ORIGINAL

Niger J Paed 2013; 40 (1): 34 –39

Brown BJ

Jacob NE

Lagunju I A

Jarrett OO

Morbidity and mortality pattern in

hospitalized children with sickle cell

disorders at the University College

Hospital, Ibadan, Nigeria

DOI:http://dx.doi.org/10.4314/njp.v40i1.6

Accepted: 16th June 2012

Abstract Objectives: To determine missions. Associated infections

the causes of hospitalization and were septicaemia in 56 (32.2 %),

outcome of children with sickle cell malaria in 49 (28.2 %), acute osteo-

disorders at the University College myelitis in 24 (13.8%), pneumonia

Brown BJ

(

) Lagunju IA

Department of Paediatrics,

College of Medicine

University of Ibadan/University College

Hospital, Ibadan

Tel: +23408051875510

Email: biosbrown@yahoo.com

Hospital, Ibadan.

in 23 (13.2%), urinary tract infec-

Methods: Case files of patients with tion in 12 (6.9%) and septic arthritis

sickle cell disease who were admit- in 10 (5.7%). Haematocrit was less

ted between March 2009 and Febru- than 15 % in 36 (20.7%) and blood

ary 2012 were analysed. Data ex- transfusion administered in 68

tracted include demographic vari- (39.1%) of admissions. There were

ables, diagnoses, types of crises, three (1.7%) deaths from cere-

associated infections, complications brovascular accident, adverse reac-

Jacob NE, Jarrett OO

Department of Paediatrics,

University College Hospital

Ibadan, Nigeria.

and outcome of treatment.

tion to blood transfusion and men-

Results: There were 174 admissions ingitis.

of 161 children with a male female Conclusion: Prevention and prompt

ratio of 1.3:1. Their ages ranged management of crises and infec-

from nine months to 18 years with tions in sickle cell disease is recom-

a mean of 7.3(4.0) years. Vaso- mended to reduce morbidity and

occlusive crisis was present in 107 mortality.

(

2

61.5%), hyper haemolytic crisis in

9 (16.7%) and acute splenic se- Key words: Admissions, sickle

questration in 12 (6.9 %) of all ad- cell, crisis

Introduction

important. Information on the causes and pattern of mor-

bidity of hospitalization may highlight the most severe

manifestations with potential for mortality. Such infor-

mation will be helpful in targeting prevention strategies₄,

healthcare planning and appropriate resource allocation.

Studies on hospitalized children with sickle cell disor-

ders in Nigeria have often focused on a single complica-

tion such as_,₆either patterns of crises alone or patterns of

Sickle cell disease is a genetic disorder characterized by

chronic haemolysis resulting from premature destruction

of brittle and poorly deformable red blood cells. It is

caused by the presence of sickle haemoglobin which

th

results from substitution of glutamic acid at the 6 posi-

1

tion of its β chains by valine. Apart from chronic hae-

5

molysis, other manifestations of sickle cell anaemia are

attributable to ischaemic changes resulting from vascu-

lar occlusion by masses of sickled red cells. The clinical

course of affected children is typically associated with

intermitt₁ent episodic events, often referred to as

infections ; few have focused on the overall pattern of

morbidity in all hospitalized children showing t₂he rela-

tive contributions of different complications .

The

value of morbidity data in resource allocation is typified

by the establishment of a Day hospital for the manage-

ment of painful crises in the Bronx, New York based on

the finding that painful crises were responsible for the

highest proportion of hospitalizations. The result was a

significant reduction in admissions and length₇ of stay,

equivalent to savings of 1.7 million US dollars.

‘crises.’ Another major problem in children with sickle

cell disease is altered splenic function resulting in in-

creased susceptibility to infections by encapsulated or-

ganisms, and causing meningitis, septicaemia and other

serious infections. Complications of sick₂le cell disease

when severe often require hospitalization.

The objectives of this study were therefore to describe

the morbidity pattern and outcome of children with

sickle cell disorders admitted at the Paediatric wards of

University College Hospital Ibadan, Nigeria.

