Factors influencing the development of evidence-based practice among nurses: a self-report survey

Data collection

The questionnaires were distributed to the nurses at work with an attached reply envelope. Each unit selected a contact person who took responsibility for collecting data. They were informed about the aims of the study and that the study was anonymised and that participation was voluntary. The designated contact persons further disseminated information about the study to the nurses at their respective units by e-mail as well as at information meetings. Completed questionnaires were collected at each unit. A person other than the investigator opened the envelopes and delivered the questionnaires for data processing. The study was approved by the ombudsperson of the Haukeland University Hospital as well as the head of the Department of Research and Development. A pseudo number was made to ensure that data could not directly be traced back to any respondent. The link between the pseudo number and personal identification number was stored separately on a secured server at the hospital.

Questionnaire

The self-reported Developing Evidence-based Practice Questionnaire was developed by Gerrish et al.[18] for mapping factors influencing the development of evidence-based practice. The questionnaire was previously shown to be a valid and reliable instrument [18]. Gerrish has done extensive research on the implementation of strategies to promote evidence-based practice and provided permission to translate and use her questionnaire. The questionnaire was translated into Norwegian in accordance with the World Health Organization’s [22] procedures for translation and adaptation before data collection started. The Norwegian version included an additional question in the second section asking whether ability in English might be a barrier for using research-based evidence. The internal consistency of the Norwegian questionnaire version measured by Cronbach`s alpha was 0.88. The first author of this publication carried out the translation.

The Norwegian version of the questionnaire includes 50 pre-structured questions and 4 open-ended questions. The pre-structured questions that were answered by using a five-point Likert scale. The open-ended questions were beyond the scope of the current study. The Questionnaire is divided into 5 sections. Section 1 consists of 22 questions about the use of sources of knowledge for support in practice, rated from never (1) to always (5). The sources of knowledge are divided into five dimensions, (1) own experience covered by items 1–5, (2) social interaction covered by items 6–8, (3) internal sources of knowledge covered by items 9–14, (4) research evidence covered by items 15–19 and (5) external sources of knowledge covered by items 20–22. Sections 2–4 include 20 statements on possible barriers to evidence-based practice. Respondents rated the statements in sections 2 and 3 from strongly agree (1) to strongly disagree (5). In contrast, section 4 has a positive value for affirmative responses, from always (1) to never (5). Section 5 includes 8 questions about skills in finding, reviewing and using research-based evidence. The responses range from complete beginner (1) to expert (5). The questionnaire ends with questions on demographic variables such as the nurse’s age, sex, current educational level and number of years of nursing practice.

Statistical analysis

We used PASW for Windows version 18.0 (SPSS Inc., Chicago, Illinois) for statistical analysis. Due to multiple testing, all P values were two-sided, and values below 0.01 were considered statistically significant. To handle missing values in regression models, we used the method of list-wise deletion. Continuous variables were reported as means (SD) and categorical variables as counts (percentage).

To examine the associations of various background variables with variables on sources of knowledge, we used linear regression models. Variables on sources of knowledge were continuous (i.e., scales) and constructed by aggregating response items within each of the five dimensions: own experience, social interaction, internal source of knowledge, external source of knowledge, and research evidence. Background variables selected for the analyses were categorical and included nurse’s age, number of years of clinical practice, and number of years since obtaining last health professional degree. The P value for trend was estimated by incorporating the background variables as linear terms in the regression models. P values were reported with and without adjustment for age, number of years of nursing practice, number of years since obtaining last health professional degree, as well as current educational level.

To examine the associations of self-reported skills in finding, reviewing and using research evidence with the use of specific sources of knowledge and evidence, we used chi-square tests. Variables on the use of specific sources of knowledge and evidence were recoded into 3 categories (seldom/never, sometimes, often/always) and included two separate response items on finding research-based evidence, two separate response items on assessing research-based evidence, and two separate response items on using research-based evidence to change practice. Similar analyses were performed for the association between self-reported skills in evidence-based practice and reported barriers.

Findings

The results showed that nurses mainly used experience-based knowledge for use in practice rather than evidence gained from research journals. Insufficient time was the greatest barrier to finding and reviewing research literature. However, skills in evidence-based practice influenced the use of knowledge sources as well as how the nurses assessed barriers to evidence-based practice.

