OBJECTIVES: To estimate the incidence, describe the mortality, and identify independent predictors of peripartum cardiomyopathy, a very serious cardiovascular complication of pregnancy associated with maternal morbidity and mortality among otherwise healthy women without prior heart disease.
METHODS: We identified all cases of diagnosed heart failure that occurred among women within 1 month before to 5 months after delivery of a liveborn neonate in Kaiser Permanente Northern California delivery hospitals between 1995 and 2004. Incident peripartum cardiomyopathy was confirmed from medical records documenting dilated cardiomyopathy with reduced left ventricular systolic function after excluding women with prior heart failure or valvular disease. Data sources included medical records, electronic clinical databases, and state birth and death files.
RESULTS: Among 227,224 eligible women, we confirmed 110 recognized peripartum cardiomyopathy cases (incidence: 4.84 per 10,000 live births, 95% confidence interval 3.98–5.83). Independent predictors included maternal age of 25 years or older, non-Hispanic African American and Filipino groups, parity of 4 or greater, multiple gestation, severe anemia, pre-existing and pregnancy-related hypertensive disorders, and hemolysis, elevated liver enzymes, low platelets syndrome. Maternal death rate (per 1,000 person-years) was higher among cases (6.12) than noncases (0.23; P<.001). Neonates whose mothers developed peripartum cardiomyopathy experienced poorer clinical outcomes.
CONCLUSION: Within a large, diverse northern California population, 1 of every 2,066 women delivering a liveborn neonate had recognized, confirmed peripartum cardiomyopathy, which was associated with higher maternal and neonatal death rates and worse neonatal outcomes. Several readily available patient characteristics can be used to identify women at risk for this severe pregnancy complication.
LEVEL OF EVIDENCE: II
Peripartum cardiomyopathy occurred at a higher rate than previously reported and was independently associated with age, race or ethnicity, severe anemia, maternal hypertension, and other complications.
From the Division of Research and the Department of Obstetrics and Gynecology, Kaiser Permanente of Northern California, Oakland, California; and the Departments of Epidemiology, Biostatistics, and Medicine, University of California, San Francisco, San Francisco, California.
Supported by research grant 115-9928 from the Kaiser Permanente of Northern California Community Benefit Program, Oakland, CA.
Corresponding author: Erica P. Gunderson, PhD, Division of Research, Kaiser Permanente of Northern California, 2000 Broadway Street, Oakland, CA 94612-2304; e-mail: Erica.Gunderson@kp.org.
Financial Disclosure The authors did not report any potential conflicts of interest.
Insights into the epidemiology of peripartum cardiomyopathy, a rare form of heart failure with no identifiable cause during late pregnancy or within a few months postpartum in women without prior heart failure, are limited.1,2 Transient hypertrophy of the left ventricle and a decrease in left ventricular systolic function as a result of increased circulating blood volume occur in the second half of pregnancy with return to normal during the early postpartum period.3 Peripartum cardiomyopathy develops in otherwise healthy individuals and may result in persistent cardiac dysfunction or death. Although the etiology of peripartum cardiomyopathy is unclear, hypothesized pathways may involve oxidative stress, cardiotoxic prolactin subfragments, and inflammatory, autoimmune, or viral processes.4–7
Diagnosing peripartum cardiomyopathy on clinical evaluation alone is challenging because many symptoms of normal pregnancy during the last month (ie, dyspnea, fatigue, lower extremity edema) can overlap with heart failure-related symptomatology. Milder peripartum cardiomyopathy may go unrecognized in the absence of echocardiographic screening. Peripartum cardiomyopathy incidence from case series reports varies widely with the highest rates in Haiti (33 cases per 10,000 live births) and South Africa (one case per 1,000 deliveries).8,9 Estimates from two US population-based studies10,11 range from 2.5 to 3.1 cases per 10,000 live births. Of these, only one10 excluded pre-existing disease conditions and confirmed case status by medical record review according to the 1997 National Heart, Lung and Blood Institute criteria, but neither assessed independent predictors of peripartum cardiomyopathy. Overall incidence rates from these studies may be underestimated by inclusion of repeat pregnancies during the study period. Maternal mortality among US women with peripartum cardiomyopathy ranges from 0% to 3.3%10–12 or slightly higher.13 Also, little is known about neonatal morbidity and infant mortality among children born to women with peripartum cardiomyopathy.
