Timing of Surgery for Congenital Diaphragmatic Hernia:
Is Emergency Operation Necessary?
By Jacob C. Langer, Robert M. Filler, Desmond J. Bohn, Barry Shandling, Sigmund H. Ein,
David E. Wesson, and Riccardo A. Superina
Toronto, Ontario
9 Congenital diaphragmatic hernia (CDHI is considered
by most researchers to be a surgical emergency. However,
early repair does not necessarily improve respiratory func-
tion or reverse fetal circulation, and many patients deterio-
rate postoperatively. As a result, in 1985, w e began to
employ a protocol in which surgery w a s delayed until the
PCO z was maintained below 40 and the child was hemody-
namically stable; children in whom these criteria could not
be achieved died without surgical repair. Sixty-one consec-
utive infants with CDH w e r e managed over 4 years; 31
from 1983 to 1984 (group 1) and 30 from 1985 to 1986
(group 2). The groups w e r e similar with respect to sex,
side of the defect, birth weight, gestational age, incidence
of pneumothorax, and blood gases. High frequency oscilla-
tion was used with increasing frequency during the study
period, for patients with refractory hypercarbia (13% in
group 1, 30% in group 2). All patients w e r e initially para-
lyzed and ventilated. Mean time from admission to surgery
w a s 4.1 hours in group 1 and 24.4 hours in group 2
(P < .05). In group 1, 87% of patients had surgical repair
(77% within eight hours of admission, 10% after eight
hours), and in group 2 only 70% of patients had surgery
(10% within eight hours, 60% after eight hours). All
patients who w e r e not operated on died. Overall mortality
w as 5 6 % in group 1 and 5 0 % in group 2; this difference was
not statistically significant. These data indicate that our
current approach has not increased overall mortality. We
believe that early repair in the face of labile respiratory and
hemodynamic function may be harmful, and that delayed
operation may allow patients with a borderline prognosis
t o survive. For these reasons w e conclude that emergency
surgery is not necessary, and that repair should be done
only when the patient has been satisfactorily stabilized.
9 1 9 8 8 b y Grune R, S t r a t t o n , Inc.
INDEX WORDS: Congenital diaphragmatic hernia; high
frequency oscillation.
ONGENITAL diaphragmatic hernia (CDH)
C through the foramen of Bochdalek usually pres-
ents in the newborn period with severe respiratory
distress. Babies presenting within 12 hours of birth
develop high pulmonary vascular resistance with per-
sistent fetal circulation, hypoxia, hypercarbia, and
acidosis. This, in turn, leads to low cardiac output,
worsening of the pulmonary hypertension, and a
vicious circle which leads to death in approximately
50% of cases. 1'2
Therapy is first aimed at optimizing ventilation and
oxygenation, and consists of paralysis, intubation, and
mechanical ventilation. Pharmacologic agents are
often used in an attempt to decrease pulmonary vascu-
lar resistance and improve cardiac output. In some
Journal of PediatricSurgery, Vo123, No 8 (August), 1988: pp 731-734
centers, infants who deteriorate in the postoperative
period with right-to-left shunting are placed on extra-
corporeal membrane oxgenation (EMCO) in an
attempt to improve oxygenation. However, one of the
basic tenets in both the literature and the textbooks has
been that CDH in the newborn is a surgical emer-
gency.
We observed in our patients that emergency repair
was often followed by deterioration rather than
improvement in clinical status, and we began to ques-
tion the wisdom of early operation. As a result, the
approach to these patients in our institution changed at
the end of 1984, so that patients were not operated on
urgently, but only when their respiratory and hemody-
namic status had been stabilized. This report docu-
ments the results of treatment in a consecutive series of
patients surrounding this change in approach.
MATERIALS AND METHODS
Sixty-one consecutive patients with CDH, presenting in the first
12 hours of life, were managed at the Hospital for Sick Children
between January 1983 and October 1986. Thirty-one patients were
seen in 1983 and 1984 (group 1) and 30 were seen in 1985 and 1986
(group 2). The two groups were similar with respect to sex,
gestational age, side of defect, birth weight, age at diagnosis,
intubation, and timing of transfer to our center (Table 1).
The principles of initial management were constant during the
entire study period. All patients were paralyzed and ventilated,
aiming to keep the PCO2 below 40 mmHg and the preductal PO 2
greater than 60 mmHg. Arterial lines, urinary catheters, and central
venous catheters were placed in all patients. Inotropes and pulmo-
nary vasodilators were used to increase cardiac output and decrease
right-to-left shunting, although the specific drugs varied over time
(Table 2).
All patients were initially placed on a pressure-controlled ventila-
tor; rates and pressures were adjusted to bring the PCO 2 below 40
mmHg. Toward the end of 1984, high-frequency oscillation (HFO) 3
became available for patients with persistent hypercarbia despite
maximum conventional ventilation. Four patients in group 1 (13%)
and 9 in group 2 (30%) required HFO preoperatively.
From the Department of Surgery and Intensive Care Unit,
Hospital for Sick Children, Toronto, Ontario.
Presented at the 19th Annual Meeting of the Canadian Associa-
tion of Paediatric Surgeons, Winnipeg, Manitoba, Canada, Septem-
ber 9-12, t987.
Address reprint requests to Robert M. Filler, MD, Surgeon-
in-Chief Hospital for Sick Children, 555 University Ave, Toronto,
Ontario, Canada M5G 1)(8.
9 1988 by Grune & Stratton, Inc.
0022-3468/88/2308-0009503.00/0
731
732 LANCER ET AL

