Obstetric hospitalizations in the United States for women with systemic lupus erythematosus and rheumatoid arthritis

Authors


Abstract

Objective

To estimate the national occurrence of pregnancies in women with systemic lupus erythematosus (SLE) and rheumatoid arthritis (RA) and to compare pregnancy outcomes in these patients with those in women with pregestational diabetes mellitus (DM) and with the general obstetric population.

Methods

We studied the 2002 Nationwide Inpatient Sample of the Healthcare Cost and Utilization Project to estimate the number of obstetric hospitalizations, deliveries, and cesarean deliveries in women with SLE, RA, pregestational DM, and the general obstetric population. Pregnancy outcomes included length of hospital stay, hypertensive disorders including preeclampsia, premature rupture of membranes, and intrauterine growth restriction.

Results

Of an estimated 4.04 million deliveries, 3,264 occurred in women with SLE, 1,425 in women with RA, and 13,574 in women with pregestational DM. Women with SLE, RA, and pregestational DM had significantly increased rates of hypertensive disorders compared with the general obstetric population (23.2%, 11.1%, 27.4%, and 7.8%, respectively), longer hospital stays, and significantly higher risk of cesarean delivery. Although women with SLE, RA, and pregestational DM were significantly older than women in the general obstetric population, disparities in the risk of adverse outcomes of pregnancy remained statistically significant after adjustment for maternal age.

Conclusion

To our knowledge, this is the first study to examine national data on pregnancy outcomes in women with common rheumatic diseases. As with underlying pregestational DM, women with SLE and RA appear to have a higher age-adjusted risk of adverse outcomes of pregnancy and longer hospital stays than do pregnant women in the general population, and careful antenatal monitoring should be performed.

Systemic lupus erythematosus (SLE) and rheumatoid arthritis (RA) are chronic autoimmune diseases that have a strong female predominance and commonly affect women of childbearing age. The prevalence of SLE in the US is estimated to be ∼1–4 per 1,000 women (1, 2), with the peak age at onset during the childbearing years. Studies have consistently shown that women comprise 90% of all lupus patients. With the exception of premature ovarian failure due to therapy with alkylating agents, fertility rates for women with SLE are normal (3). Pregnancy outcomes in women with SLE have been studied extensively in specialty clinic settings, and the results have been mixed with regard to the effects of pregnancy on disease flares in women with SLE (4, 5). Most, although not all, studies have shown an increased occurrence of pregnancy loss, premature delivery, and preeclampsia in pregnant women with SLE (6–10). The majority of studies examining SLE pregnancies have been prospective or retrospective analyses of pregnant women followed up at tertiary-care academic centers and may not reflect the spectrum of pregnancy experiences across the country (6–10). Furthermore, because of the small number of pregnant patients at any given center or hospital each year, assembled cohorts necessarily span many years in order to achieve numbers that are amenable to analysis.

The antiphospholipid antibody syndrome (APS) is an acquired autoimmune thrombophilia that is associated with significant obstetric morbidity and mortality (11). Diagnosis of APS requires repeated positive results of tests for antiphospholipid antibodies or assays for lupus anticoagulant, as well as documented arterial thromboses, venous thromboses, or specific complications of pregnancy (12). Although the prevalence of APS is unknown, it has been suggested that up to 30% of SLE patients will develop APS (13).

The prevalence of RA among women in the US is estimated to be ∼1.4% (14). Unlike SLE, the incidence of RA increases with age, with a median age at onset of 59 years in women. The prevalence of RA among women ages 16–44 years is estimated to be ∼1–2 cases per 1,000 women in the UK (15). Recent prevalence estimates for young women with RA in the US are not available. Studies have suggested that fertility and fecundity may be reduced in women with RA, but the cause for this is unknown (3). While there have been many studies evaluating the effect of pregnancy on disease activity in mothers with RA, relatively few studies have addressed pregnancy outcomes in women with the disease (16). There is 1 report of increased rates of premature delivery and lower birth weight in infants born to mothers with RA (17). As with SLE, studies of RA have involved cohorts of patients at tertiary care centers or through voluntary registries and may not reflect the national population of RA patients. Little is known about the rates of preeclampsia, cesarean section, and other pregnancy outcomes among women with RA.

