Acute respiratory distress syndrome (RDS) is an important cause of morbidity and mortality among nonpregnant1–3 and pregnant4–6 patients with acute critical illnesses. Acute RDS was first described by Ashbaugh et al in 19677; the diagnostic criteria were refined by Murray et al in 19888 and revised by a consensus development conference in 1994.9 Acute RDS, also called adult RDS, is defined as a lung disease with acute onset, bilateral infiltrates on chest x-ray, no evidence of intravascular volume overload or pulmonary artery wedge pressure no greater than 18 mmHg, and severely impaired oxygenation, with arterial oxygen tension divided by fraction of inspired oxygen of less than 200 mmHg. Acute RDS may be triggered by direct lung injury such as aspiration, drowning, or lung trauma, or indirect injury, most commonly sepsis, shock, or multiple transfusions. The frequency of acute RDS in the general population is estimated at 1.5 per 100,000 per year,1 with a fatality rate of 35–50%.1–3
The first series of cases of acute RDS in pregnancy was published in 1975; four of the six patients died.10 The majority of reports have been of one or two cases, with only three published series of over ten patients.11–13 The purpose of this study was to broaden understanding of the causes, courses, complications, and outcomes of acute RDS in pregnancy and the postpartum period.
Materials and Methods
Since 1988, one of us (VAC) has recorded all cases of suspected acute RDS in association with pregnancy that were treated by the maternal fetal medicine and pulmonary medicine groups at the University of Arkansas for Medical Sciences (from January 1988 to July 1989) and Sharp Memorial Hospital and Sharp Mary Birch Hospital for Women (from August 1989 to January 2000).
Acute RDS was defined according to consensus conference criteria.9 Only patients who required intubation and mechanical ventilation during pregnancy or within 7 days of delivery were included in the study. Patients were classified as consult if they were managed at other institutions with our consultation, transport if they were transferred to us during the hospitalization during which acute RDS occurred, and primary otherwise. A multidisciplinary team, including maternal fetal medicine, pulmonary medicine, anesthesiology, and medical and surgical subspecialists, provided patient care at our institutions. Invasive hemodynamic monitoring, ventilatory care, and ventilatory mode were individualized. Electronic fetal monitoring was used in undelivered fetuses at or beyond 26 weeks' gestation; decisions for delivery were made jointly by maternal fetal medicine, neonatology, and pulmonary medicine.
Charts were reviewed to assess criteria for acute RDS, cause or causes of acute RDS, associated injury to other organ systems, ventilator days, hospital stays, and, for maternal survivors, condition at discharge. Gestational age was taken from prenatal records; for patients without prenatal care, the gestational age estimate from sonography during the hospitalization was used. For patients who required intubation before delivery, the timing and indication for delivery, birth weight, Apgar scores for the newborn, and condition at discharge were noted. Perinatal asphyxia was defined according to ACOG criteria. Maternal mortality rates were expressed as a percentage, in accordance with the conventions of the critical care literature, rather than per 100,000 pregnancies which is the standard obstetric usage.
Proportions and confidence intervals were calculated using exact binomial technique. χ2 testing was used when appropriate. Analyses were performed using Epilnfo 6.02 (Centers for Disease Control and Prevention, Atlanta, GA, 1994).
The project was approved by the institutional review board of Sharp Healthcare.
Twenty-eight women with acute RDS required intubation during pregnancy or within 1 week postpartum. Eight of them have been included in previous reports.14–18 Eight mothers were transports, six were consults, and the remaining 14 were primary patients. Consults were treated at a total of five other institutions.
The rate of acute RDS, excluding maternal transports and consults, was one per 6277 births, or 0.016% (95% confidence interval 0, 0.027%). Maternal demographics are presented in Table 1. Twenty-three mothers had regular prenatal care, two had irregular care, and three had no prenatal care.
Initial treatment was with intubation and volume-cycled ventilation. Nine mothers were subsequently changed to pressure-controlled ventilation, two had high frequency ventilation, and two had extracorporeal carbon dioxide removal. Nitric oxide was used in one case. Swan-Ganz monitoring was used in all but one case. Pressors were used to treat maternal hypotension in all but four cases. Broad-spectrum antibiotic coverage was administered routinely.
Barotrauma requiring chest tubes occurred in ten women, including two with pneumomediastinum as well as pneumothorax. There were two major complications of chest tube placement, subclavian vein laceration resulting in maternal death and splenic laceration requiring emergency laparotomy and splenectomy.
Three mothers had cesarean hysterectomy. Indications were placenta accreta, placenta percreta, and a gas-forming intrauterine infection after fetal death.15 There were two postpartum hysterectomies. One was for postcesarean hemorrhage. The other was for persistent sepsis and acute RDS after more than 1 month of antibiotic therapy for postcesarean endometritis; there was dramatic clinical improvement. Other operations included tracheostomies and one laporotomy for a posthysterectomy retroperitoneal hematoma.
