Abstract
Our clinical research team previously demonstrated an 89% decrease in mortality rate in a before-and-after feasibility study among women with sickle cell disease (SCD) living in a low-resource setting in Ghana. In the same cohort with additional participants with and without SCD, we used a prospective cohort design to test the hypothesis that implementing a multidisciplinary care team for pregnant women with SCD in the same low-resource setting will result in similar maternal and perinatal mortality rates compared to women without SCD. We prospectively enrolled pregnant women with and without SCD and followed them up to six weeks postpartum. We recruited age and parity matched pregnant women without SCD as controls. Maternal and perinatal mortality rates were the primary outcomes. A total of 149 pregnant women with SCD (HbSS, 54; HbSC, 95) and 117 pregnant women without SCD or sickle cell trait were included in the analysis. Maternal mortality rates were1.4% and 0.9% in livebirths in women with and without SCD, respectively (p = 1.00); the perinatal mortality rates were 7.4% and 3.4%, for women with and without SCD, respectively (p = 0.164). Preeclampsia and HbSS phenotype were independently associated with increased the risk of prematurity and low birthweight. We conclude that multidisciplinary care of pregnant women with SCD may reduce maternal and perinatal mortality rates to levels comparable to that of pregnant women without SCD in low-resource settings.
Keywords: Sickle cell disease, pregnancy, maternal mortality, perinatal mortality, preeclampsia, multi-disciplinary care
INTRODUCTION
In sub-Saharan Africa, the risk of maternal death amongst pregnant women with sickle cell disease (SCD) is significantly higher (29-fold) compared to their non-SCD counterparts (1). A systematic review and meta-analysis by our team demonstrated that SCD is associated with an increased risk of maternal death compared to women without SCD in low-resource settings [odds ratio (OR) 22.8 (95% CI 14.7 – 35.5), p < 0.001] (2). Acute chest syndrome (ACS) is a major cause of death among pregnant women with SCD (3,4). Other adverse maternal outcomes in women with SCD include hypertensive disorders of pregnancy (preeclampsia, eclampsia), severe anemia, urinary tract infections, and increased odds of both antenatal and postnatal acute pain episodes (1,2,5–9). We have previously demonstrated, in a before-and-after study design, that multidisciplinary care of pregnant women with SCD reduced maternal mortality by 89% in a low-resource setting in sub-Saharan Africa (10).
In a prospective cohort study of pregnant women with and without SCD, we tested the hypothesis that multidisciplinary care of pregnant women with SCD in a low-resource setting will reduce maternal and perinatal mortality rates to similar levels in pregnant women without SCD.
METHODOLOGY
Study Design and Setting
The prospective cohort study included pregnant women with and without SCD at the Korle-Bu Teaching Hospital (KBTH), Accra, Ghana, from May 2015 to December 2016. The primary outcomes were maternal and perinatal mortality rates. The study received ethical approval from the Ethical and Protocol Review Committee of the College of Health Sciences, University of Ghana (approval number MS-Et/M.12-P4), and Vanderbilt University Medical Center (VUMC) Institutional Review Board, Nashville, TN, USA (approval number VUMC IRB 141050). The KBTH is a tertiary referral hospital in the capital, Accra, and the largest specialist hospital in Ghana. The maternity unit attends to about 11,000 pregnant women annually including approximately 200 pregnant women with SCD.
Study population
The cases were pregnant women with SCD (HbSS or HbSC) as determined by cellulose acetate hemoglobin electrophoresis at alkali pH. The controls were pregnant women with Hb AA selected from the same antenatal clinic to match cases for age and parity. Gestational age was estimated based on first trimester ultrasound scan. We excluded women who were referred with acute pregnancy-related complications, in active labour or after delivery, and those who were unable to comply with the follow-up requirements for prenatal visits.
Implementation Strategies
Our study drew on an organization model of innovation (11) which posits that the implementation strategy, via changes in the implementation climate (12) and implementation fit (13), influences implementation and clinical outcomes.
Our implementation strategy for this prospective study included a combination of four elements:
Generating buy-in for the creation of new treatment protocols: In 2012–14, members of KBTH and Vanderbilt teams met face to face multiple times to discuss the project. Five members of the KBTH team came to Vanderbilt University Medical Center for a month-long training program. During the month- long training in Nashville the team also systematically reviewed the literature to conduct a pooled analysis on maternal and perinatal outcomes in pregnancies associated with SCD (2). Also, the team conducted a combined retrospective (January 2010 to April 2015) and prospective (May 2015 to December 2016) case series of all maternal deaths in women with SCD at KBTH (3). This process highlighted service delivery, failures and challenges to delivering improved care while allowing for providers to develop new, acceptable, and feasible solutions.
