Pyuria as a diagnostic test for urinary tract infection in children with sickle cell anaemia in Zaria Nigeria.html

Niger J Paed 2015; 42 (1): 17 –20

ORIGINAL

Yauba MS

Pyuria as a diagnostic test for

Aikhionbare HA

urinary tract infection in children

Ogunrinde GO

Bugaje MA

with sickle cell anaemia in Zaria,

Nigeria

DOI:http://dx.doi.org/10.4314/njp.v42i1,5

Accepted: 3rd November 2013

Abstract: Objective: This study

in steady state (4/185; 2.2%) and

aimed at determining the signifi-

the difference was statistically

Yauba MS

(

)

cance of pyuria as a screening test

significant (^2χ = 27.323, p =

Aikhionbare HA, Ogunrinde GO

for UTI in children with sickle

0.001). Of the 22 SCA subjects

Bugaje MA

Department of Paediatrics,

cell anaemia (SCA).

with confirmed UTI, 19 (86.4%)

Ahmadu Bello University Teaching

Methods: Two hundred and sev-

had significant pyuria with a sensi-

Hospital,

enty-two children with SCA, aged

tivity of 86.4%, specificity of

PMB 06, Shika-Zaria,

6 months to 15 years, were stud-

68.8%, and positive predictive

Kaduna state, Nigeria.

ied out of which 185 (68.0%)

value of 19.6%. The most common

Email: saadyko@yahoo.co.uk

were in stable state and 87

bacterial isolates were Escherichia

(32.0%) were in crises. Their

coli 11 (50.0%) while the least

urine was assessed for microscopy

isolated were Salmonella typhi 1

culture and sensitivity. Significant

(4.6%).

pyuria and bacteriuria was deter-

Conclusion: This study indicates

mined using standard methods.

that pyuria as a screening test for

Results:

Significant

bacteriuria

UTI is not very specific but very

was detected in 22 (8.1%) of the

sensitive necessitating a more test

272 subjects with SCA. The

to be done for diagnosis of UTI.

prevalence of significant bacteri-

uria was higher among those in

Keywords : Sickle cell anaemia,

crisis (18/87; 20.7%) than in those

bacteriuria, pyuria, children

Introduction

uria was reported to be 55.4% and 22.2% respectively ⁷ .

In comparison, Akor et al in Jos, reported a much lower

Urinary tract infection (UTI) is an important contribut-

prevalence (1.5%) of significant pyuria in 650 appar-

ing factor to morbidity and mortality in children with

ently healthy primary school children aged 5 – 12 years.

SCA and may be a precipitating event for the crises as-

Therefore, this study was undertaken to determine the

sociated with the disease . Bacteriuria, presence of bac-

significance of pyuria as a screening test for UTI in chil-

teria in urine not due to contamination from sample col-

dren with SCA seen in ABUTH.

lection, can be symptomatic or asymptomatic. Bacteri-

uria can resolve spontaneously, especially when asymp-

tomatic, or can progress to chronic nephropathies like

chronic pyelonephritis, particularly in situations of al-

Subjects and methods

tered immunity such as sickle cell anaemia . Very often,

the dramatic manifestations of vasoocclusive crisis and

The study was a prospective, descriptive and cross-

life-threatening anaemia cause the attending physician to

sectional one. The study sample population consisted of

overlook coexisting infection . Urinary tract infections

consecutively selected children with SCA (in steady

(UTI) in children with sickle cell anemia (SCA) may

state and in crisis) aged 6 months to 15 years, seen in the

result in long term renal dysfunction .

Department of Paediatrics, Ahmadu Bello University

Teaching Hospital (ABUTH), Zaria, over a period of six

Various methods of screening for UTI have been em-

months. Children with SCA who had been on antibiotics

ployed globally but the most commonly used in this part

one week preceding enrolment into the study, those with

of the world is urinalysis either by dipstick or direct

confirmed (or suspected) congenital urogenital anoma-

microscopic examination of centrifuged urinary sedi-

lies and those who had recent (< 1 week) manipulative

ments . Pyuria has been found to be a useful tool in the

urogenital procedure (like catheterization and cysto-

diagnosis of UTI. The prevalence of significant pyuria in

scopy) were excluded from the study. Also those whose

children with SCA with bacteriura and without bacteri-

parents or guardians did not consent and those with

HbSC and other forms of sickle cell disease other than

subjects in steady state was 6.8 ±3.9 years while that of

SCA were excluded.

