Babirye et al.

BMC Public Health 2011, 11:723

http://www.biomedcentral.com/1471-2458/11/723

RESEARCH ARTICLE Open Access

More support for mothers: a qualitative study on

factors affecting immunisation behaviour in
Kampala, Uganda
Juliet N Babirye1,2*, Elizeus Rutebemberwa1, Juliet Kiguli1, Henry Wamani1, Fred Nuwaha1 and
Ingunn MS Engebretsen2

Abstract
Background: The proportion of Ugandan children who are fully vaccinated has varied over the years.
Understanding vaccination behaviour is important for the success of the immunisation programme. This study
examined influences on immunisation behaviour using the attitude-social influence-self efficacy model.
Methods: We conducted nine focus group discussions (FGDs) with mothers and fathers. Eight key informant
interviews (KIIs) were held with those in charge of community mobilisation for immunisation, fathers and mothers.
Data was analysed using content analysis.
Results: Influences on the mother’s immunisation behaviour ranged from the non-supportive role of male partners
sometimes resulting into intimate partner violence, lack of presentable clothing which made mothers vulnerable to
bullying, inconvenient schedules and time constraints, to suspicion against immunisation such as vaccines cause
physical disability and/or death.
Conclusions: Immunisation programmes should position themselves to address social contexts. A community
programme that empowers women economically and helps men recognise the role of women in decision making
for child health is needed. Increasing male involvement and knowledge of immunisation concepts among
caretakers could improve immunisation.

Background between kin and friends affect health decision making

Each year, over 24 million children under one year of age [13]. Often the socio-cultural context relevant for health
miss routine immunisation services. Seventy percent of seeking behaviour is not considered during programme
these children live in ten countries including Uganda [1]. implementation.
Several studies identify demographic characteristics of Immunisation outreaches have been tailored to rural
the caretakers with suboptimal utilisation of immunisa- areas, where access to health services is low due to poor
tion services [2-9]. Others have identified health system road network. The rationale for this being the majority of
factors and behavioural influences on immunisation the population live in rural areas. A growing concern with
[10-12]. Those studies on behavioural factors have been rapid urbanisation is how to provide immunisation ser-
done in settings with almost universal access to immuni- vices. This study examined influences on childhood
sation services. immunisation behaviour using the attitude-social influ-
In low income settings, immunisation programmes ence-self efficacy model.
have traditionally targeted women and neglected the role
of men. Power relations within the household and Attitude-social influence-self efficacy model
The attitude-social influence-self efficacy model (shown
in figure 1) was originally developed by de Vries et al
* Correspondence: [email protected] for smoking cessation [14]. In the model presented in
1
School of Public Health, Makerere University College of Health Sciences, P.O.
Box 7072, Kampala Uganda this paper behaviour related to childhood immunisation
Full list of author information is available at the end of the article

© 2011 Babirye et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons
Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited.
Babirye et al. BMC Public Health 2011, 11:723 Page 2 of 11
http://www.biomedcentral.com/1471-2458/11/723

Socio-
Social influence
economic &
from important others
demographic
like spouse,
factors
mother in-law,
Age peers, etc. Barriers
Gender
Religion Self-efficacy
Education Intention Behaviour Child
Beliefs & ability to To immunise Children Survival
Occupation overcome barriers to
Marital status children immunised &
immunisation dropout rate
Infant sex activities
Abilities
Attitude
Towards childhood
immunisation and its
effects

Adapted from Sandvik 2009; Nuwaha 2002

Figure 1 Attitude-social influence-self efficacy model.

is the result of behaviour intention. This intention is in demographic variables are usually not easily changeable.
turn predicted by three main factors: social influence, On the other hand, the demographic features would be
self efficacy, and attitude [14-16]. A person’s attitude valuable in identifying individuals who fail to complete
refers to the extent to which a person has a favourable immunisation schedules or do not immunise their chil-
or unfavourable evaluation of the behaviour. A person’s dren [15-17].
attitude towards childhood immunisation may be influ-
enced by personal beliefs such as misconceptions asso- Methods
ciated with immunisation of children, and by the fear Study area
associated with side effects from vaccines. This fear is a The study was conducted in Kampala from June to
barrier to optimal utilization of immunisation services. September 2010. Kampala is the largest urban area and
Social influence results from social norms related to capital city of Uganda with an average population density
immunisation of children and the support from impor- of more than 7400 persons per square kilometre and a
tant others like the partner or the mother. Self efficacy total population of about 1.6 million people. Children
refers to a person’s ability to cope with barriers that may below 5 years constitute 20% of the total population.
hinder adherence to recommended immunisation sche- Immunisation coverage from surveys in Kampala show
dules. A low perceived benefit of immunisation would BCG coverage at 91%, combined pertusis vaccine at 68%,
reduce the ability to cope with the barriers to immunisa- Polio3 at 56%, measles at 71% and those who receive all
tion services. Self efficacy not only influences behaviour expanded immunisation programme (EPI) vaccines at
intention but also directly influences behaviour. Barriers 47% [18,19]. These rates are all below the national
and abilities could influence behaviour related to immu- targets.
nisation activities. Previous behaviour or trying to per- Health services in Kampala are provided by govern-
form the behaviour has a feedback mechanism that in ment, non-governmental organization (NGO), and pri-
turn influences the attitude, social influence and self vately owned health facilities. All government and NGO
efficacy. facilities provide routine immunisation services in addi-
Thus this model infers that attitude, social influence, tion to outreach immunisation services.
and self efficacy variables can be targeted through health There are five divisions in Kampala and each division
promoting activities for improving immunisation cover- is administratively semi autonomous with a separate
age in addition to reducing delay in immunisations, work plan and budget. Three of the divisions, Central,
whereas external variables like socio-economic and Kawempe, and Rubaga, are better served with public
Babirye et al. BMC Public Health 2011, 11:723 Page 3 of 11
http://www.biomedcentral.com/1471-2458/11/723

