Antibiotics in Pregnancy: Are They Safe?

The question of whether to prescribe a course of antibiotics to a pregnant woman is a dilemma faced by obstetrics-gynecology (ob-gyn) care providers on a daily basis. In appropriate circumstances-such as the treatment of asymptomatic bacteriuria to prevent ascending infection and pyelonephritis-related adverse pregnancy outcomes-antibiotic therapy can be both effective and life saving. As with the administration of other medications, the potential benefits need to be weighed against the risk to the fetus. Some antibiotics are known to be teratogenic and should be avoided entirely during pregnancy. These include streptomycin and kanamycin (which may cause hearing loss) and tetracycline (which can lead to weakening, hypoplasia, and discoloration of long bones and teeth). How about other antibiotics? Are they safe? Can they be given with impunity?

A decade ago, a number of well-designed clinical trials^1,2 and systematic reviews³ concluded that broad-spectrum antibiotics can prolong the latency period (interval to delivery) and improve short-term perinatal outcome in pregnancies complicated by preterm premature rupture of membranes (pPROM) prior to 34 weeks of gestation, but not in women with preterm labor and intact membranes. Seven-year follow-up of the fetuses exposed to these antibiotics was recently published.^4,5 Reassuringly, broad-spectrum antibiotics given to fetuses in the setting of pPROM were not associated with any long-term disadvantage, although it is concerning to note that the short-term benefits in perinatal outcome described in the original report¹ did not appear to persist to age 7.⁴ Even more concerning, however, was the observation that fetuses exposed to broad-spectrum antibiotics in the setting of intact membranes were at significantly higher risk of cerebral palsy at age 7 (erythromycin, odds ratio [OR] 1.93; 95% confidence interval [CI], 1.21–3.09; co-amoxiclav, OR 1.69; 95% CI, 1.07–2.67).⁵ The risk was even higher when both antibiotics were given together (4.55% incidence of cerebral palsy compared with 1.97% for co-amoxiclav alone, 2.29% for erythromycin alone, and 1.63% for placebo).⁵ Exposure to co-amoxiclav was also associated with an increased risk of necrotizing enterocolitis.⁵

The mechanism of injury is not clear. The most likely explanation is an antibiotic-mediated suppression of infection and preterm birth, thereby causing the fetus to remain in a hostile proinflammatory intrauterine environment for a longer period of time. However, a direct injurious effect of the antibiotic itself cannot be excluded. Indeed, 1 reason why a significant association between antibiotics and cerebral palsy was observed with intact membranes but not in the setting of pPROM may have to do with the dose and/or duration of antibiotic exposure. Because the vast majority of women in the preterm labor and intact membrane study did not deliver within 48 hours (89.9%) or 7 days (84.6%) of enrollment, their fetuses were more likely to be exposed to the full 10-day course of antibiotic therapy.² In contrast, 30% to 40% of women in the pPROM study delivered within 48 hours and 55% to 60% within 7 days. As such, these fetuses were exposed to antibiotics for a far shorter period of time.¹ An additional adverse effect of the increased use of broad-spectrum antibiotics in the setting of pPROM is an increase in antibiotic resistance, especially erythromycin-resistant Group B β-hemolytic streptococcus (GBS).

The debate about the efficacy and safety of antibiotics in pregnancy must be seen in a larger context. It highlights the philosophical difference between 2 distinct groups of obgyn care providers: those who believe that everything possible should be offered in a given clinical setting in the hope that something will help (also known as the we don’t have all the information we need or the might as well give it, it won’t do any harm group) and those who hold out against popular opinion until there is consistent and compelling scientific evidence that an individual course of action is beneficial and has a favorable risk-to-benefit ratio (the so-called therapeutic nihilists). As protagonists of the latter camp, we offer the following suggestions to ob-gyn care providers faced with the dilemma of whether to prescribe a medication to a pregnant woman:

• Use medications only if absolutely indicated. For antibiotics, this includes treatment of confirmed infection (urinary tract infection, pyelonephritis, appendicitis, cholecystitis, chorioamnionitis), prevention of ascending infection (asymptomatic bacteriuria), and prevention of early-onset neonatal GBS sepsis.

• If possible, avoid initiating therapy during the first trimester. This is the period of fetal structural development and therefore the highest risk for iatrogenic teratogenicity.

• Select a safe medication, which often means an older drug with a proven track record in pregnancy. Certain antibiotics (streptomycin, kanamycin, tetracycline) are best avoided entirely in pregnancy because of their teratogenicity.

• Wherever possible, single-agent therapy is preferred over polypharmacy. Moreover, narrow-spectrum antibiotics are preferred over those with a broad spectrum for the treatment of established infection and intrapartum GBS chemoprophylaxis. The exception is the use of empiric broad-spectrum antibiotics to prolong latency in the setting of pPROM remote from term (discussed above).

• Use the lowest effective dose.

• Discourage the use of over-the-counter drugs, which may interfere with the efficacy and/or metabolism of prescription medications.