Recurrent antibiotic use in a small child and the effects on the family

Abstract

To determine the typical characteristics of small children on recurrent therapy with antibiotics (RTA) and the effects RTA have on the families.
Stratified randomised cluster sampling.
An unselected population-based questionnaire study in Finland.
Parents (n = 1443) expecting their first child were followed prospectively from the mother's early pregnancy until the child was 18 months of age.
The outcome measure was the number of courses of antibiotic therapy (options: 011-51 > or = 6) during the child's first 18 months of life. Six or more courses were considered RTA. Associated variables were child- and family-related factors during the child's first 18 months of life.
Children on RTA were taken to see a physician more often than other children when they had fever or a common cold. RTA was significantly associated with the child's day care outside the home, allergy and need for a special diet. The father's severe stress was associated with the child's RTA. Breastfeeding lasting more than 3 months was found to have a protective effect against RTA.
The threshold for seeking medical help was low in the families of children on RTA, which is best avoided by breastfeeding and day care at home. Health care staff should not forget to support families, especially the fathers, with children experiencing recurrent infections.

Full text

Recurrent antibiotic use in a small child and the effects
on the family
Katri Louhi-Pirkanniemi
1,2
,Pa¨ ivi Rautava
1,3
, Minna Aromaa
1
, Ansa Ojanlatva
3
, Jussi Mertsola
4
,
Hans Helenius
5
and Matti Sillanpa¨a¨
1,6
1
Department of Public Health, University of Turku,
2
Paimio-Sauvo Health Authority, Paimio,
3
Turku City Hospital, Turku,
4
Department of Pediatrics, University of Turku,
5
Department of Biostatistics, University of Turku,
6
Department of Child
Neurology, University of Turku, Turku, Finland.
Scand J Prim Health Care 2004;22:16/21. ISSN 0281-3432
Objective

/To determine the typical characteristics of small children
on recurrent therapy with antibiotics (RTA) and the effects RTA have
on the families.
Design

/Stratified randomised cluster sampling.
Setting

/An unselected population-based questionnaire study in
Finland.
Subjects

/Parents (n/1443) expecting their first child were followed
prospectively from the mother’s early pregnancy until the child was 18
months of age.
Main outcome measures

/The outcome measure was the number of
courses of antibiotic therapy (options: 0/1

/5/]/6) during the child’s
first 18 months of life. Six or more courses were considered RTA.
Associated variables were child- and family-related factors during the
child’s first 18 months of life.
Results

/Children on RTA were taken to see a physician more often
than other children when they had fever or a common cold. RTA was
significantly associated with the child’s day care outside the home,
allergy and need for a special diet. The father’s severe stress was
associated with the child’s RTA. Breastfeeding lasting more than 3
months was found to have a protective effect against RTA.
Conclusion

/The threshold for seeking medical help was low in the
families of children on RTA, which is best avoided by breastfeeding
and day care at home. Health care staff should not forget to support
families, especially the fathers, with children experiencing recurrent
infections.
Key words: antibiotic, children, families, recurrent, therapy.
Katri Louhi-Pirkanniemi, Department of Public Health, Lemmin-
ka
¨isenkatu 1, FIN-20014, University of Turku, Finland. E-mail:
katlou@netti.fi
Recurrent infections in a child may affect family life in
the same way as long-term illnesses do (1). They can
have an enduring effect on the life and health
behaviour of a family, although the effects usually
disappear as the child grows up (2). With the exception
of otitis media, hardly any studies have been published
on the relationship between recurrent courses of
antibiotics for infections and familial psychosocial
factors. During a child’s recurrent infections, the
parents eventually learn to cope with the effects of
the disease process on the family. Family members
tend to seek and rely on professional help. Parents
expect and need comfort, care and emphatic support
from professionals in addition to advice and pre-
scribed medication (3).
Our hypothesis was that, in addition to infectious
symptoms, psychosocial and behavioural factors affect
the use of antibiotics in the treatment of a child during
infections. The purpose of this study was to investigate
the relationships between intrafamilial social circum-
stances and recurrent therapy with antibiotics (RTA).
MATERIAL AND METHODS
The present study was part of the Finnish Family
Competence Study (FFC Study) (http://www.utu.fi/
med/kansanterv/ffc.html). The original FFC Study
was launched in 1985 in an effort to help improve
existing preventive health care services for young
families in Finland. The representative study sample
was collected in a stratified randomised cluster
procedure in the Province of Turku and Pori in 1986.
The study design has been described in detail else-
where (4).
Overuse of antibiotics is a growing global
problem, and small children are receiving a large
proportion of the antibiotic therapy prescribed.
.Familial psychosocial factors, in addition to
medical indications, are associated with re-
current antibiotic use in children.
.Recurrent therapy with antibiotics can be
avoided if mothers breastfeed their children
and by the families receiving day care at home
and emotional support.
.Health care personnel must be alert to
respiratory infection in children and pay close
watch to the father’s role.
æ
ORIGINAL PAPER
Scand J Prim Health Care 2004; 22 DOI 10.1080/02813430310003165
Scand J Prim Health Care Downloaded from informahealthcare.com by 61.163.165.165 on 05/20/14
For personal use only.