those in crisis was 5.6 ±3.7 years and they were not sig-

Ethical approval was obtained from the ABUTH Re-

nificantly different from each other. There were 156

search Committee and a written consent obtained from/

(57.4%) males and 116 (42.6%) females with male to

guardians of subjects.

female ratio of 1.3: 1 (Table I). There was no statisti-

cally significant difference in the distribution of age

groups between the sexes ( χ =0.59, df=2, p=0.745). One

Sample collection, transportation and culture

hundred and eighty five (68.0%) SCA subjects were in

For each patient that has achieved bladder control mid-

steady state and 87 (32.0%) were in crisis. (Table 1)

stream urine specimen (5mls) was aseptically collected

at the time of presentation into two sterile universal bot-

Table 1: Age and sex distribution of SCA subjects in crisis

tles. For adolescent girls a trained female resident doctor

and in steady state

assisted in collection of the specimens. For infants,

In steady state, n (%)

In crisis, n (%)

moribund patients and children who were not toilet

Ages

Males

Females

Males

Females

Total

trained, suprapubic bladder aspiration (SPA) was used

(Years)

for obtaining the urine specimens. All urine specimens

< 5

35 (33.3)

27 (33.8)

26 (51.0)

18 (50.0)

106 (84.1)

were obtained under aseptic procedures as described by

5-9

40 (38.1)

35 (43.8)

16 (31.4)

11 (30.6)

102 (71.5)

Anochie et al , as follows: each subject’s external geni-

10-15

30 (28.6)

18 (22.5)

9 (17.6)

7 (19.4)

64 (44.3)

Total

105(100)

80 (100)

51 (100)

36 (100)

272 (100)

talia was cleaned, midstream urine was collected, stored

in a refrigerator and then submitted to the laboratory

within an hour of collection. This procedure was carried

Of the 185 SCA children in steady state, 62 (33.5%)

out in order to minimize bacterial contamination of the

were <5 years of age and 75 (40.5%) and 48 (25.9%)

urine. Samples were analysed within an hour of collec-

were 5-9 and 10-15 years of ages respectively. The

tion, otherwise they were kept in a refrigerator and or

corresponding proportions in children in crisis were

preserved in boric acid and analysed within 3-5 hours.

50.6%, 31.0% and 18.4% (Table I). There was a statisti-

Five milliliters of the urine was centrifuged at the main

cally significant difference in the proportions of children

microbiology laboratory of the teaching hospital at 2000

in the lowest age group; the proportion of those in

rpm for 5 minutes; a wet preparation was made from the

steady state (33.5%) being less than the one for those

with crisis (50.6%; χ = 7.28, df=2, p = 0.026). Of the 87

sediment and examined under the microscope at x40

magnification. More than five pus cells per high power

subjects with SCA in crisis, 55 (63.2%) and 32 (36.8%)

field (HPF) were regarded as significant pyuria. A por-

had vaso-occlusive and anaemic crises respectively.

tion of another aliquot was inoculated onto blood and

Their corresponding mean ages were 5.5 (±3.6) and 5.0

MacConkey agar plates and incubated aerobically at

(±4.4) respectively and were not significantly different

37 C for 48 to 72 hours. A pure colony count of ≥ 10

from each other.

organisms/ ml of urine were considered a significant

growth. Other sets of culture plates were incubated in

Of the 272 children with SCA studied, 22 had signifi-

carbon dioxide extinction jar at the same temperature for

cant bacteriuria giving an overall prevalence of 8.1%.

isolation of anaerobes. In case of significant bacteriuria,

Four (2.2%) of the 185 SCA subjects in steady state had

systematic bacteriology and biochemical tests using

significant bacteriuria while 18 (20.7%) of those in

standard techniques; catalase, oxidase, sugar fermenta-

crises had significant bacteriuria, the difference in

respective prevalence being statistically significant ( χ =

tion, motility, urease, citrate, indole, hydrogen sulfide

and gas production were carried out based on bacterial

27.3, df = 1, p < 0.001). A total of 22 bacteria were iso-

gram reactions. Antimicrobial sensitivity test were car-

lated from the urine samples, mainly gram negative or-

ried out using modified Kirby-Bauer’s diffusion meth-

ganisms. The most frequently isolated organisms were

ods where zones of inhibition were measured.

Those

Escherichia coli (11; 50.0%) and Klebsiella pneumoniae

with positive culture results were treated accordingly.

(5; 22.7%). The least common isolates were Staphylo-

coccus aureus (2; 9.1%) and Salmonella typhi (1; 4.6%;

Data analysis

see Table 2). The organisms were mainly sensitive to

ceftriazone and resistant to co-trimoxazole.

