

Downes score as a clinical assessment for hypoxemia

in neonates with respiratory distress
Anita Rusmawati, Ekawati L. Haksari, Roni Naning

Abstract
Background Hypoxemia in neonates with clinical respiratory
distress has a high mortality. Downes score is used as an alternative
to evaluate clinical respiratory distress if blood gas analysis
instrument or pulse oxymetry is not available.
Objective To evaluate the validity of Downes score for assessing
hypoxemia in neonates with clinical respiratory distress.
Methods A cross sectional study was carried out on neonates with
FOLQLFDOUHVSLUDWRU\GLVWUHVVKRVSLWDOL]HGDWOHYHODQG1HRQDWDO
Care Unit and in Emergency Room of Dr. Sardjito General
+RVSLWDO <RJ\DNDUWD 'RZQHV VFRUH DQG R[\JHQ VDWXUDWLRQ
measured by a pulse oximetry were compared. Hypoxemia was
GHILQHG DV R[\JHQ VDWXUDWLRQ OHVV WKDQ  LQ WHUP DQG SRVW
WHUP LQIDQWV OHVV WKDQ  LQ SUHWHUP QHRQDWHV RU 'RZQHV
VFRUH  DFFRUGLQJ WR %DVLF (PHUJHQF\ 6HUYLFH 7UDLQLQJ IRU
2EVWHWU\DQG1HRQDWRORJ\321('LQ7KHDFFXUDF\RI
'RZQHVVFRUHLQSUHGLFWLQJK\SR[HPLDZDVDVVHVVHGE\VHQVLWLYLW\
VSHFLILFLW\ SRVLWLYHSUHGLFWLYH YDOXH QHJDWLYHSUHGLFWLYH YDOXH
and likelihood ratio.
Results Eighty nine neonates were evaluated. Downes score had
VHQVLWLYLW\RIVSHFLILFLW\RISRVLWLYHSUHGLFWLYHYDOXHRI
QHJDWLYHSUHGLFWLYHYDOXHRISRVLWLYHOLNHOLKRRGUDWLR
QHJDWLYHOLNHOLKRRGUDWLRSUHYDOHQFHRIDQGSRVW
WHVWSUREDELOLW\RI
Conclusion Downes score can be used as a clinical diagnostic
means for assessing hypoxemia in clinical respiratory distressed
QHRQDWHVZLWKVHQVLWLYLW\&,WRDQGVSHFLILFLW\
 VHQVLWLYLW\  &,  WR  [Paediatr Indones.
2008;48:342-5].
Keywords: Downes score, neonates, hypoxemia,
respiratory distress

SSUR[LPDWHO\  RI QHRQDWDO FDVHV

requiring hospitalization suffer from
respiratory distress with high morbidity
and mortality rates.1 Duke et al LQ3DSXD
1HZ *XLQHD ILQG WKDW  RI K\SR[HPLD RFFXUV LQ
neonates with acute lower respiratory infection (ALRI)
and in those without ALRI. Neonates with hypoxemia
have 3.1 times higher mortality rate. Onyango3 in
.HQ\DREVHUYHGEDELHVDQGFKLOGUHQEHORZWKUHH
\HDUVZLWK$/5,DQGIRXQGWKDWRIWKHPVXIIHUHG
from hypoxemia had a mortality rate of 4.3 times higher
depends on the degree of hypoxemia.
:RUNLQJGLDJQRVLVPXVWEHHVWDEOLVKHGLQWKHILUVW
IHZPLQXWHVDQGDQHDUO\SKDVHFDUHPXVWEHFRQGXFWHG
E\JLYLQJR[\JHQ7KHUHIRUHLWLVQHFHVVDU\WRFRQGXFW
continuous monitoring and evaluation. The majority of
conditions that cause respiratory distress can be avoided
through early detection and control.1
The degree of hypoxemia should be measured
ZLWK EORRG JDV DQDO\VLV %*$ +RZHYHU %*$ LQ
QHRQDWHVFDQFUHDWHFRPSOLFDWLRQVGXHWRSDLQGDPDJH
WRWKHDUWHU\LQFUHDVHGULVNRILQIHFWLRQWKURPERVLV

