1. Introduction
Evidence-based medicine (EBM) is the paradigm of contemporary medicine, and it consists of applying credible scientific evidence concerning the safety and efficacy of diverse therapeutic medical interventions in patient clinical management [
1,
2]. It denotes the “integration of individual skills and knowledge with the best available, external clinical evidence from methodical studies” [
3]. The aim of the EBM is to “consciously, unambiguously and reasonably apply the best and up-to-date scientific evidence when making decisions concerning the care of an individual patient” [
2]. EBM is based on randomized controlled trials (RCT) and observational studies.
The term “evidence-based nursing practice” (EBNP) is used in reference to nursing activities. “Evidence-based nursing practice is oriented towards strengthening nurses’ professional position and promoting high-quality patient care” [
4]. According to the International Council of Nurses (ICN), EBNP is defined as “an approach to solving problems in making clinical decisions which involves seeking the best and the latest evidence, clinical expertise and evaluations, and the value of patient’s care-related preferences” [
5]. According to the Canadian Nurses Association (CNA) “evidence-based practice involves making decisions and is used to optimise patient management outcomes, to improve clinical practice and to ensure accountability in nursing” [
6]. According to CNA, best practice guidelines are based on the strictest studies available, such as systematic reviews and randomised clinical trials, but they also include studies that are “well-established in expert opinions and consensus” [
6].
Modern professional nursing is focused on evidence-based practice (EBP) or evidence-based nursing (EBN). Evidence-based practice aims at applying the best methods possible in the nursing process, which are justified by the findings of studies conducted by professionals [
3].
EBNP is founded on the systematic and consistent use of the latest and most reliable scientific research findings in everyday nursing practice and eliminating costly unjustified interventions of low effectiveness [
3,
7].
The EBNP paradigm is considered in three mutually complementary dimensions: personal, organisational, and educational. The EBNP personal dimension includes: self-reflection, critical thinking, and permanent education and skill development. The organisational dimension is characterised by the clinical management model assumptions as a global quality assurance process, which guarantees the maintenance and permanent improvement of standards of care and developing public healthcare responsibility. Permanent education at the workplace according to nursing education is an education-related attribute. “It is the permanent, common, multi-level, organised and methodical improvement of knowledge and translation into practical skills of the ward staff at the workplace” [
3].
The use of the latest specialist scientific periodicals in nursing contributes largely to the improvement in nurses’ knowledge and skills, inspires them to develop their scientific careers, and to permanent education. The International Council of Nurses set the priority areas for nursing activities in accordance with the strategic plan adopted for the years 2019–2023 during the 2019 ICN Congress in Singapore, which covered the following domains: development of the nurses’ role in assuring accessibility and quality of healthcare services; professionalism and promotion of the nursing profession; socioeconomic conditions of nursing; and healthcare system development [
8]. The trends for nursing scientific research in Europe are set by the European Nursing Research Foundation (ENRF), founded by the European Federation of Nurses Associations (EFN) [
8].
Although the tradition of nursing scientific activity in Poland dates back to the inter-war period, the first scientific studies in nursing were not published until the 1960s. The current level of nurse education was significantly affected by the “European Agreement of Nurses’ Training” of 1967. Starting in the mid-1990s, the transformation process in nurse training was associated with Poland’s accession to the European Union. Beginning in 2000, professional nurse training in the majority of Polish universities was fully adapted to the European standards, and the education is now at the university level [
9].
Both undergraduate and postgraduate education are regulated by the relevant legislation. However, it was only when the Regulation of the Minister of Science and Higher Education on the standards of education was published, 12 July 2007 [
10], that an up-to-date standard of education for the first and second cycle of nursing studies was introduced, taking into account the “Research in nursing” subject in the first cycle of study. Meanwhile, the 9 May 2012 Regulation of the Minister of Science and Higher Education on the standards of education for the fields of study of medicine, medicine–dentistry, pharmacy, nursing, and midwifery [
11] provided an update on the standard of education for the nursing field of study. This Regulation placed the subject “Scientific research in nursing” in a group of sciences related to basic nursing care in the first cycle of study. In the second cycle of study, the standard included the subject “Scientific research in nursing” in a group of selected issues within the social sciences. However, only after completing the second cycle of study was an undergraduate student for the first time able to “define the principles of evidence-based medicine (EBM) and evidence-based nursing practice (EBNP)” and “knew the rules of preparing publications for nursing scientific periodicals and demonstrated knowledge of applying EBNP in one’s own professional practice or as part of a managed team” [
11].
