Use and risks of NSAIDs in pregnancy

INTRODUCTION — 

More than 20 different types of nonsteroidal antiinflammatory drugs (NSAIDs) are available in the United States, either over-the-counter or by prescription. NSAIDs may be used by pregnant individuals because of their analgesic, antipyretic, and antiinflammatory properties; however, acetaminophen is generally preferred for use during pregnancy [1]. In addition, low-dose aspirin is used to reduce the risk of developing preeclampsia in individuals at an increased risk for this disorder, and indomethacin may be used as a tocolytic [2] and occasionally for in utero management of congenital anomalies (eg, fetal Ebstein anomaly) [3].

Adverse effects of NSAIDs on the fetus and reproductive system vary based on the type of NSAID, timing of exposure, dose, and duration of use. For example, NSAIDS are associated with transiently impaired fetal and neonatal kidney function (when administered after 20 weeks of gestation) and fetal constriction of the ductus arteriosus (with the greatest risk after 30 weeks of gestation). As with the use of other medications during pregnancy, it is critical to compare the benefits of the medication against the potential risks. For many conditions, however, the relative benefits and risks are difficult to determine because of limited or unavailable data in pregnant individuals.

This topic reviews the various types of NSAIDs, as well as their effects on fertility and a broad range of obstetric and neonatal outcomes. Other issues, including the use of NSAIDs in pregnant individuals with rheumatic disease and nonpregnant individuals, and the adverse effects of NSAIDs are reviewed separately.

●(See "Safety of rheumatic disease medication use during pregnancy and lactation".)

●(See "Overview of COX-2 selective NSAIDs".)

●(See "Nonselective NSAIDs: Overview of adverse effects".)

●(See "NSAIDs: Adverse cardiovascular effects".)

PHARMACOLOGY — 

The mechanism of action of NSAIDs is through the inhibition of cyclooxygenase (COX), also known as prostaglandin synthase (the enzyme involved in the transformation of arachidonic acid to prostaglandins), thromboxanes (potent vasoconstrictors and promoters of platelet aggregation), and prostacyclins (potent vasodilators and inhibitors of platelet aggregation).

●Nonselective COX inhibitors, such as ibuprofen, naproxen, and indomethacin, inhibit both COX-1 and COX-2 isoenzymes. Different nonselective COX inhibitors vary in the amount they inhibit each of the COX isoenzymes.

●Aspirin (acetylsalicylic acid) is a specific type of nonselective COX inhibitor. The mechanism of action differs depending on its dose.

•Low dose – At lower doses (typically 60 to 150 or 162 mg/day), aspirin preferentially inhibits COX-1, resulting in lower platelet thromboxane synthesis, without altering the production of prostacyclin in the vascular wall (see "Preeclampsia: Prevention", section on 'Low-dose aspirin'). Aspirin doses up to 162 mg/day will be considered "low dose" in this topic. (See 'Low-dose aspirin' below.)

•Higher doses – At higher doses (eg, ≥325 mg/day), the ability of aspirin to inhibit both isoforms of COX increases, thus reducing levels of both thromboxane and prostacyclin. These formulations are included in the discussion of other nonselective COX inhibitors. (See 'Other NSAIDs' below.)

●Cyclooxygenase-2 (COX-2) inhibitors, such as celecoxib, selectively inhibit COX-2 isoenzymes.

●Cyclooxygenase-1 (COX-1) inhibitors selectively inhibit COX-1 isoenzymes; only a few are available.

The pharmacology and mechanism of action of NSAIDs are discussed in more detail separately. (See "NSAIDs (including aspirin): Pharmacology and mechanism of action", section on 'Cyclooxygenase inhibition'.)

LIMITATIONS OF AVAILABLE EVIDENCE — 

Evaluating the safety of NSAIDs during pregnancy is challenging because:

●Pregnant individuals are often excluded from initial trials of new medications; thus, data on the use of medications during pregnancy are typically obtained from retrospective studies after the medication is commercially available [4].

●It is difficult to differentiate the risks associated with the medication itself from the indication for its use (eg, fever) and other confounders [5].

●In case-control studies that depend on maternal recall of NSAID use during pregnancy, medications that are used as needed might be less likely to be recalled than those used chronically [6] and patients with pregnancy complications might be more likely to recall medication use than those with uncomplicated pregnancies.

●Many NSAIDs are available over-the-counter, and these medications are often poorly documented or not included in studies based on administrative claims data, a methodology often used to evaluate the risks of medications during pregnancy [7].

●Many studies evaluating NSAID use in pregnancy combine different types of NSAIDs together, which might be necessary because of the infrequent use of certain NSAIDs during pregnancy, but does not allow for the evaluation of individual medications. Different routes (topical versus systemic), doses, duration of use (eg, one time, short duration [eg, up to a few days], long duration [eg, weeks or months], and gestational timing (first, second, or third trimester) might also be combined in certain studies, which can affect the analysis of potential teratogenic effects [8]. NSAIDs might also be included as a component of a medication (eg, cold medication), which also might increase the difficulty in ascertaining exposures.

