Journal of Obstetrics and Gynaecology, January 2014; 34: 21–24 © 2014 Informa UK, Ltd. ISSN 0144-3615 print/ISSN 1364-6893 online DOI: 10.3109/01443615.2013.823387
OBSTETRICS
Is a chlorhexidine reaction test better than dipsticks to detect asymptomatic bacteriuria in pregnancy? B. O. Okusanya1, E. O. S. Aigere2, J. O. Eigbefoh3, G. B. O. Okome3 & C. E. Gigi3
J Obstet Gynaecol Downloaded from informahealthcare.com by Nyu Medical Center on 06/14/15 For personal use only.
Department of Obstetrics and Gynaecology,1College of Medicine, University of Lagos, Lagos State, 2Federal Medical Centre Katsina, Katsina State, 3Irrua Specialist Teaching Hospital, Edo State, Nigeria
Detection of asymptomatic bacteriuria (ASB) in pregnancy is important to avert the attendant morbidities. Therefore, we assessed the use of chlorhexidine reaction to detect ASB in pregnancy. This was a prospective study, which compared chlorhexidine reaction with dipstick tests and urine culture in 150 asymptomatic pregnant women. Urine cultures detected bacteriuria in seven women (4.7%). Chlorhexidine detected ASB in 72 women (48%) and had sensitivity, specificity and accuracy of 100%, 54% and 56%, respectively. Leucocyte esterase (LE) and nitrite detected bacteriuria in 31 (20.7%) women and 12 (8.0%) women, respectively. Singly, LE had a sensitivity and specificity of 14.3% and 79%, respectively, while nitrite’s sensitivity and specificity was 42.9% and 93.7%, respectively. Combined, LE and nitrite had better sensitivity (97.9%) and accuracy (94%). Since the accuracy of chlorhexidine is low, other than urine culture, combined dipstick urinalysis of leucocyte esterase and nitrite tests is good to detect asymptomatic bacteriuria in pregnancy. Keywords: Asymptomatic bacteriuria, urinary tract infection
Introduction Asymptomatic bacteriuria is a precursor for urinary tract infections (UTI), which are common during pregnancy. The combination of mechanical, hormonal and physiological changes during pregnancy cause significant changes in the urinary tract, predispose to the acquisition and alters the natural history of bacteriuria during pregnancy (Abdullah and Al-Moslih 2005; Seccon et al. 2003). Also, progesterone relaxes ureteric smooth muscle causing dilatation of the ureters and stasis of urine, which is further aggravated by pressure from the expanding uterus (Perera 2009). Asymptomatic bacteriuria, generally defined as true bacteriuria, ⬎ 105 colony forming units of bacteria/ml of urine, in the absence of specific symptoms of acute urinary tract infection, occurred in 2–24% of all pregnancies in previous studies from Nigeria (Odigie and Anugweje 2010; Oyetunji et al. 2006; Eigbefoh et al. 2008; Mandara and Shittu 1999). It has been reported in 7.3% of pregnant women in Ghana (Turpin et al. 2007). The prevalence is most closely related to low socioeconomic status, which is a common predisposition in both pregnant and non-pregnant women (Smaill and Vazquez 2007). Due to the high prevalence, a systematic review concluded that screening for and aggressively
treating pregnant women for asymptomatic bacteriuria would possibly decrease the annual incidence of pyelonephritis during pregnancy by up to 80% (Smaill and Vazquez 2007). The organisms that cause urinary tract infection are the same as those found in non-pregnant patients. Escherichia coli accounts for 80–90% of infections. Other Gram-negative rods such as Proteus mirabilis and Klebsiella pneumonia have also been cultured (Perera 2009; Eigbefoh et al. 2008; Okonofua 1989). Asymptomatic bacteriuria precedes cystitis or acute pyelonephritis. In addition, it is associated with pre-term labour, low birth weight, prematurity, pre-eclampsia and chronic renal disease, with attendant adverse obstetric and perinatal outcome (Eigbefoh et al. 2008; Smaill and Vazquez 2007). Therefore, screening for and treatment of asymptomatic bacteriuria to prevent pyelonephritis, has been shown to be cost-effective (Rouse et al. 1995). Urine culture, although the gold standard for bacteriuria, is expensive, time-consuming and laborious. It also needs microbiology laboratory infrastructure and qualified staff. Moreover, although dipsticks are cheap and rapid tests, which require little expertise, they have repeatedly been shown to be unreliable diagnostic tools for asymptomatic bacteriuria (Kacmaz et al. 2006; Khattak et al. 2004; Millar et al. 2000). The chlorhexidine test is cheap, simple to use by middle-level health personnel and requires no microscope or electricity. When used to detect bacteriuria in patients with urinary tract infection, it was very sensitive but had low accuracy and specificity, which made the authors recommend its use as a screening test (Okonkwo et al. 2006). When chlorhexidine is added to a suspension of bacterial cells, it is immediately adsorbed on the bacterial surface (Okonkwo et al. 2006). Even at low concentrations, there is a rapid and irreversible loss of cytoplasmic constitutions, which causes precipitation of cellular protein and nucleic acid (Okonkwo et al. 2006). While most clinicians agree on the need for early urine screening, the best screening test, other than urine culture, is yet to be determined. The search, however, for a cheap, easy and acceptable screening method is continuing. This was the basis for this research, which evaluated chlorhexidine urine test for screening of asymptomatic bacteriuria in pregnant women. The study objective was to determine the accuracy and correlation of chlorhexidine reaction test when compared with dipstick tests (leukocyte esterase and nitrite) and urine culture. The cost–benefit of the different screening tests was also determined.