Creation of a multidisciplinary sickle cell disease obstetric team: In 2015, the clinic expanded to a multidisciplinary team including three hematologists, two midwives, two laboratory scientists, a pediatrician, a dual-certified anesthesiologist and pulmonologist, and SCD experts from Vanderbilt and Duke University Medical Centers. A key component was the leadership of KBTH agreed to assign a nurse to the team that would not rotate off the team after several years.
Creation of an implementation climate where specific innovation was expected, supported and rewarded (14).
-
Establishment of a communication system to discuss current patients
Weekly meetings among members of the multidisciplinary team to discuss current cases, including any treatment failures.
Intervention strategies
Prior to this intervention, pregnant women with SCD were managed by obstetrician-led antenatal team which comprised a obstetricians, interns and midwives. In-patient admissions and care were scattered across five floors depending on when patients required inpatient care. There was no standard care protocol for outpatient and in-patient care and hematologist consult was sought using the hospitals procedures for internal referrals.
Outpatient and inpatient maternal assessment
After enrolment, an obstetrician and a hematologist followed all participants for inpatient and outpatient care. Pregnant women with SCD were risk-stratified based on clinical, laboratory and ultrasound maternal and fetal assessment. Women were followed bi-weekly until 34 weeks gestation and weekly thereafter until delivery; low-risk participants were followed monthly until 28 weeks gestation, biweekly until 36 weeks gestation, and thereafter weekly until delivery. The multidisciplinary team provided all outpatient medical care, and the pregnant women were admitted under the care of one of the two team obstetricians and hematologist was consulted.
Routine use of incentive spirometers, an evidence-based strategy demonstrated to prevent ACS in SCD (15), was considered essential for patient management but was not financially sustainable in a low-resource setting. Instead, we improvised using latex balloons. Repeated inflations were used as recommended routinely on the obstetric wards among pregnant women with SCD in place of incentive spirometers to prevent ACS during acute pain episodes and after cesarean sections (15). Blown balloons were inspected at the bedside as documentation of completing the task. Measurement of oxygen saturation was integrated into routine clinical care. A protocol was established for the transfer of patients with ACS to the Intensive Care Unit (ICU) for ventilatory support when necessary.
Assessment of SCD status of newborns
Newborn screening was performed for babies of pregnant women with SCD to determine their genotype. The heel prick method was used to obtain samples on filter paper between day two and six weeks of birth, and the isoelectric focusing method was used for newborn screening of SCD. Infants with a diagnosis of SCD were referred to the pediatric SCD unit for medical care.
Acute vaso-occlusive and ACS episodes.
Acute vaso-occlusive pain was distinguished from labor pain based on the absence of uterine contractions, evidence of labor progression and delivery. ACS was defined based on the presence of at least two of the following criteria including positive chest signs: temperature greater than 38⁰C, increased respiratory rate of > 20 per minute, positive chest pain or pulmonary auscultatory findings, increased oxygen requirement (SpO2 drop > 3% from a documented steady-state value on room air), and ± a new radio density on chest roentgenogram (17). In the absence of abdominal shielding, a chest x-ray was done for only postpartum women. A diagnosis of pneumonia was considered an ACS episode. All cases of acute vaso-occlusive pain and ACS episodes were adjudicated on to ensure a uniform definition. The observers and adjudicators were separate individuals.
Evaluation for malaria:
Malaria infection was defined using the WHO gold standard of examining stained thick smears under the light microscope (18) and parasite density (graded as + through to ++++) (19) together with the presence of clinical symptoms of fever, general malaise/illness and worsening anemia. All pregnant women (with and without SCD) were evaluated for malaria infection using the same criteria, and all pregnant women who did not have Glucose-6-phosphate dehydrogenase (G6PD) deficiency were given anti-malarial prophylaxis of Sulfadoxin-Pyrimethamine combination therapy starting at quickening and repeated at four-weekly intervals until 36 weeks gestational ages. The rate of asymptomatic malaria parasitemia was also assessed.
Spirometric evaluation
Spirometric evaluation was performed for all study participants on study visits at enrolment, 28 weeks, 36 weeks, and 6 weeks postpartum using the Morgan FVL spirometer according to the American Thoracic Society/European Respiratory Society (ATS/ERS) guidelines. The forced expiratory volume in 1 second (FEV1) and the percentage of the predicted normal values were determined using the Global Lung Initiative 2012 equations (20). Based on prior evidence, only the relationship between FEV1% predicted and acute pain and ACS incidence rate were included (21).
Statistical analysis:
Simple descriptive statistics were performed to determine proportions for discrete variables such as parity, marital status, and some relevant medical history. Height and weight were used to calculate body mass index (BMI) at study entry.
