- Email: [email protected]
Prospective randomized controlled trial of fetal fibronectin on preterm labor management in a tertiary care center
American Journal of Obstetrics and Gynecology (2004) 190, 358e62
www.elsevier.com/locate/ajog
GENERAL OBSTETRICS AND GYNECOLOGY: OBSTETRICS
Prospective randomized controlled trial of fetal fibronectin on preterm labor management in a tertiary care center M. Patrick Lowe, MD,a Bridget Zimmerman, PhD,b Wendy Hansen, MDa,* Departments of Obstetrics and Gynecologya and Biostatistics,b University of Iowa Hospitals and Clinics, Iowa City, Iowa Received for publication April 3, 2003; revised July 15, 2003; accepted August 27, 2003
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– KEY WORDS Preterm Fetal fibronectin Length of stay
Objective: The purpose of this study was to investigate the effect of the rapid fetal fibronectin on the length of hospital stay and the use of preterm labor interventions in a tertiary care center. Study design: Women who were seen in the Labor and Delivery Unit with symptoms of preterm labor were assigned randomly to receive fetal fibronectin (n = 46 women) or to preterm labor management without fetal fibronectin (n = 51 women). Physicians were not blinded to the result; groups were compared. Results: There was no difference between groups in demographic or obstetric characteristics, the hours spent in labor and delivery, the number of women who were admitted to the antepartum service, the length of stay, or medical interventions. When the results for women with a negative fetal fibronectin test were compared to women with a positive fetal fibronectin test, a significant difference was found in admissions to the antepartum service (P = .032) and the length of stay (P = .008). Conclusions: A negative fetal fibronectin test was associated with fewer admissions to the antepartum ward and a shorter length of stay. Ó 2004 Elsevier Inc. All rights reserved.
––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– Preterm delivery is the greatest contributor to neonatal morbidity and death, with close to 11% of all deliveries in the United States occurring before 37 weeks of gestation.1 Hospitalizations for preterm labor comprise a major cost of perinatal care and, in one study, accounted for 33% of all admissions before delivery.2 In the past, our ability to
Supported by Process Improvement Grant, University of Iowa. * Reprint requests: Wendy Hansen, MD, Division of MaternalFetal Medicine, Department of Obstetrics and Gynecology, University of Iowa Hospitals and Clinics, 200 Hawkins Dr, Iowa City, IA 52242. E-mail: [email protected] 0002-9378/$ - see front matter Ó 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.ajog.2003.08.041
make the diagnosis of preterm labor has relied solely on clinical assessment: preterm contractions with subtle cervical change. More recently, a laboratory test has become clinically available: vaginal fetal fibronectin (fFN) to help determine the risk of early delivery. fFN, first identified in 1985 by Matsura and Hakomori,3 is an extracellular matrix protein that is found between the chorion and decidua. It has a critical role in maintaining contact between the uterus and placenta. fFN is normally present in the cervicovaginal fluid in the first 22 weeks of gestation, but is typically absent until 34 weeks of gestation. Its reappearance is thought to be part of the normal physiologic process that occurs
359
Lowe, Zimmerman, and Hansen Table I
Demographics and obstetric characteristics at examination
Demographic
No fFN (n = 51)
fFN (n = 46)
P value
Mean age (y)* Median gravidity (interquartile range) Median parity (interquartile range) Median gestational age at time of test (wk) (interquartile range) R28 w !28 w Previous preterm birth Referring physician: UIHCy Family practice referral Local obstetrician Other Multiple gestation Cervical dilation in centimeters 0 1-1.5 2-2.5 3-4 Cervical effacement in centimeters 0 1-1.5 2
26.7G5.8 2 (1-3) 1 (0-1) 30.6 (26.6-32.4)
27.4G5.3 2.5 (2-3) 1 (0-1) 30.4 (27.1-32.0)
.543 .683 .337 .902
37/51 14/51 14/51 30 6 13 2 3
(73%) (27%) (27%) (59%) (12%) (25%) (4%) (6%)
30/46 16/46 12/46 28 6 11 1 5
(65%) (35%) (26%) (61%) (13%) (24%) (2%) (11%)
29 12 5 5
(57%) (23%) (10%) (10%)
23 10 9 4
(50%) (22%) (20%) (9%)
.437
2 (4%) 11 (24%) 33 (72%)
.448
3 (6%) 7 (14%) 41 (80%)
.734
.471
* Values are given as meanGSD. y University of Iowa Hospitals and Clinics.
in preparation for delivery. The primary purpose of our study was to investigate the effect of the rapid fFN on the length of hospital stay and the use of preterm labor interventions. We also evaluated the performance characteristics of the rapid assay in our population.
