Pulmonary edema is defined as the accumulation of fluid in the pulmonary interstitial spaces and alveoli, which prevents the adequate diffusion of both oxygen and carbon dioxide. Although an uncommon event in pregnancy, it is associated with an increased risk of maternal and fetal morbidity and mortality.1 Several risk factors have been identified: preeclampsia or eclampsia,1–3 use of tocolytic therapy,4–10 severe infection,3,8,9 cardiac disease,11 iatrogenic fluid overload,5,11 and multiple gestation. Furthermore, the physiologic changes associated with pregnancy may predispose the pregnant patient to pulmonary edema.12–15 Despite the multiple causes of pulmonary edema, there is no clear consensus on the role that each of these risk factors plays in the pregnant population.
A MEDLINE search for the key words “pulmonary edema” or “heart failure” and “pregnancy” from 1966 to 2001 revealed case reports of acute pulmonary edema in pregnancy and diagnosis-specific reviews. There are no reports of the etiologies and outcomes in a large cohort. We sought to describe the attributable causes and overall and cause-specific rates of pulmonary edema in a large cohort.
MATERIALS AND METHODS
After approval by the Christiana Hospital institutional review board and human experimentation committee, we reviewed our obstetric database for the diagnosis of pulmonary edema at our institution between January 1, 1989 and June 1, 1999. This obstetric database is repeatedly validated and found to be 98% accurate. From the computerized database, we identified patients using Classification of Diseases, 9th Revision codes for the following diagnoses: pulmonary edema (646.13), heart failure (669.43 and 428.9), severe preeclampsia (642.53), pulmonary embolus (673.83), pneumonia (673.83 and 481), and pulmonary complications or assisted ventilation (946.56 and 946.57). When reviewing the charts for the diagnosis of pulmonary edema, all of the following criteria were necessary for entry into the study: significant shortness of breath, tachypnea, rales or rhonchi upon auscultation, evidence of hypoxia by pulse oximetry or arterial blood gas, and chest x-ray findings consistent with pulmonary edema.
Medical records were reviewed by the authors for maternal demographics, gestational age by a reliable last menstruation and/or early sonogram, gravidity, parity, number of fetuses, admission diagnosis, preexisting maternal disease, obstetric complications, drug and medication use, tobacco use, maternal symptoms, physical findings, laboratory studies, imaging studies, fluid balance, tocolytic use and type, evidence of preeclampsia,16 infection, and therapeutic measures. The timing of pulmonary edema in pregnancy and outcomes were recorded. The probable cause of the pulmonary edema was determined by the authors based on history, physical examinations, and laboratory and radiologic findings. In a few cases, more than one attributable cause was possible, and in those cases, the most likely diagnosis was chosen. Pregnancy outcomes included gestational age at delivery, birth weight, mode of delivery, and Apgar scores. Applied therapies such as supplemental oxygen, need for ventilatory support, fluid restriction, diuretic usage, cessation of tocolytics, antihypertensive use, termination of pregnancy, and antiarrhythmic usage were recorded.
The data were analyzed using a subgroup analysis by analysis of variance, Mantel–Haenszel, or Fisher exact test where appropriate. A P value of .05 was considered statistically significant.
During the study time period, 62,917 women delivered at our institution. Fifty-one patients (0.08%) were diagnosed with pulmonary edema during pregnancy or in the immediate postpartum period (24 hours or less). The cases were evenly distributed across the 10-year period. Pulmonary edema was most commonly diagnosed during the antepartum (47%) and postpartum (39%) periods (Table 1).
The mean maternal age of those diagnosed with pulmonary edema was 27.6 ± 6.4 years, with a range of 18–42 and a mean gestational age at diagnosis of 31.5 ± 6.8 weeks. Nineteen patients were delivered by cesarean. There was no difference between the subgroups with regard to maternal age or gestational age at diagnosis (Table 1).
