From the Department of Child Health, University of Newcastle upon Tyne, Newcastle upon Tyne, United Kingdom
| ABSTRACT |
|---|
|
|
|---|
Methods. A retrospective review of cases from a regional case registry, the Northern Region Congenital Anomaly Survey, was conducted.
Results. A total of 185 cases were identified. Mortality was 62% and did not vary significantly during the study period. Mortality was unaffected by the introduction of new therapies. There was a significant inverse correlation between the rate of elective termination and survival of live borns. The presence of an additional anomaly increased mortality to 79%.
Conclusions. The mortality of CDH when complete case ascertainment is achieved is unaffected by new therapies. The survival rate is principally determined by the rate of antenatal termination and the incidence of associated anomalies. Reports of improved survival of CDH should be interpreted with caution, as variations in outcome are more likely to be explained by case selection artifact.
Key Words: congenital diaphragmatic hernia survival
Abbreviations: CDH, congenital diaphragmatic hernia ECMO, extracorporeal membrane oxygenation HFOV, high-frequency oscillatory ventilation iNO, inhaled nitric oxide CI, confidence interval
Therapies used to treat congenital diaphragmatic hernia (CDH) include delayed surgery,1 extracorporeal membrane oxygenation (ECMO),2 high-frequency oscillatory ventilation (HFOV),3 inhaled nitric oxide (iNO),4 and surfactant5 with reports of improved survival. However, the hidden mortality either by antenatal termination or by postnatal death before transfer to referral centers is ignored.
The United Kingdoms Northern health region is a geographically defined population of 3.2 million with 34 000 annual live births, within which the Northern Region Congenital Abnormality Survey registers infants with significant congenital abnormalities. The aims of this study were to identify variables that influenced survival in this population during a period when several new therapies were introduced.
| METHODS |
|---|
|
|
|---|
Outcome was classified as spontaneous abortion, elective termination, antepartum stillbirth, early neonatal death (death between birth and 24 hours), late neonatal death (death between 1 and 7 days), postnatal death (death between 7 days and 1 year), and survival to 1 year of age. Treatment of infants with congenital diaphragmatic hernia displaying respiratory distress at birth included intubation, paralysis, and ventilation, with manipulation of ventilator settings according to clinical criteria. New therapies were first used in the management of CDH as follows: HFOV (June 1995), iNO (October 1995), and surfactant (March 1998). After a prospective trial,6 delayed surgery was accepted as normal practice in 1993. ECMO was used between December 1994 and June 1999.
Statistical analysis used the
2 test. Regression analysis was performed with the Minitab statistical package.
| RESULTS |
|---|
|
|
|---|
2 = 10.0; P = .4). Of the 44 fetuses that were terminated electively, 32 had additional anomalies. Of the 58 live borns who subsequently died, 42 did so within the first 24 hours, 11 between 24 hours and 7 days, and 5 between 1 week and 1 year. Sixteen deaths occurred in peripheral hospitals. Fifty-two infants were born in the surgical center, 75 in peripheral hospitals, and 2 at home. Survival was 44%, 63%, and 100%, respectively. The better survival in peripheral hospitals was not statistically significant (
2 = 3.5; P = .06). Overall mortality was 62%, or 114 of 185 (95% confidence interval [CI]: 52.4%69%), and did not vary during the period (
2 = 9.2; P = .5; Fig 1).
|
2 = 16.6; P < .001). Patients who received a diagnosis antenatally were not more likely to have associated major congenital anomalies. Forty-three of 97 infants whose CDH was detected antenatally had an additional anomaly; 37 of 88 cases that were detected postnatally had an additional anomaly (
2 = 0.09; P = .7). There was a strong inverse relationship between the annual rate of antenatal death, principally elective termination, and the annual mortality rate of live borns (R2 = 0.41; P = .03; Fig 2).
|
2 = 26.9; P < .001). Left-sided hernias had 55% mortality; right-sided hernias had 54% mortality. Three infants had bilateral hernia with 100% mortality.
