Nigerian Journal of Paediatrics

5Prevalence and pattern of bacteraemia among HIV infected under five children in a tertiary hospital in Kano Nigeria

Niger J Paediatr 2017; 44 (1):26 – 31

ORIGINAL

Adisa AK

Prevalence and pattern of bacte-

Hassan-Hanga F

Oyelami OA

raemia among HIV-infected under

-five children in a tertiary

hospital in Kano, Nigeria

DOI:http://dx.doi.org/10.4314/njp.v44i1.5

Accepted: 6th December 2016

Abstract : Background: Bacterae-

typhi and Staphylococcus aureus

mia is an invasive bacterial dis-

were the predominant isolates,

Adisa AK (

)

ease of childhood that is associ-

each accounting for 21% of all

Department of Paediatrics,

ated with serious complications

cases

bacteraemia.

Most

Aminu Kano Teaching Hospital

and high mortality especially in

(81.3%) of the subjects were on

PMB 3452, Zaria Road

Kano, Nigeria

immunocomprised HIV infected

HAART and its use had no effect

Email: adisakolly@yahoo.com

children.

on rate of bacteraemia. Fourteen

Aim: To determine the prevalence

(73.7%) and 12(63.2%) of the iso-

Hassan-Hanga F

and pattern of bacteraemia among

lates were sensitive to ciproflox-

Department of Pediatrics,

HIV-infected Under-five children.

acin and ceftriaxone respectively.

Aminu Kano Teaching Hospital/

Design: It was a prospective cross

Sensitivities to ampicillin, clox-

Bayero University Kano, Kano Nigeria

-sectional study

acillin and co-trimoxazole were

Subjects and Methods: One hun-

0.0%, 5.3% and 5.3% respectively.

Oyelami OA

dred and thirty four febrile HIV-

Conclusion: Bacteraemia is a sig-

Department of Pediatrics and Child

infected children were recruited

nificant health problem among

Health, Obafemi Awolowo University

from the outpatient departments

HIV-infected under-five children

Teaching Hospital Complex, Ile-Ife,

and emergency room of a tertiary

despite the high rate of HAART

Osun State

hospital to determine the presence

use. Treatment adherence should

of bacteraemia, the etiologic agent

be strengthened among this popu-

and antibiotics susceptibility. An

lation. There is need for improve-

automated (BACTEC) incubator

ment in personal and food hygiene,

was used to detect bacteraemia,

environmental sanitation and pos-

subcultures were done and identi-

sibly introducing typhoid vaccine

fication and antibiotic susceptibil-

among under-five HIV-infected

ity tests were done using standard

children.

laboratory procedures.

Socio-

demographic and clinical data

Keywords: bacteraemia, under-

were obtained using a proforma

five, human immunodeficiency

and data analysis was done using

virus/acquired immune deficiency

SPSS version 17.0 for windows.

syndrome, prevalence, highly ac-

Results: The prevalence of bacte-

tive antiretroviral therapy.

raemia in HIV-infected children

was 14.2% (19/134). Salmonella

Introduction

depending on the presence of an underlying turbulent

cardiac flow or immunosuppression as in HIV infection.

Seventy one percent of world’s HIV -infected children

A form of clinical bacteraemia associated with fever but

live in Sub-Saharan Africa. Nigeria accounts for an

no evidence of sepsis or clear focus of infection is

termed “occult bacteraemia”. In this report, we shall

estimated 10% of global burden of HIV/AIDS and 30%

of the burden of mother-to-child transmission of HIV

use the term “bacteraemia” to mean all clinically signifi-

with the national HIV prevalence rate of 3.4%. Over

2,3

cant forms.

440,000 Nigerian children under the age of 15 years are

living with the infection and three out of every 100

In the HIV-infected, bacteraemia is a serious condition

deaths in children are due to HIV/AIDS directly or indi-

that often persists and could lead to potentially lethal

rectly.