Nigeria has the highest burden of sickle cell anaemia

worldwide with₃ a prevalence of about 20 per 1000 live

births annually. In order to plan for health care of af-

fected children, a knowledge of the morbidity pattern is

3

5

Methods

Table 1: Socio-demographic parameters of children

Parameter

Frequency Per cent

This was a retrospective descriptive study of children

with sickle cell disease admitted into the Paediatric

wards of the University College Hospital, Ibadan be-

tween March 2009 and February 2012. The University

College Hospital is a tertiary hospital located in the city

of Ibadan, south western Nigeria. Haemoglobin pheno-

type of all patients was established by haemoglobin

electrophoresis at a pH of 8.6 at the haematology labora-

tory of the hospital. Proguanil for malaria prophylaxis is

routinely used by all patients attending in the Paediatric

sickle cell clinic. Pneumococcal vaccines are also rou-

tinely prescribed but repeatedly questioning of carers in

our clinic indicates that some of them usually receive it

following a long interval after prescription due to its

relatively high costs. Documentation after vaccination is

therefore not consistently done so that an accurate re-

cord of the uptake would not be obtained from their case

notes. . Penicillin prophylaxis is not routinely given in

our hospital.

Sex

Male

Female

Total

0-4

90

71

161

46

66

43

6

55.9

44.1

100

28.6

41.0

26.7

3.7

Age group (years)

5

-9

10-14

15

≥

Total

I

161

23

43

55

38

2

100

14.3

26.7

34.1

23.6

1.2

Socioeconomic class

II

III

IV

V

Data was extracted from the case notes of the patients

seen during the study period. Data extracted included

demographic variables, haemoglobin phenotype, type of

crisis, associated infections and other complications of

sickle cell disease. Other information extracted included

the haematocrit on admission, number of blood transfu-

sions, duration of hospital admission, outcome of treat-

ment and final diagnosis at discharge or death. In our

hospital, children are usually discharged when there is

some resolution of the presenting symptoms and signs,

intravenous drugs are no longer necessary and carers

will be able to continue care for the children at home.

Demographic variables extracted included patients’

ages, sex and educational levels and occupations of their

parents. The families were classified into socioeco-

nomic groups using a combined score derived from the

occupation and maximum educational level of both par-

Total

161

100

Different forms of crises were present in 130 (74.7%)

the 174 hospital admissions; the most frequent form of

associated crisis was the vaso-occlusive type and the

least common was the aplastic type (Table 2). Of the 29

children that had hyperhaemolytic crises, 16 also had

associated vaso-occlusive crisis. Two patients had each

had a combination of painful and splenic sequestration

crises. All cases of acute splenic sequestration (ASS)

and aplastic crisis were of the Hb SS phenotype. Eight

(72.7 percent) of the cases of ASS were aged five years

or older. Of the 130 children with different forms of

crises, there was associated infection in 97 (74.6%) of

them. The commonest infections associated with crises

were septicaemia (46), malaria (40), acute osteomyelitis

(20), and pneumonia (15). Less common associated in-

fections were urinary tract infection (nine), aseptic ar-

thritis (eight), pharyngitis (four) chronic osteomyelitis

(three) and meningitis (one).

8

ents (or their substitutes) as described by Oyedeji. Data

were entered into a microcomputer and analyzed with

SPSS version 20.0. Means, standard deviations, medi-

ans and inter quartile range computed for continuous

variables. Categorical variables were presented as fre-

quencies.

Table 2: Frequency and age distribution of sickle cell

crisis in 174 admissions

Type of crisis

n (%)

Age (years)

Range

Results

Mean (SD)

7.1 (4.0)

8.5 (4.7)

8.9 (4.3)

Vaso-occlusive

107 (61.5) 0.75-18.0

There were 174 admissions of 161 children during the

period of study. Of these, 149 (85.6%) had one admis-

sion each, 22(12.6%) had two admissions each and

three (1.7%) had three admissions each. Their ages

ranged from nine months to 18 years with a mean (±1

Hyper haemolytic

29(16.7)

11 (6.9)

1.4-18.0

2-14

Acute splenic se-

questration

A plastic

1 (0.6)

7

standard deviation) of 7.3 (4.0) years.

Socio-

demographic parameters of the children are shown in

Table1.

SD= standard deviation

The haemoglobin phenotype was SS in 148 (91.9%) and

SC in 13 (8.1%) of the children studied.

