-
Arch. Dis. Childh., 1968, 43, 155.
Effects of Intragastric Sodium Bicarbonate in Infantswith
Respiratory Distress
MARGARET E. R. STONEMAN and ROSALIND M. OWENSFrom the Paediatric
Research Unit and Special Care Baby Unit, Exeter City Hospital
Since Usher reported a significant fall in mortalityin infants
with respiratory distress who were treatedwith intravenous glucose
and sodium bicarbonate(Usher, 1959, 1960, 1961, 1963), this form
oftreatment has been widely adopted (Warley andGairdner, 1962;
Hutchison et al., 1962, 1964;Gairdner, 1965). Intragastric glucose
and sodiumbicarbonate have also been used (Abraham andBrown,
1967).
Feeding of healthy premature babies, beginningat the age of 3 or
4 hours, was being carried out withsuccess in the Special Care Baby
Unit of ExeterCity Hospital when facilities for
biochemicalinvestigation of cases of respiratory distress
becameavailable in 1963. A decision was, therefore, madeto
undertake a trial of treatment of cases of respira-tory distress by
means of intragastric sodium bicar-bonate. Our experience has shown
that the valueof this treatment is limited, and that in some
casesit may possibly be harmful. This form of treat-ment has
therefore been discontinued in allexcept a minority of cases. The
results of ourobservations on babies treated with and with-out
intragastric sodium bicarbonate are reportedin the present paper.
The indications and contra-indications for such treatment are
discussed.
Material and MethodsAll the patients were admitted to the
Special Care
Baby Unit of Exeter City Hospital between August 1963and July
1967. A diagnosis of respiratory distress wasmade if two or more of
the following criteria werepresent after the age of 1 hour: (1)
grunting respiration;(2) sternal or intercostal recession; (3)
cyanosis; (4)respiratory rate of 65 or more, on more than one
recor-ding at 3-hourly intervals.Of 390 babies who fulfilled these
criteria, 24, whose
respiratory distress was due to miscellaneous causes suchas
diaphragmatic hernia, pneumothorax, and massivemeconium aspiration
have been excluded, as have 18babies who died from causes other
than respiratorydistress. It was also necessary to exclude the
effects of
Received October 16, 1967.
intermittent positive pressure ventilation (IPPV) onbiochemical
findings and on mortality. For this reason16 further cases have
been excluded: 7 survivors,because it was considered that their
survival wouldhave been extremely unlikely without IPPV, and 9
fatalcases in whom IPPV had been started before
biochemicalinvestigations were carried out. Of the remaining
332babies, 225 had estimations of capillary blood pH, Pco,2and
standard bicarbonate, using the Astrup micro-methods, started
before the age of 12 hours, and theseform the basis of our present
study. Of these 225cases, 132 were born in Exeter City Hospital,
and 93were admitted from outside before the age of 12 hours.In
those cases that were treated with IPPV the resultsof biochemical
investigations carried out after startingIPPV have been
excluded.
All infants were nursed in incubators, the temperature,humidity,
and oxygen concentrations being adjusted,with the aim of
maintaining rectal temperatures of36-37° C. (97-98° F.). High
oxygen concentrations wereused whenever these were necessary to
overcomecyanosis.Tube feeding was begun at the age of 3-4 hours,
in
amounts of 65 ml./kg. daily, and gradually increasedaccording to
tolerance and appetite. Feeds were given3-hourly, and indwelling
gastric tubes were rarely used.During the earlier period feeds were
of 10% dextrosewith sodium bicarbonate in amounts varying from 10
to25 mEq/100 ml., according to the capillary blood pH,as
recommended by Usher for intravenous therapy.Administration of
sodium bicarbonate in the first fewcases was continued until the pH
rose to 7 3, but inlater cases it was discontinued when the
standardbicarbonate reached 18 mEq/l., irrespective of thedegree of
respiratory acidosis. In the later part of thestudy infants were
given one feed of 10% dextrose,followed by expressed breast
milk.