To address these gaps, we identified a large diverse cohort of pregnant women receiving care from an integrated healthcare system without known heart disease associated with cardiac failure to estimate peripartum cardiomyopathy incidence, overall and within multiple racial and ethnic groups, to identify independent predictors, and describe associated maternal mortality and neonatal morbidity and mortality.
Kaiser Permanente of Northern California is a large, integrated healthcare delivery system that provides care for 3.2 million persons or more in northern California. Kaiser Permanente of Northern California members are demographically similar to the northern California population, except for slightly lower representation of the extremes of age and income.14 Pregnant women represent a higher proportion of the economically disadvantaged (approximately 10%) within Kaiser Permanente of Northern California because they are eligible for prenatal health care services under the State of California MediCal Program.
We analyzed all live births that occurred at any Kaiser Permanente of Northern California hospital between January 1, 1995, and December 31, 2004 (n=326,381). We excluded mother–newborn pairs who had inaccurate medical record numbers (n=545), no accessible delivery records (n=1,719), missing data on maternal age (n=35) or length of gestation (n=297), or for whom we could not find a match in the neonatal birth certificate files (n=4,824). For women who had more than one delivery during the study period (n=91,349), we examined the first pregnancy identified with peripartum cardiomyopathy for the women (cases) and from the group of women without cardiomyopathy (noncases) randomly selected one pregnancy. To enhance the accuracy of identifying peripartum cardiomyopathy, we excluded women who had a diagnosis of heart failure or valvular heart disease before the selected pregnancy in the study period (n=159), leaving 227,453 women for this analysis. The study was approved by the Kaiser Foundation Research Institute and the State of California institutional review boards. Waiver of informed consent was obtained because of the nature of the study.
Peripartum cardiomyopathy was defined according to National Heart, Lung and Blood Institute criteria: onset of clinical heart failure with no identifiable cause in the last month of pregnancy or within 5 months after delivery in the absence of heart disease before the last month of pregnancy with echocardiographic evidence of left ventricular systolic dysfunction such as reduced fractional shortening or ejection fraction as well as a dilated left ventricle.1
We identified possible cases of recognized peripartum cardiomyopathy by identifying physician-assigned diagnoses for heart failure from hospitalization and ambulatory visit databases for visits occurring from 1 month before through 5 months after delivery. We used the following International Classification of Diseases, Ninth Edition (ICD-9) codes: 398.91 (rheumatic heart disease), 402.01, 402.11, and 402.91 (hypertensive heart disease with heart failure), 425.4, 425.7, 425.8, and 425.9 (other heart failure for peripartum period), 428.0, 428.1, and 428.9 (left-±right-sided heart failure) to locate peripartum cardiomyopathy cases and pre-existing heart failure.15 A trained abstractor reviewed relevant medical records for data elements from the Framingham heart failure criteria16 as well as information from echocardiograms, nuclear scintigraphy, or other cardiac imaging test reports demonstrating evidence of ventricular dilatation and quantitative or semiquantitative measures of reduced left ventricular systolic function (ie, ejection fraction 0.45 or less, fractional shortening less than 0.25, or description of moderately or severely reduced systolic function). One of the physician investigators (A.S.G.) reviewed each case that met the Framingham clinical criteria for heart failure and confirmed medical record documentation of a dilated cardiomyopathy with reduced left ventricular systolic function in the absence of other possible etiologies. Patients who had pre-existing cardiac disease based on chart review or relevant ICD-9 codes (specified previously) found in health plan databases were excluded.
