Read before using any of it
- This is written for clinicians and assumes the ability to monitor a woman and a fetus, to obtain intravenous access, and to escalate. It is not a self-administration guide.
- It buys time, it does not treat the cause. The purpose is 24 to 48 hours for corticosteroids to work and for transfer to a unit with a neonatal cot.
- Local protocol and obstetric opinion override everything here.
1. Scope
Threatened preterm labour from 24+0 to 33+6 weeks in a woman meeting the entry criteria below, with intact membranes and no evidence of infection, bleeding or fetal compromise — the window in which antenatal corticosteroids are offered, which is what the delay is for.
Between 20+0 and 23+6 weeks tocolysis has no steroid window to protect and the decision is a different one: an individualised decision taken with senior obstetric and neonatal input, not a protocol.
2. Entry assessment
All of the following before the first patch. If any is outside range, GTN is not the next step.
| Parameter | Required | Note |
|---|---|---|
| Contractions | One or two lasting ~30 seconds every 10 minutes | The threshold for treating rather than observing |
| Cervical dilatation | 0–2 cm ideal; not beyond 3 cm | Assess by speculum with fetal fibronectin where available rather than digitally — NICE prefers that route, and it avoids repeated cervical handling. Do not repeat the examination |
| Membranes | Intact | Rupture is a contraindication |
| Maternal blood pressure | Systolic 90–100 or above to start; 100/70 or higher is ideal | GTN is a vasodilator and the entry floor deliberately sits above the stop line of 80/60 — starting at the threshold at which you would stop leaves no room to fall. The single most important entry criterion |
| Maternal pulse | Below 120 at rest | Tachycardia raises the question of infection or blood loss |
| Maternal temperature | Below 37.5 °C | Screening for chorioamnionitis, which is an absolute contraindication |
| Respiratory rate and SpO₂ | SpO₂ above 95% | Count the rate rather than estimating it |
| Fetal heart rate | Reassuring baseline, 110–160 bpm by guideline definition | 110–160 is the guideline normal. A tighter 120–160 may be used as a deliberately conservative band for starting treatment; if so, record it as an entry criterion rather than as abnormality |
| Urinalysis | Normal | Urinary infection is a common and treatable trigger |
Alongside, and not delaying treatment: history and abdominal examination; ultrasound for presentation; fetal fibronectin between 24 and 34 weeks; full blood count and group B streptococcus status to exclude infection.
3. Contraindications
| Absolute | Caution |
|---|---|
| Chorioamnionitis — pyrexia above 37.5 °C with maternal tachycardia and uterine tenderness, with or without offensive discharge Ruptured membranes Antepartum haemorrhage or placenta praevia Nitrate sensitivity Concurrent phosphodiesterase-5 inhibitor — sildenafil, tadalafil: profound hypotension Hypotension Maternal cardiac disease, including obstructive hypertrophic cardiomyopathy and severe aortic or mitral stenosis Raised intracranial pressure Fetal distress on CTG Major congenital fetal anomaly |
Advanced labour Hypertensive disorders of pregnancy Fetal growth restriction Any concern about fetal wellbeing Anaemia Closed-angle glaucoma Migraine, given the headache rate |
4. Treatment
Apply one GTN patch 10 mg/24 h to the anterior abdominal wall, to either side of the navel
Delivering 0.4 mg/hour. This is the site the tocolysis literature has used — published series specify the anterior abdominal wall — and it is what the protocol this page is based on used in practice. Rotate the side on each change, and keep it clear of the area a CTG transducer or an emergency abdominal field would occupy.
