• Paracetamol is the analgesic and/or antipyretic of choice during pregnancy.
• However, if during the first or second trimester paracetamol is ineffective and an NSAID is clinically indicated:
• Ibuprofen is the preferred NSAID, but should only be used before 30 weeks of pregnancy (off-label use).
• Use the lowest effective dose for the shortest duration possible.
• If regular use is clinically indicated, such as in a woman with rheumatoid arthritis, seek specialist advice.
• NSAIDs must not be used from 30 weeks of pregnancy onwards without specialist advice and regular fetal monitoring.

For more information, contact the UK Teratology Information Service (UKTIS), formerly the National Teratology Information Service (NTIS), on 0844 892 0909.

These recommendations are based on advice from the UK Teratology Information Service (UKTIS), formerly the National Teratology Information Service (NTIS) [NTIS, 2008], and expert opinion from the medical literature [Schaefer et al, 2007].

Use of paracetamol during pregnancy

• Experience suggests that paracetamol can be used at any time during pregnancy [NTIS, 2004; Schaefer et al, 2007].

Occasional use of ibuprofen before 30 weeks of pregnancy

• Most manufactures advise avoiding NSAIDs at any time during pregnancy (off-label use) [BNF 57, 2009].
• NSAIDs are contraindicated from 30 weeks of pregnancy onwards because of the risk of premature closure of the ductus arteriosus, or decreasing amniotic fluid [Schaefer et al, 2007]. If treatment is unavoidable from week 30 onwards, fetal circulation and amniotic fluid levels should be monitored regularly (by ultrasound once or twice a week) and medication should be stopped as soon as signs of ductal constriction appear or if oligohydramnios occurs [Schaefer et al, 2007].
• If occasional use of an NSAID is required during the first two trimesters, ibuprofen is preferred because there is most clinical experience with its use [NTIS, 2008].

Risks of NSAID use before 30 weeks of pregnancy

• Miscarriage
• The UK Teratology Information Service (UKTIS), (formerly the National Teratology Information Service [NTIS]), have reviewed safety data of NSAIDs from published research and postmarketing surveillance systems [NTIS, 2008]. UKTIS concluded that 'an increased risk of spontaneous abortion following maternal exposure to NSAIDs during pregnancy has been suggested, but has not been conclusively proven'. A summary of the UKTIS review is given below:
• One case-control study [Nielsen et al, 2001] found an increased incidence of miscarriage in the group prescribed NSAIDs (63 of 381 women; 16.5%) compared with the control group (4,205 of 33,637 women; 12.5%). However, the overall incidence of miscarriage reported in this study was within the range of the European background incidence of miscarriages in clinically confirmed pregnancies (10-20%). In addition, there were many methodological weaknesses with the way the data were analysed in this study, making the results difficult to interpret.
• A re-analysis of this case-control study [Nielsen et al, 2004] substantially reduced the strength of the miscarriage associated with NSAIDs. The revised data set included 1,599 women, 45 of whom had filled prescriptions for NSAIDs in the last 12 weeks prior to miscarriage. The odds ratio for miscarriage was 0.58 following NSAID exposure at 10-12 weeks prior to miscarriage, increasing to 3.35 following NSAID exposure at one week prior to miscarriage.
• A 'post-hoc' analysis (not a pre-specified endpoint) of a population-based prospective cohort study involving 1,055 pregnant women evaluated exposure to NSAIDs, aspirin and paracetamol (taken after their last menstrual period) and the risk of miscarriage [Li et al, 2003]. Use of paracetamol (n = 172) during pregnancy was not associated with a risk of miscarriage, regardless of the timing or duration of use. The use of NSAIDs (n = 53) was associated with an increased risk of spontaneous abortion (hazard ratio 1.8, 95% CI 1.0 to 3.2). The risk of miscarriage was much higher when NSAIDs were taken around conception (adjusted hazard ratio 5.6, 95% CI 2.3 to 13.7) or were used for longer than a week (hazard ratio 8.1, 95% CI 2.8 to 23.4). However, the risk estimates are from observational studies that, like all observation studies, are susceptible to biases; for example, 'confounding by indication', the bias that arises when a drug is used to treat a condition that causes what seems to be an adverse effect of the drug. Also, these risk estimates are based on small actual numbers of exposures.
• Congenital defects
• The UK Teratology Information Service (UKTIS), (formerly the National Teratology Information Service [NTIS]), have reviewed safety data of NSAIDs from published research and postmarketing surveillance systems [NTIS, 2008]. UKTIS concluded that 'while recent data have suggested that there may be an increased risk of cardiac malformations associated with NSAID use in early pregnancy, this finding has not been confirmed. The available data on NSAIDs generally do not indicate that exposure before 30 weeks of pregnancy is associated with an increased risk of malformations.' A summary of the data included in the UKTIS review is given below:
• One registry-based study [Ofori et al, 2006] observed an increased risk of cardiac septal defects. There were 93 births with congenital malformations among 1,056 women who had filled a prescription for an NSAID in the first trimester compared with 2,478 among 35,331 women who did not use NSAIDs. The odds ratio for cardiac septal defects, adjusted for confounding variables was 3.34 (95% CI 1.87 to 2.85).
• A previous registry-based study [Ericson and Kallen, 2001] found no overall increased risk of malformations associated with NSAID use (OR 1.04, 95% CI 0.84 to 1.29), but did report 36 births with cardiac defects among 2,557 women who reported using an NSAID in the first trimester (OR 1.86, 95% CI 1.32 to 2.62).
• A case-control study of 168 infants with a ventricular septal defect and 692 controls found no link between maternal use of NSAIDs and cardiac septal defects [Cleves et al, 2004].
• A study of 88 pregnant patients, of whom 45 were women with rheumatoid arthritis using regular NSAIDs, found there was no significant effect of NSAID exposure on pregnancy outcome [Ostensen and Ostensen, 1996]. No congenital anomalies were reported among those women who used NSAIDs during pregnancy. NSAIDs were withdrawn by week 36 of pregnancy in all women, and the average duration of NSAID exposure was 15.3 weeks.

Risks of NSAID used from 30 weeks of pregnancy onwards

• Premature closure of the ductus arteriosus
• Premature closure of the ductus arteriosus may lead to tricuspid regurgitation in utero, pulmonary hypertension in the fetus, and persistent pulmonary hypertension in the neonate.
• The risk of ductal closure increases with gestational age, being rare before week 29, but increasing with advancing gestational age to 50–70% at 34 weeks, rising as high as 100% from week 36 onwards [Schaefer et al, 2007].
• Based on 217 women exposed to NSAIDs in the third trimester and 221 to controls, the risk of ductal closure was 15-fold higher in the group of women exposed to NSAIDs (pooled OR 15.04; 95% CI 3.29 to 68.68) [Koren et al, 2006].
• Persistent pulmonary hypertension occurs in about 1–2 newborns per 1,000 live births, causing severe hypoxaemia. The mortality rate is around 10–20% of affected newborns [Silvani and Camporesi, 2007].
• Oligohydramnios
• Oligohydramnios is a condition in pregnancy characterized by a deficiency of amniotic fluid.
• NSAIDs are known to adversely affect renal function, and normal renal function is required to maintain amniotic fluid [Schaefer et al, 2007].