Self-reported maternal age, race, ethnicity, parity, and educational attainment were obtained from California state birth certificates, which had been previously linked to the Kaiser Permanente of Northern California electronic delivery records.17 We used health plan hospitalization, billing claims, and ambulatory visit clinical databases to characterize multiple gestation,18 trimester prenatal care began, gestational diabetes (ICD-9 code 648.8), hypertension status before and during gestation (ICD-9 codes 401–405, 642.0–642.9), and hemolysis, elevated liver enzymes, and low platelet count syndrome (ICD-9 code 642.5x). Health plan laboratory databases were used to obtain hemoglobin or hematocrit levels during the second (14–28 weeks of gestation) and third (29 or more weeks of gestation) trimesters to classify women for anemia (normal, mild, moderate, or severe) using Institute of Medicine criteria.19 We also determined vaginal or cesarean delivery from the State of California birth certificate files.17
We identified all women who died during delivery or within 36 months after delivery using a combination of health plan hospitalization and administrative databases, Social Security Administration vital status files, and California state death certificate files.20
We used electronic delivery records and the validated Kaiser Permanente Neonatal Minimum Data Set21 to identify selected neonatal outcomes including gestational age (32 or less, 33 to 36, and 37 or more weeks), birth weight (less than 1,500, 1,500–2,499, and 2,500 g or more), neonatal size for gestational age (appropriate for gestational, small for gestational [less than the 10th percentile], and large for gestation [greater than the 90th percentile]),22 and Apgar score at 5 minutes (0–6 and 7 or more).
Incidence of peripartum cardiomyopathy with associated 95% confidence intervals (CIs) was calculated per 10,000 live births. Maternal characteristics and neonatal outcomes were compared between women with peripartum cardiomyopathy and women without peripartum cardiomyopathy using t tests for continuous variables and chi-square or Fisher's exact tests for categorical variables. We used multivariable logistic regression to identify independent predictors of peripartum cardiomyopathy and included as candidate variables those characteristics previously reported to be associated with peripartum cardiomyopathy as well as variables that differed between groups in univariable analyses using P<.05. Asian or Pacific Islanders were stratified as Chinese, Filipino, and “other Asians or Pacific Islander” as a result of small sample sizes in other Asian ethnic subgroups. P values for trend were obtained by including continuous or ordinal variables in the model. Model goodness of fit was examined using the Hosmer-Lemeshow test.23 For neonatal characteristics, we used nonlinear mixed models with random effects for each pregnancy to account for clustering within multiple gestations. Finally, we calculated the rate of maternal death within 36 months postdelivery per 1,000 person-years with associated 95% CI.
Among 227,453 eligible women, we identified 357 women with possible recognized peripartum cardiomyopathy. Of these, medical records for four women (delivered at non-Kaiser hospitals) were not retrievable, leaving medical records of 353 women for review. After examining relevant medical information and diagnostic tests, peripartum cardiomyopathy diagnosis was confirmed for 110 women. Of the remaining 243 women, evidence of clinical heart failure based on Framingham criteria was present for 84 women, but detailed information of the specific level of left ventricular systolic function was unavailable, so they were considered “probable” peripartum cardiomyopathy cases and excluded from this analysis. Therefore, the final analytic sample (n=227,224) included 110 confirmed cases of peripartum cardiomyopathy and 227,114 noncases and their newborns (n=229,040).
The incidence for confirmed cases (n=110) combined with probable cases (n=84) was 8.53 per 10,000 (95% CI 7.38–9.82). The observed incidence among confirmed recognized peripartum cardiomyopathy only was 4.84 per 10,000 live births (95% CI 3.98–5.83). The incidence of confirmed peripartum cardiomyopathy was highest among women 40 years of age or older (Fig. 1). The incidence varied by race or ethnicity with the highest rates observed for non-Hispanic African American and Filipino women and the lowest for Chinese and Hispanic women (Fig. 2). There was no significant temporal trend in the incidence rate during the study period (Fig. 3). Among the 110 confirmed cases, 70 (63.6%) were diagnosed between 3 days before and 14 days after delivery.
Compared with women who did not experience peripartum cardiomyopathy, women with peripartum cardiomyopathy were more likely to be older, non-Hispanic African American or Filipino, and to have four or more total births. During the index pregnancy, women with peripartum cardiomyopathy were more likely to experience anemia, multiple gestation, gestational diabetes mellitus, pre-existing and pregnancy-related hypertension, hemolysis, elevated liver enzymes, low platelets syndrome, and a cesarean delivery (Table 1). In multivariable analyses, independent predictors of peripartum cardiomyopathy included maternal age 25 years or older, non-Hispanic African American and Filipino race, Hispanic ethnicity, parity of four or more, severe anemia, multiple gestation, pre-existing and pregnancy-related hypertension, pre-eclampsia or eclampsia, and hemolysis, elevated liver enzymes, low platelets syndrome (Table 2). Peripartum cardiomyopathy risk increased significantly with increasing maternal age, increasing parity, and severity of anemia and hypertensive disorders. The model fit using the Hosmer-Lemeshow test was adequate (P=.08).
