Trends in Pregnancy Hospitalizations That Included a Stroke in the United States From 1994 to 2007
Reasons for Concern?
Background and Purpose—Stroke is an important contributor to maternal morbidity and mortality, but there are no recent data on trends in pregnancy-related hospitalizations that have involved a stroke. This report describes stroke hospitalizations for women in the antenatal, delivery, and postpartum periods from 1994 to 1995 to 2006 to 2007 and analyzes the changes in these hospitalizations over time.
Methods—Hospital discharge data were obtained from the Nationwide Inpatient Sample, developed as part of the Healthcare Cost and Utilization Project sponsored by the Agency for Healthcare Research and Quality. Pregnancy-related hospitalizations with stroke were identified according to the International Classification of Diseases, Ninth Revision. All statistical analyses accounted for the complex sampling design of the data source.
Results—Between 1994 to 1995 and 2006 to 2007, the rate of any stroke (subarachnoid hemorrhage, intracerebral hemorrhage, ischemic stroke, transient ischemic attack, cerebral venous thrombosis, or unspecified) among antenatal hospitalizations increased by 47% (from 0.15 to 0.22 per 1000 deliveries) and among postpartum hospitalizations by 83% (from 0.12 to 0.22 per 1000 deliveries) while remaining unchanged at 0.27 for delivery hospitalizations. In 2006 to 2007, ≈32% and 53% of antenatal and postpartum hospitalizations with stroke, respectively, had concurrent hypertensive disorders or heart disease. Changes in the prevalence of these 2 conditions from 1994 to 1995 to 2006 to 2007 explained almost all of the increase in postpartum hospitalizations with stroke during the same period.
Conclusions—Our results have demonstrated an increasing trend in the rate of pregnancy-related hospitalizations with stroke in the United States, especially during the postpartum period, from 1994 to 1995 to 2006 to 2007.
Despite the encouraging decline in overall stroke incidence and mortality among adults in the United States and United Kingdom during the last decade, recent concern was raised that trends in young adults may be less favorable.1,2 Trends in stroke during pregnancy and the postpartum period are a subject of special interest, since the World Health Organization proposed to consider stroke as a life-threatening (or “near-miss”) obstetric complication3 and as a condition that may make unintended pregnancy an unacceptable health risk.4
The growing proportion of pregnant women with risk factors for stroke, such as heart disease,5 hypertensive disorders,6 diabetes,7 and postpartum hemorrhage,8 has been demonstrated by several population-based studies in the United States. Furthermore, rising rates of obesity9 and of multiple births and increasing age at birth10 have been cited as factors that have adversely affected the health status of pregnant women in high-income countries, including the United States. In addition, improvements in the management of certain chronic conditions (for example, congenital heart disease and autoimmune disorders) have resulted in growing numbers of women who are entering pregnancy with increased risks for cardiovascular complications.11,12 Thus, there would be grounds to expect that the rate in the United States of hospitalization with pregnancy-related stroke has been growing recently.
The primary objectives of the present report are to describe the prevalence and trends for hospitalizations with pregnancy-related stroke and the contribution to these trends of changes in the prevalence of heart disease, hypertensive disorders, and postpartum hemorrhage by using the 1994 to 2007 Nationwide Inpatient Sample (NIS) of the Healthcare Cost and Utilization Project.
The NIS is the largest nationwide all-payer hospital inpatient care database available in the United States. It is 1 of a family of databases and software tools developed as part of the Healthcare Cost and Utilization Project; the NIS is sponsored by the Agency for Healthcare Research and Quality in partnership with state-level data-collection organizations to provide national estimates for inpatient care.13 The NIS is a stratified sample of ≈20% of all US community hospitals, with hospitals selected according to 5 characteristics: rural/urban location, number of beds, region, teaching status, and ownership. The database includes all discharges from selected hospitals and provides information on 5 to 8 million discharges from an average of 1000 hospitals each year.13 Because the data are publicly available and do not contain direct personal identifiers, this study was exempt from review by the institutional review board of the Centers for Disease Control and Prevention.
