Obstetrics & Gynecology:
Accuracy of Diagnostic Tests to Detect Asymptomatic Bacteriuria During Pregnancy
Mignini, Luciano MD1; Carroli, Guillermo MD1; Abalos, Edgardo1; Widmer, Mariana MD2; Amigot, Susana4; Nardin, Juan Manuel MD1; Giordano, Daniel1; Merialdi, Mario MD, PhD2; Arciero, Graciela4; del Carmen Hourquescos, Maria4; for the World Health Organization Asymptomatic Bacteriuria Trial Group
From 1Centro Rosarino de Estudios Perinatales, Santa Fe, Argentina; the 2World Health Organization, World Bank Special Programme of Research, Development and Research Training in Human Reproduction, and Department of Reproductive Health and Research, World Health Organization, Geneva, Switzerland; and 4Laboratorio Central, Centro de Especialidades Médicas Ambulatorias de Rosario “M. H. Zuasnábar” (CEMAR), Dirección de servicios de Laboratorio y Análisis Clínicos (DSLyAC), Secretaria de Salud Publica de la Municipalidad de Rosario, Rosario, Argentina.
See related article on page 339.
*For a list of members of the World Health Organization Asymptomatic Bacteriuria Trial Group, see the Appendix online at http://links.lww.com/A645.
Supported by UNDP/UNFPA/WHO/World Bank Special Programme of Research, Development and Research Training in Human Reproduction. Orion Diagnostica Finland provided the dipslides, and Wiener Laboratories S.A.I.C. provided the dipsticks. The Central Laboratory, CEMAR, Department of Public Health of the Municipality of Rosario, provided the quantitative urine cultures for all women enrolled in the study.
The authors thank Dr. Pisake Lumbiganon from the Department of Obstetrics and Gynaecology, Faculty of Medicine, Khon Kaen University, Thailand, and Dr. Malinee Laopaiboon from the Department of Biostatistics and Demography, Faculty of Public Health, Khon Kaen University, Thailand, for their important contributions to this article.
Disclaimer: The views expressed are solely those of the authors and do not necessarily reflect the decisions or stated policy of the World Health Organization.
Corresponding author: Luciano Mignini, Centro Rosarino de Estudios Perinatales (CREP), Pueyrredón 985, (2000) Rosario, Argentina; e-mail: firstname.lastname@example.org.
Financial Disclosure The authors did not report any potential conflicts of interest.
OBJECTIVE: A dipslide is a plastic paddle coated with agar that is attached to a plastic cap that screws onto a sterile plastic vial. Our objective was to estimate the diagnostic accuracy of the dipslide culture technique to detect asymptomatic bacteriuria during pregnancy and to evaluate the accuracy of nitrate and leucocyte esterase dipslides for screening.
METHODS: This was an ancillary study within a trial comparing single-day with 7-day therapy in treating asymptomatic bacteriuria. Clean-catch midstream samples were collected from pregnant women seeking routine care. Positive and negative likelihood ratios and sensitivity and specificity for the culture-based dipslide to detect and chemical dipsticks (nitrites, leukocyte esterase, or both) to screen were estimated using traditional urine culture as the “gold standard.”
RESULTS: A total of 3,048 eligible pregnant women were screened. The prevalence of asymptomatic bacteriuria was 15%, with Escherichia coli the most prevalent organism. The likelihood ratio for detecting asymptomatic bacteriuria with a positive dipslide test was 225 (95% confidence interval [CI] 113–449), increasing the probability of asymptomatic bacteriuria to 98%; the likelihood ratio for a negative dipslide test was 0.02 (95% CI 0.01–0.05), reducing the probability of bacteriuria to less than 1%. The positive likelihood ratio of leukocyte esterase and nitrite dipsticks (when both or either one was positive) was 6.95 (95% CI 5.80–8.33), increasing the probability of bacteriuria to only 54%; the negative likelihood ratio was 0.50 (95% CI 0.45–0.57), reducing the probability to 8%.
CONCLUSION: A pregnant woman with a positive dipslide test is very likely to have a definitive diagnosis of asymptomatic bacteriuria, whereas a negative result effectively rules out the presence of bacteriuria. Dipsticks that measure nitrites and leukocyte esterase have low sensitivity for use in screening for asymptomatic bacteriuria during gestation.