Infants whose condition deteriorated seriously wereintubated and
given IPPV.
ResultsRadiological examination of the chest was carried
out in most cases and the findings formed a usefulbasis for
classification, as they were found to beclosely related to the
clinical course and prognosis.The first group comprised those with
normal
155
on April 7, 2021 by guest. P
rotected by copyright.http://adc.bm
j.com/
Arch D
is Child: first published as 10.1136/adc.43.228.155 on 1 A
pril 1968. Dow
nloaded from
http://adc.bmj.com/
-
Stoneman and Owensradiological appearances, and had no
mortality.The second group will be referred to as havinglocalized
radiological abnormality, with appearancesof localized atelectasis
or aspiration pneumonia, andnormal aeration of the remainder of the
lung.Cases in this group also had a good prognosis, theonly two
deaths being in infants of less than 28weeks' gestation. The third
group will be referredto as having generalized radiological
abnormality;the appearances varied from miliary mottling ofboth
lung fields, to complete opacity with appearancesof an 'air
bronchogram'. This was the only groupwith a high mortality. In the
fourth group whichcomprised cases of mild respiratory distress
whowere not examined radiologically, there was nomortality. The
number of cases and the mortalityin each group are shown in the
Table.
TABLENumber of Cases and Mortality in Infants withRespiratory
Distress, Grouped according to Radio-
logical Appearances
Appearance of Chest No. of No. of MortalityX-ray Cases Deaths*
(%)
Normal .48 0 0Localized abnormality 39 2 (2) 5Generalized
abnormality.. ill 54 (7) 49Not examined radio-
logically. 27 0 0
Total .225 56 (9) 25
* Figures in parentheses indicate the number of fatal cases of
lessthan 28 weeks' gestation.
Incidence and degree of initial metabolicacidosis. In most cases
the initial standard bicar-bonate estimation was carried out before
the age of4 hours and in all cases before 12 hours. Theincidence
and degree of initial metabolic acidosisare shown in Fig. 1.The
incidence of severe metabolic acidosis was
higher in those cases with generalized radiologicalabnormality,
but even in this group 42% of thecases had initial standard
bicarbonate levels above18 mEq/l., and of the total number of 225
babies,53% had initial standard bicarbonate levels above18
mEq/l.
Duration of metabolic acidosis in cases nottreated with sodium
bicarbonate. Fig. 2 showsthe percentage of patients with initial
standardbicarbonate levels below 18 mEq/l., whose
standardbicarbonate had risen to 18 mEq/l. or more at agesup to 60
hours without alkali treatment.
_~ Standard bicorbonate
-
Effects of Intragastric Sodium Bicarbonate in Infants with
Respiratory Distressacidosis, whereas in those with generalized
radiolo-gical abnormality recovery took much longer.
Effects of Intragastric Sodium Bicarbonate(A) Cases with
generalized radiological ab-
normality. Any beneficial effect of intragastricsodium
bicarbonate should be demonstrable in thegroup with generalized
radiological abnormality, asthis is the only group
withprolongedmetabolic acido-sisand a high mortality rate. Of the
111 babies in thisgroup, 7 were treated with THAM, 3 were only
givenintragastric sodium bicarbonate at a late stage, and1 died
within 3 hours of the initial estimation ofstandard bicarbonate,
before treatment could havehad any effect. After exclusion of these
11 babies,100 remain, of whom 31 were treated with intra-gastric
sodium bicarbonate, with 18 deaths; and 69without alkalis, with 27
deaths.
(1) Effect of intragastric sodium bicarbonate onstandard
bicarbonate levels. Of the patients withgeneralized radiological
abnormality, 56 had initialstandard bicarbonate levels below 18
mEq/l. Ofthese, 26 were treated with intragastric sodiumbicarbonate
and 30 without. The percentages ofpatients in each group whose
standard bicarbonatehad risen to 18 mEq/l. or more at ages up to
60hours are shown in Fig. 3.
