ORIGINALNiger J Paed 2016; 43 (1):30 –33

Ocheke OIJohn CC,Ogbe P,Donli A,Oguche S

The febrile child: how frequentshould we investigate for urinarytract infection

Accepted: 3rd August 2015

Ocheke OIJohn CC, Ogbe P, Donli A, Oguche SDepartment of Paediatrics,Jos University Teaching HospitalP. M. B 2076, Jos, NigeriaEmail: ieocheke@yahoo.com

( )

DOI:http://dx.doi.org/10.4314/njp.v43i1.6

Abstract: Background: Febrileillness in children remains themost common cause of emer-gency room visit. In many tropicalcountries where malaria is en-demic, children presenting withfever are treated for malaria pre-sumptuously. Current evidencesuggests however that malarialparasitaemia in febrile children isdeclining and the prevalence ofother causes of fever apparentlyon the increase. Therefore, high-lighting such causes of fever asurinary tract infection (UTI) isindispensable. This is much so asUTI not only is common inyounger children and often ne-glected but also associated withlong term complicationsMethods: Children aged 6-59months with fever of less than2weeks were consecutively re-cruited. Each child had both clini-cal evaluation and preliminarylaboratory assessment such asdipstick urinalysis. Further micro-biological and radiologicalevaluations were performed

where necessary: blood film formalarial parasite identification andcount, cerebrospinal fluid (CSF)analysis and chest X-ray.Results: Of the 303 children 180(59.4%) were males and 123 werefemales (40.6%). The mean agewas 21.7±14.0months, 54.5% wereless than 24months.ARI accounted for 44.6% (mainlytonsillitis, 61%, pneumonia, 27%and otitis media, 12%), while ma-laria and UTI were observed in38.3% and 4.6% respectively.Five (35.7%) patients with UTIwere males while 9 (64.3%) werefemales. Their combined mean agewas 25.4±18.6months, 57% ofthese children were less than 24months old. In 3(21.4%), UTI co-existed with malaria.Conclusions: Acute respiratoryinfection, malaria and UTI are thethree leading causes of fever inchildren under 5 years.

Keywords: Fever, Children, Acuterespiratory infection, Malaria,UTI.

malaria parasitaemia accounted for only 35 % of 5,217children presenting with fever.8

The pre-treatment identification of malaria parasites in asuspected case is beneficial in many ways much as onlyindividuals with the infection are treated. A new prob-lem as a result of this step however is, what appropriatetreatment would those who test negative for Plasmo-dium falciparum parasites receive, particularly at thefirst and secondary-level health facilities. Available evi-dence suggests that with the decrease in the use of anti-malarials, there has been a corresponding increase in theindiscriminate use of antibiotics in febrile children.9,10

Many times, the choice of antibiotics and their dosagesare inappropriate, inadequate and unnecessary. Thispractice can lead to increase in the development of sev-eral resistant microbial strains. It could also lead to poorcare and missed diagnosis of bacterial infections thatmay result in long term complications particularly inyounger children.11

Introduction

Febrile illness represents the most common cause ofhospital visits and admission in children globally.1,2 Theunderlying causes are predominantly infectious and UTIis increasingly recognised as an important cause.3 Ap-proximately two decades ago, it was noted that 30% ofoutpatient and 50% of inpatient children under the ageof five years living in Sub-Saharan Africa had malaria.4

Without a reliable case definition, children presentingwith febrile illness in most cases were treated presump-tively for malaria.1,5-6 Current evidence suggests how-ever that there is declining malarial transmission inmany of the malarious countries in Africa. This necessi-tated the World Health Organisation (WHO), point-of-care rapid diagnostic test which ensures that all childrenpresenting with fever are tested for malaria before em-barking on antimalarial treatment.7 In Nigeria a recentmulticentre study conducted in 2009/2010 showed that

Laboratory assessment

A dipstick urinalysis was done on a portion of urinespecimen (catheter urine and clean catch for children 2years and below, while mid-stream urine for older ones),for each subject who did not have any obvious focus ofinfection. This was to identify the presence of nitrite andor leukocyte esterase (LE). Only the urine of childrenwhose dipstick urinalysis was positive for both nitriteand LE or for either of the two were subjected to furthermicrobiological analysis according to National Institutefor Health and Care Excellence (NICE) guidelines.16

The remaining portion of such urine sample was sent tothe JUTH microbiology laboratory for microscopy, cul-ture and sensitivity within one hour of collection accord-ing to standard protocol. The confirmation of UTI wasmade only in children with a single bacterial isolate of≥105 colony forming units (CFU) per ml of mid-streamurine and 103 CFU/ml of catheter urine.17

Venous blood was obtained for peripheral smear andsubjected to Giemsa stain for Plasmodium falciparumidentification and counting. Similarly, all children withfeatures suggestive of acute central nervous system in-fections (fever with seizures, nuchal rigidity or uncon-sciousness) had lumbar punctures for cerebrospinal fluid(CSF) analysis.Data obtained were entered into EPI info version3.4.3software for analysis. The student‘t’ test was usedto compare group means and the Chi-squared test tocompare proportions. Fisher exact was used when cellscontained observations less than 5. P value less than0.05 was considered significant.