CLINICAL TRIAL REGISTRATION: ISRCTN, isrctn.org, 1196608
LEVEL OF EVIDENCE: II
A symptomatic bacteriuria occurs in 5% to 10% of all pregnancies,1–3 the prevalence being highest in women of low socioeconomic status and in diabetics.2 Pregnant women who are not treated appear more likely to develop pyelonephritis, on occasion life-threatening, and to deliver more low birth weight neonates.4 On the other hand, and as demonstrated through randomized trials, antimicrobial treatment of asymptomatic bacteriuria during pregnancy will decrease the risk of subsequent pyelonephritis from 20–35% to 1–4% and the risk of having a low birth weight baby from 15% to 5%.4 Thus, screening for and treating asymptomatic bacteriuria have become a standard component of obstetrical care, and such recommendations are part of most antenatal-care guidelines.5–8 Unfortunately, universal screening is not practiced worldwide, especially in developing countries where the costs of standard plate cultures are prohibitive to limited health care budgets and because they also may lack adequate laboratory facilities or trained microbiologists or both.9
Other approaches to overcome these problems have been proposed, including gram staining, microscopic urinalysis, urine dipsticks, dipslide urine cultures, and bioluminescence assays.10–14 Some of these methods are used widely because they are easy to perform and produce results rapidly. However, there appears to be little evidence of their accuracy as compared with standard culture techniques and quantified for positive and negative likelihood ratios and cost-effectiveness.15
The World Health Organization (WHO) has just completed a multicenter trial (ISRCTN11966080) in four countries (Argentina, Philippines, Thailand, and Vietnam) designed to test the efficacy of single-day therapy with nitrofurantoin compared with traditional 7-day treatment to eradicate asymptomatic bacteriuria in pregnancy.16 Of the approximately 25,000 women screened, 778 were recruited for the study. At all centers, dipslides were used to screen participants, and subsequent eligibility for enrollment was assessed on the basis of results of traditional cultures. As an ancillary study, at the center in Argentina, dipslides, chemical dipsticks, and agar plate cultures were performed simultaneously with the objective of testing the primary hypothesis that the diagnostic accuracy of dipslide tests (a culture-based alternative test to standard culture) is acceptable for determining whether a pregnant woman has asymptomatic bacteriuria during pregnancy, when comparing the agar plate culture as gold standard. A second objective of this study was to evaluate whether nitrite and leukocyte esterase dipsticks are useful screening tests.
MATERIALS AND METHODS
This ancillary study was conducted in parallel with the WHO treatment trial “A multicentre, double blind randomized trial to evaluate the effectiveness of a 1-day versus 7-day regimen of nitrofurantoin for the treatment of asymptomatic bacteriuria in pregnancy.” Details and results of this study, which comprised 778 women with asymptomatic bacteriuria randomized into treatment groups, are published elsewhere (Lumbiganon, in press). In Argentina, all women attending 17 antenatal care clinics who were scheduled to deliver at two maternal hospitals belonging to the Department of Public Health of the Municipality of Rosario were screened for bacteriuria at their first antenatal visits. Inclusion required presentation with a live fetus at gestational weeks 12 to 35. Because they would be ineligible for randomization if asymptomatic bacteriuria were detected, women with the following conditions were excluded from the study: underlying disease that required continuous steroid treatment or antibiotic treatment or both (eg, systemic lupus erythematosus, valvular heart disease), use of antibiotics before assessment, any treatment for urinary tract infection at any time during pregnancy, history of nitrofurantoin hypersensitivity, symptoms suggesting symptomatic urinary tract infection such as flank pain or dysuria, glucose 6 phosphate dehydrogenase deficiency or other hematological disease, previous negative urine culture or positive culture with an organism resistant to nitrofurantoin, and notation that they were not planning to deliver at the study hospital.
This trial was approved by the Scientific and Ethical Review Group at UNDP/UNFPA/WHO/World Bank Special Programme for Research, Development and Research Training in Human Reproduction and the institutional review boards of participating centers. Good clinical practice procedures17 were followed in all centers and monitored by the WHO on a regular basis. All participants provided an informed consent in writing.
