Development and validation of a machine-learning model for prediction of shoulder dystocia.

OBJECTIVES: To develop a machine-learning (ML) model for prediction of shoulder dystocia (ShD) and to externally validate the model's predictive accuracy and potential clinical efficacy in optimizing the use of Cesarean delivery in the context of suspected macrosomia. METHODS: We used electronic health records (EHR) from the Sheba Medical Center in Israel to develop the model (derivation cohort) and EHR from the University of California San Francisco Medical Center to validate the model's accuracy and clinical efficacy (validation cohort). Subsequent to application of inclusion and exclusion criteria, the derivation cohort included 686 singleton vaginal deliveries, of which 131 were complicated by ShD, and the validation cohort included 2584 deliveries, of which 31 were complicated by ShD. For each of these deliveries, we collected maternal and neonatal delivery outcomes coupled with maternal demographics, obstetric clinical data and sonographic fetal biometry. Biometric measurements and their derived estimated fetal weight were adjusted (aEFW) according to gestational age at delivery. A ML pipeline was utilized to develop the model. RESULTS: In the derivation cohort, the ML model provided significantly better prediction than did the current clinical paradigm based on fetal weight and maternal diabetes: using nested cross-validation, the area under the receiver-operating-characteristics curve (AUC) of the model was 0.793 ± 0.041, outperforming aEFW combined with diabetes (AUC = 0.745 ± 0.044, P = 1e-16 ). The following risk modifiers had a positive beta that was > 0.02, i.e. they increased the risk of ShD: aEFW (beta = 0.164), pregestational diabetes (beta = 0.047), prior ShD (beta = 0.04), female fetal sex (beta = 0.04) and adjusted abdominal circumference (beta = 0.03). The following risk modifiers had a negative beta that was < -0.02, i.e. they were protective of ShD: adjusted biparietal diameter (beta = -0.08) and maternal height (beta = -0.03). In the validation cohort, the model outperformed aEFW combined with diabetes (AUC = 0.866 vs 0.784, P = 0.00007). Additionally, in the validation cohort, among the subgroup of 273 women carrying a fetus with aEFW ≥ 4000 g, the aEFW had no predictive power (AUC = 0.548), and the model performed significantly better (0.775, P = 0.0002). A risk-score threshold of 0.5 stratified 42.9% of deliveries to the high-risk group, which included 90.9% of ShD cases and all cases accompanied by maternal or newborn complications. A more specific threshold of 0.7 stratified only 27.5% of the deliveries to the high-risk group, which included 63.6% of ShD cases and all those accompanied by newborn complications. CONCLUSION: We developed a ML model for prediction of ShD and, in a different cohort, externally validated its performance. The model predicted ShD better than did estimated fetal weight either alone or combined with maternal diabetes, and was able to stratify the risk of ShD and neonatal injury in the context of suspected macrosomia. Copyright © 2019 ISUOG. Published by John Wiley & Sons Ltd.

Maternal characteristics and mid-pregnancy serum biomarkers as risk factors for subtypes of preterm birth.

OBJECTIVE: To examine the relationship between maternal characteristics, serum biomarkers and preterm birth (PTB) by spontaneous and medically indicated subtypes. DESIGN: Population-based cohort. SETTING: California, United States of America. POPULATION: From a total population of 1 004 039 live singleton births in 2009 and 2010, 841 665 pregnancies with linked birth certificate and hospital discharge records were included. METHODS: Characteristics were compared for term and preterm deliveries by PTB subtype using logistic regression and odds ratios adjusted for maternal characteristics and obstetric factors present in final stepwise models and 95% confidence intervals. First-trimester and second-trimester serum marker levels were analysed in a subset of 125 202 pregnancies with available first-trimester and second-trimester serum biomarker results. MAIN OUTCOME MEASURE: PTB by subtype. RESULTS: In fully adjusted models, ten characteristics and three serum biomarkers were associated with increased risk in each PTB subtype (Black race/ethnicity, pre-existing hypertension with and without pre-eclampsia, gestational hypertension with pre-eclampsia, pre-existing diabetes, anaemia, previous PTB, one or two or more previous caesarean section(s), interpregnancy interval ≥ 60 months, low first-trimester pregnancy-associated plasma protein A, high second-trimester α-fetoprotein, and high second-trimester dimeric inhibin A). These risks occurred in 51.6-86.2% of all pregnancies ending in PTB depending on subtype. The highest risk observed was for medically indicated PTB <32 weeks in women with pre-existing hypertension and pre-eclampsia (adjusted odds ratio 89.7, 95% CI 27.3-111.2). CONCLUSIONS: Our findings suggest a shared aetiology across PTB subtypes. These commonalities point to targets for further study and exploration of risk reduction strategies. TWEETABLE ABSTRACT: Findings suggest a shared aetiology across preterm birth subtypes. Patterns may inform risk reduction efforts.