The study population consisted of pregnant nulli-
parous women. Nurses at 67 maternal health care
clinics recruited 1582 women, 1443 of whom (91%)
gave their informed consent to participate. The present
study is part of a project investigating factors asso-
ciated with RTA and the possible long-term effects on
children.
Under the routine well-baby care system, the
children were seen by either a designated public health
nurse or a designated physician, or both, several times
until the age of 18 months. In addition to the health
records held at the well-baby clinics, each child had a
personal health diary for the recording of dates of
immunisations, results of health visits, and medica-
tions and treatments.
The outcome measure was the number of courses of
antibiotic therapy during the child’s first 18 months of
life as documented by the parents at ages 9 months
and 18 months. The numbers of courses of antibiotic
therapy were grouped into three categories
(0/1
/5/]/6). Six or more courses were considered
RTA. Since distribution of the outcome measure was
highly skewed (the reported courses of antibiotic
therapy varied from zero to 20), it was put in a
category rather than being used as a continuous
variable. Apart from the fact that six or more courses
of therapy could be considered a large number during
the first 18 months of life, categorisation in this study
was accomplished so that the number of observations
in each category would be feasible for analysis. The
cut-off point was based on data from earlier studies
(5
/7).
The 9-month questionnaire was returned by 1111
mothers and the 18-month questionnaire by 1025
mothers. The parents of 817/1025 children (80%)
completed the questionnaire satisfactorily, including
questions on the occasions of antibiotic therapy at
both stages of the study. Of these 817 children, 202
had received no courses of antibiotic therapy, 496
children had received 1 to 5 courses and 119 children 6
or more courses of antibiotic therapy.
The data on associated variables were examined and
grouped into family-related and child-related variables
(Table I). Three separate questionnaires were used,
one for the mother, one for the father and one for the
nurse. The mother’s and the father’s questionnaires
were not identical, and there were no questions for the
parents to answer jointly. The number of questions in
the 9-month questionnaire was 61 for mothers, 26 for
fathers and 10 for nurses. In the 18-month question-
naire, there were 68, 41 and 11 questions, respectively.
The Denver Development Screening Test (8,9) was
applied to the neurodevelopmental assessment of the
children. The Denver scale for measuring the devel-
opment and growth of children was filled in by the
parents at home. The nurses assessed the children at
the 9 and 18 months follow-up, and they too filled in
the Denver scale.
In the drop-out analysis, young parents (mother’s
age B/20 years, pB/0.004; father’s age B/25 years,
p
/0.015) and mothers with a lower level of basic
education (less than high school education, p
/0.003)
were more common among non-participants than
among participants.
Statistical methods
The significance of univariate associations between the
outcome variable and the associated variables was first
tested using Pearson’s chi-square test. For multivariate
analyses of family-related factors, the associated vari-
ables were classified into subgroups by context.
Stepwise logistic regression analyses were performed
Table I. Associated variables of a child’s recurrent therapy with antibiotics.
Family-related factors
Parental well-being factors: Parents’ subjective assessments of their health status, physical condition and mental well-being, smoking
habits and stress experienced.
Parent
/infant-related factors: Parents’ subjective opinions about the child’s health, problems with eating and sleeping, and different
types of crying, the mother’s opinion about the child’s behaviour in situations of separation and on days of day care compared with
holidays, the father’s time with his children (duration and quality), parents’ adaptation to changes brought about by the infant’s
infection.
Family functioning factors: Families’ need for outside help (financial, child care, housekeeping), disagreement on child-rearing, well-
baby clinic nurses’ opinions about the family’s need for extra help.
Social factors: Frequencies of parents’ visits to their friends and relatives, leisure time activities and hobbies, information about
parents’ problems with grandparents concerning child-rearing, parents’ experiences with the information and support offered by
physicians and nurses at the well-baby clinic.
Child-related factors
The child’s weight and height, numbers of periods of hospital stay and medical visits, number of illnesses and paracentheses, scheduled
immunisation programmes, problems concerning skin care, digestive functions, allergic symptoms, need for any special diet, types of
day care. Denver scale (9,10)
Recurrent antibiotic use in a small child 17
Scand J Prim Health Care 2004; 22
Scand J Prim Health Care Downloaded from informahealthcare.com by 61.163.165.165 on 05/20/14
For personal use only.