We compared maternal (preeclampsia vs no preeclampsia) and perinatal outcomes (term birth vs low birth weight) of the cases (HbSS and HbSC combined) to the controls. We separately compared the HbSS to HbSC group. In the models, SCD phenotype HbSS vs HbSC (reference category) was considered as a covariate. We used t-tests for continuous variables that were normally distributed, a Mann-Whitney U test for those that were not normally distributed, and a chi-square test for categorical variables. Two sets of multivariable logistic regression analyses were employed. In the first model cases (SCD) versus controls (non-SCD) was included as an independent covariate with the other covariates. In the second model Hb SS was compared to HbSC as an independent covariate with the other covariates. Independent covariates included: maternal age, BMI, parity, FEV1% predicted, and preeclampsia. Multicollinearity was assessed with a variance inflation factor > 3; residual statistics were reviewed for any cases with disproportionate influence on the model results. The risk of adverse outcome was estimated as odds ratio (OR) and their 95% confidence interval (CI). A statistical test of significance was set at p value < 0.05. Analyses were conducted using IBM SPSS Statistics (Version 25, Armonk, NY IBM).
RESULTS
Over a 20-month period, we enrolled 152 pregnant women with SCD (HbSS, 55; HbSC, 97) and 122 pregnant women without SCD. During the study, three participants with SCD and five without SCD were lost to follow-up. A total of 149 pregnant women with SCD (HbSS, 54; HbSC, 95) between the ages of 18 – 41 years (mean 29.2 ± 4.9) and 117 pregnant women with HbAA phenotype (controls) between the ages of 18 – 43 years (mean 29.7 ± 4.6) were included in the analysis. All demographic characteristics of the participants are described in Table 1.
TABLE 1.
Characteristics of the study population of pregnant women with sickle cell disease compared to pregnant women without sickle cell disease (n = 266)
| Characteristic | Pregnant women with SCD (n = 149) | Pregnant women without SCD (n = 117) | P valuea |
|---|---|---|---|
| Mother’s age at enrollment, years, mean (SD) | 29.2 (4.9) | 29.7 (4.6) | .532 |
| BMI (kg/m2), mean (SD) | 25.7 (5.0) | 30.7 (6.3) | <.001 |
| Gestational age at enrollment, weeks, mean (SD) | 24.2 (8.5) | 27.6 (6.0) | <.001 |
| Total time followed, weeks, mean (SD) | 19.0 (8.0) | 16.8 (6.0) | .014 |
| FEV1% predicted at baseline, mean (SD), (n = 252) | 87.1 (14.7) | 88.6 (14.9) | .427 |
| FVC % predicted at baseline, mean (SD), (n = 252) | 92.4 (16.3) | 96.0 (17.0) | .087 |
| FEV1/FVC % predicted, mean (SD), (n = 252) | 94.2 (7.9) | 92.2 (9.5) | .076 |
| Parity (n = 265) | .649b | ||
| No previous births (%) | 45.0 | 42.2 | |
| 1–4 prior births (%) | 54.4 | 56.9 | |
| 5 or more prior births (%) | 0.7 | 0.9 | |
| Hospitalization for malaria, (%), (n = 234) | 23.5 | 11.8 | .022 |
| Preeclampsia, (%), (n = 263) | 12.3 | 6.0 | .081 |
| Cesarean birth, (%), (n = 261) | 57.6 | 49.6 | .193 |
| Late preterm birth (birth <37 weeks), (%), (n = 265) | 26.4 | 16.2 | .048 |
| Intrauterine growth restriction, (%), (n = 257) | 6.3 | 2.7 | .175 |
| Low birth weight (<2500 g), (%), (n = 259) | 25.2 | 12.9 | .014 |
| Very low birth weight (<1500 g), (%), (n = 259) | 5.6 | 2.6 | .233 |
| Fetal death, (%) | 7.4 | 3.4 | .164 |
| Maternal death, (%), (n = 262) | 1.3 | 0.9 | 1.000c |
| Pain rate (events per patient year), mean (SD) | 1.8 (2.6) | ||
| ACS rate (events per patient year), mean (SD) | 0.4 (1.1) |
Abbreviations: ACS, acute chest syndrome; BMI, body mass index; FEV, forced expiratory volume; FVC, forced vital capacity; SCD, sickle cell disease.
t Test for mean difference or chi-square test for percentage difference, unless otherwise indicated.
Fisher’s exact test.
Mann–Whitney U test.
Multidisciplinary care reduced maternal and perinatal mortality rates amongst pregnant women with SCD compared to pregnant women without SCD
Pregnancy in women with SCD was not associated with increased rates of maternal and perinatal deaths compared to controls. Two participants with SCD died; one from a massive bilateral pulmonary embolism confirmed at autopsy, and the other from ACS. One participant in the control group died from an anaesthesia-related complication postpartum. Maternal mortality rates among women with and without SCD were 1.3% and 0.9% of live births, respectively (p = 1.00). Perinatal mortality rate was comparable in women with and without SCD at 7.4% and 3.4%, respectively (p = 0.164). The small sample size of maternal (n = 3) and perinatal (n = 15) deaths, in both the women with and without SCD, precluded any multivariable analysis.
