SMFM Papers

ajog.org

Contemporary outcomes of sickle cell disease in pregnancy Kelly Kuo, MD; Aaron B. Caughey, MD, PhD

BACKGROUND: Data regarding pregnancy outcomes in sickle cell disease are conflicting. Previous studies are limited by small sample size, narrow geographic area, and a wide range of resource availability. OBJECTIVE: The purpose of this study was to examine the association between maternal sickle cell disease and adverse pregnancy outcomes in a contemporary North American cohort. STUDY DESIGN: We performed a retrospective cohort study of 2,027,323 women with singleton pregnancies delivered in California from 2005e2008. Deliveries at <24 or >42 6/7 weeks of gestation were excluded. Women with sickle cell disease were compared with control subjects. Maternal outcomes of interest included preeclampsia, preterm delivery, placental abruption, oligohydramnios, and cesarean delivery; neonatal outcomes included small for gestational age, anomalies, stillbirth, neonatal death, and infant death. RESULTS: The prevalence of sickle cell disease was 0.017%. Compared with control subjects, women with sickle cell disease were more likely to have limited prenatal care (7.4 vs 3.8%; P¼.001),

S

ickle cell disease (SCD) is one of the most common inherited genetic disorders in the world and is associated with significant lifelong morbidity.1 Approximately 1 in 500 African American and 1 in 1200 Hispanic American births in the United States are affected by SCD, with important implications for both maternal and neonatal outcomes.2,3 Maternal morbidity may occur secondary to acute SCD-related crises, in addition to venous thromboembolism, infection, or chronic end-organ dysfunction. Numerous studies have demonstrated significantly increased rates of intrauterine growth restriction, preterm delivery (PTD), and small for gestational age (SGA) infants in women with SCD, likely part to because of underlying hypertension and placental insufficiency. However, data regarding the association between SCD and gestational diabetes mellitus, preeclampsia, intrauterine fetal death (IUFD),

Cite this article as: Kuo K, Caughey AB. Contemporary outcomes of sickle cell disease in pregnancy. Am J Obstet Gynecol 2016;215:505.e1-5. 0002-9378/$36.00 ª 2016 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ajog.2016.05.032

underlying chronic hypertension (2.3% vs 1.1%; P¼.038), and fetal anomalies (14.0 vs 6.4%; P<.001). The increased odds of fetal anomalies persisted after adjustment for multiple confounders (odds ratio, 1.73; 95% confidence interval, 1.26e2.38). Women with sickle cell disease also had higher odds of severe preeclampsia (odds ratio, 3.75; 95% confidence interval, 2.21e6.38), preterm delivery (odds ratio, 2.50; 95% confidence interval, 1.93e3.21), small for gestational age (odds ratio, 1.96; 95% confidence interval, 1.18e3.25), and cesarean delivery (odds ratio, 1.93; 95% confidence interval, 1.40e2.67). CONCLUSION: Women with sickle cell disease are at high risk of maternal and neonatal morbidity. Low rates of fetal and neonatal death may reflect improved antenatal surveillance and management as compared with previous studies. The association between sickle cell disease and fetal anomalies warrants further investigation. Key words: fetal anomalies, outcome, pregnancy, sickle cell disease

neonatal death, and maternal death are conflicting, and estimates vary widely in regards to the magnitude of associated risk.4-10 A recent systematic review and metaanalysis of 19 studies from 9 different countries reported an increased risk of both IUFD and neonatal death in the setting of maternal SCD, with pooled odds ratios of 4.05 and 2.71, respectively.7 A markedly increased risk of maternal death was seen among women with SCD compared with women without SCD; however, this effect was driven largely by data from low-income countries, in which the odds of maternal death were nearly 30 times higher than that of women without SCD.11,12 Although maternal SCD status did not confer any significant additional risk for maternal death in high-income countries, such as the United States, previous studies have been limited similarly by small sample sizes, long periods of data accrual, and variations in clinical practice that may limit their generalizability.7,13,14 Given the significant impact and burden of disease associated with maternal SCD, additional data are needed regarding outcomes and risks of adverse pregnancy outcomes. We

therefore sought to investigate outcomes of SCD in pregnancy in a contemporary North American cohort.