LOW-DOSE ASPIRIN

Clinical use and safety — The safety of low-dose aspirin in pregnancy is well established, and its use can improve some pregnancy outcomes. In particular, low-dose aspirin (typically 81 to 162 mg daily) initiated between 12 and 28 weeks of gestation (optimally before 16 weeks) reduces the risk of developing preeclampsia in individuals at an increased risk for this disorder. The evidence regarding both the efficacy and safety of using low-dose aspirin for this indication is discussed in detail separately. (See "Preeclampsia: Prevention", section on 'Low-dose aspirin'.)

A reduction in the risk of placenta-mediated complications other than preeclampsia has also been reported, but efficacy data are conflicting. Safety issues would be expected to be the same as in pregnant patients taking low-dose aspirin to reduce the risk of developing preeclampsia. These data and clinical recommendations are reviewed separately.

●Recurrent pregnancy loss (see "Recurrent pregnancy loss: Management", section on 'Aspirin with or without heparin')

●Preterm birth (see "Spontaneous preterm birth: Overview of interventions for risk reduction", section on 'Low-dose aspirin')

●Growth restriction (see "Fetal growth restriction: Pregnancy management and outcome", section on 'Risk reduction in subsequent pregnancies')

●Stillbirth (see "Stillbirth: Incidence, risk factors, etiology, and prevention", section on 'Strategies for prevention of recurrent stillbirth')

●Antiphospholipid syndrome (see "Antiphospholipid syndrome: Obstetric implications and management in pregnancy", section on 'Pregnancy management')

Although long-term low-dose aspirin therapy has been used for the secondary prevention of atherosclerotic cardiovascular disease and colorectal cancer, this use is rare in females of childbearing age. Pregnant patients are typically young and healthy; therefore, safety issues for the rare patient taking low-dose aspirin for these indications during pregnancy would be expected to be low and the same as in pregnant patients taking low-dose aspirin to reduce the risk of developing preeclampsia. An increase in gastrointestinal bleeding has not been reported in pregnant patients taking low-dose aspirin. (See "Aspirin in the primary prevention of cardiovascular disease and cancer" and "Aspirin for the secondary prevention of atherosclerotic cardiovascular disease" and "Preeclampsia: Prevention", section on 'Low-dose aspirin'.)

OTHER NSAIDS

Clinical use — NSAIDs are sometimes offered to pregnant patients in specific clinical situations where the benefits may outweigh the potential risks. The decision should be determined with shared decision-making after discussing the risks and benefits of NSAID use versus no medication or an alternative medication. The decision is often based on individual factors (eg, severity of pain, acceptance of potential fetal risk [which is often, but not always, reversible]).

Examples of specific clinical situations in which NSAIDs may be given during pregnancy include the following:

●Acute pain – For pregnant patients ≤20 weeks with pain (eg, nephrolithiasis, headache, musculoskeletal pain, postoperative/procedural pain) in whom acetaminophen (the preferred analgesic during pregnancy) is ineffective, NSAID use is reasonable. For pregnant patients >20 but ≤28 weeks in whom NSAID use is deemed necessary, the shortest duration (ideally ≤48 hours) and lowest effective dose should be used. For patients >28 weeks, we avoid NSAID use given their associations with adverse outcomes at these gestational ages. When NSAIDs are avoided, opioids may be a reasonable alternative.

●Tocolysis of preterm labor – For pregnant patients with preterm labor before 32 weeks of gestation, NSAIDs may be used short term (up to 48 hours) for tocolysis to allow for administration of a course of antenatal corticosteroids. The threshold of 32 weeks is higher than that used for other indications (eg, pain) as the benefits of tocolysis (delaying the birth of a very preterm infant) are believed to outweigh the potential risks of NSAIDs at this higher gestational age. Indomethacin is the preferred NSAID in this setting, as it is the most effective tocolytic. This is consistent with guidance from the American College of Obstetricians and Gynecologists (ACOG) [9] and discussed in detail separately. (See "Inhibition of acute preterm labor", section on 'Cyclooxygenase inhibitors (eg, indomethacin)'.)

Short-term inhibition of labor can also be useful during interfacility maternal transport. (See "Inter-facility maternal transport".)

●Cerclage placement – Perioperative prophylactic indomethacin may be used for selected patients undergoing transvaginal cerclage because cerclage placement may be painful and has been associated with a transient elevation in prostaglandin levels (prostaglandins induce uterine contractions). This is described in detail separately. (See "Transvaginal cervical cerclage", section on 'Prophylactic antibiotics and tocolytics'.)

●Fetal therapy of congenital heart disease – NSAIDs (indomethacin, ibuprofen) have been used during pregnancy as an experimental approach for in utero management of severe Ebstein anomaly or tricuspid value dysplasia, which are fetal congenital heart anomalies with a high risk of perinatal mortality and where constriction of the ductus arteriosus is beneficial [3]. (See 'Second and third trimester' below.)

●Amniotic fluid reduction in polyhydramnios – Rarely, NSAIDs have been used to manage symptomatic polyhydramnios as maternal administration of prostaglandin synthetase inhibitors reduces amniotic fluid volume. However, the Society for Maternal-Fetal Medicine (SMFM) recommends against indomethacin administration for the sole purpose of decreasing amniotic fluid levels in pregnancies with polyhydramnios, given the risks of postnatal neonatal kidney dysfunction and other neonatal complications [2]. (See "Polyhydramnios: Etiology, diagnosis, and management in singleton gestations", section on 'Indomethacin' and 'Second and third trimester' below.)