During 681,387 person-years of follow-up (up to 36 months postdelivery), we confirmed two deaths among 110 women with peripartum cardiomyopathy (6.12 per 1,000 person-years, 95% CI 0.74–22.10) compared with 154 deaths among 227,114 women without peripartum cardiomyopathy (0.23 per 1,000 person-years, 95% CI 0.19–0.26), representing a substantially higher relative risk (unadjusted relative risk 26.81, 95% CI 7.32–95.12). Of note, only one of the two deaths among cases was attributed to heart failure; the other death was attributed to metastatic breast cancer.
Compared with neonates born to women without peripartum cardiomyopathy, the 122 neonates born to women with peripartum cardiomyopathy were more likely to be premature, low or very low birth weight, small for gestational age, and have lower 5-minute Apgar scores (Table 3). We observed one death (younger than 60 days of age) among the 122 infants of women with peripartum cardiomyopathy (82.0 per 10,000 births, 95% CI 2.1–460.5) and 387 deaths among the infants of women without peripartum cardiomyopathy (17.0 per 10,000 births, 95% CI 15.3–18.7; P=.08).
Within a large, multiracial and ethnic pregnancy cohort in northern California, approximately one of every 2,066 pregnant women who delivered a liveborn neonate developed confirmed peripartum cardiomyopathy as defined by National Heart, Lung and Blood Institute criteria. Incidence rates were considerably higher for non-Hispanic African American women (one in 664) and Filipino women (one in 978) and lower for Hispanic women (one in 6,729) compared with non-Hispanic white women (one in 2,450). Although independent risk factors for peripartum cardiomyopathy included maternal age 25 years or older, multiparity, and severe anemia during pregnancy, the strongest independent predictors were maternal age 40 years or older, hypertension before pregnancy, multiple gestation, or pregnancy-related hypertensive disorders, especially hemolysis, elevated liver enzymes, low platelets syndrome. Lowest risk of peripartum cardiomyopathy was associated with maternal age younger than 24 years, Hispanic ethnicity, and singleton gestations uncomplicated by hypertensive disorders or anemia. Cesarean delivery was more common among women who developed peripartum cardiomyopathy. Women with peripartum cardiomyopathy had a higher absolute risk of maternal death, and their newborns had worse neonatal morbidity, including prematurity, low birth weight, small size for gestational age, and lower Apgar scores at 5 minutes.
Our estimate of peripartum cardiomyopathy incidence is one case per 2,066 live births, which is 55–95% higher than estimates from population-based studies.10,11 Differences in peripartum cardiomyopathy incidence across studies could be explained by differences in racial and ethnic composition of the study populations. The southern California sample was comprised predominantly of Hispanics (49%),10 whereas minority women represented only 21% of the National Hospital Discharge data set.11 Other explanations include variation in case ascertainment methods and our selection of only one pregnancy per woman. A previous study that relied primarily on hospital discharge codes to identify heart failure or peripartum cardiomyopathy did not exclude pre-existing conditions, and these factors would result in overestimates of the peripartum cardiomyopathy cases,11 yet our estimates were higher although we excluded pre-existing diagnosed heart failure and confirmed clinical heart failure with dilated cardiomyopathy and reduced left ventricular systolic function from medical records. We included only one pregnancy per woman to eliminate bias resulting from differential birth rates across racial and ethnic groups (eg, lower incidence of peripartum cardiomyopathy resulting from higher parity, ie, Hispanics) and to assess the relative importance of the “independent” predictors of peripartum cardiomyopathy. Our incidence rates decreased as expected when we included repeat pregnancies (one case per 2,895 deliveries overall; one per 3,529 among non-Hispanic whites, one per 896 among non-Hispanic African Americans, one per 9,172 among Hispanics) but still remained higher than studies including repeat pregnancies.10,11 Furthermore, despite using the accepted National Heart, Lung and Blood Institute case definition (eg, exclusion of women with pre-existing heart conditions and lower left ventricular ejection fraction cutoff of 45% or less in our study compared with 50% or less by others), our overall incidence rate was still higher than previous estimates based on the same National Heart, Lung and Blood Institute definition or less stringent criteria.10,11