Methods We performed a retrospective cohort study that included all singleton pregnancies delivered in the state of California from 2005e2008. The data were derived from linked mother-infant datasets from the California Vital Statistics Birth Certificate Data, infant Vital Statistics Death Certificate Data, California Patient Discharge Data, and Vital Statistics Fetal Death File. Data linkage is performed by the California Office of Statewide Health Planning and Development Healthcare Information Resource Center, under the California Health and Human Services Agency, which used a unique “record linkage number” specific to the mother-infant pair. The state of California maintains these linked datasets that include health information from maternal antepartum and postpartum hospital records for the 9 months before delivery and 1 year after delivery, as well as birth records and all infant admissions that occur within the first year of life. We obtained human subjects approval from the Institutional Review Board at Oregon Health &

OCTOBER 2016 American Journal of Obstetrics & Gynecology

505.e1

SMFM Papers Science University, the California Office of Statewide Health Planning and Development, and the Committee for the Protection of Human Subjects. The linked dataset did not contain potential patient privacy/identification information, so informed consent was exempted. Our primary exposure of interest was a diagnosis of SCD in pregnancy. The following International Classification of Diseases, 9th Revision (ICD-9) codes were used: 282.60 (sickle cell disease, unspecified), 282.61 (Hg-S disease), 282.62 (Hg-S with crisis), 282.64 (Hg-S with vaso-occlusive pain). Three hundred forty-four cases were compared with 2,026,979 control pregnancies. Patients with sickle trait and other hemoglobinopathies were excluded, as were deliveries at <24or >42 6/7 weeks of gestation. To avoid the confounding of complications linked to multifetal gestations, we excluded these pregnancies for both the cases and the controls. Analyses were conducted with Stata software (version 12; Stata Corporation, College Station, TX). Outcomes of interest that were examined were also determined retrospectively through the use of ICD-9 codes and included preeclampsia, severe preeclampsia, eclampsia, abruption, IUFD, PTD <37 weeks of gestation, PTD <32 weeks of gestation, SGA, gestational diabetes mellitus, neonatal death, and infant death. Our analytic approach was first to conduct bivariate analyses of women with and without the exposure of interest for each of the outcomes of interest. Statistical comparisons of categoric variables were made with the use of chisquared tests. We then conducted multivariable logistic regression models to control for potential confounding. Potential confounders that were assessed included maternal age (35 years old and <20 years old), maternal education (>12 years vs 12 years), insurance status (private insurance vs public insurance or no insurance), race/ethnicity, parity, diabetes mellitus mellitus, chronic hypertension, and gestational diabetes mellitus. Additionally, we excluded those variables that were used

ajog.org

TABLE 1

Maternal demographic characteristics Characteristic

Sickle cell disease (n¼344), n (%)

Control (n¼2,026,979), n (%)

Maternal age, y

.08

20

31 (9.0)

191,244 (9.4)

21-34

266 (77.3)

1,485,717 (73.3)

35

47 (13.7)

350,018 (17.3) <.001

Maternal race African American

P value

264 (76.7)

103,359 (5.1)

White

13 (3.8)

538,457 (26.6)

Hispanic

43 (12.5)

Asian/Pacific Islander

10 (2.9)

241,760 (11.9)

Other/unknown

14 (4.1)

38,370 (1.9)

1,102,700 (54.5)

Highest education level

.13

High school or less

170 (49.4)

1,085,946 (53.6)

Some college or graduate degree

165 (48.0)

882,097 (43.5)

9 (2.6)

58,936 (2.9)

159 (46.2)

806,593 (39.8)

.011

Unknown Nulliparous Limited prenatal care

26 (7.4)

77,392 (3.8)

.011

183 (53.2)

981,322 (48.4)

.076

Tobacco use

7 (2.0)

13,039 (0.6)

.001

Chronic hypertension

8 (2.3)

23,075 (1.1)

.038

Public insurance

Kuo & Caughey. Sickle cell in pregnancy. Am J Obstet Gynecol 2016.

as the outcome of interest when appropriate. For example, we did not adjust for preeclampsia when evaluating preeclampsia as an outcome. Adjusted odds ratios were calculated for all outcomes of interest. Statistical significance was determined by a probability value of <.05 and/or 95% confidence intervals.