Data was analyzed using Epi Info version 3.5.3 statisti-

cal software. Values were expressed as frequency, mean

Table 2: Urinary bacterial isolates among patients

and standard deviation. Chi-square test was used to de-

Organisms

Frequency (%)

termine the level of significance. P-values less than 0.05

Escherichia coli

11 (50.0)

were considered significant.

Klebsiella species

5 (22.7)

Proteus species

3 (13.6)

Staphylococcus aureus

2 (9.1)

Salmonella typhi

1 (4.6)

Results

Total

22 (100.0)

A total of 272 subjects with SCA were studied and their

ages ranged from six months to 15 years with a mean

Of the 272 children with SCA, 97 (35.7%) had signifi-

age (± 1 SD) of 6.4 ±3.8 years. The mean age of SCA

cant Pyuria. There were 56 (35.9%) males and 41

(35.3%) females and the difference was not statistically

significance (p= 0.97). Pyuria was commonly found

age group. This study also revealed that pyuria was

among the under-fives occurring in 45 (46.4%) children

commoner among boys than girls and this may be due to

while 29 (29.8%) and 23 (23.7%) children with pyuria

the fact that infection is generally commoner among

were among the 5-9 years and 9-15 years respectively.

boys than girls as a result of genetic variations between

Of the 22 children with sickle cell anaemia and signifi-

the two sexes.

cant bacteriuria 19 (86.4%) had significant pyuria, while

only 78 (31.2%) out of 250 with no bacteriuria had sig-

This study revealed pyuria with a high sensitivity but

nificant pyuria. The screening test had sensitivity, speci-

low specificity of 86.4% and 68.8% respectively. These

ficity and positive predictive values of 86.4%, 68.8%

finding indicates that presence of pyuria is highly sug-

and 19.6% respectively. Pyuria was associated with

gestive of UTI but absence of pyuria does not exclude

increased risk of bacteriuria (p<0.001) (Table 3).

UTI. The positive predictive value revealed that only

19.6% of patients with confirmed UTI have the disease.

Table 3: Urine microscopy (pyuria) and culture

This suggests that pyuria is not a worthwhile screening

Urine culture

test for UTI. However, the absence of pyuria in children

Pyuria

Significant

Insignificant

Total

with UTI is rare but can occur when a child is been

growth (%)

evaluated so early in the course of the infection when

Significant pyuria

19 (86.4)

78 (31.2)

97 (35.7)

the inflammatory response has not yet developed.

Insignificant pyuria

3 (13.6)

172 (68.8)

175 (64.3)

Total

22 (100)

250 (100)

272 (100)

Many workers, on the contrary, reported low sensitivity

of pyuria among children with UTI.

7,14

The higher sensi-

p<0.001 ( χ 2)

tivity reported in this study as compared to the 55.4%

Sensitivity = 86.4%

reported by Mava et al may be due to the fact that most

Specificity = 68.8%

Positive predictive value = 19.6%

of the SCA children with bacteriuria in this study were

Negative predictive value = 98.3%

in SCA crises which could have been precipitated by

UTI. The negative and positive predictive values, for

this study, were 98.3% and 19.6% respectively. This

Discussion

implies that to use significant pyuria to diagnose or pre-

The overall prevalence of significant bacteriuria in chil-

dict bacteriuria in children with SCA will result in sig-

dren with SCA in this study was found to be 8.1%. This

nificantly large numbers of false-positive and false-

agrees with the study by Tarry et al , in USA, in which

negative results. The high sensitivity obtained in this

study is comparable to that of Smith et al

an overall prevalence of 9.0% bacteriuria was deter-

as well as

Blum et al

mined among children aged 1-18 years with SCA. Chil-

who reported sensitivities of significant

dren in steady state had a significantly lower rates

pyuria to be 64% and 98% among subjects with signifi-

(2.2%) compared to those in crises. This was lower than

cant bacteriuria. The specificity in our study was low in

that reported by Chukwu et al

who documented a

contrast with specificity of 68% reported by Blum et

higher prevalence of 6% among steady state SCA chil-

al.

The use of symptomatic women as their sample

dren aged 2 – 12 years. The slightly lower prevalence

population for the study may contribute to the observed

observed in this study could be attributed to the varia-

significant difference.

tion in the age distribution in the two studies as well as

differences in the environmental and genetic constitution

of the study populations.