Results

A total of 303 children were recruited for the study. Onehundred and eighty (59.4%) were males while 123(40.6%) were females. Table 1 show the clinical charac-teristics of the subjects. The combined mean age forboth males and females was 21.67±14.02 while the me-dian age was 17 months. One hundred and ninety nine(65.7%) were aged less than 24 months.

Fig 1 shows the causes of fever in the study population.Acute respiratory infection, comprising tonsillitis, pneu-monia and otitis media in that order, was the leadingcause of fever in the children, constituting 44.6%. Ma-laria was the second common cause of fever followed byurinary tract infection at 38.3% and 4.6% respectively.Of the children with ARI, 93 (69%) were aged 24months and below with majority, 54 (58.1%) in thisgroup aged 12 months and less. Similarly, 86 (62.4%) ofthe children who had malaria were aged 24 months andbelow. There were 14 children with UTI, 5(35.7%) inmales and 9(64.3%) in females. Their mean age was25.4 ± 18.6 months. Eight children (57%) were 24months or less while the remaining 6 (43%) patientswere older.Three of the children with UTI also had parasitologicalevidence of malaria. The commonest organism

It has been shown that urinary tract infection (UTI) iscommon in children presenting with febrile illness par-ticularly in younger age groups, and could be associatedwith long-term complications.12,13,14These complicationscould be prevented by early diagnosis and prompt treat-ment.

Considering these, a systematic approach to case man-agement of febrile children, particularly those under-fiveyears is imperative. Such a measure would include care-ful clinical evaluation and laboratory assessment to iden-tify the underlying cause of fever in each child. Thisstudy was therefore, undertaken to highlight the underly-ing causes of fever in children six months to 59 monthswith specific emphasis on UTI identification in our hos-pital.

Materials and methods

This was prospective, cross-sectional and descriptivestudy, carried out in the paediatric emergency and outpa-tient units of the Jos University Teaching Hospital, Ni-geria. Study participants were children aged 6 to 59months seen in these units from the first of April to theend of October 2012. Ethical approval was obtainedfrom the Institution's Health Research and Ethics Com-mittee. All children who presented with fever (axillarytemperature ≥ 37.5oC), that had lasted less than twoweeks and whose parental/caregiver consent had beenobtained were consecutively recruited. Children withmoderate to severe malnutrition according to worldhealth organization (WHO)15 reference charts, thosewith sickle cell anaemia or any underlying chronic ill-nesses (renal, cardiac and chronic infection such as tu-berculosis or human immunosuppressive virus) wereexcluded. Similarly, children who had used antibioticswithin one week prior to hospital visit were not in-cluded. All the children seen during the study, includingthose who participated and those who did not weregiven appropriate medical care following careful evalua-tion.

Clinical assessment

The history of illness was obtained from the patient’sparent/caregiver. Each child had physical examination toidentify possible focus of infection or any other featurethat could assist in making diagnosis. Those who haveclinical features suggestive of acute lower respiratoryinfections had chest radiography. All participating sub-jects had their weight taken using a Seca® standing orbassinet weighing scales (infants), while height/lengthwas measured in recumbent or standing positions basedon the age of the child respectively.The mid upper arm circumference (MUAC), was meas-ured in all the children using an inelastic tape and meas-urement recorded to the nearest 0.1cm.

31

responsible for UTI was Escherichia coli in 8(57.1%).The other organisms included Klebsiella species in 3(21.4%), Staphylococcus aureus 2(14.3%) and Proteusspecies in 1(7.1%). A significantly high proportion(86.4%) of the organisms isolated were sensitive toflouroquinolones, sensitivity to gentamicin was 56.8%,while to the third generation cephalosporin (ceftriaxone)was 46.2%.

Fig 1: Underlying diagnosis of fever in the children(Others- cellulitis, chicken pox, measles and mumps)

Table 1: Characteristics of the study subjects

Discussion

From our study, acute respiratory infection was identi-fied as the leading infection in children presenting withfever, followed by malaria. Urinary tract infection wasthe third common cause of fever in these children. Thisgeneral pattern is similar to findings from other studiesin Africa and elsewhere in the world.18,19,20,21 In a studyof 418 children with fever in Gabon, Bouyou-Akotet etal18 reported UTI prevalence of 4.1%, while D’Acre-mont et al19, in his review of the causes of fever in 1005Tanzanian (Zanzibar) children 2 months to 10 years re-ported a prevalence of 5.9%. These studies also showedthat acute respiratory infection was the commonestcause of fever in the children as our study demonstrated.We found that malaria was the second most commoncause of fever similar to the report from Gabon,18 butthis contrasts with the observation from Zanzibar19

where malaria was the fourth common cause of fever.

