Ladner, Heidi E. MD*; Danielsen, Beate PhD†; Gilbert, William M. MD*
Acute myocardial infarction (MI) during pregnancy and the puerperium is a rare but often catastrophic event, associated with significant maternal mortality.1,2 In prior meta-analysis of case reports, maternal mortality has been estimated to be from 19%1 to as high as 37%.2 In addition, it seems that maternal mortality is higher if it occurs in the third trimester or postpartum period.2–4 Neonatal mortality, on the other hand, has been reported to be less than that of the mother but still quite elevated (13% to 17%) when compared with the population at large.1,3
The incidence of acute MI during pregnancy has been determined to occur with a frequency of 1 in 10,000 to 1 in 30,000 pregnancies.1,3 A true population-based frequency has been hard to determine because of the rarity of the diagnosis and the long time periods required to collect sufficient cases to report.
Maternal morbidity and other pregnancy outcomes of acute MI during pregnancy have been previously reported to be increased, but not often clearly defined. Several reports have examined the associations of cardiac risk factors and women suffering an acute MI during pregnancy. In one study, cardiac risk factors were found in 58% of pregnant women with acute MI, suggesting that these high-risk patients may be identifiable.1 Of those risk factors reported, hypertension and smoking were most prevalent and found in 25% and 20% of cases respectively.1 Diabetes mellitus (not identified as gestational or pregestational) was identified in 3–5% of these patients. Hypercholesterolemia, hyperlipidemia, and obesity were all found in less than 5% of cases reported, and intuition would suggest that obesity alone should be more common than that reported in the normal population.
Previously published data on acute MI in pregnancy have largely been obtained by literature search and case report review, with cases dating as far back as 1922.1–3 Diagnosis and treatment modalities for acute MI have improved since that time, making the older data of questionable applicability to current patients. In this study, we examined the pregnancy outcomes in a population of deliveries in California with the diagnosis of acute MI during the 10-year period from 1991–2000.
MATERIALS AND METHODS
Permission was obtained from the State of California Human Subjects Protection Committee, the University of California, Davis Institutional Review Board and the California Office of Statewide Health Planning and Development for this study. We used a database that links maternal and neonatal hospital-discharge records to birth certificates for all civilian hospitals that report to the California Office of Statewide Health Planning and Development. The database represents 98% of all births occurring in California; of these births, more than 98% of birth records were successfully matched to a maternal and neonatal discharge record.5
Women in the database were identified who sustained an acute MI during pregnancy, labor, and delivery, or the 6-week postpartum period and who delivered between January 1, 1991, and December 31, 2000. The remaining non-MI pregnant population was subsequently used as a comparison group for analysis. All linked records of both acute MI and non-MI cohorts were searched for International Classification of Diseases, 9th Revision, Clinical Modification codes. The pregnancy complications searched for included preeclampsia, eclampsia, and gestational diabetes as well as identifiable cardiac risk factors, including coronary atherosclerosis, essential hypertension, pregestational diabetes, and disorders of lipid metabolism. In addition, linked records were searched for pregnancy outcomes such as mortality, cesarean delivery, prematurity, birth weight, maternal prenatal admission, maternal postpartum readmission, hospital length of stay, and total hospital charges.
The 2 groups, those with an acute MI and those without, were compared using χ2 tests for categorical variables. In addition, the 3 groups broken down into timing of acute MI were compared as well. A multivariate analysis was performed to identify risk factors for acute MI using logistic regression controlling for race, parity, diabetes, gestational diabetes, chronic hypertension, severe preeclampsia, eclampsia, and maternal age where appropriate.