Results A total of 2,027,323 pregnancies met inclusion criteria, among which 344 cases of SCD were identified. Maternal demographic characteristics are shown in Table 1. The prevalence of SCD was 0.017%. Compared with control subjects, women with SCD were more likely to be African American (76.7% vs 5.1%; P<.001), to be nulliparous (46.2% vs 39.8%; P¼.011), to have underlying chronic hypertension (2.3% vs 1.1%; P¼.038), and to have limited prenatal

505.e2 American Journal of Obstetrics & Gynecology OCTOBER 2016

care, as defined by <5 visits (7.4% vs 3.8%; P¼.001). A trend toward higher rates of public insurance was noted in the SCD group, although this was not statistically significant (53.2% vs 48.4%; P¼.07). Maternal educational status and the relative percentages of women aged <20 years or 35 years at time of delivery were also similar between groups. In univariate analyses, SCD was associated with a statistically significant increase in rates of preeclampsia, PTD, and SGA (Figure). Women with SCD had higher rates of both mild preeclampsia (10.2% vs 3.0%; P<.001) and severe preeclampsia (5.4% vs 0.9%); however, no difference in the rates of eclampsia were noted between women with or without SCD (0.0% vs 0.07; P¼.62; data not shown). The incidence of PTD at <37 weeks of gestation was nearly 2.5 times higher in women with

SMFM Papers

ajog.org

FIGURE

Maternal and neonatal outcomes in women with sickle cell disease vs control subjects

Bivariate analyses. CS, cesarean section delivery; GDM, gestational diabetes; IOL, induction of labor; IUFD, intrauterine fetal demise; NND, neonatal death; PET, preeclampsia; PTD, preterm delivery; PTD<32w, preterm delivery <32 weeks of; SCD, sickle cell disease; SGA, small for gestational age. Kuo & Caughey. Sickle cell in pregnancy. Am J Obstet Gynecol 2016.

SCD, as compared with control subjects (25.9% vs 9.8%; P<.001). An even greater difference was seen in the incidence of PTD at <32 weeks of gestation, with rates approximately 6 times higher in the setting of SCD (7.0% vs 1.3%; P<.001). The association between SGA and SCD was significant for all percentile cutoffs that were assessed: <10th percentile (16.2% vs 6.3%; P<.001),

<5th percentile (8.4% vs 2.7%; P<.001), and <3rd percentile (5.4% vs 1.6%; P<.001). Women with SCD were less likely to have gestational diabetes mellitus, as compared with control subjects (3.8% vs 6.5%; P¼.04). They were more likely to require induction of labor (21.8% vs 14.3%; P<.001) and to be delivered by cesarean section (46.2% vs 30.2%;

TABLE 2

Selected fetal anomalies in patients with sickle cell disease vs control subjects Variable

Sickle cell disease (n¼344), %

Control subjects (n¼2,026,979), %

Comment P value

Cardiac anomalies

1.74

0.52

.01a

Abdominal wall defect

0.58

0.06

.021a

Congenital lung abnormality

0.58

0.05

.013a

Neural tube defects

0

0.03

.75

Cleft lip/palate

0

0.13

1.00

Hypospadias

0

0.21

1.00

Any anomaly

14.0

6.39

<.001a

a

Statistically significant values. Kuo & Caughey. Sickle cell in pregnancy. Am J Obstet Gynecol 2016.

