Obstetrics & Gynecology Email Alerts
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Obstetrics & Gynecology 2003;102:897-903
© 2003 by The American College of Obstetricians and Gynecologists
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Gilbert, T. T.
Right arrow Articles by Scardella, A. T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Gilbert, T. T.
Right arrow Articles by Scardella, A. T.

ORIGINAL RESEARCH

Obstetric Admissions to the Intensive Care Unit: Outcomes and Severity of Illness

Tricia T. Gilbert, MD, John C. Smulian, MD, MPH, Andrew A. Martin, MD, Cande V. Ananth, PhD, MPH, William Scorza, MD and Anthony T. Scardella, MD for the Critical Care Obstetric Team*

From the Division of Pulmonary and Critical Care Medicine, Department of Medicine; Division of Maternal-Fetal Medicine and Section of Epidemiology and Biostatistics, Department of Obstetrics, Gynecology and Reproductive Sciences, University of Medicine and Dentistry of New Jersey–Robert Wood Johnson Medical School/Saint Peter’s University Hospital, New Brunswick, New Jersey. ,*

Address reprint requests to: Anthony T. Scardella, MD, Saint Peter’s University Hospital, Division of Pulmonary and Critical Care Medicine, 254 Easton Avenue, New Brunswick, NJ 08901; E-mail: scardean{at}umdnj.edu.


   ABSTRACT
TOP
ABSTRACT
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 APPENDIX
 REFERENCES
 
OBJECTIVE: To determine whether mortality prediction based on a current model of outcome prediction is accurate in obstetric patients.

METHODS: Consecutive obstetric admissions to a medical intensive care unit from 1991 to 1998 were reviewed to determine whether mortality prediction is feasible in obstetric patients based on a widely used model. The Simplified Acute Physiologic Score (SAPS II) was used to predict the probability of hospital mortality.

RESULTS: The Simplified Acute Physiologic Score overestimated mortality in all patients (19 predicted deaths, eight observed) but accurately predicted mortality in patients admitted to the intensive care unit for medical reasons (seven predicted, five observed). The Simplified Acute Physiologic Score did not predict mortality in patients admitted for obstetric indications or postpartum hemorrhage. Median SAPS II scores were significantly higher in those patients who died, compared with survivors. For all groups, SAPS II scores were correlated with intensive care unit length of stay but not hospital length of stay.

CONCLUSION: The Simplified Acute Physiologic Score accurately predicts hospital mortality in obstetric patients admitted to the intensive care unit for medical reasons but not for indications related to pregnancy and delivery. An alternate model that predicts outcomes in obstetric patients admitted for obstetric indications should be developed.

Admission of an obstetric patient to the medical intensive care unit (ICU) occurs in roughly 2–4 per 1000 deliveries.1–3 Despite this low incidence, the overall acuity of this group of patients is high, with a mortality rate that varies widely, from 2.3% to 20%.1–6 Because there are approximately 4 million births in the United States each year, medical ICU admissions for obstetric patients might be required in as many as 16,000 cases yearly. In facilities where there are few obstetric ICU admissions, underestimation of severity of illness carries the potential for poor outcomes with excess morbidity, mortality, and resource utilization.

Little is known regarding the ability to assess severity of illness and predict outcomes in obstetric patients. Several investigators have applied a variety of scoring tools derived from nonobstetric populations (Acute Physiology and Chronic Health Evaluation, Mortality Probability Model, and the Simplified Acute Physiologic Score [SAPS II]) to obstetric populations in an attempt to predict the probability of hospital mortality.7–9 There are no current models that are designed specifically for use in obstetric patients. Therefore, this study was designed to describe illness characteristics of a large cohort of obstetric patients admitted to the ICU to determine whether mortality prediction was accurate with use of the SAPS II model.


   MATERIALS AND METHODS
TOP
ABSTRACT
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 APPENDIX
 REFERENCES
 
This institutional review board–approved study took place at Saint Peter’s University Hospital in New Brunswick, New Jersey. Saint Peter’s University Hospital is a 416-bed institution with a 30-bed ICU that contains separate medical, surgical, and coronary care units. Saint Peter’s University Hospital is a major teaching affiliate of the University of Medicine and Dentistry of New Jersey–Robert Wood Johnson Medical School. It hosts approximately 7000 deliveries each year and serves as a high-risk, tertiary level III, perinatal center that receives referrals from ten area community hospitals.