●Prior to endoscopic retrograde cholangiopancreatography (ERCP) – In pregnant patients with choledocholithiasis ≤28 weeks, a single dose of a rectal NSAID (indomethacin suppository) can be used at the time of ERCP to reduce the risk of procedure-related pancreatitis. For patients >28 weeks, we avoid NSAID use given its association with adverse outcomes at these gestational ages. Nitrates may be a reasonable alternative for prophylaxis in patients in whom an NSAID is not an option, but maternal hypotension should be avoided. (See 'Safety issues by gestational age at administration' below and "Endoscopic retrograde cholangiopancreatography (ERCP) in pregnancy", section on 'Patient preparation' and "Post-endoscopic retrograde cholangiopancreatography (ERCP) pancreatitis", section on 'Other pharmacologic strategies'.)

●Treatment of rheumatic diseases – Treatment of rheumatic diseases (eg, rheumatoid arthritis, systemic lupus erythematosus, inflammatory bowel disease) during pregnancy is often challenging. When NSAIDs are used in such patients, the duration of use is typically longer than for the other indications listed above. This is discussed separately. (See "Safety of rheumatic disease medication use during pregnancy and lactation" and "Rheumatoid arthritis and pregnancy", section on 'Medication management during pregnancy'.)

Safety issues by gestational age at administration

Prepregnancy

Fertility risks — Prostaglandins play a role in multiple reproductive functions (ovulation, fertilization, tubal function, embryo implantation) and thus may adversely impact fecundability [10-14]. As an example, in a cohort study of 245 females with rheumatoid arthritis trying to conceive, more patients with prepregnancy use of NSAIDs (nonselective and COX-2 inhibitors; length of use not specified) compared with no use had a time-to-pregnancy interval of >12 months (34 versus 18 percent; hazard ratio [HR] for occurrence of pregnancy 0.66, 95% CI 0.46-0.94) [15]. However, data are conflicting, and in a prospective study of over 800 females trying to conceive (and without a history of infertility), NSAIDs were not associated with decreased fecundability [16].

First-trimester use

Risk of pregnancy loss — Use of nonaspirin NSAIDs may increase the risk for pregnancy loss when used around the time of implantation [17,18], but not when used at other times. However, meta-analyses have been limited by the lack of information on the dose, frequency, and chronicity of NSAID use.

In a 2022 meta-analysis of observational studies evaluating the relationship between nonaspirin NSAID exposure and risk of pregnancy loss before 24 weeks of gestation, NSAID use was associated with an increased risk of pregnancy loss in the group with exposure around the time of conception (odds ratio [OR] 2.32, 95% CI 1.16-4.66; 2 studies), but not in the group with first-trimester exposure (OR 1.00, 95% CI  0.83-1.20; 5 studies) or in the overall group with exposure during pregnancy (OR 1.37, 95% CI 0.99-1.88; 10 studies) [19]. Similarly, in a 2024 meta-analysis of 31 observational studies, NSAID exposure during pregnancy was not associated with an increased risk of pregnancy loss before 20 weeks (OR 1.20, 95% CI 0.93-1.55; 11 studies) [20].

Risk of congenital anomalies — NSAID use in the first trimester may result in a small increase in risk of some congenital anomalies (eg, gastroschisis, congenital heart disease), although data are conflicting. Identifying signals of potentially teratogenic medications is challenging because results obtained from animal studies do not reliably predict teratogenic effects in humans, the frequency of any individual anomaly in exposed fetuses is low, the potential for multiple related anomalies needs to be analyzed, adjustment for multiple testing is needed, and the multiple biases involved in data collection, analysis, and publication need to be considered.

●Overall risk – In the general population, the overall risk of a major congenital fetal anomaly (ie, of medical, surgical, or cosmetic significance) is between 2 to 4 percent of live births (the rate is higher in those with pregnancy loss) (see "Congenital anomalies: Epidemiology, types, and patterns", section on 'Epidemiology'). In the 2024 meta-analysis of observational studies discussed above, NSAID exposure during the first trimester was associated with an increased risk of congenital anomalies overall (OR 1.24, 95% CI 1.15-1.34; 31 studies) and major congenital anomalies (OR 1.19, 95% CI 1.08-1.30; 23 studies) [20]. As noted previously, limitations of this meta-analysis include that the dose, frequency, and chronicity of NSAID use were not known.

●Gastroschisis — Data on the risk for gastroschisis after prenatal NSAID exposure are generally reassuring, although the available data are unable to exclude a small increase in risk. Given that the birth prevalence of gastroschisis is approximately 1 in 2000 live births [21], if a small increase is present, the absolute risk would be low and exposure would have to occur before 12 weeks since the abdominal wall closes by this time of pregnancy (ie, physiologic gut herniation is no longer present beyond the 12^th week).

•In a retrospective study using data from the National Birth Defects Prevention Study, mothers of infants with gastroschisis (514 cases) were more likely than control mothers to have used a nonaspirin NSAID in the first-trimester (29 versus 23 percent, adjusted odds ratio [aOR] 1.4, 95% CI 1.1-1.7) [22]. A subsequent study using the same dataset identified an association between first-trimester ibuprofen use and gastroschisis (aOR 1.4, 95% CI 1.2-1.6) [23]. By contrast, in a large retrospective analysis of over-the-counter analgesic use during pregnancy, diclofenac consumption during pregnancy was not associated with increased rates of gastroschisis; the association between ibuprofen, aspirin, or naproxen and gastroschisis could not be assessed because no mothers of infants with gastroschisis had been exposed to these medications [24].