health facilities. Nakawa and Makindye divisions are that used a Luo speaking translator. The interviews were
relatively least served by public health facilities and moderated by one female social scientist and note taking
since most immunisation services are provided by public was by one male educated in health promotion. Both had
health facilities Nakawa and Makindye divisions are in experience in conducting FGDs, and were both fluent in
great need for improved services. This study was con- Luganda and English languages. JNB (first author) made
ducted in Nakawa and Makindye divisions of Kampala. observations at all discussions and asked questions to
clarify some of the issues raised in the FGDs.
Data collection and study population Eight key informant interviews (KIIs) were conducted in
The main data collection methods included focus group the local language, Luganda. Two of the KIIs were held
discussions (FGDs) and key informant interviews (KII). with those in charge of community mobilisation for
FGDs which are described as appropriate arenas for immunisation, three with fathers and three with mothers
discussing and airing multiple views on an issue were in the community. The KIIs were conducted by an experi-
chosen as one of the data collection methods [20]. Nine enced research assistant (health promotion). Data was
FGDs were conducted with three different categories of reviewed by JNB and the research assistant after each
respondents; three with mothers aged 18-25 years interview to assess how the questions were being answered
(referred to as ‘FGD with younger mothers’), four with before conducting the next KII.
mothers older than 25 years (referred to as ‘FGD with JNB was the principal investigator with medical back-
older mothers’) and two with fathers (referred to as ‘FGD ground and fluent in English and Luganda. She was
with fathers’). Older women were chosen because they trained in qualitative methods and had prior experience
play a significant role in society and are looked up to by with qualitative field work in other health related fields.
younger parents for advice. FGDs of fathers were con- She discussed the experiences of the interviewer and note
ducted because fathers also play a significant role in taker and whether questions were being understood by
childhood immunisation, especially when costs are to be FGD participants. The FGDs and KIIs were conducted
incurred [13,21,22]. With the assumption that men and using a guide that focused on beliefs, perceptions, experi-
women tell different stories, and that adults in different ences, actions and consequences from immunisation activ-
age groups feel more like peers within their respective ities. We asked questions such as; the reasons why
age group, different FGDs were constituted to give eva- participants took their children for immunisation, what
luation of consensus and contestation of information barriers they faced during the process, and how they over-
[20,23]. The local council leaders mobilised mothers and came these barriers. Reasons for refusal of immunisation
fathers that had children younger than five years, older were also explored.
women and men. Local council leaders are village heads The numbers of FGDs/KIIs were deemed sufficient
and lead a population of approximately 1000 adults. The when additional interviews yielded little new informa-
entire executive at the village level are nominated and tion on the core study objectives.
voted into these positions by the adults that constitute a
village. Our study team explained the inclusion criteria to Data analysis
the local council leader and to the secretary for women All the data were tape recorded after obtaining partici-
on the village committee. A later date was set for the pants’ consent. The audio data was transcribed and later
FGD. Either the local council chairman (for male FGDs) translated from the local language into English by the
or the secretary for women (for female FGDs) chose one moderators while for two FGDs (one conducted in English
participant per household. The households from which and another that used a translator) were transcribed
FGD participants were selected were geographically scat- directly into English. Audio records and local language ter-
tered across the village and not clustered in same locality. minologies were kept for consistence checks and compiled
Those mobilising participants were encouraged to bring with field notes. The researchers listened to the audio
up to 25 eligible participants. On the day of the FGD, the recordings to confirm the information on the transcripts.
study team together with the local leaders or secretary Audio recording enabled details of the KIIs and FGDs to
for women then chose individuals who would participate be obtained with accuracy that would not be got from the
in the discussion. FGDs were conducted in one of the field notes or from memory alone. Tape recording also
homes of the FGD participants (five FGDs) or inside the allowed more eye contact between the moderators and the
community centres (four FGDs). There were a total of 73 respondents [24]. The unit of analysis was the transcripts
participants for FGDs; 58 were women. Each FGD had from FGDs and KIIs. The authors JNB, ER, JK and IMSE
between six to eleven participants per group discussion. read through the transcripts and came up with meaning
All discussions lasted 1-2 hours and were conducted in units individually. They harmonised their meaning units
Luganda (local language) except for one FGD that used and went back and separately coded the meaning units.
English as the medium of communication and one FGD These codes were again shared between the four authors.
Babirye et al. BMC Public Health 2011, 11:723 Page 4 of 11
http://www.biomedcentral.com/1471-2458/11/723