The attributable causes and onset of pulmonary edema are described in Table 2. Pulmonary edema was most commonly secondary to tocolytic use and underlying cardiac disease. All women for whom tocolytic use was the probable cause of pulmonary edema were treated with multiple simultaneous tocolytics. The most common combination was intravenous magnesium sulfate and subcutaneous terbutaline (eight of 13 patients), with three of 13 receiving indomethacin with magnesium sulfate and two receiving magnesium sulfate, indomethacin, and terbutaline. Magnesium sulfate was administered as a 4–6-g bolus dose, followed by 2–5 g per hour. Indomethacin was administered as 50 mg per rectum every 6 hours and terbutaline as 250 μg every 20 minutes for a maximum of six doses, until diminution of contractions or a heart rate of at least 120 beats per minute.
Thirteen women (25.5%) diagnosed with pulmonary edema had underlying cardiac disease (Table 3). Although most women with underlying cardiac disease had valvular disease, structural cardiac disease was newly diagnosed in six women (11.7%) after the diagnosis of pulmonary edema. One woman without underlying heart disease had idiopathic supraventricular tachycardia, which was ultimately controlled with digoxin and verapamil.
Iatrogenic fluid overload was a common attributable cause, with 11 patients (21.5%) having this diagnosis. Women who had this as their etiology had no other risk factor identified. This was the most common reason for pulmonary edema in the postpartum time period (Table 2). All patients received isotonic fluid. Two patients received blood products for postpartum hemorrhage. Those women with fluid overload identified as their likely etiology had a significantly greater mean positive fluid balance during the 48 hours preceding the diagnosis (6022 ± 3340 mL versus 1017 ± 2757 mL, P = .003).
Nine patients (18%) developed pulmonary edema secondary to preeclampsia. Five of those patients had preexisting essential hypertension. The majority (56%) of pulmonary edema due to preeclampsia occurred in the antepartum period. There were no cases of eclampsia in this group.
Two patients (3.9%) had infection as the probable cause of their pulmonary edema. One patient had sepsis secondary to pyelonephritis. The other patient had left lower lobe pneumonia.
Pulmonary edema was attributed to incidents of each of the following: immunoglobulin A nephropathy and renal failure, acute fatty liver of pregnancy, and a paradoxical reaction to general anesthesia. One maternal mortality occurred in a patient with type I diabetes mellitus, who developed pulmonary edema secondary to infection. The patient recovered from pulmonary edema and was doing well when she developed diabetic ketoacidosis and adult respiratory distress syndrome.
Although pulmonary edema did not directly cause any neonatal morbidity, the underlying causes or preterm labor were associated with premature birth and the resultant neonatal morbidities. There were no neonatal deaths.
This is the first report examining a large cohort of pregnant women with the diagnosis of pulmonary edema for individual etiology and outcomes. In this cohort study, we report a review of all women diagnosed with acute pulmonary edema over a 10-year period in a busy tertiary care hospital. We found that the rate of acute pulmonary edema in pregnancy was 0.08%. Despite the low rate of occurrence, there was significant morbidity and one mortality associated with this diagnosis.
We found the most common time for the occurrence of pulmonary edema in our study was the antepartum period, closely followed by the postpartum period. Although this corroborates the findings by Mabie et al,17 it differs from two previous reports that found pulmonary edema in the postpartum period to be most common.12,18 This difference is most likely due to the different populations studied and/or different practice patterns. These reports attribute the late onset to the postpartum period mobilization of fluids combined with a 30% decrease in colloid osmotic pressure. These late physiologic changes highlight the importance of judicious use of intravenous fluids. In our report, a large portion of pulmonary edema was secondary to volume overload from vigorous intravenous hydration commonly associated with tocolytic use. This had become a common practice in our and other institutions despite any evidence to support intravenous hydration as a method to decrease preterm birth.19 Education of the obstetric staff and guidelines that limit the amount of hydration have helped us to limit the occurrence of this preventable cause of pulmonary edema since the completion of this study.