•In a case-control study of maternal medication use for colds, fever, or pain and risk of gastroschisis in newborns (206 cases from 15 cities across the United States and Canada), mothers of infants with gastroschisis were nearly three times as likely to report aspirin use compared with mothers of infants without gastroschisis (adjusted OR 2.7, 95% CI 1.2-5.9) [25]. While the aspirin dose was not ascertained, low-dose aspirin is not used for these indications. Ibuprofen was not associated with an increased risk of gastroschisis. The control group included newborns with major nongastrointestinal structural anomalies and newborns without anomalies admitted to the hospital because of a medical condition.

•Although a 1995 meta-analysis observed an association between first-trimester NSAID use and the composite outcome of exomphalos or gastroschisis (OR 1.52, 95% CI 1.26-1.83, ten studies), these are different anomalies that are etiologically heterogeneous [26].

●Congenital heart disease — Data on the overall risk of congenital heart disease suggest a small increase in risk [27,28]. In the 2024 meta-analysis discussed above, first-trimester NSAID use was associated with an increased risk of congenital heart anomalies (OR 1.20, 95% CI 1.08-1.33, 14 studies) [20].

The heart (including valves, outflow tracts, and septa) is fully developed by 12 weeks of gestation, although the fetal circulation is different from that after birth (eg, patent ductus arteriosus and foramen ovale in utero). Thus, structural anomalies would not be expected to occur from NSAID exposure after 12 weeks, but functional changes are possible. (See 'Risk of premature constriction of the ductus arteriosus' below.)

Studies have reported associations between NSAID use and individual heart anomalies, but a consistent relationship with one anomaly or a specific pattern of anomalies has not emerged. Ventricular septal defects (VSDs) are the most common congenital cardiac anomaly in the general population and are thought to occur as a result of a complex interaction between environmental exposures and genetic susceptibilities. In a population-based case-control study, no significant association between the occurrence of muscular VSDs and maternal use of NSAIDs was observed after controlling for maternal fever, regardless of the exposure window [29].

●Other – In the 2024 meta-analysis of observational studies, NSAID exposure in the first trimester was associated with an increased risk of neural tube defects (OR 1.81, 95% CI 1.22-2.68; seven studies) and hypospadias (OR 1.35, 95% CI 1.08-1.69; five studies) [20]. As the neural tube typically closes during the fifth and sixth weeks of gestation, a defect would not be expected if NSAID exposure occurred after that time. By contrast, the male urethra, foreskin, and penis develop later (into the early second trimester) and thus remain at risk [30]. The risk of cleft lip (with or without cleft palate) also appeared to be higher, but the increase did not reach statistical significance (OR 1.17, 95% CI 0.98-1.41, five studies). As the lip closes by 7 weeks and the palate by 10 weeks of gestation, a defect would not be expected if NSAID exposure occurred after that time.

Second and third trimester — Use of NSAIDs other than low-dose aspirin in the second half of pregnancy is associated with well-documented risks of fetal/neonatal kidney dysfunction and fetal constriction of the ductus arteriosus [2,3]. Other risks have been reported, but their association with NSAID use is more controversial.

Risk of fetal and neonatal kidney dysfunction — An association between fetal exposure to NSAIDs and impaired fetal and neonatal kidney function is well supported in the literature and presumably due to the effects of prostanoid inhibition on the kidney (see "NSAIDs: Acute kidney injury", section on 'Mechanism of acute kidney injury'). The association is most evident after approximately 20 weeks of gestation and increases with advancing gestational age.

In 2020, the US Food and Drug Administration (FDA) issued a warning that use of NSAIDs (other than low-dose aspirin) at or after 20 weeks of gestation could cause fetal kidney dysfunction, leading to oligohydramnios, and in some cases, functional kidney impairment in the newborn [31]. The FDA informed providers that these outcomes usually occur after days to weeks of treatment, although oligohydramnios has been infrequently reported as soon as 48 hours after NSAID initiation. They also noted that oligohydramnios is often, but not always, reversible within 72 hours to one week of NSAID discontinuation. The FDA recommended avoiding NSAIDs at ≥20 weeks of pregnancy when possible, but if use between 20 and 30 weeks of gestation is deemed necessary, it should be limited to the lowest effective dose for the shortest possible duration. In 2023, the Medicines and Healthcare Products Regulatory Agency (MHRA), the drug regulatory authority in the United Kingdom, also posted a safety update warning about prolonged use of NSAIDs after 20 weeks of gestation [32].

Concern about NSAID use at ≥20 weeks is supported by the following lines of evidence:

●A 2020 review of cases of oligohydramnios or neonatal kidney dysfunction associated with NSAID use during pregnancy conducted by the FDA included a literature review of a total of 17 case reports, randomized trials, and observational studies, as well as 35 cases in the FDA Adverse Event Reporting System (FAERS) [31].