for each subgroup separately. Multinominal logistic
analyses were applied, because the outcome measure
had three classes (10). The final logistic regression
analysis included all statistically significant associa-
tions of the previous analyses of family-related factors.
Significance level was B/0.05. Child-related factors
were examined as a separate entity. The child’s medical
visits and illnesses (Table II) were examined using the
Kruskal-Wallis test, and the data of the Denver test
(8,9) using the Anova test.
The Joint Ethics Review Committee of the Medical
Faculty of the University of Turku and the University
Central Hospital of Turku approved the design of the
study.
RESULTS
The median number of visits to a physician by children
on RTA was 4 (Q
1
/2, Q
3
/7[Q
1
/25th percentile,
Q
3
/75th percentile]) over 0/9 months, 9 (Q
1
/6,
Q
3
/13) over 10/18 months and 14 (Q
1
/11,
Q
3
/21) over 0/18 months. These children were taken
to see a physician with fever or common cold
significantly more often than children without RTA
(Table II).
Family-related factors
Table III gives the family-related factors that were
significantly associated with the child’s RTA in uni-
variate logistic regression analyses. In multivariate
stepwise logistic regression analysis, the father’s ex-
perience of severe stress, the mother’s estimation of the
child’s poor health status, the father’s opinion that
there was no need for temporary child care help in the
daytime, and the mother’s opinion that the physician
at the well-baby clinic was too busy were associated
with the child’s RTA (Table III).
In the final logistic regression analysis of family
factors, the father’s experience of severe stress re-
mained the only independent factor associated with
the child’s RTA (Table III).
Child-related factors
Table IV gives the child-associated factors that were
significantly associated with the child’s RTA in uni-
variate analyses. In multivariate stepwise logistic
regression analysis, day care outside the home, allergy,
need for a special diet, hospital stay and duration of
breastfeeding (more than 3 months) were independent
factors associated with RTA (Table IV).
No statistically significant differences were seen
between the therapy groups in speech development,
gross or fine motor functions or psychosocial activities
(Denver score).
DISCUSSION
The most interesting finding of the study about
families was the father’s psychological well-being in
families with a child on RTA. These fathers experi-
enced more stress than the other fathers. Previous
studies of children with medical problems have usually
emphasised the mother’s role (11,12). We also found
that children with fevers or common colds on RTA
were taken to see a physician more often than those
without RTA. This suggests that the threshold of
parents with children on RTA for seeking medical help
for the child’s health problems is low. The more often
they consult a physician, the more likely the child
receives the antibiotic therapy.
The original study population was a stratified
randomised cluster sample of young Finnish families
in south-western Finland with no significant skewness
in distribution and with a high participation rate (4).
The basic family characteristics of the first-born
children and the child-rearing experiences of these
parents were acceptably similar, which is an advantage
of the study setting. The Finnish health care system is
a good setting for research, because it provides all
families with an equal opportunity to receive health
care and preventive services.
The data of the present study were collected in the
late 1980s and might be considered as old. However,
the results are still valid, because no relevant changes
have since occurred in the socio-economics and
support methods of the public health system in
provision for families with small children. The father’s
role has not changed dramatically, but fathers spend
more time at home with the mother and the newborn
baby than they did before. One element has changed:
the use of antibiotics has increased (13,14). This is a
well-known problem which should make physicians
carefully consider the prescription of any course of
antibiotic therapy.
Table II. Medians (Q
1
,Q
3
) of children’s visits to physicians and
common colds and fevers at 10
/18 months of age.
No. of courses of antibiotic
therapy
01
/5]/6
p
Median
(Q
1
,Q
3
)
Median
(Q
1
,Q
3
)
Median
(Q
1
,Q
3
)
Visits to physicians B/0.05 1 (0,2) 3 (2,5) 9 (6,13)
Common colds and
fevers
3 (2,5) 5 (3,7) 8 (5,14)
Common colds and
fevers per one visit to
physician
2 (1,4) 1.5 (1,2) 1 (0.5,1.3)
Q
1
/25th percentile, Q
3
/75th percentile.
18 K. Louhi-Pirkanniemi et al.
Scand J Prim Health Care 2004; 22
Scand J Prim Health Care Downloaded from informahealthcare.com by 61.163.165.165 on 05/20/14
For personal use only.

