1

ETHIOPIAN JOURNAL OF
PEDIATRICS AND CHILD
HEALTH
July 2009, Volume V, Number 5

Original articles
The Prevalence of Nosocomial Infections and Associated Risk Factors in Pediatric Patients in Tikur
Anbessa Hospital
Mikyas Demissie ,M.D. , Sileshi Lulsesed, M.D.
Clinical Predictors of Pneumonia Among Under-five Children At Tikur Anbesa Specilaized Hospital
Kalid Asrat, MD, Amha Mekasha, MD, MSc
Gullian Barre Syndrome in Children At Tikur Anbessa Specialized Hospital
Tigist Bacha, MD
Assessment of quality of care of sick under-five children in referral hospitals in Ethiopia.
Sirak Hailu, MD, Solomon Emyu, MD/MPH, Fisseha Mamo, MPH, Tolawaq Kejela MD
Management of severe acute malnutrition in children using community based therapeutic care approach: a
review of three years data from southern Ethiopia.
Efrem Teferi, MD Shiferaw Teklemariam, MD, MPH , Lopiso Erosie, BSC, MPH , Abel Hailu, MD,
Tefera Belachew, MD, MSc, DLSHTM, Mohammed A Yassin, MD, MSc, PhD

Review article
Child Survival: Progress Towards meeting MDG4
Assaye Kassie, MD

Case report
A Case Report : Optic glioma in a child with NF1
Kalid Asrat, MD

Notes for contributors

Ethiopian Pediatric Society

E-mail: eps@ethionet.et Telephone: 251-011-8602843
Addis Ababa, Ethiopia

Ethiopian Journal of Pediatrics and Child Health

The official organ of Ethiopian Pediatric Society

The Ethiopian Journal of Pediatrics and Child Health aims to contribute towards the improvement
of child health in developing countries, particularly in Ethiopia. The journal publishes original
articles, reviews, case reports pertaining to health problems of children.

Editorial board
Amha Mekasha, MD, Msc
Assaye Kassie, MD
Sirak Hialu, MD
Tedbab Degife, MD

Editor-in-chief

Table of contents
Original articles
The Prevalence of Nosocomial Infections and Associated Risk Factors in Pediatric Patients in Tikur
Anbessa Hospital
Mikyas Demissie ,M.D. , Sileshi Lulsesed, M.D.
Clinical Predictors of Pneumonia Among Under-five Children At Tikur Anbesa Specilaized Hospital
Kalid Asrat, MD, Amha Mekasha, MD, MSc
Gullian Barre Syndrome in Children At Tikur Anbessa Specialized Hospital
Tigist Bacha, MD
Assessment of quality of care of sick under-five children in referral hospitals in Ethiopia.
Sirak Hailu, MD, Solomon Emyu, MD/MPH, Fisseha Mamo, MPH, Tolawaq Kejela MD
Management of severe acute malnutrition in children using community based therapeutic care approach: a
review of three years data from southern Ethiopia.
Efrem Teferi, MD Shiferaw Teklemariam, MD, MPH , Lopiso Erosie, BSC, MPH , Abel Hailu, MD,
Tefera Belachew, MD, MSc, DLSHTM, Mohammed A Yassin, MD, MSc, PhD

Review article
Child Survival: Progress Towards meeting MDG4
Assaye Kassie, MD

Case report
A Case Report : Optic glioma in a child with NF1
Kalid Asrat, MD

Notes for contributors

4
The Prevalence of Nosocomial Infections and Associated Risk Factors in Pediatric Patients in
Tikur Anbessa Hospital
Mikyas Demissie, MD, Sileshi Lulseged, MD, Msc

Abstract
Little is known about nosocomial infections and associated risk factors among children in Ethiopia.
The aim of the study is to generate data on nosocomial infection in children that will serve as a base for
further studies and to develop prevention interventions. A case control study was done on 111 cases and
222 controls from pediatrics wards of Tikur Anbessa Hospital, Aug 2002-Dec 2003. Nosocomial infection
rate was 5 per 100 discharges. A total of 143 nosocomial infections were detected in 111 cases. The
commonest infection was pneumonia 39.8%. Specimen for culture and sensitivity was taken from
63/143(44.1%) infections and organisms were isolated in 43 infections. And 14 different type of bacteria
were found. E coli, klebsiella pneumoniae and pseudomonas species were the most frequently isolated
organisms. The resistance to ampicillin was 91.9% , gentamycin 67.2%, , ceftriaxone 50%, norfloxacin
18.3%, and ciprofloxacin 15.4%. Age less than one year, malnutrition, admission to orthopedics unit,
peripheral intravenous line and prolonged hospitalization were significantly associated with nosocomial
infection. In conclusion surveillance for high risk patients, education of health personnel, proper isolation
technique, hand washing or use of glove and gowns, avoid prolonged hospitalization when possible and
reestablishment of infection control committee are needed for prevention of nosocomial infection. And
antimicrobial therapy should be guided based on drug susceptibility pattern of bacteria isolated from patients
with nosocomial infection in the hospital.

Introduction
Nosocomial infections result in considerable

and surgical wound infection was common in

morbidity

of

other studies. Gram negative bacteria, Gram

care

positive bacteria, and viruses were incriminated

expenditure (1).Published data about nosocomial

as the predominant causes of nosocomial

infection rate in Ethiopian pediatric patients are not

infection in various places (1, 5, 7-9).

and

hospitalization

mortality,
and

prolongation

increases

patient

currently available. But there were two reports on

outbreak of klebsiella at the Etho-Swedish

Determinants of infection include host factors,

Children's Hospital. The first report was on

prior invasive procedures, use of catheters, use

outbreak

klebsiella

of antibiotics and exposure to other patients,

bacteremia between February 1988 and February

visitors or health care providers with contagious

1990, and infection control committee was

diseases (10). Nosocomial infections can be

functioning at that time (2).The second report was

transmitted by contact, common source, air or

on outbreak of klebsiella oxytoca at Ethio-Swedish

vectors. Most nosocomial infections in infants

hospital in 1992 and 1993(3).

and children result from contact transmission via

Nosocomial infection rates in pediatric patients in

the hands of personnel (1,11-13). Malnutrition is

other countries range from 1.2 to 10.3 infections

the

per 100 discharges. In most studies the infection

immunodeficiency

rate varies by age and ward or service. The

probability that an individual will experience at

highest infection rate was seen in infants younger

least one nosocomial infection increases with

than one year of age (1,4-7).

increasing exposure to the hospital (1,11).

of

gentamycin-resistant

most

common

cause

(14-16).

of

secondary

The

cumulative

Depressed level of consciousness increases the

The predominant site of infection differs by the

likelihood of reflux of gastric contents and

population studied and the type of surveillance

aspiration

performed. Respiratory tract and gastrointestinal

Pneumonia (11, 16-19). Antimicrobial therapy

tract infections were common in some studies

reduces the concentration of normal flora and

into

lower

airway

leading

to

6
allowing antimicrobial resistant microorganisms

are rare. Nosocomial infections also are caused

to gain a foothold. Any patient taking antibiotics

occasionally by contaminated intravenous fluids

for prophylaxis or treatment may become a

or blood products (1,3,11,22,23) .

source

of

highly

resistant

organisms.

Immunocompromised hosts are more likely to

Surveillance for infection is the first step in

have

infection(11,12,18,20,21).

identifying nosocomial infection and suggesting

Percutaneously inserted peripheral intravenous

methods of prevention (10). Surveillance, by

catheters are associated with a low rate of local

itself, is an effective process to decrease the

infection in children and blood stream infections

frequency of hospital acquired infections (24).

nosocomial

Methods and Materials

A case control study was conducted reviewing

nosocomial infections for neonates are different

medical records of children who were discharged

from others. Additionally the medical records of

from pediatric wards of Tikur Anbessa Hospital

the emergency ward were incomplete and those

from August 2002 to December 2003. Neonates

who were discharged from the emergency ward

were excluded because the risk factors for

were excluded.
neonates and those who were discharged from

The sample size was calculated with assumption

the emergency ward. They were identified after

of prevalence of malnutrition in hospitalized

all charts of patients who were discharged from

children in Ethiopia to be 50% which makes

August 2002 to December 2003 were reviewed.

p2=0.5 and p1=0.66. Calculated sample size

For each case the next two discharged patients

was: total sample size=333, Cases=111 and

without nosocomial infection were taken as

controls=222 with confidence interval (CI) =95%,

controls

odds ratio (OR) =2

neonates and those discharged from emergency

, Z=1.96 and Z=0.84.

Cases are patients with nosocomial infection who

were discharged from pediatric wards excluding

ward.

from

registration

book

excluding

7
Data was collected from charts of cases and

Short duration nasotracheal or orotracheal

controls using structured format that include age,

intubation is defined as intubation only during

sex,

general

primary

diagnosis,

weight,

length,

anaesthesia

while

Long

duration

community acquired infection, the ward and the

nasotracheal

service the patient was admitted, procedure that

intubation for respiratory failure(19). Orotracheal

was done, nosocomial infection by site, culture

or nasotracheal intubation is all episodes of

and sensitivity result, use of antibiotics, use of

orotracheal or nasotracheal intubation with in the

chemotherapy,

of

period of one week prior to the onset of

consciousness, HIV status, duration of stay in

nosocomial pneumonia will be considered.

hospital, the status of admitted child at discharge.

Nasogastric intubation is the presence of a

Operational Definitions:

nasogastric intubation will be accepted when it

Nosocomial infections are infections acquired

was present for at least two days within the

during hospital care which are not present or

period one week prior to the onset of nosocomial

incubating at admission. Infections occurring

infection (pneumonia) (19).

use

of

steroid,

level

or

orotracheal

intubation

is

more than 48 hours after admission are usually

considered nosocomial. Infection that is acquired

Wasting and stunting were based on calculating

in the hospital becomes evident after hospital

weight as a percentage of reference median

discharge is considered nosocomial. Definition to

weight for height and height as a percentage of

identify

been

reference median height for age (26). Nutritional

developed for specific infection sites. These are

status for under-five was according to the

derived from those published by the Center for

welcome classification (27).

nosocomial

infections

have

Disease Control and Prevention(CDC) in the

United States of America or during international

Data was entered into SPSS 10.4 statistical

conferences and are used for surviellance of

software and it was analyzed using EPI info 6.

nosocomial infections (24,25).

After analysis result was presented using

descriptive statistics, chi-square determination,

8
and p-value and odds ratio calculation. P-value

admission, 17(11.9%) between the 4th and 7th

was considered as significant if it was less than

day, 50(35%) between 8th and 14th day,

0.05. Multiple regression analysis was done

21(14.7%) between 15th and 21st day and

using SPSS to control the effect confounding

45(31.4%) nosocomial infections were detected

factors.

after 21 days of admission.

Results

The commonest site of nosocomial infection was

A total of 1701 patients were discharge from

pneumonia 39.8% followed by gastroenteritis

pediatric wards excluding neonatal ward and

11.9% and primary blood stream infection

emergency ward in 2003. Of these 85 patients

10.9%(Table 1).The predominant site of infection

had acquired nosocomial infection at discharge.

was pneumonia in pediatrics, general surgery,

Hence, nosocomial infection rate was 5 per 100

neurosurgery, plastic surgery, tumor therapy unit,

discharges in 2003 in these wards. Charts of 111

and ENT unit. In burn unit the predominant site of

cases and 222 controls who were discharged

infection was skin and soft tissue infection. And

from pediatrics wards from August 2002 to

in orthopedics upper respiratory tract infection

December

was the commonest

2003,

excluding

neonatal

and

nosocomial infection

emergency wards were reviewed analyzed.

detected followed by skin and soft tissue

Among 111 cases 85 were discharged in 2003

infection. Surgical wound infection was the

and 26 were discharged in 2002.

second predominant site of nosocomial infection

Of all cases with nosocomial infection 86(77.5%)

in patients who were discharged from general

had nosocomial infection once, 21(18.9%) twice,

surgery.

2(1.8%) three times, and 2(1.8) four times. So,

Specimen for culture and sensitivity was taken

143 nosocomial infections were detected in 111

from

cases.

The majority of nosocomial infections

Among 63 nosocomial infections for which

were detected in the 2nd week and after the 2nd

specimen were collected for culture and

week . Out of 143 nosocomial infections 10(7%)

sensitivity test 30 (47.6%) were from blood,

of them were detected in the first three days after

11(17.5%) were from urine and 15(23.8%) were

63/143(44.1%)

nosocomial

infections.

9
from pus and the remaining 7(11.1%) were from

unit, peripheral intravenous line, and prolonged

other body fluids.

hospitalization were found to have significantly

The result of culture and sensitivity test was

increased nosocomial infection. And factors such

positive

nosocomial

as sex, weight for age according Wellcome

infections. A total of 62 different isolates were

classification, height for age according to

found from those with positive culture result. A

Waterlow classification,

single organism was isolated from each of 27

admission, surgical intervention, urgency of

nosocomial infections and two or more bacteria

surgery, type of surgical wound, duration of

were isolated from 16 nosocomial infections.

operative procedure, duration of stay before

When the isolates were classified by the type of

surgical intervention, orothracheal intubation,

organism, 14 different types of bacteria were

nasogastric intubation, urinary catheterization,

isolated. Escherichia coli was the most frequently

bronchoscopy, prior treatment with antibiotics,

isolated

chemotherapy or steroid, depressed level of

in

43/63(68.3%)

organism

pneumoniae,

followed

pseudomonas

of

by

klebsiella

species,

ward, infection at

and

consciousness, and HIV status were not

coagulase negative staphylococci as shown in

significantly associated with the development of

Table 3.

nosocomial infection. Majority of children with

Drug susceptibility test for gentamycin was done

nosocomial infection (66.7%) stayed more than

for 61 of isolates and 41/61(67.2%) were

21 days before discharge when compared to only

resistant . And 11/60(18.3%)of isolates were

7.2% of controls stayed more than 21 days.

resistant to norfloxacin . In similar way

(Table 5)

18/36(50%) were resistant to ceftriaxone. Drug

Among

susceptibility for cloxacillin was done only for

discharged after improvement, 18(8.1%) were

three isolates. (Table 4)

discharged in the same or worse condition, and

Infants under one year of age (29days-11mo),

29(13.1%) died. Of 111 cases, 77(69.4%) were

children with moderate wasting (Waterlow

discharged after improvement, 13(11.7%) were

classification), children admitted to orthopedics

discharged in the same or worse condition, and

222

controls,

175(78.8%)

were

10
21(18.9%) died. Children with nosocomial

drug susceptibility pattern of organisms isolated.

infection had no significant increased risk of

The proportion of isolated bacteria resistant to

death

floroquinolones was less when compared to the

when

compared

to

those

without

nosocomial infection ( p value=0.157).

proportion of resistance to other antibiotics.

Discussion

Changing or rotating standard group of antibiotics

Nosocomial infection rate in this study was with

used for empiric therapy has been efficacious in

in the range which was observed in other studies

limited studies. The role of floroquinolones in the

(1, 4, 6, 7). The predominant site of nosocomial

treatment of serious infections in children does

infection

by

not appear to be compromised by safety

primary blood stream

concerns since arthralgia and quenolone-induced

infection. This was similar to the study done

cartilage toxicity were low and episodes of

Welliver and Mc Laughlin. (7).Though 49(44%) of

arthralgia were mostly reversible (12,28).

was

gastroenteritis,

pneumonia
and

followed

patients with nosocomial infection were admitted

to surgical service in this study, surgical wound

In this study Gram negative bacilli were

infection was not common. It accounted for only

commonly isolated organisms which constituted

5.6% of all nosocomial infections.

46/62(74.2%) of all isolated bacteria. In reports

Specimen for culture and sensitivity was not

from the 1960s and 1970s from USA Gram

taken in 80(55.9%) of nosocomial infection and

negative bacteria accounted for more than 50%

they were treated empirically. But isolated

of the infections which is similar to our study

organisms are usually resistant to commonly

(1,5,7-9). It is not possible to differentiate

used antibiotics for nosocomial infection like

whether

gentamycin and ceftriaxone. According to this

staphylococci were contaminant or etiologic

study, it does not seem rational to treat

agent from this study since only one blood

nosocomial infection empirically with the above

culture sample was taken and details of blood

mentioned antibiotics in Tikur Anbessa hospital.

culture and clinical response to treatment was

Antimicrobial therapy should be guided based on

not available. (10,25).

the

isolated

coagulase

negative

11
Disruption of the physical barrier occurs in burn

Effective targeted surveillance for high risk

patients and in those with degloving injury.

patients, staff education, use of proper isolation

Admission to orthopedics unit was significantly

techniques and effective infection control practice

associated with nosocomial infection in this

such as hand washing before and after patient

study. But admission to burn unit in the hospital

contact or use of glove and gowns are

was not significantly associated with nosocomial

recommended for the prevention of nosocomial

infection. (10,12). Increased risk of nosocomial

infection. Prolonged hospitalization should be

infection in infants less than one year of age and

avoided as much as possible since it is

in nutritionally compromised patients was seen

significantly associated with the development of

in other studies as it was seen in this study (1, 5,

nosocomial infection. And reestablishment of

7, 14-17). Children admitted to tumor therapy

infection control committee is also required.

unit and those who took anti neoplastic

chemotherapy have no significantly increased
nosocomial infection unlike other studies (18).
The effect of peripheral intravenous line on the
occurrence of nosocomial infection needs further
prospective

study

since

canula

related

septicemia is considered when the same

organism is isolated from

canula and blood.

Even though increased mortality is expected in

patients with nosocomial infection, mortality of
patients with nosocomial infection was not
increased in this study when compared to the
control group (1, 11, 23).