2,4

diseases including pneumonia, septic arthritis, osteo-

myelitis, meningitis and severe sepsis. Factors such as

Bacteraemia is the presence of viable bacteria in the

abnormalities in humoral and cell-mediated immunity,

circulating blood. This may or may not be symptomatic

phagocytic cell dysfunction and skin and mucous mem-

brane defects all contribute to higher risk of bacteraemia

Sampling/data collection

in them. Bacterial infections are responsible for the

immediate cause of death of up to 30% of patients with

By serial recruitment, every eligible HIV-infected child

HIV infection.

who met the inclusion criteria was recruited from the

Unfortunately, 630,000 African children are receiving

PIDC or EPU. Diagnosis of HIV infection was based on

the Nigerian national protocol.

Anti-retroviral Therapy (ART) which corresponds to

only 22% of those that are eligible. In Nigeria, the

During routine clinic visit and emergency room consul-

situation is even more distressing as only 15% of the

tation, eligible participants were identified, proforma

eligible 92,200 Nigerian HIV-infected children have

administered and blood samples were obtained for blood

access to this life-saving treatment modality.

culture and blood counts. A complete physical examina-

There is a paucity of data on bacteraemia in HIV-

tion was also carried out on each child.

infected children particularly in the northern part of the

country. More so, results from few studies from other

Subjects were classified based on the history, clinical

parts of the country such as Benin may not be applica-

examination findings and most recently documented

ble to patients in our locality. This study was carried out

CD4 count into the appropriate WHO clinical and im-

to determine the prevalence and pattern of bacteraemia

munological stages. The Partec Cyflow counter serial

in HIV-infected children. It is hoped that results of this

number 050117117 was used for the estimation of CD4

study would provide the basis for cost-effective inter-

count and percentages were calculated as a fraction of

ventions such as chemoprophylaxis, vaccinations and

the Total Lymphocyte Count (TLC) obtained from full

rational use of antibiotics that will improve the manage-

blood count of the same sample. The most recent CD4

ment and outcome of children with HIV/AIDS.

percentages obtained during infection free period were

used to classify subjects into different immunologic

categories using the revised WHO staging of 2007.

Subjects and Methods

Sample collection

Study design

In each case, the procedure was explained to the parent/

caregiver. The site thoroughly cleaned, with 70% isopro-

The study was a hospital-based prospective cross-

pyl alcohol solution and tourniquet applied, followed by

sectional study conducted between August 2014 and

povidone iodine solution that was applied in a circular

June 2015.

pattern and then allowed to dry. Two to three milliliters

(2-3ml) of blood was obtained following a sterile proce-

Study population

dure by inserting an appropriate-sized vacutainer needle

into an antecubital vein in the arm or any other site

The study population included children aged between 6

deemed appropriate with the opposite end puncturing

weeks and 60 months, confirmed to have HIV infection

into the vial for direct inoculation. Prior to the inocula-

presenting on follow up to the Pediatric Infectious Dis-

tion, the flip-off cap of the commercially produced vials

containing BD BACTEC

Peds Plus

eases Clinic (PIDC) and the Emergency Pediatrics Unit

media was

(EPU).

wiped with alcohol swab and allowed to dry. An addi-

tional 2mls of blood was put in an EDTA bottle for

Inclusion criteria

blood counts which served as an initial markers of infec-

tion pending the availability of culture results.

HIV-infected children aged between 6 weeks and 60

months, on follow up at the PIDC and EPU during the

Laboratory methods

study period. Those who were either febrile or hypother-

mic at presentation and whose parents/caregivers con-

Inoculated blood culture vials were delivered to the

sented to the study.

laboratory within one hour of collection for placement in

the incubator. Samples were incubated in the automated

Exclusion criteria

BACTEC 9050 blood-culture system (Becton Dickin-

son, Temse, Belgium) for a maximum of five days.

Those who had antibiotics (other than co-trimoxazole

Whenever there was a positive signal from the incubator

which is routine in under-five HIV-infected children)

(usually within 48 hours), an aliquot was obtained from

within one week prior to enrolment.

the vial with a sterile syringe and needle and further

examined by Gram stain and sub-cultured onto appropri-

Ethical approval

ate solid media (blood, chocolate and MacConkey agars)

for 48 hours. Vials with no signals after five days of

Ethical approval for the study was obtained from the

incubation in the BACTEC system were checked by

Research and Ethics Committee (REC) of the hospital.