100
Table 1. Comparison of Groups
surgery<8 hrs
Characteristic Group 1 Group2 P
Gestational age (wk) 39.3 • 2.7 39.3 _+ 2.7 NS ~ l surgery>8 hrs
Birth weight (kg) 3.15 • 0.67 3.19 • 0.61 NS 50
no surgical repair
Sex (% male) 52 57 NS (all died)
Side (% left) 87 77 NS
Age at diagnosis (h) 1.67 • 2.4 2.05 +- 3.1 NS
Age at intubation (h) 1.24 • 2.1 1.80 • 2.9 NS
Age at transfer (h) 4.27 • 4.1 3.77 • 3.1 NS

Extracorporeal membrane oxygenation (ECMO) was not used in
any of the patients during the course of study.
Forty-eight patients underwent surgical repair. The approach was
transabdominal in 47 and transthoracic in one. A standard closure of
the defect was done in 36 patients, and either a flap of abdominal
muscle or a Marlex patch (Bard Cardiosurgery, CR Bard, Ontario)
was used in twelve [eight in group 1 (30%) v four in group 2 (19%)].
Postoperative chest tubes were used in 37 patients [23 in group 1
(85%) v 14 in group 2 (67%)].
The mean time from admission to operation was 4.1 hours (range,
0.5 to 24 hours) in group 1, and 24.4 hours (range, three to 108
hours) in Group 2 (P < .001, Student's t test).
Figure 1 shows the distribution of patients within each group,
according to the timing of surgery. In group 1, 77% were operated on
within eight hours of admission and 10% after eight hours. In group
2 only 10% were operated on within eight hours, and 60% after eight
hours. Of particular note, 13% of patients in group 1 and 30% of
patients in group 2 were never operated on; all these patients died.
RESULTS
Prognostic Factors
Preoperative respiratory function was assessed by
recording the P O E and P C O 2 upon admission, as well as
the worst value obtained preoperatively (Table 3).
There were no significant differences between groups
with respect to these parameters.
The incidence of pneumothorax 4 is shown in Table 4.
Preoperative pneumothorax was more common in
group 2 and postoperative pneumothorax was more
common in group 1; these differences likely resulted
from the fact that group 2 patients were generally
operated on later. The overall incidence of pneumotho-
rax was not significantly different between the
groups.
Ventilatory parameters such as airway pressure,
rate, and response of P C O 2 t o ventilation were difficult
to compare between groups because of the greater use
of HFO in group 2. However, of the 13 patients treated
with HFO, all but two died from increasing hypoxia,
group 1 group 2
(1983-84) (1985-86)
Fig 1. Timing of surgery.
despite improvement in PCO2 and pH. The two surviv-
ing patients were both in group 2, and were operated on
very late (94 and 108 hours after admission).
Complications and Mortality
Postoperative complications were infrequent, and
are summarized in Table 5.
Overall mortality for the entire group was 58% in
group 1 and 50% in group 2 (not significant at the 0.05
level by x 2 analysis).
Autopsies were performed on 20 of the 32 patients
who died. All of these patients had bilateral pulmonary
hypoplasia, which was always worse on the side
affected by the hernia.
DISCUSSION
From the earliest descriptions by Gross, 5 to the
current textbooks, 6 emergency repair has been the
cornerstone of the recommended management of
CDH. The rationale of this approach is to remove the
abdominal viscera from the chest, allowing for expan-
sion of the lung on the involved side. However, it has
never been proven that emergency repair of the defect
actually improves survival. In fact, there are a number
of observations that suggest emergency repair of the
hernia is unnecessary, and may even be more harmful
than beneficial in the early period of respiratory and
hemodynamic instability.
The first observation is that the abdominal viscera
appear to move easily from the chest to the abdomen.
This has been noted in utero by Adzick et al. 7 We have
Table 3. Preoperative Respiratory Function
BloodGases Group 1 Group2 P