12
Table 1. Distribution of nosocomial infection by site , Tikur Anbessa hospital, August 2002-December 2003.
.Site of nosocomial infection
Pneumonia
Gastroenteritis
Primary blood stream infection
Urinary tract infection
Skin and soft tissue infection
Upper respirator tract infection
Surgical wound infection
Central Nervous System infections
Others
Systemic infection
Total

frequency
57
17
15
13
13
9
8
2
7
2
143

%
39.9
11.9
10.9
9.1
9.1
6.3
5.6
1.4
4.9
1.4
100

8
(18.2)
2
(10.5)
0

Skin &
soft
tissue
(%)

27
1
(61.4)
(2.3)
General
5
0
7
1
surgery
(26.3)
(36.8)
(5.3)
Neuro
0
1
3
2
surgery
(10)
(30)
(20)
Plastic
1
0
0
1
0
surgery
(50)
(50)
Burn unit
1
1
0
3
4
(7.1)
(7.1)
(21.4)
(28.6)
Tumor
0
1
1
11
1
therapy
(3.2)
(3.2)
(35.5)
(3.2)
ENT
0
0
1
3
0
(16.7)
(50)
Orthopedics
1
1
6
2
4
(5.9)
(5.9)
(35.3)
(11.8)
(23.5)
Total
8
13
9
57
13
SWI=surgical wound infection UTI=urinary tract infection
URTI= upper respiratory tract infection CNS= central nervous system

primary
blood
stream
infection(
%)
5
(11.4)
0
0
0
2
(14.3)
8
(25.8)
0
0
15

Systemic
infectio(%)
(%)infection
infectiom
Total

Pneumo
nia
(%)

Other(%)

Pediatrics

Site of nosocomial infection

SWI
UTI
URTI
(%)
(%)
(%)

CNS(%)

Service

Gastro
Enteritis (%)

Table 2. Site of nosocomial infections by the service in Tikur Anbessa hospital, August 2002-December 2003.

4
(9)
2
(10.5)
1
(10)
0

44

1
(5.3)
1
(10)
0

1
(5.3)
1
(10)
0

19

10

1
(7.1)
6
(19.3)
1
(16.7)
2
(11.8)
17

14

3
(9.7)
1(16.
7)
0

1
(7.1)
0

17

1
(5.5)
2

31

143

Table 3 Types of organism isolated from children with nosocomial infection in Tikur Anbessa hospital, August 2002December 2003.
Organism isolated
Frequency
%
Escherichia coli
11
17.7
Klebsiella pneumoniae
9
14.5
Pseudomonas species
9
14.5
Coagulase negative staphylococci 6
9.7
Acinetobacter species
5
8.1
Klebsiella oxytoca
4
6.5
Salmonella species
4
6.5
Staphylococcus aureus
3
4.8
Proteus vulgaris
3
4.8
Citrobacter species
3
4.8
Shigella species
2
3.2
Group A streptococci
1
1.6
Morganella morgagne
1
1.6
Enterobacter cloaca
1
1.6
Total
62
100

14

Table 4 Drug susceptibility pattern of bacteria isolated from children with nosocomial infections in Tikur Anbessa Hospital,
August 2002- December 2003.
Antibiotics
No. of isolates
sensitive
Intermediate
Resistant
tested
sensitive
Ampicillin
62
4(6.5%)
1(1.6%)
57(91.9%)
Gentamycin
61
20(32.8%)
0
41(67.2%)
Norfloxacin
60
47(78.3%)
2(3.3%)
11(18.3%)
Trimethoprim58
10(17.2%)
2(3.5%)
46(79.3%)
Sulfamethoxazole
Tetracycline
58
13(22.4%)
2(3.5%)
43(74.1%)
Chloramphenicol
55
8(14.5%)
0
47(85.5%)
Ceftriaxone
36
13(36.1%)
5(13.9%)
18(50%)
Augmentin
31
8(25.8%)
5(16.1%)
18(58.1%)
Doxycycline
25
5(20%)
1(4%)
19(76%)
Amoxicillin
24
2(8.3%)
1(4.2%)
21(87.5%)
Naldixic Acid
22
12(54.5%)
0
10(45.5%)
Nirofurantoin
19
6(31.6%)
1(5.3%)
12(63.1%)
Amikacin
14
11(78.6%)
1(7.1%)
2(14.3%)
Ciprofloxacin
13
9(69.2%)
2(15.4%)
2(15.4%)
Penicillin G
13
0
2(15.4%)
11(84.6%)
Erythromycin
13
4(30.8%)
3(10%)
6(46.2%)
Methicillin
11
2(18.2%)
3(27.3%)
6(54.5%)
Carbencillin
8
2(25%)
3(37.5%)
3(37.5%)
Kanamycin
4
0
2(50%)
2(50%)
Cephalotin
4
0
2(50%)
2(50%)
Cloxacillin
3
0
2(66.7%)
1(33.3%)
Lincomycin
3
1(33.3%)
2(66.7%)
0

14

15

Table 5 Risk factors associated with nosocomial infection in pediatric patients, Tikur Anbessa Hospital, August 2002December 2003.
Variables
Age in month

Weight for
height

Peripheral
intravenous line
Use of
antineoplastic
chemotherapy
Duration of stay
before
discharge

Service

1-11
12-59
60-119
120-180
>=90%
80-89%
70-79%
<70%
No
Yes
No
Yes

Control
Number
35
85
52
50
143
44
22
7
23
199
215
7

%
15.8
38.3
23.4
22.5
66.2
20.4
10.2
3.2
10.4
89.6
96.8
3.2

Case
Number
31
33
27
20
58
21
19
8
3
108
89
22

%
27.9
29.7
24.3
18
54.7
19.8
17.9
7.6
2.7
97.3
80.2
19.8

<7 days
8-14 days
15-21days
>21 days

74
62
40
46

33.3
27.9
18
7.2

2
19
16
74

1.8
17.1
14.4
66.7

General surgery
Pediatrics
Plastic surgery
ENT
Neurosurgery
Orthopedics
Burn unit
Tumor therapy
unit

72
90
10
14
6
21
4
5

32.4
40.5
4.5
6.3
2.7
9.5
1.8
2.3

16
38
2
4
5
15
7
24

14.4
34.2
1.8
3.6
4.5
13.5
6.3
21.6

OR

P value

Adjusted OR (95% CI)

2.21(1.03-4.79)
0.97(0.48-1.97)
1.30(0.61-2.76)
1
1
1.18(0.62-2.24)
2.13(1.02-4.46)
2.82(0.88-9.13)
1
4.16(1.15-17.82)
1
7.59(2.94-20.35)

0.042
0.937
0.577

6.373(2.16-18.83)
0.741(0.296-1.856)
1.155(0.421-3.169)

0.709
0.0447
0.09

0.94(0.41-2.18)
3.68(1.46-9.30)
2.28(0.52-10.04)

0.025

7.566(1.009-56.708)

0.000001

0.006(0.00-12)

0.00032
0.00005
0.00000

13.43(2.65-68.08)
9.29(1.74-48.64)
70.15(13.96-352.46

0.079
0.785
0.945
0.096
0.012
0.003
0.000

1.68(0.42-6.71)
1.21(0.19-7.56)
1.83(0.36-9.21)
2.89(0.53-15.59)
4.86(1.31-18.01)
4.89(0.93-25.83)
1778.85(0.00-34)

1
11.34(2.4-73.44)
14.8(3.02-98.38)
59.52(13.41368.48)
1
1.9(0.94-3.89)
0.90(0.0-5.09)
1.29(o.31-5.001)
3.75(o.85-16.41)
3.21(1.26-8.27)
7.88(1.77-37.27)
21.6(6.47-77.05)

15

16

References
1.
2.
3.
4.
5.
6.
7.
8.
9.

10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.

Jarvis W.R : Epidemiology of nosocomial infections in pediatric patients. Pediatr. Infect. Dis. J. 1987; 6: 344-51
Moss W. An outbreak of getamycin-resistant klebsiella bacteremia at a children's hospital. Ethiop Med J. 1992;
30: 197-205
Worku B. Klebsiella oxytoca outbreak at Ethio-Swedish childrens hospital (ESCH). Ethiop Med J. 1997; 35:17783
Cooper R.G., Sumner C. Hospital infection data from a childrens hospital Med J. Aust. 1970; 2:1110-13
Gardner P, Carles D.G Infection acquired in a pediatric hospital. J. Pediatr. 1972; 81: 1205-1210
Wenzel R.P, Osterman C.A., Hunting K.J. Hospital acquired infection 11. Infection rates by site, service and
common procedures in a university hospital. Am. J. Epidemiol. 1976; 104: 645-51
Welliver R.C., Mc Laughlin S. Unique epidemiology of nosocomial infection in a childrens hospital. Am. J. Dis.
Child. 1984; 138: 131-5
Gaynes R.P., Edward J.R., Jarvis W.R., Culver D. H, Tolson J. S., Martone W. J. Nosocomial infection among
neonates in high- risk nurseries in the United States. Pediatrics. 1996; 98: 357-61
Sohn A. H., Garret D. O., Sinkovitz- Cochran R.L., Grohskopf L. A., Levine G. L., Stover B.H., Seigel J.D.,Jarvis
W.R. Prevalence of nosocomial infections in neonatal intensive care unit patients: Results from the first national
point prevalence survey.J.Pediatr .2001;133:821-827.
Behrman R, Kleigman R, Jenson H. T lymphocyte, B lymphocyte and Natural Killer cells.-In: Buckley R., ed ;
Infection control and prophylaxis.-In: Fisher, ed. Nelson, Text Book of pediatrics, 17th ed. New Delhi, Elsevier,
2004:683, 866, 1184.
Feigin, Cherry. Nosocomial infections.-In: Huskin W.C., Goldman D.A., eds., Textbook of pediatric infectious
disease, 4th ed. Philadelphia, Elsevier, 1998; 2545-85.
Fleming C.A., Balaguera H.U., Craven D.E., Risk factors for nosocomial preumonia.Med chinics of north
America 2001;85:1545-1563.
Knittle M.A., Eitzman D.V., Baer H, Role of hand contamination of personnel in the epidemiology of gram
negative nosocomial infections: J.pediatr 1975; 86:433-437.
Chandra R.K: Nutrition, immunity, and infection: Present knowledge and future direction.Lancet.1983; 1:688-91.
Puri S, Chandra P.K: Nutritional regulation of host resistance and predictive value of immunlogic test in
assessement of outcome. Pediatric Clinics of North America.1985; 32:499-516.
Bhattacharyya N, Kosloske A.M., Macarlhia C.Nosocomial infection in pediatric surgical patients: a study of 608
infants and children: J.pediatr.Surg.1993; 28:338-343.
Bhattacharyya N, Koslosk A.M., postoperative wound infectious in pediatric surgical patients: A study of 676
infants and children. J.pediatr surg.1990; 25:125-129.
Haley R.W., Houton T.M., Culver D.H., Stanley R.C., Emori T.G., Hardisan C.D., Quade D, Schauchtman
R.H.Schaberg D.R., shah B.V., Schat, G.D., Nosocomial infections in U.S. Hospitals, 1975-1976 estimated
frequency by selected characteristics of patients AM J Med 1981; 70: 947-959.
Calis R, Torres A, Gatell J.M., Almale M, Rodringuez-Roisin R.,Aqustin vidal A. Noscomial preumonia. A
multivariate analysis of risk and prognosis. Chest 1988; 93:318-24.
Price D.J., Sleigh J.D., Control of infection due to klebsiella aerogenes in a neurosurgical unit by withdrawal of all
antibiotics. Lancet 1970; 2:1213-1215.
Young L.S. Nosocomial infection in the immuocompromised adult. Am. J .Med 1981; 70:398-403.

16

17
22.
23.
24.
25.
26.
27.
28.

Garlsand I.S.,Nelson D.B., Cheah T.E., Hennes H.H., Johnson T.M. infectious complications during peripheral
intravenous therapy with Teflon catheters: a prospective study.pediatr infect dis. J. 1987; 6: 918-921.
Tully J.L., Friedland F.H., Baldini L.M., Goldmann D.A., Complications of intravenous therapy with steel needles
and Teflon catheters. Am. J Med 1981; 70:702-6.
Ducel.G, Fabry.J, Nicolle.L; Prevention of hospital acquired infections A Practical Guide 2nd
edition.WHO/CDS/EPH/2002.12
Garner J.S., Jarvis W.R., Emori T. G., Horan T.C. , Hughes J .M.. CDC definition for nosocomial infections. Am J
Infect Control 1988;16:128-140.
Waterlow. J.C. Note on the assessement and classification of protein energy malnutrition in children. Lancet
1973; 2:87-98.
Classification of infantile malnutrition. Lancet 1970; 2: 302-303.
Grady R., Safety profile of quinolone antibiotics in the pediatric population. Pediatr Infect Dis J 2003; 22: 112832.

Acknowledgements
I thank Dr Fithun Lulseged and staff at the statistics office who helped me collect data. I also extend my gratitude to Dr
Adamu Adissie and Dr Firew Mekonnen who helped me in the analysis of the data.

17

18

Guillain Barre Syndrome in Children At Tikur Anbessa Specialized Hospital

Tigist Bacha, MD

Abstract
This is a six years retrospective descriptive study conducted in the pediatric and child health department of Tikur Anbessa
Hospital from September, 2001 September, 2006 G.C to assess the pattern of acute flaccid paralysis (AFP) and the
clinical and epidemiologic features of Guillain Barre syndrome (GBS).
Data was collected from medical records of all patients admitted with diagnosis of AFP, and analyzed using standard
statistical tests with SPSS version 14 software. Out of 70 admitted cases of AFP, forty six (65.7%) were males and 24
(34.3%) were females. Sixty seven cases (95.7%) were diagnosed to have GBS and the rest three were compatible with
poliomyelitis, transverse myelitis and post injection neuritis. Out of the 70 cases 66 (94.3%) have received at least one
dose of polio vaccination and the rest 4 (5.7%) were never vaccinated.

Out of the cases diagnosed to have GBS, 44 (65.7%) met NNCDS diagnostic criteria. History of antecedent event was
obtained in 31/67 (46.3%) patients. Majority of the patients 45 (67.17%) presented with ascending reflexes quadriparesis,
2 (2.98%) patients with descending areflexic quadriparesis, 19(28.35%) only with lower limb involvement and 1 (1.5%)
with typical miller-fisher type. Sensation was affected in 4 patients.

Cytoalbuminological dissociation was found in 27(40.3%). There were 11 deaths (16.4%) of whom five were admitted to
ICU the rest six didnt. This study showed that the commonest cause of AFP is GBS which is associated with high
mortality. This high mortality rate 11/67 (16.42%) is attributed to absence of pediatrics ICU, late arrival to Hospital after
onset of illness, and poor supportive care.

18

19

INTRODUCTION
Acute flaccid paralysis (AFP) is the clinical condition

The objectives of the study are to describe pattern of

diagnosed in any child under 15 years of age with

AFP, determine the most common cause of AFP, and

acute floppiness of one or more limbs or any age in

determine the clinical and epidemiological feature of

whom clinician suspect polio (1,2).It is caused by many

GBS.

conditions including Guillain Barre syndrome (GBS) ,

poliomyelitis,

transverse

myelitis

and

metabolic

Material and method

neuropathy like hypokalemia (3,4).The most common

Retrospective descriptive study was done in pediatric

cause of AFP is GBS followed by transverse myelitis

and child health department of Tikur Anbessa Hospital,

(2,5,6,7)

Addis Ababa. Study groups were all children admitted

with the case definition of AFP from September 2001

In Ethiopia, the polio eradication initiative (PEI) was

started in 1996. Surveillance for acute flaccid paralysis
(AFP) was initiated in May, 1997. AFP surveillance is
the detection of at least one AFP case per 100,000
children under 15 years of age. AFP surveillance
depends on immediate reporting, investigation of AFP
cases, routine monthly reporting of cases including zero
reporting (1, 9).

September 2006. The medical report of Statistics Office

registered for AFP surveillance, Pediatric wards,
surgical and medical ICU and neurology clinic
registration books were reviewed to trace all cases. This
hospital is the only pediatric tertiary Hospital serving for
all the country.
In this study:
Antecedent triggering event was defined as the

GBS is an acute inflammatory polyneuropathy. The

presence of respiratory, gastrointestinal, febrile illness

cause is unknown but autoimmunity is incriminated.

or vaccination for the previous 4 weeks.

Unlike polio, GBS is usually symmetrical (asymmetrical

in 9%), and fever at presentation is not present.
Paralysis develops acutely over days, or at most weeks.
After brief plateau the patients improvement begins with
gradual resolution that lasts from weeks to months. 50%
has bulbar involvement, 33% of them require ICU
admission 25% required mechanical ventilation 5-10%

Polio-compatible case is defined as a case in which

one adequate stool specimen was not collected from a
probable case within 2 weeks of the onset of paralysis,
and there is either an acute paralytic illness with poliocompatible residual paralysis at 60 days, or death takes
place within 60 days, or the case is lost to follow-up.

mortality rate (2-3% in best ICU) and 80% complete

The diagnosis of GBS in this study is based on National

recovery. Treatment is intensive care support; early

Institute of Neurological and Communicative Disorders

intravenous immunoglobulin (IVIg) therapy and plasma

and stroke (NNCDS) diagnostic criteria. (11)

exchange hasten early recovery. Corticosteroid alone

outcome is controversial (3,8,10).

19

1. Progressive weakness of more than one limb

Results

due to neuropathy,

In this six year period 70 AFP cases (46 male and 24

2. Areflexia or hyporeflexia

female) were admitted and majority 67 (95.7%) were

3. Duration of progress less than 4 weeks,

GBS. Three cases were suspected to have a

4. The absence of a sharp sensory level on the

compatible poliomyelitis, transverse myelitis and post

trunk,

injection neuritis each accounting 1.4%. The mean age

5. The absence of other causes of acute

neuropathy

of AFP is 5.82 year with a range of 1-12 year. Cases

were reported from different regions of Addis Ababa 44,

6. Less than 50 mononuclear leukocytes per mm3

of CSF.