Rates of preeclampsia and hospitalization for malaria was higher in pregnant women with SCD compared to those without SCD
The rate of preeclampsia in women with SCD was twice as high compared to pregnant women without SCD, 12.3% and 6.0% respectively (p = 0.081). The rate of hospitalization for malaria was significantly higher in pregnant women with SCD when compared to women without SCD at 23.5% and 11.8%, respectively (p = 0.022). The rates of cesarean delivery were similar in both cohorts (SCD 57.6% and non-SCD 49.6%; p = 0.193). There were no cases of asymptomatic malaria parasitemia.
Preterm births and low birthweight infants were higher amongst pregnant women with SCD compared to women without SCD
Preterm birth (< 37 weeks) and LBW (< 2,500g) were more frequent in pregnant women with SCD compared to women without SCD. The rates of preterm birth were 26.4% and 16.2% respectively in women with and without SCD, (p = 0.048). There was a significantly higher rate of LBW infants among women with SCD compared to those without SCD, 25.2% and 12.9%, respectively (p = 0.014). Though the rate of very LBW (< 1,500g) doubled in women with SCD compared to those without SCD, 5.6% and 2.6%, respectively, the difference was not significant (p = 0.233). Early preterm birth rates (< 34 weeks) among women with SCD was almost twice that of women without SCD, though the difference was not significant, 11.5% and 6.0%, respectively (p = 0.121). The rate of intrauterine growth restriction (IUGR) in pregnant women with SCD was more than double that in women without SCD, 6.3% and 2.7%, respectively, but the difference was also not significant (p = 0.175), Table 1.
Preeclampsia and low BMI were associated with preterm birth and low birthweight in women with SCD compared to women without SCD
Among all pregnant women in the cohort, preeclampsia was a risk factor for preterm birth (< 37 weeks), [OR 5.0 (95% CI 2.0 −12.5), p < 0.001], Table 2a. Similarly, among all pregnant women, a higher BMI was associated with a lower rate of preterm delivery < 37 weeks, [OR 0.9 (95% CI 0.8–1.0), p = 0.006], but not early preterm birth < 34 weeks. Table 2a.
TABLE 2.
Multivariable logistic regression models
| Variable | Odds ratio | 95% confidence interval | P value |
|---|---|---|---|
| Multivariable logistic regression models for premature birth at 37 and 34 weeks among all pregnant women with and without sickle cell disease in the prospective cohort (N = 248)a | |||
| Birth before 37 weeks | |||
| BMI | 0.91 | 0.85–0.97 | .006 |
| Pre-eclampsia | 5.04 | 2.04–12.49 | <.001 |
| SCD | 1.11 | 0.53–2.32 | .777 |
| Birth before 34 weeks | |||
| Pre-eclampsia | 2.72 | 0.90–8.23 | .077 |
| SCD | 1.93 | 0.74–5.26 | .171 |
| Multivariable logistic regression models for low birth weight below 2500 g, among all pregnant women with and without sickle cell disease in the prospective cohort (N = 244)a | |||
| BMI | 0.95 | 0.88–1.02 | .166 |
| Parity (one or more previous births) | 0.31 | 0.15–0.63 | .001 |
| Pre-eclampsia | 6.44 | 2.45–16.88 | <.001 |
| SCD | 1.62 | 0.72–3.63 | .245 |
| Multivariable logistic regression models for pre-eclampsia, among all pregnant women with and without sickle cell disease in the prospective cohort (N = 249)b | |||
| Mother’s age | 1.15 | 1.04–1.27 | 0.006 |
| Parity (one or more previous births) | 0.49 | 0.20–1.23 | .128 |
| SCD | 2.52 | 0.94–6.73 | .065 |
Abbreviations: BMI, body mass index; FEV, forced expiratory volume; SCD, sickle cell disease.
Screening models included mother’s age, BMI, FEV,% predicted parity, and pre-eclampsia.
Screening model included mother’s age. BMI, FEV1% predicted and parity.
Among all pregnant women in the cohort, preeclampsia was a risk factor for LBW (< 2,500g) [OR 6.4 (95% CI 2.5–16.9), p < 0.001] (Table 2b). Similarly, among all pregnant women, a higher parity was associated with a lower rate of LBW [OR 0.3 (95% CI 0.2–0.6), p = 0.001]. SCD alone when compared to the absence of SCD in the control group was not a risk factor for LBW (p = 0.245), Table 2b.
Among all pregnant women, the only significant risk factor for preeclampsia was maternal age. For every year increase in maternal age, there was a 15% increase in the odds of having preeclampsia [OR 1.15 (95% CI 1.04–1.27), p = 0.006], Table 2c.