Diabetes mellitus (DM) is a chronic metabolic disease that commonly affects women of childbearing years and is associated with the development of progressive vascular complications. Type 1 DM results from autoimmune destruction of islet cells and mainly affects children and young adults, whereas type 2 DM results from a combination of insulin resistance and relative insulin deficiency. The prevalence of type 2 DM has been increasing in teenagers and young adults, likely due to an epidemic of obesity (18). The elevated rates of adverse outcomes of pregnancy in women with pregestational DM, including preeclampsia, preterm birth, macrosomia, and congenital malformations, are well established (18). Pregestational DM complicates an estimated 0.2–0.5% of pregnancies (19) and is considered to be an independent risk factor for the development of preeclampsia (20) and for cesarean delivery (21, 22). National estimates of adverse outcomes of pregnancy in women with pregestational DM have not been reported.

In this study, we sought to estimate the nationwide number of obstetric hospitalizations and deliveries occurring in women with an underlying diagnosis of SLE, RA, APS, or pregestational DM. We compared the rates of hospitalizations in each group that were complicated by adverse obstetric outcomes, including hypertensive disorders of pregnancy, cesarean delivery, premature rupture of membranes (PROM), and intrauterine growth restriction (IUGR).

MATERIALS AND METHODS

We utilized the Nationwide Inpatient Sample (NIS) of the Healthcare Cost and Utilization Project (23). The NIS, established in 1988, is the largest all-payer inpatient care database that is publicly available in the US and contains clinical and demographic information available from discharge abstracts. The database contains information from all inpatient hospitalizations in 995 hospitals, which were sampled to approximate a 20% stratified sample of community hospitals in the US. The 2002 sample was drawn from hospitals in 35 states: California, Colorado, Connecticut, Florida, Georgia, Hawaii, Illinois, Iowa, Kansas, Kentucky, Maine, Maryland, Massachusetts, Michigan, Minnesota, Missouri, Nebraska, Nevada, New Jersey, New York, North Carolina, Ohio, Oregon, Pennsylvania, Rhode Island, South Carolina, South Dakota, Tennessee, Texas, Utah, Vermont, Virginia, Washington, West Virginia, and Wisconsin. The NIS included all the non-federal short-stay hospitals listed by the American Hospital Association. These hospitals were divided into 60 strata based on region, location, teaching status, bed size category, and ownership. Within each stratum, a systematic random sample of hospitals was drawn, equal in size to 20% of that stratum. Detailed information on the design of the survey and the magnitude of sampling errors associated with the estimates is available in the technical documentation (24).

Information available from discharge abstracts included maternal age, length of hospital stay, principal diagnosis, up to 14 secondary diagnoses, principle procedure, and up to 14 secondary procedures. Data on race were not collected for 11 states and were therefore not used in the main analysis. Diagnoses and procedures were coded by International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes (25). In this data set, the unit of analysis is hospitalization, rather than patient.

All obstetric hospitalizations were included in the analysis. Hospitalizations were ascribed to the diagnosis (i.e., SLE [ICD-9-CM code 710.0], RA [714.0], APS [795.79], and pregestational DM [250.0]) that was found in the discharge summary. Hospitalizations for which both SLE and RA were reported as discharge diagnoses were excluded from further study. The control group was defined as any obstetric hospitalization in women who did not have a diagnosis of SLE, RA, or pregestational DM. Hospitalizations that were associated with delivery of an infant were defined as any hospitalization with procedure codes for cesarean section and other procedures for inducing or assisting delivery, or ICD-9-CM codes for normal delivery and delivery with complications. All other hospitalizations were considered antenatal hospitalizations that did not result in delivery. The ICD-9-CM codes used in this study are available from the authors.

Weighting functions using hospital and discharge weights were applied to the sample data to estimate the total number of hospitalizations, delivery-associated hospitalizations, and cesarean section for women with SLE, RA, APS, pregestational DM, and the entire obstetric population across the US during 2002. National estimates and 95% confidence intervals (95% CIs) were calculated.

