Methods
Following the sequential mixed-methods design, the initial results from the surveys were further explained by data gathered through the focus group. In both, we examined the
participants’ relationships to EBP beliefs, values, and behaviors while assessing factors
like years of education, employment, age, and years of research experience.
The study was approved by the Seattle University Institutional Review Board (file number FY2017-003). The participants signed an informed consent agreement (in Spanish) prior to participating, which expressed the voluntary nature of their participation, their right to refuse to participate or answer questions, and the measures taken to ensure the privacy of their responses.
Setting and Participant Recruitment . Professional nurses (Licenciadas en Enfermeria )
in three hospitals and five public health centers (Redes de Salud ) in La Paz, who had worked for at least two years were invited to participate in the study. Hospitals A, B, and C were all tertiary hospitals with approximately 358, 345, and 160 beds, respectively. The health centers, collectively referred here as D, were located in marginalized neighborhoods. The participants were recruited during brief visits to the study’s settings by two Bolivian-trained research assistants (RAs). During the survey period, the RAs could begin to identify the formal and informal leaders, who were to comprise the focus group. These leaders were later invited by e-mail and mail to participate in the focus group. No participant was excluded by any demographic variable (e.g., age, marital status, or sex). The focus group was configured to be homogenous in terms of experience and focus of practice, although each person’s characteristics, e.g., age and years of employment, provided for a diversity of opinion during the group sessions, which always had at least seven participants. Participants received means for transportation to motivate attendance and they were also offered small gifts as a gesture of appreciation.
Measures . The survey included: a) demographic questions (Table 1), b) EBP Beliefs (EBPB), c) EBP Implementation (EBPI), and d) the BARRIERS scales. The EBPB scale measures a provider’s belief about the value of EBP.19 This scale had 16 items, gauged on a five point Likert-type scale, ranging from ”strong disagreement = 1” to ”strong agreement = 5”. Total responses could thus range from 16 to 80. The EBPI scale, containing 18 items, measures the belief and confidence in implementing EBP. It asked the frequency of each item performed, ranging from ”0” (zero times)” to “4,” (> 8” times). Total scores could range from 0 to 72. The scales establish appropriate face, content, and construct validity with internal consistency reliabilities of Cronbach’s alpha coefficients > 0.90.19The BARRIERS scale assesses the provider’s perceived barriers to research utilization. This Likert-type scale is comprised of 29 items under four factors (Table 2), on which they can respond to potential barriers, from 1 (to no extent) to 4 (to a great extent), as well as a non-opinion option. This scale has demonstrated to have high face and content validity with a Cronbach’s alpha of 0.65-0.80.20The BARRIERS survey was translated into Spanish following an acceptable process,21 and pretested, along with the Spanish version of the EBPB and EBPI survey questionnaires with a group of nurse volunteers (N = 12) having similar characteristics to the study’s target population. Similarly, five nurses validated the focus group questions to produce discussion sessions of approximately two hours moderated by a skilled facilitator. The discussion questions were: a) What does evidence-based practice mean to you? b) How do you identify evidence-based interventions? c) What barriers make it difficult to implement EBP? and d) What changes (personal or institutional) do you think are necessary to implement EBP?
Data Analysis . Quantitative data were analyzed in SPSS 19 (IBM, 2010). Here, the descriptive statistics regarding the nurses’ demographic and professional characteristics and survey questionnaires were analyzed. The relationships between nurses’ demographics and professional characteristics and their identified barriers and facilitators for EBP were examined using chi-square tests for categorical measures and Pearson’s r for interval or ratio measures. For the qualitative part, all sessions dialogues, field notes, and the principal investigator’s (PI) personal notes were transcribed. The analysis used an inductive approach, beginning close to the data and moving through levels of more abstract analysis to identify the patterns and relationships explaining the phenomena under review.22 ATLAS.ti 8.0 was used to identify the repetition of phrases and words using in vivo coding to determine themes, as well as to provide verbatim statements made about those themes. The results were consistent across the three sources of data. For cross-language trustworthiness, the analysis was completed in Spanish—the participants’ language. Additionally, transcripts were reviewed by the PI and a doctorate-prepared Bolivian nurse, who negotiated differences. Furthermore, two strategies for validation of data analysis were used: a) investigator triangulation by two bilingual investigators, who independently analyzed the same data; and b) a review of the description of findings by the participants themselves.