Oromia 12, Amhara 11 and South Nations Nationalities

and Peoples region (SNPPR) 3 (Fig 1). The distribution

The data were transferred to a structured form which

over the years is shown in Fig 2 from 2001-06 (1993-

was then entered into computer data base. SPSS

1998 Eth C). 1More cases were seen in 1996 Eth C.

Version 14, 2005 was used to process the statistical

Stool sample was taken for polio in 65 cases and was

data.

not taken in 5 of them. Out of the 70 cases 66 (94.3%)

of the cases got at least one dose of polio vaccination
and the rest 4 (5.7%) never vaccinated (table 1). Prior
injection history is found in 7 (1%) of the patients.

50

NO of AFP Cases

40

30

20

10

0
Amhara

Oromiya

SNNPR

Region

Addis Ababa

21

Fig. 1 AFP case Regional distributions

25.0%

AFP cases in Percent

20.0%

15.0%

10.0%

5.0%

0.0%
1993
Fig 2. AFP cases distribution
per year

1994

1995

1996

1997

1998

Admition Year

Table 1. Cranial nerve involvement

Cranial nerve involved
Facial nerve
Opthalmolopelgia
Glossopharyngeal nerve
Vagus nerve
Multiple cranial nerve
Total

Frequency
3
1
5
1
1
11

Percentage
27.2
9.1
45.4
9.1
9.1
100

21

22
Table 2: CSF analysis Result

Frequency

Percent

4.5

10.4

27

40.3

21

31.3

CSF cell count >10/cumm

and protein >45mg
CSF cell count >10/cumm
and protein <45mg
CSF cell count <10/cumm
and protein >45mg
CSF cell count <10/cumm
and protein <45mg
Not Documented
Total

9
67

13.4
100.0

From those diagnosed as having GBS 44(62.8%) met

from 1 day to 14.0 days the mean being 3.44 days. The

NNCDS diagnostic criteria. The mean age was 5.9 with

mean duration of hospital stay is15.5days. Out of the 17

ranges of 15 months -12 years. Most of them were male.

patients whose blood pressure was measured 12 patients

History of antecedent event was obtained in 31(46.3%)

had normal measurement and 5 had transient

patients. Respiratory symptoms accounted 19(27.1%),

hypertension. Bladder and bowel dysfunction was

gastrointestinal symptoms 9 (11.4%), both gastrointestinal

reported in 44 out of 64 documented cases. Cranial nerve

and respiratory symptoms 1 (1.4%), 1 (1.4%) post

palsy was reported in 11 patients. As shown in table xxx

vaccine (Rabies), and 1 (1.4%) had prior malarial attack.

the commonest were cranial IX (45.4%) followed by facial

The pattern of progression of paralysis was ascending

nerve (27.2%). Multiple cranial nerve involvement was

areflexic quadriparesis in 45 (67.2%) patients, descending

found in 1(9.1%) patient. Sensation was affected in 4

areflexic quadriparesis in 2(3%), patients with only lower

(6%) cases. Ataxia was documented in 8(11.9%) patients.

limb involvement were 19 (28.3%) and one (1.4%) was

Cytoalbuminological dissociation was found in 27(40.3%)

Miller - Fisher variant. The mean interval between onset

patients. see table 5. EMG was done in 5 patients with

of symptoms to hospital admission was 6 days with a

GBS out of which 2 were demyelinating, 2 axonal and 1

range of 5 hrs-14days. Rapidity of the progression ranged

mixed

axonal

and

demyelinating.

Fifteen15/67 (22.39%) patients required ICU care out of

infection were cause of death in 5 of them (3 had

which 6 didnt get the service. There were 11(16.4%)

pneumonia, 1 urosepsis). Specific treatment such as

deaths. Five died in ICU and 6 patients died in the pediatric

plasmapheresis and Immunoglobulin was given to none of

ward. Four of them came within 4 days of onset of illness

the patients. Prednisolon was given only for 4 patients

and 7 of them after 4 days. Respiratory failure was

(1.0%).HIV screening was done only for one patient and

considered as a cause of death in 6, Respiratory failure and

the result was non reactive.

22

Discussion
This is the first study done on pattern of AFP and clinical and

This study showed a high need of having pediatric ICU. We

epidemiology of GBS in the pediatric age groups. The

can see the mortality rate from GBS is high some of which

predominance of male is similar to other studies (6-10). The

could have been prevented if there was a pediatric ICU.

common AFP identified was GBS which also correlates with

The care of critical patients and universal precautions for

other studies in Australia, Bangladesh, and Honduras (2,6-8).

infection prevention should be encouraged. In addition the

use of specific treatments such as plasmaphereis and

Regarding GBS the antecedent events are lower in this study

immunoglobulin should be introduced.

than other studies done in Kenya, Tanzania, Nigeria and

including the study done in this hospital in adults (12-15).

For five cases stool sample for polio was not taken out of

Similar to other studies respiratory symptoms were the

whom four died. Taking specimen as soon as possible

commonest antecedent events (12). The cranial nerve

especially on critical patients, including on weekend may

involvement are common findings similar to that of Kenya

increase the surveillance. Improving the AFP surveillance

and adult Ethiopian patients (12, 13).

increased by case investigation reporting cases as early as

possible especially on critical patients should be

In this study one patient gave history of antecedent malarial

encouraged.

attack the species not documented. From other studies the

In addition care seeking behaviour should be improved so

development of GuillainBarr syndrome was reported in 10

that patients are brought to hospital as early as possible.

patients who had had acute P. falciparum malaria during its

Finally a further large scale study with more investigation

seasonal exacerbation is reported (16).

modalities should be done in the near future.

The high mortality rate is higher to the report from other

studies done in Kenya (13) and Tanzania (9). Also a
higher mortality was observed in adult study done in the
same hospital (12). The main attributable cause is lack of
good pediatric ICU in this Hospital

which could be compounded by the late appearance of

patients to the hospital. In addition specific therapies such
as immunoglobulin and plasmapheresis are not available in
the setup.

24

References
1. Acute Flacc
2. id paral
3. ysis surveillance: Ministry of Health and WHO Ethiopia June 2006: 5
4. Koul R, Chacko A, Javed H et al Acute flaccid paralysis in Australian children. . J Paediatr Child Health.
2003; 39 (1):22-6.
5. Hahn AF. Guillain-Barre syndrome. Lancet 1998; 352: 635-41.
6. Lovecchio F, Jacobson S. Approach to generalized weakness and peripheral neuromuscular disease.
Emerg Med clin N am 1997; 15(3):605-23
7. Rehman A, Idris M, Elahi M, Arif A. Guillain-Barre syndrome. The leading cause of Acute Flaccid Paralysis
in Hazara Division. J ayub Med Coll Abbottabad 2007;19:26-28.
8. Rasul CH, Das PL, Alam S, Ahmed S, Ahmed M. Clinical profile of acute flaccid paralysis. Med J Malaysia.
2002 ; 57 (1):61-5.
9. Molinero MR, Varon D, Holden KR, Sladky JT, Molina IB, Cleaves F. Epidemiology of childhood GuillainBarre syndrome as a cause of acute flaccid paralysis in Honduras: 1989-1999.
J Child Neurol. 2003; 18(11):741-7..
10. Kuwabara S. Guillain-Barre syndrome: epidemiology, pathophysiology and management. Drugs.
2004;64(6):597-610. .
11. Berhane Beyene, Ayele Gebremariam, Tilahun Teka et al. Laboratory and Epidemiology communications,
Regional Distribution of Acute Flaccid Paralysis Cases in Ethiopia in 2000 2002. Jpn.J.Infect. Dis., 2004;
52:. 72.
12. Harvey B.Sarnat. Neuromuscular Disorders. In: Richard E. Behrman, Robert M. Kliegman and Hal B.
Jenson, Editors. Nelson Text Book of Pediatrics, 17th Edition .Philadelphia Pennsylvania, 2004:2080-81.
13. Asbury Ak. Assessment of current diagnostic criteria for Guillian-Barre syndrome. Ann Neurol
1990:27(suppl):s21-4
14. Zenebe Melaku, Guta Zenebe, Abera Bekele .Guillaian barre syndrome in Ethiopian patients. Ethiop Med J
2005; 43( 1) :21
15. Bahemuka M. Guillain-Barre syndrome in Kenya: a clinical review of 54 patients.
J Neurol. 1988; 235(7):418-21.
16. Howlett WP, Vedeler CA, Nyland H, Aarli JA. Gullian-Barrre syndrome in Northern Tanzania: a comparison
of epidemiological and clinical findings with western Norway. Acta Neurol Scand 1987:75:95-100.
17. Osuntookun BO, Agebebi K. Prognosis of GuillainBarre syndrome. Medicine in the African Region: The
Nigerian experience. Neurol Neurosurg Psychatry 1973:36:478-83.

24

25

Acknowledgments:
I extend my thanks to Dr. Ahmed Bedru for his invaluable advice in the conduct of the research. Ato Tilahun
Zimita for his assistance in data entry and analysis.

25

26

ASSESSMENT OF QUALITY OF CARE OF SICK UNDER-FIVE CHILDREN IN

REFERRAL HOSPITALS IN ETHIOPIA
Sirak Hailu, MD1, Solomon Emyu, MD/MPH2, Fisseha Mamo, MPH3, Tolawaq Kejela MD4

ABSTRACT
Background: About 10-20% of sick children presenting to a primary care facility require referral to hospital for
inpatient care. Improvement of the quality of pediatric referral care has a major contribution to the child survival
efforts by ensuring the continuum of care and averting mortality.
Objective: To assess the quality of care for children in selected referral hospitals based on the minimum standards
derived from the WHO Pocket book of Hospital Care for Children, 2005 and thereby to initiate pediatric referral care
quality improvement process in the country.
Methods: A qualitative assessment of pediatric referral care was conducted in 8 hospitals selected by convenient
sampling, January July 2008. A team composed of experienced pediatricians and health officer used an adapted
WHO hospital assessment tool to assess the quality of triage, emergency care and case management practices &
hospital infrastructure and support services.
Results: None of these hospitals were practicing the standard triaging process by assessing children immediately on
arrival for emergency and priority signs. All of them were not appropriately organized and fully equipped to handle
pediatric emergencies effectively. Overall, the case management of common neonatal and childhood illnesses was
not optimal. Generally, there was shortage of some essential drugs and lack of materials such as nasal prongs, infant
and child size bag & masks, nebulizers, heaters and oxygen concentrators. Hygienic facilities were below the
expected standard. Staff were not trained in ETAT (Emergency Triage Assessment and Treatment) and there were
no protocols for pediatric referral care. There was no clearly designated high dependency area where very sick
children receive highest attention and no special rooms for providing appropriate neonatal care in majority of the
hospitals. The overall case fatality rate was 11% (10-16%) but first 24 hours mortality could not be determined due to
problems with the recording system.

Conclusions: The quality of pediatric referral care needs serious attention and coordinated efforts utilizing the
opportunity of the national hospital management initiative and the BPR (Business Process Re-engineering) to
institutionalize ETAT and standards of hospital care for children. This has to be complemented with availing of
appropriate job aids, essential supplies and equipments, and improvement of health worker skills through training,
clinical mentoring and regular supportive supervision.
1

WHO/Ethiopia, 2 WHO/Ethiopia, 3 FMOH/Addis Ababa, 4 Medical Faculty/AAU,

26

27
practices, and care at the first-level hospital. On the
basis of current guidelines, it has been estimated that
10% to 20% of sick children who present for primary
care (i.e., the most severely ill) require referral to a
first referral or district hospital. The quality of care
provided in these hospitals is likely therefore to have
a major impact on the health and lives of millions of
children each year.

INTRODUCTION
The Ethiopian health service delivery system is
organized as a four-tier system, characterized by a
Primary Health Care Unit (PHCU) 1st tier, then the
district hospital 2nd tier, zonal hospital 3rd tier and
specialized hospital- 4th tier. Services given at each
level of these tiers have a crucial role in averting the
morbidity and mortality burden of children and
contributing a lot to the achievement of healthy
society. In relation to this, Ethiopia has developed a
comprehensive national child survival strategy, which
is part of the national HSDP-3, and is implementing
the IMNCI approach at a wide scale at health facility
and community levels. Currently the national IMNCI
coverage of the country at Health Center levels is
about 60%. These interventions at a community and
frontline health facilities (PHCU) levels are very
important to address the majority of the health
problems of children and also to make the services
close and easily available to the society.

It is estimated that 40-60% of deaths at the referral

hospitals occurs in the first 24 hours of admission.
Many of these deaths could be prevented if very sick
children are identified soon after their arrival in the
health facility, and treatment is started immediately.
The Emergency Triage Assessment and Treatment
(ETAT) tool is designed to enable health workers
triage all sick children when they arrive at a health
facility, into those with emergency signs, with priority
signs, or non-urgent cases. It also enable them to
provide emergency treatment for life-threatening
conditions. The standards of care of the ETAT
guidelines correspond to the minimum standards that
should be maintained even in small hospitals and
where resources are limited.

The IMCI/IMNCI strategy seeks to strengthen

prevention and care for children through appropriate
community and household care, primary care, referral
Early assessment and prioritization for management
of sick children attending a health service are critical
to achieving good health outcomes. Experience of
Malawi in implementing ETAT showed a reduction of
total inpatient mortality from 10-18% (median 12.4%)
to 6-8% (median 5.7%) over a 2 year period (2001-

2003) and deaths within 24 hours of admission from

36% to 12.5%.
This survey tries to assess the current situation of the
quality of pediatric care in selected government
hospitals.

OBJECTIVES & METHODOLOGY

adapted and used for data collection. Adaptation of
the tool was done by practicing pediatricians, general
practitioners and nurses and it was pretested during a
five days national orientation workshop on pediatric
referral care.

The objective of this study was to assess the quality

of care for children in selected referral hospitals
based on the minimum standards derived from the
WHO Pocket book of Hospital Care for Children,
2005 and to initiate pediatric referral care quality
improvement process by identifying key areas that
need immediate and long term action.

The assessment tool had 12 major sections: General

hospital information, Hospital layout and structure,
Hospital support systems (drugs, equipment and
laboratory), Emergency care, Pediatric wards (layout,
facilities, staff, supplies & equipments), Case
management in the ward (Cough or difficult breathing,
Diarrhoea, Fever conditions, Malnutrition, HIV/AIDS),
Supportive care & nutrition, Monitoring of patients,
Neonatal Care (layout and staff, Routine neonatal
care and Sick newborn care), Paediatric surgery and
rehabilitation, Hospital administration and Access to
hospital.

Eight hospitals (Adama, Ambo, Bishoftu, Debre

Birhan, Dessie, Yekatit 12, Yirgalem and Zewditu
hospitals); one district, six zonal and one regional
referral hospital, were included in the assessment
using convenient sampling. The main criterion used
for including a hospital was having functional
pediatrics outpatient and inpatient services. WHOs
generic assessment tool, Assessment of the
quality of care for children in hospitals, was

27

28

The adapted assessment tool was used to collect

information from all hospitals, January to July 2008.
Data was collected by two teams each composed of a
pediatrician and a health officer who were trained and
also involved in the adaptation of the data collection
tool. Four different methods of data collection were
used namely; hospital visit, case management
observations, records review and interviews of
caretakers and providers which took two full days for
assessing each hospital. Hospital visit: Direct
observation of the hospital layout and structure, OPD
attendance, admission rates, availability of essential
drugs, availability of diagnostic and therapeutic
equipments were made. Areas of doubt were clarified

by interviews. Case observations: The care for

admitted children to the hospital was observed
without interference from the assessors. This is
complemented by discussion of the cases with staff,
review of the case records and monitoring charts, and
interviewing the caretakers. Records review:
Assessors obtain information on the quality of care for
admitted and recently discharged patients by
checking records. If there are not sufficient patients
for direct case observations, assessors reviewed at
least 3-5 records for each clinical problem.
Interviews: Assessors conducted interviews with
hospital staff and caretakers to gain some idea of
their perception of care for children in the hospitals.

Each major assessment section consisted of a

number of standards and each standard was qualified
by several detailed criteria. Thus, each of the 12
sections of the assessment tool was scored based on
the standards and the criteria to meet these
standards. The standards are the minimum
requirements for good quality of care for children as
defined by the WHO Pocket book of Hospital Care
for Children, 2005. The detailed criteria of each
standard were rated/scored based on the need for
improvement as good or need to be improved,
where Good means the criteria is similar to the
Finally, a total summary evaluation score for complete
assessment of all sections was marked in the
summary evaluation sheet. The total summary score
can assist in monitoring hospital improvements over
time and to make also inter-hospital comparisons to
some extent.

standard, and To be improved means the

component is below the expectation of the WHO
standard and needs improvement. A summary score
from 5 to 1 was awarded at the end of each section
whereby 5 indicates Good practice complying with
standards of care while scores from 4 to 1 indicate
practices To be improved (4 = Little improvement, 3
= Some improvement, 2 = Considerable/Significant
improvement is needed to reach standards of care,
and 1 = Services not provided, totally inadequate care
or potentially life-threatening practices).
At the end of each hospital visit, assessors met with
hospital administration and staff for a debriefing and
to identify priority actions for improvement of the
hospital services.

RESULTS
General hospital information
All surveyed hospitals had isolated pediatrics OPD
and ward except Bishoftu hospital whose pediatric
inpatient room was part of the adult medical ward.
Separate pediatric waiting area and archive rooms
were found in only 3 and 2 hospitals respectively. Six
of the 8 hospitals had separate room for admitting
paediatric infectious cases while only 3 had separate

room for admitting newborns. In 6 hospitals, children

with surgical conditions were admitted in the general
pediatric ward. None of the hospitals had a clearly
designated high dependency area or room where
very sick children are cared for and receive closest
attention.

