2022 年 1 巻 2 号 p. 136-159
Objective: This study aimed to develop a concept that satisfies the element of “principles that are true to the clinical situation, up-to-date, and with a simple structure” in genetic and genomic nursing competencies.
Methods: Phase 1 involved a literature review and qualitative interviews with a focus group of nurses practicing genetic nursing. In Phase 2, nurses from randomly selected facilities in Japan answered a questionnaire based on the Phase 1 results. The integration of the results involved comparing the structure and content, and restructuring the concept.
Results: Nursing practices, consisting of seven structures, were extracted from 41 documents and interviews with 21 nurses. Ultimately, 88 items were used to develop the questionnaire. A survey of 137 participants revealed that 54 items had a practice rate of < 80%. Fourteen cluster structures were identified through factor analysis. Using meta-inference, the structural incongruity obtained from these qualitative and quantitative data may be due to a mixture of noise attributes resulting from infrequent practice. In total, 54 items with <80% clinical practice and 22 items whose similarity was confirmed by IT correlation were deleted. The remaining 12 items had a one-factor structure (Cronbach’s α = 0.903), and seven structures in Phase 1 were covered. Finally, factor loadings of 0.47–0.79 were determined as attribute weights, specified as an attribute of the concept, in line with clinical practice.
Conclusion: Mixed methods have helped in developing the concept of “clinical competency in oncological genetics and genomics nursing,” which combines the characteristics of intensity.
“What they do and how they do it depends upon the concepts through which they see” (Pitkin, 1967). Understanding a concept requires learning it naturally through daily experiences or systematic science (Vygotsky, 1956/2001). These experiences must be practical, particularly for learning concepts in the professional industry.
Monomethodology is the primary approach to concept development in the field of nursing in Japan (Masuno, 2013; Aoyagi, 2016; Imai et al., 2016; Nishina et al., 2019; Tanaka & Arakida, 2019). Concept validation is limited to a small number of situations, such as the development of a scale (Murayama, 2012), and the custom of concept development is problematic. Moreover, the validity of a concept is confirmed only by inexpensive and routine examinations (Yoshida et al., 2012). As a result, presently, concept development is performed using a one-sided method regardless of the type of concept, and whether the development method is tailored to the purpose is not considered important. However, given that human thoughts and behaviors depend on concepts, it may be crucial to prepare the first unknown concept that is encountered in an easy-to-incorporate manner. In nursing education, in particular, the refinement of concepts has been shown to bridge the gap between education and practice, leading to improved patient outcomes and clinical judgments (Scott Tilley, 2008). Thus, this study aimed to develop a practical concept to improve nursing quality, and we identified three issues in the field of genetic and genomic nursing.
Three Challenges in Concept Development in NursingIn Japan, nursing science, which evolved with the advancement of genome analysis technology, has attracted attention as a new research field. Presently, this field of nursing is called genetic and genomic nursing (American Nurses Association, 2016). Currently, in Japan, it is an elective subject for pre-nursing education (Tsuji et al., 2014), and attempts have been made to clarify the theoretical concept using the Delphi method, as well as a review of domestic and foreign literature (Arimori et al., 2004; Terashima, 2015).
First, in this field, there is a problem of unclear clinical suitability because nursing science rarely verifies clinical adaptation following concept development (Schwartz-Barcott & Kim, 2000; Uemura & Honda, 2005). In the United States, competency in genetic and genomic nursing is required for all nurses (American Nurses Association, 2006). However, a subsequent study reported that most nurses did not practice (Newcomb et al., 2019), raising the question of the need for this competency. Concepts that do not represent modern clinical practices have no clinical utility.
The second issue is the uncertainty of the concept of “up to date.” Concepts change over time (Association for the Study of Science, 1995); in genetic and genomic medicine, the assessment of clinical issues has not adapted to the rapid evolution of gene analysis technology (Inoue et al., 2014). For instance, in traditional genetic nursing, it is essential to heed to the resulting psychological consequences when a genetic mutation is found in a patient through genetic testing (Arimori et al., 2004). However, with the development of personalized medicine, this approach could not be applied in some cases. This is because genetic mutations can be useful determinants for determining treatment options.