Women were enrolled consecutively and instructed by trained personnel on the appropriate way to collect a clean-catch midstream urine sample into sterile containers. The samples, refrigerated locally at 4°C for a maximum of 4 hours, were transported to the Central Laboratory (Department of Public Health of the Municipality of Rosario), where each sample was divided into three aliquots and processed within 1 hour of arrival by trained laboratory microbiologists. Dipsticks, dipslides, and traditional cultures were performed by separate microbiologists, each unaware of the results of the other determinations. Women whose urine cultures were positive for potentially pathogenic organisms then were randomized to either a 1-day or a 7-day nitrofurantoin treatment regimen (200 mg/d).
One urine aliquot was analyzed using Orion Diagnostica's Uricult dipslide (Orion Diagnostica Oy, Espoo, Finland). The device consists of a 2-cm × 5-cm plastic paddle coated on both sides with agar media. (Fig. 1). One side of the Uricult dipslide is covered with green cysteine lactose electrolyte deficient medium, which supports growth of virtually all bacteria known to cause urinary tract infections. This medium also was used to determine the concentration of colony-forming units (CFU) in the urine sample. The reddish MacConkey medium on the other side of the Uricult slide is selective for gram-negative bacteria. The dipslide is attached to a plastic cap that screws onto a sterile plastic vial. The system was stored at room temperature before use. Dipslides were inoculated by dipping the agar-coated slides into the urine and incubated at 37°C for 24 hours. Results were determined by comparison of the microbial density on the slide with a model chart provided by the manufacturer. Dipslides were designated positive when the colony concentration was 105 CFU/mL or higher of a single microorganism or when two different colonies were present but one was 105 CFU/mL or higher. Dipslide tests were considered negative when the concentration was less than 105 CFU/mL and/or considered contaminated if more than two different colony species grew on them.
The second aliquot was tested for presence of nitrites and leukocyte esterase using Urine Reagent Test Strips (Wiener Laboratories S.A.I.C., Rosario, Argentina) per the manufacturer's instructions. Results of this test were interpreted visually according to standard color charts. The leukocyte esterase measurement was recorded as negative, trace, small (+1), moderate (+2), or large (+3), the latter considered positive. The nitrite measurement was read at 60 seconds and recorded as either negative or positive.
The third aliquot of the urine sample underwent classic quantitative culturing in the microbiology laboratory. MacConkey, blood, and cysteine lactose electrolyte deficient agar plates were swabbed and incubated for 24 hours at 37°C, the culture considered positive if 105 CFU/mL or more of a single potential uropathogen or of two organisms not consistent with skin flora were isolated. Cultures were considered negative if there was no growth of bacteria or if the growth was less than 105 CFU/mL. Cultures were considered contaminated if more than two microorganisms were isolated. Results of the culture were considered the gold standard for assessing the diagnostic accuracy of dipslides and dipsticks.
The likelihood ratios, sensitivity, and specificity for dipsticks and dipslides were calculated using SAS 9.1 (SAS Institute, Inc., Cary, NC). As suggested by Jaeschke et al, a likelihood ratio for a positive test result above 10 and a likelihood ratio for a negative test result below 0.1 were considered to provide convincing diagnostic accuracy. Moderate prediction was designated as likelihood ratio values of 5 to 10 and 0.1 to 0.2, whereas those below 5 and above 0.2 gave only minimal prediction.18
The study was completed between August 2004 and April 2007. A total of 3,048 eligible pregnant women were screened. All had their urine screened using agar plate urine cultures. Uricult-dipslide studies were performed on 3,047 and chemical dipstick determinations on 3,032 samples (Fig. 2). One dipslide and 16 dipsticks were omitted owing to unavailability of the samples. Escherichia coli was the most prevalent uropathogen isolated by traditional culture (Table 1). In our study, 681 urine samples had contaminated cultures (22.3%) and were not included in the analyses.