On whether the site produces a local effect, because it is a reasonable thing to assume and it is not what happens. Transdermal GTN is a systemic delivery system: the drug crosses the skin into the local capillary bed and reaches the myometrium through the circulation, the same way it would from any site. GTN does have genuine local effects where the target is millimetres deep — a patch over a vein to help cannulation, ointment for an anal fissure or for Raynaud's — but the myometrium sits centimetres below the skin behind fat, fascia and peritoneum even at term, which is well beyond the depth at which topical delivery produces meaningful tissue concentrations. So the abdomen is the right site because it is the site the trials used and it is practical, not because the drug is acting locally. Anywhere with intact, hairless, unscarred skin will deliver the drug; where the manufacturer's labelling for the cardiac indication names chest, upper arm, shoulder or back, that reflects the studied sites for that use rather than a restriction on absorption.
One real caveat about site rather than mechanism: transdermal absorption does vary with skin thickness, hydration and local blood flow, and the stretched abdominal skin of late pregnancy is not a well-characterised absorption surface. That is an argument for watching the response and the blood pressure rather than for moving the patch. A 5 mg/24 h patch (0.2 mg/hour) is the lower-dose option where blood pressure is at the lower end of acceptable. Record the time of application — every subsequent decision is timed from it.
At one hour, add a second patch of the same strength if there has been little or no effect
Reassess contraction frequency, duration and strength, and blood pressure, before adding it.
Maximum 20 mg in 24 hours
Two 10 mg patches.
Maintain the patch for 24 to 48 hours, on clinical judgement
Replace at 24 hours. If contractions have ceased completely, treatment can stop — there is nothing to be gained by continuing against a quiet uterus, and the steroid window is what mattered. Consider discharge if stable and contraction-free.
If contractions have not fully stopped by 48 hours, consider stopping
At that point the corticosteroid course is complete and its benefit secured, and continued tocolysis against persisting contractions is buying progressively less. Reassess the diagnosis rather than escalating by default: concealed abruption and infection both present as tocolysis that will not work.
Run continuously through the first 48 hours — no nitrate-free interval
This is the default. Those 48 hours are what the whole protocol exists to buy, and eight hours off inside them is eight hours without tocolytic cover while the corticosteroid course is still being completed. Replace the patch at 24 hours; do not interrupt it.
The single exception: where continuation beyond 48 hours is already planned, a 16-hours-on, 8-hours-off cycle may be started at 24 hours, because nitrate tolerance develops with continuous exposure and an interval restores responsiveness — and a course intended to run for days will lose more to tolerance than to a single planned gap. That is a deliberate exception and should be recorded as one, with the reason.
Do not escalate to second-line tocolysis before two hours
Applies equally to nifedipine and atosiban.
Given alongside
- Antenatal corticosteroids — the reason for the whole exercise. Betamethasone 12 mg intramuscularly, two doses 24 hours apart, or 12 hours apart if delivery looks imminent. Offer from 24+0 to 33+6 weeks and consider from 34+0 to 35+6. Give the first dose as early as possible: benefit is greatest between 24 hours and 7 days after the second dose. Repeat courses are not routine and should never be given weekly; a single rescue course may be considered in selected cases where the first was given some time earlier and delivery again looks imminent, which remains a senior decision rather than a protocol step.
- Magnesium sulfate for neuroprotection where delivery is expected within 24 hours below 30 weeks, per local protocol.
- Paracetamol for the headache, which should be anticipated rather than reacted to.
- Intravenous access, with fluids available.
5. Monitoring
| When | What |
|---|---|
| Every 15 minutes until stable | Typically up to 3 hours |
| Then hourly for 4 hours | Then by clinical judgement |
| Every set | Blood pressure (must stay above 80/60), pulse (below 120), temperature (below 37.5 °C), SpO₂ (above 95%), respiratory rate, contraction frequency / duration / strength, fetal heart rate, headache score out of 10, and the patch site |
| Continuously if available | CTG |
| Not at all | Repeat cervical examination, unless there is a specific reason |
Chart everything with times. The trend across sets is what identifies both response and deterioration, and a protocol that depends on removing the patch promptly depends on someone noticing.
Stop and remove the patch for
- Blood pressure at or below 80/60, or any symptomatic hypotension.