Gram stain and sub-cultured onto solid media for the

A written informed consent to enroll the patient into the

same duration of 48 hours prior to discarding as nega-

study was obtained from the parent(s) or the accompa-

tive. Blood cultures were considered positive if a defi-

nying caregiver(s) of each child.

nite non-contaminant pathogen was isolated after a

maximum of seven days.

For bacterial identification, all positive blood cultures

Table 1: Socio-demographic characteristics of the subjects

were examined directly by Gram stain microscopy and

Variable

Frequency(%)

subcultured on standard media plates. Identification of

the organisms was obtained by biochemical and sero-

Age categories (months)

logical tests. Susceptibility to ampicillin, amoxicillin-

0-11

45(33.6)

clavulanate, cefuroxime, ceftazidime, ceftriaxone, co-

12-35

55(41.0)

trimoxazole, ciprofloxacin, cloxacillin, gentamicin, ox-

36-60

34(25.4)

acillin and ofloxacin were tested using the Kirby-Bauer

Mean (SD)

25(16.6)

disc diffusion method. Preliminary results were made

Gender

available to the managing physicians within 48 hours

Male

80(59.7)

Female

54(40.3)

and the final results after subculture and sensitivity in 7-

Ethnic Group

9 days.

Hausa/Fulani

108(80.6)

Yoruba

5(3.7)

Data analysis

Igbo

10(7.5)

Others

11(8.2)

Statistical analysis was conducted using the statistical

Socio-Economic Class

software package, SPSS version 17.0 (Chicago, IL,

Lower

67(50.0)

USA). Data were presented using frequency tables and

Middle

52(38.8)

cross-tabulations. Quantitative variables were summa-

Upper

15(11.2)

Maternal Education

rized using the mean, median, range, interquartile range

Tertiary

6(4.5)

and standard deviation while qualitative variables were

Secondary

15(11.2)

summarized using frequencies and percentages. Chi

Primary

45(33.6)

square ( χ ) test was used to determine significance of

No formal education

68(50.7)

association between age and prevalence of bacteraemia

and Odds ratio (OR) for association between use of

The overall prevalence of bacteraemia was 14.2%

HAART and prevalence of bacteraemia. Confidence

(19/134). The prevalence was 15.6%, 12.7% and 14.7%

level was set at 95% and a p-value of < 0.05 was consid-

for the 0-11months, 12-35months and 36-59months age

ered significant.

groups respectively. There was no statistically signifi-

cant association between age and bacteraemia (Table 2).

The predominant organisms responsible for bacteremia

were Salmonella typhi and Staphylococcus aureus , each

Results

accounting for 4/19(21%) respectively. Streptococcus

pneumoniae was the second commonest isolate account-

A total of 134 febrile HIV-infected were recruited with a

ing for 3 /19(15.8%)(Table 3).

mean age (SD) of 25(16.6) months. There were 80

(59.7%) males and 54(40.3%) females and the male-to-

Table 2: Prevalence of bacteraemia according to age groups

female ratio was 1.5:1. Forty five (33.6%) of the sub-

among HIV-infected children

jects were infants and 108(80.6%) were of the Hausa/

Bacteremia

Fulani ethnicity. Half (50%) of them were from the

Age Group(mo)

Positive

Negative

lower socio-economic class (Table 1).

One hundred and nine (81.3%) subjects were on

0-11

7(15.6)

38(84.4)

HAART and the highest proportion (32/34) of those on

12-35

7(12.7)

48(87.3)

HAART was among the 36-59 month age group (Fig 1).

36-59

5(14.7)

29(85.3)

Total

19(14.2)

115(85.8)

Fig 1: Proportion of HIV-infected group on HAART

χ =0.17, df=2, p-value = 0.92

according to age groups

There was no statistically significant difference in the

prevalence of bacteraemia between those on HAART

and those that were ARV-naïve (Table 4).

Fourteen(73.7%) of the isolates were susceptible to

ciprofloxacin while 12(63.2%) were susceptible to cef-

triaxone. The only NTS isolated was sensitive to co-

trimoxazole, ciprofloxacin and amoxicillin-clavulanate.