For each group of patients (SLE, RA, APS, pregestational DM, and control), the maternal age and length of hospital stay were tabulated, and differences in means were compared using Student's t-test. The proportion of hospitalizations for each group with the following pregnancy outcomes was calculated: antenatal hospitalizations, hypertensive disorders including preeclampsia, PROM, cesarean section, and IUGR. The chi-square test was applied to compare the difference in rates between each group and the control group. Statistical comparisons between groups (SLE, RA, APS, and pregestational DM) were not performed.

For each dichotomous outcome of interest (antenatal hospitalization, hypertensive disorders, IUGR, PROM, and cesarean delivery), logistic regression was performed for each disease (SLE, RA, APS, and pregestational DM) using maternal age as a covariate. Multivariate logistic regression was then performed using the covariates disease group and maternal age. Multiple linear regression was used to determine coefficients for length of hospital stay. A multivariate analysis of the subset of the data using only the 24 states that reported data on race was performed in order to estimate the effects of race on pregnancy outcomes. In order to remove bias introduced by multiple hospitalizations of a single patient, the analysis was repeated using only hospitalizations that resulted in delivery of an infant.

Because of the large numbers of observations in the data set, P values less than 0.01 were considered statistically significant. All analyses were performed using Stata version 8.0 software (Stata, College Station, TX).

RESULTS

Obstetric hospitalizations.

There were 976,527 obstetric hospitalizations in the 2002 NIS database. Of these, 943 occurred in women with a discharge diagnosis of SLE, 360 occurred in women with a diagnosis of RA, 719 in women with APS, and 8,338 in women with pregestational DM. Only 1.6% of hospitalizations in women with a diagnosis of SLE also carried a diagnosis of pregestational DM, and 2.3% in women with RA had a concomitant diagnosis of pregestational DM. The majority of hospitalizations for APS (95%) occurred in women without concomitant diagnoses for SLE or RA. Only 3.1% of SLE patients carried a diagnosis of APS, and 1.1% of RA patients were also diagnosed as having APS.

After application of sampling weights, there were an estimated 4.56 million obstetric hospitalizations: approximately 4,425 occurred in women with SLE, 1,685 in women with RA, 3,398 in women with APS, and 38,713 in women with pregestational DM. Nationwide estimates of all obstetric hospitalizations, deliveries, and cesarean deliveries in patients with diagnoses of SLE, RA, APS, pregestational DM, and the entire obstetric population are presented in Table 1. Of all obstetric hospitalizations, we estimated there to be ∼4.04 million deliveries in the US in 2002, which compares with 4.02 million deliveries reported by the National Vital Statistics System for that year (26). Of these deliveries, approximately 3,264 (0.08%) occurred in women with a diagnosis of SLE, 1,425 (0.035%) in women with RA, 2,884 (0.07%) in women with APS, and 24,427 (0.6%) in women with pregestational DM.

Table 1. Nationwide estimates for all obstetric hospitalizations, deliveries, and cesarean deliveries in 2002, by diagnostic group
Diagnosis, outcomePoint estimate95% confidence interval
Systemic lupus erythematosus  
 All obstetric hospitalizations4,4253,854–4,997
 Deliveries3,2642,840–3,688
 Cesarean delivery1,2661,055–1,477
Rheumatoid arthritis  
 All obstetric hospitalizations1,6851,431–1,939
 Deliveries1,4251,204–1,646
 Cesarean delivery516403–630
Antiphospholipid syndrome  
 All obstetric hospitalizations3,3982,805–3,991
 Deliveries2,8842,354–3,413
 Cesarean delivery1,2851,027–1,543
Pregestational diabetes mellitus  
 All obstetric hospitalizations38,71334,640–42,786
 Deliveries24,42721,820–27,053
 Cesarean delivery13,57412,192–14,965
All  
 All obstetric hospitalizations4,557,2244,274,380–4,840,067
 Deliveries4,041,1513,784,988–4,297,314
 Cesarean delivery1,070,761997,670–1,143,853

Only 11.2% of obstetric hospitalizations in the control group were for complications of pregnancy that were not associated with the delivery of an infant. Patients with RA and APS had a slightly increased risk of antenatal hospitalization (15.6% and 15.2%, respectively; P = 0.006), whereas patients with SLE and pregestational DM had significantly higher rates of nondelivery hospitalization (26.3% and 38.1%, respectively versus controls; P < 0.001 versus controls).