3

6

Different forms of infections were present in 126

72.4%) of the 174 admissions with a male: female ratio

of 1.4:1. The commonest infections for which the

children were treated were septicaemia, malaria, acute

osteomyelitis and Pneumonia (Table 3).

new cases representing 6.9 percent of all admissions.

All cases of CVA were of the HbSS phenotype. There

was also one case each of Hodgkin lymphoma and non-

Hodgkin lymphoma admitted for chemotherapy.

(

Duration of admission

Table 3: Frequency of infections in 174 admissions of

children with sickle cell disease

Duration of admission ranged from 1-131 days with a

median of 8 days and inter-quartile range of 5-15 days.

There was no significant association between the type of

crises and being admitted for more than seven days

Infection*

n

%

(

Table 4). Analysis of the risk of being admitted for

Septicaemia

Malaria

56

49

32.2

28.2

longer than seven days computed for the different infec-

tions and complications revealed that patients with septi-

caemia, acute osteomyelitis, cerebrovascular accidents

and avascular necrosis of the femoral head were at in-

creased risk of prolonged admission (Table 5).

Acute osteomyelitis

Pneumonia

24

23

13.8

13.2

Septic arthritis

UTI

10

12

3

5.7

6.9

1.7

2.9

Table 4: Relationship between type of crises and dura-

tion of admission in 130 children who presented with

crises

Meningitis

Chronic Osteomyelitis

Pharyngotonsillitis

3

5

Type of crises

Admitted Admit-

Total Exact

(N=130 Sig

for ≤ 7

ted for

*multiple infections present in some patients

days

>7 days

(n= 67)

)

(2side)

(n=63)

Bacterial isolates

Pain

42

47

89

13

9

0.709

0.388

1.0

Hyper-haemolytic

8

5

Twenty five (44.6%) of the 56 children with suspected

septicaemia had blood cultures done which was positive

in 12 as follows: Staphylococcus aureus (five), Kleb-

siella species (five), Escherichia coli (one) and Coagu-

lase- negative staphylococcus (one). Diagnosis of sep-

ticaemia in the other children was based on their clinical

presentation, negative malaria parasite tests and urine

cultures, and presence of toxic granulations in neutro-

phils. Some patients could not afford the cost of blood

culture before commencement of antibiotics. Significant

bacteriuria was found in seven of the 12 children treated

for urinary tract infection (UTI) and the urinary isolates

were Klebsiella species in four cases and Escherichia

coli in three cases. The remaining children treated for

UTI were symptomatic and had pyuria.

Acute splenic

sequestration

Aplastic crises

4

5

1

0

1

0.485

0.792

1.000

Pain + Hyperhemolytic

crises

Pain + splenic seques-

tration

Total

7

9

16

2

1

63

67

130

Table 5: Duration of admission and risk factors for ad-

mission longer than seven days

Complications

Median dura-

tion of admis-

sion (days)

Risk of admission > 7 days

Odd

ratio

95 % Confi-

dence Interval

Multiple bacterial agents were isolated in some patients.

Klebsiella spp were isolated from the blood and urine of

a child with septicaemia, urinary tract infection and

radiological features of acute osteomyelitis. Another

patient had Klebsiella septicaemia and Escherichia coli

UTI. A third patient had Staphylococcus aureus septi-

caemia and Klebsiella UTI.