)URPWKH'HSDUWPHQWRI&KLOG+HDOWK0HGLFDO6FKRRO*DGMDK0DGD
8QLYHUVLW\<RJ\DNDUWD,QGRQHVLD
Request reprint to$QLWD5XVPDZDWL0''HSDUWHPHQWRI&KLOG+HDOWK
0HGLFDO 6FKRRO *DGMDK 0DGD 8QLYHUVLW\ 'U 6DUGMLWR +RVSLWDO -OQ
.HVHKDWDQQR6HNLS8WDUD<RJ\DNDUWD,QGRQHVLD7HO
H[W)D[

342Paediatr Indones, Vol. 48, No. 6, November 2008

Anita Rusmawati et al: Downes score in neonatal respiratory distress

DQGEOHHGLQJ)RUFOLQLFDOSXUSRVHVWKHHYDOXDWLRQRI
hypoxemia can be conducted by using pulse oxymetry.
$VLJQLILFDQWFRUUHODWLRQRIU LVDFKLHYHGWKURXJK
a comparison of BGA saturation and pulse oxymetry
in babies.4 Hay IRXQG D FRUUHODWLRQ RI U  LQ
premature and full-term babies. Another study finds a
VHQVLWLYLW\RIDQGDVSHFLILFLW\RI6+RZHYHU
there is limitation of pulse oxymetry due to the high
SULFHRIWKLVLQVWUXPHQW7KHUHIRUHDFOLQLFDOHYDOXDWLRQ
to determine the presence of hypoxemia at early stage
without supporting tools is very important.
,QWKH3ULQFLSOHVRI%DVLF1HRQDWDO&DUH7UDLQLQJ
&RXUVH a Downes score was used as a basis for
clinically evaluating respiratory problems. A modified
Downes score was used in Basic Emergency Service
7UDLQLQJIRU2EVWHWU\DQG1HRQDWRORJ\321('11
LQ  'RZQHV stated that if Downes score is
HYDOXDWHGLQWHQVLYHO\HYHU\PLQXWHVLWZRXOG
be very useful for evaluating the progress of respiratory
GLVWUHVV 9DULRXV VRXUFHV KRZHYHU VKRZ GLIIHUHQW
interpretations. Downes score by USAID Indonesia
VWDWHG WKDW D VFRUH RI  PHDQV QR RI UHVSLUDWRU\
GLVWUHVVVKRZVWKHSUHVHQFHRIUHVSLUDWRU\GLVWUHVV
ZKLOH!VKRZVWKHWKUHDWRIUHVSLUDWRU\IDLOXUH7KH
absence or presence of cyanosis is determined by
WKHPLQLPXPR[\JHQVLPLODUWRWKDWIRXQGE\
Downes. According to the Russian Federation-United
States of America13 DQG 321(' PLOG UHVSLUDWRU\
GLVWUHVVRFFXUVLIWKHVFRUHLVPRGHUDWHUHVSLUDWRU\
GLVWUHVVLILWLVDQGVHYHUHUHVSLUDWRU\GLVWUHVVLIWKH
VFRUH LV   ,Q F\DQRVLV HYDOXDWLRQ E\ 321(' LQ
JLYLQJR[\JHQRUQRWDQGGHWHUPLQLQJZKHWKHU
the cyanosis persists or disappears is conducted instead
RIJLYLQJPLQLPXPR[\JHQ
The aim of this study was to determine the
validity of Downes score for assessing hyperoxaemia
in neonates with respiratory distress.

Methods
:H SHUIRUPHG D GLDJQRVWLF WHVW VWXG\ WR FRPSDUH
Downes score to oxygen saturation measured by pulse
R[\PHWU\7KLVVWXG\ZDVFRQGXFWHGLQWKHOHYHOVDQG
3 neonatal care unit and the emergency unit of the Dr.
Sardjito Hospital. The inclusion criteria were neonates
with signs of respiratory distress and an approval from
SDUHQWV7KHH[FOXVLRQFULWHULDZHUHIHYHUVKRFNVHYHUH