Correspondence: B. O. Okusanya, Department of Obstetrics and Gynaecology, College of Medicine, University of Lagos, Idi-Araba, Lagos. E-mail: [email protected]
22
B. O. Okusanya et al.
J Obstet Gynaecol Downloaded from informahealthcare.com by Nyu Medical Center on 06/14/15 For personal use only.
Subjects and methods This was a prospective observational study conducted at the antenatal clinic and laboratory of Irrua Specialist Teaching Hospital (ISTH). Pregnant women were included if they had no dysuria, urine frequency, suprapubic and loin pain or vaginal discharge. In addition to the presence of any of the aforementioned symptoms, women using antibiotics for any reason or who had antibiotic treatment in the preceding 2 weeks prior to counselling on the study, were excluded. At a prevalence of 10% (Smail and Vazquez 2007) of asymptomatic bacteriuria, the sample size of 138 pregnant women was needed, based on a formula (Araoye 2003). However, 150 consecutive pregnant women attending the antenatal booking clinic who met the inclusion criteria were recruited after they gave informed consent. The Research Ethic Committee of ISTH approved the study. The method of collection of urine was explained to the women and they were assisted by a nurse if needed, in order to reduce the chances of contamination. Each patient was given a sterile, dry, wide mouth container, with an instruction to clean the vulva with clean water. Urine was then collected (approx. 20 ml) by midstream clean catch technique. The urine sample was divided into three containers (one for each screening method under consideration) and labelled accordingly. The samples to be sent to the laboratory were transported to the laboratory immediately and processed within 1 h. When there was any delay, the samples were refrigerated at 4°C. The urine specimens were processed by routine quantitative culture and were also tested by the screening methods under consideration.
Dipstick test – leukocyte esterase and nitrite This was performed within 5 min of passing urine. Urine specimen was tested with colorimetric Multistix 10 SG (Ames Division Miles Inc., Elkhart, IN) for the presence of nitrite and leukocyte esterase activity following the manufacturer’s instructions. The nitrite and leukocyte esterase portions of the test were interpreted as positive if the colour on the reagent square was positive for each portion. Tests that showed no colour change or trace results were considered as negative.
Chlorhexidine reaction In this chemical test, chlorhexidine 1:1000 solution in water was utilised. Using a laboratory dropper, 10 drops of urine from a container was placed in a clean test-tube; the colour and appearance of the urine was noted. Five drops of chlorhexidine was added to the test tube and as this was being done, the urine in the test tube was observed for any changes, e.g. cloudiness, precipitate or particulate matter. The entire contents of the test-tube was then shaken and observed after 1 min. A cloudy colour indicated a positive result. Half the volume of chlorhexidine (5 drops) was titrated against the full volume of urine (10 drops) because this was found in an earlier study to be the critical point were the colour changes were maximal (Okonkwo et al. 2006). Two laboratory scientists performed the urine culture in the laboratory, while the researcher and two trained resident doctors performed the chlorhexidine reaction test and dipstick test for leukocyte esterase and nitrite. This was done at the antenatal clinic. Chlorhexidine reaction test was evaluated using sensitivity, specificity, the positive predictive value, negative predictive value and accuracy rate, when compared with dipsticks and urine culture.
A questionnaire was used to document participants’ sociodemographic characteristics and the results of the screening tests before they were analysed. Data were analysed using the Statistical Package for Social Science v. 16 (SPSS Inc., IL). The various tests of validity were subsequently calculated.
Results A total of 150 pregnant women were recruited during the period of study. Seven women had positive urine cultures. The prevalence of asymptomatic bacteriuria in this study was therefore 4.7%. The sociodemographic characteristics of the women are in Table I. Using urine culture as the gold standard, the diagnostic performance of chlorhexidine reaction test was evaluated. Dipstick detected asymptomatic bacteriuria in 31 women (20.7%) using leucocyte esterase, and in 12 women (8.0%), with nitrite alone. The sensitivity and specificity of leucocyte esterase test was 14.3% and 79.0%, respectively, while nitrite sensitivity and specificity was 42.9% and 93.7%, respectively. The combined test (leucocyte esterase and nitrite) confirmed one out of the seven urine-culture positive samples (14.3% sensitive), missed six urine-culture positive samples and was 97.9% specific. Furthermore, the positive and negative predictive value of leucocyte esterase was 3.2% and 95%, respectively; that of nitrite was 25% and 97.1% and the combined test had positive and negative predictive values of 25% and 96%, respectively. Leucocyte esterase was 76% accurate, while the accuracy for nitrite was 91%. When combined, both tests had a slightly higher accuracy of 94%. This implies that the combined test Table I. Sociodemographic characteristics of sample population. Characteristics Parity 0 1 2 3 4 ⱖ5 Age (years) 20–24 25–29 30–34 35–39 ⱖ 40 Occupation Housewife Civil servant Trading Student Others Ethnicity Esan Afemai Edo Igbo Yoruba Others Level of education Primary Secondary Tertiary Marital status Married Single
n
(%)
35 56 30 18 10 1
23.3 37.3 20 12 6.7 7
22 52 48 24 4
14.7 34.7 32 16 2.7
20 41 56 30 3
13.3 27.3 37.3 20 2
98 18 13 10 7 4
65.3 12 8.6 6.7 4.7 2.7
3 59 90
2 38 60
150 0
100 0
J Obstet Gynaecol Downloaded from informahealthcare.com by Nyu Medical Center on 06/14/15 For personal use only.