During the discussion, the codes were merged into cate- ‘As for me, I make sure that when my wife is preg-
gories and subsequently into themes which were shared nant she attends the antenatal clinic as required and
with the other co-authors [25]. The themes were grouped is also immunised because she usually tells me when
and presented according to the attitude, social influence she is immunised. Also after she gives birth I make
and self efficacy model. The different data sources sure she takes the children for immunisation on the
informed each other during design, implementation and dates written on the immunisation card.’ (FGD with
analysis, thus the data were triangulated during the entire fathers)
research process [26].
A minority of male participants rejected immunisation
Ethics approval however and therefore hindered their wives from immu-
Ethics approval was obtained from Makerere University nising their children.
School of Public Health Higher Degrees Research and
Ethics Committee (IRB00005876FWA/Protocol 085) and ‘My wife is pregnant but she has not been immu-
independently from the Uganda National Council for nised. She has a four year old child and she talks
Science and Technology (HS 786). The interviews were about immunising the child but I stop her from
conducted only after informed consent was obtained doing it. For me I don’t believe in it. As you can see I
from the study participants. am a mature person but I did not grow up because
of that (immunisation). It was better for me to use
Results traditional medicine to treat fever for example, but
The major findings of this study were that; convictions of because these days the fever is very strong I now use
the caretakers, self efficacy, and the supportive or non- tablets (for treatment). Even these injections (from
supportive role of significant others influenced the invol- immunisation) paralyze people I know, and we also
vement or non-involvement of parents in childhood see them in books and in pictures.’ (FGD with
immunisation (see table 1). In the next section we present fathers)
two main sub-themes under social influence namely; influ-
ence from the male partner and influence from the older If a father disagreed to immunisation, the mothers
generation or peers on immunisation behaviour. The expressed less power for decision making at the house-
theme on self efficacy presents data on barriers faced hold level which made them unable to take their children
during utilisation of immunisation services and the partici- for immunisation. They said that it was the man’s prero-
pants’ ability to overcome these namely; barriers associated gative to make the decision to immunise the child. So if
with access to immunisation services and those that arose the father of the child stopped them from taking children
from personal challenges. Attitudinal factors are cate- for immunisation then they would not immunise their
gorised into three main sub-themes: trust in immunisa- children. The women who felt this way were mainly
tion, fear of side effects, and programmatic preferences. young and with lower level of education. They believed
they should submit to the men’s directives at all times.
Social influence
Supportive kin recognising the benefits of immunisation ‘Because the wife fears the husband, if you give me
were essential to the child being immunised. In this study instructions never to leave the home, can I leave it? I
setting the decision to go for immunisation was generally will have broken a rule.’ (FGD with younger
a joint decision between the mother and father of the mothers)
child. Most study participants (FGDs and KIIs) strongly
emphasized that both the child’s mother and father were If a child fell sick it would be the mothers who would
responsible for the immunisation of the child, but in rea- spend ‘sleepless nights’ while the men slept or went to
lity, only women were in charge of taking children for work. Some women, mostly in FGDs with older mothers,
immunisation. Below, we present the influences on the therefore made decisions to immunise their children
mother’s behaviour from her social context, first her male despite opposition and threats from their husbands, and
partner, second the older generation such as mothers-in- they stressed that this has to be done with determination.
law, fathers-in-law, sisters-in-law, and lastly her peers.
Most women expressed support from their partners ‘Also, it is the mother who should really make sure
when taking the child for immunisation, such as money your child is immunised. If you follow the man’s
for transport and a granted permission to take the child advice and you don’t immunise your child, when
for immunisation. This support was reiterated by male that child falls sick it is you the mother who will
participants; both KII and FGD participants. spend sleepless nights when the child is sick. He will
Babirye et al. BMC Public Health 2011, 11:723 Page 5 of 11
http://www.biomedcentral.com/1471-2458/11/723

Table 1 Emerging themes-direction of influence on immunisation behaviour

Themes Data source (FGDs, KIIs) Influence on utilisation of
immunisation services
Social influence
Influence from male partner
Father’s offer support in form of transport to immunisation unit All Positive
Father’s make the decision to immunise Female FGDs Both positive and negative
Joint decision to immunise or not All (emphasized more among male Positive
FGDs)
Less power for decision making Female FGDs Negative influence for women
Influence from older generation and peers
Decision making involves kin All Mostly negative
Decision to immunise against spousal consent encouraged by Female FGDs (older women) and Positive
experienced mothers female key informant
Intimate partner violence Female FGDs Negative
Social stigma against teenage mothers Female KII Negative

Self-efficacy
Difficult to access immunisation unit All Negative
Vaccines sometimes out of stock All Negative
Have to choose between money for transport and food All FGDs Negative
Lack of presentable clothing Female FGDs Negative
Gender roles All Negative for the men
Lack of job security leading to choice between work and All Negative
immunisation

Attitude

Trust in immunisation
Believe immunisation is beneficial All Positive

Perception of lack of trust in immunisation programme/immunisation All Negative

is harmful to the child’s health

Fear of vaccine side effects leading to drop out or delayed Mostly female FGDs Negative
immunisation