Of the 3,047 dipslides, 348 (11.4%) were positive, 1,869 (61.3%) were negative, and 830 (27.2%) were contaminated (Table 2). There were eight cases in which dipslide tests were negative but urine cultures positive, and eight urine samples showing positive dipslide tests were negative by standard culture (105 CFU/mL or less). The positive likelihood ratio was 225 (95% confidence interval [CI] 113–449), and the negative likelihood ratio was 0.02 (95% CI 0.01–0.05) (Table 3).
Of 3,032 dipsticks, 486 (16%) read positive to nitrites or leukocyte esterase or both, and 2,546 (84%) were negative to both (Table 4). Also, both false-positive and false-negative rates were high; the positive likelihood ratio for detecting asymptomatic bacteriuria was only 6.95 (95% CI 5.8–8.33), and the negative likelihood ratio was 0.50 (95% CI 0.45–0.57) (Table 3).
This study primarily tested the diagnostic accuracy of dipslide tests for detecting asymptomatic bacteriuria during pregnancy using traditional culture as the gold standard. Our results are though-provoking not only because the dipslide test proved to be very accurate, but also because, owing to its simplicity compared with traditional cultures, it could allow universal screening to be extended in those countries where limited resources, both financial and the unavailability of trained microbiologists, have heretofore limited such practices.
We have chosen to study the dipslide test's accuracy because available evidence indicated it could be a more feasible alternative test to traditional urine culture.19–21 If proven accurate, it could contribute to reducing the costs of testing women for asymptomatic bacteriuria and would be very suitable for small laboratories or medical centers in areas where health care resources are limited.22 In addition, the dipslide test is simpler to perform and can be applied in an office setting. Moreover, results can be read directly by physicians or paramedical personnel without waiting for laboratory responses. Also, there is the potential advantage of directly culturing a freshly voided urine sample, eliminating the risk of potential overgrowth of commensal flora during transport to the laboratory.22 In this respect, it is well-known that leaving urine specimens at room temperature for a prolonged amount of time may modify the final counts of microorganisms. This further suggests that a reliable dipslide test would be of greater value where delays between the clinics and the laboratory are either unavoidable, unpredictable, or both, and where facilities for refrigeration are unavailable.23
The results of our study show that a positive dipslide result increases the likelihood of asymptomatic bacteriuria from 15% (pretest probability) to 98% (posttest probability), and a negative dipslide result reduces the possibility of asymptomatic bacteriuria to less than 1%. This means that only 2% of pregnant women with asymptomatic bacteriuria will not receive antibiotic treatment when necessary and fewer than 1% of pregnant women without asymptomatic bacteriuria would be classified as positive and consequently would be treated.
There are disadvantages that must be considered. The dipslide approach does not indicate sensitivities and requires the use of antibiotics that cover a broad spectrum of sensitivities. In our study, nitrofurantoin met this requirement because most of the microorganisms isolated in our study were sensitive to this antibiotic (Widmer M, Amigot S, Sribenjalux P, Laopaiboon M, Bergel E, Beltran A. Asymptomatic bacteriuria in pregnancy: etiology and antimicrobial susceptibility. Submitted). Although dipslides appeared easy to perform and to interpret with few if any problems or pitfalls, they were performed in a specialized laboratory by trained microbiologists who were participating in a large, multicenter treatment trial. Thus, one must be cautious in extrapolating such results to heath facilities staffed with less-well-trained personnel. The dipslide is stored at room temperature before being inoculated, and this might lead overestimating the bacterial counts.22 A recent report has claimed that, under daily practice conditions, the accuracy of dipslides may be lower than under optimal laboratory conditions.19 Thus, when introducing dipslides into practice the personnel involved in every step of the procedure should be trained in all proper techniques and pitfalls to avoid. We believe this approach can be implemented with ease even in developing countries.
We used a two-medium dipslide (Uricult, Orion Diagnostica) coated with the cysteine lactose electrolyte deficient medium on one side and a MacConkey medium on the other side. Almost all bacteria causing urinary tract infections grow on the cysteine lactose electrolyte deficient medium, and the MacConkey medium is selective for gram-negative bacteria, so tentative diagnoses can be made through quantitative and qualitative interpretation of the growth of colonies on these media. However, to typify the causative microorganism and its antibiotic sensitivity, the traditional quantitative culture is still needed. Taking into account that, in our study, 65% of the samples were positive for this uropathogen (Table 1) and that their sensitivity to nitrofurantoin was high (98%), this kind of test seems to be the test of choice to detect asymptomatic bacteriuria during pregnancy.