Chlorhexidine test for asymptomatic bacteriuria in pregnancy 23 will reliably diagnose asymptomatic bacteriuria in nine out of every 10 women tested. Unlike dipsticks, chlorhexidine reaction test detected asymptomatic bacteriuria in 72 (48%) women. All seven urine-culture positive samples were detected by chlorhexidine reaction test (100% sensitive), and it was 54% specific. It had a positive and negative predictive value of 9.7% and 100%, respectively, an accuracy rate of 56%, and a false negative rate of 0%. Isolated organisms were Gram-negative with Escherichia coli being the predominant organism isolated (57.1%). Others were Klebsiella species (14.3%), Proteus species (14.3%). The only Gram-positive organism was Staphylococcus aureus (14.3%). A total of 85.7% of organisms isolated were sensitive to Nitrofurantoin, while 57.1% of isolated organisms were sensitive to Gentamicin, Ofloxacin and Augmentin®. The isolated organisms were all poorly-sensitive to Amoxicillin. Chlorhexidine reaction was done in 1–2 min, the same time as for dipsticks, but the method is much cheaper than dipsticks (US$0.31 vs US$1.25). Urine cultures cost US$3.13 and the results were not ready for use until 48 h after sample collection.
Discussion Asymptomatic bacteriuria is the presence of actively multiplying bacteria at a time when the patient has no urinary symptoms, so that diagnosis relies upon microbiological findings (GaringaloMolina 2000). An early detection and treatment of asymptomatic bacteriuria may be of considerable importance, not only to forestall acute pyelonephritis and chronic renal failure in the mother, but also to reduce prematurity and perinatal mortality in the newborn (Perera 2009). In this study, four main tests of validity were used to compare urine leucocyte esterase and nitrite tests singly, and in combination with chlorhexidine test. Urine culture served as the gold standard for their ability to detect asymptomatic bacteriuria. Of the dipsticks, nitrite had the highest sensitivity (42.9%). Although low, what we found fell within the often-quoted range of 35–85% for the nitrite test (Khattak et al. 2004; Millar et al. 2000). Since sensitivity is a measure of the false-negative rate of 57.1%, the nitrite test will test negative in approximately six out of every 10 women with significant bacteriuria. This is likely to be due to bacteriuria of organisms not having nitrate reductase. This is seen when the urine has only stayed in the bladder for ⬍ 4 h, such that nitrate is yet to be reduced to nitrite, or when dietary nitrite is absent (Kacmaz et al. 2006). Thus, it is best that the first voided urine in the morning be used (Jayalakshmi and Jayaram 2008), as it has been proven to be accurate, albeit morning urine collection was not practicable in our patient population, since the study enrolled outpatient pregnant women. This may explain the low sensitivity. Conversely, nitrite test’s high specificity meant that if the test was negative for bacteriuria, it was accurate in nine of 10 women. It should be borne in mind that a positive nitrite test indicates that nitrite has been produced by the reduction of nitrate by enteric bacteria, most commonly from the Enterobacteriaceae family (Kacmaz et al. 2006). The tests of validity for leucocyte esterase and nitrite suggest that they are unreliable when used alone to detect asymptomatic bacteriuria. However, when combined, the high specificity and accuracy rate of 95.9% mean that almost every woman with asymptomatic bacteriuria will be detected. Hence, when dipstick is the only option available, a combined test must be done to detect asymptomatic bacteriuria. Chlorhexidine reaction accurately detected asymptomatic bacteriuria in all women who had positive urine culture. The
mechanism of the reaction which precipitates cellular protein and nucleic acid explains this. Though the sensitivity of chlorhexidine was high, and correlates with an earlier study which involved symptomatic women (Okonkwo et al. 2006), its specificity of 54.6% and accuracy of 56% meant that the chlorhexidine test cannot replace urine dipsticks in routine use for the detection of asymptomatic bacteriuria. This is because chlorhexidine also reacts with other components of urine, such as cells and crystals (Okonkwo et al. 2006). Regarding the methods and costs, just as dipstick tests, chlorhexidine tests are easy to perform and do not need laboratory or trained personnel. The results of both tests were ready within 1–2 min. The cost of chlorhexidine reaction, dipstick urinalysis and urine culture was, respectively, 50 Naira (30 cents), 200 Naira (US$ 1.3) and 500 Naira (US$ 3.3). The low accuracy negates the cheaper price of chlorhexidine. This implies that other than urine culture, dipstick tests of leucocyte esterase and nitrites will remain the choice, especially when combined to detect asymptomatic bacteriuria in pregnant women. The prevalence of asymptomatic bacteriuria of 4.7% of this study, was much lower than the 22% prevalence found 6 years earlier in the same institution (Eigbefoh et al. 2008). In the 6 years between the two studies, the community has transited from predominantly rural to a suburban setting. Since the prevalence of bacteriuria is closely related to socioeconomic status (Smaill and Vazquez 2007), improvement in the status of women attending the antenatal clinic may be responsible for the different prevalence. The most common organism isolated was Escherichia coli. This conforms to previous studies, which showed that Escherichia coli is the most common cause of bacteriuria in pregnancy (Ezechi et al. 2003; Abdul and Onile 2001). Other organisms isolated were Klebsiella species, Proteus species and Staphylococcus aureus, as previously reported (Ezechi et al. 2003; Abdul and Onile 2001; Akerele et al. 2001). When bacteriuria is detected in pregnant women, they should be treated. The choice of antibiotic should be effective for the most common infective organisms (i.e. Gram-negative gastrointestinal organisms) and should be safe for the mother and fetus. Nitrofurantoin was identified in this study as having the highest and widest spectrum of activity against the organisms responsible for bacteriuria. This was a good choice, since nitrofurantoin has high urinary concentration (Sanders and Sanders 1992). Gentamycin, Augmentin and Ofloxacin are other effective antibiotics in this study. Ofloxacin, a fluoroquinolone, although with fairly good spectrum, has possible toxic effects on the fetus and should not be prescribed to pregnant women. Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
References Abdul IF, Onile BA. 2001. Bacterial isolates from urine of women in Ilorin and their antibiotic susceptibility patterns. Tropical Journal of Obstetrics and Gynaecology 18:61–65. Abdullah AA, Al-Moslih MI. 2005. Prevalence of asymptomatic bacteriuria in pregnant women in Sharjah, United Arab Emirates. East Mediterranean Health Journal 11:1045–1052. Akerele J, Abhulimen P, Okonofua F. 2001. Prevalence of asymptomatic bacteriuria among pregnant women in Benin City, Nigeria. Journal of Obstetrics and Gynaecology 21:141–144. Araoye MO (ed.). 2003. Subject selection. In: Research methodology with statistics for health and social sciences. Ilorin: Nathadex Publications. p. 115–139.