- Unbearable headache. Score it each time; roughly a third of women get one shortly after application, and it is the commonest reason for discontinuation.
- Rash at the site or elsewhere.
- Any fetal heart rate abnormality, maternal pyrexia, bleeding, or rupture of membranes.
- Established labour despite treatment.
6. Managing complications
- Any adverse effect: remove the patch. Wipe the site, because a depot of drug remains in the skin and absorption continues for a period after removal. Most effects settle within about 30 minutes, and that residual absorption is the reason it is not immediate.
- Hypotension. Remove the patch, position her left lateral — not supine, which compresses the great vessels and worsens it — and raise the legs. Give intravenous isotonic crystalloid: 500 ml over 30 minutes, faster if she is compromised, repeated to a maximum of 2,000 ml with reassessment between each. Escalate if she does not respond to the second bolus.
- Suspected infection. Chorioamnionitis is a contraindication and a reason to stop tocolysis and deliver, not a reason to add an oral antibiotic — it needs intravenous broad-spectrum treatment and senior obstetric input. Separately, group B streptococcus prophylaxis in established preterm labour is intravenous benzylpenicillin per local protocol, and if membranes have ruptured, erythromycin is the agent of choice — co-amoxiclav is specifically avoided in preterm rupture because of the association with neonatal necrotising enterocolitis.
- If she does not respond to two patches within two hours, move to second-line tocolysis and reconsider the diagnosis, in particular concealed abruption and infection.
7. Fetal heart rate interpretation, briefly
| Normal baseline | 110–160 bpm |
|---|---|
| Accelerations | A rise of at least 15 bpm for at least 15 seconds — but 10 bpm for 10 seconds below 32 weeks, which matters in exactly this population. They indicate a well-oxygenated fetus |
| Early decelerations | Mirror the contraction, from head compression. Usually benign |
| Late decelerations | Begin at or after the peak of the contraction; shallow, smooth, and slow to recover. Suggest the fetus is not coping |
| Variable decelerations | Irregular dips, from cord compression. Common; concerning if repetitive |
8. The alternatives, for comparison
| Agent | Regimen | Notes |
|---|---|---|
| Nifedipine NICE first line | 10–20 mg orally, immediate release, repeated at 30 minutes and again at 30 minutes if contractions persist; then 10–20 mg modified release three times daily for 48–72 hours. An alternative regimen gives 10 mg every 15 minutes for the first hour, then 10–20 mg modified release 8-hourly for 48 hours. Bite or chew the first doses to speed absorption | Maximum 120 mg/day, and there is no clear consensus on the optimal regimen. Above 60 mg/day, headache and hypotension rise roughly fourfold. Around 1 in 10 get headache, hypotension, nausea or tachycardia |
| Atosiban | 6.75 mg bolus over 1 minute, then 18 mg/hour for 3 hours, then 6 mg/hour for up to 45 hours. Maximum 330 mg/day | Licensed in the UK and named by NICE where nifedipine is contraindicated. The Cochrane review of oxytocin receptor antagonists found it no better than placebo or betamimetics for the outcomes that matter, with a concern about infant deaths in one trial that had a gestational-age imbalance. Treat it as an option with weak supporting evidence rather than as ineffective |
| GTN | As above | Reported adverse effects against nifedipine, from Kashanian, Zamen and Sheikhansari, J Perinatol 2014;34:683–687, with the direction supported by a 2025 meta-analysis and the NICHD tocolytic synthesis: headache comparable at roughly 28% each; hypotension roughly 25% against 28%; nausea and tachycardia more common with GTN at around 8% against near zero. Overall burden lower and recovery faster on withdrawal, which is the practical argument for it |
Reported prolongation of pregnancy is on the order of 7 days for nifedipine and 4 days for atosiban. Set against the goal — 24 to 48 hours for steroids and transfer — all three agents clear the bar on average, which is why tolerability, route and availability carry as much weight as potency.