The results were summarized as showing corroborating
evidence for the association between the index of support
and QOL aspect if at least 60% of all original studies
(addressing a respective support source) indicated significant
associations between support and QOL indices (e.g.,
two out of three studies referring to support from family/
friends and emotionalQOL yielded positive findings). Again,
the 60% threshold has been applied in earlier reviews [22,23]
as the indication of corroborating evidence. The results were
summarized as showing preliminary evidence for the role of
the social support index if (1) 50–59% of the studies
discussing the social support variable and a respective
outcome showed significant associations, or (2) the
association was tested in only one study, which revealed
significant effects [22,23]. To our knowledge, there is a
lack of alternative thresholds used to analyze data in
systematic reviews than those applied in the present study.
Quality assessment was conducted using the quality
evaluation tool developed by Kmet et al. [18]. Respective
standard quality assessment criteria for evaluating primary
research papers [18] are included in several quality evaluation
tools, such as TREND [24]. The quality evaluation
tool [18] applies quantitative methods, and it allows to
investigate whether the study adheres to the following
14 criteria: sufficiently described objectives, evident/
appropriate design, clear description of participant selection
and measures, participant description, random allocation
(experimental trials), blinding of interventionists
(experimental trials), blinding of participants (experimental
trials), selection of outcomes, appropriate sample size,
analytic methods (selection and description), an estimate
of variance reported in main results, controlling analyses
for confounders, reporting results in sufficient detail, and
conclusions supported by results. Each criterion is rated
using a 3-point response scale. The summary scores
(Table 1) are reported as percentages, representing a ratio
of total score obtained to a total possible sum score [18].
The concordance coefficients for quality assessment were
high (all Kappas≥.76, ps<.01).
The cut-off score for the acceptable quality of studies
was twofold: (1) quality score≥60% (55% and 60% are
suggested as relatively liberal thresholds, indicating acceptable
quality [18]) and (2) the study should at least partially
meet the criteria referring to the methods, analyses, and
results [19]. Meeting at least 75% of quality criteria is considered
a conservative quality threshold [18], indicating
minor flaws [19] and thus showing relatively high quality.
In case of longitudinal studies, data from the latest
available follow-up were included into analysis. For
experimental studies investigating the influence of a social
support intervention, the effect of the manipulation was
accounted for in our analyses. In case of multiple analyses
dealing with the same QOL and support indices reported
in the original study, we included units that controlled
for potential confounders.
การแปล กรุณารอสักครู่..