Pregnancy outcomes.

Pregnancy outcomes for delivery-associated hospitalizations are listed in Table 2. The mean maternal age in the control group was 27.5 years, and the mean length of hospital stay was 2.5 days. Rates of outcomes calculated for the control population approximated the established rates in the US. Approximately 7.8% of pregnancies were complicated by hypertensive disorders including preeclampsia. Of the pregnancies in the control population, 26.5% resulted in cesarean section, which approximates the rate of 26.1% reported by the National Vital Statistics System for 2002 (26).

Table 2. Rates of pregnancy outcomes (delivery-associated hospitalizations) by diagnostic group
Diagnostic groupAge, mean yearsLength of stay, mean no. of days% with hypertensive disorders% with premature rupture of membranes% with intrauterine growth restriction% with cesarean delivery
  • *

    P < 0.001 versus control, by Student's t-test for means and by chi-square test for proportions.

  • P < 0.01 versus control, by Student's t-test for means and by chi-square test for proportions.

Systemic lupus erythematosus29.6*4.2*23.2*55.3*39.4*
Rheumatoid arthritis29.8*3.1*11.16.4*3.437.2*
Antiphospholipid syndrome31.3*3.6*13.0*1.85.1*44.6*
Pregestational diabetes mellitus30.1*4.0*27.4*4.21.855.5*
Control27.52.57.83.91.626.5

Women with a diagnosis of SLE had significantly higher rates of adverse outcomes of pregnancy. Maternal age was higher in SLE patients (29.6 years) compared with the controls. Even after controlling for maternal age, SLE patients had a 3-fold higher rate of hypertensive disorders than did women in the control group (23.2% versus 7.8%; P < 0.001) and a 3-fold higher rate of IUGR (5.3% versus 1.6%; P < 0.001). Rates of cesarean section were significantly higher in the SLE patients than in the control group (39.4% versus 26.5%; P < 0.001).

Although to a lesser degree than with SLE, women diagnosed as having RA had higher rates of pregnancy complications than did the control group. Again, maternal age was slightly higher (29.8 years) than that in the control group. Rates of hypertensive disorders of pregnancy, PROM, and IUGR were slightly, but significantly, higher than in the control population. Length of hospital stay in patients with RA was longer, and this may be a reflection of the increased rate of cesarean section (37.2%; P < 0.001 versus controls).

Similarly, women diagnosed as having APS had higher rates of adverse outcomes of pregnancy compared with controls. They were, on average, 4 years older than women in the control group, and more of them experienced hypertensive disorders (13.0%) and IUGR (5.1%) than did the women in the control group. As expected, women with pregestational DM had higher rates of hypertensive disorders and cesarean deliveries than did the women in the control population.

Risk of adverse obstetric outcomes.

Results of multivariate analyses examining the risk of adverse outcomes in hospitalizations associated with delivery are shown in Table 3. Logistic regression analyses were adjusted for maternal age. A diagnosis of SLE significantly increased the odds of hypertensive disorders (odds ratio [OR] 3.3 [95% CI 2.8–4.0]), IUGR (OR 3.5 [95% CI 2.5–4.9]), and cesarean delivery (OR 1.6 [95% CI 1.4–1.9]). Likewise, patients with SLE had an increased length of hospital stay even after adjustment for cesarean delivery.

Table 3. Risk of adverse obstetric outcomes in the US for delivery-associated hospitalizations, by diagnostic group*
 Systemic lupus erythematosusRheumatoid arthritisAntiphospholipid syndromePregestational diabetes mellitus
  • *

    Values are the odds ratio (95% confidence interval). P values were determined by logistic regression, except for length of stay, which were determined by linear regression.

  • Univariate coefficient for length of stay, adjusted for age and cesarean delivery.

  • P < 0.001.

  • §

    Multivariate analysis includes age and group. Cesarean delivery was an additional covariate in the linear regression for length of stay.

  • P < 0.01.