It will be seen that the standard bicarbonate levelsrose more
quickly and in a higher percentage ofcases in those given
intragastric sodium bicarbonatethan in those given no alkali
treatment.
(2) Effect of intragastric sodium bicarbonate onPco2 levels. In
the same 56 babies with generalizedradiological abnormality, the
percentages in thegroups treated with and without intragastric
sodiumbicarbonate whose Pco2 levels had fallen below70 mm. Hg and
did not rise again are shown inFig. 4. (One baby in whom
estimations werediscontinued after 28 hours, before the Pco2
hadfallen below 70 mm., has been omitted.)
It will be seen that the percentage of cases withprolonged
severe respiratory acidosis was higher inthose treated with
intragastric sodium bicarbonate.
In order to eliminate any possible errors due todifferences in
initial Pco2 or differences in thenumbers of fatal cases in the two
groups, a moredetailed study of the behaviour of the Pco2 levelsin
two groups of suvivors with generalized radio-logical abnormality
was carried out. Only survivingbabies with a large number of
readings were used.There were 7 such babies who were given
intra-gastric sodium bicarbonate and on the average theyhad a
fairly high initial Pco2. Therefore, 7
vb
0
0
100
90-
80
70-6o-
50-40-
30-20-
10-
00 10 20 30 40 50 6OAge (hours)
FIG. 3.-Percentage of patients with generalized radio-logical
abnormality and initial standard bicarbonate below18 mEqll., whose
standard bicarbonate had risen to 18mEqll. at different ages. *, 26
cases receiving intragastricsodium bicarbonate; 0, 30 cases not
given sodium bicar-
bonate.
untreated babies who were found to have a similarlyhigh average
initial Pco2 level, were selected. Thereadings in every four-hour
period of 48 hours wereaveraged in each group and the
arithmeticalregressions calculated (Simpson, Roe, and
Lewontin,1960). Results are shown in Fig. 5.The resulting lines
show a tendency for untreated
babies to have a falling Pco2 in the first 48 hours
100
90 -
80 -
70-
6o -0u 50-
.60 4n-,
0 10 20 30 40 50 60Age (hours)
FIG. 4.-Percentage of patients with generalized radio-logical
abnormality and initial standard bicarbonate below18 mEqll., whose
PCO2 hadfallen below 70 mm. at differentages. *, 26 cases receiving
intragastric sodium bicar-
bonate; 0, 29 cases not given sodium bicarbonate.
157
%a
on April 7, 2021 by guest. P
rotected by copyright.http://adc.bm
j.com/
Arch D
is Child: first published as 10.1136/adc.43.228.155 on 1 A
pril 1968. Dow
nloaded from
http://adc.bmj.com/
-
Stoneman and Owens
100-
90
80-
70 -
50 -
40
A
B
2 6 10 14 18 22 26 30 34 ~8 42 46Age (hours)
FIG. 5.-Change in Pco.0 levels as a function of time(arithmetic
regression) in 14 cases of respiratory distresswith generalized
radiological abnormality, treated withand without intragastric
sodium bicarbonate. A, 7 casesreceiving intragastric sodium
bicarbonate; B, 7 cases not
receiving sodium bicarbonate.
of life while those babies treated with intragastricsodium
bicarbonate tended to retain their respiratoryacidosis.
(3) Relation between total dosage of intragastricsodium
bicarbonate and highest Pco2 levels. Therelation between total
dosage of intragastric sodiumbicarbonate and the highest subsequent
Pco2 levelsreached is shown in Fig. 6. The babies weredivided into
4 groups according to total dosage ofintragastric sodium
bicarbonate given per kg. bodyweight. In each group the average was
taken of
150,
140 -
130 -
120
110-
100 -
90 -
80 -
70-
60
(4)(12)
('5)
(69)
0 12Total dose of sodium bicarbonate (mEq/1. per kg.)