Septicaemia

Malaria

11

6

4.13

0.65

2.06, 8.29*

0.33, 1.26

Acute osteomyelitis

Pneumonia

14

8

6.03

1.32

4.2

1.97, 18.49*

0.54, 3.19

Septic arthritis

18

0.87, 20.38

Urinary Tract Infection 14

2.05

1.98

0.48

0.59, 7.08

0.18, 22.21

0.04, 5.42

Non-infectious complications

Meningitis

51

5

Chronic osteomyelitis

Among the 174 admissions, non-infective complications

of sickle cell disease observed were cerebrovasular

accident (CVA) in 16 (9.2%), transient ischaemic attack

in 1(0.6), afebrile seizure in 1(0.6%), priapism in

Pharyngotonsillitis

7

0.64

5.39

0.105, 3.95

Cerebrovascular acci-

dents

18

1.14, 25.35*

Avascular necrosis of

the femoral head

82

2.0

1.7, 2.3*

3

2

(1.7%), avascular necrosis of the femoral head in

(1.1%) and acute kidney injury in 2(1.1%). Four of

the 16 cases of CVA were previously diagnosed cases

on chronic blood transfusion programme and 12 were

*significant

3

7

Anaemia and Blood Transfusions

prior use of over-the-counter antibiotics and decreased

susceptibility to Streptococcus pneumoniae a₆s_,₁₄a result of

Haematocrit levels during admission ranged from 7-36

percent with a mean (SD) 20.7 (6.4) percent. There was

severe anaemia i.e. a haematocrit of less than 15 percent

in 36 (20.7%) episodes of admissions. Blood transfusion

was administered in 68 (39.1%) of admissions. The

numbers of blood transfusions received was one in each

of 51 (29.3%) admissions, two in each of 12 (6.9%) ad-

missions, 3 in each of 4 (2.3%) admissions and five in

splenomegaly which persists due to malaria.

There is

a need for large scale studies across equatorial Africa to

firmly establish the role of Streptococcus pneumoniae in

causing infections in children with SCD and to evaluate

the role, if any, of pneumococcal vaccines in preventing

infections and reducing mortality. The low yield of mi-

crobial aetiological agents for infections in this retro-

spective study due to inability of carers to pay for cul-

tures prior to commencement of antibiotics represents a

challenge to appropriate management. There is need for

prospective studies aimed at identifying aetiological

agents and antimicrobial sensitivity patterns for infec-

tions in children with sickle cell disorders to guide an-

timicrobial choices in their treatment.

1

(0.6%) admission.

Mortality

There were three deaths yielding a case fatality rate of

1

.9 percent of the 161 children. The causes of death

were cerebrovascular accident in one child and adverse

reaction to blood transfusion in a child on chronic blood

transfusion programme for secondary stroke prevention

in another. The third child was on chemotherapy for

Hodgkin lymphoma and died from meningitis with cere-

bral oedema and pulmonary haemorrhage confirmed at

autopsy.

Early diagnosis of infections in SCD and timely institu-

tion of appropriate treatment is paramount in reducing

morbidity and mortality. The present study showed the

co-existence of multiple infections in the same patient,

sometimes by different aetiological agents. This has

previously b₅een reported in children with sickle children

1

with SCD. A high index of suspicion of the possibility

of multiple infections is therefore necessary in evaluat-

ing this subset of children to guide complete treatment.

However, a potential obstacle to this process is the out

of pocket expenditure for laboratory investigations

which many parents are unable to afford as typified in

the present study. This may be addressed by strengthen-

ing of the Nation’s Health Insurance Scheme.

Discussion

The present study has shown that Sickle cell disease is a

common cause of hospital admission in Ibadan with

male preponderance similar to observations in Enugu;

2

Eastern Nigeria. The pattern of hospital admission of

The prominence of vaso-occlusive crises above other

forms of sickle₉_,c₁₀ell crises observed in the present study

children with sickle cell disease (SCD) varies in differ-

ent parts of the world. Whilst vaso-occlusive crises ac-

count for majority of admissions in some₉_,₁p₀arts of the

is well known.

Most cases of Acute Splenic Seques-

tration (ASS) are in children with H₁b₆ SS and are known

to occur below three years of age. The occurrence of

splenic sequestration crisis up to the age of 14 with ma-

jority occurring at five years of age or more in our study

contrad₁ic₆ts the previously known facts in the United

States. . Possible explanations of this different pattern

in the present study may be death of younger children

from ASS before presenting in hospital. Another possi-

bility may be persistence of enlarged spleens in children

with SCD in the tropics beyond the age when autosple₁-₄

nectomy should have set in, attributable to malaria.

This might leave room for ASS to occur.

world particularly in developed countries

infections

account for most a₂d_,₁m₁ issions in others, particularly de-

veloping countries. Given, the reduction in invasive

pneumococcal disease associated with penicillin prophy-

laxis, it is possible that the predominance of vaso-

occlusive crises and under representation of infections in

developed countries may be related to use of₉_,p₁₂neumo-

coccal prophylaxis in prevention of infections.