Material and methods The study was approved by the University of Iowa Hospitals and Clinics Institutional Review Board. Women with signs and symptoms of preterm labor (uterine contractions and/or cervical change) who were examined at the University of Iowa Labor and Delivery Unit or those women who were transferred already receiving tocolytic medications were eligible for the study. Inclusion criteria were gestational age between 23 and 34 weeks, O16 years of age, cervical dilatation of %3 cm for primigravid women, and %4 cm for multiparous women. Exclusion criteria were higher order (more than twins) multifetal gestations, cerclage, preterm premature rupture of membranes, and vaginal bleeding. Gestational age was assigned by the patient’s last menstrual period and first- or early second-trimester ultrasound. If the last menstrual period was unknown or a discrepancy existed regarding dates, the gestational age was assigned on the basis of the ultrasound. Randomization was achieved through the use of a computer-generated table
in blocks of 10. Separate randomization tables were used for gestations of !28 weeks and R28 weeks. The results of randomization were concealed through the use of opaque, sealed envelopes that were numbered sequentially. The consent was attached to the outside of the envelope. Randomization results were revealed to the physician and patient only after informed consent was obtained and the envelope was opened. Informed consent was obtained by one of the residents who was assigned to labor and delivery. Randomization was to one of two groups: preterm labor management with fFN as soon as criteria were met or preterm labor management without fFN. Preterm labor management was left to the discretion of the treating physician. The University of Iowa seemed particularly well suited for such a study because we were introducing fFN into clinical practice and all patients are treated by the same faculty practice. The fFN assay was performed when the following criteria were met: 24 hours after sexual intercourse, digital examination, transvaginal ultrasound scanning, or use of creams or lubricants. Women with a cervical examination, transvaginal ultrasound scan, or intercourse within the previous 24 hours were enrolled in the study; however, the fFN assay was delayed until these criteria were met. FFN was collected in the following way. A Dacron swab was rolled against the posterior lip of the cervix. The collected specimen was then placed into a buffer
360 Table II
Lowe, Zimmerman, and Hansen Outcome statistics
Outcome
No fFN (n = 51)
fFN (n = 46)
P value
Administration of magnesium Administration of antibiotics Administration of betamethasone Median labor and delivery in hours (interquartile range) Antepartum: admitted, yes or no Length of stay: d Antepartum: d 0 1 2-3 R4 Median gestational age at delivery in wk (interquartile range)
23 21 22 16 12
22 (48%) 17 (37%) 23 (50%) 8.5 (3-36) 16 (35%)
.840 .683 .545 .417 .265 .224
(45%) (41%) (43%) (3-33) (24%)
39 (76%) 9 (18%) 1 (2%) 2 (4%) 37.4 (35.0-39.0)
30 (65%) 11 (24%) 4 (9%) 1 (2%) 38.3 (36.0-38.9)
.258
livery and length of stay on antepartum service were compared with the use of the Wilcoxon rank sum test; antepartum admission and administration of betamethasone, magnesium, and antibiotics was compared with the use of the Fisher exact test. Cox proportional hazard regression was used to assess the effect of fFN on the time to discharge from labor and delivery while controlling for previous preterm birth, cervical dilation, and gestational age (!or R28 weeks) between groups.
Results Figure Estimate of ‘‘survival’’ (not discharged) distribution, based on Cox proportional hazard regression, for those women with no previous preterm birth, with cervical dilatation, and at R28 weeks of gestational age.