This variation may be related to capacities for furtherlaboratory investigations, environmental and weatherconditions. For instance, in mainland Tanzania, the com-monest cause of fever among 870 paediatric and adultpatients was bacterial zoonoses while malaria was re-sponsible in only 1.6%.20 Similarly, among 1180 hospi-talized Cambodian children under 8 years with 1225febrile episodes, Chheng et al21 reported that acute respi-ratory infection was the foremost cause of fever fol-lowed by different kinds of viral infections. The geo-

graphical location may also have contributed as the find-ing from Tanzania showed; the pattern of infection infebrile children was different between the mainland andZanzibar Island.

The prevalence of UTI in our study of 4.6% is muchlower than two previous and similar studies from otherparts of Nigeria.22,23 Ibeneme et al22 reported a UTIprevalence of 11% in febrile children aged 1 to 59months in South East Nigeria. Their study included chil-dren of much younger age than ours. It is known that theprevalence of UTI is higher in younger infants, particu-larly in the first few months of life.12,14 Furthermore,urine culture for microbiological confirmation of UTIwas done for all their study population without initialscreening for nitrite and LE, while we relied on positiveurine nitrite and/or LE for further urine culture. Conver-sion of nitrate to nitrite does not occur with all bacteriaand it takes about 4 hours or more for conversion ofurine nitrate to nitrite in the bladder.24 Therefore, if achild had not retained urine for such period, it was pos-sible the screening will be falsely negative even whereUTI was present. In their study from north western Ni-geria, Wammanda et al23 reported a UTI prevalence of24.3% in 185 febrile children who had symptoms refer-able to the urogenital system. This figure is equallymuch higher than our finding. The higher rate they re-ported is likely attributable to the fact that only thosechildren who had urinary signs and symptoms were re-cruited. In that study, they also compared the ability ofpositive urine nitrite to detect UTI with urine culture andnoted that nitrite was less sensitive but had an excellentspecificity.23 In other words, there is likelihood for thistest not to detect UTI even where infection is present.

Our study evaluated children for UTI only if they had nofocus of infection and where preliminary screening fornitrite and LE was positive. It has been shown that nei-ther absence nor presence of a focus of infection neces-sarily affects the existence of UTI in children.14This mayhave contributed to the lower prevalence rate observedin this study.The prevalence of UTI in febrile children in our studymay have come in a distant third position, but it empha-sise a significant point that this condition is to be lookedfor in children presenting with fever when there is nofocus of infection. This is more so that managing thelong term complications associated with renal scarringfrom pyelonephritis is much more challenging and diffi-cult in our environment.

Our study has some limitations: urine culture was car-ried out only in children whose dipstick urinalysis waspositive for nitrite and/or LE. There may have beensome of these children whose dipstick urinalysis wasnegative for nitrite and LE but who may actually haveUTI. However, this study has shown that, for childrenpresenting with fever in our environment, acute respira-tory infection is the commonest condition, followed bymalaria and then UTI, stressing the need to evaluatesuch children with this order in mind.

Variable Total Male Female P value

Study populationAge(months)(Mean±SD)

30321.7 ±14.02

18021.1 ±13.9

12322.6 ±14.3

-0.72

Axillary Temp (oC)(Mean±SD)

38.2 ±0.78 38.2 ±0.76 38.2 ±0.81 0.73

Duration of feverbefore presentation(days) (Mean±SD)

5.6±4.3 5.3±3.6 5.4±3.5 0.82

32

Acknowledgments

We are grateful to all the children and their parents thatparticipated in this study. We also acknowledge

Mrs Carol Okorie for carefully analysing the entireblood specimen for malaria parasites for the children.

Conflict of interest: NoneFunding: None

9. Roll Back Malaria. World malariareport. 2005. http://www.rollbackmalaria.org/wmr2005.

10. O'Meara WP, Bejon P, MwangiTW, Okiro EA, Peshu N, SnowRW, Newton CRJC, Marsh K.Effect of a fall in malaria transmis-sion on morbidity and mortality inKilifi, Kenya. Lancet. 2008; 70:1555–1562

11. Guerra CA, Gikandi PW, TatemAJ, Noor AM, Smith DL, Hay SI,Snow RW. The limits and intensityof Plasmodium falciparum trans-mission: implications for malariacontrol and elimination worldwide.PLoS Med. 2008;70:e38. doi:10.1371/journal.pmed.0050038.