Length of stay    
 Univariate1.4 (1.3–1.6)0.4 (0.2–0.6)0.7 (0.5–0.9)1.0 (1.0–1.1)
 Multivariate§1.4 (1.3–1.6)0.4 (0.1–0.6)0.7 (0.5–0.8)1.0 (1.0–1.1)
Hypertensive disorders    
 Univariate3.4 (2.9–4.1)1.4 (1.0–2.0)1.7 (1.3–2.1)4.4 (4.2–4.7)
 Multivariate§3.3 (2.8–4.0)1.4 (0.9–1.9)1.6 (1.3–2.0)4.4 (4.2–4.7)
Intrauterine growth restriction    
 Univariate3.7 (2.7–5.2)2.3 (1.2–4.3)3.7 (2.6–5.3)1.2 (1.0–1.5)
 Multivariate§3.5 (2.5–4.9)2.2 (1.2–4.1)3.4 (2.4–4.9)1.2 (1.0–1.5)
Cesarean delivery    
 Univariate1.6 (1.4–1.9)1.5 (1.2–1.9)1.9 (1.6–2.2)3.2 (3.0–3.4)
 Multivariate§1.6 (1.4–1.9)1.5 (1.2–1.9)1.9 (1.6–2.2)3.2 (3.0–3.4)

Women with RA had increased risks of adverse outcomes of pregnancy compared with the general population, including a greater risk of cesarean delivery (OR 1.5 [95% CI 1.2–1.9]), IUGR (OR 2.2 [95% CI 1.2–4.1]), and hypertensive disorders (OR 1.4, 95% CI 0.9–1.9). Women with APS had increased risks of hypertensive disorders (OR 1.6 [95% CI 1.3–2.0]), IUGR (OR 3.4 [95% CI 2.4–4.9]), and cesarean delivery (OR 1.9 [95% CI 1.6–2.2]).

Outcomes in patients with pregestational DM appeared to be of similar magnitude to those in patients with SLE, with increased odds of having hypertensive disorders (OR 4.4 [95% CI 4.2–4.7]), cesarean delivery (OR 3.2 [95% CI 3.0–3.4]), and increased length of stay. In contrast, there did not appear to be an increased risk of IUGR in diabetic women.

Race/ethnicity of the patients.

Of the 35 states participating in the 2002 NIS, only 24 reported data on race. A subset analysis limited to the states that reported race variables is shown in Table 4. A higher proportion of SLE patients were African American as compared with the control population, and fewer patients were Hispanic. The majority of obstetric patients with RA or with APS were Caucasian, whereas the majority of patients with pregestational DM were non-Caucasian.

Table 4. Race/ethnicity of patients, by diagnostic group*
Diagnostic group% Caucasian% African American% Hispanic% other
  • *

    Includes data from the 24 states that report data on race/ethnicity.

  • Includes Asian, Pacific Islander, Native American, and other ethnicity.

  • P < 0.001 versus control, by chi-square test.

Systemic lupus erythematosus51221710
Rheumatoid arthritis6414158
Antiphospholipid syndrome68101111
Pregestational diabetes mellitus4123269
Control51142411

Impact of race/ethnicity on pregnancy outcomes.

In order to understand the impact of race upon outcomes of pregnancy, we performed logistic regression analyses of delivery-associated hospitalizations limited to this subset of the data (Table 5). Although African American race was associated with an increased risk of all adverse obstetric outcomes examined, after adjustment for race/ethnicity, the risks for each of the disease groups remained similar to those in the models that did not include race as a covariate. Odds ratios for hypertensive disorders were significantly increased in patients with SLE (OR 2.9 [95% CI 2.3–3.6]), APS (OR 1.6 [95% CI 1.2–2.1]), and pregestational DM (OR 4.3 [95% CI 4.0–4.7]), as were odds for cesarean delivery (OR 1.5 [95% CI 1.3–1.8], OR 1.8 [95% CI 1.5–2.2], and OR 3.0 [95% CI 2.8–3.2], respectively). SLE, RA, and APS were associated with an increased risk of IUGR (OR 3.9 [95% CI 2.7–5.7], OR 2.8 [95% CI 1.4–5.5], and OR 3.3 [95% CI 2.2–5.1], respectively).