Programmatic preferences
Preference for routine immunisation services All Positive
Reasons for preference of routine services:
Health workers take responsibility if complications develop after All Positive
immunisation
Routine services have a fixed address All Positive
Forceful methods of conducting mass immunisation exercises All Negative

be snoring and the doctors will abuse you as he is key informant told how she pretended she was going to
not around the hospital. Yet you followed his advice. the market, but went to the nearby outreach centre for
You the mother have to stick to your guns. Let him immunisation. This ‘rebellious’ behaviour had conse-
fight with you, but after your child has been immu- quences such as intimate partner violence which included
nised.’ (FGD with older mothers) emotional, verbal and physical violence. The violence
after immunisation was experienced by a minority of
Some women who opposed their husbands’ decision women both in the FGDs and KIIs. However, all female
not to immunise reported that they had to be discreet participants in the FGDs and KIIs reported it since they
about the whole process of immunisation. One female had witnessed or heard about this occurrence. An older
Babirye et al. BMC Public Health 2011, 11:723 Page 6 of 11
http://www.biomedcentral.com/1471-2458/11/723

mother in tears reported an incident after her child got young women. This was highlighted by one of the female
the first injection on the thigh, ‘the baby cried all night’. key informants in charge of community mobilisation. She
Her husband sent her and the baby out of the house in reported that teenage mothers were stigmatized by the
the night saying it was her decision to immunise the older women who laughed at them for giving birth at an
child and he could not tolerate the noise. ‘early age’. In addition some teenage mothers were told by
Even amidst spousal violence some women derived the older women in their community that the fathers of
satisfaction from the fact that their children were immu- their children were HIV positive. These younger mothers
nised particularly the older women that had disobeyed therefore stayed away from all immunisation activities
the husband’s instructions and they intended to take where it was possible to meet these older women from
their children for subsequent immunisations. whom they faced social stigma.
Only women who were convinced about the benefits Female participants felt that mostly men had a non-
of immunisation were willing to endure the conse- conforming attitude towards childhood immunisation. So
quences of opposing their spouses. they described these men as having ‘weak brains’, lazy, and
irresponsible. This was also reflected among male partici-
‘The mother will say, “Let me immunise my children pant’s judgement against those that did not immunise
for their good because when my child is disabled, my children: terms like they are ’ignorant’ or ’uneducated’
husband can have other children with another were used to describe them. The female respondents espe-
woman. It is me to suffer with my children who would cially felt that this non-conforming attitude should have
have helped me in future.” (Female key informant) consequences. They stressed that individuals in this cate-
gory should be given some form of punishment by the
It was reported by participants in FGDs and KIIs that government. Some male respondents were in agreement
if the father was against immunisation it was also com- with the women because they reasoned that these indivi-
mon to be influenced from his elder relatives and perso- duals were cruel to the innocent children whose future
nal experiences. they were ’sabotaging’ or ’ruining’ by refusing immunisa-
tion. Other men strongly opposed the idea of punishment
‘Like I explained before about some elderly women however arguing that it would be difficult to identify such
who claim children will become lame after immuni- individuals in the community.
sation, some men use that excuse because they had In general, both FGDs and KIIs supported that the
ever heard of it while still young. So when they grow mother was under strong social influence affecting deci-
up and get children they say the children will become sion making on immunisation.
lame or get brain damage. That is why you see some
children when they get measles they almost die Self -efficacy
because the husband refused the wife to take children Not only the mother’s social context influenced immu-
for immunisation.’ (FGD with older mothers) nisation behaviour, but also her own ability to overcome
barriers, defined as self-efficacy, affected behaviour.
The men who disagreed to immunisation were put Major hindrances reported included financial depriva-
under pressure by his elderly relatives and they had tion which made the cost of going for immunisation a
strong union. considerable decision to make.

‘Like for the old people who have previously heard that ’If I don’t have food, how can I use Uganda shillings
children died, when you tell them that you are taking a 2000 (approximately US$1) for a boda-boda (means
child for immunisation, they will not like it. No, his of transport using motorcycle/bicycle) to go for
father will tell him that why did you let her take the immunisation?’ (FGD with younger mothers)
child for immunisation. And he will answer that I
refused her but she insisted.’ (FGD with older mothers) With lack of money, walking could be the only alter-
native with distances of up to 4 km and having to cross
The older generation exerted influence on the mother’s two motor highways in some instances. This was
behaviour indirectly through the husband as shown above, reported from all FGD participants as a major challenge
and also directly. For instance, the younger mothers were especially for women in the post-partum period if they
persuaded to take their children for immunisation against needed to take their children for immunisation. This
spousal consent by older experienced women in their challenge was compounded by frequent reports by
neighbourhood. They were supported in breaking some mothers and fathers that they were not given the antici-
household rules to protect the wellbeing of their children. pated services due to vaccines being out of stock or due
However, older women were sometimes not supportive of to absent health workers.
Babirye et al. BMC Public Health 2011, 11:723 Page 7 of 11
http://www.biomedcentral.com/1471-2458/11/723

Another expense for the mothers was not only finan- physical disability or death. These diseases were per-
cial, but also related to time. Vaccination could easily ceived by most FGD participants and key informants as
take one day and they would have lost the potential common and the children as vulnerable to get the dis-
income for that day. eases unless immunised.