Effect of New Therapies
Before the use of ECMO, there were 69 cases; 51 were live births, with 22 surviving (68% overall mortality, 57% mortality of live births). During the period when ECMO was available, there were 73 cases; 48 were live births, 31 of whom survived (58% overall mortality, 45% mortality of live births). Six received ECMO, and there were 2 survivors. Since discontinuing ECMO, there have been 43 cases, 30 live births, and 18 survivors (58% overall mortality, 40% mortality of live births). There was no statistical difference in survival between any period when considering survival of all cases. When only live births are considered, the difference between the pre-ECMO and ECMO periods is significant (
2 = 4.5; P = .03), but there is no difference in survival since discontinuing the use of ECMO (
2 = 0.1; P = .7; 95% CIs for survival of live borns: pre-ECMO: 29.2%78.8%; ECMO: 2.9%66.6%; post-ECMO: 7.9% to 83.9%; Table 1, Fig 1).
|
2 = 1.4; P = .22). Thirteen infants have received HFOV. Mortality before and after the introduction of HFOV was 65% (95% CI: 52.4%77.2%) and 58% (95% CI: 44.9%69.5%), respectively (
2 = 1.1; P = .2). Nine infants have received surfactant. Mortality before and after the introduction of surfactant was 63% (95% CI: 55.6%70.1%) and 57% (95% CI: 24.6%81.3%), respectively (
2 = 0.69; P = .4). The mortality before and after the advent of delayed surgery was 69% and 58%, respectively (
2 = 1.8; P = .16). | DISCUSSION |
|---|
|
|
|---|
First, half of the cases are not detected by antenatal scanning,1517 leading to the birth of a infant with a life-threatening condition in an unprepared obstetric unit. Some die immediately after birth, are not subject to autopsy, and therefore remain undiagnosed; the mortality is worse than we have demonstrated.18,19 Time of death was largely in the first 24 hours, with 28% of liveborn infants who died doing so in peripheral hospitals. Only infants with stable pathophysiology survive to transfer, resulting in case selection bias. For eliminating this bias, infants who are delivered and die in peripheral hospitals or are terminated electively must be identified. When this is done, the mortality rate is unchanging across different decades. Thus, although we report a mortality rate of 62% in the 1990s, in the 1980s, Wenstrom et al20 recorded a 58% mortality, the current author reported 66% mortality,19 and Harrison et al21 recorded 66% for the early 1970s. In the early 1990s, Harrison revisited the subject of hidden mortality and reported 58% for antenatally diagnosed cases in an ECMO center.18 The close approximation of these figures suggests that without case selection bias, survival is unchanged. Because the new therapies that we have studied treat pulmonary hypertension and cannot affect pulmonary hypoplasia, this is unsurprising.
The second confounding feature of reports of survival in CDH is the unreliability of historical controls. In our study, the introduction of ECMO led to a fall in mortality in comparison with the previous period. However, having ceased to offer ECMO therapy, improved survival persisted. Comparison of current results with previous eras takes no account of general improvements in care or changes in referral patterns. The use of historical controls misleads investigators to believe that changes in outcome are the direct result of new therapies, without the rigor of prospective randomization of such therapies. Despite this, the majority of studies reporting improved survival of CDH continue to use historical controls from the same institution.2,11,14,2225
Strict criteria for the use of ECMO led to 6 infants from 48 cases (12.5%) receiving this therapy. The experience of Great Ormond Street Hospital, using the same criteria, was 11 (15%) of 72 infants receiving ECMO.26 We ceased to offer this therapy because of poor outcome. Reports from North America record up to two thirds of infants with CDH receiving ECMO.22,27 Whether this reflects severe pathology or different criteria for use of this therapy is speculative.