Table 5. Risk of adverse obstetric outcomes for delivery-associated hospitalizations that included data on race/ethnicity, by diagnostic and ethnic groups*
Adverse outcome, modelDiagnostic groupEthnic group
SLERAAPSPregestational DMCaucasianAfrican AmericanHispanicOther
  • *

    The univariate model was adjusted for age. The multivariate model included disease group and maternal age. The multivariate with ethnicity model included disease group, maternal age, and ethnicity. Values are the odds ratio (95% confidence interval). SLE = systemic lupus erythematosus; RA = rheumatoid arthritis; APS = antiphospholipid syndrome; DM = diabetes mellitus.

  • P < 0.001 versus control, by logistic regression.

  • P < 0.01 versus control, by logistic regression.

Hypertensive disorders        
 Univariate3.1 (2.5–3.8)1.5 (1.0–2.3)1.7 (1.3–2.2)4.4 (4.1–4.8)    
 Multivariate3.0 (2.4–3.7)1.4 (0.9–2.2)1.6 (1.2–2.1)4.4 (4.1–4.8)    
 Multivariate with ethnicity2.9 (2.3–3.6)1.4 (0.9–2.2)1.6 (1.2–2.1)4.3 (4.0–4.7)Reference1.5 (1.5–1.5)0.9 (0.8–0.9)0.7 (0.7–0.7)
Intrauterine growth restriction        
 Univariate4.2 (2.9–6.2)3.0 (1.5–5.8)3.7 (2.5–5.6)1.1 (0.9–1.5)    
 Multivariate4.0 (2.7–5.9)2.9 (1.5–5.6)3.4 (2.3–5.2)1.1 (0.9–1.5)    
 Multivariate with ethnicity3.9 (2.7–5.7)2.8 (1.4–5.5)3.3 (2.2–5.1)1.1 (0.9–1.5)Reference1.2 (1.2–1.3)0.7 (0.6–0.7)1.0 (1.0–1.0)
Cesarean delivery        
 Univariate1.6 (1.3–1.9)1.3 (1.0–1.8)1.8 (1.5–2.2)3.0 (2.9–3.3)    
 Multivariate1.5 (1.3–1.8)1.3 (1.0–1.7)1.8 (1.5–2.2)3.1 (2.9–3.3)    
 Multivariate with ethnicity1.5 (1.3–1.8)1.3 (1.0–1.7)1.8 (1.5–2.2)3.0 (2.8–3.2)Reference1.2 (1.2–1.2)1.1 (1.1–1.1)0.9 (0.9–0.9)

DISCUSSION

Pregestational DM is the most common underlying medical disorder in pregnant women in the US, and the perinatal risks to both mother and infant have been well described (27). In contrast, the scope of pregnancies affected by maternal autoimmune diseases is not well established, and the magnitude of pregnancy complications remains unknown. We used data from the NIS database to estimate the number of pregnancies in women with SLE, APS, and RA in the US during 2002, and we compared the outcomes of those pregnancies with the outcomes in the general obstetric population as well as with the outcomes of pregnancies complicated by pregestational DM. National estimates in this study approximated 4.04 million deliveries, an estimate similar to that reported by National Vital Statistics System for 2002 (26). Additionally, the proportion of pregnancies complicated by pregestational DM (0.6% of deliveries) was similar to that reported in the literature (27). We estimated the number of deliveries in women with a diagnosis of SLE to be ∼3,264, and the number in women with APS to be ∼2,884. In contrast, far fewer pregnancies occurred in women with a diagnosis of RA (n = 1,425). This was not unexpected, given that the incidence of RA in women peaks in the sixth decade, well after most women have completed their families. The prevalence of RA in women of childbearing age (∼1–2 per 1,000 women in the UK) (15) appears to be close to that of SLE, and it is possible that women with RA may have reduced fertility and fecundity as compared with healthy women (3).

We found significantly increased rates of hypertensive disorders of pregnancy, IUGR, length of hospital stay, and cesarean delivery in patients with a diagnosis of SLE compared with pregnant women without this diagnosis. The increased risk for the development of these pregnancy-related complications persisted even after adjustment for maternal age, race, and comorbid conditions, including pregestational DM and APS. Rates of pregnancy complications including hypertensive disorders and cesarean delivery in the control population in this data set were comparable with those reported in the literature.