Pregnant women with HbSS had higher rates of perinatal morbidities compared to pregnant women with HbSC disease
Preterm birth, LBW and IUGR rates were increased in pregnant women with HbSS compared to those with HbSC disease. Over 40% of participants with HbSS delivered preterm compared to 18% in those with HbSC (p = 0.003). In pregnant women with HbSS, the rate of early preterm birth was increased more than three times that in women with HbSC, 20.4% versus 6.4% respectively, (p = 0.010). The rate of LBW was significantly higher in pregnant women with HbSS compared to pregnant women with HbSC disease, 42.2% and 16.1%, respectively (p = 0.001). Approximately 13% of pregnant women with HbSS and 2.2% of pregnant women with HbSC had babies with IUGR (p = 0.013), Table 3.
TABLE 3.
Characteristics of pregnant women with HbSC and H bSS patients (n = 149)
| Characteristic | HbSC (n = 95)a | HbSS (n = 54)a | P valueb |
|---|---|---|---|
| Mother’s age at enrollment years, mean (SD) | 29.8 (4.7) | 28.2 (5.2) | .050 |
| BMI at baseline, mean(SD) | 26.6 (5.0) | 24.0 (4.5) | .002 |
| FEV1%predicted at baseline, mean (SD), (n = 148) | 89.5 (15.0) | 82.8 (13.1) | .007 |
| FVC %predicted at baseline, mean (SD), (n = 148) | 94.0 (16.9) | 89.6 (14.8) | .112 |
| FEV1/FVC %predicted at baseline, mean (SD), (n = 148) | 95.3 (7.5) | 92.3 (8.3) | .025 |
| Parity | .097c | ||
| No previous births (%) | 40.0 | 53.7 | |
| 1–4 prior births(%) | 58.9 | 46.3 | |
| 5 or more prior births(%) | 1.1 | 0.0 | |
| Hospitalization for malaria,(%), (n = 132) | 18.3 | 32.0 | .072 |
| Pre-eclampsia, (%). (n = 146) | 13.8 | 9.6 | .458 |
| Cesarean birth, (%), (n = 144) | 53.8 | 647 | .204 |
| Prematurity (birth <37 weeks), (%), (n = 148) | 18.1 | 40.7 | .003 |
| Prematurity (birth <34 weeks), (%). (n = 148) | 6.4 | 20.4 | .010 |
| Intrauterine growth restriction, (%), (n = 144) | 2.2 | 13.2 | .013d |
| Birth weight <2500 g, (%), (n = 143) | 16.1 | 42.0 | .001 |
| Birth weight <1500 g, (%), (n = 143) | 4.3 | 8.0 | .451d |
| Fetal death, (%), (%) | 4.2 | 13.0 | .098d |
| Maternal death, (%), (%) | 2.1 | 0.0 | .535d |
| Pain rate (events per patient year), mean (SD) | 1.6 (2.6) | 2.1 (2.6) | .184e |
| ACS rate (events per patient year), mean (SD) | 0.4 (1.1) | 0.4 (1.0) | .928e |
Abbreviations: ACS, acute chest syndrome; BMI, body mass index; FEV, forced expiratory volume; FVC, forced vital capacity.
Mean and SD for continuous variable, or percentage for categorical variable.
t Test for mean difference or chi-square test for percentage difference, unless otherwise indicated.
Mann-Whitney U test.
Fisher’s exact test.
Mid-p exact test.
HbSS phenotype, preeclampsia and low BMI were associated with premature delivery (< 37 weeks) in women with HbSS compared to HbSC
In a multivariable logistic regression model to determine the predictors of preterm birth in only women with SCD, preeclampsia and HbSS phenotype, compared to HbSC, increased the risk of preterm birth at < 37 weeks. Increasing BMI decreased the risk of preterm delivery. Preeclampsia was associated with a four-fold increased odds of preterm birth [OR 4.0 (95% CI 1.4 – 12.2), p = 0.013]; HbSS phenotype independently increased the risk of preterm birth by nearly three-fold [OR 2.7 (95% CI 1.2 – 6.3), p = 0.018]. A one-unit increase in BMI in all women was associated with an 11% reduction in preterm birth [OR 0.89 (95% CI 0.80 – 0.99), p = 0.026]. None of the covariates evaluated, including mother’s age, BMI, FEV1 % predicted, parity, preeclampsia, or SCD phenotype were associated with early preterm birth before 34 weeks, Table 4a.
TABLE 4.