P<.001). Of note, among multiparous women with 1 previous cesarean delivery, the rate of subsequent successful vaginal birth was significantly lower in women with SCD, as compared with control subjects (49.7% vs 70.3%; P<.001; data not shown); however, the rates of attempted trial of labor after previous cesarean delivery in each group were unavailable for analysis. Women with SCD were significantly more likely to have a pregnancy complicated by a fetal anomaly (14.0% vs 6.4%; P<.001). Specific subsets of major anomalies were examined and demonstrated that the incidence of cardiac malformations, abdominal wall defects, and congenital lung abnormalities were increased significantly among women with SCD (Table 2). However, rates of IUFD and neonatal death were not significantly different between groups. An increased risk of infant death, as defined as death within the first year of life, was noted in the SCD group (1.17% vs 0.33%; P¼.007); however, this association was no longer significant after multivariable analysis. Similarly, the increased risks of oligohydramnios and placental abruption in the SCD group did not remain statistically significant after adjustment for potential confounding variables. Multivariable regression analysis results are shown in Table 3. The increased odds of multiple adverse maternal and neonatal outcomes of interest persisted after adjustment for confounding factors, with the highest adjusted odds ratios seen for severe preeclampsia and PTD <32 weeks of gestation.

In this large retrospective cohort study, significantly increased incidence and odds of both mild and severe preeclampsia were noted in women with SCD, as compared with non-SCD controls. These findings are consistent with the previous literature, although the rates of eclampsia in each group were too low to draw meaningful conclusions. The nearly 4-fold increased odds of severe preeclampsia is likely to be associated with significant maternal morbidity and death, although maternal mortality

OCTOBER 2016 American Journal of Obstetrics & Gynecology

505.e3

SMFM Papers

ajog.org

TABLE 3

Maternal and neonatal outcomes associated with sickle cell disease

Outcome

Sickle cell disease Control Adjusted 95% Confidence (n¼344), % (n¼2,026,979), % odds ratio interval

Preeclampsia

10.2

3.0

2.07

1.25e3.42a

Severe preeclampsia

5.3

0.9

3.75

2.21e6.37a

Placental abruption

1.7

0.9

1.66

0.74e3.74

Oligohydramnios

4.4

2.6

1.36

0.79e2.33

Preterm delivery

25.9

9.8

2.50

1.94e3.21a

7.0

1.3

2.99

1.89e4.74a

<10%

16.2

6.30

1.96

1.18e3.25a

<5%

8.4

2.7

2.00

1.33e2.90a

<3%

5.4

1.6

1.96

1.18e3.25a

Neonatal death

0.6

0.2

2.10

0.52e8.52

Infant death

1.2

0.3

2.21

0.82e5.97

Induction of labor

21.8

14.3

1.63

1.25e2.13a

Cesarean delivery

46.2

31.2

1.93

1.40e2.67a

Fetal anomalies

14.0

6.4

1.73

1.26e2.38a

Preterm delivery <32 wk Small for gestational age

a

Statistically significant value. Kuo & Caughey. Sickle cell in pregnancy. Am J Obstet Gynecol 2016.

data were unable to be assessed in the current study, because the linked datasets used in our study do not include maternal death certificate data. Neonates born to mothers with SCD were more likely to be delivered preterm at <37 weeks of gestation and, in particular, at <32 weeks of gestation. A significant portion of these preterm births are likely attributable to the increased odds of preeclampsia; however, whether a preterm birth was spontaneous or iatrogenic is not specifically coded in these data. The higher incidence and odds of pregnancies that were affected by fetal anomalies in women with SCD in our study was unexpected and remained significant in multivariable regression analysis that controlled for multiple potential confounders. We hypothesize that any underlying mechanisms are likely multifactorial, rather than purely because of hypoxia alone, because early embryogenesis normally occurs under hypoxic conditions until approximately 11e13 weeks of gestation.15,16 Because

prolonged hypoxia has been demonstrated to affect fetal vasculogenesis and cardiac remodeling in animal models,17,18 it is possible, however, that chronic hypoxia, placental infarctions, and systemic hypertension predispose fetuses to vascular injury that may contribute to the higher incidence of fetal cardiac, lung, and abdominal wall defects that were seen in women with SCD. A significant portion of the fetal anomalies in our linked datasets were otherwise unclassified and warrant further investigation. In contrast to previous studies of pregnancy outcomes with maternal SCD in high-income countries, in our cohort there were no significant differences in the risk of IUFD, neonatal death, and infant death in women with and without SCD. We hypothesize that low rates of fetal and neonatal death may reflect improved antenatal surveillance and management, as compared with previous studies. Despite the high prevalence of chronic pain and disability that is associated with