Structural or functional cardiac defects predispose a patient to developing pulmonary edema.20 However, cardiac disease is most likely under-diagnosed and under-reported because of under-utilization of diagnostic modalities such as echocardiography. In one report, 47% of women who had pulmonary edema had clinically unsuspected findings on echocardiography.17 The majority of these cases revealed systolic dysfunction or left ventricular hypertrophy rather than structural disease. There were no cases of valvular disease. In our report we also found that unsuspected findings on echocardiography are common, but more importantly, we found a high incidence of undiagnosed valvular disease.
Preeclampsia was the cause of pulmonary edema in 17.4% of patients in this report.3 These patients are at an increased risk for the development of pulmonary edema due to underlying endothelial damage and decreased colloid osmotic pressure, which cause leakage into the pulmonary interstitium or alveolar space. Combined with the left ventricular dysfunction and increase in peripheral vascular resistance found in preeclamptic patients, pulmonary edema develops. The development of pulmonary edema appears to be influenced by maternal age, parity, and preexisting essential hypertension.3 Two women in this report with chronic hypertension and superimposed preeclampsia developed pulmonary edema.
Despite the lack of convincing evidence that tocolytics improve neonatal outcomes, they are still commonly used.20 Thirteen patients (25.5%) in this report had their pulmonary edema attributed to the use of tocolytics. In our report, β-mimetics were the tocolytics used most frequently. The incidence of pulmonary edema with β-mimetic use alone has been estimated to be 5%.21 In our report, all women who had pulmonary edema attributed to tocolytics had simultaneous use of multiple tocolytics. The most common combination was subcutaneous terbutaline and intravenous magnesium sulfate; although magnesium sulfate has been associated with the occurrence of pulmonary edema,22 the etiology is less clear.23
This study has several strengths including the large population studied and strict outcome and diagnostic criteria. However, there are limitations:
1. We reviewed our database for the diagnosis codes as described above. Although this was an exhaustive review with broad diagnoses, the possibility of under-reporting exists.
2. Maternal echocardiograms were performed in 14 of the patients in this report. Although the cases where echocardiography was not performed appeared straightforward, the possibility of women with undetected heart disease exists. In those patients in whom an underlying etiology could not be detected, it is reasonable to expect that maternal echocardiography would have contributed a diagnosis. This would further reinforce the need for maternal echocardiography, especially in patients in whom the diagnosis was not clear.
3. Some patients had multiple possible etiologies for their pulmonary edema. Although we chose the most logical and apparent cause in each case, the possibility of incorrect assignment or multiple contributing factors must be considered.
4. This report was performed at a large tertiary care referral center, limiting the application of these findings to a low-risk population.
Several potential risk factors for the development of pulmonary edema in pregnancy exist. Many such factors often coexist in a given patient, though often one factor predominates. Our review of almost 63,000 pregnancies found the most common etiologies to be the use of tocolytic agents, cardiac disease, fluid overload, and preeclampsia. We encourage limiting treatment of preterm labor with a single tocolytic agent, if possible. The use of simultaneous multiagent therapy should be used with caution, especially with the lack of convincing evidence of its efficacy. When used, particular attention should be given to fluid balance, symptomatology, and physical examination. Although a case could be made for performing maternal echocardiography in all pregnant or postpartum women who have pulmonary edema, this may not be practical in many cases. However, in cases where no immediate and obvious cause for pulmonary edema is readily apparent, echocardiography should be performed early to evaluate the patient for possible underlying cardiac disease.
1. Sibai BM, Mabie WC, Harvey CJ, Gonzalez AR. Pulmonary edema in severe preeclampsia-eclampsia: Analysis of 37 consecutive cases. Am J Obstet Gynecol 1987;156:1174–9.
2. Benedetti TJ, Kates R, Williams V. Hemodynamic observations in severe preeclampsia complicated by pulmonary edema. Am J Obstet Gynecol 1985;152:330–4.