Sources of information reported for supporting practice

Articles published in medical, nursing and other research journals were the least frequently used sources of information. This result is in accordance with other studies [23–25]. A surprising result was that the use of research-based evidence as a source in practice increased with the age of the nurse and with number of years of nursing practice. This contrasts with the findings of Milner et al.[26], who found that increasing age indicated a lower score in the use of research-based evidence. The reason why age and number of years of practice were associated with various responses could be that experienced nurses have basic knowledge-based experience that provides confidence in how to carry out routine tasks and manage unforeseen events [27, 28]. Experienced nurses often pose other types of questions that can be answered using research-based sources of evidence. These are called foreground questions, in which someone wants specific evidence as the basis for decisions. Less experienced nurses tend to seek general knowledge within a topic and pose background questions that can be answered by using textbooks, knowledge from education and training and by asking more experienced colleagues [19]. A review shows, however, no association between age or the number of years of practice and the use of research-based evidence [29]. Estabrooks et al.[23] found that the nursing profession and how practice is organized may be reasons why informal sources are used so extensively. The nursing profession has traditionally been based on a theory, guided practice, in contrast to medicine, which is dominated by empirical models for evidence, especially obtained from randomized controlled trials [18, 30]. The findings in this study suggest that experienced nurses may be a resource for implementing evidence-based nursing practice. Experienced nurses can act as facilitators assisting colleagues to frame precise questions as a basis for search in the literature and critically assess relevant findings. Facilitation is crucial for successful implementation of evidence-based practice [13]. A prerequisite for this success is also that facilitators are offered training in evidence-based practice methods [31].

This study and others [24, 32] show that nurses often use hospital policy and protocols as a source of information. Good evidence-based professional policy, guidelines and protocols should therefore be worked out using the best available evidence and clinical knowledge tools for ensuring that research is integrated into clinical practice [7]. An erroneous perception has been that introduction of evidence, for example evidence-based guidelines or protocols are seen as a linear, technical process at individual level [13]. Having guidelines based on the best available evidence and easily accessible in the organization’s document management system does not guarantee that they will be used. Effective measures for implementation are required for this, but the research conducted so far does not indicate which measures are most effective [33]. A survey conducted among Norwegian hospitals showed that the quality of the protocols and procedures is uncertain [34]. There is a need to ensure that the local procedures meet established quality criteria if the protocols and policy should be a tool for implementation of evidence-based practice in health care in Norway. Further, organizational and professional awareness-raising is needed concerning the effects of not using the approved hospital guidelines and procedures, even if they are not based on evidence [32].

Barriers to evidence-based practice

In accordance with other studies [16, 35], the greatest barrier using research-based evidence was insufficient time to find research. The culture of changing the clinical practice of the hospital team was considered the least barrier. Thompson et al.[36] investigated the relationships between “insufficient time” and evidence-based practice. Insufficient time reflects the mental time and energy required to use research and the culture of busyness rather than the actual amount of time required. Nurses, their colleagues and the organization itself maintain a culture in which busyness is valued and rewarded. Such a culture does not favour participants who sit down and read and reflect over research results but instead largely rewards observable tasks carried out among patients.

In contrast, a culture that supports the implementation of research in practice is considered a strong motivational factor for using research-based evidence [37]. Findings [36] indicate that administrative and collegial support probably has more strongly influence on the use of research than lack of time resulting from a high pace of work. Cultural resistance may be more related to indifference and lack of action than active resistance [37]. Younger nurses and nurses with fewer years of practice considered lack of time a greater barrier than others. This may be related to personal characteristics such as experience and confidence and how tasks can be solved in a complex organization [36].

Nurses’ skills in finding and assessing research-based evidence and implementing it in practice

Most responding nurses rated their skills in evidence-based practice as being beginner/novice. Previous studies [15, 32, 38] have explained the limited use of research-based evidence by lack of competence in finding, assessing and understanding research. This would imply that courses in evidence-based practice would lead to greater use of research-based evidence. Nevertheless, the effect of teaching as the sole measure for increasing the use of research-based evidence is very uncertain [25, 39, 40]. Whether a person has higher skills in evidence-based practice does not mean that that this person implements the evidence in practice [41].

Associations between managing evidence and the sources of information used and the barriers to evidence-based practice

This study found that skills in evidence-based practice were statistically significantly associated with the sources of information used. Nurses who reported higher skills in evidence-based practice used research papers more often than nurses with lower skills. This suggests that an implementation process should include courses in evidence-based practice for nurses [31].

Similarly, skills were statistically significantly associated with how the barriers to the use of research-based evidence were assessed, also after being adjusted for age. Nurses with better skills reported fewer barriers to evidence-based practice. Research is not used in isolation but is influenced by factors at the individual level, collaboration between multidisciplinary groups, management and organizational structure [7]. A recently updated systematic review found association between the use of research and individual factors such as attitudes toward evidence-based practice, completed course in evidence-based practice, education of undergraduates, working in an intensive care unit and job satisfaction [29].

Limitations

This was a cross-sectional study which describes a nurse population in a specific time period. Although it may indicate possible associations, more studies should be conducted to confirm our findings. Furthermore, our results may not be representative for nurses in general, but biased towards nurses in units where the management implements evidence-based practice. Finally, the Developing Evidence-based Practice Questionnaire is relatively new and has to date only been used in a few studies for assessing factors influencing the development of evidence based practice. Our results could therefore not be directly compared to other studies.