solution and sent to the laboratory. The results were available within 1 hour. Results were reported as positive if the assay measured O50 ng/mL and negative if !50 ng/mL. This study was supported by a process improvement grant sponsored by the University of Iowa. All fFN tests were paid for by the grant. We calculated that a sample size of 50 in each group was necessary to detect a reduction in the length of stay of at least 1.3 days at the .05 significance level, with 0.80 power. Using ICD-9 Diagnosis Code 6440-Threatened Premature Labor, our Joint Office for Compliance calculated an average length of stay of 2.8 G 2.3 days between January through June of 2000, which was the 6-month period that preceded the study. Data were analyzed with SAS software (version 8.1; SAS Institute Inc, Cary, NC). Demographic and clinical characteristics of the two groups (fFN vs no fFN) were compared with the use of the two-sample t test or the Wilcoxon rank sum test for the continuous and ordinal variables and with the Fisher exact test for the categoric variables. The outcome variables of time in labor and de-
A total of 110 patients were enrolled into the study between August 2000 and May 2002. Four women were excluded because of readmissions, 1 woman was found to have preterm premature rupture of membranes, 1 woman had an urgent cesarean delivery, and data from 1 woman was unable to be obtained because of blood in the vagina. Three women were assigned randomly to receive the fFN test but were discharged before it could be performed. One woman was found to be at O34 weeks of gestation, 1 woman was O4 cm dilated, and 1 woman had inadequate documentation. Ninety-seven women were available for the study analysis. Forty-six women were assigned randomly to receive fFN, and 51 women were assigned randomly to preterm labor management without a fFN. Demographic and obstetric characteristics of the study population are detailed in Table I. There was no statistically significant difference between groups with regard to age, gravidity, parity, previous preterm birth, referring physician, multiple gestations, or gestational age at the time of the test. There were no significant differences between groups in cervical dilation or effacement on examination. We evaluated the effect of fFN on the time spent in labor and delivery, admission to the antepartum unit, and length of stay on the antepartum ward, along with the use of tocolysis, betamethasone, and antibiotics between
361
Lowe, Zimmerman, and Hansen Table III !7 d !14 d !37 wk
Fetal fibronectin as a predictor of preterm delivery Sensitivity
Specificity
Predictive value of positive test
Predictive value of negative test
2/3 66.7% (9.4, 99.2) 3/4 75.0% (0.6, 80.6) 5/13 38.5% (13.9, 68.5)
34/43 79.1% (64.0, 90.0) 34/42 81.0% (65.9, 91.4) 27/33 81.8% (64.5, 93.0)
2/11 18.2% (2.3, 51.8) 3/11 27.3% (6.0, 61.0) 5/11 45.5% (16.8, 76.6)
34/35 97.1% (85.1, 99.9) 34/35 81.0% (85.1, 99.9) 27/35 77.1% (59.9, 89.6)
those who received fFN and those who did not. There was no significant difference between groups (Table II). Furthermore, no significant difference was found when the subgroups at !28 weeks of gestation or at R28 weeks of gestation were analyzed. With the use of the Cox proportional hazard and an adjustment for previous preterm birth, cervical dilation, and gestational age, those women who received fFN had a shorter length of stay in labor and delivery (hazard ratio of being discharged, 1.7; 95% CI, 1.1-2.7; P = .017) than those women who did not receive fFN (Figure). Of the 46 women who were assigned randomly to receive the assay, 35 women received negative test results, and 11 women received positive test results. These two groups were compared for time in labor and delivery, antepartum admission, and length of stay. Although there was no statistically significant difference in the time spent in labor and delivery (for women with a negative fFN result: median, 9 hours; interquartile range, 336 hours; for women with a positive fFN result: median, 8 hours, interquartile range, 4-36 hours [P = .806]), there were significantly fewer admissions to the antepartum ward (25.7% vs 63.6%, P = .032) and significantly shorter lengths of stay on the antepartum service (median, 0; interquartile range, 0-1 day; vs median, 1; interquartile range, 0-3 days; P = .008) with a negative fFN result. The sensitivity, specificity, negative predictive value, and positive predictive value in our population can be found in Table III. FFN had a negative predictive value of 97.1% (95% CI, 85.1%-99.9%) and a positive predictive value of 27.3% (95% CI, 6.0%-61.0%) for delivery within 14 days.
Comment In 1991, Lockwood et al4 published the first study in which the presence of fFN in preterm labor patients was associated with preterm delivery. Since then, several investigators have studied the ability of fFN to predict preterm delivery in women with symptoms of preterm labor and have found a negative predictive value for delivery within 1 week near 99% and 95% for 2 weeks.5,6 These studies relied on the fFN enzyme immunoassay with a minimum turnaround time of 24 hours. Our study used the rapid test, and decision-making was not compromised by a 24-hour waiting period.