Considering the essence of research in nursing at every stage of developing professional qualifications, the “Scientific research in nursing” module was included in the 2003 framework syllabi in the general vocational training block for nurses [
12]. New framework programmes of specialisation training for nurses have been in use since 2017 [
13].
This study examines sociodemographic and work-related variables to determine their impact on the knowledge of, attitudes toward, and skills in EBNP of nurses according to the Evidence-Based Practice Profile Questionnaire (EBPPQ).
2. Materials and Methods
2.1. Settings and Design
The research was conducted from November 2018 to June 2019 among nurses working in various regions of Poland (Dolnośląskie, Łódzkie, Podlaskie, and Zachodniopomorskie voivodeships).
The criterion for inclusion in the study was age > 18 years, a valid license to practise a profession, at least one year of work experience in nursing, and informed consent to participate in the study. The exclusion criteria were: no consent to participate in the study, age < 18, an invalid license to practise a profession, and work experience < 1 year. The study was carried out using the random selection method.
The study was conducted in accordance with the guidelines of the Helsinki Declaration and was approved by the Ethics Committee of the Pomeranian Medical University in Szczecin (resolution no. KB-0012/256/06/18). The respondents were informed about the purpose of the research, about the possibility of resignation and withdrawal of consent at each stage of the study, and had the opportunity to ask questions and obtain comprehensive explanations. After agreeing to participate in the study, the respondents received a set of paper questionnaires.
In total, 1000 sets of questionnaires were distributed. The respondents could withdraw from the study at any time, without giving a reason. The respondents did not receive any remuneration for participating in the research. The completion of the questionnaires took about 15 min. After collecting the data and eliminating the incorrectly completed questionnaires, 830 packages (i.e., 83% of the total) were qualified for further statistical analysis.
2.2. Participants
The study included 830 nurses from four voivodships in Poland: Dolnośląskie (19.2%), Łódzkie (17.3%), Podlaskie (33.7%) and Zachodniopomorskie (29.8%). Women accounted for a vast majority of the participants (95.2%); they were mostly 41–50 years old (41.9%), lived in a city with a population exceeding 100 thousand (35.4%), were in a formal relationship (64.3%), and were university graduates (43.9%) (
Table 1).
2.3. Research Instruments
The study was conducted by the diagnostic survey method, using the questionnaire technique. The following research tools were applied in the study:
An Evidence-Based Practice Profile Questionnaire (EBPPQ) developed by a team of authors from the School of Health Sciences, University of South Australia: Maureen P. McEvoy, Marie T. Williams, and Timothy S. Olds, intended for examining nurses’ knowledge of, attitudes toward, and skills in EBP. The psychometric properties (in terms of reliability and accuracy) of the Polish language version of the EBPPQ were confirmed by a team at the Medical University of Warsaw, and it can be used for scientific purposes in groups of nursing students and practitioners [
14]. All of the EBPPQ domains in the Polish language version are highly reliable, as the Cronbach alpha ranged from 0.800 to 0.972 [
14].
A survey questionnaire developed by the authors, containing questions on sociodemographic data (sex, age, place of residence, family and marital status, and educational background) and work-related variables (work experience, place of work, and work organisation system).
2.4. Statistical Analysis
Metric and non-metric variables were characterised by descriptive statistic parameters. The following were determined for variables expressed on a metric (quantitative) scale: measures of central tendency (M—mean), measures of variability (SD—standard deviation, CV—coefficient of variation). The measures of structure, sample size (N) and purity (%), were determined for the variables expressed on a non-metric scale.
The classic statistics based on testing zero hypotheses were used for statistical inference. The choice of statistical tests was affected by the type of the dependent variable measurement scale and the number of factor levels (number of independent variable variants). Testing zero hypotheses was complemented by determining the measures of effect (with 95% confidence of intervals) matched to the type of the statistical test used. The mathematical statistics methods concerning inter-sample (inter-group) differences included: Student’s t-test for unrelated samples with an estimation of Cohen’s d coefficient; a one-way analysis of variance (ANOVA) with an estimation of the η2 coefficient, and Scheffe’s post hoc test.
The statistical analysis methods used to assess the relationships between two or more variables involved the Pearson linear correlation coefficient (r) defining the linear correlation between two metric variables.
The default level of statistical significance—0.05—was adopted for all analyses. The calculations were performed with STATISTICA v. 13.3 (TIBCO Software Inc., Palo Alto, CA, USA).