12. Lin DS, Huang SH, Lin CC, TungYC, Huang TT, Chiu NC et al.Urinary tract infection in febrileinfants younger than eight weeksof age. Pediatrics 2000;105:E20

13. Crain EF, Gershel JC. Urinarytract infection in febrile infantsyounger than eight weeks of age.Pediatrics 1990; 86:363-367.

14. O’Brien K, Edwards A, Hood K,Butler CC. Prevalence of urinarytract infection in acutely unwellchildren in general practice: a pro-spective study with systematicurine sampling. Br J Gen Pract2013; 63(607):156-164.

15. World Health Organisation, UNI-CEF. WHO growth standards andthe identification of severe acutemalnutrition in infants and chil-dren. A joint statement by theWorld Health Organisation and theUnited Nations Children's Fund.Geneva: WHO, 2009.URL:www.who.int/nutrition/publications/severemal nutri-tion/9789241598163/en/index.html(accessed 21 May 2012).

16. National Institute for Health andCare Excellence (NICE). Urinarytract infection in children. 2007.http://guidance.nice.org.uk/CG05.

18. Duguid JP, Marmaion BP, SwainRH. Medical Microbiology. 13th

ed. Edinburgh: Churchill Living-stone; 1978. p. 327-33

19. Bouyou-Akotet MK, Mawili-Mboumba DP, Kendjo E, EkoumAE, AbdouRaouf O, Allogho EEet al. Complicated malaria andother severe febrile illness in apediatric ward in Libreville, Ga-bon. Infect Dis 2012; 12:216-225.

20. D’Acremont V, Kilowoko M,Kyungu E, Philipina S, Sangu W,Kahama-Maro J et al. Beyond ma-laria—causes of fever in outpatientTanzanian children. N Engl J Med2014; 370:809-817.

21. Wolter N, Cohen C, von GottbergA. Causes of fever in outpatientTanzanian children. N Engl J Med2014;370(23):2243-2248.

22. Chheng K, Carter MJ, Emary K,Chanpheaktra N, Moore CE,Stoesser N et al. A prospectivestudy of the causes of febrile ill-ness requiring hospitalization inchildren in Cambodia. PLoS One2013;8(4): e60634.

23. Ibeneme CA, Oguonu T, OkaforHU, Ikefuna AN, Ozumba UC.Urinary tract infection in febrileunder five children in Enugu,South Eastern Nigeria. Niger J ClinPract 2014;17:624-628.

24. Wammanda RD, Aikhionbare HA,Ogala WN. Use of nitrite dipsticktest in the screening for urinarytract infection in children. West AfrJ Med 2000; 19: 206-208.

25. Jodal U. Urinary tract infection:significance, pathogenesis, clinicalfeatures and diagnosis. In: Pos-tlethwaite RJ, editor. Clinical Pae-diatric Nephrology. Oxford: But-terworth - Heinemann; 1994. pp.151–9.

References

1. Kallander K, Nsungwa-Sabiiti J,Peterson S. Symptom overlaps formalaria and pneumonia-policyimplications for home manage-ment strategies. Acta Trop.2004;70:211–214.

2. Liu L, Johnson HL, CousensS,Perin J, Scott S, et al. Global,regional, and national causes ofchild mortality: an updated sys-tematic analysis for 2010 withtime trends since 2000. Lancet2012; 379: 2151–2161

3. Rougemont A, Breslow N, Bren-ner E, Moret AL, Dumbo O, DoloA, Soula G, Perrin L. Epidemiol-ogical basis for clinical diagnosisof childhood malaria in endemiczone in West Africa. Lancet1991;70:1292–1295

4. Akpede G.O. and Skyes R.M.(1992) Relative contribution ofbacteremia and malaria to acutefever without localizing signs ofinfections in under five children. JTrop Pediatr38, 295-298.

5. English M., Reyburn H., Good-man C. and Snow R.W. (2009)Abandoning presumptive antima-larial treat-ment for febrile chil-dren aged less than five years- acase of running before we canwalk? PLoS Med 6, e1000015.

6. D’Acremont V, Lengeler C, Gen-ton B. Reduction in the proportionof fevers associated with Plasmo-dium falciparum parasitaemia inAfrica: a systematic review. MalarJ 2010;9:240-249.

7. Gething PW, Patil AP, Smith DL,Guerra CA, Elyazar RF, JohnstonGL et al. A new world malarialmap: Plasmodium falciparumendemicity in 2010. Malar J2011; 10: 378-386.

8. Oguche S, Okafor HU, Watila I,Meremiku M, Agomo P, Ogala Wet al. Efficacy of Artemisinin-Based Combination Treatments ofUncomplicated Falciparum Ma-laria in Under-Five-Year-Old Ni-gerian Children. Am. J. Trop. Med.Hyg., 91(5), 2014, pp. 925–935

33