Electricity and running water were available in all

hospitals even though most of the hand washing and
toilet facilities were non-functional in half of the

hospitals. Five hospitals had backup power supply. All

hospitals had appropriate sharp disposal boxes.

28

29
The pediatric service in these hospitals caters for
children up to the age of 14 years. Based on available
data for the period July 2006 to June 2007, there
were on average 29 (range 12-66) children seen in
the Paediatric OPD with about 5 admissions per day
(range 1-9). The two commonest causes of OPD
attendance and admission were severe pneumonia
and diarrhoea with severe dehydration or dysentery.
The mean bed capacity of the assessed hospitals
was 39 (range 11-68) and the average bed
occupancy rate was 58% (range 31% in Zewditu &
90% in Ambo). Children under 5 account for 64% of
admissions to the pediatric wards. Based on available
data from 5 hospitals, the average daily emergency
patient load was 7 and the average all cause mortality
in under fives was 11% (10-16%).

As shown in Tables 1 & 2, the availability of

emergency drugs and equipments in the emergency
area and the wards was inadequate. Drugs like
parenteral
Phenobarbitone
(long
acting
anticonvulsant) were non-existent and those drugs
that were available were often not immediately
accessible. Only 3 hospitals had ambu bags (big
mask) in the emergency area and the wards while
infant size masks, nasal prongs and nebulizers were
totally absent. Oxygen concentrator and heat sources
were found in only one hospital. However, all
hospitals had good laboratory facilities to perform the
five basic laboratory tests for managing emergencies
(RBS, blood film, HGB, CSF microscopy and blood
group and cross match) even though results were
delivered timely in half of the hospitals. Bilirubin and
other chemistries were being done in 5 while culture
facilities were available in only 2 hospitals.

Hospital support systems (drugs, equipments

and laboratory)
Table 1: Availability of Essential Drugs
No
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23

Drugs
Glucose 40% i.v.
Glucose 5% i.v.
Normal saline i.v.
Ringers lactate i.v.
Epinephrine (Adrenaline) s.c.
Corticosteroids i.v. or p.o.
Furosemide i.v.
Diazepam i.m., i.v.
Phenobarbital i.m., i.v.
Paracetamol
Ampicillin inj.
Benzyl penicillin
Cloxacillin
3rd generation Cephalosporins
Chloramphenicol
Gentamicin
*All anti-malaria drugs
Digoxin
F75 milk
F100 milk
Ready to Use Therapeutic Food
Vitamin K i.m. injection
ORS

Emergency area

Ward

Pharmacy/store

5 (63%)
6 (75%)
6 (75%)
8 (100%)
8 (100%)
6 (75%)
5 (63%)
5 (63%)
0 (0%)
6 (75%)
6 (75%)
5 (63%)
5 (63%)
3 (38%)
3 (38%)
5 (63%)
4 (50%)
3 (50%)
2 (25%)
3 (50%)
2 (25%)
0 (0%)
6 (75%)

6 (75%)
7 (88%)
6 (75%)
7 (88%)
7 (88%)
5 (63%)
5 (63%)
5 (63%)
0 (0%)
6 (75%)
4 (50%)
5 (63%)
5 (63%)
4 (50%)
3 (38%)
5 (63%)
4 (50%)
4 (50%)
6 (75%)
7 (88%)
5 (63%)
0 (0%)
7 (88%)

7 (88%)
8 (100%)
7 (88%)
8 (100%)
7 (88%)
7 (88%)
5 (63%)
5 (63%)
0 (0%)
8 (100%)
7 (88%)
7 (88%)
6 (75%)
5 (63%)
4 (50%)
6 (75%)
8 (100%)
6 (75%)
6 (75%)
7 (88%)
6 (75%)
3 (38%)
8 (100%)

29

30
Table 2: Availability of essential Equipments & supplies
No
1
2
3
4
5
6
7
8
9

10
11

12
13

14
15
16
17

Equipments
Resuscitation table/area
Torch
Otoscope
Scales for children
Stethoscopes
Thermometers
Heat source
Lumbar puncture set
Oxygen source:
oxygen cylinder
oxygen concentrator
central supply
Flow-meters for oxygen?
Equipment for the administration of oxygen?
nasal prongs
catheters
Masks
Self inflating bags for resuscitation
Masks
infant size
child size
adult size
Butterflies and/or cannulas of paediatric size
NG-tubes, paediatric size
Suction equipment
Nebulisers for administration of Salbutamol

Emergency area

Ward

Pharmacy/ store

5 (63%)
4 (50%)
5 (63%)
7 (88%)
7 (88%)
8 (100%)
1 (13%)
5 (63%)
7 (88%)
7 (88%)
1 (13%)
0 (0%)
7 (88%)
7 (88%)
0 (0%)
7 (88%)
0 (0%)
3 (38%)

2 (25%)
4 (50%)
6 (75%)
8 (100%)
7 (88%)
8 (100%)
2 (25%)
6 (75%)
7 (88%)
7 (88%)
1 (13%)
0 (0%)
7 (88%)
7 (88%)
0 (0%)
7 (88%)
1 (13%)
3 (38%)

3 (38%)
4 (50%)
6 (75%)
7 (88%)
8 (100%)
7 (88%)
0 (0%)
3 (38%)
4 (50%)
4 (50%)
0 (0%)
0 (0%)
5 (63%)
6 (75%)
0 (0%)
6 (75%)
0 (0%)
2 (25%)

0 (0%)
3 (38%)
1 (13%)
7 (88%)
4 (50%)
5 (63%)
0 (0%)

0 (0%)
4 (50%)
1 (13%)
7 (88%)
7 (88%)
7 (88%)
0 (0%)

0 (0%)
2 (25%)
0 (0%)
7 (88%)
7 (88%)
6 (75%)
0 (0%)

Emergency care
Most of the surveyed hospitals (5/8) were not
appropriately organized, fully staffed and equipped to
handle pediatric emergencies effectively. Most of
these hospitals were not practicing the standard
triage process by assessing children immediately on
arrival for emergency or priority signs before
administrative procedures. However, three hospitals,
two of them implementing the new hospital
improvement initiative, had assigned qualified nurses

at the reception to facilitate smooth patient flow and

prioritization of the management of emergency cases.
Even in these three hospitals the triaging and
management of emergency cases was not up to the
expected standards of ETAT. None of the staff
working in the OPD and emergency areas had been
trained in ETAT and there was lack of job aids and
standard protocol for pediatric referral care (Table 3).

30

31
Table 3: Summary of grading of the emergency setup in the hospitals
Standards and criteria

No
1
2
3

4
5
6

7
8

9
10

Children are assessed for severity/ priority signs (triaged) immediately on arrival as per
the ETAT standard
Patients do not have to wait for their turn, registration, payment etc. before a first
assessment is done and action taken.
A wall chart or job aid for identifying children by severity of condition is located in the
emergency admissions area.
Drugs, equipment and supplies
Essential drugs for emergency conditions (anticonvulsants, glucose, iv fluids) are
always available and free of charge to the family
Essential lab tests (glucose, Hb or PCV) are available and results are obtained timely
Essential equipment (needles and syringes, naso-gastric tubes, oxygen equipment,
self-inflating resuscitation bags with masks of different sizes, nebulisers or spacers) is
available
Staffing
A qualified staff member is designated to carry out triage.
A health professional is available without delay to manage children determined to have
an emergency condition.
Case management
Staff doing triage is trained in the ETAT guidelines and can implement them
appropriately when the emergency room gets busy during peak hours
Staff is skilled in the management of common emergency conditions and starts
treatment without delay: Management of convulsions, lethargy, severe respiratory
distress, shock and severe dehydration.

Good

To be
improved

0 (0%)

8 (100%)

0 (0%)

8 (100%)

0 (0%)

8 (100%)

0 (0%)

8 (100%)

4 (50%)

4 (50%)

0 (0%)

8 (100%)

3 (38%)

5 (63%)

3 (38%)

5 (63%)

0 (0%)

8 (100%)

0 (0%)

8 (100%)

Pediatrics ward
Table 4 summarizes the status of the 8 hospitals against the WHO standards and criteria for pediatric wards.

31

32

32

33
Table 4: Standard's for children ward
No

Standards and criteria

Good

Children are seen in OPD only by the designated health professional in the designated
room/area.
Closest attention for the most seriously ill
The most seriously ill children are cared for in a section where they receive closest
2
attention.
Separate ward for children.
4
Children are kept in a separate ward or separate area of a ward.
Severely ill children are kept apart from adults in wards such as for infectious diseases
5
or intensive care.
Children with surgical conditions are at least kept in a separate room, with staff aware of
6
the special needs for children such as feeding and warmth.
7
Arrangements are made to meet these needs.
8
In cold climates, the ward has an efficient and safe heat source.
Separate room for sick neonates with mothers
9
Sick new-borns are kept separate from healthy babies.
10
Mothers of sick new-borns are rooming in with their babies, and have adequate facilities.
Hygiene and accident prevention
Staff has access to hand washing facilities The ward is kept clean and dangerous items
11
are inaccessible for children
12
Sharps are disposed of in a special container preventing accidents
Hygienic and sufficient services facilitate the stay of mother and child
13
There are sufficient and adequate toilets which are easily accessible
Mothers have access to running water and to an appropriate space, near the ward, to
14
wash themselves and their child.
1

To be
improved

6/7 (86%)

1/7 (14%)

0 (0%)

8 (100%)

7 (88%)

1 (13%)

4 (50%)

4 (50%)

3 (38%)

5 (63%)

1 (13%)
0 (0%)

7 (88%)
8 (100%)

3 (38%)
0 (0%)

5 (63%)
8 (100%)

3 (38%)

5 (63%)

8 (100%)

0 (0%)

0 (0%)

8 (100%)

0 (0%)

8 (100%)

15

Mothers have access to a washing facility, in order to wash her and her childs clothes.

0 (0%)

8 (100%)

16
17

Patients are kept in a bed/cot with a clean mattress.

Patients receive bed sheets

4 (50%)
6 (75%)

4 (50%)
2 (25%)

33

34
In the majority of cases, differential diagnosis of fever
was not considered exhaustively and investigations
were incomplete. In some instances planned LP tests
were not done and patients were treated empirically.
Inappropriate choice and administration of antibiotics
was observed and documentation of patient progress
and treatment given was not up to the standard in
almost all hospitals. Overall, patient records and
monitoring charts were poorly recorded.

Case management practices

Overall, the case management of common childhood
problems including pneumonia, diarrhea, fever
conditions and malnutrition was not optimal in almost
all hospitals.
Pneumonia was diagnosed and its severity correctly
classified more or less based on diagnostic signs in 2
hospitals. Similarly, appropriate use of antibiotics and
oxygen was observed in a third of the hospitals (3/8).
Patient monitoring and supportive care were
inadequate in all hospitals. None of the hospitals had
nebulizers for inhalation therapy.
Correct assessment of dehydration was documented
in half of the hospitals and the monitoring and
management of dehydration was inadequate in
almost all hospitals. Inappropriate use of antibiotics
for diarrhea was observed and supportive care
especially feeding was inadequate.

Most of the hospitals had the dietary supplies needed

for the management of severe malnutrition. Even
though routine antibiotics were given for
malnourished children as per the national guideline,
feeding was not given according to the recommended
schedule especially at night. Prevention and
management of other complications like dehydration
and hypothermia was inadequate. None of the
surveyed hospitals had rooms with heaters for
malnourished children. Monitoring of patients and
recording of progress was poor in the majority of
cases.

All the necessary guidelines and tools were in place

for counseling, diagnosing and staging of paediatric
HIV and for the treatment and monitoring of
antiretroviral therapy. There were well trained staff
who were regularly mentored and most of the HIV

services were conducted in line with the national

standard especially at the outpatient level. However,
inpatient management of some opportunistic
infections and the supportive care of these patients
need improvement.

Supportive care and nutrition

Multiple antibiotics were often prescribed some times
for longer durations. Screened blood was used in all
hospitals but the indications for its use and the
administration procedure needs improvement.

Although breastfeeding was encouraged, none of the

hospitals provide appropriate routine pediatric diet for
children. Misuse of intravenous fluids was common
including in children with severe acute malnutrition.

Monitoring of patients
All admitted children were assessed by a doctor at
admission and majority of them were re-evaluated
about once a day during working days and upon
consultation during weekends. A qualified nurse was
available 24 hours per day in the children's wards in
all hospitals and a medical doctor in 6 of them.
Nutritional status was assessed at admission using
weight for age in most hospitals. All patients had
individual charts, vital sign and medication sheets, but
the case histories, progress notes and the other

monitoring charts were not properly and regularly

recorded in the majority of cases both by physicians
and nurses. There was no designated area where
very ill children receive highest attention. Except for
the Severe Acute Malnutrition Chart, there were no
standardized monitoring charts consisting of all
relevant parameters which could simplify patient
monitoring.

34

35
Neonatal care
Early and exclusive breastfeeding, skin to skin
contact and proper thermal protection was not
practiced adequately in almost all hospitals. Eye and
Vitamin K prophylaxis and immunizations were given
routinely in only 2 hospitals. Neonatal resuscitation
flow chart was available in only one hospital.

one hospital was well staffed and equipped (heaters,

oxygen sources, phototherapy and suction machines,
resuscitation materials and other supplies) to closely
monitor seriously ill newborns with 24 hour availability
of skilled nursing staff. However, rooming-in and
hygienic facilities for mothers were inadequate in all
hospitals. As shown on Table 5, the case
management of common neonatal problems is
inadequate and needs strong effort to improve the
situation.

Overall, only three hospitals had a separate room for

sick newborn babies with bed capacity ranging from 3
to 17. In the other 5 hospitals neonates were kept in
the general pediatric ward or the maternity room. Only
Table 5: Sick newborn case management summary table
No
1
2
3
4

Neonatal sepsis is appropriately diagnosed.

Neonatal sepsis is appropriately treated.
Specific feeding needs of sick young infants and those with
low birth weight, are met.
Jaundice is adequately recognized and managed.

Good
2/6 (33%)
1/6 (17%)

To be improved
4/6 (67%)
4/6 (67%)

1/6 (17%)

5/6 (83%)

1/6 (17%)

5/6 (83%)

Pediatric surgery & rehabilitation

Hospital administration & access to hospital
Of the 8 hospitals surveyed, one (Ambo Hospital) was
not doing any major surgery for children. In 6 of the
hospitals, children with surgical conditions were
admitted in the general pediatric ward. Pre- and postoperative starving was kept to a minimum in the
majority of the hospitals which perform major surgery.
Overall, frequent post-operative monitoring of vital
signs and the readiness for resuscitation was
inadequate in nearly half of the hospitals. Only one
hospital had basic rehabilitation facilities.

Nationally, there were no pediatric treatment

guidelines for referral hospitals and none of the
hospitals practice pediatric death audits. Availability of
essential drugs and equipment that are basic for
provision of quality pediatric emergency and inpatient
services are inadequate in almost all of the hospitals
surveyed. However, the handling of available drugs
including the stock management was good and old
drugs were used first.

Economic and transportation barriers were serious

concerns for all the care takers interviewed and most
of them had tried traditional medicine before seeking
care from providers. Caretakers were not satisfied
with the communication from providers regarding the
problems of their sick children, explanations about
ward procedures and the treatments and follow-up
details. Almost all care takers indicated that the

condition of the toilet and washing facilities was very

poor.

Summary
The following table and figures 1 & 2 show the
summarized score by major sections of the
assessment and by facility. Table 6 is a summary
sheet showing the details of the summary scores of
the 8 hospitals in the different service areas.

35

36

Table 6: Details of summary score of the 8 hospitals in the different service areas
No
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17

Summary score of different service areas

Essential drugs, supplies and equipment **
Laboratory support
Emergency area and management
Children's ward and facilities
Cough or difficult breathing child management
Diarrhoea patient management
Febrile child management
Malnourished child management
Management of child with HIV/AIDS
Summary score of supportive care
Summary score in monitoring
Routine neonatal care service
Neonatal Care Unit facilities
Case management and sick newborn care
Paediatric surgery and rehabilitation
Summary score hospital administration
Summary score access to hospital
Total score = 405
Hospital summary score = total score / 8*17

Good
5

4
6

1
3

1
2
2
5

1
1
1
2

30

To be improved
3
2
1
5
3
2
6
2
8
5
6
5
1
7
6
4
6
1

4
4
5
84
222
2.98

2
4
1
6
7
1
4
3
68

NB: - ** = Five of 8 hospitals had a summary score of 3 out of 5, while 3 hospitals has a summary score of 2 out of 5 for
emergency drugs, supplies and equipment section.

Similarly, Figure 1 shows average scoring of the

quality of the different pediatric care services in the 8
hospitals out of a maximum score of 5. As it can be
seen clearly from the figure, there are marked
variations in quality of the different pediatric

care/services rendered. The quality of care given to

children with HIV/AIDS is relatively in far better
condition compared to services given to neonates,
scores of 4.4 and 2.3 out of 5 respectively.

36

37
Figure 1: Average summary score of the different pediatric care services out of a maximum score of 5

2.3

Sick newborn care

2.4

Neonatal Care Unit facilities

2.5

Hospital administration

2.5

Monitoring
Access to hospital

2.6

Ess drugs & equipment

2.6
2.8

Supportive care

2.8

Emergency area & mgt

2.9

Surgery and rehabilitation

3.0

Routine neonatal care

3.0

Ward and facilities

3.1

Febrile child mgt

Malnutrition

3.3

Diarrhoea mgt

3.3
3.6

Cough mgt

3.8

Laboratory support

4.4

HIV/AIDS mgt

0.0

1.0

2.0

As shown in Figure 2 below, the hospitals quality of

care total summary score ranged from 2.6 to 3.44 out
of 5 with an average of 2.98 (~3.0) which indicates
that all hospitals need substantial actions for
improvement to reach defined standards. This figure

3.0

4.0

also shows the difference in the performance of the

eight hospitals.