Our study identified several risk factors for peripartum cardiomyopathy that are consistent with prior studies.24,25 Moreover, we applied analytic methods to identify independent risk factors and evaluate their relative importance to incident peripartum cardiomyopathy. Prior population-based studies described crude associations with peripartum cardiomyopathy among a few risk factors such as age and hypertension in pregnancy10 or compared crude incident rates among racial and ethnic groups.11 These studies did not control for potential confounding by age, repeated pregnancies, and parity differences among racial and ethnic groups (ie, higher parity in Hispanics that underestimate incidence) or evaluate strength of the associations.10,11 Case series or case–control studies also identified advanced maternal age, multiple gestation, pre-eclampsia and gestational hypertension, and African American race as risk factors for peripartum cardiomyopathy.26,27 A case–control study reported a higher risk of peripartum cardiomyopathy among African American women independent of sociodemographic and clinical risk factors such as any hypertension.27 Our study used multivariable statistical modeling within a large, diverse population-based sample to identify independent risk factors for peripartum cardiomyopathy and confirmed that African American race, older maternal age, multiple gestation, and various hypertensive disorders of pregnancy are important independent predictors of peripartum cardiomyopathy. Our study expands on the existing knowledge base by identifying Filipino race and severe anemia as important predictors of peripartum cardiomyopathy. We also report a progressive increase in risk of peripartum cardiomyopathy with severity of hypertensive disorders of pregnancy independent of other risk factors. This association may support the hypothesis that development of hypertensive disease of pregnancy28 and peripartum cardiomyopathy share common pathways (ie, antiangiogenic prolactin subfragments).7
The 3-year postdelivery mortality rate of 1.8% we observed among women with peripartum cardiomyopathy was slightly lower but similar to 2.1% (95% CI 0.3–10.8%)11 and 3.3% (95% CI 0.6–12.5%)10 from prior studies. The US Centers for Disease Control and Prevention reported that African American women are 6.4 times more likely than white women to die from peripartum cardiomyopathy,29 and one study reported high mortality (15.9%) in indigent African American women.30 In our study, two deaths among cases included one African American woman and one woman of another race.
The neonatal morbidity and mortality for children born to women with peripartum cardiomyopathy were not reported in previous population-based cohorts.10,11 A study of 14 women reported that 21% of peripartum cardiomyopathy deliveries were premature.31 Neonates in our cohort had significantly greater risk of adverse outcomes and borderline significant higher mortality. Given the much higher rates of pregnancy-related and other hypertensive complications as well as severe anemia among peripartum cardiomyopathy cases, the associations with multiple gestation, fetal growth restriction, and newborn prematurity were expected.
Our study strengths include a large, well-characterized source population with age and racial or ethnic diversity, electronic clinical databases within a large, integrated healthcare system, and linkage to sociodemographic data from birth certificates. Our cohort had proportions of Hispanic (27%) and non-Hispanic white women (41%), which were similar to US women giving birth in 2008 (25% Hispanic, and 54% non-Hispanic white women).32 Ninety-two percent overall retained membership in the Kaiser Permanente of Northern California healthcare system at 5 months postpartum (96.4% of cases, 92.2% of noncases; P=.10), supporting the accuracy of our estimates of incidence rates. We used an effective search strategy to identify all potential cases of peripartum cardiomyopathy from inpatient and ambulatory settings during the study period and confirmed cases of clinically recognized peripartum cardiomyopathy by medical records and selected one pregnancy per woman to compare incidence rates across race or ethnicity groups. We were also able to ascertain history of medical conditions to exclude those pregnant women with pre-existing heart failure or valvular disease. We used multivariable analytic methods in a large population-based and racially and ethnically diverse sample to identify the independent maternal characteristics related to peripartum cardiomyopathy and to examine both maternal and neonatal outcomes.