FIG. 6.-Highest Pco2 values reached, in relation to totaldosage
of intragastric sodium bicarbonate, in cases withgeneralized
radiological abnormality. The figures inbrackets are the numbers of
cases from which a mean value
was calculated.
the highest Pco2 levels reached after the administra-tion of
intragastric sodium bicarbonate. (Babieson whom estimations of Pco2
were not continuedfor at least 6 hours after the start of feeding,
with orwithout sodium bicarbonate, have been excluded.)
It will be seen that the highest Pco2 levels werefound in those
babies who had the highest totaldosage of sodium bicarbonate.
(4) Effect of intragastric sodium bicarbonate on pH.In the same
56 babies with generalized radiologicalabnormality, the percentages
of infants treated withand without intragastric sodium bicarbonate,
whosepH had risen to 7 2 at ages up to 60 hours, areshown in Fig.
7.
It will be seen that there is little differencebetween the two
groups, as the effect of sodiumbicarbonate in correcting the
metabolic acidosis iscounteracted by the prolongation of
respiratoryacidosis.
(5) Effect of intragastric sodium bicarbonate onmortality.
Mortality was 58% in the cases givenintragastric sodium bicarbonate
and 39% in casesgiven no alkali treatment. Babies born outsidehad a
slightly higher mortality than babies born inExeter City Hospital
but were equally distributedbetween the two treatment groups,
accounting for420 of those given intragastric sodium
bicarbonate,and 43% of those not given alkalis. Mortality washigher
in babies with severe initial metabolicacidosis and also in those
with lower birthweights.They were, therefore. divided into two
weightgroups, those below and those above 1500 g., andeach was then
divided into those with initialstandard bicarbonate levels below
and above 15mEq/1. This gave 4 groups, in each of which
themortality in cases treated with and without intra-gastric sodium
bicarbonate could be compared.The results are shown in Fig. 8.
It will be seen that, whatever the initial standardbicarbonate
level, treatment with intragastricsodium bicarbonate did not reduce
the mortalityin infants of 1500 g. and less. In the higher
weightgroup, mortality in infants with severe metabolicacidosis was
lower in those treated with intragastricsodium bicarbonate, but the
difference is barelysignificant (x2 = 2*042, 0 * 1
-
Effects of Intragastric Sodium Bicarbonate in Infants with
Respiratory Distress
o -d
U 50-0o 40-
30 -
20-
10
0
0 10 20 30 40 50 60
Age (hours)
FIG. 7.-Percentage of cases with generalized
radiologicalabnormality and initial standard bicarbonate below
18mEqll., whose pH had risen to 7 2 at different ages. *,26 cases
receiving intragastric sodium bicarbonate; 0, 30
cases not given sodium bicarbonate.
tion of sodium bicarbonate had no effect on thetime taken for
standard bicarbonate levels to reach18 mEq/l. or more. Recovery
from respiratoryacidosis was also much more rapid in these
infantsthan in those with generalized radiological abnor-mality,
but was slightly delayed in those givenintragastric sodium
bicarbonate as compared withthose not given alkali treatment.
Mortality was nilin infants over 28 weeks' gestation in both
thesegroups, irrespective of the method of treatment.
800- to{) [L490-
70-
b, 60
.r soz: 40
snt.aIs r 15mEq/1.stardbicorboncteBirthweight c 15009.
With sodiumbicarbonate.Wifhoutsodium.bicarbonate
I1SmEq/1.
>1500q.
FIG. 8.-Mortality in cases with generalized
radiologicalabnormality according to birthweight and initial
standardbicarbonate levels. The number of cases is shown in
brackets.
Later Metabolic AcidosisMany infants whose standard bicarbonate
levels
had reached 18 mEq/l. had a later fall in standardbicarbonate
below this level. In the case of infantswho survived, the standard
bicarbonate never fellbelow 15 mEq/l., whereas in fatal cases a
suddenfall to levels below 15 mEq/l. was a frequentoccurrence,
usually following apnoeic attacks.