Susceptibility to co-trimoxazole was 1(5.3%) and the

overall sensitivity was lowest for ampicillin(0.0%) and

oxacillin(0.0%)(Table 5).

Table 3: Etiologic agents of bacteraemia in subjects

Organism

S typhi

S aureus

Other K.

Kleb

HIb

NTS

E.coli

Proteus

CoNS

pneumo

spp

pneumo

spp

Total (%)

No of isolates

(%)

(21.0)

(15.8)

(10.5)

(5.2)

(5.3)

(100.0)

Organisms: S. typhi-Salmonella typhi, S. aureus-Staphylococcus aureus, S.pneumoniae - Streptococcus pneumonia, K. pneumo –

Klebsiella pneumoniae, Hib – Haemophilus influenza type b, NTS-Non-typhoidal Salmonella, K. spp – Klebsiella species, E. coli-

Escherichia coli, Proteus spp – Proteus species CoNS- Coagulase-Negative Staphylococcus.

Table 5: Antibiotic sensitivity pattern by isolates in HIV-infected subjects

Isolates

No.

CIP

OFL

CFZ

AMP

CXO

CFU

GEN

AUG

CLO

OXA COT

S. typhi

S.aureus

S. Pneumo 3

Other

Kleb sp

Kleb.

pneumo

Hib

NTS

E. coli

Proteus

spp

CoNS

Total

19(100.0) 14(73.7)

1(5.3) 7(36.8) -(0.0)

12(63.2) 5(26.3) 7(36.8) 6(31.6)

1(5.3)

0(0.0) 1(5.3)

Isolates:

S. typhi- Salmonella typhi , S. aureus- Staphylococcus aureus , S.pneumo - Streptococcus pneumoniae , Kleb. pneumo –Klebsiella

pneumoniae , Hib – Haemophilus influenza e, NTS-Non-typhoidal Salmonella, Kleb. spp – Klebsiella species, E. coli- Escherichia

coli , Proteus spp – Proteus species, CoNS- Coagulase-Negative Staphylococcus.

Antibiotics:CIP- Ciprofloxacin, OFL- Ofloxacin ,CFZ- Ceftazidime, AMP- Ampicillin, CXO- Ceftriaxone, CFU- Cefuroxime,

GEN- Gentamicin, AUG- Amoxi-clavulanate, CLO- Cloxacillin, OXA-Oxacillin, COT- Co-trimoxazole.

Table 4: Relationship between the use of HAART and preva-

positive cultures associated with other methods of cul-

lence of bacteraemia

ture that were used in many of the other studies that re-

HAART

Bacteraemia

ported higher prevalence.

Positive (%)

Negative(%)

Total (%)

The prevalence of bacteraemia was not related to age

Yes

15(13.8)

94(86.2)

109(100.0)

among the HIV-infected subjects. This is not consistent

4(16.0)

21(84.0)

25(100.0)

with the fact that the immune system is normally most

Total

19(14.2)

115(85.8)

134(100.0)

naïve at birth and develops over time. Our finding may

be explained by the presence of HIV infection which is

OR[95% CI]=0.84[0.25, 2.79]

likely to have distorted the trend.

A previous HIV study suggested that apart from the

Discussion

higher burden of bacteraemia in sub-Saharan Africa, the

spectrum of aetiologic agents differ from those in devel-

oped world. Such differences may also be seen within

In this study, the prevalence of bacteraemia in HIV-

infected under-five children was 14.2%. This is similar

developing countries. An increase in bacteraemia due to

to the report by Madhi et al. with high rate of bacterae-

non-typhoidal Salmonella (NTS) and Mycobacteria spe-

mia in HIV infected children of comparable age group.

cies was observed among HIV-infected individuals in a

2010 study in Uganda, but we observed that Salmo-

The prevalence from our study is also similar to that in

the Anti Retroviral Research of Watoto (ARROW)

nella typhi and Staphylococcus aureus are the most pre-

study (14.5%) and also the 15.5% found in a study at

dominant causes of bacteraemia in HIV-infected in our

the Kwa-Zulu Natal hospital in South Africa.