Nearly 25% of pregnancies in women diagnosed as having SLE were complicated by hypertensive disorders, a rate similar in magnitude to that seen in cohorts followed up at academic centers (10, 28–30) as well as in women with pregestational DM. Previous studies have suggested that several factors, including preexisting hypertension, renal insufficiency, presence of APS, and active SLE, may increase the risk of preeclampsia in pregnancies complicated by SLE (29). We found surprisingly low rates of APS in the SLE population in this study, and we believe that the diagnosis of APS was missed in many cases. Based on previous reports, we expected a 10-fold higher rate of APS in SLE patients (13). Unfortunately, the nature of the database precluded evaluation of women for the presence of antiphospholipid antibodies or positive results on lupus anticoagulant tests. Additionally, we were surprised to find that the rate of hypertensive disorders in women with a diagnosis of APS (13%) was much lower than that in women with SLE (23%). Again, this is likely due to classification bias, in which SLE patients may not have been appropriately diagnosed as also having secondary APS.

In the diabetic population, obesity, pregestational hypertension, proteinuria, and severity of underlying disease have been identified as risk factors for preeclampsia (18, 19). Our study confirms the findings of other large epidemiologic studies showing an increased risk of preeclampsia among African Americans (31); however, in multivariate regression models, the increased risk of hypertensive disorders in women with SLE or pregestational DM remained after controlling for race.

Very few population-based studies have been published that estimate the frequency of adverse obstetric outcomes among women with SLE. One study evaluating 1993–1994 data from the California Health Information for Policy Project, which links records from birth certificates and maternal and infant hospital discharges in California, was recently reported (32). Yasmeen et al identified 555 deliveries in women with SLE and compared those pregnancy outcomes with outcomes in a control group of 600,000 deliveries in women without SLE. The results showed that women with SLE had higher rates of adverse outcomes of pregnancy, including hypertensive complications, preterm delivery, cesarean delivery, IUGR, and fetal death, than did women without SLE. Additionally, women with SLE had a longer length of hospital stay and delivered lower birth-weight babies than did the control group. The majority of results from their study were of similar magnitude to our data, including the mean length of hospital stay (3.8 versus 4.2 days), rates of IUGR (6.7% versus 5.3%), and rates of cesarean delivery (38.2% versus 39.4%). However, the rate of hypertensive disorders of pregnancy found in their study (2.9%) was significantly lower than the rate found in our study (23.2%). Additionally, the rate of hypertensive disorders found in the control population (0.4%) in their study was well below the rate of 7.8% in our study, which approximated the established rate of 6–8% in the general population of the US (20).

The strengths of the study by Yasmeen et al (32) were the large number of hospitalizations available for analysis as well as the ability to link maternal and neonatal/infant discharge records. Limitations resulted from the extremely low rates of hypertensive disorders seen in both the SLE and control populations and the fact that regression analyses of associations between SLE and adverse outcomes of pregnancy were restricted to univariate analyses. Potentially important confounders, such as maternal age, ethnicity, and comorbid conditions including APS, were not included in the models. In contrast, in our study, which included data from 35 states, adjustments were made for important confounding variables. Additionally, rates of hypertensive disorders and cesarean delivery in our control population approximated those reported in the general population (26), which increases the confidence in our estimates of risk.

The impact of underlying RA upon pregnancy outcomes is less well studied. There has been a suggestion of an elevated rate of preeclampsia in women with inflammatory arthritides in some studies, but the difference often did not meet statistical significance for RA alone (33, 34). We found a small, but statistically significant, increased prevalence of hypertensive disorders in women with a diagnosis of RA. The risk of hypertensive disorders in women with RA was elevated, but it did not meet our definition of statistical significance in bivariate and multivariate models (ORs 1.38–1.44). Early data from the Organization of Teratology Information Services (OTIS) Rheumatoid Arthritis in Pregnancy study showed a higher rate of preterm births in women with RA (26%) as compared with women without RA (4.3%) (17). Other studies have shown decreased birth weight of infants born to mothers with RA as compared with that of infants born to mothers in control groups (33, 35) and a possible association between disease activity and lower birth weight (35). Because we were unable to link maternal and infant discharge records, we were not able to directly examine birth weight in this study. However, we did find a higher rate of IUGR in women with RA as compared with the control population (3.4% versus 1.6%).