Although fFN is an effective marker of preterm delivery, the limitations of testing have compromised its clinical use. Both sexual activity and cervical examination within 24 hours of sample collection have been associated with false-positive results.7 In contrast, the manufacturer reports that the use of lubricants, soaps, or disinfectants (such as povidone-iodine, miconazole, or lubricant) may physically interfere with the absorption of the specimen onto the swab, which renders the test invalid. These limitations, coupled with the cervical examination, which is the cornerstone of an evaluation for preterm labor, has challenged the most effective point in time to perform an fFN assay. Our study enrolled all women: women who we examined primarily, women who were referred from the clinic, and women who were transferred to our tertiary care center. Many of the women had received magnesium, betamethasone, and antibiotics before the transfer; other women had this regimen started on the basis of clinical examination when an fFN could not be done at first because of previous vaginal examinations. Our aim was to study the effect of fFN prospectively in actual clinical practice. Our constant was the single group of treating physicians. Interestingly, our study found no difference between groups in the use of medical interventions or in length of stay, a finding that was similar to those studies that blinded the practitioner and relied on the 24-hour test. One reason for this lack of difference is the difficulty of the implementation of fFN testing up front because the presentations of preterm labor at a tertiary care center are so varied. Given the different scenarios in clinical practice, we had to tailor the timing of the test to each situation. Giles et al8 recently studied the effect of an fFN test in the community hospitals and its effect on maternal transfers in New South Wales, Australia. They found that a negative fFN result was not helpful if cervical dilation was present because 5 of 106 women (4.7%) were delivered within 7 days of a negative test result. However, the use of the test was associated with a 90% reduction in maternal transfer and a substantial cost savings. Another reason for the lack of difference may be sample size. Our study was only large enough to detect a 1.3-day reduction (almost 50%) in the length of stay and may not detect smaller differences. Our study did find a significant difference in admissions to the antepartum service and length of stay on
362 the antepartum ward with a negative fFN result compared with a positive fFN result, which highlights the potential impact of a negative fFN result on treatment even when done later, after the decisions for tocolysis and betamethasone have been made. In actual clinical practice, there is more to the management of preterm labor than tocolysis, betamethasone, and hospitalization. Often, women are taken out of work and placed on some modification of bed rest with patient instructions that are as varied as the number of physicians who are administering the treatment. Although we did not study the effect of fFN on bed rest and work reduction recommendations, the overall consensus was to allow a woman to return to work and to care for her family if the fFN result was negative. This remains a potential area of study. fFN performed favorably in our patient population. Tests were performed by many different residents, which supports its wide clinical application. Only one woman was delivered within 7 days of a negative fFN result. This was the third woman who was enrolled in the study. This prompted a review of the testing criteria with the residents. Further safety monitoring did not reveal any other unexpected or potentially harmful effects. Its sensitivity, specificity, positive predictive value, and negative predictive value are similar to other published findings (Table III). Although the implementation of fFN testing in our institution did not change the overall use of tocolysis
Lowe, Zimmerman, and Hansen or length of stay, a negative fFN result did decrease admissions to the antepartum ward and lengths of stay compared with a positive fFN result.
References 1. Guyer B, Strobino DM, Ventura SJ, Singh GK. Annual summary of vital statistics: 1994. Pediatrics 1995;96:1029-39. 2. Nicholson WK, Frick KD, Powe NR. Economic burden of hospitalizations for preterm labor in the United States. Obstet Gynecol 2000;96:95-101. 3. Matsura H, Hakomori SI. The oncofetal domain of fibronectin defined by monoclonal antibody FDC-6: its presence in fibronectins from fetal and tumor tissues and its absence in those from normal adult tissues and plasma. Proc Natl Acad Sci U S A 1985;82:6517-21. 4. Lockwood CJ, Senyei AE, Dische MR, Shah KD, Thung SN, Jones L, et al. Fetal fibronectin in cervical and vaginal secretions as a predictor of preterm delivery. N Engl J Med 1991;325:669-74. 5. Iams JD, Casal D, Macgregor JA, Goodwin TM, Kreaden US, Lowensohn R, et al. Fetal fibronectin improves the accuracy of diagnosis of preterm labor. Am J Obstet Gynecol 1995;173:141-5. 6. Peaceman AM, Andrews WW, Thorp JM, Cliver AP, Lukes A, Iams JD, et al. Fetal fibronectin as a predictor of preterm birth in patients with symptoms: a multicenter trial. Am J Obstet Gynecol 1997;177:13-8. 7. Lukes AS, Thorp JM Jr, Eucker B, Pahel-Short L. Predictors of positivity for fetal fibronectin in patients with symptoms of preterm labor. Am J Obstet Gynecol 1997;176:639-41. 8. Giles WMB, Bisits AD, Knox M, Madsen G, Smith R. The effect of fetal fibronectin testing on admissions to a tertiary maternal-fetal medicine unit and cost savings. Am J Obstet Gynecol 2000;182: 439-42.