J Obstet Gynaecol Downloaded from informahealthcare.com by Nyu Medical Center on 06/14/15 For personal use only.
24
B. O. Okusanya et al.
Eigbefoh JO, Isabu P, Okpere E, Abebe J. 2008. The diagnostic accuracy of the rapid dipstick test to predict asymptomatic urinary tract infection of pregnancy. Journal of Obstetrics and Gynaecology 28:1–6. Ezechi OC, Fasuba OB, Dare FO. 2003. Antibiotic sensitivity pattern of microbial isolates from the urine of pregnant women with urinary tract infections. Tropical Journal of Obstetrics and Gynaecology 20: 133–135. Garingalo-Molina FD. 2000. Asymptomatic bacteriuria among pregnant women: Overview of diagnostic approaches. Philippines Journal of Microbiology and Infectious Disease 29:177–186. Jayalakshmi J, Jayaram VS. 2008. Evaluation of various screening tests to detect asymptomatic bacteriuria in pregnant women. Indian Journal of Pathology and Microbiology 51:379–381. Kacmaz B, Cakir O, Aksoy A, Biri A. 2006. Evaluation of rapid urine screening tests to detect asymptomatic bacteriuria in pregnancy. Japan Journal of Infectious Diseases 59:261–263. Khattak AM, Khan H, Akhtar W, Mahsud I, Ashiq B. 2004. Accuracy of Griess test to predict asymptomatic bacteriuria during pregnancy. Gomal Journal of Medical Sciences 2:20–23. Mandara MU, Shittu SO.1999. Asymptomatic bacteriuria in antenatal patients at Ahmadu Bello University Teaching Hospital Zaria, Nigeria. Tropical Journal of Obstetrics and Gynaecology 16:41–45. Millar L, Debuque L, Leialoha C, Grandinetti A, Killeen J. 2000. Rapid enzymatic urine screening test to detect bacteriuria in pregnancy. Obstetrics and Gynaecology 95:601–604. Odigie JO, Anugweje KC. 2010. Asymptomatic bacteriuria in pregnancy in Port Harcourt. Asian Pacific Journal of Tropical Medicine 3:580–583.
Okonkwo CA, Okpere EE, Ande AB. 2006. Evaluation of chlorhexidine in the detection of bacteriuria in pregnancy. Tropical Journal of Obstetrics and Gynaecology 23:14–16. Okonofua FE, Adediran AA, Okonofua B. 1989. Incidence and pattern of asymptomatic bacteriuria of pregnancy in Nigerian women. Nigerian Medical Practitioner 17:35–38. Oyetunji JA, Ahmed Y, Nwobodo EI, Airede LR, Ekele BA. 2006. Asymptomatic bacteriuria in pregnancy in Sokoto, Nigeria. Sahel Medical Journal 9:1–6. Perera J. 2009. Asymptomatic bacteriuria in pregnancy. Sri Lanka journal of Obstetrics and Gynaecology 31:108–109. Rouse DJ, Andrews WW, Goldenberg RL, Owen J. 1995. Screening and treatment of asymptomatic bacteriuria in pregnancy to prevent pyelonephritis: a cost effectiveness and cost- beneficial analysis. Obstetrics and Gynaecology 86:119–123. Sanders CC, Sanders NE Jnr. 1992. Beta lactam resistance in gram negative bacteria: Global trends and clinical impact. Clinical Infectious Diseases 15:824–839. Seccon NC, Garingalo-Molina FD, Yoasiano CJ, Saniel MC, Manalastaas RM. 2003. Prevalence of asymptomatic bacteriuria and associated risk factors in pregnant women. Philippines Journal of Microbiology and Infectious Disease 32:63–69. Smaill F, Vazquez JC. 2007. Antibiotics for asymptomatic bacteriuria in pregnancy. Cochrane Database of Systematic Reviews (2):CD000490. Turpin CA, Minkah B, Danso KA, Frimpong EH. 2007. Asymptomatic bacteriuria in pregnant women attending Antenatal Clinic at Komfo Anokye Teaching Hospital, Kumasi, Ghana. Ghana Medical Journal 41:26–29.