Identification of Pregnancy-Related Hospitalizations With Stroke
Our analysis included all pregnancy-related hospitalizations in 1994 to 2007, which were identified according to the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) diagnostic codes 630 to 677, V22, V23, V24, V28, and 792.3; ICD-9-CM procedure codes 72 to 75; and diagnosis-related group codes 370 to 384. These hospitalizations were further classified hierarchically into delivery, postpartum, and antenatal hospitalizations as previously described in detail elsewhere.14,15 Delivery hospitalizations were identified by an enhanced method that incorporates delivery-related ICD-9-CM and diagnosis-related group codes and that has been shown to improve the accuracy of identifying deliveries, especially those with severe complications.14 In addition, ICD-9-CM codes were used to identify 6 types of stroke hospitalizations: subarachnoid hemorrhage (430), intracerebral hemorrhage (431), ischemic stroke (433.01, 433.10, 433.11, 433.21, 433.31, 433.81, 433.91, 434.00, 434.01, 434.11, 434.91, and 436), transient ischemic attack (435, 435.1, 435.2, 435.3, 435.8, and 435.9), cerebral venous thrombosis (CVT; 671.5), and unspecified stroke (674.0). The following comorbidities were also identified: chronic heart disease, hypertensive disorders (ICD-9-CM codes 642, including preeclampsia and eclampsia but excluding 642.3x, transient [gestational] hypertension, 401 to 405, and 437.2), diabetes (ICD-9-CM codes 648.0 and 250), postpartum hemorrhage (ICD-9-CM code 666), and postpartum infection (ICD-9-CM codes 670 and 672). The list of chronic heart disease and ICD-9-CM codes used to identify them are described in detail elsewhere.5
All statistical analyses were conducted by using SAS 9.2-callable SUDAAN (Research Triangle Institute, Research Triangle Park, NC) to account for the complex sampling design of the NIS. Stroke hospitalizations were described by demographic and hospital characteristics. Changes in the prevalence of demographics, hospital characteristics, chronic conditions, complications in the postpartum period, and in-hospital mortality in 2006 to 2007 compared with 1994 to 1995 among pregnancy hospitalizations with stroke were compared by using χ2 tests.
Rates per 1000 deliveries were calculated for each type of stroke (subarachnoid hemorrhage, intracerebral hemorrhage, ischemic, transient ischemic attack, CVT, and unspecified) by hospitalization type (antenatal, delivery, and postpartum). Although we calculated rates of stroke and tested linear trends across all 7 intervals, 1994 to 1995, 1996 to 1997, 1998 to 1999, 2000 to 2001, 2002 to 2003, 2004 to 2005, and 2006 to 2007, we present the rates and their differences for 1994 to 1995 and 2006 to 2007 intervals only.
Multivariable logistic-regression models included all 7 study periods and were run separately among antenatal, delivery, and postpartum hospitalizations. The odds ratios showing the likelihood of stroke for 1996 to 1997, 1998 to 1999, 2000 to 2001, 2002 to 2003, 2004 to 2005, and 2006 to 2007 compared with 1994 to 1995 were simultaneously adjusted for age, hospital region (northwest, midwest, south, west), insurance status, heart disease, hypertensive disorders, postpartum hemorrhage (for delivery hospitalizations only), and postpartum infection (for delivery hospitalizations only).
For all 3 kinds of pregnancy hospitalization status (antenatal, delivery, and postpartum), the 25- to 34-year age group accounted for more hospitalizations than either the youngest (<25) or oldest (35 or older) age group in both time frames (Table 1). Hypertensive disorders were common, with an estimated 11.3% (antenatal), 23.4% (delivery), and 27.8% (postpartum) of hospitalizations with stroke having an indication of such a disorder in 1994 to 1995. In 2006 to 2007, the corresponding rates were higher: 17.0%, 28.5%, and 40.9%, but only the increase among postpartum hospitalizations (P=0.003) was statistically significant. In 1994 to 1995, 15.7% (antenatal), 7.6% (delivery), and 9.1% (postpartum) and in 2006 to 2007, 15.3% (antenatal), 7.7% (delivery), and 12.0% (postpartum) of hospitalizations were complicated by heart disease. No significant differences were seen in the prevalence of heart disease between 1994 to 1995 and 2006 to 2007 among antenatal (P=0.90), delivery (P=0.94), or postpartum (P=0.31) hospitalizations with stroke. In 2006 to 2007, stroke was classified as hemorrhagic (including subarachnoid hemorrhage and intracerebral hemorrhage) in 15.7% of the antenatal, 10.8% of the delivery, and 35.6% of the postpartum stroke hospitalizations, respectively, while CVT accounted for 30.6% and 43.1% of antenatal and delivery hospitalizations with stroke, respectively. Finally, 16.0% (antenatal), 30.9% (delivery), and 18.6% (postpartum) of hospitalizations were classified as unspecified.
Between 1994 to 1995 and 2006 to 2007, the rate (per 1000 deliveries) of any stroke among pregnancy hospitalizations increased, from 0.15 to 0.22 (P<0.001) for antenatal hospitalizations and from 0.12 to 0.22 (P<0.001) among postpartum hospitalizations while remaining unchanged at 0.27 for delivery hospitalizations (Table 2). In 2006 to 2007, pregnancy-related stroke hospitalizations in the NIS totaled 6293 (weighted N), up from 4085 in 1994 to 1995. Among delivery hospitalizations, rates per 1000 deliveries did not increase significantly between the 2 time frames for any type of stroke, but significant increases were found for antenatal hospitalizations with transient ischemic attack, CVT, and unspecified stroke and for postpartum hospitalizations with hemorrhagic stroke, ischemic stroke, and CVT. Of interest, in the antenatal group, the rate of CVT almost doubled between 1994 to 1995 and 2006 to 2007 (P<0.001), and in the postpartum group, the rate of hemorrhagic stroke quadrupled (P<0.001).