www.elsevier.com/locate/ajog
GENERAL OBSTETRICS AND GYNECOLOGY: OBSTETRICS
Prospective randomized controlled trial of fetal fibronectin on preterm labor management in a tertiary care center M. Patrick Lowe, MD,a Bridget Zimmerman, PhD,b Wendy Hansen, MDa,* Departments of Obstetrics and Gynecologya and Biostatistics,b University of Iowa Hospitals and Clinics, Iowa City, Iowa Received for publication April 3, 2003; revised July 15, 2003; accepted August 27, 2003
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– KEY WORDS Preterm Fetal fibronectin Length of stay
Objective: The purpose of this study was to investigate the effect of the rapid fetal fibronectin on the length of hospital stay and the use of preterm labor interventions in a tertiary care center. Study design: Women who were seen in the Labor and Delivery Unit with symptoms of preterm labor were assigned randomly to receive fetal fibronectin (n = 46 women) or to preterm labor management without fetal fibronectin (n = 51 women). Physicians were not blinded to the result; groups were compared. Results: There was no difference between groups in demographic or obstetric characteristics, the hours spent in labor and delivery, the number of women who were admitted to the antepartum service, the length of stay, or medical interventions. When the results for women with a negative fetal fibronectin test were compared to women with a positive fetal fibronectin test, a significant difference was found in admissions to the antepartum service (P = .032) and the length of stay (P = .008). Conclusions: A negative fetal fibronectin test was associated with fewer admissions to the antepartum ward and a shorter length of stay. Ó 2004 Elsevier Inc. All rights reserved.
––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– Preterm delivery is the greatest contributor to neonatal morbidity and death, with close to 11% of all deliveries in the United States occurring before 37 weeks of gestation.1 Hospitalizations for preterm labor comprise a major cost of perinatal care and, in one study, accounted for 33% of all admissions before delivery.2 In the past, our ability to
Supported by Process Improvement Grant, University of Iowa. * Reprint requests: Wendy Hansen, MD, Division of MaternalFetal Medicine, Department of Obstetrics and Gynecology, University of Iowa Hospitals and Clinics, 200 Hawkins Dr, Iowa City, IA 52242. E-mail: [email protected] 0002-9378/$ - see front matter Ó 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.ajog.2003.08.041
make the diagnosis of preterm labor has relied solely on clinical assessment: preterm contractions with subtle cervical change. More recently, a laboratory test has become clinically available: vaginal fetal fibronectin (fFN) to help determine the risk of early delivery. fFN, first identified in 1985 by Matsura and Hakomori,3 is an extracellular matrix protein that is found between the chorion and decidua. It has a critical role in maintaining contact between the uterus and placenta. fFN is normally present in the cervicovaginal fluid in the first 22 weeks of gestation, but is typically absent until 34 weeks of gestation. Its reappearance is thought to be part of the normal physiologic process that occurs
359
Lowe, Zimmerman, and Hansen Table I
Demographics and obstetric characteristics at examination
Demographic
No fFN (n = 51)
fFN (n = 46)
P value
Mean age (y)* Median gravidity (interquartile range) Median parity (interquartile range) Median gestational age at time of test (wk) (interquartile range) R28 w !28 w Previous preterm birth Referring physician: UIHCy Family practice referral Local obstetrician Other Multiple gestation Cervical dilation in centimeters 0 1-1.5 2-2.5 3-4 Cervical effacement in centimeters 0 1-1.5 2
26.7G5.8 2 (1-3) 1 (0-1) 30.6 (26.6-32.4)
27.4G5.3 2.5 (2-3) 1 (0-1) 30.4 (27.1-32.0)
.543 .683 .337 .902
37/51 14/51 14/51 30 6 13 2 3
(73%) (27%) (27%) (59%) (12%) (25%) (4%) (6%)
30/46 16/46 12/46 28 6 11 1 5
(65%) (35%) (26%) (61%) (13%) (24%) (2%) (11%)
29 12 5 5
(57%) (23%) (10%) (10%)
23 10 9 4
(50%) (22%) (20%) (9%)
.437
2 (4%) 11 (24%) 33 (72%)
.448
3 (6%) 7 (14%) 41 (80%)
.734
.471
* Values are given as meanGSD. y University of Iowa Hospitals and Clinics.
in preparation for delivery. The primary purpose of our study was to investigate the effect of the rapid fFN on the length of hospital stay and the use of preterm labor interventions. We also evaluated the performance characteristics of the rapid assay in our population.