In the literature review, oligohydramnios associated with NSAIDs was mostly observed with third-trimester use but was observed as early as 20 weeks of gestation. NSAID duration ranged from 48 hours to multiple weeks. In most cases, oligohydramnios was reversible within 72 hours to 6 days following discontinuation. In many reports, oligohydramnios reappeared after reinitiation of treatment with the same NSAID. In some reports, oligohydramnios did not recur when a different NSAID was used. The FDA did not provide details of the affected pregnancies.

In FAERS, oligohydramnios and neonatal kidney failure were rare: 32 cases of oligohydramnios of which eight also had neonatal kidney dysfunction and three cases of neonatal kidney dysfunction without report of oligohydramnios. Kidney failure was noted in neonates in the first few days of life following 2 days to 11 weeks of in utero exposure to NSAIDs. The neonatal course ranged from resolution of oligohydramnios at three days to persistent anuria requiring dialysis or exchange transfusion. Five newborns died due to neonatal kidney failure. All cases reported a temporal association between NSAID use and oligohydramnios or neonatal kidney dysfunction, with oligohydramnios occurring no earlier than 20 weeks gestation. Definitions/criteria for kidney failure and kidney dysfunction were not provided.

●A 2019 literature review identified many of the same reports of fetuses with oligo- or anhydramnios associated with NSAID exposure, but this analysis was limited to exposures at ≤28 weeks of gestation [33]. In two pregnancies, oligohydramnios was first noted at 21 3/7 weeks and 22 5/7 weeks, respectively, both following long-term exposure to diclofenac of at least 150 mg/day [34]. In five pregnancies, it was associated with indomethacin use at 21 to 26 weeks of gestation for 7 to 28 days; where reported, the dose was 25 mg every 6 to 12 hours [35-37]. The same review also identified five reports of neonatal kidney failure associated with in utero ketoprofen or indomethacin exposure beginning at 24 to 28 weeks of gestation and continuing for two days to several weeks. Several of these cases had other potential reasons for kidney dysfunction (eg, twin-twin transfusion syndrome, fetal growth restriction), so the role of the NSAID is unclear. Definitions/criteria for kidney failure were not provided.

●A 2022 review of the German Embryotox Center of Clinical Teratology and Drug Safety in Pregnancy (Embryotox) database used a cohort design to compare outcomes of pregnancies exposed to NSAIDs and metamizole (a medication that also acts by inhibiting prostaglandin synthesis) in the second and/or third trimester with outcomes following the same exposures only in the first trimester [38]. Headache was the most common treatment indication (38 percent) in second- or third-trimester exposures, followed by back or joint pain (24 percent) and other pain symptoms (21 percent); tocolysis was an uncommon indication (1.5 percent).

Major findings were:

•Oligohydramnios was diagnosed in 8 out of 904 (0.9 percent) second- and/or third-trimester-exposed pregnancies, compared with 2 out of 1154 (0.2 percent) first-trimester-exposed pregnancies (risk ratio [RR] 5.1; 95% CI 1.1-24); the definition of oligohydramnios used in the study was not provided. In at least four of the eight second-trimester exposures, the exposure lasted for several weeks.

Oligohydramnios was diagnosed in 23 out of 1092 (2.1 percent) second- and/or third-trimester-exposed pregnancies, compared with 20 out of 1154 (1.7 percent) first-trimester-exposed pregnancies.

When cases in which trimester of exposure could not be clearly assigned, oligohydramnios was diagnosed in 41 out of 1092 (3.8 percent) second- and/or third-trimester-exposed pregnancies, compared with 29 out of 1154 (2.3 percent) first-trimester-exposed pregnancies (RR 1.5, 95% CI 0.9-2.4).

•The frequency of a neonatal kidney disorder was 1 out of 1098 (0.1 percent) for second- and/or third-trimester-exposed pregnancies and 1 out of 1033 (0.1 percent) first-trimester exposed pregnancies (RR 1, 95% CI 0.1-16.5). One child in each cohort had unilateral kidney hypoplasia.

A limitation of the analysis is the inclusion of exposure to metamizole, a medication that was banned by the FDA in the late 1970s due to a risk of agranulocytosis [24]. Although the mechanism of action of metamizole is not fully understood, it is believed to work similarly to NSAIDs by inhibiting prostaglandin synthesis.

●A proof of concept study in 15 pregnancies with fetal Ebstein anomaly or tricuspid valve dysplasia showed that NSAIDs can mitigate circular shunt physiology by ductus arteriosus constriction and improve pulmonary regurgitation [3]. Therapy began at 22 to 33 weeks of gestation, and serial fetal echocardiography and obstetric ultrasounds to monitor for ductus arteriosus constriction and oligohydramnios, respectively, were performed. Ten of the 15 fetuses developed oligohydramnios 3 to 42 days of therapy after NSAID therapy was started, resulting in a change of therapy from indomethacin to ibuprofen or brief cessation of therapy and return of normal amniotic fluid volume. Two neonates developed kidney failure requiring peritoneal dialysis; one had 24 days of in utero NSAID exposure and the other had 22 days.