Upon Admission
Table 2. Use of Pharmacologic Agents PO2 (mmHg) 87.5 _+ 78.3 100.7 • 82.8 NS
Agent Group 1 (%) Group 2 (%) PCO2 (mmHg) 47.8 • 28.3 41.4 _+ 19.1 NS

Dopamine 16 45 At Worst
Isopreterenol 55 40 PO2 (mmHg) 72.3 • 61.1 68.5 + 65,4 NS
Tolazoline 10 3 PCO2 (mmHg) 48.8 • 28.6 44.2 • 20.6 NS
SURGERY IN DIAPHRAGMATIC HERNIA 733

Table 4. Incidence of Pneumothorax Table 5. Complications and Mortality

Time Group 1 (%) Group 2 (%) Problem Group 1 (n = 31) Group 2 (n = 30)

Preoperatively 13 30 Wound infection 2

Postoperatively 23
Total 36
found that the hernia partially reduces on chest x-ray
after paralysis and positive pressure ventilation in
almost all patients (Fig 2), making it seem unlikely
that the bowel exerts any significant pressure on the
ipsilateral lung. This observation weakens the ration-
ale for emergency repair of the hernia.
The second observation is that these children all
have some degree of bilateral pulmonary hypoplasia
and immaturity, as documented in our series and in the
literature. 8 It is this hypoplasia which seems to be the
direct cause of respiratory and hemodynamic instabil-
ity in these patients, 9 and this is clearly not affected by
the timing of surgery.
The third observation is that children with CDH
rarely improve after emergency surgery; in fact, many
of them are significantly worse postoperatively, l~ This
deterioration has often been attributed to the delayed
onset of persistent fetal circulation, the cause of which
is not clear. The other factor that appears to worsen the
condition of these babies in the early period is baro-
trauma from high-pressure ventilation, resulting in
pneumothorax and pulmonary vascular changes. H It
has been our feeling that emergency repair tends to
decrease thoracic compliance by raising intraabdomi-
nal pressure, thereby increasing the pressure required
to ventilate the child; this, in turn, leads to worsening
hypercarbia, acidosis, and pulmonary vascular resis-
tance.
10 Systemic sepsis 2
Hepatic failure 1
40
Chylothorax 1
Pulmonary hemorrhage
Apneic spells
Congestive heart failure 1
8rachial artery thrombosis 1
Seizures 1
Death* 18 (58%) 15 (5O%)
*P = NS at .05 level.
In support of this latter theory, investigators at our
center have measured total thoracic compliance and
PCO2 values before and after repair in a series of
neonates with CDH. 12 In all but one patient, com-
pliance decreased and PCO2 increased after repair of
the defect. These data support the concept that emer-
gency repair in an already compromised newborn
makes it more difficult to adequately ventilate the
baby and to minimize airway pressure.
The use of ECMO has been widely advocated in
recent years for poor-risk infants with CDH. ]3 Delayed
surgery does not rule out the use of ECMO; in fact it
offers several advantages. According to our protocol of
preoperative stabilization, the course of E C M O would
be completed prior to the operation; this would elimi-
nate significant bleeding complications from the oper-
ative site. In addition, we have identified a potential
borderline group with persistent hypercarbia that we
have salvaged with HFO and markedly delayed sur-
gery (the two patients referred to previously); a similar
borderline group with persistent hypoxia may also be