Serum relaxin levels and kidney function in late pregnancy with or without preeclampsia.

OBJECTIVE: Relaxin, a potent pregnancy-related hormone, has been proposed to be a major mediator of renal physiology in normal pregnancy. We wished to test relaxin levels in pregnancy and preeclampsia. METHODS: We performed precise physiologic measurements of kidney function in 38 normal peripartum women and 58 women with preeclampsia. We measured serum relaxin levels prior to delivery and over the first 4 postpartum weeks utilizing a modern, validated ELISA. Results were compared to those of 18 normal women of childbearing age. RESULTS: Relaxin levels were substantially elevated in women prior to delivery (364 ± 268 vs. 15 ± 16 pg/ml) and fell rapidly over the first postpartum week reaching normal non pregnant levels by Week 2 (32 ± 64 vs. 15 ± 16 pg/ml). No differences were seen between relaxin levels in normal pregnancy as compared to preeclampsia (364 ± 268 vs. 376 ± 241 pg/ml) despite substantial and persistent abnormalities in GFR (149 ± 33 vs. 89 ± 25 ml/min), albuminuria (14 vs. 687 mg/g) and mean arterial pressure (80 ± 8 vs. 111 ± 18). Furthermore no correlation could be established between physiologic measures (GFR, MAP, RBF, RVR) and relaxin levels (p > 0.3), either in the overall population or any of the subgroups. CONCLUSION: Relaxin is indeed significantly elevated in the serum of women during late pregnancy and the early puerperium. However, serum relaxin does not appear to influence BP, renal vascular resistance, renal blood flow or GFR in late pregnancy or in women with preeclampsia.

Out-of-hospital births in California 1991-2011.

OBJECTIVE: We investigated the frequencies and characteristics of out-of-hospital births in a 20-year period in California, where 1 of every 7 births in the United States occurs. STUDY DESIGN: Birth certificate records of deliveries in California between 1991 and 2011 were analyzed. Out-of-hospital births were assessed by year, parity, gestational age and maternal race/ethnicity. RESULTS: In the 20-year period there were 10 593,904 deliveries, of which 46 243 occurred out of hospital (0.44%). Out-of-hospital births decreased from 0.54 to 0.38% per year between 1991 and 2004, and increased from 0.41% in 2005 to 0.61% in 2011. In contrast, preterm out-of-hospital births declined from 7.2% in 2006 to 5.0% in 2011. The frequency of vaginal birth after cesarean in the out-of-hospital birth cohort increased from 1.2% (n=19) in 1996 to 4.2% (n=82) in 2011. CONCLUSION: California birth records from a 20-year period show an increase in out-of-hospital births from years 2005 to 2011, following a period of decline from 1991 to 2004.

Effect of L-arginine therapy on the glomerular injury of preeclampsia: a randomized controlled trial.

OBJECTIVE: To assess the benefit of l-arginine, the precursor to nitric oxide, on blood pressure and recovery of the glomerular lesion in preeclampsia. METHODS: Forty-five women with preeclampsia were randomized to receive either l-arginine or placebo until day 10 postpartum. Primary outcome measures including mean arterial pressure, glomerular filtration rate, and proteinuria were assessed on the third and 10th days postpartum by inulin clearance and albumin-to-creatinine ratio. Nitric oxide, cyclic guanosine 3'5' monophosphate, endothelin-1, and asymmetric-dimethyl-arginine and arginine levels were assayed before delivery and on the third and 10th days postpartum. Healthy gravid women provided control values. Assuming a standard deviation of 10 mm Hg, the study was powered to detect a 10-mm Hg difference in mean arterial pressure (alpha .05, beta .20) between the study groups. RESULTS: No significant differences existed between the groups with preeclampsia before randomization. Compared with the gravid control group, women with preeclampsia exhibited significantly increased serum levels of endothelin-1, cyclic guanosine 3'5' monophosphate, and asymmetric-dimethyl-arginine before delivery. Despite a significant increase in postpartum serum arginine levels due to treatment, no differences were found in the corresponding levels of nitric oxide, endothelin-1, cyclic guanosine 3'5' monophosphate, or asymmetric-dimethyl-arginine between the two groups with preeclampsia. Further, there were no significant differences in any of the primary outcome measures with both groups demonstrating similar levels in glomerular filtration rate and equivalent improvements in both blood pressure and proteinuria. CONCLUSION: Blood pressure and kidney function improve markedly in preeclampsia by the 10th day postpartum. Supplementation with l-arginine does not hasten this recovery. LEVEL OF EVIDENCE: I.