3. Donnelly JF, Lock FR. Causes of death in 533 fatal cases of toxemia of pregnancy. Am J Obstet Gynecol 1954;68:184–7.
4. Davies AE, Robertson MJS. Pulmonary edema after administration of IV salbutamol and ergometrine. Br J Obstet Gynaecol 1980;87:539–41.
5. Hankins GD, Havith JC, Kuehl TJ. Ritodrine hydrochloride infusion in pregnant baboon sodium and water compartment alterations. Am J Obstet Gynecol 1983;147:254–9.
6. Katz M, Robertson PA, Creasy RK. Cardiovascular complications associated with terbutaline treatment for preterm labor. Am J Obstet Gynecol 1981;139:605–8.
7. Pisani RJ, Rosenow ED. Pulmonary edema associated with tocolytic therapy. Ann Intern Med 1989;110:714–8.
8. Perry KG Jr, Morrison JC, Rust OA, Sullivan CA, Martin RW, Naef RW III. Incidence of adverse cardiopulmonary effects with low-dose continuous terbutaline infusion. Am J Obstet Gynecol 1995;73:1273–7.
9. Cunningham FG, Lucas MJ, Hankins GD. Pulmonary injury complicating antepartum pyelonephritis. Am J Obstet Gynecol 1987;156:797–807.
10. Towers CV, Kaminsdas CM, Garite TJ, Nageotte MP, Dorchester W. Pulmonary injury associated with antepartum pyelonephritis: Can patients at risk be identified? Am J Obstet Gynecol 1991;164:974–8.
11. Clark SL. Structural cardiac disease in pregnancy. In: Clark SL, Phelan JP, Cotton DB, eds. Critical care obstetrics. Oradell, New Jersey: Medical Economics Books, 1987:92.
12. Benedetti TJ, Carson RW. Studies of colloid osmotic pressure in pregnancy inducted hypertension. Am J Obstet Gynecol 1979;135:308–11.
13. Metcalfe JL, Ueland K. Maternal cardiovascular adjustments to pregnancy. Prog Cardiovasc Dis 1974;16:363–74.
14. Pritchard JA. Changes in blood volume during pregnancy and delivery. Anesthesiology 1965;26:39–42.
15. Scott DE. Anemia during pregnancy. Obstet Gynecol Annu 1972;1:219–44.
16. American College of Obstetricians and Gynecologists. Hypertension in pregnancy. ACOG technical bulletin no. 219. Washington: American College of Obstetricians and Gynecologists, 1996.
17. Mabie WC, Hackman BB, Sibai BM. Pulmonary edema associated with pregnancy: Echocardiographic insights and implications for treatment. Obstet Gynecol 1993;81:227–34.
18. Hankins GD. Acute pulmonary injury and respiratory failure during pregnancy. In: Clark SL, Phelan JP, Cotton DB, eds. Critical care obstetrics. Oradell, New Jersey: Medical Economics Books, 1987:290.
19. Guinn DA, Goepfert AR, Owen J, Brumfield C, Hauth JC. Management options in women with preterm uterine contractions: A randomized clinical trial. Am J Obstet Gynecol 1997;177:814–8.
20. Keirse MJ. New perspectives for the effective treatment of preterm labor. Am J Obstet Gynecol 1995;173:618–28.
21. Benedetti TJ. Life threatening complications of beta-mimetic therapy from preterm labor inhibition. Clin Perinatol 1986;13:843–52.
22. Ogburn PL Jr, Julian TM, Williams PP, Thompson TR. The use of magnesium sulfate for tocolysis in preterm labor complicated by twin gestation and betamimetic-induced pulmonary edema. Acta Obstet Gynecol Scand 1986;65:793–4.
23. Yeast JD, Halberstadt C, Meyer BA, Cohen GR, Thorp JA. The risk of pulmonary edema and colloid osmotic pressure changes during magnesium sulfate infusion. Am J Obstet Gynecol 1993;169:1566–71.