The study was based on all obstetric patients who were admitted to the ICU during the antepartum or postpartum (within 6 weeks of delivery) periods between January 1, 1991, and December 31, 1998. Medical records were obtained for all such admissions and were reviewed by one of two investigators (TTG, RLH). Coding protocol was established by the team before chart abstraction. Abstraction forms were designed a priori and were tested and revised before the abstraction of full data. Abstracted data included demographics, past medical history, prenatal history, delivery data, indications for ICU transfer, physiologic parameters used to calculate SAPS II, ICU complications, hospital length of stay, ICU length of stay, and death during the hospitalization. Patients were grouped by antepartum or post-partum status and by admitting diagnosis (obstetric, obstetric hemorrhage, or medical). Obstetric hemorrhage was considered a distinct entity in this study because the acute physiology in these patients differs markedly from patients with nonhemorrhagic complications, such as preeclampsia and other hypertensive disorders.

All critically ill obstetric patients at our hospital are required to be evaluated by consultants from maternal–fetal medicine and critical care and are admitted to the medical ICU. Additionally, any pregnant patient requiring telemetry monitoring was admitted to the medical ICU to ensure proper fetal monitoring. A working group that consists of members from both disciplines routinely reviews all admissions. The goal of this group is to assure cohesive, coordinated care.

The Simplified Acute Physiologic Score was used to assess the probability of hospital death upon ICU admission in each patient, as described previously.10 This scoring system was developed as an alternative to the Acute Physiology and Chronic Health Evaluation system to predict the risk of hospital mortality after admission to a medical–surgical ICU and is based on data generally available on admission to the ICU. It consists of 15 variables, each assigned a weighted score based on its deviation from normal values. The sum of the individual weighted scores for these variables yields the SAPS II score. The SAPS II score is then converted to probability of hospital mortality with a logistic regression equation, in which a higher score correlates with higher predicted mortality.

To obtain SAPS II data, we recorded the most abnormal value for each variable documented in the 24 hours before and through the first 24 hours after ICU admission. The variables were then assigned a value according to the weighted scoring system. The SAPS II variables included patient’s age, heart rate, systolic blood pressure, temperature, ratio of arterial oxygen tension to inspired oxygen fraction (PaO2/FIO2), urine output, blood urea nitrogen, white blood cell count, serum potassium, serum sodium, serum bicarbonate, total bilirubin, Glasgow Coma Score, type of admission (scheduled surgical, unscheduled surgical, medical), presence of a chronic illness (acquired immunodeficiency syndrome, hematologic malignancy, or metastatic cancer). Because the Glasgow Coma Score is not routinely documented in our ICU, a normal score was assigned unless there was clear documentation of a specific neurologic deficit that would allow the reviewer to retrospectively assign an accurate score.

Student t test and Mann–Whitney U test for continuous variables according to distribution were performed to assess the relationship between predicted and actual mortality. Pearson correlation coefficient was calculated to describe the association between the SAPS II score and length of ICU and hospital stay. Variability in the derived proportions was expressed as exact 95% confidence intervals. The population was evaluated in its entirety, followed by subgroup analyses. Subgroup analyses were performed by patients’ admission status (antepartum or postpartum) as well as by admission diagnosis (medical, obstetric, and obstetric hemorrhage).

A small proportion of patients (11 of 233) were admitted to the ICU on more than one occasion, thereby giving rise to nonindependent observations. Failure to account for this nonindependence typically leads to biased estimates of variance (although the point estimates remain unchanged). We replicated the analysis pertaining to mortality and SAPS II with and without accounting for this nonindependence. Because the results were essentially unchanged between the two approaches, we present the one based on ignoring the clustering. P values less than .05 were considered to indicate statistical significance, and all statistical tests were two-tailed.11


   RESULTS
TOP
ABSTRACT
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 APPENDIX
 REFERENCES
 
There were 233 obstetric patients (4.8 per 1000 deliveries) admitted to the medical ICU from a total of 49,349 hospital deliveries over an 8-year period. The mean (± standard deviation) of patients’ age was 30.7 ± 5.8 years. Approximately half of the patients were white, and one fourth were black. Approximately two thirds of the patients had private insurance, and one fourth were Medicaid recipients. Demographic characteristics for antepartum and postpartum subgroups are reported in Table 1Go.