Finally, there is an issue related to translation mismatches. Concepts are expressed in words (Walker & Avant, 2005/2008). As there are many languages worldwide, presenting concepts in each country’s language helps promote the understanding of complex concepts. However, there are very few genetic or genomic nurses in Japan; thus, it is necessary to create concepts using overseas literature. Therefore, at both the concept development and use stages, the interpretation is not as smooth as in the native language, increasing the risk of misunderstanding the concept. For instance, the English word “genetic” signifies inheritance and diversity. When this is translated into Japanese, it becomes “iden”; however, the term only includes inheritance in its meaning.
To address these issues, we first conducted a conceptual analysis using qualitative research (Moriya et al, 2019). For the purposes of this study, we believe that this concept is being promoted. However, we have not yet confirmed its clinical suitability, adaptability to the current time period, or ease of understanding. In other words, concepts in the field of genetic nursing are riddled with problems; for instance, they do not fit the clinical situation and do not evolve with time, in addition to being difficult to understand.
Significance of the Concept Consisting of Language and Numerical ValueHow can we develop a concept that is in line with the clinical situation, up-to-date, simple, and easy to understand, that is, a “practical concept” suitable for the clinical situation? To determine this, we focused on the expression of this concept. To date, the concept of nursing science has been examined and expressed in terms of language. However, if our horizons are expanded to other disciplines, this concept can be expressed in terms of numbers and symbols as well. For example, as part of engineering, a language processing research field develops a language association function used for web searches. In this field, it is necessary to express appropriate concepts of complex search words as search results. Therefore, when expressing concepts, we shifted from the conventional method of considering multiple languages to adding a quantitative viewpoint. Thus, a concept (meaning of a word) in this field is defined by an attribute (a word that represents the nature and characteristics of the concept) and a weight (a quantity that represents the strength of the relationship with the concept) (Kasahara et al., 1997). To put this into practical use, noise-attributed cut-offs and attributed weighting were performed (Kojima et al., 2002).
In this study, we focus on the concept development of oncological genetic and genomic nursing competency, which is yet to be developed in Japan. This concept can solve the abovementioned problems using a simple and practical model. Thus, we presumed that this model would be useful in correcting the gap between theoretical concepts that represent the ideal and clinical concepts being practiced. In this model, an attribute can be refined by eliminating the noise attribute based on the actual condition of the clinical situation. In addition, internal balance should be adjusted according to the actual situation by distinguishing the degree of explanation of the principal component by weight.
This study has four objectives: 1) to express the practical concept of oncological genetic and genomic nursing in Japan in a manner closer to the clinical truth through language and numerical values, 2) to make this concept easy to understand through a simple view and illustration in their native language, by medical personnel and the public, 3) to obtain basic information for creating useful measurement tools and developing teaching materials, and 4) to establish a practical way of developing concepts, especially the importance of assessing modifications, noise cut-off, and weighting, through mixed methods.
Exploratory Sequential Design to Create Practical ConceptsMixed methods research is optimal for generating the concept of oncological genetic and genomic nursing competencies with the qualitative and quantitative properties previously mentioned. Therefore, we assigned cut-offs for noise attributes and attribute weights, which are implemented in the language processing field and expressed using mixed methods research with an exploratory sequential design (Creswell & Plano Clark, 2007/2010), proposed as a concept development method in the field of health science.