‘The nature of work for some people at times makes ‘When the child gets measles, he will not be bedrid-
them miss these immunisation schedules since some- den. He will just get a rash or cough. He may also
body leaves home at around 6.00 a.m. and comes get red eyes or mouth rash but he will be able to
back at around 6.00 p.m. During the day this person play as usual. But if he was not immunised he will
is at the stone quarry (work place) about 2 kms from get very high temperatures, fever, diaorrhea and you
here, now the person will not take the child for become worried the child may even die.’ (FGD with
immunisation although the person will be willing to younger mothers)
take the child.’ (FGD with fathers)
The societal value of having a healthy child population
It was reported from Female FGDs especially that was strongly held among most study participants. If
poor mothers often felt stigmatised and bullied from their children survived vaccine preventable diseases they
other women and health workers if they did not show could contribute to building a strong society and
up in good clothing or with presentable clothes or shawl become ’doctors and teachers who would be able to treat
for their children. or teach the population.’
The fear of perceived ill effects of immunisation
‘Some young women fear going for immunisation underpinned the strong belief against immunisation. All
because they don’t have a baby shawl for carrying study participants perceived that a lack of trust towards
the children to hospital so when you reach at the vaccines existed among community members. Common
hospital with some sheets which are not clean some beliefs were that vaccines were ‘expired’ and could cause
nurses will sometimes begin abusing you.’ (FGD with ‘physical disability and/or death’ among their children.
older mothers) The perceived susceptibility of their children to suffer
from severe effects of the vaccines led some to decline
Gender roles were perceived as a barrier to male immunisation.
involvement in child immunisation activities and it was
hard for the men to overcome these barriers. They ‘At one time our neighbour in ’rural geographical
would ‘feel out of place’ at the immunisation centres as area’ immunised a child in the morning and by 5.00
it was considered a ‘female arena.’ Even if the men were p.m. the child was dead. From that time I fear taking
willing to take their children for immunisation they did children for immunisation and all my children are
not have time to do this because they had to go for not immunised.’ (FGD with fathers)
work. This competing demand for time was emphasised
in all FGDs and supported by the KIIs as an important A lack of trust was also observed against the health
barrier to immunisation activities. Lack of job security personnel believed not to check the drugs properly and
and high unemployment rates forced parents to serve only give ’expired’ vaccines which might cause disability
their employer if they were on private ad-hoc or longer or death.
contracts at the expense of personal activities such as Vaccine side effects
taking the child for immunisation. Among those who accepted the benefits of immunisa-
tion, side effects were recognised as a constraint. This
Attitudinal factors fear of vaccine side effects was more commonly held
The convictions of the respondents towards childhood among female than male respondents. Many had experi-
immunisation were classified into three sub-themes: 1) enced or were afraid of vaccine side effects such as
trust in immunisation 2) fear of vaccine side effects 3) fever, temporary ’paralysis of the leg’ and excessive cry-
programmatic preferences. ing after the ’first injection given on the thigh’. The con-
Trust in immunisation sequences were either declining or delaying subsequent
There were two opposing beliefs among our study parti- immunisations.
cipants: those who trusted in immunisation as a child
survival strategy and those who feared or refused immu- ‘Sometimes after immunisation children get fever and
nisation. Those who trusted vaccines were generally bet- spend the whole night crying so the health worker
ter educated and older. They recognised the diseases must tell the mother in advance what will happen to
that could cause severe outcomes in children such as the baby, “that the baby might become weak, or get a
Babirye et al. BMC Public Health 2011, 11:723 Page 8 of 11
http://www.biomedcentral.com/1471-2458/11/723