Our study showed that infections and various forms of

crises made almost equal contributions to admissions

with infections having a slight lead. In Nigeria like

many sub-Saharan African countries, penicillin prophy-

laxis and pneumococcal immunization in children with

SCD is not routine. This is attributed to purported lack

of evidence in support of the role of Streptococcus pneu-

The burden of anaemia and blood transfusion in the pre-

sent study contrast with those₂in a previous study in

Enugu, south-eastern Nigeria. In the present study,

severe anaemia was present in 20.7 percent of children,

blood transfusion given in 39.1 percent of admissions

and multiple blood transfusions in less than 10 per cent

of admissions; the corresponding figures in Enugu were

39.4 percent, 73.2 percent and 40.8 percent respec-

1

3

moniae in causing infections in SCD in tropical Africa.

Although the proportion of patients who had blood cul-

tures in our study was small due to financial constraints

faced by parents, the paucity of Pneumococcus as an

aetiological agent in bacteraemia observed is similar₅_,t₁o₃

findings from other studies in equatorial Africa.

Rather, Staphylococus aureus and Gram negative organ-

isms seem to be prominent in our study, in keeping with

2

tively. Possible explanations for these marked differ-

ences could be variations in compliance of patients with

routine haematinics and malaria prophylaxis as well as

variations in blood transfusion guidelines in the two

hospitals. Blood transfusion is usually life saving but

may also be associated with risks as typified by one

6

findings in a previous Nigerian study. The rarity of

Pneumococcal isolates in Nigeria has been attributed to

3

8

death in the present study attributed to blood transfusion

reaction. It is therefore necessary to counsel and support

parents and patients on prompt health seeking habits

that minimize the risk of severe anaemia and subsequent

blood transfusion.

Conclusion

Hospital admission of children with sickle cell disease in

Ibadan, Nigeria results from a wide variety of clinical

conditions but mainly infections and various forms of

crises. There is a need to clearly establish the aetiologic

agents of infections in these children to guide appropri-

ate antimicrobial therapy. The use of prophylaxis

against pneumococcal infections if supported by evi-

dence, has a great potential of reducing morbidity and

mortality ₁as has been shown in the United States of

Hospitalization of children with SCD constitutes a sig-

nificant burden on the caregivers. The median duration

of hospitalization in the present study is similar to find-

9

ings reported in the United Kingdom. In addition, the

findings of the present study indicate that individuals

with septicaemia, acute osteomyelitis, cerebrovascular

accident and avascular necrosis of the femoral head

were at increased risk of prolonged admission. In an

attempt to lighten the burden of admissions, some hos-

pitals instituted carrying out elective blood transfusion₁s₇

in day care settings without overnight admissions.

Considering the finding that almost 40 percent of our

study patients required blood transfusion, provision and

utilization of daycare settings may go a long way in re-

ducing the frequency of hospital admissions particularly

in children in whom no major infection or complication

needing prolonged admission is found.

1

America. There is also need for establishment of com-

prehensive programmes for sickle cell disease in Nigeria

to improve the quality of life of affected children. This

should include neonatal diagnosis and early institution

of health promoting measures and regular screening tests

for possible complications such as risks for stroke.

Unhindered access to health care is also paramount in

relieving the financial burden borne by caregivers and

therefore giving the children a better chance of survival.

Conflict of interest: None

Funding: None

The case fatality observed in this study was lower than

6

cell anaemia in Zambia and 8.5 per cent in Enugu Ni-

geria. Although the small number of deaths in our

.61 percent observed in₈ admitted children with sickle

1

2

study makes it difficult to make strong conclusions on

the pattern of mortality, the fact that two out of the three

deaths were related to cerebrovascular accidents calls for

the need to strengthen measures towards primary stroke

prevention. This need is further heightened by the rela-

tively high contribution of stroke to admissions. Routine

screening of children with SCD by Transcranial Doppler

ultrasonography to detect high risk cases and institution

Acknowledgment

The role of resident, consultant and laboratory staff in

the management of children with SCD in the department

is acknowledged. The resident doctors were also of as-

sistance in data collection.

Limitation

1

9

of appropriate preventive measures is recommended.

Blood transfusion greatly reduces the risk of a first

stroke in children with sickle cell anaemia who have

abno₁r₉mal results on transcranial Doppler ultrasonogra-

phy.

The low yield of bacterial isolates among patients

treated for sepsis partly due to inability of the parents to

pay for microbial cultures make the results on aetiologi-

cal organisms of limited value for generalization.