Our study also had some limitations. As a result of the retrospective cohort design, systematic screening of all pregnant women was not feasible. However, given that peripartum cardiomyopathy typically presents with symptomatic heart failure, it is unlikely that we missed a large number of true, clinically meaningful cases. Although we were unable to review medical records for four women identified through electronic databases, our estimates were still higher than prior studies.10,11 Hypertensive disorders were ascertained using ICD-9 codes, although any bias (ie, misclassification) would likely be nondifferential at discharge for the cases and noncases of peripartum cardiomyopathy.
Within a large, northern California population of women delivering liveborn neonates from 1995 to 2004, we found that peripartum cardiomyopathy affected approximately one in every 2,066 deliveries, a much higher incidence rate than previously reported in epidemiologic studies. Peripartum cardiomyopathy risk was independently associated with several maternal demographic features and selected pregnancy-related clinical complications. Women with peripartum cardiomyopathy were more likely to die within 3 years after delivery, and their infants also had poorer health outcomes. If confirmed in other studies, systematic assessment of these independent risk factors during pregnancy may assist providers with earlier identification of women at high risk for peripartum cardiomyopathy and the development of effective prevention and early treatment strategies.33
1. Pearson GD, Veille JC, Rahimtoola S, Hsia J, Oakley CM, Hosenpud JD, et al.. Peripartum cardiomyopathy: National Heart, Lung, and Blood Institute and Office of Rare Diseases (National Institutes of Health) workshop recommendations and review. JAMA 2000;283:1183–8.
2. Sliwa K, Hilfiker-Kleiner D, Petrie MC, Mebazaa A, Pieske B, Buchmann E, et al.. Current state of knowledge on aetiology, diagnosis, management, and therapy of peripartum cardiomyopathy: a position statement from the Heart Failure Association of the European Society of Cardiology Working Group on peripartum cardiomyopathy. Eur J Heart Fail 2010;12:767–78.
3. Geva T, Mauer MB, Striker L, Kirshon B, Pivarnik JM. Effects of physiologic load of pregnancy on left ventricular contractility and remodeling. Am Heart J 1997;133:53–9.
4. Gleicher N, Elkayam U. Peripartum cardiomyopathy, an autoimmune manifestation of allograft rejection? Autoimmun Rev 2009;8:384–7.
5. Sliwa K, Forster O, Libhaber E, Fett JD, Sundstrom JB, Hilfiker-Kleiner D, et al.. Peripartum cardiomyopathy: inflammatory markers as predictors of outcome in 100 prospectively studied patients. Eur Heart J 2006;27:441–6.
6. Bultmann BD, Klingel K, Nabauer M, Wallwiener D, Kandolf R. High prevalence of viral genomes and inflammation in peripartum cardiomyopathy. Am J Obstet Gynecol 2005;193:363–5.
7. Hilfiker-Kleiner D, Sliwa K, Drexler H. Peripartum cardiomyopathy: recent insights in its pathophysiology. Trends Cardiovasc Med 2008;18:173–9.
8. Fett JD, Christie LG, Carraway RD, Murphy JG. Five-year prospective study of the incidence and prognosis of peripartum cardiomyopathy at a single institution. Mayo Clin Proc 2005;80:1602–6.
9. Desai D, Moodley J, Naidoo D. Peripartum cardiomyopathy: experiences at King Edward VIII Hospital, Durban, South Africa and a review of the literature. Trop Doct 1995;25:118–23.
10. Brar SS, Khan SS, Sandhu GK, Jorgensen MB, Parikh N, Hsu JW, et al.. Incidence, mortality, and racial differences in peripartum cardiomyopathy. Am J Cardiol 2007;100:302–4.
11. Mielniczuk LM, Williams K, Davis DR, Tang AS, Lemery R, Green MS, et al.. Frequency of peripartum cardiomyopathy. Am J Cardiol 2006;97:1765–8.
12. Amos AM, Jaber WA, Russell SD. Improved outcomes in peripartum cardiomyopathy with contemporary. Am Heart J 2006;152:509–13.