The definitions of recurrent infections and recurrent
courses of antibiotic therapy remain inadequate.
Children have an average of 1.1
/1.2 episodes of acute
otitis media during their first year of life, and 0.7
episodes annually at the age of 1 to 5 years (15,16).
The Social Insurance Institution of Finland grants
disability allowance to a family when a child has had
six or more infections per year with at least as many
courses of antibiotic therapy (www.kela.fi). Based on
these facts and on data from earlier studies (5
/7), our
criterion of RTA (six or more courses) can be
considered reasonable.
In drop-out analysis, younger parents were more
common among non-participants. Without these
drop-outs, the findings would have been equally or
perhaps even more notable.
Family-related factors
The families of children on RTA seem to need
frequent health visits and adequate time with the
same, familiar, physician that they can call on in case
of need. A child’s recurrent infections can affect family
life. The parents of otitis-prone children have been
found to make changes in their daily routines (working
hours, child day care, housing and smoking habits) in
an effort to improve the child’s health status (17), i.e.
changes that may be considered positive parenting
behaviour. However, a child’s illnesses may affect the
Table III. Univariate and multivariate stepwise logistic regression analyses of family-related factors associated with RTA. Calculated
using multinomial logistic regression analysis with zero course as reference class.
Univariate
analyses
Multivariate analysis Final logistic analysis
1
/5 courses ]/6 courses 1/5 courses ]/6 courses
Variable p p OR (95% CI) OR (95% CI) p OR (95% CI) OR (95% CI)
Parental well-being (n
/645)
The mother’s smoking
1
0.037
The father’s stress measured with stress
points
1
0.003 0.008 0.012
1 vs 0 1.0 (0.6 
/1.6) 1.5 (0.7 /3.0) 1.4 (0.7 /2.9) 3.0 (1.1 /7.9)
]/1 vs 0 2.1 (1.0 /4.4) 4.6 (2.0 /11) 2.1 (0.8 /5.4) 5.9 (1.9 /18)
Parent

/infant-related factors (n/713)
The father’s reduced ability to recognise
different types of child’s crying
1
0.036
Awake at night
2
0.043
The mother’s estimation of the child’s
physical condition
1
B/0.001 B/0.001
Ill vs healthy 3.1 (1.2 
/8.3) 6.4 (2.2 /18)
Eating problems
1
0.021 0.010
Moderate vs none 1.7 (1.1 
/2.5) 0.9 (0.5 /1.8)
Many vs none 0.8 (0.5 
/1.3) 1.1 (0.6 /2.0)
Family functioning (n
/452)
The mother’s opinion about help
No need for temporary help
1
0.031
Need for financial help
1
0.018
No need for temporary help
2
0.007
Nurse’s opinion of the families’ need of
support outside home (yes vs no)
2
B/0.001
Parents are not married
1
0.001
Parents’ common-law marriage
1
0.013 0.037
Yes vs no 0.8 (0.4 
/1.5) 1.9 (0.9 /4.2)
The father’s opinion about temporary
help in child care in the daytime
1
0.011 0.010
No vs yes 1.1 (0.7 
/1.7) 2.5 (1.3 /5.1)
Social relations (n
/615)
The mother’s opinion that the physician
in the well-baby clinic was too busy
1
0.028 0.010
Yes vs no 1.8 (1.1 
/2.8) 2.3 (1.2 /4.2)
The mother’s disagreements with
grandparents about child-rearing
2
0.040 B/0.001
Yes vs no 0.5 (0.4 
/0.8) 1.0 (0.6 /1.8)
OR
/odds ratio, CI/confidence interval.
1
Data from the 3/9-month period.
2
Data from the 9/18-month period.
Recurrent antibiotic use in a small child 19
Scand J Prim Health Care 2004; 22
Scand J Prim Health Care Downloaded from informahealthcare.com by 61.163.165.165 on 05/20/14
For personal use only.