Risk of premature constriction of the ductus arteriosus — An association between fetal exposure to NSAIDs other than low-dose aspirin and constriction of the ductus arteriosus is also supported by evidence but appears to be reversible in the setting of short-term use. Fetal constriction of the ductus arteriosus is clinically important because it has been associated with persistent pulmonary hypertension and other adverse outcomes in newborns [39]. Most cases are associated with oral NSAIDs, particularly indomethacin, but ductus arteriosus constriction has been reported after exposure to a topical formulation of diclofenac (single case) and following exposure to a lozenge form of benzydamine (three cases) [40-42]. The proposed mechanism is prostanoid inhibition on the ductus (prostaglandins play a role in the relaxation of smooth muscle cells in the ductus arteriosus and thus help to keep it open).

Concern about NSAID use and fetal ductal constriction is supported by the following lines of evidence:

●In a 2006 meta-analysis of randomized trials evaluating ductal outcomes in 217 patients exposed to indomethacin (duration of exposure was 48 to 72 hours in six trials, four days in one trial, and up to 24 days in one trial) between 23 and 35 weeks of gestation versus 221 patients exposed to a placebo or non-NSAID medication at the same gestational age, NSAID exposure was associated with a 15-fold increase in risk of premature closure of the ductus arteriosus (22 events in 217 treated patients versus 0 events in 221 controls; pooled OR 15.04, 95% CI 3.29-68.68; eight trials) [43]. The risk associated with indomethacin exposure versus other NSAIDs was not statistically significant (OR 2.12; 95% CI 0.48-9.25; four trials).

●A 2019 literature review found 33 reported cases of narrowing or closure of the ductus arteriosus at ≤28 weeks of gestation [33]. The earliest recorded premature closure in this review was at 25 weeks and occurred in five fetuses. Three cases involved several weeks of ibuprofen exposure, a few days of indomethacin exposure, or a single dose of indomethacin six hours before mild ductal constriction was diagnosed, and two cases followed an unknown length of indomethacin treatment. The review identified one twin born at 27 weeks of gestation with a diagnosis of primary pulmonary hypertension after two doses of indomethacin within 16 hours before birth. However, most adverse events occurred after NSAID exposures lasting at least seven days and ductus arteriosus constriction usually normalized after NSAIDs were discontinued [33].

Subsequently, a case report described constriction of the ductus arteriosus three days after exposure to ibuprofen and dexamethasone at 18 weeks of gestation [44].

●In a 2022 review of the Embryotox Center of Clinical Teratology and Drug Safety in Pregnancy (Embryotox) database cohort study comparing outcomes of pregnancies exposed to NSAIDs or metamizole in the second or third trimester with outcomes following the same exposures in the first trimester, ductus arteriosus constriction was noted in 5 out of 1092 (0.5 percent) second- or third-trimester-exposed pregnancies versus 0 out of 1154 first-trimester-exposed pregnancies [38]. These events included one pregnancy with late-second-trimester exposure after long-term NSAID exposure, two pregnancies with third-trimester exposure, and two pregnancies where timing was not clearly assigned. No differences were seen in the rates of primary pulmonary hypertension between the two groups.

●A study that used fetal echocardiography to follow 72 fetuses exposed to continuous indomethacin administration for >24 hours for tocolysis at 21 to 35 weeks of gestation (mean 28.9 ± 3.2 weeks) reported ductal constriction in 35 fetuses, of whom 18 also had tricuspid regurgitation [45]. Echocardiography was performed within 48 hours after initiation of therapy and weekly thereafter. Ductal constriction was detected 1 to 27 days (mean 5 ± 6 days) after initiating therapy. Affected fetuses were between 24 0/7 and 35 0/7 weeks (mean gestational age at detection was 30.9 ± 2.3 weeks); the highest incidence of ductal constriction was in fetuses at 31 to 32 weeks of gestation. After indomethacin was discontinued, all follow-up echocardiograms showed a return of normal ductal flow velocities. Of note, both fetuses were affected in five of the seven twin gestations and one fetus was affected among the two sets of triplet gestations.

●In 15 pregnancies with fetal Ebstein anomaly or tricuspid valve dysplasia treated with NSAIDs beginning at 22 to 33 weeks of gestation, constriction of the ductus arteriosus occurred in 12 of 15 fetuses, and all 12 had improvement in pulmonary regurgitation [3].

Risk of adverse neurodevelopmental outcomes — Studies examining the effects of NSAID exposure during pregnancy and neurobehavioral outcomes have reported reassuring findings but are limited.

●In a 2019 systematic review evaluating the effects of analgesic medications on neurodevelopmental outcomes in children [46]:

•Three studies examined the effects of exposure to indomethacin for tocolysis of preterm labor. In the first study, 15 preterm infants exposed prenatally to indomethacin compared with infants born to mothers who did not receive indomethacin had similar neurodevelopment assessments at 6 to 12 months [47]. In the second study, 93 infants exposed to either indomethacin or nylidrin (a beta-agonist resulting in peripheral vasodilation) in utero had similar neurologic assessments at 18 months [48]. In the third study, 29 infants exposed to indomethacin in utero had similar neurodevelopment assessments compared with controls after adjustment for confounders (eg, use of antenatal corticosteroids) [49].

•Two studies evaluated the effects of NSAIDs for uses other than tocolysis. In a study using the Norwegian Mother and Child Cohort Study and including 155 pairs of same-sex siblings, children exposed to ibuprofen in utero compared with those not exposed had similar developmental outcomes (ie, gross motor, fine motor, communication, externalizing behavior, internalizing behavior problems) [50]. In the second study, which evaluated the effect of prenatal NSAID use (based on prescriptions and maternal self-report) on developmental outcomes for problems with attention, exposed children e had increased attention problems based on maternal reports, but not based on teacher reports, compared with unexposed children; intelligence quotient (IQ) scores were similar between groups [51].