37

5.0

38
Figure 2: Total summary score of the 8 hospitals, out of a total score of 5
5.0

4.0

3.4
3.1
2.9

3.0

3.0

2.9

3.1

3.0

Yirgalem

Zewditu

score

2.6

2.0

1.0

0.0
Adama

Ambo

Bishoftu

Debre Birhan

Dessie

Yekatit 12

Hospitals
Despite the potential limitations, this study provides
some basic information about the status of pediatric
hospital care that could guide the national efforts in
improving the quality of referral care for children.

CONCLUSIONS AND RECOMMENDATIONS

This assessment has the following limitations. First,
sampling of the hospitals was by convenience;
majority of them being zonal hospitals the results may
not fully reflect the situation in the more peripheral
district hospitals. Second, the study was largely
observational and assessments were in part based on
the judgments of the observers. Besides, the quality
of available hospital statistics was inadequate to
disaggregate patient load at OPD, emergency and
admission levels by appropriate age categories (0-28
days, 1-12 months, 1-5 years and > 5 years). The
statistics problem definitely under-estimates the
reported hospital mortality rate. Similarly, it was
impossible to determine the mortality rate within the
first 24 hours of admission which is a very sensitive
indicator of the quality of emergency care in any
hospital setting.

The overall case fatality rate of 11% (10-16%) in this

survey was comparable to that reported from Zambia
(12-15.8%) and Kenya (4-15%) but the possibility of
under-reporting is there.
All of the surveyed hospitals were not appropriately
organized and fully equipped to handle pediatric
emergencies effectively. None of these hospitals were
practicing the standard triaging process by assessing
children immediately on arrival for emergency and
priority signs & the overall management of emergency
cases was not up to the expected standards of ETAT.
In the majority of the surveyed hospitals (5/8), there
were no separate pediatric waiting areas which are

38

39
essential for the proper implementation of ETAT.
Even though the current quality improvement process
through the hospital BPR initiative is a very good
opportunity for the overall improvement of hospital
care for children, the centralized triaging mechanism
that keeps patients of all age groups together in one
waiting area is not conducive for the effective
implementation of the ETAT standards.
Generally, there was lack of some essential drugs
and materials such as nasal prongs, infant and child
size bag & masks, nebulizers, heaters and oxygen
concentrators. The total absence of long acting
parenteral
anti-convulsants
poses
serious
shortcoming in the management of pediatric
neurologic emergencies.
In almost all hospitals, there was no clearly
designated and properly arranged/equipped high
dependency area where very sick children receive
highest attention. Majority of the hospitals (5/8) do not
have any special arrangement and facilities for
providing appropriate neonatal care.
Hygienic facilities were below the expected standard
in majority of the hospitals and none of the hospitals
provide appropriate routine pediatric diet for children.
Overall, the case management of common neonatal
and childhood illnesses was not optimal. None of the
hospital staff had been trained in ETAT and there
were no job aids or protocols for pediatric referral
care which could partly contribute to the problems
observed in the case management.
In summary, the quality of pediatric referral care
needs serious attention and coordinated systematic
improvement effort using the opportunity of the
national hospital management initiative and the BPR
process to institutionalize ETAT and standards of
hospital care for children. This has to be
complemented with availing of appropriate job aids,
essential supplies and equipment, and improvement
of health worker skills through training, clinical
mentoring and regular supportive supervision.

39

40

RECOMMENDATIONS
1. The ETAT standards should be incorporated into the Ethiopian Hospital Management Initiative blue print and the
BPR documents by the FMOH to ensure its systematic implementation.
2. National plan of action should be developed for systematic and phased implementation of ETAT in referral
hospitals.
3. Staff should be trained in ETAT through in-service courses & ETAT should also be introduced into the pre-service
teaching to ensure its sustainability.
4. National pediatric referral care protocol & job aids should be availed in all hospitals
5. Coordinated efforts and more resources needed for availing essential drugs, supplies and equipments in all
hospitals.
6. All hospitals should have arrangements/rooms for the care of sick neonates, high-dependency areas for critically
sick children and isolation rooms for infectious cases.
7. Efforts should be made to improve the hygienic facilities & the quality of routine hospital diet for children.
8. The recording and reporting system in the hospitals need to be improved and pediatric data should be
disaggregated by appropriate age categories.
9. Mechanisms for regular supportive supervision and mentoring needed to achieve sustained improvement in the
quality of pediatric referral care.
10. The Ethiopian Pediatric Society, the Medical teaching institutions, WHO, UNICEF and other partners should
support the quality improvement process.

40

41

ACKNOWLEDGEMENTS
We would like to thank the FMOH especially the former Family Health and Health Services Departments, the
Directors and staff of the surveyed hospitals for facilitating the whole process. We are very grateful to the World
Health Organization for the financial and technical support provided for the assessment. Sincere gratitude is
expressed to Dr Neghist Tesfaye (FMOH), Dr Teshome Desta (WHO/ICST-ESA), Dr Wilson Were (WHO/HQ)
and Prof. Elizabeth Molyneux (Queen Elizabeth Central Hospital, College of Medicine, Blantyre/Malawi) for their
valuable technical support.

41

42
Management of severe acute malnutrition in children using community based therapeutic care approach:
a review of three years data from southern Ethiopia.
Efrem Teferi, MD1, Shiferaw Teklemariam, MD, MPH1 , Lopiso Erosie, BSC, MPH1 , Abel Hailu, MD1
Tefera Belachew, MD, MSc, DLSHTM2, Mohammed A Yassin, MD, MSc, PhD1,3

Abstract
The objective of the study is to assess the outcome of community-based therapeutic care (CTC) for children with severe
malnutrition in Southern Ethiopia. Diagnosis of severe malnutrition was made based on anthropometric measurement and
all children received therapeutic food according to the protocol and were discharged from the feeding program when they
their weight for height was more than 80% of the reference for 2 consecutive weeks. Data on the number of admissions,
discharges, weight gain and length of stay in the program were recorded using standard formants and reports were sent to
the Regional Health Bureau monthly. The data was entered using EPI-Info proportions and means were compared using
chi-square test. This is a retrospective review of reports retained in the Bureau. A total of 12,316 patients with severe acute
malnutrition, 56.2% marasmic and 43.8% kwashiorkor cases were treated in CTC program from 2003 to 2005. The average
cure and death rates were 91% (9871) and 2.5% (217) and the average weight gain was 5.3 and 5.8 grams /kg/day and the
average length of stay was 49 and 42 days for cases of Marasmus and Kwashiorkor, respectively. Except for weight gain
and length of stay, our findings exceeded the minimum sphere standards for treatment outcome measures. In conclusion
the CTC approach has a comparable outcome to Therapeutic feeding centers and could be expanded quickly during
emergency situation. As majority of patients are treated at home, the workload for the health worker would be reduced, so it
is an alternative approach for management of severe malnutrition where human resource and space in health facilities are
limited.

42

43

special Woredas (districts) with a total of 104 Woredas.

There are 56 nationalities in the region , which gives it
the unique feature in encompassing rich and diverse
culture. According to the 2005 Regional Health Bureau
annual report, there were 16 hospitals, 160 health
centers, 1336 health posts with potential health service
coverage of 50%. The regional ; DPT3 coverage was
90%, family planning 47%, antenatal care 60% and
deliveries attended by health professionals of 18%. The
major causes of mortality among children under five
years of age are pneumonia, diarrhea and malaria.

Introduction
Ethiopia has a long history of food insecurity and
nutritional problems affecting a large proportion of the
population. (1) Even during a relatively non-drought
years, malnutrition in children in Ethiopia is extremely
high exposing the survival of this group of the population
at a great threat. An estimated 47% of Ethiopias underfive children are moderately or severely stunted
contributing to an under five-mortality rate of 123/1000
live births. (2) Ethiopia stands 6th among countries with
the highest number of under-five deaths in the world, with
more than 472,000 under-five dying each year, (3)
Malnutrition, even in its milder form, accounts directly or
indirectly for 53% of all under-five deaths in Ethiopia (4).

Malnutrition has become a common feature occurring in

the region in recent years with large-scale famines
reported in 2003 and cases continuing to be reported.
Dependency on rainfall, poor resource management,
uncontrolled population growth, poor farming planning
and limited food reserves of households complemented
with failure of rains in 2002 exacerbated the living
condition of vulnerable communities leading to
exhaustion of coping mechanisms.

The Southern Nations, Nationalities and Peoples

Regional State (SNNPR) is located in southwestern part
of the country and has a population of about 14 million.
The region is divided into 14 administrative Zones and 8

In response to the recent famine, many international

humanitarian agencies and the government opened
Therapeutic Feeding Centers (TFC) to mitigate the crisis.
A Decentralized therapeutic program was initiated with
Concern, Save the Children US in collaboration with
other international organizations. In 2004, the Ministry of
health in cooperation with UNICEF initiated enhanced
outreach strategy (EOS) for child survival. This involves
screening for malnutrition, Vitamin A supplementation,
strengthening immunization and health education for
child survival and was started in 54 woredas of the region
(17) and expanded to the entire region in 2006.

patients receive formula 75 (F75) and formula 100 (F100)

milk with routine drugs (18). A high cure rate (87.2%) and
a relatively low death rate (3.6%) was reported from 25
TFCs opened in the region in 2003-05 with an average
length of stay of patients of 21-25 days (8). This high
intensive care phase of treatment is very important for
patients with complicated malnutrition associated with
anorexia, septicemia, hypothermia, hypoglycemia and
severe dehydration. However, management of children
with malnutrition in TFC requires skilled staff, room for
inpatient care, materials including bedding, cooking
utensils and caretakers food (8-11). The centers were
opened and ran by non-governmental organizations
(NGO) and coordinated by the regional health bureau.
When the activities were integrated into the routine
health services and handed over to health facilities, it
was difficult to continue with the limited work force, space
and resource in the facilities.

Improvement in child survival is strongly associated with

decrease in malnutrition in countries characterized by
high rates of general malnutrition such as in Ethiopia (57). To reduce mortality due to severe acute malnutrition,
TFCs were established in SNNPR in 2003, which
provided a high quality individual inpatient care and
-4The traditional model of inpatient treatment of severe
acute malnutrition (SAM) does not consider the social
aspects of management of malnutrition and hence has
high opportunistic costs to mothers/care givers (9).
Mothers/care givers had to stay in centers for several
weeks leaving their other children and family members at
home and hindering them from their daily activities.
Community therapeutic care (CTC) is a nutritional

intervention designed with the capacity to address SAM

in both emergency and development contexts (9). Its
underlying aims are to maximize coverage and access.
In practice, this means giving priority to provision of care
for acutely malnourished over inpatient care for a few
extreme cases.

43

44
CTC was introduced in SNNPR in 2003 and expanded in
2004 and 2005. It integrates supplementary and
therapeutic feeding with an emphasis on outreach and
community based support. In out patient therapeutic
program (OTP), the therapeutic product used is ready to
use (plumpy nut and BP 100 biscuit) with outpatient drug
treatment protocol. There were few referrals for inpatient
treatment, which includes those with complication and
who were admitted in stabilization centers (SC) and

treated for few days and referred to OTP like in phase

one TFC (13). During emergency, CTC approach can
quickly provide good coverage and treatment for a large
number of severely malnourished people. This paper
reviews reports of CTC to assess the outcome of
malnourished children treated in community therapeutic
care established in the southern Ethiopia and compared
with the results of children treated in TFC.

Materials and methods

Children with severe acute malnutrition were admitted to
therapeutic care established in response to the famine
encountered in the region.
Diagnosis of severe
malnutrition was made based on anthropometric
measurement and brief examination for bilateral pitting
pedal edema. At admission, patients were assessed for
hydration, anemia and signs of infection. Patients were
given oral doses of Vitamin A, Folic acid, Amoxicillin (5day course), Mebendazol, treated for dehydration with
Resomal and given ready to use therapeutic food (RUTF)
according to the protocol (18). Patients were discharged
from the outpatient therapeutic feeding program when
their weight for height was more than 80% of the
reference for 2 consecutive weeks and if they did not
have signs of infection. They were followed in
supplementary feeding program until they reach 85% of
the reference for 2 consecutive weeks. At each follow up
visit, weight, extent of pitting edema, presence of
infection and treatment were recorded. In CTC, plumpy

nut was used instead of F100. The main difference

between F100 and plumpy nut is that part of dried
skimmed milk in the F100 was replaced with peanut
butter (with a 25% total weight). Plumpy nut has an
energy density 5 times more than that of F100. Plumpy
nut was used in all the centers except for few weeks
when BP100 was used due to shortage of plumpy
nut.Data on the number of monthly admissions,
discharges, average weight gain and length of stay in the
program were recorded in each therapeutic centre using
standard formants and reports were compiled by the
health facilities and NGOs supporting the programs and
were sent to the Regional Health Bureau (RHB) monthly.
The data was entered using EPI- Info programme(CDC
Atlanta) and analyzed using descriptive statistics,
proportions and means were compared using chisquared test and p values value <0.05 was considered
as significant.

This is a retrospective review of reports retained in the

RHB. All reports from CTC sent to RHB were included in
this paper, except reports of 243 cases from Kembata
Tembaro zone and 44 admissions from Sidama Zone

because of incompleteness. Patients who were

defaulters, referrals and non-respondents were excluded
as their outcome was not known. Ethical approval was
obtained from the SNNPR Health Bureau.

Results
A total of 12,316 patients with severe acute malnutrition,
56.2% marasmic and 43.8% kwashiorkor cases were
treated in CTC program from 2003 to 2005. Of these,
1540 (12.5%) were treated in 2003, 1955 (16%) in 2004
and 8791(71%) in 2005. The majority (90%) of the cases
were age between 6 months to 5 years old. The average
cure and death rates for the region were 91% (9871) and
2.5% (217) respectively. The highest (99%) cure rate

was recorded in Boricha and Damot Gale districts

(Woredas), while the lowest cure rate (80.6%) was
observed in Malgano in Sidama zone. Death rate above
minimum standard was not reported from any of the
CTC, the highest death rate (8.4%) was reported from
Bedessa and the lowest (0.4%) in Arbegona as shown in
Table 1.
number of cases discharged higher than the sum of
deaths and cured cases.
The average weight gain for the region was 5.3 and 5.8
grams /kg/day for cases with Marasmus and
Kwashiorkor, respectively. The average length of stay
was 49 and 42 days for marasmic and Kwashiorkor
cases, respectively (p > 0.05 for both) as shown in Table

The number of admissions was not the same as the

number of cases discharged, as some of the cases were
on follow up when the report was compiled. There were
also cases that defaulted from the program, referred to
other places and non respondents making the total

44

45
2. The outcome of patients treated in CTC was
comparable to those treated in TFC in the region during
However, there was statistically significant difference in
the mean length of stay which were 21 and 25 vs. 42 and
49 days for marasmus and kwashiorkor patients
respectively and the average weight gain of 13.4 and14

the same period with cure rate of 91% vs. 87% (p>0.05)
and death rates of 2.5% vs. 3.6% (p>0.05), respectively.
vs. 5.3 and 5.8 g/kg/day for patients treated in TFC than
in CTC (p<0.01 for both) (Table 3). A similar trend was
observed when the outcome from the CTC program was
compared to the sphere standard as shown in Table 4.

Discussion
Given the spatial arrangement of health service
units and their limited capacity to handle a large
number
of
cases
of severe acute malnutrition during emergency and
crisis in Ethiopia, seeking an alternative solution for
management of such cases cannot be overlooked.
Community base therapeutic care provides a
promising alternative option to the TFCs and facilitybased stabilization centers (9-12). Though CTC
cannot totally replace an inpatient therapeutic care
as some cases with complications, such as
infections, may still need an inpatient care, it is
complementary to therapeutic and supplementary
feeding programs (9).

Ethiopia (Amhara, Oromia) and other African

countries (Sudan, Malawi, and Niger) (13,17). The
average length of stay was also shorter than results
in other parts of Ethiopia, which reported 36-91 days
with a similar average weight gain from studies in
other parts of Ethiopia (3-6.5 g/kg/d) but less than
those reported from other parts of Africa (4.210g/kg/day). (13, 17)
The cure and death rates of patients treated in CTC
was comparable with those treated in TFC in
SNNPR during a similar period. However, the
duration of stay was shorter and the mean weight
gain was much better for patients treated in TFC
than those treated in CTC program. This could be
due to the unavoidable sharing of RUTF with
siblings and even adults where the bulk of the
treatment period in CTC approach is based at home.
The fact that most of the data from TFC were
collected during emergency and data from the CTC
were collected during non-emergency situation
makes comparison difficult. However, according to
reports (data not shown) from Boricha district from
where data from TFC and CTC programs were
collected during emergency situations in 2003 and
2005, the treatment outcome was similar confirming
that CTC program was effective even during
emergency situation.