DiscussionIn the treatment of respiratory distress in the
newborn correction of acidosis by administeringsodium
bicarbonate intravenously has becomecommon practice, and a
reduction in mortality hasbeen reported (Usher, 1963; Hutchison et
al., 1964).A recent comparison of intragastric vs.
intravenousroutes of glucose/bicarbonate administration showeda
comparable rise in pH in the two groups, with alower mortality in
cases treated by the intragastricroute (Abraham and Brown,
1967).
In most reported series, whatever the route ofadministration of
sodium bicarbonate, treatmenthas been continued until pH has
reached normallevels, entailing the induction of a
metabolicalkalosis in order to compensate for the
respiratoryacidosis. The correctness of this principle has yetto be
proven.
In the present study the incidence and degree ofmetabolic
acidosis in a series of infants withrespiratory distress have been
investigated, and theeffects of treatment with or without
intragastricsodium bicarbonate have been compared.
Standardbicarbonate and Pco2 rather than pH levels havebeen
compared in the two groups of cases in orderto distinguish between
the metabolic and respiratorycomponents of the acidosis.
In infants of 28 weeks' gestation or more, whosechest x-rays
were normal or showed only localizedabnormality, recovery from
metabolic acidosis wasrapid and mortality was nil. Treatment
withalkalis is therefore unnecessary in such cases.
Infants whose chest x-rays showed generalizedabnormality had a
higher incidence of severe meta-bolic acidosis initially, but when
cyanosis could berelieved the metabolic acidosis slowly
correcteditself. A more rapid recovery from metabolicacidosis in
many of these infants could be achievedby intragastric
administration ofsodium bicarbonate,but this treatment delayed
recovery from respiratoryacidosis.
Persisting metabolic acidosis in cases of severerespiratory
distress is almost always an indication ofcontinuing hypoxia.
Unless oxygen tensions arebeing monitored frequently, the raising
of standard
159
on April 7, 2021 by guest. P
rotected by copyright.http://adc.bm
j.com/
Arch D
is Child: first published as 10.1136/adc.43.228.155 on 1 A
pril 1968. Dow
nloaded from
http://adc.bmj.com/
-
160 Stoneman and Owensbicarbonate levels by administering sodium
bicar-bonate may prevent the recognition of continuingmetabolic
acidosis as a valuable indication thatoxygen concentrations need to
be increased.
It might be expected that any beneficial effect oftreatment with
intragastric sodium bicarbonate onmortality would be less
noticeable in cases wherethe initial metabolic acidosis was mild.
This hasbeen confirmed in the present study, which hasshown no
reduction in mortality following intra-gastric administration of
sodium bicarbonate incases with generalized radiological
abnormality andinitial standard bicarbonate levels of 15 mEq/l.
ormore. Treatment with intragastric sodium bicar-bonate is not
therefore indicated in this group ofcases.
Mortality was high in those cases with generalizedradiological
abnormality and severe initial metabolicacidosis, whether
intragastric sodium bicarbonatewas given or not. Correction of the
metabolicacidosis by intragastric administration of
sodiumbicarbonate did not reduce the mortality in infantsweighing
1500 g. and less. Although the mortalitywas lower in infants over
1500 g. who were givenintragastric sodium bicarbonate, the
difference wasnot significant.
Since severe metabolic acidosis in cases of respira-tory
distress is indicative of hypoxia and is usuallyaccompanied -by
respiratory acidosis, correction ofmetabolic acidosis alone,
without efforts to relievehypoxia and to improve ventilation would
not beexpected to reduce mortality greatly. Treatmentby means of
IPPV is the rational approach insuch cases, when 100% 02 has failed
to relievehypoxia and metabolic acidosis, and particularlywhen
respiratory acidosis is also severe. Ourexperience with IPPV so far
confirms that of Reidand Tunstall (1966) that in most cases
IPPVcorrects both respiratory and metabolic acidosis,and leads us
to conclude that many of the infantstreated during the earlier part
of this study shouldhave been given IPPV earlier in the course of
theirillness. In a very small minority of cases metabolicacidosis
is not completely relieved by IPPV, and inthese cases alone there
may be an indication fortreatment with intragastric sodium
bicarbonate.