Material and methods The study was approved by the University of Iowa Hospitals and Clinics Institutional Review Board. Women with signs and symptoms of preterm labor (uterine contractions and/or cervical change) who were examined at the University of Iowa Labor and Delivery Unit or those women who were transferred already receiving tocolytic medications were eligible for the study. Inclusion criteria were gestational age between 23 and 34 weeks, O16 years of age, cervical dilatation of %3 cm for primigravid women, and %4 cm for multiparous women. Exclusion criteria were higher order (more than twins) multifetal gestations, cerclage, preterm premature rupture of membranes, and vaginal bleeding. Gestational age was assigned by the patient’s last menstrual period and first- or early second-trimester ultrasound. If the last menstrual period was unknown or a discrepancy existed regarding dates, the gestational age was assigned on the basis of the ultrasound. Randomization was achieved through the use of a computer-generated table
in blocks of 10. Separate randomization tables were used for gestations of !28 weeks and R28 weeks. The results of randomization were concealed through the use of opaque, sealed envelopes that were numbered sequentially. The consent was attached to the outside of the envelope. Randomization results were revealed to the physician and patient only after informed consent was obtained and the envelope was opened. Informed consent was obtained by one of the residents who was assigned to labor and delivery. Randomization was to one of two groups: preterm labor management with fFN as soon as criteria were met or preterm labor management without fFN. Preterm labor management was left to the discretion of the treating physician. The University of Iowa seemed particularly well suited for such a study because we were introducing fFN into clinical practice and all patients are treated by the same faculty practice. The fFN assay was performed when the following criteria were met: 24 hours after sexual intercourse, digital examination, transvaginal ultrasound scanning, or use of creams or lubricants. Women with a cervical examination, transvaginal ultrasound scan, or intercourse within the previous 24 hours were enrolled in the study; however, the fFN assay was delayed until these criteria were met. FFN was collected in the following way. A Dacron swab was rolled against the posterior lip of the cervix. The collected specimen was then placed into a buffer
360 Table II
Lowe, Zimmerman, and Hansen Outcome statistics
Outcome
No fFN (n = 51)
fFN (n = 46)
P value
Administration of magnesium Administration of antibiotics Administration of betamethasone Median labor and delivery in hours (interquartile range) Antepartum: admitted, yes or no Length of stay: d Antepartum: d 0 1 2-3 R4 Median gestational age at delivery in wk (interquartile range)
23 21 22 16 12
22 (48%) 17 (37%) 23 (50%) 8.5 (3-36) 16 (35%)
.840 .683 .545 .417 .265 .224
(45%) (41%) (43%) (3-33) (24%)
39 (76%) 9 (18%) 1 (2%) 2 (4%) 37.4 (35.0-39.0)
30 (65%) 11 (24%) 4 (9%) 1 (2%) 38.3 (36.0-38.9)
.258
livery and length of stay on antepartum service were compared with the use of the Wilcoxon rank sum test; antepartum admission and administration of betamethasone, magnesium, and antibiotics was compared with the use of the Fisher exact test. Cox proportional hazard regression was used to assess the effect of fFN on the time to discharge from labor and delivery while controlling for previous preterm birth, cervical dilation, and gestational age (!or R28 weeks) between groups.
Results Figure Estimate of ‘‘survival’’ (not discharged) distribution, based on Cox proportional hazard regression, for those women with no previous preterm birth, with cervical dilatation, and at R28 weeks of gestational age.