This design verifies the hypothesis derived from the qualitative and quantitative data. Analyzing quantitative data with qualitative data helps “interpret how quantitative results provide new results, better instruments, and better interventions’ (Creswell, 2013). Typical and traditional examples of scales (Bichi & Talib, 2019; Biniaz et al., 2019, Keshavarzian et al., 2020; Taghipoorreyneh & Run, 2020) and the development of intervention strategies (Wu, 2015; Ramaraj & Nagammal, 2017; Firouzan et al., 2018; Moriya, 2019) have been presented. Recently, a mixed-methods arrangement model of grounded theory, which is a well-known qualitative research method, has been reported (Howell-Smith et al., 2020), and its application to instrument development has been proposed. Thus, it was necessary to search for unknown concepts and verify whether they were simple and suitable for the actual clinical situation; therefore, we used an exploratory sequential design. This study showed that an exploratory sequential design contributes to the problem of concept development, which is complicated and inconsistent with the reality of nursing practice. This study is part of an empirical methodological mixed methods research and aims to clarify nursing competency as a concept in a way that matches the actual situation due to synergistic effects.
PurposeIn this study, the purpose of empirical research was to develop the concept of practical oncological genetic and genomic nursing competency by converting a theoretical concept into a clinical concept. The purpose of this methodological study was to explain how mixed approaches apply to the concept development of practical oncological genetic and genomic nursing competency. This study consists of three phases.
The first is to determine the purpose of the qualitative method, which is used to clarify the theoretical concept of language for oncological genetic and genomic nursing competency (what is the theoretical concept of language in oncological genetic and genomic nursing competency?). The second phase determines the purpose of the quantitative method, which is to clarify the clinical reality, consisting of numerical values regarding oncological genetic and genomic nursing competency (What is the clinical reality in terms of numerical values regarding oncological genetic and genomic nursing competency?). The integration phase determines the purpose of mixed methods, which is to assess the necessity or non-necessity of concept renewal by comparing theoretical and conceptual structures with actual clinical structures. In the case of “incongruity,” it will be renewed to a clinical (appropriate denoising and weighting) oncological genetic and genomic nursing competency. Moreover, approaches have been applied to the concept development of practical oncological genetic and genomic nursing competency (to what extent are the qualitative themes concerning competency in oncological genetics and genomics nursing consistent with quantitative data?).
Operational Definition of TermsIn this study, the concept of clinical competency in oncological genetics and genomics nursing was defined as “principles that are true to the clinical situation, up to date, and have a simple structure.”
Theoretical FrameworkThis study uses a concept-based model as a framework (Kasahara et al., 1997). In this model, the concept consists of headwords, word groups (attributes) that represent the characteristics, and criteria (weights) that quantitatively represent the strength of association with the concept. That is, arbitrary concept A is defined using the following formula by attribute ai and weight wi:
A ＝ {(a1, w1),(a2, w2),(a3, w3),･･･, (aｋ, wｋ)}
In other words, the concept of cancer-related genetic and genomic nursing competency is a chained set with k attributes and weights assigned to each attribute. Figure 1 shows the concept of clinical competency in oncological genetics and genomics nursing in this study in two dimensions with attributes and weights. Kojima et al. (2002) demonstrated clues for increasing concept validation. These were as follows: 1) consider the appearance frequency of the attribute, 2) consider the degree of association between the concept and the attribute, and 3) consider whether the attribute contains the same kanji (ideographic characters) as the concept. These approaches are referred to as “noise attribute removal” and “attribute weighting.”
Figure 1
Image of the components of the concept "clinical competency on oncological genetics and genomics nursing" Note. This figure is a two-dimensional simplification of the conceptual base of the language created by Kasahara et al. (1997). It indicates that the elements of the concept have attributes and weights. The vertical axis represents the attributes. As seven clusters have been confirmed in the concept of cancer-related genetic practical nursing ability from previous studies, for convenience, we expressed that there are five or more k attributes in total. The horizontal axis represents the weight. The maximum influence in explaining the concept is expressed as 1.
This is a type of mixed-method research with an exploratory sequential design, which is a qualitative study followed by a quantitative study. This design collects qualitative data for prototypes of deliverables, and consequently uses quantitative data to refine deliverables. Figure 2 shows the diagram of the study. Phase 1 below was reported in our study (Moriya et al., 2019).