●Subsequent studies evaluating NSAID use in pregnancy on neurodevelopmental outcomes have reported mixed findings:

•In a study of mother-infant pairs evaluating the effects of prenatal ibuprofen and acetaminophen use in early and mid-pregnancy on mid-childhood executive function and behavior, children exposed prenatally to both ibuprofen and acetaminophen had higher parent-related and teacher-related scores (indicating greater problems); these observations could also be due to confounding by indication (ie, reason for NSAID use [eg, fever]) [52].

•In a study examining the association between NSAID exposure during pregnancy and attention deficit hyperactivity disorder (ADHD) using data from the Norwegian Mother, Father and Child Cohort, children with prenatal NSAID exposure were not at an increased risk of an ADHD diagnosis or ADHD symptoms, regardless of duration or timing (first, second, or third trimester) of exposure [53].

•In a study evaluating the relation between prescriptions for NSAIDs during pregnancy and diagnosis of autism spectrum disorder or intellectual disability using Medicaid-insured mother-child dyads, prescriptions for NSAIDs during pregnancy were not associated with autism spectrum disorder; however, there was an association between NSAID prescriptions and an increased risk for intellectual disability (aOR 1.26, 95% CI 1.1-1.46) [54]. Because exposure to NSAIDs was based on prescription fills, nonprescription NSAIDS would be missed, resulting in a potential for exposure misclassification.

Risk of other adverse outcomes — The association between NSAID use later in pregnancy and other adverse outcomes (eg, low birth weight, asthma, preterm birth) has been less well studied.

●In one of the largest cohort studies, which included over 6500 patients with NSAID exposure and nearly 84,000 without NSAID exposure in pregnancy [27]:

•Second-trimester use of ibuprofen or diclofenac was associated with low birth weight (aOR 1.7, 95% CI 1.3-2.3 and aOR 3.1; 95% CI 1.1-9, respectively).

•Third-trimester use of diclofenac was associated with vaginal bleeding (aOR 1.8, 95% CI 1.1-3).

•Second- or third-trimester use of ibuprofen was associated with asthma in offspring at 18 months (aOR 1.5, 95% CI 1.2-1.9 and aOR 1.5, 95% CI 1.1-2.1, respectively).

No significant associations were observed for several other outcomes (eg, preterm birth, intracranial hemorrhage, intraventricular hemorrhage).

Some of the observed associations might have been due to chance, given the multiple comparisons performed in this study.

●Data from the Collaborative Perinatal Project, a large prospective cohort study in the United States, showed an association between maternal use of regular-dose aspirin and asthma by 7 years of age among children without a maternal history of asthma [55]. The aOR for the association was not significantly elevated in the first or second trimesters but was increased in the third trimester (aOR 1.4; 95% CI 1.1-1.6).

●In a population-based cohort study, exposure to NSAIDs in the three months prior to delivery was not associated with preterm birth (aOR 1.18, 95% CI 0.84-1.66), after adjustment for maternal autoimmune diseases, concomitant medication use, and other risk factors [56]. However, an association was observed between use of COX-2 inhibitors in the three months prior to delivery and risk for preterm birth (adjusted OR 2.46, 95% CI 1.28-4.72).

MONITORING AFTER USE — 

When NSAIDs are administered, we take the following approach to monitoring.

●Fetal ultrasound – For pregnancies ≥20 weeks of gestation exposed to an NSAID for >48 hours, we use ultrasound to monitor amniotic fluid volume and assess for signs of ductal narrowing; we do not perform monitoring if the NSAID was used for ≤48 hours. Data to assess the frequency of surveillance are not available; however, performing an ultrasound examination at least weekly while on therapy and an additional ultrasound one week after NSAID discontinuation to assess for delayed effects is reasonable. This is consistent with recommendations from the US Food and Drug Administration (FDA) and the Medicines and Healthcare products Regulatory Agency (MHRA) in the United Kingdom [31,32]. (See 'Risk of fetal and neonatal kidney dysfunction' above and 'Risk of premature constriction of the ductus arteriosus' above.)

●Management of oligohydramnios or ductal narrowing – If oligohydramnios or signs of premature closure of the ductus arteriosus are identified on ultrasound, the NSAID should be discontinued (except in the rare cases where the goal was one of these outcomes). Amniotic fluid levels are expected to return to normal within three to six days. For those pregnancies where oligohydramnios persists, once- or twice-weekly nonstress tests and/or biophysical profile scoring are appropriate until the return of normal amniotic fluid levels. This is consistent with American College of Obstetricians and Gynecologists (ACOG) recommendations for fetal surveillance of pregnancies complicated by isolated oligohydramnios [57]. Management of oligohydramnios is discussed in detail separately. (See "Oligohydramnios: Etiology, diagnosis, and management in singleton gestations", section on 'Prenatal care'.)

LACTATION — 

The use of aspirin and nonaspirin NSAIDS in lactation are discussed separately. (See "Overview of the postpartum period: Normal physiology and routine maternal care", section on 'Safety of common analgesics in breastfeeding individuals'.)