psychological status of the parents, their work activ-
ities and work performance, as well as the economic
situation of the family (18).
Child-related factors
Day care outside the home has been shown to be a risk
factor for recurrent infections (19
/21), and the present
study supports this finding. Diagnosed allergies and
the need for a special diet were more common among
RTA children than among non-RTA children. Allergy
(22) has been found to be associated with recurrent
otitis media. The causes of any connections between
allergy and recurrent otitis media remain to be
determined (23) and need further study.
A previous study has illustrated that a short period
of breastfeeding (less than 3 months) is a risk factor
for recurrent otitis media (19). The present study
illustrated the same correlation between RTA and the
duration of breastfeeding. Some studies, however, have
not shown similar correlations (24). Human milk
appears to have a beneficial effect by stimulating the
immune system of the breastfed infant (25).
A history of otitis media can contribute to a delay in
the development of language skills and other devel-
opment during the first 2 years of life (26). In the
present study, language and other developmental skills
as assessed using the Denver test were equal in each
therapy group at 18 months of age. This finding may
be comforting news to families with a child with RTA
problems. The age of 18 months, however, may be too
low for any final conclusions to be drawn.
CONCLUSIONS
RTA is best avoided by breastfeeding, day care at
home and emotional support for the families involved.
Our study shows that, in addition to medical indica-
tions, familial psychosocial factors are associated with
a child’s RTA. Health care personnel should give
support and be alert during a child’s respiratory
infections and pay particular attention to the role of
the father within the family.
REFERENCES
1. Newacheck PW, Taylor WR. Childhood chronic illness:
Prevalence, severity, and impact. Am J Public Health
1992;82:364
/71.
2. Alho OP. How common is recurrent acute otitis media?
Acta Otolaryngol 1997; Suppl 529:8 
/10.
3. Wuest J, Stern PN. The impact of fluctuating relationships
with the Canadian health care system on family manage-
ment of otitis media with effusion. J Adv Nurs 1990;15:556 
/
63.
4. Rautava P, Sillanpa
¨a
¨M. The Finnish Family Competence
Study: Knowledge of childbirth of nulliparous women seen
at maternity health care clinics. J Epidemiol Commun
Health 1989;43:253 
/60.
5. Howie VM, Ploussard JH, Sloyer J. The ‘‘Otitis-Prone’’
condition. Am J Dis Child 1975;129:676
/8.
Table IV. Univariate and multivariate stepwise logistic regression analyses of child-related factors associated with RTA (n/608).
Calculated using multinomial logistic regression analysis with zero course as reference class.
Univariate
analyses
Multivariate analysis
1
/5 courses ]/6 courses
Variable p p OR (95% CI) OR (95% CI)
Gaps in scheduled immunisations
1
0.005
A large number of paracentheses
1
B/0.001
Problems of skin care
1
0.006
Allergic symptoms
1
0.001
Special diet 0.007
Types of day care
2
B/0.001 B/0.001
Private families vs home 1.7 (1.1 
/2.6) 2.7 (1.4 /5.2)
Day care centre vs home 4.4 (2.1 
/9.3) 13 (5.1 /32)
Periods of hospital stay
1
B/0.001 0.002
Yes vs none 3.0 (1.1 
/8.0) 6.7 (2.2 /20)
Diagnosed allergy
2
0.001 0.032
Yes vs none 2.8 (1.1 
/6.9) 3.3 (1.1 /9.7)
Need for a special diet
1
0.003 0.016
Yes vs none 1.6 (0.9 
/2.9) 3.1 (1.4 /6.6)
Duration of breastfeeding measured
for every 3-month period
2
0.017 0.025
Yes vs none 1.0 (0.9 
/1.1) 0.8 (0.6 /0.9)
OR
/odds ratio, CI/confidence interval.
1
Data from the 3/9-month period.
2
Data from the 9/18-month period.
20 K. Louhi-Pirkanniemi et al.
Scand J Prim Health Care 2004; 22
Scand J Prim Health Care Downloaded from informahealthcare.com by 61.163.165.165 on 05/20/14
For personal use only.