Maternal serum markers, characteristics and morbidly adherent placenta in women with previa.

OBJECTIVE: To examine associations with morbidly adherent placenta (MAP) among women with placenta previa. STUDY DESIGN: Women with MAP (cases) and previa alone (controls) were identified from a cohort of 236,714 singleton pregnancies with both first and second trimester prenatal screening, and live birth and hospital discharge records; pregnancies with aneuploidies and neural tube or abdominal wall defects were excluded. Logistic binomial regression was used to compare cases with controls. RESULT: In all, 37 cases with MAP and 699 controls with previa alone were included. Risk for MAP was increased among multiparous women with pregnancy-associated plasma protein-A (PAPP-A) ⩾95th percentile (⩾2.63 multiple of the median (MoM); adjusted OR (aOR) 8.7, 95% confidence interval (CI) 2.8 to 27.4), maternal-serum alpha fetoprotein (MS-AFP) ⩾95th percentile (⩾1.79 MoM; aOR 2.8, 95% CI 1.0 to 8.0), and 1 and ⩾2 prior cesarean deliveries (CDs; aORs 4.4, 95% CI 1.5 to 13.6 and 18.4, 95% CI 5.9 to 57.5, respectively). CONCLUSION: Elevated PAPP-A, elevated MS-AFP and prior CDs are associated with MAP among women with previa.

Analysis of prenatal and gestational care given to women with epilepsy.

OBJECTIVE: To assess past care practices of neurologists and obstetricians to identify areas in which practice patterns differ from currently accepted optimal care. METHODS: Retrospective chart review of 155 women identified as having a diagnosis of epilepsy (or seizure disorder) who had been pregnant any time between January 1988 and December 1995 and were admitted to Stanford University Hospital for delivery. A total of 161 pregnancies (132 women) were selected for study. RESULTS: An obstetrician was seen at some point during the pregnancy in 99% of the pregnancies, whereas a neurologist was seen at least once in only 64% of the pregnancies. In the 3 months before conception, an obstetrician was seen in 5% of the pregnancies and a neurologist was seen in 15%. Seventy-five percent of the patients taking antiepileptic medication and 65% of the untreated patients had documentation of folate supplementation at any time during pregnancy. Vitamin K supplementation in the final month of pregnancy was documented for only 41% of those receiving antiepileptic drugs. In over one-third of the pregnancies the mother did not have a maternal serum alpha-fetoprotein measure documented and a similar percentage did not receive genetic counseling. Monitoring of the maternal serum concentration of the non-protein-bound fraction of the prescribed antiepileptic drugs was not documented. CONCLUSIONS: We identified specific omissions of appropriate vitamin supplementation, genetic counseling, and drug level monitoring. Educational efforts should be targeted to improve the management of pregnancy in women with epilepsy.

Lupus pregnancy. Case-control prospective study demonstrating absence of lupus exacerbation during or after pregnancy.

To assess whether pregnancy is associated with exacerbation of systemic lupus erythematosus (SLE), a variety of clinical markers of disease activity in 28 pregnant patients with SLE (33 pregnancies) were compared with the same markers in age-, race-, organ system-, and disease severity-matched nonpregnant women with SLE. Both groups were followed up for periods of up to one year after delivery. Eight patients elected abortion for nonmedical reasons. In all patient groups, there were no differences between pregnant and nonpregnant patient groups in frequency of any disease activity marker studied including therapy. However, new proteinuria occurred in four pregnant patients compared with one nonpregnant patient, and thrombocytopenia attributable to SLE occurred in five pregnant patients and one nonpregnant patient. Renal disease, when it occurred, more closely resembled pregnancy-induced hypertension than lupus nephritis. It is concluded that pregnancy complications are frequent, but the assertion that pregnancy causes exacerbation of SLE remains unproved.