solution and sent to the laboratory. The results were available within 1 hour. Results were reported as positive if the assay measured O50 ng/mL and negative if !50 ng/mL. This study was supported by a process improvement grant sponsored by the University of Iowa. All fFN tests were paid for by the grant. We calculated that a sample size of 50 in each group was necessary to detect a reduction in the length of stay of at least 1.3 days at the .05 significance level, with 0.80 power. Using ICD-9 Diagnosis Code 6440-Threatened Premature Labor, our Joint Office for Compliance calculated an average length of stay of 2.8 G 2.3 days between January through June of 2000, which was the 6-month period that preceded the study. Data were analyzed with SAS software (version 8.1; SAS Institute Inc, Cary, NC). Demographic and clinical characteristics of the two groups (fFN vs no fFN) were compared with the use of the two-sample t test or the Wilcoxon rank sum test for the continuous and ordinal variables and with the Fisher exact test for the categoric variables. The outcome variables of time in labor and de-
A total of 110 patients were enrolled into the study between August 2000 and May 2002. Four women were excluded because of readmissions, 1 woman was found to have preterm premature rupture of membranes, 1 woman had an urgent cesarean delivery, and data from 1 woman was unable to be obtained because of blood in the vagina. Three women were assigned randomly to receive the fFN test but were discharged before it could be performed. One woman was found to be at O34 weeks of gestation, 1 woman was O4 cm dilated, and 1 woman had inadequate documentation. Ninety-seven women were available for the study analysis. Forty-six women were assigned randomly to receive fFN, and 51 women were assigned randomly to preterm labor management without a fFN. Demographic and obstetric characteristics of the study population are detailed in Table I. There was no statistically significant difference between groups with regard to age, gravidity, parity, previous preterm birth, referring physician, multiple gestations, or gestational age at the time of the test. There were no significant differences between groups in cervical dilation or effacement on examination. We evaluated the effect of fFN on the time spent in labor and delivery, admission to the antepartum unit, and length of stay on the antepartum ward, along with the use of tocolysis, betamethasone, and antibiotics between
361
Lowe, Zimmerman, and Hansen Table III !7 d !14 d !37 wk
Fetal fibronectin as a predictor of preterm delivery Sensitivity
Specificity
Predictive value of positive test
Predictive value of negative test
2/3 66.7% (9.4, 99.2) 3/4 75.0% (0.6, 80.6) 5/13 38.5% (13.9, 68.5)
34/43 79.1% (64.0, 90.0) 34/42 81.0% (65.9, 91.4) 27/33 81.8% (64.5, 93.0)
2/11 18.2% (2.3, 51.8) 3/11 27.3% (6.0, 61.0) 5/11 45.5% (16.8, 76.6)
34/35 97.1% (85.1, 99.9) 34/35 81.0% (85.1, 99.9) 27/35 77.1% (59.9, 89.6)
those who received fFN and those who did not. There was no significant difference between groups (Table II). Furthermore, no significant difference was found when the subgroups at !28 weeks of gestation or at R28 weeks of gestation were analyzed. With the use of the Cox proportional hazard and an adjustment for previous preterm birth, cervical dilation, and gestational age, those women who received fFN had a shorter length of stay in labor and delivery (hazard ratio of being discharged, 1.7; 95% CI, 1.1-2.7; P = .017) than those women who did not receive fFN (Figure). Of the 46 women who were assigned randomly to receive the assay, 35 women received negative test results, and 11 women received positive test results. These two groups were compared for time in labor and delivery, antepartum admission, and length of stay. Although there was no statistically significant difference in the time spent in labor and delivery (for women with a negative fFN result: median, 9 hours; interquartile range, 336 hours; for women with a positive fFN result: median, 8 hours, interquartile range, 4-36 hours [P = .806]), there were significantly fewer admissions to the antepartum ward (25.7% vs 63.6%, P = .032) and significantly shorter lengths of stay on the antepartum service (median, 0; interquartile range, 0-1 day; vs median, 1; interquartile range, 0-3 days; P = .008) with a negative fFN result. The sensitivity, specificity, negative predictive value, and positive predictive value in our population can be found in Table III. FFN had a negative predictive value of 97.1% (95% CI, 85.1%-99.9%) and a positive predictive value of 27.3% (95% CI, 6.0%-61.0%) for delivery within 14 days.
Comment In 1991, Lockwood et al4 published the first study in which the presence of fFN in preterm labor patients was associated with preterm delivery. Since then, several investigators have studied the ability of fFN to predict preterm delivery in women with symptoms of preterm labor and have found a negative predictive value for delivery within 1 week near 99% and 95% for 2 weeks.5,6 These studies relied on the fFN enzyme immunoassay with a minimum turnaround time of 24 hours. Our study used the rapid test, and decision-making was not compromised by a 24-hour waiting period.
