Priority Updates from the Research Literature from the Family Physicians Inquiries Network

Jean Moon, PharmD; Shailendra Prasad, MBBS, MPH; Mari Egan, MD, Treating UTIs in reproductive-age MHPE North Memorial Family women—Proceed with caution Medicine Residency, University of Minnesota, Minneapolis (Drs. Moon A new study suggesting that nitrofurantoin and and Prasad); Department of Family Medicine, sulfonamides are teratogenic highlights the need for a The University of Chicago (Dr. Egan) risk-benefit analysis when your patient is a woman of

PURL s E d i t o r reproductive age. Bernard Ewigman, MD, MSPH Department of Family Medicine, The University of Chicago Practice Changer erichia coli contributing to roughly 90% of Nitrofurantoin and sulfonamides may cause the infections.2 Screening for and treating major birth defects and should be used with asymptomatic bacteriuria is also recom- caution—if at all—in women of reproductive mended during pregnancy to prevent pyelo- age.1 nephritis and increased maternal and fetal morbidity.3 In addition, UTIs are common in strength of recommendation reproductive-age women who may not know B: Population-based case-control study they are pregnant or who become pregnant during treatment with antibiotics. And nitro- Crider KS, Cleves MA, Reefhuis J, et al. Antibacterial medication use during pregnancy and risk of birth defects: National Birth Defects furantoin and sulfonamides are commonly Prevention Study. Arch Pediatr Adolesc Med. 2009;163:978-985. prescribed antibiotics for the treatment of UTIs, both in pregnant women and women of reproductive age. Illustrative Case A 24-year-old woman comes to your clinic be- Prior warnings only address cause of frequent urination for the past 2 to near-term pregnancy 3 days. She is not taking any medication, but Prior to the study detailed in this PURL, no does take a daily prenatal vitamin because she clinical trials had reported a teratogenic risk and her husband are trying to conceive. After associated with either nitrofurantoin (current your examination, you order a urinalysis and pregnancy category B) or sulfonamide (cur- perform a urine pregnancy test. The urinaly- rent pregnancy category C).4 It is recommend- sis shows bacteriuria ≥100,000 cfu/mL, and the ed, however, that both of these antibacterials pregnancy test is positive. be avoided in pregnant women who are near What will you prescribe to treat her uri- term because of the risk of hemolytic disease nary tract infection? in patients with glucose-6-phosphate dehy- drogenase deficiency associated with nitrofu- ntibacterial agents are among the rantoin and the risk of kernicterus in neonates most commonly used medications exposed to sulfamethoxazole.5 A during pregnancy because treat- But a rise in E coli resistance to penicil- ment of infections is critical to both maternal lins (resistance to amoxicillin, for example, can and fetal well-being.1 Urinary tract infec- be as high as 30-40%6) has led to greater use of tions (UTIs) are the most common medical nitrofurantoin. The drug has been viewed as a complication during pregnancy, with Esch- safe and effective alternative treatment for UTIs

