We will evaluate a risk assessment with fibronectin testing for women with clinical signs of threatened preterm labor. We will assess whether triage of these women with fibronectin testing, cervical length measurements or a combination is cost-effective. The results of this study can be used to support a policy of expectant management without immediate referral to a neonatal centre and without tocolysis in women with a low risk for preterm birth.
All women between 24 and 34 weeks of gestational age with primary complaints associated with preterm labor and intact membranes will be eligible for inclusion in the study. Women with vaginal bleeding, a cerclage, cervical dilatation of more than 3 cm or previous treatment for threatened preterm labor in the current pregnancy are excluded. Other exclusion criteria are hypertension, contra indications for nifedipine and fetal distress or maternal conditions that may lead to pregnancy termination within 7 days. The study will be limited to women admitted to a perinatal centre with neonatal intensive care facilities.
Procedures, recruitment, randomisation and collection of baseline data
Women meeting the eligibility criteria will be informed and invited to participate in the study and asked for written informed consent. At study entry baseline demographics, obstetric and medical history will be recorded. Before cervical length measurement, women will be tested for fibronectin. Therefore a sample of cervicovaginal secretion will be taken by rotating a swab in the posterior vaginal fornix for 10 seconds. Analysis will be performed using the FullTerm TLiIQ
system of Hologic [26
Women with a cervical length below 10 mm and women with a cervical length between 10-30 mm in combination with a positive fibronectin test will be treated with tocolysis. These women are considered to be at high risk for preterm delivery. Women with a cervical length above 30 mm will be treated according to the discretion of the physician. Women with a cervical length between 10-30 mm and a negative fibronectin test are considered to be at low risk, they will enter a double blind trial in which they will be randomised by a central web-based system between tocolytics or placebo (1:1). Details of delivery and maternal assessments during pregnancy of all women are recorded in case record forms (CRF) that are accessible through a closed part of a central website. All data will be collected, coded and processed with adequate precautions to ensure patient confidentiality.
All high risk women will receive tocolytics according to local protocol. Fibronectin negative women with a cervical length between 10-30 mm will be randomly allocated to tocolytics or placebo (figure ). Active treatment will be performed with the calcium channel blocking agent nifedipine with a total daily dose between 80 and 120 milligrams for 48 hours. Active drug and placebo will be administered orally at the same volume and rate. Corticosteroids will be administered at the discretion of the attending physician.
Flow chart APOSTEL-I study, an overview of which patients in the APOSTEL-I cohort will be randomized based on their cervical length and fibronectin status.
Progression of labor will be monitored every 12 hours by means of clinical observation and/or vaginal examinations for 48 hours. If labor seems inevitable in that period, study medication can be stopped and another tocolytic (excluding nifedipine) may be administered.
Follow up of women and infants
Details of admission of newborns to the neonatal intensive care unit and neonatal complications will be recorded. Long term follow up is not part of this study.
The primary outcome measure is number of days to delivery truncated at 7 days after study entry. Secondary endpoints are neonatal mortality, neonatal morbidity, maternal morbidity (side effects of nifedipine), costs and health related quality of life.
The randomised trial in our study will be a non-inferiority effectiveness trial. Based on our pilot study [25
], we anticipate the probability on preterm birth within 7 days in the group of women with a negative fibronectine test and a cervical length 10-30 mm to be 5%. We need 220 fibronectin negative women in the cohort (110 per arm) to assure with 80% power that in case of non-inferiority the upper limit of the 95% one-sided CI for the difference in the proportion of preterm deliveries < 7 days will be within a prespecified boundary of 7.5%. According to our pilot we will need a cohort of approximately 660 women [25
The results of the randomised trial will be analyzed according to the intention to treat principle. We will analyze costs and effects of a strategy based on fibronectin measurement and cervical length, in which only fibronectin positive women will be transferred to a perinatal center and/or and treated. Costs and effects of this strategy will be compared to a strategy in which these women will be transferred and treated merely based on a clinical diagnosis by e.g. vaginal examination and/or cervical length measurement.
After 100 fibronectin negative, low risk women have entered the randomised trial, an interim analysis will be performed by an external independent safety committee. Interim analysis will be performed on the primary outcome only, e.g. percentage of preterm deliveries within 7 days after fibronectin testing. If there is a significant difference between the groups treated with tocolytics and placebo, the independent safety committee will decide about further unblinding of the interim results and (dis)continuation of the trial.
The aim of the economic evaluation is to assess whether fibronectin testing as a triage for women with threatened preterm labor is cost-effective. For the cost analysis we distinguish cost within the antenatal period, during delivery and childbirth and within the postnatal period. As the study will be performed from a societal perspective, three cost categories will be included: direct medical costs, direct non-medical costs and indirect costs. Volumes of health care resource use during the index admission are measured prospectively alongside the clinical study in all participating centers as part of the CRF. Questionnaires addressing health related quality of life will be administered in a random subsample.
If the randomised trial indicates that a particular subgroup of women does not benefit from treatment with tocolytics, the economic analysis will be a cost-minimization analysis. If tocolysis and transferring low risk women would not be necessary, the costs of risk assessment (int.al. fibronectin testing) of the whole cohort of women will be compared to the potential cost saving. Furthermore, scenario analyses for relevant subgroups will be performed. Before analyzing the data from the study, a detailed cost analysis model will be developed.
The trial results will be incorporated in a diagnostic model to compare a risk assessment strategy for women with threatening preterm labor within current practice, which does not include additional risk assessment with fibronectin.
The study design will enable us to compare the costs and effects of the following strategies:
I. transfer and treat all women with threatening preterm labor
II. transfer and treat women based on the results of fibronectin measurement. As we also measure cervical length, we can also evaluate the following strategies:
III. transfer and treat women based on the results of cervical length measurement
IV. transfer and treat women based on a combination of cervical length measurement and fibronectin measurement
V. sequential combinations, for example first cervical length measurement, and fibronectin measurement in a part of that women, or vice versa.
In our cohort, we will obtain data on the moment of delivery in women treated with tocolytics and not treated with tocolytics. In addition to our cohort data, we will obtain data from medical literature on the occurrence of neonatal mortality and morbidity in relation to duration of pregnancy, both with and without treatment of corticosteroids. We will use this to calculate the expected neonatal mortality and morbidity for each of the five strategies mentioned above.
This study has been approved by the ethics committee of the Academic Medical Centre Amsterdam (ref.no MEC 08/363).