SummaryThe incidence of metabolic acidosis has been
investigated in 225 infants with respiratory distress.In infants
who initially had standard bicarbonate
levels below 18 mEq/l. (less than half the total)spontaneous
recovery from metabolic acidosis wasrapid in those whose chest
radiographs were normalor showed only localized abnormality, but
wasdelayed in many with generalized radiologicalabnormality.
Administration of intragastric sodium bicarbonateto a group of
infants with generalized radiologicalabnormality speeded recovery
from metabolicacidosis, but was often followed by a more severeand
prolonged respiratory acidosis. Mortality wasnot reduced.The value
of treatment with intragastric sodium
bicarbonate in neonatal respiratory distress isstrictly limited.
It may be of some value in thesmall minority of cases in which
metabolic acidosispersists despite treatment by means of
intermittentpositive pressure respiration.
We wish to thank Dr. F. S. W. Brimblecombe andDr. L. Haas for
allowing us to study patients admittedunder their care and for much
helpful advice; Dr. J.Sheach for radiological opinions; and Sister
J. Boxalland the staff of the Special Care Baby Unit for
theirdevoted care of the patients.
This work was made possible by a grant from theSouth Western
Regional Hospital Board, and a ResearchFellowship from Vickers
Ltd.
REFERENCES
Abraham, J. M., and Brown, R. J. K. (1967). Comparison
ofintragastric and intravenous routes of
glucose/bicarbonateadministration in respiratory distress syndrome.
Brit. med J.,3, 640.
Gairdner, D. (1965). Respiratory distress in the newborn.
InRecent Advances in Paediatrics, 3rd ed., p. 54. Ed. by
D.Gairdner.Churchill, London.
Hutchison, J. H., Kerr, M. M., Douglas, T. A., Inall, J. A.,
andCrosbie, J. C. (1964). A therapeutic approach in 100 cases ofthe
respiratory distress syndrome of the newborn infant.Pediatrics, 33,
956.
-, -, McPhail, M. F. M., Douglas, T. A., Smith, G., Norman,J.
N., and Bates, E. H. (1962). Studies in the treatment of
thepulmonary syndrome of the newborn. Lancet, 2, 465.
Reid, D. H. S., and Tunstall, M. E. (1966). The respiratory
distresssyndrome of the newborn. Anaesthesia, 21, 72.
Simpson, G. G., Roe, A., and Lewontin, R. C. (1960).
QuantitativeZoology. Harcourt, Brace, New York.
Usher, R. (1959). The respiratory distress syndrome
ofprematurity.I. Changes in potassium in the serum and the
electrocardio-gram and effects of therapy. Pediatrics, 24,
562.(1960). Management of metabolic changes in the respiratory
distress syndrome of prematurity. Amer. J. Dis. Child.,
100,485.- (1961). The respiratory distress syndrome of
prematurity.
Clinical and therapeutic aspects. Pediat. Clin. N. Amer.,
8,525.
- (1963). Reduction of mortality from respiratory
distresssyndrome of prematurity with early administration of
intra-venous glucose and sodium bicarbonate. Pediatrics, 32,
966.
Warley, M. A., and Gairdner, D. (1962). Respiratory
distresssyndrome of the newborn-principles in treatment. Arch.Dis.
Childh., 37, 455.
on April 7, 2021 by guest. P
rotected by copyright.http://adc.bm
j.com/
Arch D
is Child: first published as 10.1136/adc.43.228.155 on 1 A
pril 1968. Dow
nloaded from
http://adc.bmj.com/