Kazemier BM, Koningstein FN, Schneeberger C, et alAsymptomatic Bacteriuria in Pregnancy
Lancet Infect Dis. 2015;15:1324-1333
Antibiotics are frequently prescribed even when they are not indicated. A common area of antibiotic misuse is the treatment of asymptomatic bacteriuria (ASB). ASB is typically harmless, but treating ASB can increase rates of antibiotic resistance.Pregnancy, however, is felt to be one of the few circumstances when treating ASB is beneficial—a recommendation supported by randomized controlled trials (RCTs) from the 1960s and 1970s.
Not all obstetricians, however, screen and treat for ASB. In The Netherlands, for example, prenatal screening for ASB is not the standard of care. Kazemier and colleagues seized this unique opportunity to study the natural history of untreated ASB. They designed a multicenter prospective cohort study with an embedded RCTs to determine whether screen-and-treat strategies for ASB during pregnancy prevent pyelonephritis and preterm births.
For this study, ASB screening was offered to pregnant women aged ≥ 18 years who were between 16 and 22 weeks of gestation at 13 clinical sites in The Netherlands, representing different levels of obstetric care and patient socioeconomic status. Several exclusion criteria ensured that eligible women would all be at low risk for obstetric and infectious complications. Women who tested positive for ASB and agreed to enroll were randomly assigned to receive a 5-day course of nitrofurantoin or placebo.
More than 5000 pregnant women were eligible, and 250 (5%) tested positive for ASB. Only 85 of the ASB-positive women agreed to enroll in the RCT. Despite these low numbers, the investigators were able to define outcomes in 4283 women by electronically extracting data from a national perinatal registry. In the overall cohort, pyelonephritis occurred in 2.4% of untreated ASB-positive women and 0.6% of ASB-negative women. Although this was a statistically significant difference, the overall risk for pyelonephritis was low.
There was no difference in preterm delivery between the two groups (ASB-positive and untreated, 1.0%; ASB-negative, 1.3%). No differences in outcomes were found between women with treated ASB and those with untreated ASB.
These findings contradict more than a dozen older RCTs, so skeptics may be quick to highlight this study's limitations. The obvious weakness is the small number of women who underwent randomization. The frequency of the primary outcome was drastically lower than anticipated, so the trial was underpowered to identify a significant difference. On the other hand, because the incidence of pyelonephritis and preterm labor is low, preemptive strategies may not be warranted.
Skeptics may also argue that untreated ASB-positive women could have developed symptomatic urinary tract infections that were undetected by the investigators because they were treated as outpatients and were never hospitalized. Although women enrolled in the RCT self-reported their outpatient antibiotic use, there was no comparable monitoring mechanism for the other patients in the cohort. Even if cases such as this were missed, it's noteworthy that not treating ASB at the time of screening did not lead to worse perinatal outcomes in either the cohort study or the RCT.
Physicians long for RCTs to guide their decision-making. A Cochrane review from 2015, published before the release of this trial, reviewed 14 RCTs and concluded that treating ASB in pregnancy is beneficial, on the basis of "very low-quality evidence." Now that a modern trial has reassessed the issue, will it change practice?