How-

population. The present study was among under-five

ever, prevalence observed by Imade and colleague in

children with poorer hygiene practices (compared to

Benin was much higher(37.1%). This difference may be

adults) while the Ugandan study was among older popu-

due to the fact that most of the subjects in our study

lation in addition to the fact that aerobic culture media

were on HAART and may have been able to reconstitute

used in our study is not suitable for isolating Mycobac-

their immune system thereby having a lower predisposi-

teria. The finding of Salmonella typhi as a common iso-

tion to opportunistic infections. In addition, we found a

late may be explained by the low level of water and food

very low rate of bacteraemia due to Coagulase-Negative

hygiene that characterize most poor and developing so-

cieties like Nigeria . However, this finding is in con-

Staphylococcus (CoNS) which is a notable contaminant.

This might have significantly reduced the rate of falsely

trast to Tanzanian studies that reported and postulated

low risk for Salmonella typhi among HIV-infected

in separate studies by Imade et al.

and Madhi et al.

population.

18,19

Understandably, Staphylococcus aureus

this was the case. Furthermore, nearly all the isolates

is a ubiquitous organism that commonly colonizes the

were resistant to co-trimoxazole just as Madhi et al.

skin and nostril of most individuals regardless of the

reported in Soweto. This is largely attributable to its

immune status, they become pathogenic when they find

routine use in the HIV group as prophylaxis against op-

their way into deeper body structures which is easier in

portunistic infections like Pneumocyctis jiroveci pneu-

HIV infection as a result of defective mucosal barriers.

monia (PCP), toxoplasmosis and malaria.

Our findings agree with reports of Imade et al. in Be-

nin Nigeria and Nchabeleng et al . in Kwa-Zulu Natal,

South Africa where Staphylococcus aureus was found to

be the commonest isolate.

Conclusion

Apart from the Benin study where they found a high rate

Bacteraemia is an important cause of morbidity and

of CoNS in the ARV-naïve HIV-infected subjects,

mortality among under-five HIV-infected children. Sal-

CoNS is not a prominent cause of bacteraemia in HIV-

monella typhi and Staphylococcus aureus are important

infected children.

13,14

We found only one case (5.3%) of

etiologic agents in children with HIV infection. The use

CoNS infection in this study and that was detected in a

of the highly active antiretroviral therapy is expected to

subject with advanced immunosuppression. This may be

reduces the rate of bacteraemia but compliance is more

because they are mainly opportunistic bacteria that are

important. We recommend the use of ciprofloxacin or

expected to be more frequent in the immunosuppressed

ceftriaxone as the empiric first line antibiotics in any

and those with prosthetic devices and indwelling cathe-

suspected case of bacteraemia in under-five HIV-

ters.

[21]

In addition, the automated culture system is as-

infected children. We also suggest the need to improve

sociated a low rate of contamination by organisms such

food and water hygiene and, in the short-term, consider

as CoNS.

introduction of mass vaccination against Salmonella

Even though some studies have report that Mycobacteria

typhi to help reduce the rate of Salmonella typhi bactere-

plays notable role in bacteraemia among HIV infected

mia in HIV-infected children. Ampicillin, oxacillin,

children, our study cannot corroborate or refute their

cloxacillin and co-trimoxazole as empiric first-line anti-

findings since we only utilized aerobic media which is

biotics in the management of suspected bacteraemia

incapable of isolating Mycobacteria

should be discouraged.

The use of HAART significantly reduces the progres-

Limitation

sion of paediatric HIV/AIDS.

22,23

Over four-fifths (81%)

of our HIV-infected subjects were on HAART which

The culture media used in this study is not sensitive for

may be another reason for the lower rate of bacteraemia

detecting mycobacteria and anaerobic bacteria which

when compared to findings of Imade et al and Madhi et

may play some role in significant bacteraemia

al. Hospital-based studies in Africa prior to the advent of

HAART showed that bacteraemia was three times more

frequent among the HIV-infected than in the HIV-

Authors’ Contribution

uninfected people and five times more likely to cause

Adisa AK and Hassan-Hanga F conceptualized the

death.