Reference

5

6

7

8

.

Juwah AI, Nlemadim E U, Kaine

9. Brozovic M, Davies S C, Alison I

Brownell AI. Acute admissions of

patients with sickle cell disease

who live in Britain. Br Med J

1987; 294:1206-8.Affiliations

10. Address correspondence and re-

print requests to Marian B. Fosdal,

MA, RN, CPON®, Riley Chil-

dren's Cancer Center, 702 Barnhill

Drive, #1960 Indianapolis, IN

46202.

11. Fosdal M B, Woiner-Alexandrov

A W. Events of Hospitalization

among Children with Sickle Cell

Disease. J Pediatr Nurs 2007;

22:342-346.

1

2

.

Honig GR. HemogIobin disorders.

In Behrman RE, Kliegman RM,

Jenson HB, eds. Nelson Textbook

W. Types of anaemic crises in

paediatric patients with sickle cell

anaemia seen in Enugu, Nigeria.

Arch Dis Child 2004; 89:572–576.

Akuse R M. Variation in the pat-

tern of bacterial infection in pa-

tients with sickle cell disease re-

quiring admission. J Trop Pediatr

th

of Pediatrics. 16 ed. Philadelphia.

WB Saunders, 2000:1478-1483.

Ikefuna AN, Emodi IJ. Hospital

admission of patients with sickle

cell anaemia pattern and outcome

in Enugu area of Nigeria. Niger J

Clin Pract 2007; 10:24-29.

World Health Organization.

Sickle cell anaemia. WHO Fifty-

Ninth World Assembly A59/9,

1

996; 42:318-23.

Benjamin L J, Swinson G I, Nagel

R L. Sickle cell anemia day hospi-

tal: approach for the management

of uncomplicated painful crises.

Blood 2000; 95:1130-6.

Oyedeji GA. Socio-economic and

cultural background of hospital-

ized children in Ilesa. Nig J Paed

3

4

.

2

006.

Eck C, Pierre RB, Hambleton IR.

Medical paediatric admission pat-

terns at the University Hospital of

the West Indies: issues for future

planning West Indian Med J 2006;

1

985; 12: 111-17.

5

5:340-5.

3

9

1

2. Mabiala-Babela JR, Nkanza-

Kaluwako SA, Ganga-Zandzou

PS, Nzingoula S, Senga P. Effects

of age on causes of hospitalization

in children suffering from sickle

cell disease. Bull Soc Pathol Exot

15. Bwibo NO, Kasili EG. Clinical

aspects of sickle cell disease in

Nairobi children. Am J Pediatr

Hematol Oncol 1982; 4:187.

16. Asinobi A O, Fatunde O J, Brown

BJ, Osinusi K, Fasina N A. Uri-

nary tract infection in febrile chil-

dren with sickle cell anaemia in

Ibadan, Nigeria. Annals Trop Pae-

diatr 2003; 23:129-134.

18. Day TG, Thein SL, Drasar E, Dick

MC, Height SE, O'Driscoll S, Rees

DC. Changing pattern of hospital

admissions of children with sickle

cell disease over the last 50 years.

J Pediatr Hematol Oncol 2011;

33:491-5.

19. Athale UH, Chintu C. Clinical

analysis of mortality in hospital-

ized Zambian children with sickle

cell anaemia. East Afr Med J 1994;

71:388-91.

2

005; 98:392-3.

1

3. Gaston MH, Verter JI, Woods G,

et al. Prophylaxis with oral peni-

cillin in children with sickle cell

anemia. A randomized trial. N

Engl J Med 1986; 314:1593-9.

1

7. Driscoll C M. Sickle cell disease.

Pediatr Rev 2007; 28:259-268.

2

0. Adams RJ, McKie VC, Hsu L, et

al. Prevention of a first stroke by

transfusions in children with sickle

cell anemia and abnormal results

on transcranial Doppler ultrasono-

graphy. N Eng J Med 1998; 339:5

1

4. Kizito M E, Mworozi E, Ndugwa

C, Serjeant G R. Bacteraemia in

homozygous sickle cell disease in

Africa: is pneumococcal prophy-

laxis justified? Arch Dis Child

-11.

2

007; 92:21–23.