Etiology for acute RDS and maternal deaths are summarized in Table 2. The maternal mortality rate was 11 of 28 (39.3%, 95% confidence interval 21.5%, 59.4%). Mortality rates were three of six among consult patients, three of eight among transport patients, and six of 14 among primary patients.
Six of the 11 maternal deaths occurred in mothers with multiple organ system failure in whom hypotension that could not be corrected with fluids and pressors ultimately developed. There were two deaths from intractable hypoxia without multiple organ system failure, and one patient was brain dead at transfer. Two maternal deaths occurred as complications of therapy, one as a result of subclavian vein laceration at chest tube placement and another due to a hypoxic event during tracheostomy that ultimately resulted in brain death.
Multiple organ system failure was common. Central nervous system injury occurred in five patients (two with intracranial hemorrhage, two with cerebral edema from preeclampsia, and one with intractable postpartum seizures); all died. Hepatic involvement, manifest as transaminase values over twice the upper limit of normal, was seen in 12 patients; six died. Hepatic failure (bilirubin over 6 mg/dL or prothrombin time above control +4 seconds) occurred in seven patients. Two with acute fatty liver of pregnancy survived, and the remaining five women with hepatic failure died. Renal insufficiency, manifest as serum creatinine more than 1.5 mg/dL, occurred in 17 patients; eight died. Renal failure, defined as prolonged oliguria (less than 160 mL/8 hours) or creatinine more than 5.0 mg/dL, occurred in eight women, of whom five died. Two of four women who required dialysis survived. Combined renal and hepatic failure occurred in four women; none survived.
Hematologic involvement, manifest as disseminated intravascular coagulation, white blood cell count under 1000/mm3, or platelet count under 20,000/mm3, occurred in 14 patients, of whom eight died. Hemolytic uremic syndrome or thrombotic thrombocytopenic purpura was diagnosed in three mothers; two had plasmapheresis, and one survived. Deaths occurred in seven of 14 with intubation for 1–7 days, in two of six with intubation for 8–14 days, and in two of eight with intubation over 14 days. Three mothers were intubated for 6 weeks or more; all survived.
In patients with acute RDS from infection, the range of causative organisms was broad. Five mothers had amnionitis or endometritis, including two with group A streptococci and one with Escherichia coli. One developed amnionitis with multiple organisms, including Klebsiella, E coli, Enterococcus, and possibly Clostridia perfringens after rupture of membranes at 16 weeks. In one mother with endometritis, cultures were negative, but Gram stain at the time of hysterectomy weeks later was suggestive of C perfringens. Nonobstetric infections included E coli pyelonephritis, Streptococcus pneumoniae septicemia, spontaneous Staphylococcus aureus sepsis, influenza A pneumonia, and Varicella pneumonia and encephalitis. One patient had presumed viral pneumonia but the causative organism was not identified.
Among the 17 survivors, three had serious sequelae. One had a myopathy of unknown origin during her 72-day hospital stay and was discharged to a rehabilitation facility. Three months after delivery, she could ambulate only with the assistance of a walker. One was dependent on dialysis at discharge and at 9-month follow-up. One was discharged home on day 19 but was hospitalized elsewhere 2 weeks later for a small bowel obstruction that required surgery.
Intubation for acute RDS was required before delivery in 15 cases. Fetal death preceded intubation in three women, one with a cord prolapse during expectant management for premature rupture of membranes (PROM), one during hospitalization for viral pneumonia, and one with eclampsia and fetal death. Only one patient was discharged undelivered; she had pyelonephritis and acute RDS at 16 weeks, was intubated for 10 days, and ultimately delivered at term. One fetus died in utero at 24 weeks with the death of the mother.
There were ten cases of third-trimester acute RDS with a living fetus at the time of intubation, which are summarized in Table 3. Three were delivered because of concerns regarding the maternal condition; all neonates survived and none had evidence of perinatal asphyxia. One infant was delivered to a brain-dead mother when Torolopsis glabrata amnionitis developed at 28 weeks; the infant did well. Fetal heart rate tracings with absent variability and repetitive late decelerations after spontaneous contractions occurred within 4 days of admission in each of the remaining six pregnancies. Among these six, there was one neonatal death; one newborn chart was unavailable for review, three infants had evidence of perinatal asphyxia, and one had a smooth neonatal course and was well at age 6 years.