The 2 factors that clearly relate to postnatal mortality are the rate of termination and the presence of additional anomalies. The explanation of the inverse relationship between termination and postnatal survival is probably that the cases that are detected antenatally have more severe pathology. Previous reports have identified the poor prognosis of antenatally detected cases,7,8,18 but this is the first series in which the direct relationship between the rate of antenatal termination and survival of live births has been documented. It is noticeable that when the effects of therapies are tabulated, for each apparent improvement in survival, there is a concomitant increase in the rate of antenatal death (Table 1). Although we have analyzed these therapies as independent variables, we recognize that they were not introduced independent of each other and that analysis by use of a statistical model would be appropriate. It is unlikely that any of the therapies that we have analyzed would be independently significant taking into account the 2 principal variables that we have identified of antenatal termination and the rate of associated anomalies. We urge all future reports to give details both of numbers dying in peripheral hospitals and of termination rates. Apparent improvements in survival are likely to have their explanation in these 2 areas.
With an additional anomaly, mortality is 79%, which is in accordance with previous reports.2830 Why these anomalies should have such a devastating effect on survival is less clear. However, in terms of antenatal counseling, obstetricians, pediatricians, and pediatric surgeons should be aware of the implications of such an association.
This study has shown that when complete ascertainment of cases is attempted, there has been no improvement in the outcome of CDH in the past decade. The 2 factors that are related to survival are the rate of antenatal termination and the presence of associated anomalies. New therapies continue to be promulgated for this condition.31 We urge caution in accepting reports of improved survival after the introduction of new therapies where rates of antenatal termination and details of deaths in peripheral hospitals are omitted or where historical controls are used. Such reports have profound flaws.
| ACKNOWLEDGMENTS |
|---|
| FOOTNOTES |
|---|
Address correspondence to Bruce Jaffray, ChM, Department of Child Health, Sir James Spence Institute, University of Newcastle upon Tyne, Queen Victoria Rd, Newcastle upon Tyne NE1 4LP, United Kingdom. E-mail: bruce.jaffray{at}ncl.ac.uk
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
K. Tsukimori, K. Masumoto, S. Morokuma, T. Yoshimura, T. Taguchi, T. Hara, Y. Sakaguchi, S. Takahashi, N. Wake, and S. Suita The Lung-to-Thorax Transverse Area Ratio at Term and Near Term Correlates With Survival in Isolated Congenital Diaphragmatic Hernia J. Ultrasound Med., May 1, 2008; 27(5): 707 - 713. [Abstract] [Full Text] [PDF] |
||||
![]() |
V. Datin-Dorriere, E. Walter-Nicolet, V. Rousseau, P. Taupin, A. Benachi, S. Parat, P. Hubert, Y. Revillon, and D. Mitanchez Experience in the Management of Eighty-Two Newborns With Congenital Diaphragmatic Hernia Treated With High-Frequency Oscillatory Ventilation and Delayed Surgery Without the Use of Extracorporeal Membrane Oxygenation J Intensive Care Med, March 1, 2008; 23(2): 128 - 135. [Abstract] [PDF] |
||||
![]() |
K. W. Neff, A. K. Kilian, T. Schaible, E.-M. Schutz, and K. A. Busing Prediction of Mortality and Need for Neonatal Extracorporeal Membrane Oxygenation in Fetuses with Congenital Diaphragmatic Hernia: Logistic Regression Analysis Based on MRI Fetal Lung Volume Measurements Am. J. Roentgenol., December 1, 2007; 189(6): 1307 - 1311. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Santos, R. S. Moura, S. Gonzaga, C. Nogueira-Silva, S. Ohlmeier, and J. Correia-Pinto Embryonic Essential Myosin Light Chain Regulates Fetal Lung Development in Rats Am. J. Respir. Cell Mol. Biol., September 1, 2007; 37(3): 330 - 338. [Abstract] [Full Text] [PDF] |