During the period from January 1, 1991, to December 31, 2000, 151 cases of acute MI were identified in a population of approximately 5.4 million deliveries, giving an incidence of 1 in 35,700 pregnancies. During the 10 years there was an increase in incidence of acute MI from 1 in 73,400 initially, to 1 in 24,600 in the final year (P < .08). Eleven women (7.3%, P < .001) died from their acute MI and all were diagnosed in the antenatal and intrapartum period. The timing of diagnosis of the acute MI demonstrated equal numbers in the antepartum and postpartum periods, with only 21% diagnosed in the intrapartum period (Table 1). In Table 1, the demographics of the population are characterized. Hispanic women had fewer acute MIs compared with the rest of the population, whereas the remaining ethnic groups were not different between groups. Women with an acute MI were more likely to be multiparous than those without (Table 1); however, parity was not associated with timing of acute MI.
The maternal age distribution for those with and without an acute MI is demonstrated in Figure 1. Both groups had a normal distribution, with the acute MI group shifted to the older age by approximately 5 years (P < .05). Maternal age was not associated with timing of acute MI. Six women had multiple gestations (4%), and no demographic or timing of acute MI factors were different as compared with singleton women.
In Table 2, maternal outcomes are separated into timing of diagnosis of acute MI and into those patients who did not have an acute MI. When examining the timing of diagnosis of acute MI, several interesting findings are demonstrated. There were higher rates of maternal and pregnancy complications (prematurity, pregestational diabetes, and mild preeclampsia) associated with the diagnosis of acute MI in the antenatal period as compared with the postpartum period (Tables 2 and 3). Intrapartum diagnosis of acute MI is associated with the highest risk of eclampsia, severe preeclampsia, and maternal mortality and the lowest risk of being diagnosed with coronary artery disease (Table 2). Overall, classic risk factors (coronary artery disease, disorders of lipid metabolism) associated with acute MI were increased in those patients with any acute MI as compared with those without one (Table 2)
Table 3 displays neonatal outcomes. There was 1 neonatal death that occurred in the postpartum acute MI group, and this rate was not different from the nonacute MI group (Table 3). Prematurity and low birth weight (LBW) were increased in antenatal cases as compared with postpartum cases, with the latter (LBW) most likely a reflection of the former (prematurity).
Multivariate analysis was performed using logistic regression to determine which factors were associated with the greatest risk of an acute MI. Advancing maternal age independently increased the risk of acute MI when all ages were compared with those aged 21–25 years: aged 30–35 years, odds ratio (OR) 2.6 (95% confidence interval [CI]1.6–4.5); aged 35–40 years, OR 3.4 (95% CI 1.9–6.0); aged 40 years or older, OR 4.5 (95% CI 2.0–9.8). Two preexisting medical conditions, diabetes and chronic hypertension, were independently associated with an increased risk of acute MI: diabetes, OR 4.3 (2.3–7.9) and chronic hypertension, OR 24.5 (95% CI 14.8–40.3). Both eclampsia and severe preeclampsia were associated with an increased risk of acute MI: eclampsia, OR 15.3 (95% CI 5.3–44.1) and severe preeclampsia, OR 6.9 (95% CI 3.7–13.1).
The incidence of acute MI (1 in 35,700) in our study was similar to the range reported in the literature and most likely reflects a true population frequency.1,3 The incidence of acute MI increased during the study period and suggests that we may see more pregnant women with an acute MI in the future as more high-risk women become pregnant. The 3 strongest independent predictors or factors associated with an acute MI were chronic hypertension, diabetes, and advancing maternal age. This older population of pregnant women with preexisting medical conditions may represent a unique group for early diagnosis, treatment, and further study.
In our population-based study of 151 women with acute MI diagnosed before or during labor, or up to 6 weeks postpartum, there was significant maternal morbidity and mortality compared with women without an acute MI. Our in-hospital mortality rate of 7.3% was much lower than previously reported (19–37%).1,2 Our figure may represent a true population-based frequency, or it could reflect improved diagnosis and treatment for women with acute MI. In addition, those women who may have died at home after discharge and were not readmitted to a hospital would not have been included in our data set. Besides mortality, there was a significant degree of morbidity with acute MI, which included all types of hypertensive disorders of pregnancy, including eclampsia, preeclampsia, chronic hypertension, coronary artery disease, and pregestational diabetes. Timing of the acute MI in relation to delivery showed distinct patterns as well. The majority (59%) of cases of acute MI in our study were diagnosed in the antenatal or intrapartum periods (Table 1). Roth and Elkayam4 similarly found that the majority of their 125 cases of acute MI collected from the literature occurred during the last one third of pregnancy. Our ability to link antenatal and postpartum hospitalizations to the delivery hospitalization allows us to identify very accurately the timing of acute MI and identify cases which might have been missed in their collection of cases due to lack of reporting.4 Furthermore, Roth and Elkayam4 had a 21% maternal mortality rate, which they thought was due in part to the physiologic processes associated with the labor and delivery itself. In our postpartum-diagnosed cases (41%), there were no maternal mortalities and including these cases in our study could partially explain our overall lower maternal mortality rate.