OBSTETRICS
Is a chlorhexidine reaction test better than dipsticks to detect asymptomatic bacteriuria in pregnancy? B. O. Okusanya1, E. O. S. Aigere2, J. O. Eigbefoh3, G. B. O. Okome3 & C. E. Gigi3
J Obstet Gynaecol Downloaded from informahealthcare.com by Nyu Medical Center on 06/14/15 For personal use only.
Department of Obstetrics and Gynaecology,1College of Medicine, University of Lagos, Lagos State, 2Federal Medical Centre Katsina, Katsina State, 3Irrua Specialist Teaching Hospital, Edo State, Nigeria
Detection of asymptomatic bacteriuria (ASB) in pregnancy is important to avert the attendant morbidities. Therefore, we assessed the use of chlorhexidine reaction to detect ASB in pregnancy. This was a prospective study, which compared chlorhexidine reaction with dipstick tests and urine culture in 150 asymptomatic pregnant women. Urine cultures detected bacteriuria in seven women (4.7%). Chlorhexidine detected ASB in 72 women (48%) and had sensitivity, specificity and accuracy of 100%, 54% and 56%, respectively. Leucocyte esterase (LE) and nitrite detected bacteriuria in 31 (20.7%) women and 12 (8.0%) women, respectively. Singly, LE had a sensitivity and specificity of 14.3% and 79%, respectively, while nitrite’s sensitivity and specificity was 42.9% and 93.7%, respectively. Combined, LE and nitrite had better sensitivity (97.9%) and accuracy (94%). Since the accuracy of chlorhexidine is low, other than urine culture, combined dipstick urinalysis of leucocyte esterase and nitrite tests is good to detect asymptomatic bacteriuria in pregnancy. Keywords: Asymptomatic bacteriuria, urinary tract infection
Introduction Asymptomatic bacteriuria is a precursor for urinary tract infections (UTI), which are common during pregnancy. The combination of mechanical, hormonal and physiological changes during pregnancy cause significant changes in the urinary tract, predispose to the acquisition and alters the natural history of bacteriuria during pregnancy (Abdullah and Al-Moslih 2005; Seccon et al. 2003). Also, progesterone relaxes ureteric smooth muscle causing dilatation of the ureters and stasis of urine, which is further aggravated by pressure from the expanding uterus (Perera 2009). Asymptomatic bacteriuria, generally defined as true bacteriuria, ⬎ 105 colony forming units of bacteria/ml of urine, in the absence of specific symptoms of acute urinary tract infection, occurred in 2–24% of all pregnancies in previous studies from Nigeria (Odigie and Anugweje 2010; Oyetunji et al. 2006; Eigbefoh et al. 2008; Mandara and Shittu 1999). It has been reported in 7.3% of pregnant women in Ghana (Turpin et al. 2007). The prevalence is most closely related to low socioeconomic status, which is a common predisposition in both pregnant and non-pregnant women (Smaill and Vazquez 2007). Due to the high prevalence, a systematic review concluded that screening for and aggressively
treating pregnant women for asymptomatic bacteriuria would possibly decrease the annual incidence of pyelonephritis during pregnancy by up to 80% (Smaill and Vazquez 2007). The organisms that cause urinary tract infection are the same as those found in non-pregnant patients. Escherichia coli accounts for 80–90% of infections. Other Gram-negative rods such as Proteus mirabilis and Klebsiella pneumonia have also been cultured (Perera 2009; Eigbefoh et al. 2008; Okonofua 1989). Asymptomatic bacteriuria precedes cystitis or acute pyelonephritis. In addition, it is associated with pre-term labour, low birth weight, prematurity, pre-eclampsia and chronic renal disease, with attendant adverse obstetric and perinatal outcome (Eigbefoh et al. 2008; Smaill and Vazquez 2007). Therefore, screening for and treatment of asymptomatic bacteriuria to prevent pyelonephritis, has been shown to be cost-effective (Rouse et al. 1995). Urine culture, although the gold standard for bacteriuria, is expensive, time-consuming and laborious. It also needs microbiology laboratory infrastructure and qualified staff. Moreover, although dipsticks are cheap and rapid tests, which require little expertise, they have repeatedly been shown to be unreliable diagnostic tools for asymptomatic bacteriuria (Kacmaz et al. 2006; Khattak et al. 2004; Millar et al. 2000). The chlorhexidine test is cheap, simple to use by middle-level health personnel and requires no microscope or electricity. When used to detect bacteriuria in patients with urinary tract infection, it was very sensitive but had low accuracy and specificity, which made the authors recommend its use as a screening test (Okonkwo et al. 2006). When chlorhexidine is added to a suspension of bacterial cells, it is immediately adsorbed on the bacterial surface (Okonkwo et al. 2006). Even at low concentrations, there is a rapid and irreversible loss of cytoplasmic constitutions, which causes precipitation of cellular protein and nucleic acid (Okonkwo et al. 2006). While most clinicians agree on the need for early urine screening, the best screening test, other than urine culture, is yet to be determined. The search, however, for a cheap, easy and acceptable screening method is continuing. This was the basis for this research, which evaluated chlorhexidine urine test for screening of asymptomatic bacteriuria in pregnant women. The study objective was to determine the accuracy and correlation of chlorhexidine reaction test when compared with dipstick tests (leukocyte esterase and nitrite) and urine culture. The cost–benefit of the different screening tests was also determined.