Multivariable logistic regression models
| Variable | Odds ratio | 95% confidence interval | P value |
|---|---|---|---|
| Multivariable logistic regression models for premature birth at 37 and 34 weeks among pregnant women with HbSC and HbSS in the prospective cohort study (N = 145).a | |||
| Birth before 37 weeks | |||
| BMI | 0.89 | 0.80–0.99 | .025 |
| Pre-eclampsia | 4.05 | 1.35–12.18 | .013 |
| SS | 2.74 | 1.19–5.32 | .018 |
| Birth before 34 weeks | |||
| EMI | 0.88 | 0.76–1.01 | .075 |
| SS | 2.97 | 0.98–8.94 | .053 |
| Multivariable logistic regression models for low birth weight below 2500 g among pregnant women with HbSC and HbSS in the prospective cohort study (N = 142).a | |||
| BMI | 0.93 | 0.83–1.03 | .172 |
| Parity (one or more previous births) | 0.25 | 0.10–0.62 | .003 |
| Pre-eclampsia | 5.49 | 1.66–18.15 | .005 |
| SS | 3.46 | 1.41–8.48 | .007 |
| Multivariable logistic regress on models for pre-eclampsia among pregnant women with HaSC and HsSS in the prospective cohort study (N = 146).b | |||
| Mother’s age | 1.08 | 0.97–1.20 | .154 |
Abbreviation: BMI, body mass index.
Screening models included mother’s age BMI FEV,% predicted, parity, and pre-eclampsia. Reference category is SC.
Screening model induced mother’s age, BMI. FEV,% predicted, and parity. Reference category is SC.
HbSS phenotype, maternal age and preeclampsia were associated with low birthweight in pregnant women with HbSS compared to HbSC
In women with SCD, the risk of delivering LBW babies was associated with HbSS phenotype, parity, and preeclampsia. Preeclampsia increased the risk of LBW by five-fold in pregnant women with HbSS [OR 5.5 (95% CI 1.7 – 18.2), p = 0.005] when compared to women with HbSC, while HbSS was independently associated with greater than three-fold odds of having babies with LBW [OR 3.5 (95% CI 1.4 – 8.5), p = 0.007)]. Also, among women with SCD, increasing parity was associated with a much lower risk of having LBW babies, [OR 0.3 (95% CI 0.1 – 0.6), p = 0.003), Table 4b.
In another model to identify predictors of preeclampsia in pregnant women with SCD, none of the covariates, including maternal age, parity, hemoglobin phenotype, FEV1% predicted, or BMI predicted preeclampsia, Table 4c.
High prevalence of SCD among newborns of pregnant women with SCD
An extremely high prevalence of SCD and sickle cell trait (HbAS) occurred among newborns screened for SCD in the SCD cohort. Approximately 75% (104/138 livebirths) of newborns delivered by pregnant women with SCD (HbSS, 38; HbSC, 66) were screened for SCD. The samples of two babies (one mother each with HbSS and HbSC) did not elute and hence were excluded from the analysis. Among the mothers with SCD, approximately 15% (15/102) of newborns had SCD, with no difference by phenotype [HbSS 18.9% (7/37); HbSC 12.3% (8/65); p = 0.365]. Among mothers with HbSS and HbSC the prevalence of HbAS newborns were 64.9% (24/37) and 43.1% (28/65), respectively (p = 0.03). A total of 23 newborns delivered to mothers with HbSC disease had hemoglobin C trait. All the pregnant women without SCD had no abnormal hemoglobin detected on cellulose acetate Hb electrophoresis at alkali pH, so their newborns were not tested as part of the protocol.
DISCUSSION
We have prospectively confirmed that a multidisciplinary care approach reduced maternal and perinatal mortality in women with SCD to similar levels in those without SCD. We had previously demonstrated in a before-and-after study that a similar multidisciplinary care model reduced maternal mortality in women with SCD by nearly 90% when compared to historical controls (10).
The implementation climate for this project at KBTH was primed from 2011 to 2015, a high-risk antenatal clinic for pregnant women with SCD at KBTH in Accra, Ghana was started. At the time, the clinic was staffed exclusively with obstetricians and midwives to care for pregnant women with SCD. Despite the presence of the clinic, the maternal mortality rate remained approximately 10% (10) or 100deaths per 1,000 live births. As a comparison, the maternal mortality rate in the United States in 1910 was approximately 8.5% or 85 per 1,000 live births (22). Given the lack of any significant change in the maternal mortality rate after several years of the SCD-OB clinic, the leader of the team (SAO) sought alternative approaches to decrease the maternal mortality rate. A key component to promotion of a implementation climate where specific innovation was expected, supported and rewarded (14).
To support innovation, KBTH providers adapted Vanderbilt-Meharry Center of Excellence in Sickle Cell Disease protocols for management of acute vaso-occlusive pain and ACS. Additionally, the KBTH team members were supported and rewarded with two grants (American Society of Hematology, Clinical Research Training Institute, recipient EVA; and University of Ghana, Office of Research, Innovation and Development (ORID), recipient SAO, along with philanthropy funds). We cannot state with certainty which of the multiple new supportive care strategies were most effective for decreasing maternal and perinatal mortality rates. Only an implementation trial with random allocation of the most intensive interventions, would determine the added benefit of the joint outpatient hematologist and obstetrician clinics.