SOCIETY GUIDELINE LINKS — 

Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: General prenatal care".)

INFORMATION FOR PATIENTS — 

UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●(See "Patient education: Taking medicines during pregnancy (The Basics)".)

SUMMARY AND RECOMMENDATIONS

●Clinical significance – Nonsteroidal antiinflammatory drugs (NSAIDs) include aspirin, ibuprofen, and approximately 20 other types of over-the-counter or prescription medications. The effects of NSAIDs on the fetus and reproductive system vary based on the type of NSAID, gestational age at exposure, dose, and duration of use; the relative benefits and risks are difficult to determine because of limited or unavailable data in pregnant individuals. (See 'Introduction' above and 'Limitations of available evidence' above.)

●Low-dose aspirin – Low-dose aspirin has an acceptable safety profile across pregnancy. It is commonly used to improve some pregnancy outcomes (eg, reduce the chance of developing preeclampsia), and less commonly used for nonobstetric indications. An increase in gastrointestinal bleeding has not been reported in pregnant patients taking low-dose aspirin. (See 'Low-dose aspirin' above.)

●Other NSAIDS

•Clinical use – NSAIDs are sometimes offered to pregnant patients in specific clinical situations where the benefits may outweigh the potential risks, such as (see 'Clinical use' above):

-Acute pain – For pregnant patients ≤20 weeks with pain (eg, nephrolithiasis, headache, musculoskeletal pain, post-operative/procedural pain) in whom acetaminophen (the preferred analgesic during pregnancy) is ineffective, NSAID use is reasonable. For patients >20 but ≤28 weeks with acute pain, short-term use (≤48 hours) of an NSAID is a reasonable option; the shortest duration and lowest effective dose should be used. The decision should be determined with shared decision-making after discussing the risks and benefits of NSAID use versus no analgesic or an opioid and is often based on individual factors (eg, severity of pain, acceptance of potential fetal risk [which is often, but not always, reversible]). For patients >28 weeks with acute pain in whom acetaminophen (the preferred analgesic during pregnancy) is ineffective and analgesia is desired, we offer an opioid rather than an NSAID.

-Preterm labor – For patients with preterm labor before 32 weeks of gestation, NSAIDs may be used short term (≤48 hours) for tocolysis to allow for administration of antenatal corticosteroids and/or interfacility maternal transport. The threshold of 32 weeks is higher than that used for other indications (eg, pain) as indomethacin is probably the most effective tocolytic and the benefits of tocolysis are believed to outweigh the potential risks of NSAIDs at this higher gestational age. (See "Inhibition of acute preterm labor", section on 'Cyclooxygenase inhibitors (eg, indomethacin)'.)

-Cerclage placement – Perioperative use prior to placement of a transvaginal cerclage. (See "Transvaginal cervical cerclage", section on 'Prophylactic antibiotics and tocolytics'.)

-Rheumatic diseases – Management of rheumatic diseases (eg, rheumatoid arthritis, systemic lupus erythematosus, inflammatory bowel disease) in pregnancy is discussed in detail separately. (See "Safety of rheumatic disease medication use during pregnancy and lactation".)

•Safety

-Prepregnancy – While the effects of NSAIDs on fertility are unclear, prepregnancy use of NSAIDs may reduce fertility by preventing or delaying ovulation. (See 'Prepregnancy' above.)

-First trimester – Prenatal use of NSAIDs may increase the risk for pregnancy loss when used around the time of implantation. They may also result in a small increase in risk of some congenital anomalies (eg, gastroschisis, congenital heart disease), although data are conflicting. (See 'First-trimester use' above.)

-Second and third trimester – Use of NSAIDs other than low-dose aspirin in the second and third trimesters is associated with fetal/neonatal kidney dysfunction and fetal constriction of the ductus arteriosus; the risk increases with advancing gestational age and with use for >48 to 72 hours. The benefits of taking an NSAID at ≥20 weeks gestation (eg, inhibition of acute preterm labor) should be carefully weighed against these risks. The association between NSAID use later in pregnancy and other outcomes (eg, low birthweight, preterm birth, childhood asthma) has not been well studied. (See 'Second and third trimester' above.)

-Neurodevelopmental outcomes – The effects of NSAIDs on neurodevelopmental outcomes (eg, attention deficit hyperactivity disorder, autism spectrum disorder, intelligence quotient [IQ] score) are reassuring, but limited. (See 'Risk of adverse neurodevelopmental outcomes' above.)

●Monitoring after use – If an NSAID is taken at ≥20 weeks of gestation and used for >48 hours, we suggest using ultrasound to monitor amniotic fluid volume and assess for signs of narrowing of the ductus arteriosus. While data to assess the frequency of surveillance are not available, we perform an ultrasound examination at least weekly while on therapy and an additional ultrasound one week after NSAID discontinuation to assess for delayed effects. If oligohydramnios or signs of premature closure of the ductus arteriosus are identified (and undesired), the NSAID should be discontinued. (See 'Monitoring after use' above.)

●Lactation – The use of aspirin and nonaspirin NSAIDs in lactation is discussed separately. (See "Overview of the postpartum period: Normal physiology and routine maternal care", section on 'Safety of common analgesics in breastfeeding individuals'.)