Figure 2
The study diagram
Note. This figure is a simplified diagram of this research process. Phase 1, qualitative research, aims to extend non-verbalized concepts. The Phase 1 results were used to create the survey. This step aims to create a questionnaire that covers cancer genetics/genomic nursing. Phase 2, quantitative studies, aims to understand the clinical situation using the questionnaire created based on Phase 1. Then, in the Integration Phase, Phases 1 and 2 results are combined by the method of structural and content comparison. Structural comparisons determine whether a conceptual structure can be explained equally by the two phases. Content comparison uses meta-inference to explore strategies that bring extended concepts up-to-date and make them simple and easy to understand. If a contradiction is found between the two structures, we aim to converge and reconstruct the extended concept. Finally, we derive “clinical competency on oncological genetics and genomics nursing,” enhancing the intensity of the concept.
Qualitative data were collected through a literature search and focus-group interviews. The specific period of the literature review was from April to June 2018. Excluding proceedings from 1965 to 2018 in the Japan Medical Abstracts Society web search, we extracted 369 references, excluding duplications such as “cancer (tumor)” AND “genome (genetic)” AND “nursing” via the search terms. We excluded 41 references.
The focus group interviews were based on a semi-structured interview guide. Recruitment was conducted from February to March 2018 at a large hospital providing traditional and systematic genetic and genomic nursing education. The focus groups were formed by five to six nurses who belonged to the same department (purposeful sampling), and the results of the simple analysis were confirmed by the focus group representatives of each department. This process was repeated until theoretical saturation was confirmed, and 21 individuals from the four departments were recruited. The interview guide included the experiences and prospects of oncological genetic and genomic nursing. Here, we roughly defined oncological genetic and genomic nursing competencies in Japan.
Qualitative data analysisEach of the collected papers and interview transcripts was analyzed according to the content analysis procedure in the Funashima model (Funashima, 2007). The raw data were segmented to the extent that they did not impair their meaning. The segmented data were made more abstract by clustering them in the order of nursing practices, titles, and clusters. At the end of the analysis, nursing practices that were representative of the sectioned data were labeled. The final process of integrating data from the two sources confirmed the theoretical saturation of the cluster and involved the exploration of the boundaries and opposition to the concept proposed by Walker and Avant (2005/2008). This concept was truncated except for genetic nursing competency (Arimori et al., 2004) and general nursing competency (Sato et al., 2007; Maruyama et al., 2011; Takase et al., 2011). In the draft, seven physicians in genetic medicine and four nurses exchanged opinions and approved of the results. Additionally, a clinical genetic physician and genetic counselor responsible for educating standard-level nurses at a cancer care facility approved the clinical adequacy of 88 nursing competency items.
Integrating the Qualitative into the Quantitative Phase: Development of the QuestionnaireIn this step, a questionnaire was created from the seven clusters of oncological genetics and genomics nursing practice competencies identified in Phase 1. The questionnaire was based on seven clusters of competencies and 88 nursing items. Interpretive verification was conducted as a collaborative effort among nurses responsible for oncology at a regional hospital. They confirmed the difficulty in understanding terms and scenes, ambiguity that values were divided among subjects, ambiguity due to questions with multiple meanings, and difficulty in reading due to the difference in abstraction. The choices were based on the meaning structure of the quantitative expression terms for the questionnaire survey (Takeya et al., 1992); ultimately, “Every time (5 points),” “Often (4 points),” “Sometimes (3 points),” “Rarely (2 points),” and “Never (1 point)” were established. The questionnaire included basic attributes such as age, sex, and academic background. We included questions regarding years of clinical experience in cancer-related nursing, specialty, position, and advanced qualifications. Responses to the questionnaire were completed only in Japanese. For the Japanese-English translation used in this study, a cross-check was conducted by the researchers in addition to having been professionally translated into English.
Phase 2: Quantitative Study Quantitative data collectionThe survey was conducted between August 2019 and February 2020. The participants were regular nurses who belonged to a hospital that collaborated on cancer treatment and had a relationship with cancer patients in their daily work. The facilities were selected using the stratified random sampling method from the list of 2019 cancer treatment cooperation locations comprising approximately 667 hospitals that were published on the Ministry of Health, Labor, and Welfare of Japan’s website. The survey was conducted with nurses at 116 consenting hospitals using stratified random sampling, and the hospital was added from Phase 1 to the list by the convenience sampling method. We used the internet to confirm consent to participate, which was determined by returning the questionnaire.