Conclusion

The 35.7% prevalence of significant pyuria in this study

was lower than that obtained by Mava et al where they

This study has demonstrated that presence of pyuria is

reported the prevalence of pyuria to be 55.4% among 65

only suggestive of UTI but it is not a reliable screening

SCA children with bacteriuria aged 6 months to 12

test for UTI in children with SCA in crises as revealed

years. This may be due to the fact that most children

by the very low positive predictive value. Routine

with bacteriuria in steady state may have contributed to

screening of all such cases for pyuria is, therefore, not

the lower prevalence as Mava et al studied febrile SCA

strongly advocated.

children. The relationship between pyuria and bacteri-

uria was found to be statistically significant since pyuria

Authors Contributors

was found to be associated with increased risk of bacte-

riuria (p< 0.001). The lower prevalence, 31.2%, of sig-

MS conceived the study, participated in the design and

nificant pyuria among SCA children without bacteriuria

coordination of the study, collected the samples, contrib-

obtained in this study was similar to that reported by

uted in the laboratory work, analyzed the result and

Mava et al , 22.2%, among similar group of SCA chil-

wrote the final manuscript. HA, GO and MA partici-

dren. However, the much lower prevalence obtained by

pated in the design, writing, analyzing and supervision

Akor et al could be because children less than 5 years

of the study. A participated in the design and coordina-

of age, who were more prone to developing UTI com-

tion of the study and collected the samples. A assisted

pared to older children, were not included in the study.

in the aspect of laboratory coordination of the study.

The finding of significant pyuria to be commoner among

All authors read and approved the final manuscript.

children less than five years of age in this study might

Conflict of interest: None

be explained by the immune-compromised nature of this

Funding: None

Acknowledgements

immense technical assistance as well as Dr. Shamsu of

the department of microbiology of ABUTH for assisting

We wish to acknowledge Dr. Ibrahim MS of the depart-

in the laboratory aspect of this study.

ment of community medicine and Dr Ali AZ for their

References

Mark T, Gladwin MD, Elliott V.

Knox-Macaulay. Historical Intro-

10. Smith P, Morris A, Reller LB.

Pulmonary complications of sickle

duction, Molecular Biology and

Predicting urine culture results by

cell disease. N Engl J Med 2008;

Inheritance. In: A. F Fleming (ed).

dipstick testing and phase contrast

359: 2254-65.

Sickle Cell Disease: A Handbook

microscopy. Pathol. 2003; 35:161-

Okafor HU, Okoro BA, Ibe BC,

for the General Clinicians. Church

Njoku-Obi NU. Prevalence of

hill Livingstone: Edinburgh, 1982:

11. Tarry WF, Dukket JW, Mc Synder.

asymptomatic bacteriuria among

1-17.

Urological complications of sickle

nursery school children. Nig J

Mava Y, Timothy SY, Zailani SB,

cell disease in a paediatric popula-

Paediatr 1993; 20: 84-88.

Ambe JP. Significance of pyuria in

tion. J Urol 1987; 138: 592-594 .

Asscher AW. Natural history of

the diagnosis of urinary tract infec-

12. Chuckwu BF, Okafor HU, Ikefuna

significant bacteriuria. Proc Roy

tion in Children with sickle cell

AN. Asymptomatic bacteriuria in

Soc Med 1977; 70: 1439-1450.

anaemia in maiduguri, Nigeria.

children with sickle cell anaemia at

Akuse RM. The pattern of bacte-

Afr. J.Cln.Exper. Microbiol 2012;

University of Nigeria Teaching

rial infection and intercurrent mor-

13: 99-105. http://

Hospital, Enugu, South East, Nige-

bidity in sickle cell disease pa-

dx.doi.org/10.4314/ajcem.v13i2.8

ria. Italian J Paed 2011: 37-45.

tients. A dissertation submitted to

Akor F, Okolo FN, Agaba EI,

13. Blum RN, Wright RA. Detection

the National Postgraduate Medical

Okolo A. Urine examination find-

of pyuria and bacteriuria in symp-

College of Nigeria in part fulfill-

ings in apparently healthy new

tomatic ambulatory women. J.

ment of the requirements for the

school entrants in Jos, Nigeria.

Gen. Intern. Med. 1992; 7: 140-4

Fellowship of the College 1990:

SAJCH July 2009; 3: 60-63

14. Akinyanju OO. Profile of sickle

318-323.

Anochie IC, Nkanginieme KEO,

cell disease in Nigeria. Ann N Y

Young SP. Renal scar formation

Eke FU. The influence of instruc-

Acad Sci 1989; 565: 126-136

after urinary tract infection in

tion about the methods of urine

children. Korean J Pediatr. 2012;

collection and storage on preva-

55: 367–370.

lence of urinary tract infection. Nig

J Paediatr 2002; 28: 39-42.