In the adjusted analysis, the likelihood of any stroke was higher among patients with heart disease and in those with hypertensive disorders, and the strength of these associations varied by type of hospitalization (Table 3). For example, compared with antenatal, delivery, and postpartum hospitalizations without hypertensive disorders (the referent), hospitalizations with hypertensive disorders were 1.8, 5.6, and 3.5 times more likely to have indication of stroke, respectively. In addition, compared with hospitalizations without heart disease, antenatal and delivery hospitalizations that were complicated by heart disease were 9.4 and 5.4 times as likely, respectively, to be complicated by stroke. In logistic-regression models, adjustment for heart disease and hypertensive disorders in the postpartum period explained almost all increases in the estimated risk of stroke in 2006 to 2007 compared with 1994 to 1995 (unadjusted odds ratio=1.41; 95% CI, 1.15 to 1.73; and adjusted odds ratio=1.11; 95% CI, 0.90 to 1.36). Adjustment for other factors among antenatal and delivery hospitalizations had little effect on the observed changes in stroke in 2006 to 2007 compared with 1994 to 1995. We also ran models excluding cardiovascular disease from the outcome “any stroke.” Because the direction and magnitude of associations were the same, the results are presented with cardiovascular disease included.
Using a large, nationwide hospital discharge dataset, we have demonstrated that the number of pregnancy-related stroke hospitalizations grew by 54%, increasing from ≈4000 in 1994 to 1995 to ≈6000 hospitalizations in 2006 to 2007. We also found that pregnancy-related stroke hospitalizations increased by 47% and 83% of deliveries in antenatal and postpartum hospitalizations, respectively. At the end of the study period (2006 to 2007), the overall prevalence of pregnancy-related stroke hospitalizations was 0.71 per 1000 delivery hospitalizations. The point prevalence of hospitalizations with pregnancy-related stroke in the United States has been previously reported.16,17 However, variation in the methodology for case ascertainment makes comparison among studies and the use of estimates from previous studies to monitor trends unreliable.
In our study, roughly 2 of 3 of the pregnancy-related hospitalizations that were complicated by stroke were diagnosed during delivery or postpartum hospitalizations in 2006 to 2007. By contrast, a study with data obtained from the same nationwide database for 2000 and 2001 found that up to 90% of pregnancy-related strokes were identified during delivery and postpartum hospitalizations and thus, not in antenatal inpatient stays.17 The differences in results may reflect real changes in the characteristics of pregnancy hospitalizations with stroke over time, or they could be due to differences in the methodology used to identify delivery hospitalizations. We note that delivery hospitalizations in our study were identified by an enhanced method that was validated, thus reducing the likelihood of misclassification of antenatal hospitalizations as delivery hospitalizations. Our findings that hypertensive disorders are a leading cause of stroke in pregnancy are consistent with results from single institutions,1819,20 but the relation between heart disease and stroke in pregnancy has not been so well documented. In the second of our 2 time frames (2006 to 2007) of special interest, heart disease was common among pregnancy-related stroke hospitalizations, and it was most common among antenatal hospitalizations (15%). In addition, antenatal and delivery hospitalizations complicated by heart disease were much more likely to be complicated by stroke than were hospitalizations without heart disease (here, the highest odds ratio for stroke was 9.4 for antenatal hospitalizations).
The most current recommendations for clinical management of pregnant women at risk for stroke or with stroke are based on small and/or observational studies, with the benefits and risks remaining uncertain.21 To decrease the risk of stroke, current guidelines recommend the use of anticoagulation therapy for pregnant women with a prior history of noncardioembolic stroke and women who are at increased risk for cardioembolic stroke.22 Because the use of vitamin K antagonists may result in embryopathy, birth defects of the central nervous system, fetal bleeding, and increased rates of fetal death, in the absence of high risk for thromboembolic conditions, the use of unfractionated heparin or low-molecular-weight heparin until week 13 followed by low-dose aspirin (50 to 150 mg/d) for the remainder of the pregnancy is the current recommendation for anticoagulation therapy.22 For pregnant women who are at high risk of thromboembolic events (for example, existence of a known thrombophilia or prosthetic cardiac valve), unfractionated heparin or low-molecular-weight heparin may be used until week 13, followed by vitamin K antagonists until the middle of the third trimester, followed by unfractionated heparin or low-molecular-weight heparin for the remainder of the pregnancy.22
Currently available studies also indicate that pharmacologic treatment with heparin is widely accepted today and may be effective in reducing adverse long-term outcomes among patients with CVT.23,24 The benefits of timely (within 3 hours of the onset of symptoms) administration of intravenous thrombolytic therapy (recombinant tissue plasminogen activator or tissue plasminogen activator) for patients with acute ischemic stroke have been demonstrated in several clinical trials. However, pregnancy is considered 1 of the conditions that might lead to increased risk of bleeding or unfavorable outcomes.25 In a recent review of studies on stroke among pregnant women, only 11 patients treated with thrombolysis for acute ischemic stroke in pregnancy were identified in the literature.26 Thus, in addition to the issues of the benefits, safety, and efficacy of anticoagulation and thrombolytic therapy in pregnancy, our results indicating a growing number of pregnancy hospitalizations with stroke highlighted other areas that demand attention of the clinical and research communities. They include monitoring the prevalence of pregnant women with indications for anticoagulation therapy; estimating physicians' compliance with and patients' adherence to current evidence-based recommendations for prevention and treatment of stroke and their effects on fetal, neonatal, and maternal outcomes; and identifying new, emerging risk factors for stroke among pregnant women.