Although fFN is an effective marker of preterm delivery, the limitations of testing have compromised its clinical use. Both sexual activity and cervical examination within 24 hours of sample collection have been associated with false-positive results.7 In contrast, the manufacturer reports that the use of lubricants, soaps, or disinfectants (such as povidone-iodine, miconazole, or lubricant) may physically interfere with the absorption of the specimen onto the swab, which renders the test invalid. These limitations, coupled with the cervical examination, which is the cornerstone of an evaluation for preterm labor, has challenged the most effective point in time to perform an fFN assay. Our study enrolled all women: women who we examined primarily, women who were referred from the clinic, and women who were transferred to our tertiary care center. Many of the women had received magnesium, betamethasone, and antibiotics before the transfer; other women had this regimen started on the basis of clinical examination when an fFN could not be done at first because of previous vaginal examinations. Our aim was to study the effect of fFN prospectively in actual clinical practice. Our constant was the single group of treating physicians. Interestingly, our study found no difference between groups in the use of medical interventions or in length of stay, a finding that was similar to those studies that blinded the practitioner and relied on the 24-hour test. One reason for this lack of difference is the difficulty of the implementation of fFN testing up front because the presentations of preterm labor at a tertiary care center are so varied. Given the different scenarios in clinical practice, we had to tailor the timing of the test to each situation. Giles et al8 recently studied the effect of an fFN test in the community hospitals and its effect on maternal transfers in New South Wales, Australia. They found that a negative fFN result was not helpful if cervical dilation was present because 5 of 106 women (4.7%) were delivered within 7 days of a negative test result. However, the use of the test was associated with a 90% reduction in maternal transfer and a substantial cost savings. Another reason for the lack of difference may be sample size. Our study was only large enough to detect a 1.3-day reduction (almost 50%) in the length of stay and may not detect smaller differences. Our study did find a significant difference in admissions to the antepartum service and length of stay on
362 the antepartum ward with a negative fFN result compared with a positive fFN result, which highlights the potential impact of a negative fFN result on treatment even when done later, after the decisions for tocolysis and betamethasone have been made. In actual clinical practice, there is more to the management of preterm labor than tocolysis, betamethasone, and hospitalization. Often, women are taken out of work and placed on some modification of bed rest with patient instructions that are as varied as the number of physicians who are administering the treatment. Although we did not study the effect of fFN on bed rest and work reduction recommendations, the overall consensus was to allow a woman to return to work and to care for her family if the fFN result was negative. This remains a potential area of study. fFN performed favorably in our patient population. Tests were performed by many different residents, which supports its wide clinical application. Only one woman was delivered within 7 days of a negative fFN result. This was the third woman who was enrolled in the study. This prompted a review of the testing criteria with the residents. Further safety monitoring did not reveal any other unexpected or potentially harmful effects. Its sensitivity, specificity, positive predictive value, and negative predictive value are similar to other published findings (Table III). Although the implementation of fFN testing in our institution did not change the overall use of tocolysis
Lowe, Zimmerman, and Hansen or length of stay, a negative fFN result did decrease admissions to the antepartum ward and lengths of stay compared with a positive fFN result.
References 1. Guyer B, Strobino DM, Ventura SJ, Singh GK. Annual summary of vital statistics: 1994. Pediatrics 1995;96:1029-39. 2. Nicholson WK, Frick KD, Powe NR. Economic burden of hospitalizations for preterm labor in the United States. Obstet Gynecol 2000;96:95-101. 3. Matsura H, Hakomori SI. The oncofetal domain of fibronectin defined by monoclonal antibody FDC-6: its presence in fibronectins from fetal and tumor tissues and its absence in those from normal adult tissues and plasma. Proc Natl Acad Sci U S A 1985;82:6517-21. 4. Lockwood CJ, Senyei AE, Dische MR, Shah KD, Thung SN, Jones L, et al. Fetal fibronectin in cervical and vaginal secretions as a predictor of preterm delivery. N Engl J Med 1991;325:669-74. 5. Iams JD, Casal D, Macgregor JA, Goodwin TM, Kreaden US, Lowensohn R, et al. Fetal fibronectin improves the accuracy of diagnosis of preterm labor. Am J Obstet Gynecol 1995;173:141-5. 6. Peaceman AM, Andrews WW, Thorp JM, Cliver AP, Lukes A, Iams JD, et al. Fetal fibronectin as a predictor of preterm birth in patients with symptoms: a multicenter trial. Am J Obstet Gynecol 1997;177:13-8. 7. Lukes AS, Thorp JM Jr, Eucker B, Pahel-Short L. Predictors of positivity for fetal fibronectin in patients with symptoms of preterm labor. Am J Obstet Gynecol 1997;176:639-41. 8. Giles WMB, Bisits AD, Knox M, Madsen G, Smith R. The effect of fetal fibronectin testing on admissions to a tertiary maternal-fetal medicine unit and cost savings. Am J Obstet Gynecol 2000;182: 439-42.