fever or the injection is painful so he will cry a lot.” Discussion

But some health workers don’t warn the parents so Significant others particularly the male partner played
when they reach home the mother will notice the an important role on the mother’s decision making that
child has got a fever or is crying uncontrollably. And could be prohibitive for immunisation. A few of the
this makes her worried.’ (Male key informant) husbands held an opposing view to immunisation and
stopped their wives from immunising their children.
Home remedies against side effects were frequently Lack of clothing, money for transport, or time influ-
reported. Some treated the child with paracetamol or enced whether parents took their children for immuni-
‘junior aspirin.’ However many female FGD participants sation or not. Most respondents expressed their trust in
said that the side effects were persistent even after drugs the benefits of immunisation but had reservations
were tried out, so they applied ice, cold water, onions, towards the associated side effects. Others expressed a
oranges, or ‘black’ shoe polish at the injection site. They total distrust of immunisation programmes and vaccines.
explained that when the child was injected on the thigh, Those with less education showed more scepticism
the vaccine remained ’stationary’ at the injection site towards immunisation. Study respondents also preferred
and that is why the child suffered. When applying the routine to mass immunisation activities. Our findings
“black shoe polish” or any other home remedy the vac- create deeper understanding and complement quantita-
cine ’was moved’ (absorbed) and the child got some tive studies highlighting associations between immunisa-
relief. tion behaviour and maternal education, wealth, age and
Programmatic preferences parity [3,4,6-9].
Routine immunisation was distinguished from mass In our study, husbands and other kin sometimes hin-
immunisation services. Routine immunisation services dered the mother from immunising her child. A qualita-
involve individually scheduled immunisations according to tive study from Hong Kong reported that most
the expanded programme for immunisation. Caretakers participants did not receive advice from family members
then have to bring the child to the immunisation units. although siblings and peers encouraged participants to
On the other hand mass immunisation is the distribution take more immunisations than what was provided in the
of one particular or a group of vaccines meant for every- EPI programme [5]. The difference in study findings from
body within a certain age range in the area. It is meant to our study could be due to better access to technology and
compensate for not achieving full immunisation coverage health information, and due to mandatory immunisation
by including those who have got and those who would for school entry in Hong Kong. In developing countries
have missed the routine immunisation. Mass immunisa- such as Uganda, decision making in health seeking beha-
tion activities mostly promote measles as well as polio viour is a very complex process, and earlier studies
vaccination among children under five years and are done strongly emphasize the social dimensions [13]. Newer
as community outreaches, in schools, and in conjunction research focusing on the influence of the husband on the
with religious groups like churches. Among study respon- mother’s health behaviour show that existing organisation
dents that accepted the benefits of immunisation, most of services and lack of contextualisation in health pro-
preferred routine compared to mass immunisations grammes causes reduced male involvement. Lack of male
because they believed that routine immunisation was involvement is prohibitive for successful maternal and
’safe’. Several reasons were fronted for preferring routine child health programmes [27-29].
services: first, if the child developed complications after Uganda as a patrilineal society directly increases the
immunisation it would be easier to trace the health work- man’s power to inhibit child immunisation, for instance,
ers conducting routine services, which was not the case in he can stop the mother from spending money on trans-
mass immunisation services. The health worker could port for immunisation. In our study, the younger mothers
’take responsibility’ during routine services compared to particularly reported being disempowered economically
mass immunisation services which lacked a ’fixed physical and at decision making at the household level sometimes
address.’ Second, it was difficult for the parents to under- resulting into intimate partner violence. This supports the
stand that there was need to take their children for further literature on widespread intimate partner violence in
immunisations when they had completed the routine Ugandan communities [30,31]. The violence is often a
immunisation schedule. Only a few study participants had result of disturbing social relations. Social relations are
voluntarily taken their children for mass immunisation particularly affected when illness or death occurs, and
activities. The misgivings towards mass immunisation quarrels and blaming may arise placing guilt on partners.
were worsened by the reported methods used to conduct Thus, a fear of disturbing the social equilibrium greatly
the mass immunisation including officials forcefully influences the health decision making processes [13]. Yet
immunising the children. the health system has traditionally targeted mothers
Babirye et al. BMC Public Health 2011, 11:723 Page 9 of 11
http://www.biomedcentral.com/1471-2458/11/723

because they spend more time with the children and are Methodological considerations
caretakers of the sick [21,22]. It is important therefore to In this study triangulation was achieved by using different
carefully consider the social contexts during programme study methods and respondents, and by having research-
design and implementation for child immunisation. ers from different backgrounds, from social science and
Respondents in our study expressed fear of child death medicine. Although qualitative methods do not give the
or physical disability and these were categorised into two: magnitude and variations across the different categories
for and against immunisation. Either category was moti- of the respondents, the triangulation in this study aimed
vated by a preference for a state free of illness [13]. at reaching an objective view of the data and was useful
Those that were against immunisation wanted a state to check the consistency and contradictions across and
free of disability or illness from the vaccines. Those that within groups [26,33]. However, the following study lim-
believed in immunisation wanted a state free of illness itations need to be considered in the interpretation of
caused by the vaccine preventable diseases. In the Hong these results: first, there was a potential for the local
Kong study mentioned earlier, participants’ decision to council leaders to omit eligible respondents and to
immunise their children was often the result of weighing choose individuals that were close to them. This was
the benefits against the risks and most felt that the bene- mitigated by explaining the criteria for study inclusion to
fits outweighed the risks [5]. Another study showed that the leaders. These leaders mobilised potential partici-
personal experiences, value systems and level of trust in pants which tended to be large groups of participants of
health professionals were fundamental to parental deci- up to 25 study participants. The research team together
sion making about immunisation even when challenged with the local leader then selected FGD participants from
by anti-immunisation messages [10]. The ones that cause this large group. Second, JNB’s presence (a medical doc-
most concern are those that are against immunisation tor) could have led to eliciting socially desirable answers
due to perceived health risks from vaccines. Their con- in the discussions. This was mitigated by assuring the
cerns should be addressed systematically by the health respondents of confidentiality. However, participants’ cri-
workers since inefficient response to public debates ticisms of immunisation indicate that the potential bias
about vaccines has the potential to be amplified and to arising from this limitation was greatly reduced. Third,
significantly reduce immunisation coverage as was seen the use of a translator for one of the FGDs may have led
in the case of measles-mumps-rubella vaccine and autism to loss of depth of the issues being discussed. However
[32]. the similarity of the themes with the other FGDs is testi-
Even among study participants that accepted the bene- mony that the key findings were similar. In addition, JNB
fits of immunisation in our study, side effects were recog- who was present at this FGD asked whether key issues
nised as a constraint. The reported behaviour of the were being probed. Fourth, there was a tendency in this
mothers after experiencing side effects indicates that the study setting for participants to sometimes speak about
health workers need to strengthen the strategy of increas- their experiences in the third person. Therefore it was
ing immunisation coverage. High levels of immunisation sometimes difficult to decipher whether individuals were
can be achieved through use of information, education, speaking about themselves or other people. This was
and persuasion on immunisation concepts and the ratio- mitigated during data analysis through taking quotes
nale for immunisation rather than use only images of within their context using an FGD or KII as a unit of ana-
children suffering from vaccine preventable diseases or lysis. In addition, the quotes presented here were repre-
historic images relating to outbreaks [32]. sentative of ideas which were common across data
Behavioural or social science theories provide the basis sources and in other settings these would be presented in
for understanding health behaviour. Several models have the first person.
been utilized in this sense; such as the health belief model,
theory of planned behaviour, the theory of reasoned Conclusions
action, the attitude-social influence-self-efficacy (ASE) The non-supportive role of husbands played an impor-
model. Although the ASE model resembles the theory of tant role on the mother’s decision making that could hin-
planned behaviour, it has evolved as a separate model, der utilisation of immunisation. In addition, gender roles
with a different methodological nature [15,16]. The ASE in this setting were prohibitive for male involvement in
model was better suited and provided a useful framework immunisation services. Lack of clothing, money for trans-
to analyse the factors associated with immunisation beha- port or competition between work and child immunisa-
viour in our study because it addresses self efficacy. Self tion influenced whether parents took their children for
efficacy is better suited to dyadic behaviour such as taking immunisation or not. Distrust of immunisation pro-
the child for immunisation than volitional control which is grammes and vaccines was expressed. Study respondents
assumed by the theory of reasoned action. also preferred routine to mass immunisation activities.
Babirye et al. BMC Public Health 2011, 11:723 Page 10 of 11
http://www.biomedcentral.com/1471-2458/11/723