220 The Journal of Family Practice | APRIL 2010 | Vol 59, No 4 associated with E coli.7 Indeed, ni- and periconceptional smoking trofurantoin has been considered and/or alcohol use. to be the preferred antibiotic for z Nitrofurantoin was as- bacteriuria suppression, as both sociated with anophthalmia or ampicillin and cephalosporins microphthalmos (adjusted odds can interfere with the normal gas- ratio [AOR]= 3.7; 95% confidence trointestinal flora. Thus, nitrofurantoin is used interval [CI], 1.1-12.2), hypoplastic left heart extensively in pregnant women. Sulfonamides syndrome (AOR=4.2; 95% CI, 1.9-9.1), atrial are also prescribed for pregnant women, al- septal defects (AOR=1.9; 95% CI, 1.1-3.4), and though not as frequently.7,8 cleft lip with cleft palate (AOR=2.1; 95% CI, 1.2-3.9). z Sulfonamides were associated with Study Summary anencephaly (AOR=3.4; 95% CI, 1.3-8.8), hy- First trimester use linked poplastic left heart syndrome (AOR=3.2; 95% to many defects CI, 1.3-7.6), coarctation of the aorta (AOR=2.7; The study by Crider et al1 was based on the 95% CI, 1.3-5.6), choanal atresia (AOR=8.0; National Birth Defects Prevention Study, an 95% CI, 2.7-23.5), transverse limb deficiency ongoing, population-based case control study (AOR=2.5; 95% CI, 1.0-5.9), and diaphragmatic of an estimated annual birth population of hernia (AOR=2.4; 95% CI, 1.1-5.4). roughly 482,000, including cases identified Some links between other antibiotics and by birth defects surveillance registries in 10 birth defects were also found. For example, eryth- states.9 The researchers identified pregnan- romycins were associated with anencephaly and cies affected by any of 30 types of birth defects transverse limb deficiency, penicillins with inter- from 1997 to 2003 (n=13,155). The controls calary limb deficiency, and cephalosporins with (n=4941) were randomly selected from simi- atrial septal defects. The authors noted, however, Instant lar geographic locations, and matched for that these agents, which are commonly pre- Poll Question race/ethnicity, age, and prepregnancy body scribed for pregnant women, were not associ- mass index. Exposure to antibacterials from ated with many birth defects—and that because 1 month prepregnancy through the end of the of limited sample sizes for these drug classes, the Do you prescribe first trimester was recorded. associations may be spurious. nitrofurantoin for Crider et al interviewed all the participants UTIs in women of up to 24 months after delivery to obtain their reproductive age? What’s New exposure history to penicillins, erythromycins, ❑ Frequently; it is safe nitrofurantoin, sulfonamides, cephalosporins, A large-scale study and effective for quinolones, tetracyclines, other miscellaneous provides evidence of risk infections associated beta-lactams, aminoglycosides, antimycobac- Previous case studies and meta-analysis have with E coli terial agents, and other antibiotics. (Exposure shown no link between the use of nitrofuran- ❑ Only after ruling to antivirals, antifungals, and antiparasitic toin and congenital abnormalities.8 Similarly, out pregnancy agents was not addressed.) Women who didn’t sulfonamides have not appeared to pose sig- know whether they had been exposed to these nificant teratogenic risk. This is the first large- ❑ For most women, except those who agents or could not remember the timing of ex- scale study evaluating the risk of birth defects are pregnant and posure were excluded. associated with antibiotic use during preg- near term Overall, antibacterial use ranged from nancy, and therefore provides evidence of risk 2% to 5.8%, and peaked in the third month of not previously available. ❑ Infrequently; pregnancy. Penicillins were the most com- I usually choose a cephalosporin monly used antibiotics. Odds ratios obtained instead for birth defects were adjusted for confound- Caveats ers such as maternal age, race, education lev- Study design raises ❑ Other ______el, prepregnancy body mass index, time from questions of recall bias estimated date of delivery to the interview, use The retrospective case-control methodology Go to jfponline.com of folic acid or multivitamins from 1 month used in this study leaves open the possibil- and take our instant poll prior to pregnancy through the first month, ity of recall bias, misclassification bias, and

