As the healthcare environment has become more complicated and detail-oriented and health professions have become more advanced, more nursing professionalism has been expected in recent years. To be more competent, nurses should be critical thinkers who can effectively cope with advancing technologies, human resource limitations, and the high level of acuity required in diverse healthcare settings. Critical thinking (CT) is considered to be a crucial element for clinical decision-making by nurses, and improved empowerment to engage in CT is considered to be a core program outcome in nursing education. However, recent studies have reported difficulties in applying CT to nursing practice [1,2], moderately low levels of CT among nurses [3], and differences in the understanding of the meaning of CT among nursing educators and scholars [4].

Since CT was emphasized as an essential component of the nursing process in the 1970s, numerous nursing scholars have attempted to define the concept of CT for nursing [5]. During the introductory period of CT, intellectual or cognitive skills were mostly emphasized. A decade later, affective disposition was also noted as an important component of CT in the context of a caring relationship [6]. Emotional involvement enables nurses to genuinely feel the suffering and pain that patients experience [7]. In 2000, Scheffer and Rubenfeld [8] identified essential components of CT, including 10 affective habits of the mind and 7 cognitive skills, by using the Delphi method to arrive at a consensus on an acceptable definition of CT. In recent years, nurses have been increasingly expected to develop both CT affective dispositions and CT cognitive skills [9]. Affective dispositions such as being open-minded, inquisitive, and seeking truth can stimulate an individual towards using CT through a reasoning process [10]. Meanwhile, cognitive skills may help nurses analyze their inferences, explain their interpretations, and evaluate their analyses [11]. Knowledge is also necessary to strengthen and support the cognitive process of CT [3,10].

To our knowledge, the most recent scoping review on the concept of CT in the nursing field was reported in 2015 [5]; according to this comprehensive review [5], there was growing interest in the study of the concepts and dimensions of CT experienced by nurses and nursing students, as well as in the development of training strategies for both students and professionals. However, a direction for further research into CT among clinical nurses was to specifically focus on its features or tendencies and changes in the CT phenomenon as time goes by, because confusing perspectives and poor knowledge of CT among nurse-educators can threaten the nursing profession [12]. Furthermore, an extensive review of quantitative research findings on CT among nurses is lacking, since only a scoping review was published in 2015 [5]. Thus, it is necessary to better understand how clinical nurses exercise CT to cultivate their clinical decision-making skills by reflecting on the contemporary nursing context.

The purpose of this study was to obtain a more comprehensive understanding of CT in the clinical nursing context. In this review, we specifically addressed the following research questions: what are the levels of CT among clinical nurses?; what are the antecedents of CT?; and, what are the consequences of CT?

This study did not have human subjects; therefore, neither institutional review board approval nor informed consent was required.

A narrative literature review was used. We followed the methodologies described by the Center for Reviews and Dissemination for undertaking reviews [13] and by Petticrew and Roberts [14], who addressed the practical guide as an alternative to systematic reviews in the social sciences, since our major goal was to synthesize the individual studies narratively and not to evaluate the efficacy and safety of interventions or programs.

We evaluated the included studies using assessment sheets prepared and tested by Hawker et al. [15], who developed an instrument that is capable of appraising methodologically heterogeneous studies. The data extraction sheet explores 9 components in detail: title and abstract, introduction and aims, method and data, sampling, data analysis, ethics and bias, results, transferability or generalizability, and implications and usefulness. In our review, each of these areas was assessed using the criteria developed by Hawker et al. [15] and rated on a scale of 1 (very poor) to 4 (very good). The scores for each assessment were then summed to obtain an overall score and rating, which ranged from very poor (9) to very good (36). Any article scoring less than 18 was considered to be of poor to very poor quality.

Using the assessment described above, the selected studies had scores ranging from 23 to 33 out of 36. Hence, all studies were included in the review (Table 1). All studies mentioned either a research question or an objective. In each article, the study design was described. Procedures or interventions were described in all studies, including three quasi-experimental studies [16-18]. Random sampling and purposive sampling were most commonly employed. Some studies, however, failed to state their sampling methods. Five studies determined the sample size by using power analysis [16,19,20], the Raosoft sample size calculator [21], or Solvin’s formula [22]. All studies addressed ethical considerations except for 1 study [23]; however, no studies described or elaborated on whether the researchers had received permission to use an original or translated version of the research instruments.