Timing of the diagnosis of acute MI in relation to the pregnancy demonstrated interesting characteristics. Women who were diagnosed during the antenatal period had higher mortality and pregnancy complications (Table 2) as compared with having the diagnosis in the postpartum period. Forty-three percent of the former group delivered preterm, and 40% were of LBW as compared with 17.7% and 14.2%, respectively, for the latter group. Clearly the antenatal acute MI results in worse pregnancy outcomes when compared to the postpartum diagnosis. The women diagnosed postpartum still had major increases in morbidity (Table 2) when compared with women who did not have an acute MI, suggesting the obvious, that any acute MI is a high-risk condition.
Many distinctive characteristics of women with an acute MI were demonstrated. In California, the frequency of women delivering over the age of 40 is only 2–3%.6 Seven percent of the acute MI patients were older than 40 years and 66% were older than 30 years. With older women delaying child birth into the fifth decade of life, we need to be more aware of the possibility of acute MI in this population. A most striking finding in patients with an acute MI was the degree of hypertensive comorbidities (Table 2). All types of hypertension, either preexisting or pregnancy-related, were significantly increased. Clearly, hypertensive disorders must predispose pregnant women to acute MI as they do for older nonpregnant parous women who had preeclampsia in the past.7
Both pregestational and gestational (data not shown) diabetes were significantly increased in women with an acute MI, which could be expected due to the fact that diabetes in the general population is associated with an increased risk of acute MI. After correcting for potential confounders, pregestational diabetes was still associated with an increased risk of acute MI, whereas gestational diabetes was not. In our study, coronary atherosclerosis was diagnosed in approximately 28% of women diagnosed with an acute MI before labor and in 39% of postpartum patients (Table 2), which is greatly increased over those women without an acute MI. The rate of atherosclerosis in our study was in between that reported by Hankins et al2 (13%) and Roth and Elkayam3 (43%), suggesting that our data are consistent with the literature. Obviously, if a woman is diagnosed with an acute MI, some workup will be undertaken to determine a cause for the acute MI, whereas a women without an acute MI would be far less likely to undergo a similar workup.
There are certain limitations to using administrative databases to perform population-based studies. A limitation of this study is that data are abstracted from a database of diagnoses entered by hospital personnel. Coding of diagnoses and procedures may vary from institution to institution, but because hospital reimbursement depends on complying with particular diagnostic and procedural codes, it is intuitive that hospitals will want to be as accurate as possible for billing purposes and to prevent fraud. The linkage of discharge and birth certificate databases has a reported accuracy of successful matches more than 97% of the time.5 Birth certificates are a reliable source of information for mother's age, race or ethnicity, marital status, education, parity, principal source of payment, and mode of delivery, with sensitivity, specificity, and positive predictive values of 95–99%.8 High agreement rates have also been reported for birth weight.8
With more women delaying child birth until the fifth and sixth decades of life, healthcare providers should be aware of the risk factors associated with an acute MI (chronic hypertension, diabetes, eclampsia, and preeclampsia) and potentially perform diagnostic testing on these women in an effort to identify and potentially treat women at high risk of an acute MI. In addition, with many women using assisted reproduction with donated ovum, women aged in their 50s and 60s are regularly becoming pregnant, placing them at an even greater risk of acute MI than at a younger age.