Immunisation programmes in developing countries have 3. Drain PK, Ralaivao JS, Rakotonandrasana A, Carnell MA: Introducing auto-
disable syringes to the national immunization programme in
traditionally employed interventions that target mothers Madagascar. Bull World Health Organ 2003, 81:553-560.
to improve immunisation coverage. These interventions 4. Arevshatian L, Clements C, Lwanga S, Misore A, Ndumbe P, Seward J,
need to be re-designed in order to address socio-cultural Taylor P: An evaluation of infant immunization in Africa: is a
transformation in progress? Bull World Health Organ 2007, 85:449-457.
contexts, involve the community and not only target 5. Tarrant M, Thomson N: Secrets to success: a qualitative study of
mothers for childhood immunisation. There is need for a perceptions of childhood immunisations in a highly immunised
multidisciplinary community programme that empowers population. J Paediatr Child Health 2008, 44:541-547.
6. Rwashana AS, Williams DW, Neema S: System dynamics approach to
women economically and in decision making at the immunization healthcare issues in developing countries: a case study of
household level, and helps men recognise the role of Uganda. Health Informatics J 2009, 15:95-107.
women in decision making for child health. Further 7. Nankabirwa V, Tylleskar T, Tumwine JK, Sommerfelt H: Maternal education
is associated with vaccination status of infants less than 6 months in
research identifying effective and innovative strategies for Eastern Uganda: a cohort study. BMC Pediatr 2010, 10:92.
increasing male involvement in immunisation services in 8. Fadnes LT, Nankabirwa V, Sommerfelt H, Tylleskar T, Tumwine JK,
this setting is essential. A general need for increased Engebretsen IM: Is vaccination coverage a good indicator of age-
appropriate vaccination? A prospective study from Uganda. Vaccine
understanding on management of vaccine side effects, 2011, 29:3564-3570.
the rationale for immunisation and mass immunisation 9. Mutua MK, Kimani-Murage E, Ettarh RR: Childhood vaccination in informal
were observed in this study. Health education aiming at urban settlements in Nairobi, Kenya: who gets vaccinated? BMC Public
Health 2011, 11:6.
informing rather than creating fear of vaccine preventa- 10. Leask J, Chapman S, Hawe P, Burgess M: What maintains parental support
ble diseases could empower parents to make informed for vaccination when challenged by anti-vaccination messages? A
decisions and would increase male support for their part- qualitative study. Vaccine 2006, 24:7238-7245.
11. Odusanya OO, Alufohai EF, Meurice FP, Ahonkhai VI: Determinants of
ners/spouses. Health workers require additional in-ser- vaccination coverage in rural Nigeria. BMC Public Health 2008, 8:381.
vice training on medical ethics to promote the culture of 12. Wallace C, Leask J, Trevena LJ: Effects of a web based decision aid on
treating people with dignity and respect. In addition, parental attitudes to MMR vaccination: a before and after study. BMJ
2006, 332:146-149.
effective communication skills should be strengthened to 13. Igun UA: Stages in health-seeking: a descriptive model. Soc Sci Med 1979,
address the various issues which could potentially reduce 13A:445-456.
immunisation coverage such as health worker’s response 14. Hein de Vries MD, Kuhlman Piet: Self-efficacy: the third factor besides
attitude and subjective norm as a predictor of behavioural intentions.
to mother’s appearance, the fear of vaccines and vaccine Health Educ Res 1988, 3:273-282.
side effects. 15. Nuwaha F, Faxelid E, Wabwire-Mangen F, Eriksson C, Hojer B: Psycho-social
determinants for sexual partner referral in Uganda: quantitative results.
Soc Sci Med 2001, 53:1287-1301.
Acknowledgements and funding 16. Sandvik C, Gjestad R, Samdal O, Brug J, Klepp KI: Does socio-economic
We thank the study participants, local council leaders, the research assistants, status moderate the associations between psychosocial predictors and
and the management of the immunisation programme in Kampala. This fruit intake in schoolchildren? The Pro Children study. Health Educ Res
study was part of the OMEVAC project NFR no. 185777 (GlobVac 2010, 25:121-134.
programme: The programme for Global Health and Vaccination Research) 17. Nuwaha F, Kabatesi D, Muganwa M, Whalen CC: Factors influencing
funded by the Research Council of Norway. acceptability of voluntary counselling and testing for HIV in Bushenyi
district of Uganda. East Afr Med J 2002, 79:626-632.
Author details 18. UBOS: Uganda Bureau of Statistics (UBOS) and Macro International Inc:
1 Uganda Demographic and Health Survey Calverton, Maryland, USA; 2006.
School of Public Health, Makerere University College of Health Sciences, P.O.
Box 7072, Kampala Uganda. 2Centre for International Health, University of 19. Uganda government: Immunisation practise in Uganda: a manual for
Bergen, Bergen, Norway. operational level health workers. 2 edition. Kampala; 2007.
20. Patton MQ: Qualitative Research & Evaluation Methods. 3 edition. London:
Authors’ contributions Sage Publications; 2002.
JNB, IMSE contributed to the conception and design of the study, data 21. Deeney K, Lohan M, Parkes J, Spence D: Experiences of fathers of babies
analysis, interpretation of data and in drafting the paper. FN contributed to in intensive care. Paediatr Nurs 2009, 21:45-47.
the conception and design of the study; interpretation of data; and in 22. Demmer C: Experiences of women who have lost young children to
drafting the paper. ER, JK contributed to data analysis, interpretation of data AIDS in KwaZulu-Natal, South Africa: a qualitative study. J Int AIDS Soc
and in drafting the paper. HW contributed to interpretation and drafting of 2010, 13:50.
the manuscript. All authors approved the final manuscript. 23. Rice PLED: Qualitative Research Methods - A Health Focus Oxford, New York:
Oxford University Press; 1999.
Competing interests 24. Holstein JGJ: The Active Interview Thousand Oaks: Sage; 1995.
The authors declare that they have no competing interests. 25. Graneheim UH, Lundman B: Qualitative content analysis in nursing
research: concepts, procedures and measures to achieve
Received: 17 May 2011 Accepted: 25 September 2011 trustworthiness. Nurse Educ Today 2004, 24:105-112.
Published: 25 September 2011 26. Flick U: Triangulation revisited: Strategy of Validation or alternative?
Journal for the Theory of Social Behaviour 1992, 22:175-197.
27. Byamugisha R, Tumwine JK, Semiyaga N, Tylleskar T: Determinants of male
References
involvement in the prevention of mother-to-child transmission of HIV
1. World Health Organisation: Global Immunization Data. Annual report
programme in Eastern Uganda: a cross-sectional survey. Reprod Health
Geneva; 2009.
2010, 7:12.
2. Tarrant M, Gregory D: Exploring childhood immunization uptake with
28. Engebretsen IM, Moland KM, Nankunda J, Karamagi CA, Tylleskar T,
First Nations mothers in north-western Ontario, Canada. J Adv Nurs 2003,
Tumwine JK: Gendered perceptions on infant feeding in Eastern Uganda:
41:63-72.
Babirye et al. BMC Public Health 2011, 11:723 Page 11 of 11
http://www.biomedcentral.com/1471-2458/11/723