Nine of 28 cases of acute RDS were potentially preventable. One patient had PROM at 16 weeks but was not offered prophylactic antibiotics, hospital admission, or delivery; amnionitis and sepsis developed the next day. One woman with recurrent urinary tract infections was noncompliant with her prophylactic drug regimen and developed acute RDS resulting from pyelonephritis. One had acute RDS from influenza when immunization was not routinely recommended. One woman had persistent hemorrhage after cesarean delivery, embolization was attempted to avoid hysterectomy. She had a cardiac arrest, was resuscitated, and had a hysterectomy, with resolution of the bleeding. One patient was admitted with status epilepticus postpartum. She had continuous seizures for over 4 hours and severe hyperthermia and multiple organ system failure developed.
In three women, preeclampsia was expectantly managed after 32 weeks' gestation despite 3+ or 4+ proteinuria. They subsequently were delivered for eclampsia or worsening preeclampsia and acute RDS developed. One mother with severe preeclampsia was fluid restricted for 18 hours postpartum despite oliguria and hypotension. She subsequently developed multiple organ system failure and acute RDS and died.
In this series, the rate of pregnancy-related acute RDS, excluding maternal transports, was one per 6227 deliveries. There have been three prior series with more than ten patients with obstetric acute RDS. Smith et al11 reported 14 cases with a rate of one per 3371 pregnancy-related admissions; excluding nine maternal transports, the rate at their institution is calculated as one per 10,113. Mabie et al12 reported 16 cases, with an overall incidence of one per 2893 deliveries; excluding maternal transports, their institution's rate was similar to ours, one per 6612 deliveries. Perry et al13 reported 41 cases, with a much higher rate (.7 per 1000 live births) but did not differentiate between primary versus transported patients. Even after correcting for the duration of pregnancy, each of these figures is substantially higher than the rate for the general population, about 1.5 per 100,000 per year.1
The leading causes of acute RDS in our population were infection, preeclampsia, aspiration, and hemorrhage, as in previous studies. However, in contrast to the series of Smith et al,11 where 12 of 14 cases had obstetric causes, most of our patients with infection and 13 of 28 overall had nonobstetric conditions as the cause of respiratory failure. Causes reported among the obstetric series are quite different from those in the general population, where infection, aspiration, and trauma account for about 40%, 30%, and 25% of cases respectively.2
The mortality rate in our series was 11 of 28 (39.3%), which is similar to that of Smith et al,11 with six of 14 (43.8%), and Mabie et al,12 with a rate of seven of 16 (44%). Perry et al13 had 41 patients with ten maternal deaths (24.4%) and speculated that the improvement relative to the experiences of Smith et al11 and Mabie et al12 might be due to differences in patient population or improvements in the care of acute RDS. However, each of the reported series is small; none of the differences between their outcomes and ours reaches statistical significance.
Multiple organ system failure was the most common cause of maternal death in our patients (six of 11 deaths). It was also the most common cause in prior series, occurring in three of six deaths in Smith et al,11 five of seven in Mabie et al,12 and four of ten deaths in Perry et al.13 There were two deaths in our series due to intractable hypoxemia without multiple organ system failure and one that followed barotrauma and chest tube placement.
A surprising finding was the high frequency of unusual bacterial pathogens. Two patients had group A streptococcal infections, one had S aureus septicemia of unknown etiology, and one had persistent sepsis with probable C perfringens despite prolonged antibiotic therapy. The remarkable response to hysterectomy in this latter patient's case reinforced the concept, previously promoted by Smith et al,11 that aggressive surgical control of infection can be lifesaving in patients with acute RDS.
In a previous review article,18 we commented that the maternal mortality risk appeared to be independent of the length of ventilatory support, but we tempered our conclusion because of the potential bias toward publishing reports of cases, particularly cases with prolonged courses, with positive outcomes. Among patients intubated for 15–28 days, there were four survivors and one death in the series of Smith et al11 and two survivors and three deaths in series of Mabie et al.12 Corresponding figures for intubation longer than 28 days were one survivor and three deaths11 and one survivor and no deaths,12 respectively. Perry et al13 did not give duration of mechanical ventilation but found no difference between intensive care unit stays among survivors (8.3 ± 7.9 days) versus nonsurvivors (9.3 ± 6.6 days). The present series strongly supports aggressive ventilatory support; we found survival rates of 75% in both the 15–28 and longer than 28 days groups.
The issue of the effects of acute RDS and maternal therapy for acute RDS on the fetus has generated considerable commentary.18–20 There have been numerous case reports in which the fetus remained undelivered despite maternal acute RDS requiring intubation and ventilation. However, our prior review,18 the experience of others,12,13 and the present series suggest that the pregnant woman with acute RDS in the third trimester seldom remains undelivered for more than a few days, except in cases when the underlying cause is maternal pyelonephritis or Varicella pneumonia. In our review, only ten of 39 patients with antepartum acute RDS (including one patient from the present series) were discharged undelivered, and all had pyelonephritis or Varicella. The high rates of fetal death, spontaneous preterm labor, and fetal heart rate abnormalities in labor, as well as the high risk of perinatal asphyxia among surviving infants, suggest that the best option for treatment of maternal acute RDS after 28 weeks is expeditious delivery.