Fig 2. Chest x-ray immediately after intubation, and after 12 hours of positive pressure ventilation.
734
salvageable with E C M O , followed by late operative
repair.
T h e data presented here d e m o n s t r a t e that delayed
surgery does not adversely affect overall mortality. I n
addition, we have suggested a n u m b e r of theoretical
reasons why emergency operation m a y be inappro-
priate or even harmful. For these reasons, we conclude
t h a t emergency surgery is not necessary, and t h a t
REFERENCES
1. Simson JNL, Eckstein HB: Congenital diaphragmatic hernia:
A 20-year experience. Br J Surg 72:733-736, 1985
2. ReynoldsM, Luck SR, Lappen R: The "critical" neonate with
congenital diaphragmatic hernia: A 21-year perspective. J Pediatr
Surg 19:364-369, 1984
3. Karl SR, Ballantine TVN, Snider MT: High-frequency venti-
lation at rates of 375 to 1,800 cyclesper minute in four neonates with
congenital diaphragmatic hernia. J Pediatr Surg 18:822-828, 1983
4. Gibson C, Fonkalsrud EW: Iatrogenic pneumothorax and
mortality in congenital diaphragmatic hernia. J Pediatr Surg
18:555-559, 1983
5. Gross RE: Congenital hernia of the diaphragm, in The Surgery
of Infancy and Childhood. Philadelphia, Saunders, 1953, pp 428-
444
6. AndersonKD: Congenital diaphragmatic hernia, in Welch KJ,
Randolph JG, Ravitch MM, et al (eds): Pediatric Surgery. Chicago,
Year Book Medical, 1986, pp 589-601
7. Adzick NS, Harrison MR, Glick PL, et al: Diaphragmatic
hernia in the fetus: Prenatal diagnosis and outcome in 94 cases. J
Pediatr Surg 20:357-361, 1985
LANGER ET AL
repair should be done only when the p a t i e n t has been
stabilized using conventional ventilation, H F O , or
E C M O if necessary. Proof of this approach will await
controlled trials with large n u m b e r s of patients.
ACKNOWLEDGMENT
The authors thank Ms Helen Altena for help in the preparation of
the manuscript.
8. Nguyen L, Guttrnan FM, DeChardarevian JP: The mortality
of congenital diaphragmatic hernia: Is total pulmonary mass inade-
quate, no matter what? Ann Surg 198:766-770, 1983
9. Berdon WE, Baker DH, Amoury R: The role of pulmonary
hypoplasia in the prognosisof newborn infants with diaphragmatic
hernia and eventration. Am J Roentgenol Radium Ther Nucl Med
103:413-421, 1968
10. Geggel RL, Murppy JD, Langleben D, et al: Congenital
diaphragmatic hernia: Arterial structural changes and persistent
pulmonary hypertension after surgical repair. J Pediatr 107:457-
464, 1985
11. Hansen J, James S, Burrington J, et al: The decreasing
incidence of pneumothorax and improving survival of infants with
congenital diaphragmatic hernia. J Pediatr Surg 19:385-388, 1984
12. Sakai H, Tamura M, Hosokawa Y, et al: The effect of
surgical repair on respiratory mechanics in congenital diaphrag-
matic hernia. J Pediatr 3:432-438, 1987
13. German JC, Gazzaniga AB, Amlie R, et al: Management of
pulmonary insufficiencyin diaphragmatic hernia using extracorpor-
eal circulation with a membrane oxygenator (ECMO). J Pediatr
Surg 12:905-912, 1977