Quality assessmentAll articles were

Quality assessment
All articles were screened for quality using the published
‘Quality Assessment and Validity Tool for Correlational
Studies’ adapted from previous systematic
reviews (Cummings & Estabrooks 2003, Estabrooks
et al. 2003, Wong & Cummings 2007, Cummings
et al. 2008, 2010, Wagner et al. 2010, Cowden et al.
2011) (Figure 2).
The instrument included 13 questions to scrutinize
and score the research design, sample, measurement
and statistical analysis of the studies. The
questions were in the dichotomous answer format
and a total of 14 points could be assigned for the
13 criteria. Twelve items were scored as 0 (=not
met) or 1 (=met) and the items related to outcomes
measurement were scored out of two. Based on
scores assigned, the instrument classifies the articles
into three quality categories: low (0–4), medium (5–9)
and high (10–14).
Data extraction
Following the assessment for quality, 12 studies
remained for data extraction.
Data extracted from selected studies included:
author, year, journal, research question presented
(aim of the study), sample, response rate, independent
variable, dependent variable, measures, reliability and
validity of the instrument used, analysis and main
results.
Results
The electronic database search yielded over 1500
abstracts and titles. Following removal of duplicates,
the potentially eligible studies were screened. Some
573 studies were excluded. Quality assessments were
completed on 13 articles and following quality assessment,
one quantitative paper was excluded (Laschinger
et al. 2003). Twelve studies were included in the
final analysis.
Figure 3 illustrates the search and retrieval process.
Summary of quality review
Five studies included in the review were rated
high quality (Laschinger et al. 2004, Laschinger et al.
2001a,b, 2011, Ning et al. 2009). The seven remaining
studies were rated medium quality (Manojlovich
& Laschinger 2002, Laschinger 2008, Laschinger et al.
2009a, Lautizi et al. 2009, Cai & Zhou 2009, Cai
et al. 2011, Ahmad et al. 2010).
Most of the studies utilized a non-experimental,
cross-sectional design. Only one study used a longitudinal
design (Laschinger et al. 2004). All studies
were prospective in design as data were collected
prospectively.