Correspondence: B. O. Okusanya, Department of Obstetrics and Gynaecology, College of Medicine, University of Lagos, Idi-Araba, Lagos. E-mail: [email protected]
22
B. O. Okusanya et al.
J Obstet Gynaecol Downloaded from informahealthcare.com by Nyu Medical Center on 06/14/15 For personal use only.
Subjects and methods This was a prospective observational study conducted at the antenatal clinic and laboratory of Irrua Specialist Teaching Hospital (ISTH). Pregnant women were included if they had no dysuria, urine frequency, suprapubic and loin pain or vaginal discharge. In addition to the presence of any of the aforementioned symptoms, women using antibiotics for any reason or who had antibiotic treatment in the preceding 2 weeks prior to counselling on the study, were excluded. At a prevalence of 10% (Smail and Vazquez 2007) of asymptomatic bacteriuria, the sample size of 138 pregnant women was needed, based on a formula (Araoye 2003). However, 150 consecutive pregnant women attending the antenatal booking clinic who met the inclusion criteria were recruited after they gave informed consent. The Research Ethic Committee of ISTH approved the study. The method of collection of urine was explained to the women and they were assisted by a nurse if needed, in order to reduce the chances of contamination. Each patient was given a sterile, dry, wide mouth container, with an instruction to clean the vulva with clean water. Urine was then collected (approx. 20 ml) by midstream clean catch technique. The urine sample was divided into three containers (one for each screening method under consideration) and labelled accordingly. The samples to be sent to the laboratory were transported to the laboratory immediately and processed within 1 h. When there was any delay, the samples were refrigerated at 4°C. The urine specimens were processed by routine quantitative culture and were also tested by the screening methods under consideration.
Dipstick test – leukocyte esterase and nitrite This was performed within 5 min of passing urine. Urine specimen was tested with colorimetric Multistix 10 SG (Ames Division Miles Inc., Elkhart, IN) for the presence of nitrite and leukocyte esterase activity following the manufacturer’s instructions. The nitrite and leukocyte esterase portions of the test were interpreted as positive if the colour on the reagent square was positive for each portion. Tests that showed no colour change or trace results were considered as negative.
Chlorhexidine reaction In this chemical test, chlorhexidine 1:1000 solution in water was utilised. Using a laboratory dropper, 10 drops of urine from a container was placed in a clean test-tube; the colour and appearance of the urine was noted. Five drops of chlorhexidine was added to the test tube and as this was being done, the urine in the test tube was observed for any changes, e.g. cloudiness, precipitate or particulate matter. The entire contents of the test-tube was then shaken and observed after 1 min. A cloudy colour indicated a positive result. Half the volume of chlorhexidine (5 drops) was titrated against the full volume of urine (10 drops) because this was found in an earlier study to be the critical point were the colour changes were maximal (Okonkwo et al. 2006). Two laboratory scientists performed the urine culture in the laboratory, while the researcher and two trained resident doctors performed the chlorhexidine reaction test and dipstick test for leukocyte esterase and nitrite. This was done at the antenatal clinic. Chlorhexidine reaction test was evaluated using sensitivity, specificity, the positive predictive value, negative predictive value and accuracy rate, when compared with dipsticks and urine culture.
A questionnaire was used to document participants’ sociodemographic characteristics and the results of the screening tests before they were analysed. Data were analysed using the Statistical Package for Social Science v. 16 (SPSS Inc., IL). The various tests of validity were subsequently calculated.
Results A total of 150 pregnant women were recruited during the period of study. Seven women had positive urine cultures. The prevalence of asymptomatic bacteriuria in this study was therefore 4.7%. The sociodemographic characteristics of the women are in Table I. Using urine culture as the gold standard, the diagnostic performance of chlorhexidine reaction test was evaluated. Dipstick detected asymptomatic bacteriuria in 31 women (20.7%) using leucocyte esterase, and in 12 women (8.0%), with nitrite alone. The sensitivity and specificity of leucocyte esterase test was 14.3% and 79.0%, respectively, while nitrite sensitivity and specificity was 42.9% and 93.7%, respectively. The combined test (leucocyte esterase and nitrite) confirmed one out of the seven urine-culture positive samples (14.3% sensitive), missed six urine-culture positive samples and was 97.9% specific. Furthermore, the positive and negative predictive value of leucocyte esterase was 3.2% and 95%, respectively; that of nitrite was 25% and 97.1% and the combined test had positive and negative predictive values of 25% and 96%, respectively. Leucocyte esterase was 76% accurate, while the accuracy for nitrite was 91%. When combined, both tests had a slightly higher accuracy of 94%. This implies that the combined test Table I. Sociodemographic characteristics of sample population. Characteristics Parity 0 1 2 3 4 ⱖ5 Age (years) 20–24 25–29 30–34 35–39 ⱖ 40 Occupation Housewife Civil servant Trading Student Others Ethnicity Esan Afemai Edo Igbo Yoruba Others Level of education Primary Secondary Tertiary Marital status Married Single
n
(%)
35 56 30 18 10 1
23.3 37.3 20 12 6.7 7
22 52 48 24 4
14.7 34.7 32 16 2.7
20 41 56 30 3
13.3 27.3 37.3 20 2
98 18 13 10 7 4
65.3 12 8.6 6.7 4.7 2.7
3 59 90
2 38 60
150 0
100 0
J Obstet Gynaecol Downloaded from informahealthcare.com by Nyu Medical Center on 06/14/15 For personal use only.