13. Felker GM, Thompson RE, Hare JM, Hruban RH, Clemetson DE, Howard DL, et al.. Underlying causes and long-term survival in patients with initially unexplained cardiomyopathy. N Engl J Med 2000;342:1077–84.
14. Krieger N. Overcoming the absence of socioeconomic data in medical records: validation and application of a census-based methodology. Am J Public Health 1992;82:703–10.
15. Go AS, Hylek EM, Phillips KA, Chang Y, Henault LE, Selby JV, et al.. Prevalence of diagnosed atrial fibrillation in adults: national implications for rhythm management and stroke prevention: the AnTicoagulation and Risk Factors in Atrial Fibrillation (ATRIA) Study. JAMA 2001;285:2370–5.
16. McKee PA, Castelli WP, McNamara PM, Kannel WB. The natural history of congestive heart failure: the Framingham study. N Engl J Med 1971;285:1441–6.
17. Croen LA, Gunderson EP, Escobar GJ, Yoshida CK, Jones T. Use of linked data bases to assess selected quality indicators for maternal health. Final report to the Agency for Healthcare Research and Quality. 2002. Report No. IDSRN Task Order No. 3, Contract No. 290-00-0015.
18. Williams obstetrics. 22nd ed. New York (NY): McGraw-Hill Medical Publishing Division; 2005.
19. Institute of Medicine. Nutrition during pregnancy. Part I. Weight gain. Washington (DC): National Academy of Sciences; 1990.
20. Arellano MG, Petersen GR, Petitti DB, Smith RE. The California Automated Mortality Linkage System (CAMLIS). Am J Public Health 1984;74:1324–30.
21. Escobar GJ, Fischer A, Kremers R, Usatin MS, Macedo AM, Gardner MN. Rapid retrieval of neonatal outcomes data: the Kaiser Permanente Neonatal Minimum Data Set. Qual Manag Health Care 1997;5:19–33.
22. Brenner WE, Edelman DA, Hendricks CH. A standard of fetal growth for the United States of America. Am J Obstet Gynecol 1976;126:555–64.
23. Hosmer DW, Hjort NL. Goodness-of-fit processes for logistic regression: simulation results. Stat Med 2002;21:2723–38.
24. Abboud J, Murad Y, Chen-Scarabelli C, Saravolatz L, Scarabelli TM. Peripartum cardiomyopathy: a comprehensive review. Int J Cardiol 2007;118:295–303.
25. Goland S, Modi K, Bitar F, Janmohamed M, Mirocha JM, Czer LS, et al.. Clinical profile and predictors of complications in peripartum cardiomyopathy. J Card Fail 2009;15:645–50.
26. Demakis JG, Rahimtoola SH. Peripartum cardiomyopathy. Circulation 1971;44:964–8.
27. Gentry MB, Dias JK, Luis A, Patel R, Thornton J, Reed GL. African-American women have a higher risk for developing peripartum cardiomyopathy. J Am Coll Cardiol 2010;55:654–9.
28. Leanos-Miranda A, Marquez-Acosta J, Cardenas-Mondragon GM, Chinolla-Arellano ZL, Rivera-Leanos R, Bermejo-Huerta S, et al.. Urinary prolactin as a reliable marker for preeclampsia, its severity, and the occurrence of adverse pregnancy outcomes. J Clin Endocrinol Metab 2008;93:2492–9.
29. Whitehead SJ, Berg CJ, Chang J. Pregnancy-related mortality due to cardiomyopathy: United States, 1991–1997. Obstet Gynecol 2003;102:1326–31.
30. Modi KA, Illum S, Jariatul K, Caldito G, Reddy PC. Poor outcome of indigent patients with peripartum cardiomyopathy in the United States. Am J Obstet Gynecol 2009;201:171.e1–5.
31. O'Connell JB, Costanzo-Nordin MR, Subramanian R, Robinson JA, Wallis DE, Scanlon PJ, et al.. Peripartum cardiomyopathy: clinical, hemodynamic, histologic and prognostic characteristics. J Am Coll Cardiol 1986;8:52–6.
© 2011 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.
33. Elkayam U, Goland S. Bromocriptine for the treatment of peripartum cardiomyopathy. Circulation 2010;121:1463–4.