DQHPLDDEQRUPDOFHQWUDOQHUYHV\VWHPSRO\F\WHPLD
and subjects who were not using the breathing support
equipment. This research had been approved by the
Medical Research Ethics Committee and the Medical
School of Gadjah Mada University.
All subjects were evaluated using the pulse
R[\PHWU\DQGPRGLILHG'RZQHVVFRUHE\321('DW
the same time. The subjects were grouped into those
with hypoxemia and those without hypoxemia based
on both systems. Hypoxemia was considered present
LI6S2ZDVLQIXOOWHUPQHRQDWHVZHHN
JHVWDWLRQDODJHDQGLQSUHWHUPEDELHV%DVHG
RQ'RZQHVVFRUH the presence of hypoxemia was
YDOXHGLIWKHVFRUHZDV
The sample size was calculated based on
 FRQILGHQFH LQWHUYDO DEVROXWH DFFXUDF\ 
DQG VHQVLWLYLW\ RI  6XEMHFWV ZHUH UHFUXLWHG
consecutively to a neonate was identified as suffering
UHVSLUDWRU\ GLVWUHVV LI KHVKH SRVVHVVHG RQH RU PRUH
VLJQVRIQDVDOIODUHUHWUDFWLRQWDFK\SQHDJUXQWLQJ
or cyanosis. Cyanosis was evaluated by observing
WKHEOXLVKQHVVRIWKHOLSVWRQJXHDQGPRXWKPXFRVD
under sufficient lighting. Mild retraction was defined
as there was subcostal retraction and it was still
apparent when the child made a movement. Severe
retraction was defined as there were intercostal
UHWUDFWLRQ VXSUDVWHUQDO RU SDUDGR[ EUHDWKLQJ VHH
saw). Grunting was observed and listened at the end
of expiration. Breathing sound was listened from the
WRSPHGLDOFKHVWRQWKHD[LOODU\OLQHWRHYDOXDWHDLU
flow into the lung.
Oxygen saturation was measured using the
Nellcor pulse oxymetry by placing it on the left
RU ULJKW IRRW IRU  VHFRQGV DQG WKH YDOXH RQ WKH
PRQLWRUGLGQRWFKDQJHIRUDPLQLPXPRIVHFRQGV
6HYHUHDQHPLDZDVGLDJQRVHGDVKHPRJORELQOHYHO
JG/DQGSRO\F\WHPLDZKHQYHLQKDHPDWRFULWOHYHO
ZDV ! :H FRQVLGHUHG IHYHU ZKHQ WKH D[LOODU
WHPSHUDWXUH ZDV !& PHDVXUHG E\ D GLJLWDO
thermometer for one minute.

Results
7KHUH ZHUH  QHRQDWHV UHFUXLWHG IURP 2FWREHU
nd  WR 'HFHPEHU th  7KH H[DPLQHUV
DJUHHPHQW WHVW VKRZHG D .DSSD YDOXH RI 
indicating a good agreement rate.

Paediatr Indones, Vol. 48, No. 6, November 2008343

Anita Rusmawati et al: Downes score in neonatal respiratory distress

The characteristics of study subjects are shown

in Table 1.
,QWKLVVWXG\WKHK\SR[HPLDOLPLWIURP'RZQHV
score was identified by determining the cut off point
in Receiver Operator Curve (ROC) at Downes score
RITable 2 shows predictions of hypoxemia using
comparative cut off point from Downes score.

Total
30 (54 )
26 (46)

Sensitivity
97
97
97
88
44
31
9
3

95%
CI
91 to 100
91 to 100
91 to 100
79 to 99
27 to 61
15 to 47
(-1) to 19
(-1) to 9

Value

95% CI

88
81
72
92
4.53
0.15
36

79 to 99
70 to 91
58 to 86
84 to 100
2.62 to 7.83
0.06 to 0.39
26 to 46

25 (45)
31 (55)
21 (38 )
35 (62)

Discussion

38 (68 )
18 (32 )
140 (13.7)
36.7 (0.51)
17.1 (2.25)
6.2 (6.03)

Table 2. Diagnostic means of hypoxemia using cut-off point from

Downes score.
Downes
Score
2
3
4
5
6
7
8
9

Table 3. Prediction of hypoxemia in neonates by using Downes

score

Sensitivity (%)
5RGEKEKV[

Positive Predictive Value (%)
Negative Predictive Value (%)
Positive Likelihood Ratio
Negative Likelihood Ratio
Pretest probability (%)

Table 1. Characteristics of research subjects

Characteristics
1. Gender (%):
Male
Female
2. Weight at birth (%):
< 2500 g
I
3. Gestational age (%):
< 37 weeks
YGGMU
4. Method of delivery (%):
Spontaneous
Cesarean
5. Heart rate per min., mean (SD)
6. Temperature, 0C, mean (SD)
7. Hemoglobin, g/dL, mean (SD)
8. Age , yrs, mean (SD)

'RZQHVVFRUHFRXOGGHWHFWK\SR[HPLDLQ
neonates with respiratory distress (Table 3).
The results of diagnostic test based on Downes
score parameters are shown in Table 4.