The data abstraction and synthesis process consisted of re-reading, isolating, comparing, categorizing, and relating relevant data. Included articles were read repeatedly to obtain an overall understanding of the material. Relevant data were gathered and classified into 3 categories: levels of CT, antecedents of CT, and consequences of CT.

Our review included 13 publications (Table 1). The studies were conducted in 7 different countries: Korea and the United States (n=3, for each country), Spain and Taiwan (n=2, for each country); and Malaysia, Turkey, and Egypt (n=1, for each country). The research settings were hospitals (n=6), intensive or critical care units (n=3), acute care units (n=2), and psychiatric care units (n=1). One study included nurses working in critical care and emergency units [24]. In all studies, the sample consisted of only nurses.

The methodological features of the included studies are summarized in Table 1. Twelve studies implemented quantitative research to examine the phenomenon of CT, while 1 study used a mixed-methods research approach [18]. Six of the included studies implemented a quasi-experimental design to evaluate the effect of their programs on CT among clinical nurses [16-18,23-25]. All included studies described institutional review board approval, except for 3 studies, which either stated that researchers verbally obtained the consent of the participants to be included in the research [22,26] or did not mention this issue [23].

CT was evaluated employing the California Critical Thinking Disposition Inventory [21,22,24,26], Critical Thinking Disposition Inventory [18,19], Critical Thinking Disposition [17,20], Nursing Critical Thinking in Clinical Practice Questionnaire [27,28], Clinical Critical Thinking Skill Test [16], Health Sciences Reasoning Test (HSRT) [25], or a self-evaluation tool to measure 5 key indicators of the development of CT [23]. All studies, except for 1, utilized a validated version of the original instruments in the appropriate language or validated the instruments in their research [22]. All studies except for 3 reported internal consistency reliability [22,23,25].

All studies measured the levels of CT among nurses, except for 1 study [20] (Table 1). Clinical nurses in 4 studies reported low [26], moderate [22,27], and high [21] levels of CT. Chen et al. [19] reported that experienced nurses, with an average of 18.38 years of work experience, had higher CT scores than novice registered nurses did. Similarly, Zuriguel-Perez et al. [28] reported that the level of CT among more experienced nurse managers was higher than among other nurses.

Five studies showed that their developed programs significantly improved the levels of CT among nurses in the experimental group compared to nurses in the control group [16-18,23,24]. One study presented a significant increase in the mean overall CT score for the HSRT on the posttest using a 1-group pretest-posttest design [25]. In particular, 5 programs—a work-based critical reflection program [16], a scenario-based simulation training program [17], case studies with videotaped vignettes [25], and concept mapping [23]—had positive effects on CT levels among novice nurses. Zori et al. [24] reported significant effects of a reflective journaling exercise to strengthen CT dispositions among nurses with diverse work experience. Hung et al. [18] developed a problem-based learning program for mental health care nurses for 3 hours every week, for a total of 5 weeks (15 hours total).

Seven studies reported inconsistent findings regarding the influence of 1 or more sociodemographic variables on CT (Table 1). According to these studies, there were significant differences in CT across sociodemographic variables, including age [19,21,27,28], gender [21], ethnicity [21], years of experience [19,21,27,28], and educational level [21,28]. In 3 studies, older nurses, those with more clinical experience, or those with higher levels of education had higher levels of CT [19,21,28]. Although Ludin [21] reported significant differences in the levels of CT according to gender, ethnicity, and educational level, detailed information was not provided. In contrast, Mahmoud and Mohamed [22] reported that none of the sociodemographic variables or job characteristics had statistically significant relationships with the total CT disposition and, in 2 studies, there were no significant relationships between CT levels and educational level [26,27], years of experience [26], gender [27], or work units [27]. Nurses had higher levels of CT when they had higher levels of self-reflection [19] and lower levels of perception of barriers to research use [20].

Only 1 study investigated the consequences of CT [20], and found that CT disposition of nurses positively influenced evidence-based practice (Table 1). In this study, Kim et al. [20] found that the relationship between barriers to research use and evidence-based practice was mediated by CT disposition.

Studies from 6 different countries were included. Most of the studies were done in Asian countries; only 2 of the studies were conducted in Europe. Synthesizing and integrating data from different countries and cultures is a complex and challenging task [29], especially since differences in cultural attitudes on CT extend beyond our expertise. The restricted professional autonomy perceived by nurses, which impeded CT, may be different for each culture or country. For instance, several studies in Asia have reported that nurses lack or have limited authority in providing care for their patients [30,31]; furthermore, while CT allows nurses to generate new ideas quickly, become more flexible, and act independently and confidently, the scope of their action is still ultimately limited by the physician’s clinical decisions [10]. Thus, several Asian nursing scholars have stressed the growth of professional autonomy among nurses through exercising higher levels of CT as an area that needs support to improve nurses’ clinical competence.