24-26

study, Adisa AK recruited the subjects and analyzed the

The isolates were sensitivity to ciprofloxacin and cef-

data.

tazidime and highly resistant to co-trimoxazole. There

Oyelami OA reviewed the concept and edited the manu-

was significant sensitivity of Salmonella to ciproflox-

script.

acin but it was not as susceptible to amoxicillin-

Conflict of Interest: None

clavulanate and gentamicin as observed among a Thai

Funding: None

population of HIV-infected subjects.

Until recently,

ciprofloxacin had not been recommended for use in chil-

dren due to concerns about its musculoskeletal adverse

effects. Recent studies have shown that these are mild,

Acknowledgement

transient and reversible.

This concern over the years

may have relatively preserved ciprofloxacin and other

Our sincere appreciation goes to Prof SK Obaro for his

fluoroquinolones (like ofloxacin) thereby retaining their

support for the automated blood culture.

efficacy and preventing the development of resistance.

Ampicillin resistance was high in our study. Similarly,

Reference

UNAIDS. Global Report: UN-

Nigeria, Federal Ministry of

AIDS Report on the Global AIDS

Health. National guideline for

Health. National human immuno-

epidemic 2015 [Internet].2015

Paediatric HIV and AIDS treat-

deficiency virus and acquired

[updated 2015 Nov; Accessed

ment and care. Federal Ministry of

immune deficiency syndrome and

2015 Aug].

Health. Nigeria 2010.

Reproductive Health Survey 2012

(plus II): Human immunodeficiency

virus Testing. J HIV Hum Reprod

2014; 2:15-29.

United Nations General Assembly

14. Nchabeleng M, Yeung S, Escott S,

24. Gordon MA, Walsh AL,

Special Session (UNGASS) Coun-

Wilkinson D, Sturm AW. (2000).

Chaponda M, Soko D, Mbuwinji,

try Progress Report [internet].

Bacteraemia in HIV-infected chil-

Molyneux ME, et al. Bacteraemia

Nigeria. NACA/UNAIDS; 2012;

dren in a rural Kwa-zulu Natal

and mortality among adult medi-

Accessed 2013 Apr 18.

Hospital. Int. Conf. AIDS. July 9-

cal admissions in Malawi – pre-

Spraycar M, ed. Stedman's Medi-

14; 13: abstract No. MoPeB2202.

dominance of non-typhi salmo-

cal Dictionary . 26th ed. Baltimore,

15. Ward JL, Zangwill KM . Haemophi-

nellae and Streptococcus pneu-

Md: Lippincott Williams & Wil-

lus influenzae . In: Feigin RD,

moniae. J Infect 2001; 42: 44-9.

kins; 1995.

Cherry JD, editors. Textbook of

25. Gilks CF, Brindle RJ, Otieno LS,

6. Thwaites GE, Edgeworth JD, Ghra-

Pediatric Infectious Diseases. 4

Watkins WM, Waiyaki PG, Were

nia-Klotas E, Kirby A, Tilley R,

edition. Philadelphia, PA:WB

JB, et al. Life-threatening bacte-

Torok ME et al. Clinical manage-

Saunders; 1998:1464-82.

raemia in HIV-1 seropositive

ment of Staphylococcus aureus

16. Grant A, Djomand G, De Cock K.

adults admitted to hospital in

bacteremia. Lancet Infect Dis

Natural history and spectrum of

Nairobi, Kenya. Lancet 1990;

2011; 11:208-22.

disease in adults with HIV ⁄ AIDS

336:545-9.

Benneth NJ. Bacteremia. Med-

in Africa. AIDS 1997; 11:S43 – S54.

26. Vugia DJ, Kiehlbauch JA, Ye-

scape [internet] 2015; [updated

17. Umeh E, Agbulu C. Distribution

boue K, N’Gbichi, JM, Lacina D,

2015 Jun. 22].Accessed 2016 Mar

pattern of Salmonella typhoidal

Maran M et al. Pathogens and

23.Availablefrom:http://

serotypes in Benue state central

predictors of fatal septicemia

www.emedicinemedscape.com/

Nigeria. Int. J. Epidemiol 2013; 8

associated with human

article/961169.overview.