The experience in implementing therapeutic feeding

programs in the region in the last few years enabled
us to understand and expand CTC very quickly. The
total number of cases treated in the region (12,316)
was sizable enough to assess the outcome of the
therapeutic feeding programs. Overall, the outcome
of patients treated in the CTC approach were
promising considering the minimum sphere
standards for the outcome indictors of therapeutic
feeding programs (9-11,15,). In addition, CTC
handles the majority of cases by creating access
and capacity which is very difficult to meet during
emergency situation (9-11). The outcome of
children treated by the CTC approach in the SNNPR
exceeded the minimum standard for both cure and
death rates although the average weight gain was
low and the average length of stay was longer. The
latter may be due to sharing of RUTF with siblings
within the family as patients in CTC are treated at
home. Our results are similar to the findings in study
done in Badewacho in Hadiya in SNNPR, which
reported a cure rate of 85% and death rate of 4%,
mean weight gain of 4.8 g/kg/day with a mean
length of stay of 42 days among 170 children treated
in CTC (16). The results of our study were better
compared to findings reported from other parts of

It was observed that inpatient treatment schemes

had no additional advantage over CTC except the
higher average weight gain obtained compared to
the workload it incurs to health workers and the
difficulty to admit all patients with the limited space
and health professionals available in the health
facilities.

45

46
was no information on the follow up of individual
cases and suffers incompleteness and missing data.

The limitation of this study was the fact that it was

based on a retrospective record review and there
In conclusion CTC as a new strategy for management of
severe malnutrition was successfully implemented in
SNNPR. It has a high coverage, low cost and could be
expanded very quickly in emergency situations. The
majority of patients can be treated in the outpatient
program without disrupting caregivers from their daily
activities and only very few patients who had infections
and very severe malnutrition required inpatient care for
few days. The product used in CTC was ready to use
therapeutic food (plumpy nut, BP 100 biscuit) and could
be implemented as an outreach and by Health Extension
Workers.

CTC is an alternative approach for management of

severe malnutrition where shortage of health
professionals and limited space for admission are the
major hurdles. However, problems including shortage of
health professionals, drugs, therapeutic products,
transportation, lack of space to admit complicated cases,
misunderstanding among some who consider CTC as an
NGO business, lack of training and high turnover of
health workers should be anticipated during integration of
CTC into the routine health system. As CTC is a
community based approach, deployment of health
extension workers in the region in the recent years would
be a good opportunity for outreach activities including
follow up and referral of cases in the villages.
For the CTC to be taken, scaled up and to be part of the
routine service, continuous in-service training of health
professionals, inclusion of CTC to the pre-service
training, facilitation of transportation of materials required
for CTC, local production of therapeutic food and
integration of CTC into the routine health service is
recommended. The program should be supported with
proper nutrition behavioral change communication on
infant and young children feeding so that to prevent
malnutrition and its recurrence.

46

47

-11Table 1: Outcome of cases of severe acute malnutrition treated in the CTC program from 20032005, South Ethiopia

Woredas

Number of cases Treatment Outcomes

Admitted
Cured
Death
N
Percent N
Percent

Total Discharged

Dalocha
Lanfuro
Malgano
Boricha
Arbegona
Bensa
Bedessa
Shebedino
Awssa Zu.
Boloso Sore
Offa
Damot Gale
Humbo
Sodo Zurria
Modulla
Karat
Total

762
907
1159
1255
248
213
399
575
748
2593
801
578
159
278
313
1328
12316

680
901
1011
1152
230
68
237
518
535
2586
741
521
149
207
163
1132
10831

Zone

Silte
Sidama

Wolaita

Kembata T.
Konso

631
795
815
1140
215
61
206
495
479
2408
641
514
142
185
141
1003
9871

92.79
88.24
80.61
98.96
93.48
89.71
86.92
95.56
89.53
93.12
86.50
98.66
95.30
89.37
86.50
88.60
90.9

19
21
14
16
1
2
20
7
11
43
20
4
5
9
6
19
217

2.8
2.3
1.4
1.4
0.4
2.9
8.4
1.4
2.1
1.7
2.7
0.8
3.4
4.3
3.7
1.7
2.5

CTC = Community based Therapeutic care, N = Number

47

48

-12Table 2: Mean length of stay and average weight gain for cases of severe acute malnutrition
admitted to the Community based Therapeutic care program from 2003 to 2005, South Ethiopia.

Centers

Average length of stay in

Average weight gain in gram/kg/day

days
Marsmus

Kwashiorkor

Marasmus

Kwashiorkor

Dalocha

63

47

5.2

4.0

Lanfuro

61

40

5.6

5.1

Malgano

37

32

7.8

10.0

Shebedino

40

32

4.6

5.0

Awassa Zu.

59

47

3.0

2.8

Arbegona

58

50

5.2

4.4

Boricha

45

32

6.0

5.5

Bedessa

70

66

3.0

18.0

Offa

62

52

3.9

3.1

Sodo Zur.

40

42

7.5

4.8

Boloso So.

58

52

5.0

4.2

Damot Ga.

36

33

5.4

4.4

Humbo

41

30

5.5

3.8

Modulla

32

30

5.3

6.3

Karat

39

37

6.0

5.6

Regional average

49.4

41.5

5.3

5.8

48

49

-13-

Table 3: Comparison of outcome of patients treated in TFC and CTC

Program

Cure
rate

Death
rate

Average length of stay Average

(day)
(g/kg/d)

weight

gain

Marasmus

Kwashiorkor

Marasmus

Kwashiorkor

TFC

87.2

3.6

25

21

14

13.4

CTC

91

2.5

49

42

5.3

5.8

in the southern region treated during similar period (2003-2005)

CTC = Community based Therapeutic care, TFC = Therapeutic Feeding Centre
-14-

Table 4: Outcome of children treated by the Community based Therapeutic care e program from
2003- 2005 compared to the minimum sphere standard, South Ethiopia.
Indicator
Cure rate
Death rate
Mean weight gain
Mean duration of stay

Sphere standard* Our Study findings

>75%

90.9%

<10%

2.5%

>8grams/kg/day

5.34-5.8 gram/ kg/ day

30-40 days

42-49 days

49

50
*Guideline for management of severe acute malnutrition,FMOH,2007, Referrence.22

References
1. Birhane G. Running a national early warning system: the Ethiopian experience, Addis Ababa Relief
and Rehabilitation commission, 1991.
2. CSA & ORC Macro (Central Statistical Authority (Ethiopia) and OCR Macro), 2006, Ethiopia
Demographic and Health Survey 2005. Addis Ababa, Ethiopia and Calverton, Maryland, U.S.A:
Central Statistical Authority and OCR Macro.
3. Robert E Black, Saul S Morris, Jennifer Bryce. Where and why are 10 million children dying every
year? The Lancet 2003;28: 361,
4. BASICS II. Basic support for institutionalizing child survival. The Second Child Survival Revolution,
Summary of the Lancet Child Survival Series: BASICS II, 2003.
5.

Pelletier, D., & E. Frongillo. 2002. Changes in Child Survival are Strongly Associated with
Changes in Malnutrition in Developing Countries. FANTA, Academy for Educational Development:
Washington DC, USA.

6. Federal Ministry of Health, Family Health Department. National strategy for child survival in
Ethiopia, Addis Ababa, July 2005.
7. WHO. Reducing severe and moderate malnutrition in Children. Bull WHO 1995, 73(4): 443-48.
8. Teferi E et al. Treatment Outcome of children admitted to Therapeutic Feeding centers in Southern
Region. (in press).
9. De Waal, A. Taffesse, L. Carruth . Child survival during the 20022003 droughts in Ethiopia,
Special Issue:

Humanitarian Crises: The Emergency Rooms of Global Health, Taylor &

Francis,June 2006, 1(2).

10. Collins S, Dent N, Binns P etal. Management of Sever acute Malnutrition in children: Review,
online www.thelancet.com September 25, 2006.

50

51
11. Collins, S. Changing the Way we Address Severe Malnutrition during Famine. The Lancet 2001,
358:498-501.
12. Collins, S. and Sadler, K. The Outpatient care for severely malnourished children in emergency
relief programs: a retrospective cohort study. The Lancet 2002, 360:1824-30.
13. Community based therapeutic care (CTC) in Ethiopia. Proceeding of workshop in Addis Ababa,
Ethiopia, 22- 23 June 2004.
14. Federal MoH/UNICEF/MOST. Guideline for enhanced outreach strategy (EOS) for child survival
interventions, revised version, Addis Ababa. July 2005.
15. EL HadjiIssakhaDiop, Nicolle Idohou, Marieb am Adeline Ndour, Andre Brined and Salimata Wade.
Comparison of
the efficacy of ready to use food and liquid milk Based diet for rehabilitation of severely
malnourished children. Am Clin Nutr 2003; 78: ; 302-7.
16. Steve Collins, Kate Sadler, Outpatient care for severely malnourished children in emergency relief
programmes; a retrospective Cohort study. The Lancet 2002;360;1824-30.
17. Community based approach to managing severe malnutrition, Proceedings of an interagency
workshop, Dublin October 2003.
18. Federal Ministry of Health . National guideline for the management of severe acute malnutrition for
Ethiopia. MoH, Addis Ababa, May 2004.

51

52

Acknowledgments
We would like to thank members of Laboratory and Research Department for their valuable comments on
the draft. We also thank save the children USA, International Medical Corpus, Action Contra La Faim for
helping to implement the program and health workers in zones and woreda for working hard to save the
lives of patients. I t would have been difficult to implement the program with out the help of UNICEF (United
Nations Childrens Fund), who provided therapeutic products drugs, and conducted trainings in cooperation
with Regional Health Bureau.

52

53
CHILD SURVIVAL: PROGRES TOWARDS MEETING MDG4
Assaye Kassie1

Abstract
Few causes are responsible for the majority of under-five deaths in Ethiopia: pneumonia (28%), neonatal causes
(25%), malaria (20%), diarrhea (20%), measles (4%) and AIDS (1%). To prevent these deaths, and to achieve
Millennium Development Goal 4 (MDG4) (to reduce by two thirds, between 1990 and 2015, the under 5 mortality
rate), it is necessary to ensure the implementation of cost-effective interventions that are listed in the National Child
Survival Strategy.
There are examples of remarkable achievements in coverage increase within short time periods, including training of
30,000 Health Extension Workers (HEWS) in the last 4 years, rapid increase, from 1% in 2005 to 42% in 2007, in
percentage of children under the age of five years who slept under a Long Lasting Insecticide-treated Nets (LLINs),
and increase in coverage of Vitamin A supplementation from 45 % in 2005 to 91% in 2007. These successes can
serve as benchmarks to scaling up of other interventions.
Among the major killers, the ones that are poorly addressed are childhood pneumonia and perinatal problems which
are the leading causes of under-five mortality in Ethiopia. Realizing the continuum of care approach at delivery level
and sustaining it over time, and searching for an alternative way of improving access to treatment of childhood
pneumonia and essential newborn care, are crucial challenges for child survival in Ethiopia.
Furthermore, there is growing consensus that a primary bottleneck to achieving MDGs in low-income countries is
health systems that are too fragile and fragmented to deliver the volume and quality of services to those in need. Major
shortfalls are identified in the health workforce, lack of donor coordination, and week information systems. It is for this
reason that the Health Sector Development Programme (HSDP) in Ethiopia is focusing on cost-effective health
interventions and on health systems strengthening in order to achieve the dual goals of improving population health
and reducing health inequalities.

53

54

1) Introduction: an historical perspective

Following the success of the 1979 International
Year of the Child, in the early 1980s signs of
hope were emerging for childrens causes. The
evolution of Basic Service and Primary Health
Care (PHC) approaches gave the practitioners of
child health and human development a new
sense of purpose and in 1982 an initiative known
as the Child Survival Revolution (later including
child development) was launched (19). The
Child Survival Revolution was initiated to
promote Growth monitoring, Oral rehydration
therapy, Breast feeding, Immunization, the
provision of Food and Family planning. These
interventions have collectively come to be known
as GOBIFF (19).
Ethiopia was one of the first countries to
implement these high-impact child survival
interventions.
The
Ethiopian
Expanded
Programme on Immunization (EPI) was launched
in 1980 but until recently coverage of all of the
above key interventions, including EPI, remained
very low.
Regardless of the fact that so many lives could
have been saved with the implementation of such
simple, high-impact interventions, around the
mid-nineties, it was noticed that child mortality
was not yet receiving
The National Child Survival Strategy has been
instrumental in the scaling-up of child survival
interventions through the active participation of
partners, relevant sectors and the community at
large (6). This has wide implications also in
terms of poverty reduction. In fact, the focus on
child (and maternal) care provides more of a
poverty orientation than reliance on other
services, since the disease burden at an early
age or at childbirth is particularly important
among the poor. In fact, not only are death rates
higher among the poor, compared to the rich, but
the highest poor-rich mortality ratio is observed

1 UNICEF, Addis Ababa

enough attention. This was mainly because the

worlds attention had been, understandably,
focused on the growing HIV pandemic and HIV
associated opportunistic infections such as
tuberculosis and the like. While progress in
reducing under-five mortality has in many low
income countries slowed, in others it has totally
stopped declining and in some cases even
regressed significantly (17;16).
Following the Lancet Child Survival publications
in 2003 (4), the second global Child Survival
Revolution was launched. In 2004, the National
Child Survival Conference was organised in
Addis Ababa. The Federal Ministry of Health
(FMOH) and its partners all participated and in
2005 the National Child Survival Strategy was
developed (6). The overall objective of this
strategy was to reduce under five mortality by
two thirds between 1990 and 2015 to achieve the
Millennium Development Goal 4 (MDG4).
Primarily, the National Child Survival Strategy
focuses on the health sector, but important
distant determinants of child survival, like
reducing poverty, improving household food
security, raising levels of maternal education and
providing safe water and sanitation, are also
recognized in the document.
for complications of pregnancy and childhood
infectious diseases (13).
2) Causes of mortality: what are Ethiopian
Children dying from?
The Child Survival Strategy identifies the direct
causes that are responsible for under-five
mortality (Figure 1): pneumonia (28%), neonatal
causes (25%), malaria (20%), diarrhea (20%),
measles (4%) and AIDS (1%). Underlying
conditions are also identified, manly malnutrition
and HIV/AIDS (6).

54

55

Other, 2%
Measles, 4%
AIDS, 1%

Neonatal, 25%

Diarrhea, 20%

Malnutrition
57%
HIV/AI
DS

Malaria, 20%

Pneumonia,
28%

Figure 1. Causes of under-5 deaths in Ethiopia (1).

In Ethiopia, neonatal mortality contributes to

about one fourth of the overall under-five
mortality. Newborns die mainly due to infections,

Congenital
4%

Causes of
neonatal
deaths

followed
by
perinatal
asphyxia
prematurity/low birth weight (Figure 2).

Other
7%

Diarrhoea
3%
Tetanus
7%

and

Asphyxia
25%

Preterm birth
17%

Infection
37%

Infections and tetanus

account for 47% of
neonatal deaths in
Ethiopia

Figure 2. Causes of neonatal deaths in Ethiopia (1).

Fortunately, for each target condition there are

both preventive and curative interventions that
can prevent around 72% of deaths in under five
children. To prevent these deaths and to achieve

MDG4, it is necessary to ensure effective

implementation of a limited number of
interventions that are listed in the National Child
Survival Strategy. In 2003, the Lancet child

55

56
survival series estimated that, with 99% coverage
sufficient evidence for effect in prevention or
treatment, it would be possible to prevent 65% of
deaths
due to pneumonia, 55% of deaths due to
neonatal complications, 91% of deaths due to
malaria, 88% of deaths from diarrhea, 100% of
deaths from measles, and 48% of those due to
AIDS (12).

of the high impact interventions for which there is

3) Millennium Development Goal 4: where do

we stand?
According to the Ethiopia Demographic and
Health Survey carried out in 2005, there was an
improvement in under 5 mortality rates, with a
decrease from 165 to 123 per 1,000 live births
between 1990 and 2005 (3). The plan is to
decrease U5MR to 54 per 1,000 in the year 2015
to meet MDG4 (Figure 3).

Achieving the MDG4 for child survival in Ethiopia

demands focused and coordinated action to
strengthen the health systems, improve nutrition
and reduce inequities in access to effective
interventions against all the diseases which kill
under-five children (5; 20).

Under 5 Mortality Rate

160

165
153
140
123

120

109
89

80
HSDP I

95

54

II

40

0
1990

1995

Current Trend

2000

2005
Years

2010

2015

MDG Trend

Figure 3. Trend in under 5 Mortality Rate in the period 1990-2005 and projections until
2015 in Ethiopia.

56

57
4) Opportunities to achieve MDG4:
implementation of child survival interventions
An integrated approach is in place in order to
achieve MDG 4, with:
Focus on the community: Ethiopia is investing
a lot in the Health Extension Program (HEP) to
reach the poorest of the poor with basic and
essential life saving care by focusing mainly on
the mothers, newborns and children of the rural
population.
HEP
institutionalizes
and
standardizes the village health-care delivery
system to empower care-takers, families and
communities to take care of their own health. It
Child Survival Strategy: HSDP III is
incorporated as the de facto health component of
the Plan for Accelerated and Sustained
Development to End Poverty (PASDEP) (15).
Maternal and Child Health (MCH) are major
focuses of HSDP III, with Child Survival Strategy
being a major component of HSDP III.
Availability of Experienced Programs: there
are experienced programs relevant to child
health in Ethiopia, including EPI, Integrated
5) Benchmarks: scaling-up is possible
Through proper and rational utilization of
available opportunities in the country, it is
possible to scale up the implementation of high
impact child survival interventions. These are
some examples of successes and remarkable
achievements in coverage increase within short
time periods that can serve as benchmarks to

increases access and utilization of most of the

high impact preventive and curative interventions
which are listed in the National Child Survival
Strategy.
Accelerated Expansion of Primary health
Care: within the context of HSDP III, the FMOH
has initiated an accelerated expansion of PHC
services. This new initiative aims to accelerate
physical infrastructure expansion, a base for
improving access to basic health care services in
rural Ethiopia. Besides physical infrastructure
expansion, the initiative also entails an increase
in the number of health professionals mainly at
primary health care unit level.
Management of Neonatal and Childhood
Illnesses (IMNCI), Nutrition, Safe Motherhood
and Malaria control programs.
Partnership for Maternal, Newborn and Child
Health (PMNCH):
there is a growing partnership between the
Government, UN organizations, bilateral
partners, Private institutions and NonGovernmental Organizations for Child Survival.
scaling up of other interventions. These include
training of 24,600 HEWs within a period of 3
years, and rapid increase in percentage of
children under the age of five years who slept
under Long Lasting Insecticide-treated Nets
(LLINs) from 1% in 2005 (3) to 42% in 2007 (10).
Furthermore, through the Enhanced Outreach
Strategy, it has been possible to scale up biannual supplementation of Vitamin A from the
2005 baseline level of 45 % to 91% in 2007
(Figure 4).

