BMC Medical Research Methodology 2008, 8:45
http://www.biomedcentral.com/1471-2288/8/45
Stage three: generating analytical themes
Up until this point, we had produced a synthesis which
kept very close to the original findings of the included
studies. The findings of each study had been combined
into a whole via a listing of themes which described chil-
dren's perspectives on healthy eating. However, we did
not yet have a synthesis product that addressed directly
the concerns of our review - regarding how to promote
healthy eating, in particular fruit and vegetable intake,
amongst children. Neither had we 'gone beyond' the find-
ings of the primary studies and generated additional con-
cepts, understandings or hypotheses. As noted earlier, the
idea or step of 'going beyond' the content of the original
studies has been identified by some as the defining char-
acteristic of synthesis [32,14].
This stage of a qualitative synthesis is the most difficult to
describe and is, potentially, the most controversial, since
it is dependent on the judgement and insights of the
reviewers. The equivalent stage in meta-ethnography is
the development of 'third order interpretations' which go
beyond the content of original studies [32,11]. In our
example, the step of 'going beyond' the content of the
original studies was achieved by using the descriptive
themes that emerged from our inductive analysis of study
findings to answer the review questions we had temporar-
ily put to one side. Reviewers inferred barriers and facilita-
tors from the views children were expressing about
healthy eating or food in general, captured by the descrip-
tive themes, and then considered the implications of chil-
dren's views for intervention development. Each reviewer
first did this independently and then as a group. Through
this discussion more abstract or analytical themes began
to emerge. The barriers and facilitators and implications
for intervention development were examined again in
light of these themes and changes made as necessary. This
cyclical process was repeated until the new themes were
sufficiently abstract to describe and/or explain all of our
initial descriptive themes, our inferred barriers and facili-
tators and implications for intervention development.
For example, five of the 12 descriptive themes concerned
the influences on children's choice of foods (food prefer-
ences, perceptions of health benefits, knowledge behav-
iour gap, roles and responsibilities, non-influencing
factors). From these, reviewers inferred several barriers
and implications for intervention development. Children
identified readily that taste was the major concern for
them when selecting food and that health was either a sec-
ondary factor or, in some cases, a reason for rejecting
food. Children also felt that buying healthy food was not
a legitimate use of their pocket money, which they would
use to buy sweets that could be enjoyed with friends.
These perspectives indicated to us that branding fruit and
vegetables as a 'tasty' rather than 'healthy' might be more
effective in increasing consumption. As one child noted
astutely, 'All adverts for healthy stuff go on about healthy
things. The adverts for unhealthy things tell you how nice they
taste.' [[56], p75]. We captured this line of argument in the
analytical theme entitled 'Children do not see it as their
role to be interested in health'. Altogether, this process
resulted in the generation of six analytical themes which
were associated with ten recommendations for interven-
tions.
Results
Six main issues emerged from the studies of children's
views: (1) children do not see it as their role to be inter-
ested in health; (2) children do not see messages about
future health as personally relevant or credible; (3) fruit,
vegetables and confectionery have very different mean-
ings for children; (4) children actively seek ways to exer-
cise their own choices with regard to food; (5) children
value eating as a social occasion; and (6) children see the
contradiction between what is promoted in theory and
what adults provide in practice. The review found that
most interventions were based in school (though fre-
quently with parental involvement) and often combined
learning about the health benefits of fruit and vegetables
with 'hands-on' experience in the form of food prepara-
tion and taste-testing. Interventions targeted at people
with particular risk factors worked better than others, and
multi-component interventions that combined the pro-
motion of physical activity with healthy eating did not
work as well as those that only concentrated on healthy
eating. The studies of children's views suggested that fruit
and vegetables should be treated in different ways in inter-
ventions, and that messages should not focus on health
warnings. Interventions that were in line with these sug-
gestions tended to be more effective than those which
were not.
Discussion
Context and rigour in thematic synthesis
The process of translation, through the development of
descriptive and analytical themes, can be carried out in a
rigorous way that facilitates transparency of reporting.
Since we aim to produce a synthesis that both generates
'abstract and formal theories' that are nevertheless 'empiri-
cally faithful to the cases from which they were developed'
[[53], p1371], we see the explicit recording of the develop-
ment of themes as being central to the method. The use of
software as described can facilitate this by allowing
reviewers to examine the contribution made to their find-
ings by individual studies, groups of studies, or sub-pop-
ulations within studies.
Some may argue against the synthesis of qualitative
research on the grounds that the findings of individual
studies are de-contextualised and that concepts identified
Page 7 of 10
(page number not for citation purposes)