jfponline.com Vol 59, No 4 | APRIL 2010 | The Journal of Family Practice 221 confounding bias. The length of time from birth defect such as those reported in this study actual exposure to data collection could affect is still a rare risk. There may be clinical situa- the accuracy of participants’ memories. The tions in which the benefits of using nitrofuran- data gathered were not cross-verified against toin or sulfonamides in women who are or may medical records, and other issues, such as the become pregnant outweigh the potential risks. possible effect of medications for other infec- tions (eg, antivirals and antifungals), could not be measured. However, women who did Challenges to Implementation not know or were unsure of their medica- Finding an alternative treatment tion exposure history were excluded from the The main challenge to implementing this analysis, which should reduce the risk of this new recommendation lies in choosing alter- potential bias. native antibiotics with which to treat UTIs in The investigators also controlled for sev- reproductive-age women and bacteriuria in eral important sources of potential confound- pregnancy. Obtaining a pregnancy test in sex- ing bias, and the reporting rates were similar ually active patients of reproductive age who among participants in both the case and con- are not using a reliable form of contraception trol groups. These measures provide some as- seems like a prudent first step. surance that the outcomes are valid. If the pregnancy test is positive, cephalexin It would be unethical (and extraordinarily should be a good initial choice until the results If the pregnancy expensive) to conduct a prospective random- of culture and sensitivities are available. In test is ized controlled trial to confirm these findings. the event of Enterococcus infection (for which positive, Case-control methodology is the most practical cephalosporins are not active) or other organ- cephalexin way to assess for the risk of birth defects, and isms resistant to cephalosporins, the sensitivity should be a our literature review suggests that this is the results should provide guidance.3 JFP good initial most rigorous study to date. In our view, the choice until potential harm from continuing to use these Acknowledgement the results antibiotics for pregnant women and women The PURLs Surveillance System is supported in part by Grant Number UL1RR024999 from the National Center for Research of culture and who may become pregnant far outweighs the Resources; the grant was a Clinical Translational Science sensitivities risk that these findings may be erroneous. Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily repre- are available. That said, a final caveat is the fact that even sent the official views of the National Center for Research a several-fold increase in the risk of a rare major Resources or the National Institutes of Health.

References 1. Crider KS, Cleves MA, Reefhuis J, et al. Antibacterial medication use 6. ABXguide. Urinary tract infections in pregnancy. Avail- during pregnancy and risk of birth defects: National Birth Defects able at: http://prod.hopkins-abxguide.org/diagnosis/ Prevention Study. Arch Pediatr Adolesc Med. 2009;163:978-985. genitourinary/urinary_tract_infections_in_pregnancy. 2. Gilstrap LC 3rd, Ramin SM. Urinary tract infections during preg- html?contentInstanceId=255490. Accessed February 15, 2010. nancy. Obstet Gynecol. 2001;28:581-591. 7. Huang ES, Stafford RS. National patterns in the treatment of uri- 3. Macejko AM, Schaeffer AJ. Asymptomatic bacteriuria and symp- nary tract infections in women by ambulatory care physicians. tomatic urinary tract infections during pregnancy. Urol Clin Arch Intern Med. 2002;162:41-47. North Am. 2007;34:35-42. 8. Shrim A, Garcia-Bournissen F, Koren G. Pharmaceutical 4. Thomson Reuters (Healthcare). Micromedex® Healthcare Series agents and pregnancy in urology practice. Urol Clin North Am. Intranet. 5.1. 2007;34:27-33. 5. Czeizel AE, Rockenbauer M, Olsen J. Use of antibiotics dur- 9. Yoon PW, Rasmussen SA, Lynberg MC, et al. The National Birth ing pregnancy. Eur J Obstet Gynecol Reprod Biol. 1998;81: Defects Prevention Study. Public Health Rep. 2001;116(suppl 1-8. 1):32-40.

Erratum The March PURL (When is it safe to forego a CT in kids with head trauma?, J Fam Pract. 2010;59;159-164) included an acknowl- edgement of the guidance provided by Sarah-Anne Schumann, MD, in the preparation of the manuscript. Dr. Schumann was actually a coauthor. The byline should have read: Kohar Jones, MD; Gail Patrick, MD, MPP; Sarah-Anne Schumann, MD; Depart- ment of Family Medicine, University of Chicago (Drs. Jones and Schumann); Department of Family and Community Medicine, Northwestern University, Chicago (Dr. Patrick).

222 The Journal of Family Practice | APRIL 2010 | Vol 59, No 4