57

58

Child Survival: Children (6-59 months) supplemented with Vitamin A

Achievements per round as of November 2007
11,926,235
11,423,171
11,567,721 11,423,171 11,926,235
10,154,375
11,567,721

12,000,000
10,000,000

8,269,753
8,000,000

10,154,375
8,269,753

6,000,000

4,946,811
4,946,811

4,000,000

2,576,620
2,000,000

1,389,438

0
2004 - 1st
ROUND

2004 - 2nd
ROUND

2005 - 1st
ROUND

2005 - 2nd
ROUND

2006 - 1st
ROUND

2006 - 2nd
ROUND

2007 - 1st
ROUND

2007 - 2nd
ROUND

Achievements: # of children covered with vitamin A

National target children in Child Survival

Figure 4. Coverage in Vitamin A supplementation among children aged 6-59 months in Ethiopia during
the period 2004-07.

To realize the continuum of care that starts from

the household and continues up to the facility
level, the Family Health Department (FHD) of the
Federal Ministry of Health (FMOH) launched the
case management IMCI training in 1996 and,
following that, community IMCI was adopted by a
national workshop in 2001.
In 2004, 36 % of the Health Centers had at least
one health worker trained in IMCI (5).
Assessment, classification and management of
early neonatal problems were incorporated into
the formal IMCI training guideline in 2006 and
since then IMCI was renamed as Integrated
Management of Newborn and Childhood
illnesses (IMNCI). According to the March 2008

IMNCI annual review meeting report of the FHD,

60% of the Health Centers have at least one
person trained in IMNCI (11). On the other side,
Community-integrated
IMNCI
(C-IMNCI)
coverage has shown a significant increase from 2
woredas in 2004 to 180 woredas in 2008 (11).
In the past five years, immunization coverage is
also showing an increasing trend. In 2003, DPT3
coverage was 50 % and in 2007 the Pentavalent
coverage reached 73% (8). Most importantly, the
introduction of Haemophilus influenzae type b
(Hib) vaccine in 2007 had brought a paramount
benefit to prevent childhood pneumonia and
meningitis in under five children (9).
6) Lessons learned and way forward: more of
the same is not enough

Recent evidence suggests that, based on current

trends, many low-income countries are unlikely to
achieve the MDG health target by 2015 (21). This
is despite the fact that there are a growing
number of cost-effective interventions, as well as
increasing international assistance for specific
disease control programmes. There is growing
consensus that a primary bottleneck to achieving
MDGs in low-income countries is health systems
that are too fragile and fragmented to deliver the

volume and quality of services to those in need

(18). Major shortfalls are identified in the health
workforce, lack of donor coordination, and week
information systems as critical challenges to
achieving MDGs. It is for this reason that HSDP
in Ethiopia is focusing on cost-effective health
interventions and on health systems
strengthening in order to achieve the dual goals
of improving population health and reducing
health inequalities.

58

59

Training a sufficient number of health

professionals and construction of an adequate
number of health facilities in a very short time are
some of the grand achievements of the FMOH.
However, the anatomy alone may not take us to
the destination until and unless it is
complimented with the physiology. The health

workers that we have trained until now must have

the necessary knowledge and skills to fulfill their
jobs, and, equally, health facilities should be
equipped and supplied with essential items.
Moreover, there should be regular and
continuous supervision to achieve the ultimate
goal of child survival.

Among the major killers, the ones that are poorly

addressed are childhood pneumonia and
perinatal problems which are the leading killers of
under-five children in Ethiopia. According to the
2005 EDHS, only 4.9% of pneumonia cases have

had access to antibiotic therapy and, from the

same source, skill delivery coverage was also
alarmingly low, with only 6% of total deliveries
being conducted by a skilled attendant (3).

The Child Survival Strategy identifies Health

Extension Program as an important vehicle that
carries most high impact child survival
interventions to the community. At this juncture,
the potential of HEWs to implementing most of
the child survival interventions is not fully
exploited. There are multiple reasons for this,
including problems related to competency,
shortage of supplies, lack of regular supportive
supervision, and lack of community ownership.
Therefore, the conclusion is that achieving MDG4
in Ethiopia is feasible, but demands addressing
the following serious concerns:
Mobilizing adequate amounts of resources
to fully implement the Health Extension
Program;
Realizing the continuum of care approach at
delivery level and sustaining it over time;
Searching for an alternative way of improving
access to treatment of childhood pneumonia
and essential newborn care, which are the two
major killers of the under five children: this may
include provision of community-based
pneumonia and neonatal infection treatment;
Strengthening Monitoring and Evaluation,
which is the life blood of the child survival
strategy to track progress towards the goal and
to ensure quality of service rendered to the
community.

59

60

References
1. Child health in Ethiopia. Background document for the National Child Survival Conference April
22-22, 2004, Addis Ababa, Ethiopia. World Health Organization, Addis Ababa.
2. [http://www.afro.who.int/cah/documents/situational_analysis/et_final_cs_situation_analysis.pdf.
Accessed June 2008].
3. CSA, 2006. Ethiopia Demographic and Health Survey 2005. Central Statistic Agency, Addis Ababa
and ORC Macro, Calverton.
4. Editorial, 2003. The worlds forgotten children. The Lancet, 361: 1.
5. FMOH, 2004. National review and planning meeting. Report of the Family Health Department of
the FMOH. Federal Ministry of Health, Addis Ababa.
6. FMOH, 2005a. National strategy for child survival. Family Health Department. Federal Ministry of
Health, Addis Ababa.
7. FMOH, 2005b. HSDP III. Health Sector Strategic Plan 2005/06-2009/10. Federal Ministry of
Health, Addis Ababa.
8. FMOH, 2007a. Health and health related indicators2006/07. Federal Ministry of Health, Addis
Ababa.
9. FMOH, 2007b. Annual performance report of HSDP III. EFY 1999 (2006/2007). Federal Ministry of
Health, Addis Ababa.
10. FMOH, 2008a. National Malaria Indicator Survey. Federal Ministry of Health, Addis Ababa.
11. FMOH, 2008b. IMNCI and C-IMNCI national review meeting. Report of the Family Health
Department of the FMOH. Federal Ministry of Health, Addis Ababa.
12. Gareth, J., Steketee, R., W., Black, R., E., Bhutta, Z. A., Morris S. S., and the Bellagio Child
Survival Study Group, 2003. How many children deaths can we prevent this year? The Lancet,
362: 65-71.
13. Gwatkin, D.R., 2000. Health inequalities and the health of the poor: What do we know? What can
we do? Bull. World Health Organ. 78, 3-17.
14. Lawn J., Kerber, K., 2006. Opportunities for Africas newborns. Practical data, policy and
programmatic support for newborn care in Africa. The partnership for maternal newborn and child
health, Cape Town.
15. MOFED, 2007. Ethiopia: Building on progress. A Plan for Accelerated and Sustained Development
to End Poverty (PASDEP). Annual Progress Report 2006/07. Ministry of Finance and Economic
Development, Addis Ababa.
16. Save The Children, 2006. New Report shows improvements in child survival in Africa , but more
than a million African babies die in the first month of life. Save The Children, US.

60

61

17. Schuftan C., 1990. The child survival revolution: a critique. Family practice, 7(4):329-332.
18. Travis, P., Bennet, S., Haynes, A., Pang, T., Bhutta, Z., Hyder, A.A., Pielemeier, N.R., Mills, A.,
Evans, T., 2004. Overcoming health-systems constraints to achieve Millennium Development
Goals. Lancet, 364: 900-906.
19. UNICEF, 1996. The state of the World Children 1996. The 1980s: Campaign for child survival.
page 1-5. United Nations Childrens Fund, New York.
20. UNICEF, 2006. Preliminary Report. Trends in Child Mortality in Eastern and Southern Africa 19902005. United Nations Childrens Fund, New York.
21. World Bank, 2003. The Millennium Development Goals for Health: Rising to the Challenges (draft).
World Bank, Washington, DC.

61

62

A CASE REPORT: Optic glioma in a child with NF1

Kalid Astrat, 1 MD

Abstract
A 10 year old female patient presented with progressive right eye proptosis ( Fig 1) and skin
rash of three years duration was seen at Tikur Anbessa Specialized Hospital, department of
pediatrics hematology /oncology unit. Physical examination showed mildly decreased visual acuity
and cafe au lait spot ( Fig 2) AND axillary freckling and Orbital CT ( Fig 3) showed right
intraorbital mass with an assessment of right optic nerve glioma she is to be started on weekly
vinblastine at a dose of 6mg/m2.

62

63
INTRODUCTION
Optic nerve glioma (also known as optic pathway
glioma) is the most common primary neoplasm of
the optic nerve. Along with reducing visual acuity
in the affected eye, the tumor sometimes
produces additional symptoms as it grows. A lowgrade form of this neoplasm, benign optic glioma,
occurs most often in pediatric patients. Another
form, aggressive glioma, is most common in
adults; it is frequently fatal, even with treatment.1
Optic-pathway glioma accounts for 1-5% of all
brain tumors in children [1]. About half of these
cases occur in children with neurofibromatosis
type 1 [2]. The diagnosis is usually rendered
before age 6 years, although there are some
reports of older ages [3,4]. The vast majority of
optic-pathway gliomas in children are pilocytic
astrocytomas [2,5]. The tumor may arise
anywhere along the optic pathway, from just
behind the globe to the lateral geniculate body
[5,6]. In patients with neurofibromatosis type 1,
the tumor is usually smaller than in sporadic
(non-neurofibromatosis type 1-associated) cases
[5,7].
The clinical presentation is variable. In patients
with neurofibromatosis type 1, 40-80% of opticpathway gliomas are asymptomatic at diagnosis,
whereas in sporadic cases, they are symptomatic
[3,5,8-10]. The most common signs are visionrelated: mainly visual loss, decreased visual
acuity, and strabismus. Other findings include
endocrine disturbances, signs of increased
intracranial pressure, and hydrocephalus
[2,5,8,9,10-13]. Ophthalmologic examination
may reveal decreased visual acuity, pathologic
visual fields, proptosis and more [9,10,13,14].
The generally young age of the children and high
prevalence of neurofibromatosis
type 1associated attention deficit hyperactivity disorder
render the ophthalmologic examination difficult,
and decrease its sensitivity and specificity [2,15].
Early diagnosis is important so that the tumor can
be carefully monitored and treatment can be
administered early, before visual deterioration.
The diagnosis can be made functionally by
visual-evoked potentials, but as is

the case for eye examinations, their efficacy is

limited, and specificity is low [2,16]. Modern
neuroimaging modalities
provide excellent
characterization of optic-pathway gliomas,
obviating the need for biopsy [17]. Magnetic
resonance imaging was found to be superior to
computed tomography for the detection and
evaluation of extensive tumor involvement. It has
a higher specificity, and can be used to assess
disease progression [18,19]. The biological
behavior of optic-pathway glioma varies. The
tumor may progress rapidly, remain stable for
years, or
even shrink spontaneously or after biopsy, mostly
with clinical improvement [2,5,7,20]. Regrowth
after a stable period or after biopsy was also
reported [2]. Tumor progression is apparently
affected by the presence of neurofibromatosis
type 1, patient age, and tumor location [21]. Less
progression was evident in patients with
neurofibromatosis type1 than in sporadic cases,
and in children who were older at diagnosis
[2,5,7,9,10,19,22]. Tumors situated at the optic
nerve or chiasma tend to grow more slowly and
less aggressively than chiasmatic/hypothalamic
gliomas, with lower mortality [23,24]. Posterior
involvement may also lead to significant
morbidity and mortality [6,14,25].The natural
1

Department of Pediatrics and Child Health, Medical faculty, Addis

Ababa University.

history
of
optic-pathway
gliomas
in
neurofibromatosis type 1 is considered
unpredictable [2,26]. Hence deciding
whether, or when, to initiate treatment becomes
difficult. The presence of an optic-pathway
glioma in a patient without neurofibromatosis
type 1 is considered an indication for treatment
[5]. Although neurofibromatosis type 1 is thought
to be relatively benign, given the risk of visual
impairment, blindness, neurologic deficits, or
death [27,28], patients affected by the tumor
should be monitored routinely for its size and
visual function, and an adverse change in either
should be considered an indication for treatment
[2]. If a glioma tends to remain stable, the
63

64
intervals between
magnetic resonance
examinations can be gradually increased [10].
Nevertheless, the subjective timing of imaging
scans and the lack of objective references to
identify deviations from normality place patients
at risk of either unnecessary or insufficient
neuroimaging. An optic-pathway glioma tends to
grow along the optic pathways by increasing its
width, rather than as one concentric mass that
grows in all directions [5]. As such, stereotypical
patterns of growth as seen on imaging scans of
children with optic-pathway glioma are often
highly comparable, because the tumor tends to
involve the same brain structure, and the
manner of growth is very similar.
CASE REPORT

A 10 year old female patient presented with

progressive right eye proptosis and skin rash
of three years duration was seen at Tikur
Anbessa Specialized Hospital, department of
pediatrics hematology /oncology unit.
Since
the last 6 months the proptosis was more
progressive to attain the current size. Family
history is positive for paternal unilateral loss
of vision unrelated to
trauma. Physical
examination showed mildly decreased visual
acquity (OD =6/9, OS =6/6) and cafe au lait
spot( Fig 2) and axillary freklings other wise
no other findings on the musculoskletal
and CNS. Orbital CT( Fig 3) showed right
intraorbital mass with an assessment of right
optic nerve glioma she was started on weekly
vinblastine at a dose of 6mg/m2.

Fig 1

64

65

Fig 2

65

66

Fig 3
Discussion

ophthalmologic
feasible.

and

Orbital

CT/MRI

if

This is one of the rarely reported case of a

10 year old Ethiopian child with type 1
neurofibromatosis and right optic nerve glioma.
In most young patients with optic glioma the
presenting symptom is painless proptosis. Optic
atrophy is common, as is reduced visual acuity,
although the latter may be a late symptom. A
large lesion may compress the optic chiasm,
causing nystagmus or other symptoms.
Hypothalamic symptoms, such as changes in
appetite or sleep, also may occur. Massive
lesions may compress the third ventricle,
resulting
in
obstructive
hydrocephalus
accompanied by headache, nausea, and
vomiting also may occur but these findings were
not found in this patient. Historically, surgery and
radiotherapy have played a primary role in
management, however, in the last 15 years,
chemotherapy has evolved into the first-line
treatment of choice. The case presented was
started on weekly Vinblastine at a dose of 6
mg/m2
for one year with regular

66

67
References
1.

Alshail E, Rutka JE, Becker LE, Hoffman HJ. Optic chiasmatic-hypothalamic glioma. Brain Pathol
1997;7:799-806.

2.

Shuper A, Horev G, Kornreich L, et al. Visual pathway glioma: An erratic tumor with therapeutic
dilemmas. Arch Dis Child 1997;76: 259-63.

3.

Thiagalingam S, Flaherty M, Billson F, North K. Neurofibromatosis type 1 and optic pathway

gliomas: Follow-up of 54 patients. Ophthalmology 2004;111:568-77.

4.

Listernick R, Ferner RE, Piersall L, Sharif S, Gutmann DH, Charrow J. Late-onset optic pathway
tumors in children with neurofibromatosis1. Neurology 2004;63:1944-6.

5.

Kornreich L, Blaser S, Schwarz M, et al. Optic pathway glioma: Correlation of imaging findings with
the presence of neurofibromatosis. AJNR 2001;22:1963-9.

6.

Liu GT, Brodsky MC, Phillips PC, et al. Optic radiation involvement in optic pathway gliomas in
neurofibromatosis. Am J Ophthalmol 2004;137:407-14.

7.

Astrup J. Natural history and clinical management of optic pathway glioma. Br J

Neurosurg 2003;17:327-35.

8.

Guillamo JS, Creange A, Kalifa C, et al. Prognostic factors of CNS tumors in neurofibromatosis 1
(NF1): A retrospective study of 104 patients. Brain 2003;126:152-60.

9.

Czyzyk E, Jozwiak S, Roszkowski M, Schwartz RA. Optic pathway gliomas in children with and
without neurofibromatosis 1. J Child Neurol 2003;18:471-8.

10.

Listernick R, Charrow J, Greenwald M, Mets M. Natural history of optic pathway tumors in children
with neurofibromatosis type 1: A longitudinal study. J Pediatr 1994;125:63-6.

11.

Cnossen MH, Stam EN, Cooiman LC, et al. Endocrinologic disorders and optic pathway gliomas in
children with neurofibromatosis type1. Pediatrics 1997;100:667-70.

12.

Shuper A, Kornreich L, Michowitz S, Schwartz M, Yaniv I,Cohen IJ. Visual pathway tumors and
hydrocephalus. Pediatr Hematol Oncol 2000;17:463-8.

13.

Khafaga Y, Hassounah M, Kandil A, et al. Optic gliomas: A retrospective analysis of 50 cases. Int J
Radiat Oncol Biol Phys 2003;56:807-12.

14.

