Objective:

To update the USPSTF’s previous recommendation statement on Screening for Asymptomatic Bacteriuria in Adults, we systematically reviewed evidence on the benefits and harms of screening for asymptomatic bacteriuria (ASB) and treatment for pregnant women, nonpregnant women, and men.

Data Sources:

MEDLINE, PubMed Publisher-Supplied Records, and the Cochrane Collaboration Central Registry of Controlled Trials for literature published through September 7, 2018.

Study Selection:

Two researchers independently reviewed 4,318 titles and abstracts and 288 full-text articles against prespecified inclusion criteria, then abstracted data from included studies. English-language randomized trials and observational studies were included to assess the direct health benefits and potential harms of screening for ASB. Randomized trials with control conditions of placebo or no treatment were included to evaluate the benefits and harms of ASB treatment, with observational studies also included for assessment of potential treatment harms among pregnant women. Included study populations were asymptomatic community-dwelling adults (ages 18+), not undergoing treatment or specialized care related to surgical or urologic procedures, including catheterization. Pregnant women of any age were also included and studied as a separate population. Due to the historical nature of the evidence, more lenient quality rating of studies was employed to allow for changes in trial reporting standards over time.

Data Analysis:

We synthesized data on the benefits and harms of ASB screening and treatment for general adult populations separately from studies of pregnant women. Health outcomes and harms were sparsely and inconsistently reported in the studies conducted among general adult populations and in studies of screening conducted among pregnant women, precluding meta-analysis. For these outcomes, we described findings in the review text and tables and conducted narrative synthesis. Outcomes for the treatment of screen-detected ASB in pregnancy were analyzed with random effects meta-analysis to calculate the pooled differences when data were sufficient. We examined statistical heterogeneity among the pooled studies using standard χ2 tests and estimated the proportion of total variability in point estimates using the I² statistic. We generated funnel plots and conducted the Egger tests for small-study effects for all pooled analyses that included at least 10 studies. Using established methods, we assessed the strength of evidence for each question.

Results:

We included 19 studies of screening or treatment for ASB reported in 36 publications. Fourteen of the included studies were conducted among pregnant women; two of them examining the effectiveness and/or harms of screening (N=5,289) and 12 examining the effectiveness and harms of treatment (N=2,377). Five included studies examined the effectiveness and harms of treatment among adult men and nonpregnant women (N=777), with most primarily focused on women. Reporting on the characteristics of study participants was sparse in the included literature, and all but one included were judged to be fair quality in risk of bias assessments.

Screening:

Of the two cohort studies on screening in pregnant women, one conducted in Spain (N=4,917) identified a three-fold reduction in risk for pyelonephritis in unadjusted comparisons on a retrospective unscreened and screened cohort. The other cohort study of screening in pregnant women was conducted in Turkey (N=372) and had low statistical power for comparisons of health outcomes in a screened and unscreened cohort due to rarity of outcome events. For health outcomes related to ASB screening in adult men/nonpregnant women, no eligible studies were identified for inclusion in the review.

Treatment:

Twelve trials of ASB treatment among pregnant women (N=2,377) and five trials of ASB treatment among general adult populations (N=777) were included. Screening with culture testing was used in all but one recent included study. Antibiotic treatment was the most common intervention, but the treatment protocols varied considerably across studies. Data from 12 trials provided evidence that treatment of ASB in pregnancy reduces the risk of pyelonephritis (pooled relative risk [RR], 0.24 [95% CI, 0.14 to 0.40], k=12, n=2,068, I2 56.9%). Seven treatment studies reported infant outcomes, demonstrating a reduction in low birthweight (<2500g or or small for gestational age [SGA; weight below the 10th percentile for gestation age]) (pooled RR, 0.64 [95% CI, 0.46 to 0.90], k=7, n=1,522, I2 15.8.6%). Data on potential harms and adverse effects of antibiotic treatment of ASB in pregnancy were sparsely reported in the trials, and power was low for observing rare outcomes. A pooled analysis from five studies reporting congenital malformations was null (pooled RR, 0.44 [95% CI, 0.16 to 1.22], k=5, n=961, I2 0%). Adverse reactions to medications were reported, including vaginitis, diarrhea, rashes, and nausea.

Five trials (N=777) addressed the benefits of treating screen-detected ASB general adult populations, focused on women and older adults. Four trials were conducted only in women, and the fifth trial was primarily among older adult women (84%). Treatment was variable across the trials, ranging from a single dose to 3 months of daily antibiotics. Overall, no study found a difference in mortality, mobility, or rates of symptomatic infections between treated and untreated individuals. Data were inconsistently reported in the four studies reporting harms because they did not report any adverse events or identified few or no patients who withdrew from the study based on adverse events.

Limitations:

This review was limited to English-language evidence, primarily from trials conducted in high and very high HDI countries. Risk of bias was judged to be high or difficult to assess due to limitations in reporting in many of the included studies. Most of the trials among pregnant women were conducted over 40 years ago, many using treatment protocols and scientific methods that are no longer commonly employed.

Conclusions:

In pregnancy, there is some evidence that treatment of urine culture screen-detected ASB confers a benefit to maternal and infant health, but most of the evidence is from an earlier era. We did not find evidence that treatment of ASB in nonpregnant populations is beneficial to health, based on a limited number of trials conducted mainly among older women. Information on harms was limited in the included studies, but established and emerging evidence highlights the importance of antibiotic stewardship to limit the development of antibiotic resistance and rising awareness of potential harms associated with antibiotic exposure, including changes to the microbiome that increasingly are found to have consequences for health.