5RGEKEKV[
28
53
65
81
98
98
98
98

95%
CI
16 to 40
40 to 66
53 to 77
70 to 91
95 to 100
95 to 100
95 to 100
95 to 100

Lack of control in evaluating neonates with respiratory

distress can lead to late detection of hypoxemia.
Based on the study conducted by Downes a linear
correlation was found between Downes score and
3D2 +\SR[HPLD RFFXUUHGLQ  PP+J 3D2 and
'RZQHVVFRUHRI,QWKLVVWXG\WKHFXWRIISRLQWIRU
K\SR[HPLDLQ52&ZDVDW'RZQHVVFRUHRIZKLFK
was consistent with that found by the Downes original
research.
The results of this study showed a sensitivity
6QDQGVSHFLILFLW\6S7KHFRPELQDWLRQ
of several parallel clinical parameters showed that
it could increase sensitivity with good specificity.
7KHUHIRUH 'RZQHV VFRUH FDQ EH XVHG WR HYDOXDWH
hypoxemia accurately in neonates with respiratory

Table 4. The results of diagnostic test based on Downes score parameter

Parameter

Sn (%)

Sp (%)

Respiratory rate
(95% CI)
Cyanosis
(95% CI)
Retraction
(95% CI)
Grunting
(95% CI)
Respiratory sound
(95% CI)

94
(85 to 100)
56
(39 to 73)
97
(91 to 100)
38
(21 to 54)
88
(76 to 99)

9
(1 to 16)
82
(73 to 92)
35
(23 to 47)
95
(89 to 100)
63
(51 to 76)

PPV (%)

NPV (%)

PLR (%)

37
71
1.03
(26 to 47) (38 to 100) (0.91 to 0.16)
64
77
3.21
(47 to 82)
(66 to 88) (1.69 to 6.08)
36
46
1.49
(26 to 46)
(34 to 57) (1.22 to 1.82)
80
73
7.13
(60 to 100) (63 to 83) (2.17 to 23.39)
57
90
2.38
(43 to 71)
(81 to 99) (1.65 to 3.42)

344Paediatr Indones, Vol. 48, No. 6, November 2008

NLR (%)
0,71
(0.15 to 3.46)
0.53
(0.35 to 0.80)
0.09
(0.09 to 0.63)
0.66
(0.50 to 0.87)
0.20
(0.08 to 0.51)

Post Test
Probability
37
64
46
80
57

Anita Rusmawati et al: Downes score in neonatal respiratory distress

distress and increase the probability of diagnosing

K\SR[HPLDWLPHVFRPSDUHGWRQHRQDWHVZLWKRXW
hypoxemia.
Evaluation of single clinical signs of neonatal
hypoxemia tends to provide different results. Duke
et al found that cyanosis had good sensitivity and
specivicity. Retraction and grunting showed low
VHQVLWLYLW\ ZKLOH F\DQRVLV DQG JUXQWLQJ VKRZHG
high specivicity. The study conducted by Onyango
et al 3 RQ EDELHV  PRQWKV ZLWK $/5, VKRZHG
that tachypnea and retraction correlated with
KLJK VHQVLWLYLW\ JUXQWLQJ ZLWK ORZ VHQVLWLYLW\ DQG
VSHFLILFLW\DQGF\DQRVLVZLWKKLJKVSHFLILFLW\,QWKLV
VWXG\DQDQDO\VLVRQHDFK'RZQHVVFRUHSDUDPHWHU
VKRZHG WKDW WDFK\SQHD UHVSLUDWRU\ UDWH ![
PLQXWHUHWUDFWLRQDQGUHVSLUDWRU\VRXQGSRVVHVVHG
KLJK VHQVLWLYLW\ ZLWK ORZ VSHFLILFLW\ &\DQRVLV
and grunting resulted in low sensitivity with high
specificity.
This study showed that a scoring system which
requires only basic skills and minimum training
for evaluating clinical signs can be used to assess
K\SR[HPLDDQGWKHUHIRUHFDQDOVREHXVHGDVDEDVLV
for giving oxygen or referrals.
The limitation of this study was that it was not
EOLQGHGVLQFHLWZLOOQRWEHHWKLFDOLIWKHH[DPLQHUVGR
not assess clinical and pulse oxymetry in respiratory
distress patients. The size of the sample was not based
on the prevalence of hypoxemia in Indonesia. The
result of prevalence in this research was estimated
DW  ZKHUHDV WKH SUHYDOHQFH RI K\SR[HPLD LQ
neonates based on the study conducted by Duke
used as sample size calculation for this research was
7KXVRXUVDPSOHGLGQRWWRWDOO\UHSUHVHQWWKH
population of Indonesia. This was proven by the wide
confidence interval. It was conducted in a tertiary
KHDOWKFDUHFHQWUHDQGPD\WKHUHIRUHOHDGWRDUHIHUUDO
bias.
,QFRQFOXVLRQ'RZQHVVFRUHFDQEHXVHGDVD
clinical diagnostic means for assessing hypoxemia in
QHRQDWHVZLWKUHVSLUDWRU\GLVWUHVVZLWKVHQVLWLYLW\
DQGVSHFLILFLW\