In the studies we reviewed, 7 different instruments were used. Due to this diversity of instruments, it was difficult to compare and integrate quantitative data. In addition, some studies utilized instruments without testing reliability and validity; thus, it is recommended to validate CT assessment instruments used in future research to ensure their reliability. A large range in sample sizes and response rates, possible non-responder bias, and validation of the instruments restricted to small populations limited the representativeness of our study results.

Although the assessment tools used to measure the level of CT varied across the studies reviewed, the level of CT was mostly moderate or high among the nurses evaluated. This may be partly due to the emphasis of CT in nursing education in recent years; furthermore, CT is now recognized as an essential competency among nurses and is required for the accreditation of nursing education [9,32]. However, this result contrasts with other research that reported a low level of CT among nursing students [33]. Further research is needed to verify the differences between nurses and nursing students according to factors influencing CT disposition and skills.

Our review complements the results of a previous review that scoped the concept of CT in the nursing field [5]. For instance, as antecedents of CT, the association between sociodemographic variables and CT can only be revealed by quantitative studies. It is necessary to examine the relationships between them in the future since the influence of sociodemographic variables on CT was found to be inconsistent in our study, except for years of work experience, which showed a consistent association with CT capacity. This finding may be associated with the significant experience gained by more senior nurses, which complements their theoretical knowledge and clinical decision-making [34] and enables them to be capable of better reflecting on past experiences, which may foster a deeper understanding of the situation [19]. On the contrary, less-experienced nurses had difficulties in exercising CT because of their perceptions of a gap between theory and practice with reference to their education and the real workplace setting [16]. Thus, it can be useful for senior nurses to share and reflect on their successful experiences of applying CT for patient care through group discussions; meanwhile, for novice nurses, a clear and detailed approach on exercising CT to reduce the gap between theories and the clinical setting may be beneficial. For this reason, a tailored education program on CT should be developed according to nurses’ years of work experience.

Self-reflection was also significantly related to CT among nurses in our review. This finding can be explained in terms of genuine self-reflection which can help them develop their CT dispositions and skills by balancing a lack of confidence and professional autonomy [34]. CT encourages nurses to generate new ideas quickly, be flexible, and act independently and confidently [29]. In contrast, nurses’ CT becomes more limited when they are more dependent on physicians’ clinical decisions. Meanwhile, expert nurses are not confined to or constrained by theoretical knowledge and are able to interpret situations by actively utilizing their nursing care experiences with past patients through exercising CT in their decision-making process [9,40]. To promote and encourage CT, nurses need to be more independent, confident, and responsible. As nurses’ autonomy develops, the need to think critically is further promoted [29,33]. However, nurses in some nursing environments have reported limited or restricted professional autonomy due to existing rigid and hierarchical cultures, as well as physician-centered paradigms in hospitals, which can hinder nurses from exercising CT [35-38]. More research is required regarding autonomy and CT among nurses in relation to their perceptions of the organizational atmosphere.

Our review revealed that there is limited empirical research on the consequences of CT, since only 1 of the included studies investigated the consequences of CT. CT stimulates nurses to explore related knowledge and establish priorities for solving patients` clinical problems [39]. As a method of assessing, planning, implementing, reevaluating, and reconstructing nursing care, a CT approach encourages nurses to challenge established theory and practice [5,9]. In addition, good clinical judgement results from exercising CT by advancing nursing competence in contemporary healthcare environments, where the complexity of data and amount of newly developed knowledge increases daily [10]. Papathanasiou et al. [9] emphasized that nurses’ ability to find specific solutions to certain problems is easily achieved when creativity and CT work together. Nevertheless, an integrated review found no relationship between CT and clinical decision-making in nursing [40]. Further research is recommended to explore the consequences of CT in nursing.

Despite complementing the findings of a previous scoping review [5], our review has 2 major limitations. First, a more synthesized approach should be attempted, including both quantitative and qualitative studies, in order to facilitate a more in-depth examination of our research topic. Second, access to all resources via electronic databases was not possible and only studies written in English were included in our review.