Dipslides are proposed as office screening methods; however, there are still delays in obtaining results (up to 24 hours). A separate problem appears to be the lack of ability of dipslides to differentiate asymptomatic bacteriuria caused by microorganisms other than E. coli. For example, the prevalence of asymptomatic bacteriuria caused by beta-hemolytic streptococcus was 8% in our population. Of 26 urine cultures showing streptococcal growth, 20 (77%) were positive to dipslides, but the etiologic agent was detected only by traditional cultures. This should be taken into account in populations with a high prevalence of beta-hemolytic streptococcus or other microorganisms because of the possibility of missing specific treatment.
Although many studies have evaluated the accuracy of dipstick tests for rapid diagnosis of bacteriuria and urinary tract infections,24–27 there is still an ongoing debate regarding their accuracy.28 A published meta-analysis of a urine dipstick test's ability to rule out urine infections concluded that the accuracy of nitrites was high and that a negative test for both nitrites and leukocyte esterase could rule out urinary tract infections in asymptomatic pregnant women.28 In our study, the performance of dipsticks for the diagnosis of asymptomatic bacteriuria was poor (Tables 3 and 4). A positive test result increases the likelihood of asymptomatic bacteriuria from 15% (pretest probability) to only 54% (posttest probability). A negative test reduces that probability to 8%, meaning that, if the dipsticks were the primary screening method, 46% of all pregnant women with asymptomatic bacteriuria would be missed. This rate is unacceptable because pregnant women with untreated bacteriuria have significantly increased rates of low birth weight neonates, preterm delivery, and pyelonephritis.4
Taking into account the high accuracy for detection of asymptomatic bacteriuria and the advantages regarding their use, dipslide devices should be considered as valid alternatives to traditional cultures. However, further research in more pragmatic conditions is desirable before recommending this method for use as an office screening test. Dipstick tests performed only once at the first antenatal visit should be discouraged because of their low accuracy for detecting asymptomatic bacteriuria. Multiple testing strategies should be explored further.
1. Millar LK, Cox SM. Urinary tract infections complicating pregnancy. Infect Dis Clin North Am 1997;11:13–26.
2. Whalley P. Bacteriuria of pregnancy. Am J Obstet Gynecol 1967;97:723–38.
3. Little PJ. The incidence of urinary infection in 5000 pregnant women. Lancet 1966;2:925–8.
4. Smaill F, Vazquez JC. Antibiotics for asymptomatic bacteriuria in pregnancy. The Cochrane Database of Systematic Reviews 2007, Issue 2. Art. No.: CD000490. DOI: 10.1002/14651858. CD000490.pub2.
5. Screening for asymptomatic bacteriuria: a brief evidence update for the U.S Preventive Services Task Force. Agency for Healthcare Research and Quality. Available at http://www.preventiveservices.ahrq.gov
. Retrieved May 8, 2008.
6. Nicolle LE, Bradley S, Colgan R, Rice JC, Schaeffer A, Hooton TM, et al. Infectious Diseases Society of America guidelines for the diagnosis and treatment of asymptomatic bacteriuria in adults [published erratum appears in Clin Infect Dis 2005;40:1556]. Clin Infect Dis 2005;40:643–54.
7. National Collaborating Centre for Women's and Children's Health. Clinical Guideline 6. Antenatal Care. Routine care for the healthy pregnant woman. London. National Institute for Health and Clinical Excellence 2003.
8. The Scottish Intercollegiate Guidelines Network (SIGN). Management of suspected bacterial urinary tract infection in adults. A national clinical guideline. Edimburg. The Scottish Intercollegiate Guidelines Network (SIGN) 2006.
9. Carroli G, Rooney C, Villar J. How effective is antenatal care in preventing maternal mortality and serious morbidity? An overview of the evidence. Paediatr Perinat Epidemiol 2001;15 (Suppl 1):1–42.
10. Bachman JW, Heise RH, Naessens JM, Timmerman MG. A study of various tests to detect asymptomatic urinary tract infections in an obstetric population. JAMA 1993;270:1971–4.