6. Tainio VM, Savilahti E, Salmenpera
¨L, Arjomaa P, Siimes
MA, Perheentupa J. Risk factors for infantile recurrent
otitis media: Atopy but not type of feeding. Pediatr Res
1988;23:509 
/12.
7. Harsten G, Prellner K, Heldrup J, Kalm O, Kornfa
¨lt R.
Recurrent acute otitis media. A prospective study of
children during the first three years of life. Acta Otolaryngol
1989;107:111 
/9.
8. Frankenburg WK, Dodds JB. The Denver Developmental
Screening Test. J Pediatr 1967;71:181 
/91.
9. Frankenburg WK, Goldstein AD, Camp BW. The revised
Denver Development Screening Test: Its accuracy as a
screening instrument. J Pediatr 1971;79:988
/95.
10. Agresti A. Categorical data analysis. New York: John Wiley
& Sons, 1990.
11. Westbom L. Impact of chronic illness in children on
parental living conditions. A population-based study in a
Swedish primary care district. Scand J Prim Health Care
1992;10:83 
/90.
12. Darke PR, Goldberg S. Father 
/infant interaction and
parent stress with healthy and medically compromised
infants. Infant Behav Dev 1994;17:3 
/14.
13. National Agency for Medicines and the Social Insurance
Institution. The Finnish Statistics on Medicines 1987.
Helsinki, 1988.
14. National Agency for Medicines and the Social Insurance
Institution. The Finnish Statistics on Medicines 2001.
Helsinki, 2002.
15. Teele DW, Klein JO, Rosner B. Epidemiology ofotitis media
during the first seven years of life in children in greater
Boston: A prospective, cohort study. J Infect Dis
1989;160:83 
/94.
16. Alho O-P, Koivu M, Sorri M, Rantakallio P. The occur-
rence of acute otitis media in infants: A life-table analysis.
Int J Pediatr Otorhinolaryngol 1991;21:7 
/14.
17. Bexell A, Ra˚stam L, Isacsson S-O, Ingvarsson L. Parents’
response to recurrent middle ear infection in their children.
Scand J Soc Med 1990;18:25 
/30.
18. Midence K. The effects of chronic illness on children and
their families: An overview. Genet Soc Gen Psychol Monogr
1994;120:311 
/26.
19. Alho O-P, Koivu M, Sorri M, Rantakallio P. Risk
factors for recurrent acute otitis media and respiratory
infection in infancy. Int J Pediatr Otolaryngol 1990;19:
151
/61.
20. Collet JP, Burtin P, Kramer MS, Floret D, Bossard N,
Ducruet T. Type of day-care setting and risk of repeated
infections. Pediatrics 1994;94 Suppl:997 
/9.
21. Lanphear BP, Byrd RS, Auinger P, Hall CB. Increasing
prevalence of recurrent otitis media among children in the
United States. Pediatrics 1997;99:El.
22. Juntti H, Tikkanen S, Kokkonen J, Alho O-P, Niinima
¨ki A.
Cow’s milk allergy is associated with recurrent otitis media
during childhood. Acta Otolaryngol 1999;119:867
/73.
23. Fireman P. Otitis media and its relationship to allergy.
Pediatr Clin North Am 1988;35:1075 
/90.
24. Rubin DH, Leventhal JM, Krasilnikoff PA, Kuo HS, Jekel
JF, Weile B, et al. Relationship between infant feeding and
infectious illness: A prospectivestudy of infants during the
first year of life. Pediatrics 1990;85:464 
/71.
25. Hanson LA
˚, Korotkova M, Ha˚versen L, Mattsby-Baltzer I,
Hahn-Zoric M, Silferdal S-A, et al. Breast-feeding, a
complex support system for the offspring. Pediatrics
2002;44:347.
26. Feldman HM, Dollaghan CA, Campbell TF, Colborn DK,
Kurs-Lasky M, Janosky JE, et al. Parent-reported language
and communications skills at one and two years of age in
relation to otitis media in the first two years of life.
Pediatrics 1999;104:E52.
Recurrent antibiotic use in a small child 21
Scand J Prim Health Care 2004; 22
Scand J Prim Health Care Downloaded from informahealthcare.com by 61.163.165.165 on 05/20/14
For personal use only.