(1):5749-61.

immunode ﬁciency virus infection

8. Kovacs A, Leaf HL, Simberkoff

18. Crump JA, Ramadhani HO, Morris-

in Ivory Coast, West Africa. J

MS. Bacterial infection s. Med Clin

sey AB, Saganda W, Mwako MS,

Infect Dis 1993; 168:564-70.

North Am 1997; 81:319-43.

Yang LY, et al. Invasive bacterial

27. Srifuengfung S, Chokephaibulkit

Stein M, O'Sullivan P, Wachtel T,

and fungal infections among hospi-

K. Bacteraemia and antimicrobial

Fisher A, Mikolich D, Sepe S et al.

talized HIV-infected and HIV-

susceptibility in HIV-infected

Causes of death in persons with

uninfected adults and adolescents

patients at Siriraj hospital. South-

human immunodeficiency virus

in northern Tanzania. Clin Infect

east Asian J Trop Med Public

infection. Am J Med 1992; 93:387-

Dis 2011;52:341 – 8.

Health 2005; 36(2):347-51.

90.

19. Levine MM, Farag TH. Invasive

28. World Health Organisation.

10. WHO. Global Update on HIV treat-

salmonella infections and HIV in

Fluoroquinolone use in children.

ment 2013: Results, Impact and

Northern Tanzania. Clin Infect Dis

Second Meeting of the Subcom-

Opportunities. WHO report in

2011; 52:349 – 51.

mittee of the Expert Committee

partnership with UNICEF and

20. Todd JK. Staphylococcus. In:

on the Selection and Use of Es-

UNAIDS.2013 [Updated 2013

Kleigman RM, Stanton BF, Schor

sential Medicines [internet] 2008.

Nov; Accessed 2014 Aug].

NF, St. Geme III JW, Behrman

Accessed Nov 2015.

11. Imade PE, Eghafona NO. Inci-

RE, editors. Nelson Textbook of

29. Adefurin A, Sammons H, Jacqz-

Paediatrics. 19 edition, Pennysla-

dence of bacteraemia in antiretro-

Aigrain E, Choonara I. Ciproflox-

viral-naive HIV-positive children

vania: W.B Saunders Company,

acin safety in paediatrics. Arch

less than five years of age in Benin

2011:904.

Dis Child 2011; 96:874-

-city, Nigeria. Libyan J Med 2010;

21. Henderson KL, Johnson AP, Mul-

880 doi:10.1136/

5:10.

ler-Pebody B, Charlett A, Gilbert

adc.2010.208843

12. Madhi SA, Petersen K, Madhi A,

R, Sharland M. The changing aeti-

Khoosal M, Klugman KP. In-

ology of paediatric bacteraemia in

creased Disease Burden and Anti-

England and Wales, 1998-2007. J

biotic Resistance of Bacteria Caus-

Med Microbiol. 2010; 59(2):213-9 .

ing Severe Community-Acquired

22. Tumbarello M, Tacconelli E, Do-

Lower Respiratory Tract Infections

nati KG, Citton R, Leone F, Spanu

in Human Immunodeficiency Vi-

T et al. HIV-associated bactere-

rus Type 1-Infected Children. Clin

mia: how it has changed in the

Infect Dis 2000; 31: 170-6.

highly active antiretroviral therapy

13. Musiime V, Cook A, Bakeera-

(HAART) era. J Acquir Immune

Kitaka S, Vhembo T, Lutakome J,

Defic Syndr 2000; 23:145-51.

Keishanyu R et al. Bacteraemia,

23. Manfredi R, Nanetti A, Ferri M,

Causative Agents and Antimicro-

Chiodo F. HIV-associated non-

bial Susceptibility Among HIV-1-

mycobacterial sepsis-bacteraemia,

infected Children on Antiretroviral

before and during the highly active

Therapy in Uganda and Zim-

antiretroviral therapy era. AIDS

babwe. Pediatr Infect Dis 2013;

1999; 13:1274-76.

32:85-62.