11. Blondeau JM, Yaschuk Y, Galenzoski D, Hrabok D, Isaacson M, Lee L, et al. Evaluation of the Cult-Dip Plus dip slide method for urinary tract infection. J Clin Pathol 1995 Aug;48:710–3.
12. Graninger W, Fleischmann D, Schneeweiss B, Aram L, Stockenhuber F. Rapid screening for bacteriuria in pregnancy. Infection 1992;20:9–11.
13. Kacmaz B, Cakir O, Aksoy A, Biri A. Evaluation of rapid urine screening tests to detect asymptomatic bacteriuria in pregnancy. Jpn J Infect Dis 2006;59:261–3.
14. Lenke RR, Van Dorsten JP. The efficacy of the nitrite test and microscopic urinalysis in predicting urine culture results. Am J Obstet Gynecol 1981;140:427–9.
15. Abalos EJ, Gulmezoglu AM, Carroli G. Assessing the scientific value of screening for antenatal infections. In: Newell ML, McIntyre JE, editors. Congenital and perinatal infections. Cambridge (UK): Cambridge University Press; 2000. p. 64–79.
. Lumbiganon P, Villar J, Laopaiboon M, Widmer M, Thinkhamrop J, Carroli G. One-day compared with 7-day nitrofurantoin for asymptomatic bacteriuria in pregnancy: a randamized controlled trial. Obstet Gynecol 2009;113:339–345.
17. World Health Organization. Guidelines for good clinical practice for trials on pharmaceutical products. WHO Technical Report Series No 850, Geneva, Switzerland 1995;97–137.
18. Jaeschke R, Guyatt GH, Sackett DL. Users' guides to the medical literature. III. How to use an article about a diagnostic test. B. What are the results and will they help me in caring for my patients? The Evidence-Based Medicine Working Group. JAMA 1994;271:703–7.
19. Winkens R, Nelissen-Arets H, Stobberingh E. Validity of the urine dipslide under daily practice conditions. Fam Pract 2003;20:410–2.
20. Olsen BE, Hinderaker SG, Lie RT, Gasheka P, Baerheim A, Bergsjø P, et al. The diagnosis of urinary tract infections among pregnant women in rural Tanzania; prevalences and correspondence between different diagnostic methods. Acta Obstet Gynecol Scand 2000;79:729–36.
21. Hallander HO, Kallner A, Lundin A, Osterberg E. Evaluation of rapid methods for the detection of bacteriuria (screening) in primary health care. Acta Pathol Microbiol Immunol Scand [B] 1986 Feb;94:39–49.
22. Morandi PA, Mauris A, Deom A, Rohner P. External quality control results of urine dip-slide devices. Diagn Microbiol Infect Dis 2007;57:235–41.
23. European Confederation of Laboratory Medicine. European urinalysis guidelines. Scand J Clin Lab Invest Suppl 2000;231:1–86.
24. McNair RD, MacDonald SR, Dooley SL, Peterson LR. Evaluation of the centrifuged and Gram-stained smear, urinalysis, and reagent strip testing to detect asymptomatic bacteriuria in obstetric patients. Am J Obstet Gynecol 2000;182:1076–9.
25. Millar L, DeBuque L, Leialoha C, Grandinetti A, Killeen J. Rapid enzymatic urine screening test to detect bacteriuria in pregnancy. Obstet Gynecol 2000;95:601–4.
26. Robertson AW, Duff P. The nitrite and leukocyte esterase tests for the evaluation of asymptomatic bacteriuria in obstetric patients. Obstet Gynecol 1988;71:878–81.
27. Shelton SD, Boggess KA, Kirvan K, Sedor F, Herbert WN. Urinary interleukin-8 with asymptomatic bacteriuria in pregnancy. Obstet Gynecol 2001;97:583–6.
28. Devillé WL, Yzermans JC, van Duijn NP, Bezemer PD, van der Windt DA, Bouter LM. The urine dipstick test useful to rule out infections. A meta-analysis of the accuracy. BMC Urol 2004;4:4.
Supplemental Digital Content
© 2009 by The American College of Obstetricians and Gynecologists.