View this table:
[in this window]
[in a new window]
  		Table 1. Demographics by Subgroups: Time of Admission (Antepartum/Postpartum) and Indication for Admission (Obstetric/Obstetric Hemorrhage/Medical)
 
Forty-two percent of patients (n = 98) had prior medical conditions. Cardiac conditions, including valvular heart disease and supraventricular arrhythmias, asthma, and hypertension were most common. Sixty-four percent (n = 149) had pregnancy-specific abnormal conditions or risk factors. Advanced maternal age and pregnancy-induced hypertension were the most common, followed by multiple gestations, preterm labor, and gestational diabetes. Six percent of the patients had abused either alcohol or drugs during pregnancy. Ten percent had either inadequate prenatal care (as stated on hospital admission note) or no prenatal care. Thirteen women had more than one admission to the ICU, and each admission was considered separately.

Sixty-five admissions (28%) were antepartum, and 168 (72%) were postpartum. Admitting diagnoses are shown in Table 2Go. The mean gestational age for antepartum admissions was 26.2 ± 8.0 weeks. The majority of antepartum admissions were for medical reasons, compared with postpartum admissions. Among the antepartum patients, pneumonia was the most common medical reason for admission, followed by arrhythmia and drug overdose. The median postpartum day for admission was 1 (admission on the day of delivery), with a range from 1 to 25 days postpartum. The most common reasons for postpartum admission were postpartum hemorrhage, preeclampsia, and eclampsia. Of all admissions, 92 (39.5%) were admitted for medical reasons, 85 (36.5%) for obstetric indications not related to obstetric hemorrhage, and 56 (24%) for obstetric hemorrhage. The most common etiologies for obstetric hemorrhage were placenta previa or accreta, ruptured uterus, and disseminated intravascular coagulopathy. There were four cases each of hemorrhage due to uterine atony, retained products of conception, abruption, and operative site bleeding. There were 15 cases of hemorrhage with no documented etiology for the bleeding.


View this table:
[in this window]
[in a new window]
  		Table 2. ICU Admitting Diagnoses of All 233 Obstetric Patients, by Type of Admission (Antepartum, Postpartum) and Indication Type (Medical, Obstetric)
 
There were eight deaths among obstetric patients admitted to the ICU between 1991 and 1998. The median SAPS II score for the deceased patients was 58 (range, 11–74), which is higher (P = .002) than that for the surviving population, which was 16 (range, 4–70). Diagnoses, SAPS II scores, and length of stay are shown in Table 3Go. While in the ICU, three developed sepsis, four developed acute respiratory distress syndrome (ARDS), and six developed disseminated intravascular coagulopathy (DIC). All of the deceased patients were delivered by cesarean. Five women died within 1 week of admission to the ICU.


View this table:
[in this window]
[in a new window]
  		Table 3. Summary of Deceased Obstetric Patients Admitted to the ICU, 1991–1998
 
Mortality predicted by SAPS II scores and actual mortality rates for the entire population were compared based on type of ICU admission (antepartum or postpartum) as well as admission indication (medical, obstetric, postpartum hemorrhage) and are shown in Table 4Go. For the entire cohort, SAPS II overestimated mortality, predicting 19 deaths (8.3%), whereas the actual mortality rate was 3.4% (eight deaths). After subgroup analysis, SAPS II accurately predicted hospital mortality for patients admitted with medical diagnoses (seven predicted, five observed, P =.22). The observed mortality rate was significantly lower than the predicted mortality for patients admitted with obstetric diagnoses and postpartum hemorrhage. The mean (standard deviation) ICU and hospital lengths of stay were 4.1 ± 9.5 and 12.9 ± 13.4 days, respectively. For all patients, SAPS II score correlated significantly with ICU length of stay (P < .01) but not hospital length of stay, for which patients with higher SAPS II scores at ICU admission had longer ICU stays.