A regression model (generalized linear model (GLM)) was used to explain the perception of, knowledge of, and skills in EBNP. Two regression models were analysed: the full model (all predictors) and the stepwise model (the stepwise model that best explains the variance of the dependent variable). All the independent variables (predictors) were entered into the full regression model simultaneously. Then, predictors were eliminated backwards from the full regression model in order to obtain the model that best explained the variance of the dependent variable. We tested the assumption of the noncollinearity of the predictors by calculating the tolerance and the variance inflation factor (VIF). The standardized regression coefficient (β) with 95% confidence interval (95% CI) was estimated for each predictor. The adjusted coefficient of determination (adjusted R-squared) was calculated to define the percentage of the explained variance of the dependent variable by the independent variables.
3. Results
3.1. Educational and Work-Related Profile of the Study Group
The majority of the 830 nurses graduated from a medical university (48.2%), followed by those who graduated from a non-medical university (18.6%) or from a vocational school of higher education (16.9%). Moreover, 84% of the respondents completed postgraduate training (specialisation in a specific field of nursing).
The mean work experience was 18.9 years (SD = 11.86 years). The majority of the respondents were employed under a contract of employment (84.7%), others worked in a two-shift system (61.2%) and worked in a clinical hospital (25.7%), in a voivodship (22.7%), county (20.7%), or a municipal hospital (13%). The other respondents worked in primary healthcare facilities (16.9%), curative care centres (3.5%), or nursing homes (1.3%). Further analysis of the employment profile revealed data on the place of work: a medical treatment ward (33.3%), noninvasive treatment ward (19.9%), paediatric ward (15.5%), or intensive care unit and emergency unit (total: 14.3%); work at a different ward was declared by 0.5% of the respondents, and 16.5% of them did not choose any of the options (
Table S1).
More than half of the surveyed nurses (64.5%) did not use EBNP in their work, 29.6% answered affirmatively, while 5.9% did not give an answer. In total, 61.9% of the respondents declared that they learned about the subject of EBNP in the course of their education, 34.1% did not have knowledge of this issue, and 4% of the respondents did not provide any answers. The majority of nurses, 40.5% learned about EBNP during their studies, 14% during postgraduate education, and 10.7% during self-education. When examining the use of EBNP in the work of the respondents, the following data were obtained: half of the respondents 418 (50.4%) did not use it, 356 (42.9%) used it, and 56 people (6.7%) did not provide an answer (data not shown).
3.2. The Level of Knowledge of, Attitudes toward, and Skills in EBNP According to the EBPPQ
An EBPPQ questionnaire was used to evaluate the knowledge of, attitudes toward, and skills in EBNP. According to the results, the highest scores were given to “EBNP-related skills” (M = 5.56 ± 1.84), followed by “Level of knowledge of scientific research terminology” (M = 5.50 ± 2.03) and “Other aspects of EBNP” (M = 5.50 ± 1.80). The other subscales included: “Attitude towards expanding one’s competence in EBNP”, “Attitude towards selected aspects of EBNP at work”, and “Frequency of using various EBNP components in everyday clinical work” and were estimated at similar levels (M from 5.48 to 5.49, SD from 1.84 to 1.91) (
Table 2).
3.3. Impact of Sociodemographic Variables on Knowledge of, Attitudes toward, and Skills in EBNP among Nurses According to the EBPPQ
This study analysed the impact of sociodemographic variables (age, marital status, education, and work experience) on the knowledge of, attitudes toward, and skills in EBNP according to the EBPPQ.
The results revealed a statistically significant negative correlation (
p < 0.05) between age and the subscales “Frequency of using various EBNP components in everyday clinical work” (r = −0.13), “Level of knowledge of scientific research terminology” (r = −0.07), and “EBNP-related skills” (r = −0.12) (
Table 3).
Three groups of nurses, single, in a formal relationship, and those in an informal relationship, were analysed in terms of the marital status impact on the knowledge of, attitudes toward, and skills in EBNP according to the EBPPQ. A statistically significant difference in the mean level of “EBNP-related skills” was demonstrated in the study (F
(2,811) = 3.644;
p = 0.027; η
2 = 0.009; 95% CI [0.001; 0.024]). The mean level of these skills was found to be significantly higher in individuals in informal relationships than in those in formal relationships (M: 5.88 vs. 5.44; Scheffe test:
p = 0.032). However, no significant differences were found between the group of single individuals and those in formal or informal relationships (
Table 4).
The impact of the respondents’ educational background on their knowledge of, attitudes toward, and skills in EBNP was considered in two groups: respondents with secondary (medical secondary or postsecondary school) or university education (bachelor or master of nursing studies).
The results revealed a significant difference (p < 0.05) between the mean level of “The attitude towards expanding one’s competence in EBNP”, “Level of knowledge of scientific research terminology”, “Frequency of using various EBNP components in everyday clinical work”, “EBNP-related skills”, and “Other aspects of EBNP” in the groups under study.
A higher mean level of a subscale under study was observed in all of the statistically significant correlations in a group of respondents with university education compared with those who completed secondary or postsecondary school. The strongest impact of the educational background on the attributes under study was observed for the “EBNP-related skills” subscale (d = 0.46; 95% CI [0.26; 0.63]), (
Table 5).
3.4. Impact of Work-Related Variables on Knowledge of, Attitudes toward, and Skills in EBNP among Nurses According to the EBPPQ
A weak negative correlation was revealed in terms of the work experience impact on the knowledge of, attitudes toward, and skills in EBNP according to the EBPPQ, concerning: “Frequency of using various EBNP components in everyday clinical work” (r = −0.13;
p < 0.001) and “EBNP-related skills” (r = −0.10;
p = 0.004) (
Table 6).
This study analysed the impact of specialisation training on the knowledge of, attitudes toward, and skills in EBNP according to the EBPPQ. The results revealed statistically significant differences (
p < 0.05) in the knowledge of, attitudes toward, and skills in EBNP among nurses according to the EBPPQ as affected by specialisation training in a specific field of nursing. A higher level of the subscale was demonstrated in the group of respondents with a specialisation. The strongest impact of specialisation on the subscales was observed for “The attitude to expanding one’s competence in EBNP” (M: 5.74 vs. 5.37;
p = 0.007; 95% Cl [0.05; 0.35]) (
Table 7).
3.5. Impact of Work-Related Variables on the Knowledge of, Attitudes toward, and Skills in EBNP among Nurses According to the EBPPQ and Regression Analysis
We used regression analysis, introducing variables to the full multivariate model and the stepwise model in order to generate the best model explaining the highest possible level of the variance of the dependent variable.
3.5.1. Regression Analysis for the Subscale ‘Attitude towards Expanding One’s Competence in EBNP’
The positive attitude toward expanding one’s own EBNP competencies was shown to increase with work experience (β = 0.28,
p = 0.009). Those who completed training organized by the Board of Nursing had significantly worse attitudes toward expanding their EBNP competencies than those who completed training organized by a university (β = −0.11,
p = 0.018). People who worked in the Primary Nursing system had a significantly better attitude towards expanding their EBNP competencies than those working in other systems (β = 0.17,
p = 0.028). People who had the subject of EBNP in the course of their education had a significantly better attitude towards expanding their competencies regarding EBNP than individuals without such experience during their training (β = 0.31,
p < 0.001). People using EBNP in their work had a significantly better attitude towards expanding their EBNP competencies than individuals without such opportunities in the workplace (β = 0.34,
p < 0.001) (
Table 8). The presented model was statistically significant (F (25, 366) = 8.707,
p < 0.001). The independent variables together explained approximately 33% of the variation in the dependent variable (adjusted-R
2 = 0.33, SEE = 1.51).
The stepwise model additionally confirmed that the negative attitude toward expanding one’s own EBNP competencies increased with age (β = −0.20,
p = 0.046), and individuals who worked in a hospital had significantly worse attitudes towards expanding their EBNP competencies than those who worked in other workplaces (β = −0.11,
p = 0.011) (
Table 9). The presented model was statistically significant (F (7, 384) = 28.079,
p < 0.001). The independent variables together explained approximately 33% of the variation in the dependent variable (adjusted-R
2 = 0.33, SEE = 1.52).
3.5.2. Regression Analysis for the Subscale ‘Attitude towards Selected Aspects of EBNP at Work’
People who worked in clinical hospitals (β = 0.19,
p = 0.038), district hospitals (β = 0.17,
p = 0.031), and primary care centres (β = 0.15,
p = 0.018) had significantly better attitudes towards selected aspects of EBNP at work than people who worked in other workplaces. Moreover, people who worked in surgical wards had significantly better attitudes towards selected aspects of EBNP at work than those who worked in other wards (β = 0.21,
p = 0.045). People who came into contact with the topic of EBNP in the workplace had significantly better attitudes towards using selected aspects of EBNP at work than people without such experience (β = 0.16,
p < 0.010) (
Table 10). The presented model was statistically significant (F (25, 366) = 1.777,
p = 0.013). The independent variables together explained approximately 5% of the variation in the dependent variable (adjusted-R
2 = 0.05, SEE = 1.90).
In the multivariate model with backward elimination, it was additionally confirmed that people who worked in hospitals had significantly better attitudes towards selected aspects of EBNP at work than those in other workplaces (β = 0.12,
p = 0.033) (
Table 11). The presented model was statistically significant (F (25, 366) = 4.192,
p = 0.001). The independent variables together explained approximately 4% of the variation in the dependent variable (adjusted-R
2 = 0.04, SEE = 1.90).
3.5.3. Regression Analysis for the Subscale ‘Level of Knowledge of Scientific Research Terminology’
People who came into contact with the topic of EBNP in the workplace had a significantly better knowledge of the terminology related to scientific research than those without such experience (β = 0.18,
p = 0.001) (
Table S2). The presented model was statistically significant (F (5, 386) = 3.334,
p < 0.001). The independent variables together explained approximately 13% of the variation in the dependent variable (adjusted-R
2 = 0.13, SEE = 1.93).
It was confirmed in the stepwise model that people working under a hospital employment contract had a significantly better level of knowledge of terminology related to scientific research than those working under a civil contract (β = −0.10,
p = 0.034) (
Table S3). The presented model was statistically significant (F (3, 386) = 21.919,
p < 0.001). The independent variables together explained approximately 14% of the variation in the dependent variable (adjusted-R
2 = 0.14, SEE = 1.92).
3.5.4. Regression Analysis for the Subscale ‘Frequency of Using Various EBNP Components in Everyday Clinical Work’
People who used EBNP in their work were significantly more likely to apply individual elements of EBNP in their daily clinical work than those without such opportunities in the workplace (β = 0.17,
p = 0.009). The presented model was statistically significant (F (25, 366) = 1.707,
p = 0.020). The independent variables together explained approximately 4% of the variation in the dependent variable (adjusted-R
2 = 0.04, SEE = 1.896) (
Table S4).
It was noted in the stepwise model that the older the nurses, the less frequent the use of individual EBNP elements in their daily clinical work (β = −0.17,
p = 0.001), and people who used EBNP in their work used individual elements of EBNP in everyday clinical work definitely more often than those without such opportunities in the workplace (β = 0.20,
p < 0.001) (
Table S5). The presented model was statistically significant (F (2, 389) = 14.015,
p < 0.001). The independent variables together explained approximately 6% of the variation in the dependent variable (adjusted-R
2 = 0.062, SEE = 1.877).
3.5.5. Regression Analysis for the Subscale ‘EBNP-Related Skills’
People who completed training organized by the Board of Nursing had significantly worse EBNP-related skills than those who completed training organized by a university (β = −0.14,
p = 0.010). Similarly, people employed in several places had worse EBNP-related skills than those employed only in one workplace (β = −0.14,
p = 0.013). People who used individual elements of EBNP in everyday clinical work had significantly better EBNP-related skills than those without such opportunities in the workplace (β = 0.17,
p = 0.009). The presented model was statistically significant (F (25, 366) = 2.076,
p = 0.002). The independent variables together explained approximately 6% of the variation in the dependent variable (adjusted-R
2 = 0.06, SEE = 1.697) (
Table S6).
It was observed in the stepwise model that people who completed training organized by the Board of Nursing had significantly worse EBNP-related skills than those who had completed training organized by a university (β = −0.12,
p = 0.018). Similarly, people employed in several places had worse EBNP-related skills than those employed only in one workplace (β = −0.10,
p = 0.040). People who used EBNP in their work had better EBNP-related skills (β = 0.24,
p < 0.001) than those without such opportunities in the workplace (
Table S7). The presented model was statistically significant (F (4, 385) = 8.812,
p < 0.001). The independent variables together explained approximately 7% of the variation in the dependent variable (adjusted-R
2 = 0.074, SEE = 1.688).
6. Limitation
Considering the nurse training system in Poland, it should be emphasised that the study findings concerning the sociodemographic data with an impact on the knowledge of, attitudes toward, and skills in EBNP may have been affected by the respondents’ age and—closely related to it—their educational background, which depends on the under- and postgraduate training system in Poland. The study findings with respect to work-related variables with an impact on the knowledge of, attitudes toward, and skills in EBNP may have been affected by diverse work organisation systems, organizational culture, regional characteristics within a healthcare facility, as well as the composition of nursing staff. We will improve these variable factors in our future studies.
Another limitation of our research may be the random selection of the study sample. Further, our research was conducted in only some parts of Poland. Hence, the results of this study may not be representative of all nurses and cannot be generalized. However, the results of this study should be carefully incorporated into the curriculum of nursing studies.