The rate of cesarean delivery was higher in all of the autoimmune disease groups compared with the control population. Increased rates of cesarean delivery have previously been reported for SLE (32), RA (34), and pregestational DM (18, 22). Current indications for cesarean delivery, both emergent and elective, are numerous and varied: previous cesarean delivery, fetal distress, failed labor, multiple gestation, macrosomia, malpresentation, active maternal infections, maternal medical conditions, and maternal request (36). Certainly, women with diagnoses of SLE, RA, APS, or pregestational DM had higher rates of acute maternal conditions, including hypertensive disorders of pregnancy, which may have led to the performance of a cesarean delivery. However, the patients may have had several concomitant indications for cesarean delivery, and the decision to operate may have not been made solely on the basis of underlying maternal disease. The nature of this data set did not allow for analysis of specific indications for cesarean delivery, and therefore, causal associations between underlying disease and the increased cesarean rates seen in our study populations remain matters of speculation.

Several limitations of this study need to be addressed. The first limitation concerns the inability to validate discharge diagnoses for autoimmune diseases or pregnancy outcomes. Because the database used discharge diagnoses provided by the treating physicians, no standardized criteria were used to define cases, which could lead to overreporting or underreporting of cases. Thus, the sensitivity and specificity of these diagnoses could not be determined. A second limitation is that the unit of analysis was hospital discharge, rather than the patient. Since there is no way to link multiple records to a given individual, patients hospitalized several times for the same conditions were counted as separate units for analysis. To address this issue, we performed separate analyses limited to hospitalizations associated with delivery of an infant, and this did not alter the conclusions drawn from the data. It was assumed that childbirth occurs once during each pregnancy and that each delivery represents data from a separate pregnancy. Because no patient-identifying information was provided, we were unable to link the discharge records of the infants to the maternal records, so information such as birth weight, congenital anomalies, and other infant-specific outcomes could not be evaluated.

A third limitation was that the NIS contained only data from hospitalizations. All deliveries that occurred at home or in birthing centers were not captured in the data set, nor was information about pregnancies that ended in early fetal loss that did not necessitate hospitalization. And last, the database did not contain any information about disease severity, medication use, prenatal care, obstetric history, smoking status, or other factors that may be important confounders that influence pregnancy outcomes.

Despite these limitations, the strengths of this study rest in the large numbers of pregnancies from an ethnically diverse nationwide sample of hospitalizations available for analysis. To our knowledge, this is the first report of estimates of the number of deliveries occurring in women in the US who have SLE, APS, or RA. Although the estimated numbers of pregnant women with SLE and RA did not approach those with pregestational DM, they did suggest that relatively large numbers of women with these rheumatic conditions do choose to bear children and that the majority appear to have good pregnancy outcomes.

We found increased age-adjusted risks of hypertensive disorders, IUGR, and cesarean delivery associated with a diagnosis of SLE, APS, or RA, and we confirmed reports of increased complications of pregnancy in women with pregestational DM. These results remained unchanged after adjustment for maternal race/ethnicity. Despite differences in the rates of secondary APS, the rates of hypertensive disorders of pregnancy in SLE patients reported in this nationwide sample approximated those reported by tertiary care specialty clinics (6–10), suggesting that the results of detailed analyses of pregnancy outcomes from these centers may have applicability to the broader range of SLE patients seen nationwide. In addition, we found slightly elevated rates of hypertensive disorders and IUGR in the less well-studied population of pregnant RA patients. Rates of cesarean delivery were significantly increased in all groups compared with the general population; however, indications for surgical delivery were likely to be multifactorial. As is seen among women with pregestational DM, women with common rheumatic diseases appeared to have an increased risk of adverse outcomes of pregnancy and should undergo careful antenatal monitoring in order to minimize any complications.

Ancillary