View this table:
[in this window]
[in a new window]
  		Table 4. Mean SAPS II Score, Predicted and Actual Mortality
 

   DISCUSSION
TOP
ABSTRACT
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
APPENDIX
 REFERENCES
 
Advances in clinical medicine have made it possible for women with underlying medical conditions or with advanced maternal age to conceive and carry a fetus to term. This has led to an increase in the number of women requiring a greater degree of medical care in both the antepartum and postpartum periods. Medical conditions might present differently during pregnancy and might not only complicate pregnancy but also the disease state. Necessary treatments can impact the developing fetus and delivery management. From the obstetric standpoint, there are several acute complications specific to pregnancy that, although uncommon, require a high level of expertise and familiarity for proper management.12 Knowledge of these special circumstances is essential for both intensivists and obstetricians who are responsible for the care of these women.

Mabie and Sibai13 reported their experience of 200 obstetric patients admitted to an obstetric ICU. Their hospital had a designated obstetric ICU (admitted 0.9% of 7500 deliveries per year) staffed by specialists in maternal–fetal medicine. Nine of these patients (4.5%) required transfer to a medical ICU. Reasons for transfer to the general medical ICU included initiation of mechanical ventilation, thrombolytic therapy for pulmonary embolism, and chronic management of thrombotic thrombocytopenic purpura. In contrast to our series, the majority of women in their series were admitted with hypertensive disorders (46%), followed by medical complications of pregnancy (44%). Obstetric hemorrhage, the cause of 24% of our admissions, only comprised 10% of their obstetric ICU admissions.

Many smaller series, ranging from 20 to 131 patients, have been reported.1–4,6 The distribution of diagnoses were similar to our population, except for the series reported by Afessa et al,6 who had a predominance of postpartum admissions for respiratory failure and a high incidence of sepsis or systemic inflammatory response syndrome. Differences in patient populations, management protocols, and hospital nursing policies at each institution might account for the observed variations. More recently, Hazelgrove et al9 reported 210 obstetric ICU admissions with data pooled from 14 hospitals in southern England. The majority of their admissions were a result of postpartum obstetric complications. They noted that 37.5% of the women were admitted for less than 2 days and received no specific ICU interventions, suggesting that they might have been managed effectively outside the ICU. This is a legitimate concern, because resource utilization and cost-effective therapy are an integral part of modern medical practice.

Early and aggressive intervention is an important strategy in the care of these patients to ensure the best possible outcome for both mother and baby. Clinical experience is currently the only way to assess risk and predict the possibility of a poor outcome in these young women. A reliable model to predict outcomes on initial presentation would alert the care team as to the patient’s acuity. Such a model would allow proper mobilization of resources (notification of consultants, blood bank, transfer center) and would be especially helpful during initial assessments at centers that treat a small number of critically ill obstetric patients.

Models such as the Acute Physiology and Chronic Health Evaluation, the Mortality Probability Model, and SAPS II have been developed in the nonpregnant ICU population to predict adverse outcomes based on data available in the first 24 hours of admission.10,14,15 Many investigators have applied these models to obstetric patients with inconsistent results, and SAPS II has been the model most frequently studied. We found that SAPS II predicted mortality in the patients admitted with general medical conditions but not in patients with obstetric illnesses. El-Sohl et al7 found that SAPS II performed equally well in the obstetric population compared with a similar number of age- and sex-matched controls. Their patient population had a higher percentage of women admitted with medical diagnoses (70%) compared with our study (39%). Because SAPS II was created in a general medical population, it is not surprising that it would reflect severity of illness in an obstetric population that consisted mostly of patients with general medical conditions not related to pregnancy. Bhagwanjee et al8 assessed the ability of the Acute Physiology and Chronic Health Evaluation, the Glasgow Coma Scale score, and the organ system failure score to predict mortality in eclamptic patients. A low Glasgow Coma Scale score or organ system failure score (more than two organ systems failing for greater than 48 hours) were associated with increased likelihood of death, but no variables in the Acute Physiology and Chronic Health Evaluation predicted adverse outcome. This further suggests that obstetric patients are unique and that outcomes cannot be predicted based on currently available models.