||||
![]() |
The Congenital Diaphragmatic Hernia Study Group Defect Size Determines Survival in Infants With Congenital Diaphragmatic Hernia Pediatrics, September 1, 2007; 120(3): e651 - e657. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. C. Jani, M. Cannie, C. F. A. Peralta, J. A. Deprest, K. H. Nicolaides, and S. Dymarkowski Lung Volumes in Fetuses with Congenital Diaphragmatic Hernia: Comparison of 3D US and MR Imaging Assessments Radiology, August 1, 2007; 244(2): 575 - 582. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. R. Rempel and M. J. Harrison Safeguarding Precarious Survival: Parenting Children Who Have Life-Threatening Heart Disease Qual Health Res, July 1, 2007; 17(6): 824 - 837. [Abstract] [PDF] |
||||
![]() |
O. Boucherat, A. Benachi, A.-M. Barlier-Mur, M.-L. Franco-Montoya, J. Martinovic, B. Thebaud, B. Chailley-Heu, and J. R. Bourbon Decreased Lung Fibroblast Growth Factor 18 and Elastin in Human Congenital Diaphragmatic Hernia and Animal Models Am. J. Respir. Crit. Care Med., May 15, 2007; 175(10): 1066 - 1077. [Abstract] [Full Text] [PDF] |
||||
![]() |
E. J. Grethel, J. Farrell, R. Ball, H. Lee, and K. K. Nobuhara Congenital Diaphragmatic Hernia Associated With Spinal Anomalies Obstet. Gynecol., February 1, 2007; 109(2): 485 - 488. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. D. Clugston, J. Klattig, C. Englert, M. Clagett-Dame, J. Martinovic, A. Benachi, and J. J. Greer Teratogen-Induced, Dietary and Genetic Models of Congenital Diaphragmatic Hernia Share a Common Mechanism of Pathogenesis Am. J. Pathol., November 1, 2006; 169(5): 1541 - 1549. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. E. Larson and J. C. Cohen Improvement of pulmonary hypoplasia associated with congenital diaphragmatic hernia by in utero CFTR gene therapy Am J Physiol Lung Cell Mol Physiol, July 1, 2006; 291(1): L4 - L10. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. R. Bourbon and A. Benachi CFTR gene therapy, a method to rescue lung hypoplasia in congenital diaphragmatic hernia? Am J Physiol Lung Cell Mol Physiol, July 1, 2006; 291(1): L1 - L3. [Full Text] [PDF] |
||||
![]() |
H. G. Folkesson, C. J. Chapin, L. L. Beard, R. Ertsey, M. A. Matthay, and J. A. Kitterman Congenital diaphragmatic hernia prevents absorption of distal air space fluid in late-gestation rat fetuses Am J Physiol Lung Cell Mol Physiol, March 1, 2006; 290(3): L478 - L484. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Colvin, C. Bower, J. E. Dickinson, and J. Sokol Outcomes of Congenital Diaphragmatic Hernia: A Population-Based Study in Western Australia Pediatrics, September 1, 2005; 116(3): e356 - e363. [Abstract] [Full Text] [PDF] |
||||
![]() |
N P Smith, E C Jesudason, N C Featherstone, H J Corbett, and P D Losty Recent advances in congenital diaphragmatic hernia Arch. Dis. Child., April 1, 2005; 90(4): 426 - 428. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. P. Babiuk, B. Thebaud, and J. J. Greer Reductions in the incidence of nitrofen-induced diaphragmatic hernia by vitamin A and retinoic acid Am J Physiol Lung Cell Mol Physiol, May 1, 2004; 286(5): L970 - L973. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. R. Chess The Effect of Gentle Ventilation on Survival in Congenital Diaphragmatic Hernia Pediatrics, April 1, 2004; 113(4): 917 - 917. [Full Text] [PDF] |
||||
![]() |
M. R. Harrison, R. L. Keller, S. B. Hawgood, J. A. Kitterman, P. L. Sandberg, D. L. Farmer, H. Lee, R. A. Filly, J. A. Farrell, and C. T. Albanese A Randomized Trial of Fetal Endoscopic Tracheal Occlusion for Severe Fetal Congenital Diaphragmatic Hernia N. Engl. J. Med., November 13, 2003; 349(20): 1916 - 1924. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||