The clinical significance of predictions based on screening second trimester mean arterial pressure: adverse maternal [corrected] and infant outcomes.

The design of a trial of primary prevention of hypertension in pregnancy rests on both the ability to identify women who are at risk and the definition of a clinically important outcome. The risk of developing antepartum hypertension can now be assessed nonivasively by the midpoint of pregnancy. However, maternal hypertension is not always associated with a clinically important adverse outcome for either mother or infant. The purpose of this study was to prospectively assess whether increasing risk of antepartum hypertension is associated with increasing rates of clinically important maternal and/or infant morbidity. We assembled a prospective cohort of 720 women with singleton pregnancies. The proportion of pregnancies complicated by both antepartum hypertension and maternal and/or infant morbidity increased significantly between low, moderate, and high risk groups (0.2, 6 and 58.8%, respectively, p less than 0.0001). We conclude that a trial of primary prevention of hypertension in pregnancy should include a measure of significant morbidity in mother and infant.

Development and validation of a multivariate predictor of mortality in very low birth weight.

Accurate prognosis is critical to the design of all prospective research aimed at improving survival. Predictions based on birth weight, gestational age, or any other single variable, fail to take into account the potentially important contribution of other factors. In order to develop a practical and accurate multivariate model, we studied all singleton pregnancies resulting in viable liveborn infants who weighed less than or equal to 1500 g at birth during 1984 and 1985 at the New York Hospital-Cornell Medical Center. When gestational age, birth weight, and/or crown-heel length were considered, no maternal characteristics were significant predictors of mortality. The model with the maximal predictive accuracy (84.5%) used birth weight and 5-minute Apgar score to calculate a probability of mortality. This prognostic model was then validated in a separate cohort of singletons born in 1986. We conclude that clinical trials should require stratification before randomization, using the calculated probability of mortality, rather than birth weight or gestational age alone. Given the ability of models, such as the one presented here, to generate reasonable estimates of mortality, this information might also be used in the clinical setting to assist parents and physicians in individualized decision-making processes for a given infant.

Effect of maternal glucose ingestion compared with maternal water ingestion on the nonstress test.

Ninety-three nonstress tests were performed on 57 nondiabetic patients at greater than 34 weeks' gestation. Maternal whole blood glucose levels were measured before beginning the nonstress test and within five minutes of the second fetal heart rate acceleration. There was a significant rise in maternal whole blood glucose levels in the maternal glucose ingestion group but not in the maternal water ingestion group. There was no significant difference in the mean time to reactivity between the two groups. These results suggest that maternal glucose ingestion does not affect time to reactivity or the incidence of reactive nonstress tests.

The use of vibroacoustic stimulation during the abnormal or equivocal biophysical profile.

OBJECTIVE: To determine whether vibroacoustic stimulation during the biophysical profile can change the fetal behavioral state and thus improve the score without increasing the false-negative rate of the test. METHODS: Eighty-one patients whose biophysical profile scores were 6 or lower after 15 minutes of observation had an electronic artificial larynx applied to the maternal abdomen in the region of the fetal head for 3 seconds, followed by continued observation for fetal movement, tone, and breathing for 15 minutes. We compared the obstetric and neonatal outcomes of 41 patients whose biophysical profile scores improved to normal after vibroacoustic stimulation with those of 283 patients whose scores were normal without vibroacoustic stimulation. RESULTS: Vibroacoustic stimulation did improve an abnormal or equivocal biophysical profile score to normal in 67 of 81 cases (82%). No antepartum stillbirths or perinatal deaths occurred. There was no increase in the obstetric and neonatal complication rates of cesarean delivery for fetal distress, meconium staining of the amniotic fluid, and the incidence of small for gestational age infants. CONCLUSION: Vibroacoustic stimulation improved the biophysical profile scores in most cases, an effect seen throughout the third trimester. Vibroacoustic stimulation did not appear to increase the false-negative rate of the biophysical profile and may reduce the incidence of unnecessary obstetric intervention.

Intrapartum uterine rupture.

The association between diethylstilbestrol (DES) exposure in utero and uterine malformations resulting in poor reproductive performance is well established. A case is presented of uterine rupture in a patient exposed to DES in utero who had no known predisposing factors for uterine rupture.