Quantitative data analysisThe statistical package IBM SPSS Statistics 26.0 was used for the analysis. In this analysis, basic attributes and descriptive statistical information for each item were obtained, and 88 items were used to determine the structure of the concept using exploratory factor analysis (principal factor method and maximum likelihood method).
Interpretation and Integration of Qualitative and Quantitative Data Comparison of structure and contentCombining Phases 1 and 2 results in structural and content comparisons. Structural comparisons determine whether a conceptual structure can be explained equally in both phases. Content comparison uses meta-inference to explore strategies that bring extended concepts up-to-date, simple, and easy to understand. A joint display is attempted to make it easier to interpret whether there is a conflict between the two structures and their content.
Reconstruction of the concept by noise cut-off and weightingFirst, we performed a noise-attribute cut-off. We excluded items for which ≥ 20% of the responses indicated that they were “not in clinical situations.” Of the remaining Pearson correlations of ≥ 0.65 among the remaining attributes, we excluded items for which the attribute meanings and kanji were similar to typical ones. As a criterion, we first retained independent kanji items, and if there were multiple items containing the same kanji keyword in the same cluster in Phase 1, items containing more keywords or those that are frequently used in clinical practice were adopted. To maintain consistency with Phase 1, one or more attributes from each cluster extracted in Phase 1 remain the basis for the cutoff. Finally, the factor structure, Cronbach’s alpha, and IT correlations were confirmed. Additionally, we determined the concept attributes by confirming that the items representing the conceptual structure of Phase 1 remained and conformed to the definition created in Phase 1. Subsequently, we performed a principal component analysis to weigh the attributes. We confirmed the concept’s stability using the retest method after two weeks based on the correlation between the two groups. Among the basic attributes, we conducted an independent t-test for “sex” and “qualification” and centrally assigned “age,” “education,” “years of clinical experience,” “years of clinical cancer experience,” “specialty,” and “position.” In the analysis of variance, the validity of the construct was confirmed by performing Tukey’s test, assuming equal variances. Finally, we confirmed the reproducibility of the definition created in Phase 1 at our researcher’s meeting.
Ethical StatementThe study conformed to the principles outlined in the Declaration of Helsinki and was approved by the Clinical Research Review Committee of the Tokai University School of Medicine (No. 17R214. Approved on December 14, 2017, No. 18R226, approved on January 16, 2019).
Through a review of 41 types of analyzed literature, 419 nursing practices were extracted, and seven were clustered into 64 titles. Regarding the basic attributes of the 21 nurses who participated in the interview, there were 17 women and four men, and the average years of nursing experience was 11.0 (SD = 8.5, min = 2, max = 30). There were three inpatient units (gynecology, respiratory, and gastroenterology) and one outpatient unit. The average voice data duration was 63 min (SD = 10.5). Through nurses’ narratives, 424 nursing practices were extracted and summarized into 49 titles. In the process of integration into the previous structure, the titles increased to 88; however, theoretical saturation in the seven clusters was validated, as confirmed by empirical verification. The seven clusters were as follows: A. use of genetic and genomic information (32 items); B. prevention of cancer, including family members (12 items); C. adjustment of genome-related resources (10 items); D. attitude towards diversity in individuals (10 items); E. fulfillment of basic responsibilities (10 items); F. acquisition of specific medical knowledge (12 items); and G. awareness of the contribution of genetic or genomic medicine (2 items). The definition of this concept was determined as follows: “Competency related to the impact of genetic and genomic information on the onset persons, non-onset persons, and their families who receive cancer treatment, and it is a competency that accepts evolving genetic and genomic medicine and creates care according to the current circumstances.”