505.e4 American Journal of Obstetrics & Gynecology OCTOBER 2016

SCD, women with SCD in our cohort were less likely to have sufficient prenatal care, as defined by 5 prenatal visits. It is unclear whether this was due to higher rates of emergency department use and/ or inpatient hospitalization, other comorbid conditions (such as substance abuse), or socioeconomic factors and limited access to care; however, these findings highlight some of the challenges that women with SCD and their obstetric providers may face. Our study has several important limitations. Our data are retrospective in nature and relied on ICD-9 codes for categorizing all outcomes of interest. Validation studies of the California Registry and Surveillance System for Hemoglobinopathies Project, which also used linked datasets from Vital Statistics data, administrative databases, and California Patient Discharge Data, suggest high accuracy rates in correct identification of cases of SCD.19,20 The prevalence of SCD in our study population was 0.017%, which is lower than national estimates of approximately 0.03% but are very close to published estimates of SCD in California of 0.0174%.21,22 However, inaccuracies in such coding may exist and can lead most commonly to misclassification bias. Because of nondifferential misclassification of a dichotomous exposure biases toward the null, the magnitude of the positive associations in our study may represent the lower bounds of the actual impact of SCD, and our negative findings may be too conservative. Additional information regarding severity of maternal disease (such as rates of concomitant cardiac or pulmonary dysfunction, number of sickle crises experienced during pregnancy, or frequency of transfusions) were unavailable with the use of our linked datasets. Thus, we were unable to consider severity of illness and examine it in a dose-response fashion. We were similarly unable to control for additional factors, such as the prevalence of maternal opioid dependence and substance abuse, which would conceivably affect fetal and neonatal outcomes, particularly the frequency and severity of neonatal abstinence syndrome. Because SCD is not a reportable condition,

SMFM Papers

ajog.org neonatal SCD status is not available from birth certificate data. However, strengths of our study include the use of linked datasets that include information from both antepartum and postpartum hospitalization records, the inclusion of an ethnically diverse population within the state of California, and an analysis of multiple cutoffs of outcome severity. In summary, pregnancies that are complicated by SCD are at high risk of maternal and neonatal morbidity, although rates of stillbirth and neonatal death are not increased above the general population. These findings highlight the importance of fetal surveillance, particularly in the setting of coexisting maternal hypertension, vaso-occlusive crises, placental insufficiency, and growth restriction. Because women with SCD are more likely to have limited prenatal care, patients should be counseled in the preconception period, whenever possible, regarding the risks that are associated with pregnancy and should be offered reliable contraceptive options, so that pregnancies may be undertaken in a planned manner with coordination between maternal-fetal medicine, hematology, genetic counseling, and additional subspecialties as needed. Finally, genetic screening should be offered, and a detailed anatomy scan should be performed at 18e20 weeks of gestation, given the association between maternal SCD and fetal anomalies. Future studies are needed to further investigate the association between SCD and fetal anomalies, in addition to the optimal timing of delivery in women with SCD. n References 1. Stuart MJ, Nagel RL. Sickle-cell disease. Lancet 2004;364:1343-60.