continued need for exclusive breastfeeding support. Int Breastfeed J 2010,

5:13.
29. Falnes EF, Moland KM, Tylleskar T, de Paoli MM, Msuya SE, Engebretsen IM:
“It’s her responsibility": partner involvement in prevention of mother to
child transmission of HIV programmes, northern Tanzania. J Int AIDS Soc
2011, 14:21.
30. Karamagi CA, Tumwine JK, Tylleskar T, Heggenhougen K: Intimate partner
violence and infant morbidity: evidence of an association from a
population-based study in eastern Uganda in 2003. BMC Pediatr 2007,
7:34.
31. Speizer IS: Intimate partner violence attitudes and experience among
women and men in Uganda. J Interpers Violence 2010, 25:1224-1241.
32. Burgess DC, Burgess MA, Leask J: The MMR vaccination and autism
controversy in United Kingdom 1998-2005: inevitable community
outrage or a failure of risk communication? Vaccine 2006, 24:3921-3928.
33. Rutebemberwa E, Ekirapa-Kiracho E, Okui O, Walker D, Mutebi A, Pariyo G:
Lack of effective communication between communities and hospitals in
Uganda: a qualitative exploration of missing links. BMC Health Serv Res
2009, 9:146.

Pre-publication history
The pre-publication history for this paper can be accessed here:
http://www.biomedcentral.com/1471-2458/11/723/prepub

doi:10.1186/1471-2458-11-723
Cite this article as: Babirye et al.: More support for mothers: a
qualitative study on factors affecting immunisation behaviour in
Kampala, Uganda. BMC Public Health 2011 11:723.

Submit your next manuscript to BioMed Central

and take full advantage of:

• Convenient online submission

• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution

Submit your manuscript at

www.biomedcentral.com/submit