Questionnaire Development and Pilot TestingSeven nurses participated in the pilot study and two issues were noted for this item. First, they did not understand terms related to genetic genomic medicine. In response to this, we discussed whether it might be possible to measure the ability of respondents by intentionally avoiding specialized technical terms related to genomic medicine, and ultimately decided to retain these items. Second, nurses indicated that this item was not relevant to clinical practice. In response, we confirmed that all respondents may have encountered items in clinical practice. However, it was assumed that some people were unable to answer because they could not perceive the object in question. Therefore, we added the item "Not in clinical situations" to the questionnaire. Through this process, we developed the principles of the questionnaire protocol.
Phase 2: Based on Quantitative DataOf the 117 facilities nationwide for which the survey was requested, cooperation was obtained from 17 regions other than the northernmost and southernmost. Of the 1,482 regular nurses involved with cancer patients at these facilities, 155 responded to the questionnaire (recovery rate: 10.5%). Of these, 137 responses were analyzed, of which 18 were excluded because of missing values of ≥ 40% (valid response rate, 88.4%). Most of the participants were in their 40s (37.2%) and the majority were females (94.2%). More than half of the participants (64.2%) had an educational background of a diploma level education. The units in which the participants worked were gynecology (27.9%), gastroenterology (24.3%), and others (47.8%). Table 1 presents the participants’ characteristics.
Some respondents indicated that not all items were relevant in a clinical setting, and the average number of respondents was 23.9 ± 9.5% (max = 44.9%, min = 7.5%). Among these, 20.0% or more respondents answered that 54 items were not relevant in a clinical setting. As a result of the factor analysis, both the main factor method and maximum likelihood method had a structure of 14 clusters. Based on the factor analysis, the 88 items were classified into 14 factors, and 85 of the 88 items had the highest contribution rate for the first factor, at 49.6% to 49.8% (no rotation, 49.6%; varimax rotation, 49.8%; promax rotation, 49.6%). Using the maximum likelihood method, they were classified into 14 factors, and 85 of the 88 items had the highest contribution rate for the first factor at 32.0% for no rotation, varimax rotation, and promax rotation.
Interpretation and Integrating Qualitative and Quantitative Data Mixed methods inferencesThe results of Phases 1 and 2 were combined by structural and content comparisons. In the structural comparison, we inferred whether the conceptual structure could be explained in the same way in both phases.
Comparing the conceptual structures of Phases 1 and 2, it was found that there were seven clusters in Phase 1 and 14 clusters via both the principal factor method and the maximum likelihood method in Phase 2. Table 2 shows the structural differences between the phases. There may have been incongruity because the conceptual structures of the two were different; thus, we determine that this concept includes noise attributes. Furthermore, the main factor method showed that 85 of the 88 items could be explained by one factor (contribution rate: 49.8%). Thus, we inferred that this concept can be explained by a single cluster.
The content comparison inferred why the “incongruity” occurred, and we examined strategies that make the extended concept the latest, simplest, and easiest-to-understand by meta-inference. This study has two points of interest. The first is the separation of nursing clusters dealing with molecular-targeted drugs. In Phase 1, this point was included in the utilization of genetic genomic information; however, in Phase 2, it was extracted independently. A feature in which the practice frequency was higher than that of other items was also noted. This suggests that molecular-targeted drugs may not be recognized as nursing using genetic genomic information, and that their use may be practiced in a state of separation from genetic genomic nursing without sufficient association. The second factor was the flattening of nursing. As a result, one cluster was formed from basic practice that can be sufficiently considered nursing, even with existing knowledge of applications that cannot be considered nursing without knowledge of the genetic genome and interest in family diseases. In other words, this nursing practice suggests that both traditional genetic and new genomic nursing are rarely practiced, which could be why only one cluster could explain these concepts. In these situations, complex concepts presented in many languages are impractical. Finally, we derived a concept suitable for clinical situations.
In the cutoff process, the response “not in clinical situations” was confirmed for all items. Thus, we deleted 54 competencies observed at a rate of ≥ 20%, leaving 34 items. The correlation coefficients between items ranged from 0.10 to 0.89. For convenience, a correlation coefficient ≥ 0.65, which was found for 30 items, was determined. Thus, we compared the items that were related to each other and deleted 22 items with similar kanji meanings and expressions. When applied to the seven clusters extracted in Phase 1, it was confirmed that all clusters were satisfied.