Sigorini M, Zuccoli G, Ferrozzi F, et al. Magnetic resonance findings and ophthalmologic

abnormalities are correlated in patients with neurofibromatosis type 1(NF1). Am J Med Genet
2000;93:269-72.

67

68
15.

Wolsey DH, Larson SA, Creel D, Hoffman R. Can screening for optic nerve gliomas in patients with
neurofibromatosis type I be performed with visual-evoked potential testing? J AAPOS 2006;10:30711.

16.

North K, Cochineas C, Tang E, Fagan E. Optic gliomas in neurofibromatosis type 1: Role of visual
evoked potentials. Pediatr Neurol 1994;10:117-23.

17.
Pepin SM, Lessell S. Anterior visual pathway gliomas: The last 30 years. Semin Ophthal.
2006;21:117-24.
18.

Van Es S, North KN, McHugh K, De Silva M. MRI findings in children with neurofibromatosis type
1: A prospective study. Pediatr Radiol 1996;26:478-87.

19.

Chateil JF, Soussotte C, Pedespan JM, Brun M, Le Manh C, Diard F. MRI and clinical differences
between optic pathway tumours in children with and without neurofibromatosis. Br J Radiol
2001;74:
24-31.

20.

Parsa CF, Hoyt CS, Lesser RL, et al. Spontaneous regression of optic gliomas: Thirteen cases
documented by serial neuroimaging. Arch Ophthalmol 2001;119:516-29.

21.

Chan MY, Foong AP, Heisey DM, Harkness W, Hayward R, Michalski A. Potential prognostic
factors of relapse-free survival in childhood optic pathway glioma: A multivariate analysis. Pediatr
Neurosurg
1998;29:23-8.

22.

Grill J, Laithier V, Rodriguez D, Raquin MA, Pierre-Kahn A, Kalifa C. When do children with optic
pathway tumors need treatment? An oncological perspective in 106 patients treated in a single
center. Eur
J Pediatr 2000;159:692-6.

23.

Schroder S, Baumann-Schroder U, Hazim W, Haase W, Mautner VF. Long-term outcome of

gliomas of the visual pathway in type 1 neurofibromatosis. Klin Monatsbl Augenheilkd
1999;215:349-54.

24.

Tow SL, Chandela S, Miller NR, Avellino AM. Long-term outcome in children with gliomas of the
anterior visual pathway. Pediatr Neurol 2003;28:262-70.

25.

Balcer LJ, Liu GT, Heller G, et al. Visual loss in children with neurofibromatosis type 1 and optic
pathway gliomas: Relation to tumor location by magnetic resonance imaging. Am J Ophthalmol
2001;131:442-5.

26.

Serova NK, Lazareva LA, Gorelychev SK, Ozerova VI, Pronin IN. A follow-up of patients with
anterior optic tract glioma concurrent with type 1 neurofibromatosis. Vestn Oftalmol 2006;122:3942.

68

69
27.

Luh GY, Bird CR. Imaging of brain tumors in the pediatric population. Neuroimag Clin North Am
1999;9:691-716.

28.

Kosa E, Csakvary V. Neurofibromatosis type 1 in childrenWith special consideration of

ophthalmologic symptoms. Orv Hetil 2004;145:473-8.

Acknowledgments
I would like to thank Dr David N Korones pediatric oncologist at Rocester medical center , New
York and Dr Ibrhaim Qaddumi , oncologist at St Jude Childrens Research hospital for their
constructive ideas and providing me with important references.

69

70

Clinical Predictors of Pneumonia Among Under-five Children

At Tikur Anbesa Specilaized Hospital
Kalid Asrat1, MD, Amha Mekasha1, MD, MSc

Abstract
The aim of the study is to identify simple clinical signs and symptoms in under five children which are
predictors of pneumonia at Tikur Anbesa Specialized Hospital (TASH), Department of Pediatrics and
Child Health, emergency and regular out-patient units. The design of the study is a prospective crosssectional study carried out during Aug 2004 Sep 2005. All children between the age of 2-59 months who
attended the regular and emergency pediatrics units at TASH during the study period who had either cough
or difficulty of breathing or chest x-ray evidence of pneumonia were included in the study. A calculated
sample of 164 was taken. Data analysis was done using SPSS and EPINFO version 1.1.2. Chi-square test
was used to calculate the differences in distribution of clinical signs and symptoms between groups with
and without chest x-ray (CXR) evidence of pneumonia.
A total of 179 patients were studied of whom 102 were males and 77 females (M:F 1:0.75). Clinical
symptoms and signs were related to CXR pneumonia. Tachypnea (94.5% with CXR pneumonia Vs 59.3%
without CXR pneumonia) and retraction (86.3% CXR Vs 40.6% without CXR pneumonia) were the best
clinical predictors of pneumonia. Tachypnea, flaring of alae nasi and retraction were also independently
associated with CXR pneumonia. Vomiting, refusal to feed, history of rapid breathing , grunting and chest
findings did not predict pneumonia.
In conclusion pneumonia is a significant cause of morbidity and mortality in developing countries like
Ethiopia . Using
simple clinical indicators pneumonia can be diagnosed by primary health workers and mortality can be
reduced
significantly.

70

71

INTRODUCTION
Around 10.6 million children die every year before reaching
their 5th birth day. Almost all of these deaths occur in low
income and middle income countries mainly in Africa and
south east Asia. Most deaths among under fives are still
attributable to just a handful of conditions, acute respiratory
infection(ARI) mostly pneumonia accounts for 19% of all
deaths(1,2,3,). Age Specific mortality
from lower
respiratory infection( LRI) in young Gambian children has
been estimated at 10 per 1000 each year(4 ) . In Ethiopia
infant mortality rate is 77 per 1000, under five mortality is
120 per 1000 (5), ARI being one of the major cause of
morbidity and mortality.
Taking these into consideration WHO in 1981 initiated a
programme for the control of ARI on a case management
of pneumonia. One of the strategies is to improve case
detection and patient management by primary health care
workers (7) and so reduce mortality. The aim of this study
is to evaluate the usefulness of simple clinical symptoms
and signs in the diagnosis of LRI, mainly pneumonia, which
can easily be used by primary health care workers. Despite
LRI (mainly pneumonia) being a condition commonly
encountered by clinicians, uncertainty remains over the

diagnosis, investigation and treatment of the condition.

Infants and children may present with a number of different
clinical symptoms and signs such as fever, cough and
tachypnea. Minority of children may present with fever of
unknown origin ( FUO) and may have no respiratory
symptoms or signs.
WHO has developed algorithm (8) to aid medical and nonmedical health care workers in diagnosis of LRI with out
radiological confirmation. The WHO algorithm stresses the
importance of tachypnea which has a 74% sensitivity and
67% specificity for radiologically defined pneumonia (8).
However, in children who had diseases for less than three
days (9), tachypnea had a lower sensitivity and specificity
of illness. Clinicians must be aware that the absence of
tachypnea does not necessarily mean the absence of
pneumonia (10 ). Grunting and nasal flaring increase the
chance of pneumonia, but their absence cannot be relied
upon to rule out the chance of pneumonia (9). Other signs
that relate to severity of pneumonia are chest indrawing,
nasal flaring and cyanosis. High fever in young children
(age up to 3 years) was found to be a sign of pneumonia
(11,12).The British thoracic

as significant amount of alveolar type of consolidation. So

does a normal x-ray rule out pneumonia? There is
anecdotal evidence for having pneumonia with a normal
CXR. Fever and tachypnea may present before CXR
changes are seen.

Department of Pediatrics and Child Health, Medical faculty, Addis

Ababa University.

society ( BTS) guidelines suggested that in children less

than 3 years combination of fever > 38.5c,chest indrawing
and RR >50/min indicate pneumonia, breathing difficulty is
more reliable sign in older children (18).
Chest x-ray (CXR) is still considered to be the gold
standard for diagnosing pneumonia in the developed world.
However, there is poor concordance between radiological
changes which constitute pneumonia( inter and intra
observer variation), consolidation on CXR was most
commonly identified by the radiologist was generally
agreed to represent pneumonic changes(12) .
WHO has recognized the difficulties with CXR
interpretation and developed a tool to standardize the
reporting of CXR use in epidemiological studies of
pneumonia (8) this system classifies CXR as: normal
appearance, infiltrates and end stage consolidation defined
Study population: All children between the age of 2-59
months who attended the regular as well as the
emergency pediatrics OPD at TASH during the study
period(AUG 2004- SEP 2005) who have either cough or
difficulty of breathing or CXR evidence of pneumonia
were included. A calculated sample of 164 was taken.
Data analysis: Data was entered and analysis was done
using SPSS and EPINFO version 1.1.2 software. ChiResults

Subjects and methods

Study setting: The study was conducted at Tikur Anbessa
Specialized Hospital (TASH), department of pediatrics and
child health emergency and regular OPD which is one of
the very few centers in the country which gives pediatric
out patient and in patient services.
Study design: a cross sectional survey was under-taken
among children under five children in the period of (AUG
2004- SEP 2005) who have either cough or difficulty of
breathing or CXR evidence of pneumonia
square test was used to calculate the differences in
distribution of clinical signs and symptoms between groups
with and without CXR pneumonia. Signs and symptoms
which were significant on univariate analysis were taken for
multiple logistic regression analysis.
Ethical consideration: Permission to undertake the study
was obtained from the department of pediatrics and child
health of TASH.

71

72
A total of 179 children among whom 102 (57%) males and
77 (43%) females who fulfilled the inclusion criterias were

included in this study. The age distribution is shown in

table 1 where half of the patients were infants.

21 patients have severe PEM,13 patients have rickets,3

have sero-proven pediatrics HIV infection and one patient
each have gastroenteritis, malaria, downs syndrome and
meningitis (table 3).
Table 2 shows clinical signs that best correlate with CXR
pneumonia, these are tachypnea, flaring and retraction ,the
other clinical signs and symptoms like vomiting, refusal to
feed, cough, fever and grunting and chest findings did not
correlate with CXR pneumonia. Table 4 shows the

sensitivity, specificity, 95% confidence intervals and

prevalence of selected symptoms and signs and table 5
shows the frequency of clinical signs and symptoms.
In children between the age of 2-59 months, tachypnea
(94.5% with CXR pneumonia Vs 59.3% without) and
retraction (86.3% Vs 40.6%) were the best predictors of
CXR pneumonia. Tachypnea, flaring and retraction were
independently associated with CXR pneumonia.

72

73

73

74

74

75

Discussion
In rural areas of developing countries the case fatality rate
from LRI in children is most likely to be reduced if primary
health care workers can identify the most serious forms of
LRI and deal with them appropriately, accurate guidelines
to detect LRI(pneumonia) who can be safely treated with
antibiotics as outpatients from those who require
immediate referral must be based on symptoms and signs
that can be readily assessed.
This study has attempted to look into clinical signs and
symptoms predictors of pneumonia in less than 5 year
children such as vomiting rapid breathing, tachypnea and
chest retraction, age and sex distribution ,frequency of
each clinical signs and symptoms at emergency and
regular pediatric out-patient department of TASH over one
year period. On a study done by Campbell, et al showed
that in infants a fever of >38.5c, refusal to feed (breast
feeding) or the presence of vomiting best correlated with
CXR pneumonia. In children aged 1-4 year, a fever of >
38.5c, RR >60/min are best correlated with CXR
pneumonia(11).Another study done by Cherian et al
showed that RR >50/min in infants and RR >40/min in
The Bangladesh study suggested that chest indrawing was
more specific as a sign
of
severe pneumonia
(17).Tachypnea has been exhaustively demonstrated to be
an excellent predictor of radiologically defined pneumonia
in a study done by Levantine J et al (10,14). In the present
study tachypnea and retraction are very specific signs of
CXR pneumonia, this is also consistent with other studies
done on the subject(10,14,17,18) and their usefulness is
further increased by their high sensitivity and are also
easier to teach to non medical staff like primary health care
workers. Vomiting, rapid breathing, fever, cough and
grunting were not satisfactory predictors of CXR
pneumonia.
Since our study is a hospital based study it should be
interpreted with caution when community health workers
(CHW) training and national guidelines are setup. Such
programmes must adopt policies that take into consideration
the health resource s available to implement them. In rural
areas of many developing countries including Ethiopia the
referral systems are poorly developed and much of the
primary care must be developed to CHWs, this should come
from community based studies.
In conclusion pneumonia is still the significant cause of
morbidity and mortality in developing countries like
Ethiopia . If pneumonia is diagnosed earlier using simple
clinical parameters which can also be used by non medical
staff like community health workers and other health
professional at each level mortality could be reduced
significantly. Tachypnea (94.5%) and retraction(86.3%) were
the best predictors of pneumonia as evidenced by

children 12-35 months of age, as well as history of rapid

breathing and the presence of chest retraction in both age
groups were found to be sensitive and specific indicators
of LRI. Increased RR and history of rapid breathing were
also sensitive in diagnosing less severe LRI that did not
necessitate admission to the wards, whereas chest
retraction was not. All these clinical signs had a lower
sensitivity in diagnosing LRI in children aged 36 months
and over (19).Study done by Shan et al showed that chest
indrawing was a reliable sign in children between the age
of 0-4 years with cough for assessment of severe LRI
(pneumonia) and among other children with cough a RR of
>50/min is a reliable basis for diagnosing LRI (14). A
multicentric study done by the WHO young infant study
group indicated that the best threshold for predicting
pneumonia in infants aged less than 2 months was RR >
60/min( 15).The best combination of sensitivity( 78-82%)
and specificity (73-89%) was achieved by wing thresholds
of 50 and 40 breaths/min for children aged 2-11 months
and 1-4 years, respectively (16).
suggestive CXR (CXR pneumonia). Further studies should
be conducted on the subject especially at the community
level.

75

76
References
1.

2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.

13.
14.
15.
16.
17.
18.
19.
2:125-8

World health report 2005

Demographic estimates of the annual number of infants and child defense in the
world.The pediatric health 1984.
Handbook of health information of India. Central Bureaus of health Intelligence, New
Delhi,1984.
Deaths in infancy and early childhood in well vaccinated rural west African population.
Ped 1994 ,91-99.
MOH. Health and health related indicators.1998.
Williams B et al estimate of world wide distribution of deaths from ARI. Lancet inf dis J
2002,2:25-32.
WHO. A programme for controlling ARI in children memorandum,WHO meeting. Bull
WHO 1994 ,62:45-58.
WHO ,management of ARI in children ,practical guidelines for outpatient care.
Geneva WHO 1995.
Plafox M etal diagnostic value of tachypnea in pneumonia defined radiologically. Arch
Dis child 2000,82:41-5
Leventhal J. clinical predictors of pneumonia as a guide to ordering chest X-ray.
Clinical pediatr 1982;21; 730-40
Campbell H et al .Assessment of clinical criteria for identification severe LRI in
children ,the lancet 1989,297-9
Campbell SM, Shann F, et al. The effect of panel membership and feedback in ratings
in a two round Delphi survey ,results of randomized controlled trial ,medicinal care
1999,37,964-8
Davies H et al. Reliability of chest X-ray of LRI in young children. pediatr inf dis
1996,16,600-4
Shann F et al. LRI in children, possible criteria for selection of patients for antibiotic
therapy and hospital admission, bull of the WHO 1984,62,749-53
The WHO young infant study group, clinical prediction of serious bacterial infection in
young infants in developing countries. Pediatr inf dis J 1999,18 suppl 8: 523-31
Mulholland EK et al. Standardized diagnosis of pneumonia in developing countries.
Pediatr inf dis J 1992,11,77-81
Kalteer HD et al. Identifying sick children requiring referral to hospital in Bangladesh
.Bull WHO 1997,75 supp1,65-75
BTS/SIGN ,British guideline on the management of Asthma Thorax 2003,58: supp 1
Cherian T et al evaluation of simple clinical signs for the diagnosis of LRI Lancet 1988

76

77
Instructions to contributors
1. All material submitted to the considered for publication must be submitted
exclusively to the Journal. All manuscripts should be typed in double spacing
and should be sent in triplicates accompanied by an electronic copy to the
Editor- in- chief of Journal.
2. Abstracts should be prepared in a structured format of about 250 words must
be included. The abs tract should include: objectives(S) ,design, setting,
subjects of participants invitations, interventions, main outcome measures,
result and conclusion.
3. Format and style: Original articles should include: abstract, Introduction
materials and methods, Results and Discussion, Acknowledgements and
References.
4. Abbreviations may be used after the long form has been written, e.g. the
world health Organization (WHO) united nation (UN). Drugs should be
referred to be approved and proprietary names. Scientific measurements
should be given in SI units. Tables and Figure should be typed on separated
Sheet. It is advisable that table s should not exceed five in number.
5. The title page should contain (i) the title of the article, (ii) Author (s) name (s)
and signature (s), degrees, designation, name of institution where the work
was carried out, present mailing address, telephone and E- mail and (iii)
name and address of the author to whom all correspondences should be
addressed.
6. References made to Journal should use the name of the journal abbreviated
according to the style used in Index Medic us, IT should follow the Vancouver
style: Year, volume, first and last page number E.g. Samson T. Lakech G
.Malnutrition and enteric parasitizes among under five children in Aynalem
Village, Tigray, Ethiop J HLth Dev 2000; 14:67-75
7. References made to books should include the authors or editors last name.
Initils, titles edition number, place of publication, publisher and year E.g.
Hailu B,Ali H. Chronic cough in children. In the tropics. Addis Ababa Medical
books pub 2006: 110-120

77

78
8. All manuscripts submitted to the Journal per peed reviewed. Manuscripts
are sent to two external referees selected from pool of experts.
9. All authors must give signed consent to publication. Authors are responsible
for the views, opinions and authenticity of the material published in the
journal.
10. Rejected articles will not be returned to the authors

Acknowledgements
The printing of the Journal was made possible through the generous contribution of
UNICEF Ethiopia

78