Chlorhexidine test for asymptomatic bacteriuria in pregnancy 23 will reliably diagnose asymptomatic bacteriuria in nine out of every 10 women tested. Unlike dipsticks, chlorhexidine reaction test detected asymptomatic bacteriuria in 72 (48%) women. All seven urine-culture positive samples were detected by chlorhexidine reaction test (100% sensitive), and it was 54% specific. It had a positive and negative predictive value of 9.7% and 100%, respectively, an accuracy rate of 56%, and a false negative rate of 0%. Isolated organisms were Gram-negative with Escherichia coli being the predominant organism isolated (57.1%). Others were Klebsiella species (14.3%), Proteus species (14.3%). The only Gram-positive organism was Staphylococcus aureus (14.3%). A total of 85.7% of organisms isolated were sensitive to Nitrofurantoin, while 57.1% of isolated organisms were sensitive to Gentamicin, Ofloxacin and Augmentin®. The isolated organisms were all poorly-sensitive to Amoxicillin. Chlorhexidine reaction was done in 1–2 min, the same time as for dipsticks, but the method is much cheaper than dipsticks (US$0.31 vs US$1.25). Urine cultures cost US$3.13 and the results were not ready for use until 48 h after sample collection.
Discussion Asymptomatic bacteriuria is the presence of actively multiplying bacteria at a time when the patient has no urinary symptoms, so that diagnosis relies upon microbiological findings (GaringaloMolina 2000). An early detection and treatment of asymptomatic bacteriuria may be of considerable importance, not only to forestall acute pyelonephritis and chronic renal failure in the mother, but also to reduce prematurity and perinatal mortality in the newborn (Perera 2009). In this study, four main tests of validity were used to compare urine leucocyte esterase and nitrite tests singly, and in combination with chlorhexidine test. Urine culture served as the gold standard for their ability to detect asymptomatic bacteriuria. Of the dipsticks, nitrite had the highest sensitivity (42.9%). Although low, what we found fell within the often-quoted range of 35–85% for the nitrite test (Khattak et al. 2004; Millar et al. 2000). Since sensitivity is a measure of the false-negative rate of 57.1%, the nitrite test will test negative in approximately six out of every 10 women with significant bacteriuria. This is likely to be due to bacteriuria of organisms not having nitrate reductase. This is seen when the urine has only stayed in the bladder for ⬍ 4 h, such that nitrate is yet to be reduced to nitrite, or when dietary nitrite is absent (Kacmaz et al. 2006). Thus, it is best that the first voided urine in the morning be used (Jayalakshmi and Jayaram 2008), as it has been proven to be accurate, albeit morning urine collection was not practicable in our patient population, since the study enrolled outpatient pregnant women. This may explain the low sensitivity. Conversely, nitrite test’s high specificity meant that if the test was negative for bacteriuria, it was accurate in nine of 10 women. It should be borne in mind that a positive nitrite test indicates that nitrite has been produced by the reduction of nitrate by enteric bacteria, most commonly from the Enterobacteriaceae family (Kacmaz et al. 2006). The tests of validity for leucocyte esterase and nitrite suggest that they are unreliable when used alone to detect asymptomatic bacteriuria. However, when combined, the high specificity and accuracy rate of 95.9% mean that almost every woman with asymptomatic bacteriuria will be detected. Hence, when dipstick is the only option available, a combined test must be done to detect asymptomatic bacteriuria. Chlorhexidine reaction accurately detected asymptomatic bacteriuria in all women who had positive urine culture. The
mechanism of the reaction which precipitates cellular protein and nucleic acid explains this. Though the sensitivity of chlorhexidine was high, and correlates with an earlier study which involved symptomatic women (Okonkwo et al. 2006), its specificity of 54.6% and accuracy of 56% meant that the chlorhexidine test cannot replace urine dipsticks in routine use for the detection of asymptomatic bacteriuria. This is because chlorhexidine also reacts with other components of urine, such as cells and crystals (Okonkwo et al. 2006). Regarding the methods and costs, just as dipstick tests, chlorhexidine tests are easy to perform and do not need laboratory or trained personnel. The results of both tests were ready within 1–2 min. The cost of chlorhexidine reaction, dipstick urinalysis and urine culture was, respectively, 50 Naira (30 cents), 200 Naira (US$ 1.3) and 500 Naira (US$ 3.3). The low accuracy negates the cheaper price of chlorhexidine. This implies that other than urine culture, dipstick tests of leucocyte esterase and nitrites will remain the choice, especially when combined to detect asymptomatic bacteriuria in pregnant women. The prevalence of asymptomatic bacteriuria of 4.7% of this study, was much lower than the 22% prevalence found 6 years earlier in the same institution (Eigbefoh et al. 2008). In the 6 years between the two studies, the community has transited from predominantly rural to a suburban setting. Since the prevalence of bacteriuria is closely related to socioeconomic status (Smaill and Vazquez 2007), improvement in the status of women attending the antenatal clinic may be responsible for the different prevalence. The most common organism isolated was Escherichia coli. This conforms to previous studies, which showed that Escherichia coli is the most common cause of bacteriuria in pregnancy (Ezechi et al. 2003; Abdul and Onile 2001). Other organisms isolated were Klebsiella species, Proteus species and Staphylococcus aureus, as previously reported (Ezechi et al. 2003; Abdul and Onile 2001; Akerele et al. 2001). When bacteriuria is detected in pregnant women, they should be treated. The choice of antibiotic should be effective for the most common infective organisms (i.e. Gram-negative gastrointestinal organisms) and should be safe for the mother and fetus. Nitrofurantoin was identified in this study as having the highest and widest spectrum of activity against the organisms responsible for bacteriuria. This was a good choice, since nitrofurantoin has high urinary concentration (Sanders and Sanders 1992). Gentamycin, Augmentin and Ofloxacin are other effective antibiotics in this study. Ofloxacin, a fluoroquinolone, although with fairly good spectrum, has possible toxic effects on the fetus and should not be prescribed to pregnant women. Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
References Abdul IF, Onile BA. 2001. Bacterial isolates from urine of women in Ilorin and their antibiotic susceptibility patterns. Tropical Journal of Obstetrics and Gynaecology 18:61–65. Abdullah AA, Al-Moslih MI. 2005. Prevalence of asymptomatic bacteriuria in pregnant women in Sharjah, United Arab Emirates. East Mediterranean Health Journal 11:1045–1052. Akerele J, Abhulimen P, Okonofua F. 2001. Prevalence of asymptomatic bacteriuria among pregnant women in Benin City, Nigeria. Journal of Obstetrics and Gynaecology 21:141–144. Araoye MO (ed.). 2003. Subject selection. In: Research methodology with statistics for health and social sciences. Ilorin: Nathadex Publications. p. 115–139.