A unique attribute of our study of pregnant women with SCD is the high proportion of women with HbSS and HbSC allowing for a direct comparison of maternal and perinatal risk factors and outcomes between the two SCD phenotypes. Previous studies have reported higher rates of preterm birth and LBW among pregnant women with SCD compared to non-SCD controls (8,23–26). In this study, preterm birth and LBW were increased in pregnant women with HbSS when compared to women with HbSC disease. In a Jamaican cohort study of 113 pregnant women with HbSC disease, preeclampsia was also associated with low birth weight (27).
Regardless of SCD phenotype, preeclampsia increased the risk of preterm birth and low birthweight in all pregnant women with SCD. Thus, targeting a strategy to prevent and treat preeclampsia among women with SCD in both low- and high-income countries would likely improve perinatal outcomes. Further, among women with SCD, each subsequent pregnancy was associated with a lower odds of low birthweight newborns.
Our study has anticipated strengths and weaknesses. One distinctive strength of the study is that we have provided one of the few estimates for the prevalence of SCD (specifically HbSS and HbSC) in the offspring of pregnant women with SCD living in West Africa. Based on the observation that at least 15% of the newborns of mothers were diagnosed with SCD, as opposed to the expected 2% (28) in this region, targeting pre-conceptual genetic counselling in young women with SCD may have a significant impact in making informed decisions about having a child with SCD. Potentially, lessons learned in this population can be used to inform effective genetic counselling strategies in both low and high resource settings.
One of the weakness of the study is that we did not demonstrate whether our approach is reproducible in a non-tertiary facility in low-resource settings. We are currently attempting to implement our multidisciplinary care strategy at other non-academic hospitals in Accra, Ghana.
In conclusion, implementation of a multidisciplinary care strategy for pregnant women with SCD reduced maternal and perinatal mortality to comparable levels in pregnant women without SCD in a low-resource setting. Further efforts to reduce perinatal morbidity must target preeclampsia and its impact on IUGR and prematurity.
Key point summary.
Multidisciplinary care for women with SCD in a low-resource setting decreases the maternal and perinatal mortality rates to levels in the general population
ACKNOWLEDGEMENTS
The authors acknowledge the leadership, Department of Obstetrics and Gynecology, KBTH for administrative support, funds from University of Ghana, Office of Research, Innovation and Development (ORID); Doris Duke Charitable Foundation; Burroughs Wellcome Foundation, Phillips Family Donation; Aaron Ardoin Foundation for Sickle Cell Anemia; and J.C. Peterson, M.D. endowed chair funds from Vanderbilt University School of Medicine. We also acknowledge the donation of pulse oximetry machines by the team from Vanderbilt University Medical Center.
RESEARCH FUNDING:
Financial support from the University of Ghana Research Fund/8/LMG-008; Doris Duke Charitable Foundation; Burroughs Wellcome Foundation; Phillips Family Donation; Aaron Ardoin Foundation for Sickle Cell Anemia; Vanderbilt University School of Medicine (J.C.Peterson, M.D.); Vanderbilt University Medical Center Gift Funds; the National Center for Research Resources, Grant/Award Number: Grant UL1 RR024975-01; the National Center for Advancing Translational Sciences, Grant/Award Number: Grant 2 UL1 TR000445-06.