Three additional studies have shown that SAPS II overestimates mortality in obstetric patients.9,16,17 Hazel-grove et al9 compared median SAPS II scores of obstetric patients with age-matched, nonpregnant controls and found that scores were lower for obstetric patients. They concluded that, in their series, there was a lower threshold for admitting obstetric patients to the ICU. We think that the physiologic variables that are normal in obstetric patients are assigned points in the SAPS II scoring system. For example, SAPS II assigns 5 points for a systolic blood pressure between 70 and 99 mm Hg, a range of values that are normal in pregnancy. Additionally, SAPS II assigns 6 points for intubation and mechanical ventilation, which is a frequent occurrence in obstetric patients requiring surgery for postpartum hemorrhage. We think that because normal physiologic variables in obstetric patients are scored as abnormal in SAPS II, a more inclusive model might result in more representative severity of illness scores and assist in the identification of patients who truly have a higher mortality risk.

There are numerous reasons why the physiologic abnormalities in obstetric patients are not reflected in SAPS II and other traditional scoring systems. For obstetric patients, tests of liver function, platelet count, and uric acid levels can be important in the assessment of patients with hypertensive disorders of pregnancy, such as hemolysis, elevated liver enzymes, and low platelets (HELLP) and eclampsia. Some quantification of transfusion requirements might have a predictive value in patients admitted for postpartum hemorrhage. These variables are neither reflected in SAPS II nor any other available scoring system for critical illness. Another inadequacy of applying SAPS II to an obstetric population is that it was created in an older (mean age 57.2 years), predominantly male (59.6%) population.10 Its weighted scoring system assigns points based on deviation from values that are normal for this population. These normal values do not apply to pregnancy. For example, no points are assigned for patients older than 40 years. Because 27% of the patients in our study were between 35 and 40 years of age, an established risk marker for several adverse outcomes of pregnancy, the SAPS II scores might not adequately reflect the true severity of illness of these patients. The Simplified Acute Physiologic Score also uses blood urea nitrogen as a marker for renal dysfunction and considers values less than 28 mg/dL to be within normal limits. In pregnancy, the blood urea nitrogen normally decreases to 8 to 9 mg/dL, and any higher level might indicate a prerenal or renal abnormality. As a result of these and other differences between pregnancy and the nonpregnant state, models for predicting outcome in the general medical ICU patients might be inadequate to predict outcomes in obstetric patients.

Our study indicates that the SAPS II scoring rule accurately predicted hospital mortality among patients admitted to the ICU for medical reasons but performs poorly in predicting deaths for patients admitted for obstetric reasons as well as for those admitted for postpartum hemorrhage (Table 4Go). A post hoc power analysis indicated that our study had a power of 97.1% for analysis related to the predicting of mortality for medical reasons (Type I error of 5%). Thus, the large size of our study allowed us to draw valid (and robust) statistical inferences for predicting hospital mortality with the SAPS II scoring rule.

We have described the ICU admissions of 233 critically ill obstetric patients with a tremendous diversity in demographic characteristics and disease processes. In our population, antepartum patients are more likely to be admitted with general medical conditions unrelated to pregnancy. Postpartum patients are more likely to be admitted on the day of delivery because of hemorrhagic complications or hypertensive disorders. The Simplified Acute Physiologic Score accurately predicts hospital death in obstetric patients admitted for medical reasons but not in obstetric patients with disorders related to pregnancy and delivery.


   APPENDIX
TOP
ABSTRACT
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 APPENDIX
REFERENCES
 
Members of the Critical Care Obstetric Team include the following: Critical Care Medicine: Amar Bukhari, MD, Andrew Martin, MD, Tricia Gilbert, MD, Helen Miley, PhD, and Anthony Scardella, MD; Maternal Fetal Medicine/Obstetric Medicine: Cande Ananth, PhD, MPH, Michael Carson, MD, William Scorza, MD, John Smulian, MD, MPH, and Anthony Vintzileos, MD.