Early prediction of antepartum hypertension.

We validated a mid-pregnancy screening mean arterial pressure (MAP2) of 85 mmHg or higher as a significant predictor of hypertension in pregnancy. During the 17-month period from October 1984 through February 1986, 730 women, or 16% of all women cared for and delivered at our institution, were screened at or near 20 weeks of amenorrhea. Of the 139 women with a MAP2 of 85 mmHg or higher, 21.6% developed antepartum hypertension, compared with only 0.7% of the 591 women with a MAP2 below 85 mmHg. The screening MAP2 level of 85 mmHg was the optimal cutoff for MAP2 as a screening test. Controlling for the value of the screening MAP2, the only other important predictors of antepartum hypertension were chronic hypertension and diabetes mellitus. Using these three variables, the probability that an individual pregnant woman will develop antepartum hypertension can be assessed with a high degree of accuracy (84.5%) by 20 weeks of amenorrhea. This assessment is noninvasive and simple to use. Three distinct levels of risk have been defined; the moderate- and high-risk groups warrant careful surveillance during pregnancy and may be reasonable groups in which to test preventive interventions.

Should all pregnant patients be offered prenatal diagnosis regardless of age?

OBJECTIVE: To assess the acceptance of prenatal genetic diagnosis by patients younger than 35 years old who are therefore not yet at great risk for non-disjunction trisomies based on maternal age. METHODS: The patients were counseled regarding the following: 1) the age-related risk of chromosomal abnormalities, 2) the procedure-related risk of fetal loss, 3) clinical implications of chromosomal abnormalities, 4) the need for complete counseling by a certified genetic counselor, and 5) the patient expense of $600-1200 if third-party reimbursement was not available. Patients were recruited from the private practice of the senior author at the New York Hospital--Cornell Medical Center. Five hundred ninety-one patients were offered prenatal genetic diagnosis. The outcome measure was the patient's decision to undergo prenatal diagnosis even though the risk of a non-disjunction trisomy was expected to be low based on maternal age. Amniocentesis was performed in 128 patients and chorionic villus sampling in five. RESULTS: One hundred thirty-three patients (22.5%) chose prenatal diagnosis. Karyotype was obtained in 131 procedures, but two were unsuccessful. One of the 131 karyotypes was abnormal and the patient chose to terminate the pregnancy. CONCLUSIONS: The data showed the following: 1) Inappropriate influence of patients by the health provider was not evident; 2) routine offering of genetic diagnosis enhanced the autonomy of pregnant women; 3) the potential increase in the loss of pregnancies that accompanies this practice is ethically justified; and 4) there are no compelling cost-benefit objections to such a practice.

Pregnancy complicated by primary antiphospholipid antibody syndrome.

BACKGROUND: Primary antiphospholipid antibody syndrome is a clinical entity that may threaten the health of both fetus and mother. CASE: We report a fatal case of primary antiphospholipid antibody syndrome in a woman who developed catastrophic disease due to multisystem thrombosis in the postpartum period following a fetal death. Three years before her admission, primary antiphospholipid antibody syndrome was diagnosed on the basis of high titers of immunoglobulin G anticardiolipin antibody, a positive lupus anticoagulant, a false-positive VDRL, and fibrin deposits in the biopsy of a palmar lesion. CONCLUSION: The physician must recognize the potentially catastrophic complications of pregnancy and the postpartum period in patients with antiphospholipid antibodies, and appropriate patient counseling should be provided.

Cost-effectiveness of a trial of labor after previous cesarean.

OBJECTIVE: To determine the cost-effective method of delivery, from society's perspective, in patients who have had a previous cesarean. METHODS: We completed an incremental cost-effectiveness analysis of a trial of labor relative to cesarean using a computerized model for a hypothetical 30-year old parturient. The model incorporated data from peer-reviewed studies, actual hospital costs, and utilities to quantify health-related quality of life. A threshold of $50,000 per quality-adjusted life-years was used to define cost-effective. RESULTS: The model was most sensitive to the probability of successful vaginal delivery. If the probability of successful vaginal birth after cesarean (VBAC) was less than 0.65, elective repeat cesarean was both less costly and more effective than a trial of labor. Between 0.65 and 0.74, elective repeat cesarean was cost-effective (the cost-effectiveness ratio was less than $50,000 per quality-adjusted life-years), because, although it cost more than VBAC, it was offset by improved outcomes. Between 0.74 and 0.76, trial of labor was cost-effective. If the probability of successful vaginal delivery exceeded 0.76, trial of labor became less costly and more effective. Costs associated with a moderately morbid neonatal outcome, as well as the probabilities of infant morbidity occurring, heavily impacted our results. CONCLUSION: The cost-effectiveness of VBAC depends on the likelihood of successful trial of labor. Our modeling suggests that a trial of labor is cost-effective if the probability of successful vaginal delivery is greater than 0.74. Improved algorithms are needed to more precisely estimate the likelihood that a patient with a previous cesarean will have a successful vaginal delivery.