J Obstet Gynaecol Downloaded from informahealthcare.com by Nyu Medical Center on 06/14/15 For personal use only.
24
B. O. Okusanya et al.
Eigbefoh JO, Isabu P, Okpere E, Abebe J. 2008. The diagnostic accuracy of the rapid dipstick test to predict asymptomatic urinary tract infection of pregnancy. Journal of Obstetrics and Gynaecology 28:1–6. Ezechi OC, Fasuba OB, Dare FO. 2003. Antibiotic sensitivity pattern of microbial isolates from the urine of pregnant women with urinary tract infections. Tropical Journal of Obstetrics and Gynaecology 20: 133–135. Garingalo-Molina FD. 2000. Asymptomatic bacteriuria among pregnant women: Overview of diagnostic approaches. Philippines Journal of Microbiology and Infectious Disease 29:177–186. Jayalakshmi J, Jayaram VS. 2008. Evaluation of various screening tests to detect asymptomatic bacteriuria in pregnant women. Indian Journal of Pathology and Microbiology 51:379–381. Kacmaz B, Cakir O, Aksoy A, Biri A. 2006. Evaluation of rapid urine screening tests to detect asymptomatic bacteriuria in pregnancy. Japan Journal of Infectious Diseases 59:261–263. Khattak AM, Khan H, Akhtar W, Mahsud I, Ashiq B. 2004. Accuracy of Griess test to predict asymptomatic bacteriuria during pregnancy. Gomal Journal of Medical Sciences 2:20–23. Mandara MU, Shittu SO.1999. Asymptomatic bacteriuria in antenatal patients at Ahmadu Bello University Teaching Hospital Zaria, Nigeria. Tropical Journal of Obstetrics and Gynaecology 16:41–45. Millar L, Debuque L, Leialoha C, Grandinetti A, Killeen J. 2000. Rapid enzymatic urine screening test to detect bacteriuria in pregnancy. Obstetrics and Gynaecology 95:601–604. Odigie JO, Anugweje KC. 2010. Asymptomatic bacteriuria in pregnancy in Port Harcourt. Asian Pacific Journal of Tropical Medicine 3:580–583.
Okonkwo CA, Okpere EE, Ande AB. 2006. Evaluation of chlorhexidine in the detection of bacteriuria in pregnancy. Tropical Journal of Obstetrics and Gynaecology 23:14–16. Okonofua FE, Adediran AA, Okonofua B. 1989. Incidence and pattern of asymptomatic bacteriuria of pregnancy in Nigerian women. Nigerian Medical Practitioner 17:35–38. Oyetunji JA, Ahmed Y, Nwobodo EI, Airede LR, Ekele BA. 2006. Asymptomatic bacteriuria in pregnancy in Sokoto, Nigeria. Sahel Medical Journal 9:1–6. Perera J. 2009. Asymptomatic bacteriuria in pregnancy. Sri Lanka journal of Obstetrics and Gynaecology 31:108–109. Rouse DJ, Andrews WW, Goldenberg RL, Owen J. 1995. Screening and treatment of asymptomatic bacteriuria in pregnancy to prevent pyelonephritis: a cost effectiveness and cost- beneficial analysis. Obstetrics and Gynaecology 86:119–123. Sanders CC, Sanders NE Jnr. 1992. Beta lactam resistance in gram negative bacteria: Global trends and clinical impact. Clinical Infectious Diseases 15:824–839. Seccon NC, Garingalo-Molina FD, Yoasiano CJ, Saniel MC, Manalastaas RM. 2003. Prevalence of asymptomatic bacteriuria and associated risk factors in pregnant women. Philippines Journal of Microbiology and Infectious Disease 32:63–69. Smaill F, Vazquez JC. 2007. Antibiotics for asymptomatic bacteriuria in pregnancy. Cochrane Database of Systematic Reviews (2):CD000490. Turpin CA, Minkah B, Danso KA, Frimpong EH. 2007. Asymptomatic bacteriuria in pregnant women attending Antenatal Clinic at Komfo Anokye Teaching Hospital, Kumasi, Ghana. Ghana Medical Journal 41:26–29.