   Footnotes
 
* For other members of the Critical Care Obstetric Team, see the Appendix. Back

This project was funded in part by the James O’Connell Fellowship Award. Drs. Smulian and Ananth were supported, in part, by a grant from the National Institutes of Health (NICHD, R01-HD-38902) awarded to Dr. Ananth.

The authors thank Robin L. Hardie, MD, for her contribution during data acquisition.

doi:10.1016/S0029-7844(03)00767-1

Received April 10, 2003. Received in revised form June 19, 2003. Accepted June 26, 2003.


   REFERENCES
TOP
ABSTRACT
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 APPENDIX
 REFERENCES
 
1. Kilpatrick S, Matthay M. Obstetric patients requiring critical care, a five-year review. Chest 1992;101:1407–12.[Abstract/Free Full Text]

2. Mahutte N, Murphy-Kaulbeck L, Le Q, Solomon J, Benjamin A, Boyd M. Obstetric admissions to the intensive care unit. Obstet Gynecol 1999;92:263–6.

3. Lapinsky S, Kruczynski K, Seaward G, Farine D, Grossman R. Critical care management of the obstetric patient. Can J Anaesth 1997;44:325–9.[Abstract/Free Full Text]

4. Collop NA, Sahn S. Critical illness in pregnancy. Chest 1993;103:1548–52.[Abstract/Free Full Text]

5. Lewinsohn G, Herman A, Leonov Y, Klinowski E. Critically ill obstetrical patients: Outcome and predictability. Crit Care Med 1994;22:1412–4.[Medline]

6. Afessa B, Green B, Delke I, Koch K. Systemic inflammatory response syndrome, organ failure and outcome in critically ill obstetric patients treated in an ICU. Chest 2001;120:1271–7.[Abstract/Free Full Text]

7. El-Solh A, Grant B. A comparison of severity of illness scoring system for critically ill obstetric patients. Chest 1996;110:1299–304.[Abstract/Free Full Text]

8. Bhagwanjee S, Paruk F, Moodley J, Muckart D. Intensive care unit mortality from eclmpsia: An evaluation of the Acute Physiology and Chronic Health Evaluation II score and the Glasgow Coma Scale score. Crit Care Med 2000; 28:120–4.[Medline]

9. Hazelgrove J, Price C, Pappachan VJ, Smith G. Multi-center study of obstetric admissions to 14 intensive care units in southern England. Crit Care Med 2001;29:770–5.[Medline]

10. Le Gall J, Lemeshow S, Saulnier F. A new simplified acute physiology score (SAPS II) based on a European/North American multi-center study. JAMA 1993;270:2957–63.[Abstract]

11. Riffenburgh RH. Statistics in medicine. San Diego: Academic Press, 1999.

12. Lapinsky S, Kruczynski K, Slutsky A. Critical care in the pregnant patient. Am J Respir Crit Care Med 1995;152: 427–55.[Medline]

13. Mabie WC, Sibai BM. Treatment in an obstetric intensive care unit. Am J Obstet Gynecol 1990;162:1–4.[Medline]

14. Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: A severity of disease classification system. Crit Care Med 1985;13:818–29.[Medline]

15. Lemeshow S, Teres D, Klar J, Avrunin J, Gelbach S, Rapoport J. Mortality probability models (MPM II) based on an international cohort of intensive care unit patients. JAMA 1993;270:2478–86.[Abstract]

16. Arabi Y, Goraj R, Horsfall D, Itani M, Djazmati W, Armstrong D, et al. Scoring systems in obstetric patients requiring intensive care in Saudi Arabia: A ten-year review. Chest 1999;116:286S.

17. Scarpinato L. Critically ill obstetric patients: Outcome and predictability utilizing the SAPS II score in a 314 bed community hospital. Chest 1995;108:184S.





This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Gilbert, T. T.
Right arrow Articles by Scardella, A. T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Gilbert, T. T.
Right arrow Articles by Scardella, A. T.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS