INTRODUCTION

Technology drives both people and society to change and develop. With the developing technology, it is necessary to follow more innovative processes to meet the increasing requirements of the education processes [1]. Innovation is defined as the renewal of science and technology in a way that will provide economic and social benefits, creating inventions and being different. According to TDK (2023), “innovation is the adaptation and application of new creative ideas or inventions to economic fields” [2].

Rogers (2003) defines innovation as "an idea, practice, object that is perceived as new by the individual or society" and lists its characteristics as follows [3]: relative utility, compatibility, complexity, trialability and observability. In addition, in this process, attitudes towards innovation as Innovators, Pioneers, Curiosity, Skepticism, Traditionalists are specified. (looking with prejudice against change, tending to adopt innovations last, waiting for the innovation to be tried by others and the results to be observed before adopting the innovation) defined [4]. The International Union of Nurses-ICN (2023) described innovation in the health care system as transforming a good idea into a viable/achievable outcome for health promotion, disease prevention, and higher quality patient care [5].

Developing new ideas, technologies and techniques, suggesting new ways/methods for realizing goals, testing new work procedures in the work area, changing work routines and applying new methods are defined as "innovative behavior" [6]. In global trend analysis studies, it is prominent that innovative practices in the health field will increase with the developing technology in the future [7-9]. The National Nursing Association (NLN) has called for dramatic reform and innovation in nursing education [10]. NLP recommended the preparation of evidence-based curricula that can act with technology in nursing and respond to the needs of students and health care systems. For this purpose; the theme of Nursing Week in 2009 was determined as "nursing and innovation". The International Council of Nursing advocated that nurses should be pioneers in developing innovative care practices. It supported the creative behaviors of nurses, the products and methods developed by nurses as an innovation, and the revealing of the proactive role of nurses in health care, in cooperation with educational institutions and professional organizations, both in public health and public health. It recommended the development of professional knowledge [11, 12]. In the report published by the American Institute of Medicine (IOM) in 2010, health informatics, innovation, technology applications, and applications for developing healthcare technologies come to the fore among the competencies nurses should have in the future [13]. This shows that innovation is still on the agenda for the development and advancement of nursing. Personal entrepreneurship is essential in the health care system. Entrepreneurship has been defined as a concept that encompasses all processes of entrepreneurs to take risks, pursue opportunities, implement, and innovate [14]. Entrepreneurship is the expenditure of a great deal of energy in establishing an enterprise or organization [5, 15]. Entrepreneurship is essential to nurses' decision-making or career-planning processes [14]. When this information is evaluated, innovation in nursing is significant. When the studies in the literature were examined, no meta-analysis examined the studies on determining individual innovativeness characteristics (innovation) in nursing. This research was conducted to determine the personal innovativeness characteristics of the invention in nursing.

Research Hypotheses

H1: Innovation in nursing affects individual innovativeness.

MATERIALS AND METHODS

This research was done by meta-analysis method. Literature review method was used in the study. Since the literature review does not directly affect animals or humans, ethics committee. Keywords for article searches were determined by searching the database of Turkish Science Terms and medical topics (MeSH Browser) [16]. Identified keywords “nurse”, “nursing” and “innovation” articles related to the subject were searched from databases in English and Turkish. In the searches made from Google Scholar, Medline, Turk Medline, ULAKBIM, and Cinahl databases, related research articles published between 2009 January and February 2020 were included in the evaluation. After the repetitions in the themes downloaded from separate databases were deleted, the articles' title, abstract, and full-text reading stages were started, respectively. A total of 856 articles were found in the searches performed, with the search strategy determined in five databases. After removing duplicates and those whose title and abstract did not match, the remaining articles were evaluated within the scope of title and abstract reading. Articles that were found to be irrelevant to the subject were classified in detail and excluded from the study. The data obtained in the MOOSE studies were displayed based on the criteria of authors, year, type of study, sample size, quality assessment score of innovation. The implementation steps of the study were classified based on PRISMA [17] and MOOSE criteria. These are presented in Figure 1 and Table 1.

Implementation Steps of the Study.

Criteria for Inclusion in Meta-Analysis and Methodological Quality Assessment

1. Quantitative articles,
2. Only those written in English and Turkish,
3. Accessible full text and published in a national/international peer-reviewed journal
4. Nursing and innovation articles,
5. Articles from the last 11 years (2009-2020)

Two researchers evaluated the articles separately to avoid publication bias.  Information about the articles is as follows.

1. Publication year of the study
2. Types of work
3. The sample size of studies
4. Individual innovative features
5. Quality evaluation score

Kappa statistics [15] suggested by Polit and Beck and 12 of the research quality evaluation criteria were used in all studies. Each study was evaluated on all requirements and separately by the researchers, and if each item fully met, "1 point" was given; if not, "0 points" was given, and studies that met the inclusion criteria were included in the study. It was evaluated as an article with a weak quality of 0-4 points, medium quality with 5-9 points, and robust quality with 9-12 points [15]. As a result of the evaluation, the highest score was 11, and the lowest was 7. A total score of 12 indicates that the study is of good quality. The quality evaluation of the seven studies included in the meta-analysis and the distribution of the scores they received from each domain are given in Table 1. Cohen kappa statistic was used to evaluate the agreement between researchers for selecting articles and bias scoring made independently by the three authors. A kappa statistic between 0.41 and 0.60 was considered moderate, between 0.61 and 0.80, a significant degree of agreement, and a score over 0.80 was an excellent agreement. The agreement between the scorers was calculated with the SPSS-25 program, and the kappa values based on articles were Cohen's kappa of 0.718 95% in the confidence interval [Confidence interval: 0.645-0.986)]. The general agreement rate kappa value in this study was calculated as 0.718, and the reliability was high.

Analysis of Data

"Comprehensive Meta-Analysis Academic/Non-profit Pricing (Version 3)" was used. "Cochran's Q statistic" was used for heterogeneity. In the heterogeneity assessment, if the heterogeneity rate (fI²) is below 25%, it does not exist; 25-50% is low; 51-75% is moderate, and above 75% is considered high [18]. “Random effects model was used in group analyses with p ≤0.05 in the heterogeneity test for overall effect sizes, and fixed effects model was used in analyses with p>0.05”."RR and OR" values were taken as the basis for evaluating the overall effect size in the binary data analyses. In evaluating the overall effect, the limit of statistical significance was accepted as p≤0.05. “Classic Fail-Safe N and Tau coefficient calculation results were used to test the publication bias”.

Research Findings

A total of 856 research articles were reached. Ten studies that met the research criteria were reviewed. Four of these studies are cross-sectional, five are descriptive and one is descriptive correlation type.

Evaluation of Methodological Quality

Kappa statistics [19] and quality assessment criteria consisting of 12 items were used by Polit and Beck for the quality assessment of all studies. All articles were evaluated separately and received "1 point" if each item was fully met and "0" if not. It was evaluated as an article with a weak quality of 0-4 points, medium quality with 5-9 points, and robust quality with 9-12 points [15]. As a result of the evaluation, the highest score was 11, and the lowest score was 7. Studies with a total score of 12 are quality studies. Seven of the studies (n=10) whose quality was evaluated by the independent evaluators included in the study were rated as "strong" and three as "moderate." Ten studies that met the evaluation criteria and were of good quality were included in the study. The chart showing the quality evaluation of the 10 studies included in the meta-analysis and the distribution of the scores they received from each domain is given in Table 1.

The article was evaluated by three authors independently. The "Cohen kappa" statistic was used to evaluate the scores given by the researchers. The agreement between raters was calculated with the SPSS-25 program, and the kappa values based on the reports were in the 95% confidence interval of “Cohen's kappa 0.718 (Confidence Interval: 0.645-0.9869”. In this study, the kappa value showed good agreement between raters [20].

Effect Sizes and Heterogeneity

These research results, individual innovative features in nursing and innovation; innovativeness was the covariate for all leading, questioning, and skeptical articles and “Cochran's Q” statistic was used to test for cross-study heterogeneity. As a result of the heterogeneity test, the p-value was found to be “less than 0.05, and the Q value” (6093,496) is more remarkable than the value corresponding to the df value. All articles are heterogeneous. "I²" statistic value 99,360 The overall effect size was evaluated according to the random effects model.

Analysis according to the random effects model showed that the overall effect size of individual innovativeness traits in nursing was 0.080 (CI; 0.029-0.217; p<0.05) Table 2. According to the confidence interval results; if we repeat this study with 100 different samples selected from the same population, the odds ratio value we calculated in 95% of them could be between 0.029 and 0.217.  The odds ratio value is 0.080 <1. Although the calculated odds ratio value is statistically smaller than 1, it is significant (P=0.000).  The nurses' characteristics of making individual initiation have a decreasing effect on general initiation. H1 hypothesis is rejected.

   ^askeptical of innovation, ^b pioneer innovation, ^c innovation questioner, ^d innovative, Odds(R)>1.000, p<0,05

The random effects model was used in Table 3, which shows the effect size of the study for the individual innovativeness characteristics variable. According to the results of the analysis made according to this model, it was concluded that the overall effect size of individual innovativeness characteristics in nursing was 0.080 (I.A; 0.029-0.217; p < 0.05), which was not statistically significant. This is thought to be due to the limited number of studies on this subject in the literature.

Reliability and Validity of the Research

Funnel plot, Rosenthal's Safe N method, and Orwin's Safe N method were used to demonstrate that the meta-analysis study was reliable and valid and to determine publication bias. The effect sizes of 10 studies examining nursing and innovation were evaluated according to the funnel scatter plot. If the effect sizes of individual studies are distributed symmetrically in the funnel plot, it does not cause publication bias; Although it is distributed asymmetrically, it causes publication bias [21]. In line with this information, when examined in Figure 2, it can be said that the effect sizes of the studies are distributed on the graph close to a symmetrical shape. This distribution indicates the absence of publication bias.