A crucial point is that of prevention of acute RDS in pregnancy. Although many of the cases reported in the literature and in this series occurred as a result of acute events, some cases of acute RDS, particularly those due to infection and to complications of preeclampsia, are preventable. Among the nine patients in this series with potentially preventable acute RDS, there were three maternal deaths and one maternal survivor with renal failure, and one death in utero.
1. Luce JM. Acute lung injury and the acute respiratory distress syndrome. Critical Care Med 1998;26:369–76.
2. Hudson LD, Steinberg KP. Acute respiratory distress syndrome: Clinical features, management, and outcome. In: Fishman AP, Elias JA, Fishman J, Kaiser R, eds. Fishman's pulmonary diseases and disorders. McGraw-Hill, New York, 1998:2549–65.
3. Milberg JA, Davis DR, Steinberg KP, Hudson LD. Improved survival of patients with acute respiratory distress syndrome (ARDS), 1983–1993. JAMA 1995;273:306–9.
4. Collop NA, Sahn SA. Critical illness in pregnancy: An analysis of 20 patients admitted to a medical intensive care unit. Chest 1993;103:1548–52.
5. Monaco TJ Jr, Spielman FJ, Katz VL. Pregnant patients in the intensive care unit: A descriptive analysis. South Med J 1993;86:414–7.
6. Kirshon B, Hinkley CM, Cotton DB, Miller J. Maternal mortality in a maternal-fetal medicine intensive care unit. J Reprod Med 1990;35:25–8.
7. Ashbaugh DG, Bigelow DB, Perry TL, Levine BE. Acute respiratory distress in adults. Lancet 1967;2:319–23.
8. Murray YF, Mathay MA, Luce JM, Flick MR. Pulmonary perspectives: An expanded definition of the adult respiratory distress syndrome. Am Rev Respir Dis 1988;138:720–33.
9. Bernard GR, Artigas A, Brigham KL, Carlet J, Falla K, Hudson L, et al. The American-European consensus conference on ARDS. Definitions, mechanisms, relevant outcomes, and clinical trial coordination. Am J Respir Crit Care Med 1994;149:818–24.
10. Futoran JM, Hill JD. Pulmonary insufficiency associated with pregnancy. Am J Obstet Gynecol 1975;121:637–40.
11. Smith JL, Thomas F, Orme JF Jr, Clemmer TP. Adult respiratory distress syndrome during pregnancy and immediately postpartum. Am J Med 1990;153:508–10.
12. Mabie WC, Barton JR, Sibai BM. Adult respiratory distress syndrome in pregnancy. Am J Obstet Gynecol 1992;167:950–7.
13. Perry KG Jr, Martin RW, Blake PG, Roberts WC, Martin JN Jr. Maternal mortality associated with adult respiratory distress syndrome. South Med J 1998;91:441–4.
14. Catanzarite VA, Steinberg SM, Mosley CA, Landers CE, Cousins LM, Schneider JM. Severe preeclampsia with fulminant and extreme elevation of aspartate aminotransferase and lactate dehydrogenase levels: High risk for maternal death. Am J Perinatol 1995;12:310–3.
15. Catanzarite V, Schibanoff JM, Chinn R, Mendoza A, Weiss R. Overwhelming maternal sepsis due to a gas-forming Escherichia coli
chorioamnionitis. Am J Perinatol 1994;11:205–7.
16. Hurley TJ, Montgomery R, Waldron JA Jr, Catanzarite V, Quirk G Jr. Fatal course of malignant non-Hodgkin's lymphoma of T-cell type during pregnancy with metastasis to the fetus. J Matern Fetal Med 1994;3:69–74.
17. Catanzarite VA, Willms DC, Holdy KE, Gardner SE, Ludwig DM, Cousins LM. Brain death during pregnancy: Tocolytic therapy and aggressive maternal support on behalf of the fetus. Am J Perinatol 1997;14:431–4.
18. Catanzarite VA, Willms D. Adult respiratory distress syndrome in pregnancy: Report of three cases and review of the literature. Obstet Gynecol Surv 1997;52:381–92.
19. Van Hook JW. Acute respiratory distress syndrome in pregnancy. Semin Perinatol 1997;21:320–7.
20. Hankins GDV. Acute pulmonary injury and respiratory failure during pregnancy. In: Clark SL, Cotton DB, Hankins GDV, Phelan JP, eds. Critical care obstetrics. 2nd ed. Boston: Blackwell Scientific Publications, 1991;340–70.