The results of the present analysis, while offering important insights into the problem of stroke among pregnant women, should be interpreted in the light of several limitations of the study. First, stroke and other diseases were identified by ICD-9-CM codes, which may be incomplete, as they were not validated by a review of the actual medical record, or they may have been miscoded owing to different hospital practices; it is not known whether these codes were applied on the basis of the presence of clinical symptoms and signs or on the results of diagnostic procedures, such as computed tomography or magnetic resonance imaging, or necropsy.27 Second, our data did not allow us to account for multiple pregnancy hospitalizations for the same woman during the study period, because hospitalizations were the units of analyses, not individual women. However, given that the majority of our results were highly significant (P<0.01) and that the year-to-year sampling of hospitals was random, any adjustment for a clustering effect (multiple hospitalizations per 1 patient) in testing for trends would be unlikely to change our conclusions. Third, the rate of hospitalizations for stroke in the postpartum period may be underestimated, because some hospitalizations occurring after the 6-week time frame that is defined as the “postpartum period” may not be coded as postpartum hospitalizations and thus, may have been misclassified as nonpregnancy hospitalization in our study. Fourth, although our results showed strong associations with stroke for both heart disease and hypertensive disorders, causality cannot be established from our cross-sectional data. Finally, it is possible that some change over time could be related to an increased use of or improvements in diagnostic imaging as well as changes in coding practices. This may be especially true for changes in CVT over time.
Our results have demonstrated an increasing trend in the rate of pregnancy-related hospitalizations with stroke in the United States, especially during the postpartum period, from 1994 to 1995 to 2006 to 2007. This trend can be explained in substantial measure by the increase in hospitalizations of patients with heart disease and hypertensive disorders among postpartum hospitalizations. The surveillance of cardiovascular disease and the evaluation of measures designed to enhance the quality of care that target pregnant women specifically remain essential for informing the design of clinical research and ultimately for improving the clinical care of women in the United States.
The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.
- Received December 4, 2010.
- Accepted April 8, 2011.
- © 2011 American Heart Association, Inc.
- McCarron MO,
- Davey Smith G,
- McCarron P
- Russo CA,
- Andrews RM
- Farr S,
- Folger SG,
- Paulen M,
- Tepper N,
- Whiteman M,
- Zapata L,
- et al
- Albrecht SS,
- Kuklina EV,
- Bansil P,
- Jamieson DJ,
- Whiteman MK,
- Kourtis AP,
- et al
- Heron M,
- Sutton PD,
- Xu J,
- Ventura SJ,
- Strobino DM,
- Guyer B
Healthcare Cost Utilization Project (HCUP). Data use agreement training. http://www.hcup-us.ahrq.gov/techassist/dua.jsp. Accessed December 23, 2008.
- Lanska DJ,
- Kryscio RJ
- Furie KL,
- Kasner SE,
- Adams RJ,
- Albers GW,
- Bush RL,
- Fagan SC,
- et al
- Guenther G,
- Arauz A
Specifications Manual for National Hospital Inpatient Quality Measures Discharges; 10-01-10 (4Q10) through 03-31-11 (1Q11). http://www.qualitynet.org/dcs/ContentServer?c=Page&pagename=QnetPublic%2FPage%2FQnetTier4&cid=1228754600169. Accessed March 4, 2011.
- Del Zotto E,
- Giossi A,
- Volonghi I,
- Costa P,
- Padovani A,
- Pezzini A
MORGAM Project. MORGAM Manual (2001-). http://www.ktl.fi/publications/morgam/manual/contents.htm, URN:NBN:fi-fe20041529. http://www.ktl.fi/publications/morgam/manual/followup/form23.htm. Accessed March 4, 2011.