References
 0DWKDL665DMX8.DQLWNDU00DQDJHPHQWRIUHVSLUDWRU\
GLVWUHVVLQWKHQHZERUQ0-$),
 'XNH7%ODVFKNH$-6LDOLV6%RQNRZVN\-/Hypoxaemia in
acute respiratory and in non-respiratory illnesses in neonates
and children in a developing country. Arch Dis Child.

 2Q\DQJR)(6WHLQKRII0&:DULXD60XVLD-.LWRQ\L-
Hypoxemia in young children with acute lower respiratory
LQIHFWLRQ%0-
 -HQQLV 06 3HDERG\ -/ 3XOVH R[LPHWU\  DQ DOWHUQDWLYH
method for the assessment of oxygenation in newborn.
3HGLDWULFV
 +D\::%URFNZD\-(\]DJXLUUH1HRQDWDOSXOVHR[LPHWU\
DFFXUDF\DQGUHOLDELOLW\3HGLDWULFV
 /HH::0D\EHUU\.&UDSR5-HQVHQ5/7KHDFFXUDF\RI
SXOVHR[LPHWU\LQWKHHPHUJHQF\GHSDUWPHQW$P-(PHUJ
0HG
 %DVQHW6$GKLNDUL5*XUXQJ&+\SR[HPLDLQFKLOGUHQZLWK
SQHXPRQLD DQG LWV FOLQLFDO SUHGLFWRUV ,QGLDQ - 3HGLDWU

 /RGKD5%KDGDXULD3.XWWLNDW$3XUDQLN0*XSWD63DQGH\50.DEUD6.Can clinical symptoms or signs accurately
predict hypoxemia in children with acute lower respiratory
WUDFWLQIHFWLRQV,QGLDQ3HGLDWULFV
 8VHQ 6 :HEHUW 0 0XOKROODQG . &OLQLFDO SUHGLFWRUV RI
hypoxemia in Gambian children with acute lower respiratory
WUDFWLQIHFWLRQSURVSHFWLYHFRKRUWVWXG\%0-
 $O\+1HRQDWDOUHVSLUDWRU\GLVRUGHU,Q86$,',QGRQHVLD
+63 'HSDUWHPHQ .HVHKDWDQ 5, ,'$, 3ULQFLSOHV RI
EDVLF QHRQDWDO FDUH WUDLQLQJ FRXUVH -DNDUWD 'HSDUWHPHQ
.HVHKDWDQ
 321(' 3HODWLKDQ 3HOD\DQDQ 2EVWHWUL GDQ 1HRQDWDO
(PHUJHQVL'DVDU-DNDUWD'HSDUWHPHQ.HVHKDWDQ
 'RZQHV -- 9LG\DVDJDU ' 0RUURZ *0 %RJJV 75 5HVSLUDWRU\ GLVWUHVV V\QGURPH RI QHZERUQ LQIDQW &OLQ 3HGLDWU

13. Russian Federation-United States of America. (Re) Designing
the system of care for neonates suffering from respiratory
GLVWUHVVV\QGURPH5XVVLD7KH86$5XVVLD-RLQW&RPPLVVLRQ
RQ(FRQRPLFDQG7HFKQRORJLFDO&RRSHUDWLRQ

Paediatr Indones, Vol. 48, No. 6, November 2008345