Diagnosis of fetal death in utero by real-time ultrasound.

The efficacy of real-time ultrasound for the diagnosis of fetal death or hydatidiform mole was evaluated during a 1-year period. During this time, 116 patients were referred to the obstetric ultrasound service for the confirmation of clinical diagnoses. In 24 of 46 patients (52%) presenting in the first half of pregnancy, the referring diagnosis was confirmed. In 1 case of an early intrauterine pregnancy with a degenerating myoma, the ultrasound diagnosis of molar pregnancy was in error. In 48 of 70 patients (69%) referred after 20 weeks' gestation, the clinical diagnosis was confirmed. In no instance was either a false-positive or false-negative diagnosis made with real-time ultrasound in the last half of pregnancy. This method should prove to be the method of choice in diagnosing intrauterine fetal death.

Randomized comparison of intravenous nitroglycerin and magnesium sulfate for treatment of preterm labor.

OBJECTIVE: To compare the safety and efficacy of high-dose intravenous (IV) nitroglycerin with those of IV magnesium sulfate for acute tocolysis of preterm labor. METHODS: Thirty-one women with preterm labor before 35 weeks' gestation were assigned randomly to IV magnesium sulfate or IV nitroglycerin for tocolysis. Preterm labor was defined as the occurrence of at least two contractions in 10 minutes, with cervical change or ruptured membranes. Acute tocolysis was defined as tocolysis for up to 48 hours. Magnesium sulfate was administered as a 4-g bolus, then at a rate of 2-4 g/h. Nitroglycerin was administered as a 100-microg bolus, then at a rate of 1- to 10-microg/kg/min. The primary outcome measure was achievement of at least 12 hours of successful tocolysis. RESULTS: Thirty patients were available for analysis. There were no significant differences in gestational age, cervical dilation, or incidence of ruptured membranes between groups at the initiation of tocolysis. Successful tocolysis was achieved in six of 16 patients receiving nitroglycerin, compared with 11 of 14 receiving magnesium sulfate (37.5 versus 78.6%, P = .033). Tocolytic failures (nitroglycerin versus magnesium sulfate) were due to persistent contractions with cervical change or rupture of previously intact membranes (five of 16 versus two of 14), persistent hypotension (four of 16 versus none of 14), and other severe side effects (one of 16 versus one of 14). Maternal hemodynamic alterations were more pronounced in patients who received nitroglycerin, and 25% of patients assigned to nitroglycerin treatment had hypotension requiring discontinuation of therapy. CONCLUSION: Tocolytic failures were more common with nitroglycerin than with magnesium sulfate. The hemodynamic alterations noted in patients receiving nitroglycerin, including a 25% incidence of persistent hypotension, might limit the usefulness of IV nitroglycerin for the acute tocolysis of preterm labor.

Maternal platelet count at delivery in patients with idiopathic thrombocytopenic purpura, not related to perioperative complications.

BACKGROUND: The relationship between maternal platelet count at the time of cesarean section and perioperative and postpartum operative complications was examined. STUDY DESIGN: A retrospective analysis of 46 pregnancies in 41 women with histories of idiopathic thrombocytopenic purpura was performed. Thirty-five patients had platelet counts greater than 100,000 before delivery and 11 had counts less than 100,000. Statistical comparisons were made using Student's t test and chi-square test. RESULTS: The perioperative and postpartum course for patients in the two groups differed significantly only in platelet counts at delivery. Change in hematocrit (from admission to postpartum), estimated blood loss at cesarean section, incidence of wound complications or transfusion, were not significantly different. There were no neonatal complications. CONCLUSIONS: Mild thrombocytopenia in patients with idiopathic thrombocytopenic purpura is unlikely to be associated with an increase in blood loss, infection, wound complication, or need for transfusion.