2. Centers for Disease Control and Prevention. Sickle Cell Disease (SCD): Data & Statistics. 2015 7/8/2015 2/11/16]; Available from: http:// www.cdc.gov/ncbddd/sicklecell/data.html. Accessed February 22, 2016. 3. Hassell KL. Population estimates of sickle cell disease in the U.S. Am J Prev Med 2010;38(suppl):S512-21. 4. Boulet SL, Okoroh EM, Azonobi I, Grant A, Craig Hooper W. Sickle cell disease in pregnancy: maternal complications in a Medicaidenrolled population. Matern Child Health J 2013;17:200-7. 5. Villers MS, Jamison MG, De Castro LM, James AH. Morbidity associated with sickle cell disease in pregnancy. Am J Obstet Gynecol 2008;199:125.e1-5. 6. Smith JA, Okoroh EM, Azonobi I, Grant A, Craig Hooper W. Pregnancy in sickle cell disease: experience of the Cooperative Study of Sickle Cell Disease. Obstet Gynecol 1996;87: 199-204. 7. Boafor TK, Olayemi E, Galadanci N, et al. Pregnancy outcomes in women with sickle-cell disease in low and high income countries: a systematic review and meta-analysis. BJOG 2016;123:691-8. 8. Alayed N, Kezouh A, Oddy L, Abenhaim HA. Sickle cell disease and pregnancy outcomes: population-based study on 8.8 million births. J Perinat Med 2014;42:487-92. 9. Odum CU, Anorlu RI, Dim SI, Oyekan TO. Pregnancy outcome in HbSS-sickle cell disease in Lagos, Nigeria. West Afr J Med 2002;21: 19-23. 10. Wilson NO, Ceesay FK, Hibbert JM, et al. Pregnancy outcomes among patients with sickle cell disease at Korle-Bu Teaching Hospital, Accra, Ghana: retrospective cohort study. Am J Trop Med Hyg 2012;86:936-42. 11. Afolabi BB, Iwuala NC, Iwuala IC, Ogedengbe OK. Morbidity and mortality in sickle cell pregnancies in Lagos, Nigeria: a case control study. J Obstet Gynaecol 2009;29:104-6. 12. Muganyizi PS, Kidanto H. Sickle cell disease in pregnancy: trend and pregnancy outcomes at a tertiary hospital in Tanzania. PLoS One 2013;8: e56541. 13. Barfield WD, Barradas DT, Manning SE, Kotelchuck M, Shapiro-Mendoza CK. Sickle cell disease and pregnancy outcomes: women of African descent. Am J Prev Med 2010;38(suppl): S542-9. 14. Sun PM, Wilburn W, Raynor BD, Jamieson D. Sickle cell disease in pregnancy:

twenty years of experience at Grady Memorial Hospital, Atlanta, Georgia. Am J Obstet Gynecol 2001;184:1127-30. 15. Hutter D, Kingdom J, Jaeggi E. Causes and mechanisms of intrauterine hypoxia and its impact on the fetal cardiovascular system: a review. Int J Pediatr 2010;2010:9. 16. Jauniaux E, Watson AL, Hempstock J, Bao YP, Skepper JN, Burton GJ. Onset of maternal arterial blood flow and placental oxidative stress: a possible factor in human early pregnancy failure. Am J Pathol 2000;157: 2111-22. 17. Ba S, Xiao Y, Li G, Casiano CA, Zhang L. Effect of maternal chronic hypoxic exposure during gestation on apoptosis in fetal rat heart. Am J Physiol Heart Circ Physiol 2003;285: H983-90. 18. Nichol D, Stuhlmann H. EGFL7: a unique angiogenic signaling factor in vascular development and disease. Blood 2012;119:1345-52. 19. Lane PA, Theodore RS, Quarmyne MO, Eckman JR, Zhou M, Snyder AB. Accuracy of ICD-9 coding for SCD in children and adolescents: results from the Georgia (GA) Rush Surveillance Project. Blood 2014;124:48564856. 20. Paulukonis ST, Harris WT, Coates TD, Neumayr L, Treadwell M, Vichinsky E. Population based surveillance in sickle cell disease: methods, findings and implications from the California registry and surveillance system in hemoglobinopathies project (RuSH). Pediatr Blood Cancer 2014;61:2271-6. 21. Brousseau DC, Panepinto JA, Nimmer M, Hoffmann RG. The number of people with sicklecell disease in the United States: national and state estimates. Am J Hematol 2010;85:77-8. 22. US Census Bureau; Census 2005: National Tables. Annual Estimates of the Population for the United States and States, and for Puerto Rico: April 1, 2000 to July 1, 2005 (NSTEST2005-01). Available at: http://www.census. gov/popest/data/historical/2000s/vintage_2005. Accessed May 6, 2016.

Author and article information From the Department of Obstetrics and Gynecology, Oregon Health & Science University, Portland, OR. Received March 4, 2016; revised May 10, 2016; accepted May 20, 2016. The authors report no conflict of interest. Corresponding author: Kelly Kuo, MD. kuok@ohsu. edu

OCTOBER 2016 American Journal of Obstetrics & Gynecology

505.e5