REFERENCES
- 1.Muganyizi PS, Kidanto H. Sickle Cell Disease in Pregnancy: Trend and Pregnancy Outcomes at a Tertiary Hospital in Tanzania. PLOS ONE. 2013;8(2):e56541. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.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 Int J Obstet Gynaecol. 2015;123(5):691–8. [DOI] [PubMed] [Google Scholar]
- 3.Asare EV, Olayemi E, Boafor T et al. A case series describing causes of death in pregnant women with sickle cell disease in a low-resource setting. Am J Hematol. 2018;93(7):E167–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Asnani MR, McCaw-Binns AM, Reid ME. Excess Risk of Maternal Death from Sickle Cell Disease in Jamaica: 1998–2007. PLOS ONE. 2011;6(10):e26281. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.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(6):936–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Villers MS, Jamison MG, Castro LMD, James AH. Morbidity associated with sickle cell disease in pregnancy. Am J Obstet Gynecol. 2008;199(2):125.e1–125.e5. [DOI] [PubMed] [Google Scholar]
- 7.Ngô C, Kayem G, Habibi A et al. Pregnancy in sickle cell disease: maternal and fetal outcomes in a population receiving prophylactic partial exchange transfusions. Eur J Obstet Gynecol Reprod Biol. 2010;152(2):138–42. [DOI] [PubMed] [Google Scholar]
- 8.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(6):1127–30. [DOI] [PubMed] [Google Scholar]
- 9.Oteng‐Ntim E, Ayensah B, Knight M, Howard J. Pregnancy outcome in patients with sickle cell disease in the UK – a national cohort study comparing sickle cell anaemia (HbSS) with HbSC disease. Br J Haematol. 2015;169(1):129–37. [DOI] [PubMed] [Google Scholar]
- 10.Asare EV, Olayemi E, Boafor T et al. Implementation of multidisciplinary care reduces maternal mortality in women with sickle cell disease living in low-resource setting. Am J Hematol. 2017;92(9):872–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Jacobs SR, Weiner BJ, Reeve BB, Hofmann DA, Christian M, Weinberger M. Determining the predictors of innovation implementation in healthcare: a quantitative analysis of implementation effectiveness. BMC Health Serv Res. 2015;15(1):6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Jacobs SR, Weiner BJ, Bunger AC. Context matters: measuring implementation climate among individuals and groups. Implement Sci. 2014;9(1):46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Weiner BJ, Lewis CC, Stanick C et al. Psychometric assessment of three newly developed implementation outcome measures. Implement Sci. 2017;12(1):108. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Klein KJ, Sorra JS. The Challenge of Innovation Implementation. Acad Manage Rev. 1996;21(4):1055–80. [Google Scholar]
- 15.Bellet PS, Kalinyak KA, Shukla R, Gelfand MJ, Rucknagel DL. Incentive Spirometry to Prevent Acute Pulmonary Complications in Sickle Cell Diseases. N Engl J Med. 1995;333(11):699–703. [DOI] [PubMed] [Google Scholar]
- 16.World Health Organization, editor. The WHO application of ICD-10 to deaths during pregnancy, childbirth and the puerperium, IDC MM. Geneva: World Health Organization; 2012. 68 p. [Google Scholar]
- 17.Yawn BP, Buchanan GR, Afenyi-Annan AN et al. Management of Sickle Cell Disease: Summary of the 2014 Evidence-Based Report by Expert Panel Members. JAMA. 2014;312(10):1033–48. [DOI] [PubMed] [Google Scholar]
- 18.Warhurst DC, Williams JE. ACP Broadsheet no 148. July 1996. Laboratory diagnosis of malaria. J Clin Pathol. 1996;49(7):533–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Organization WH. Basic Malaria Microscopy. World Health Organization; 2010. 58 p. [Google Scholar]
- 20.Quanjer PH, Hall GL, Stanojevic S, Cole TJ, Stocks J. Age- and height-based prediction bias in spirometry reference equations. Eur Respir J. 2012;40(1):190–7. [DOI] [PubMed] [Google Scholar]
- 21.Ivy ZK, Kassim AA, Payne AB et al. Lower Airway Obstruction Is Associated with Increased Vaso-Occlusive Pain Episodes in Adults with Sickle Cell Anemia. Blood. 2015;126(23):978–978. [Google Scholar]
- 22.Flaherty Laurie. Achievements in Public Health, 1900–1999: Healthier Mothers and Babies. Journal of Emergency Nursing. 2000;26(2):151–152. [Google Scholar]
- 23.Kahtani MAA, AlQahtani M, Alshebaily MM, Elzaher MA, Moawad A, AlJohani N. Morbidity and pregnancy outcomes associated with sickle cell anemia among Saudi women. Int J Gynecol Obstet. 2012;119(3):224–6. [DOI] [PubMed] [Google Scholar]
- 24.Jama FEA, Gasem T, Burshaid S, Rahman J, Suleiman SAA, Rahman MS. Pregnancy outcome in patients with homozygous sickle cell disease in a university hospital, Eastern Saudi Arabia. Arch Gynecol Obstet. 2009;280(5):793–7. [DOI] [PubMed] [Google Scholar]
- 25.Serjeant GR, Loy LL, Crowther M, Hambleton IR, Thame M. Outcome of pregnancy in homozygous sickle cell disease. Obstet Gynecol. 2004;103(6):1278–85. [DOI] [PubMed] [Google Scholar]
- 26.Thame MM, Osmond C, Serjeant GR. Fetal growth in women with homozygous sickle cell disease: an observational study. Eur J Obstet Gynecol Reprod Biol. 2013;170(1):62–6. [DOI] [PubMed] [Google Scholar]
- 27.Thame M, Lewis J, Trotman H, Hambleton I, Serjeant G. The mechanisms of low birth weight in infants of mothers with homozygous sickle cell disease. Pediatrics. 2007;120(3):e686–693. [DOI] [PubMed] [Google Scholar]
- 28.Ohene-Frempong K, Oduro J, Tetteh H, Nkrumah F. Screening newborns for sickle cell disease in Ghana. Pediatrics. 2008;121(Supplement 2):S120–S121. [Google Scholar]