As a result, we deleted 76 items and reconstructed the remaining 12 items by factor analysis using the main factor method, calculated the eigenvalue as one, and confirmed one-dimensionality. None of the items had less than the load | .40 | for the first principal component, and the contribution rate of the 12 items was 49.2%. The item-total correlations ranged from 0.40 to 0.73. Cronbach’s α ranged from 0.89 to 0.90, with a total of 0.90.
The breakdown of the 12 items is as follows. A to G are the identification symbols of the cluster created in Phase 1. Questions in the questionnaire are given in parentheses. In “A. Use of genetic and genomic information” (32 items), “Assess a patient or a patient’s family by performing an overall cancer evaluation” (Q87 / A), “Confirm the response to molecular targeted drugs based on imaging findings” (Q35 / A), “Confirm that the treatment provided for patients with genetic issues is based on their values and idea of life” (Q16 / A), and “Pay attention to the psychological aspects of children who have parents with hereditary tumor” (Q5 / A) were selected. In “B. Prevention of cancer, including family members” (12 items), “Continuously update family history” (Q60 / B) was selected. For “C. Adjustment of genome-related resources” (10 items), “Understand the role of a certified nurse specialist in genetic nursing” (Q24 / C) was selected. In “D. Attitude toward diversity in individuals” (10 items), “Consider the living conditions of patients with genetic issues” (Q7 / D) and “Reflect on my prejudice against and discrimination of genetic information” (Q77 / D) were selected. For “E. Fulfillment of basic responsibilities” (10 items), “Determine if I can accept a task when I provide treatment for patients with genetic issues, based on my abilities” (Q75 / E) was selected. In “F. Acquisition of specific medical knowledge” (12 items), “Evaluate a type of variants in cancer, specifically, somatic or germline mutations” (Q57 / F) and “Understand how to obtain the latest information on genetic/genomic medicine” (Q79 / F) were selected. Finally, in “G. Awareness of the contribution of genetic or genomic medicine” (2 items), “Take action for the improvement of genetic/genomic nursing by collaborating with other members of the medical team” (Q85 / G) was selected.
Attribute weighting and validationAt this point in the process, the abovementioned 12 items and attributes of the concept were determined. For attribute weighting, the factor loading of the main factor method was referenced. The maximum factor loading was Q75 (0.79) and the minimum was Q60 (0.47).
Eleven of the 155 respondents cooperated in the retest method after two weeks. Of these, 10 had no missing values and were analyzed. The correlation between the first and second responses was 0.72, which was considered strong.
The basic attribute test targeted 128 individuals with no missing values. The weighted scores were significantly different in “qualified (p = 0.000),” “master’s degree (p = 0.001),” and “more than 11 years of clinical cancer experience (p = 0.007).” These 12 attributes included the target person, behavior, and intention, as defined in Phase 1. Figure 3 shows the integrated conceptual model of clinical competency in oncological genetics and genomics nursing in Japan. Finally, it was confirmed that it has the characteristics of “principles that are true to the clinical situation, up to date, and have a simple structure.” This is a complete version with noise cut-off and weighting, using the concept-based model as a framework.
Figure 3
Integrated conceptual model of the clinical competency in oncological genetics and genomics nursing in Japan
Note. This figure uses a concept-based model to represent cancer inheritance/genomic nursing. By comparing each structure, the conceptual structure was "incongruity" in two phases. Using meta-inference, we couldn't take advantage of the characteristics of the practice content because of the infrequent practice, and interpreted the better concept as a simpler structure. Therefore, we aimed to converge the concept expanded to 88 items and reconstruct it into a simple feature. Finally, we derived "clinical competence in tumor genetics and genomics nursing" and strengthened the concept. The vertical axis represents the attribute. Prior to the attributes, the questionnaire number(Q), cluster symbol（A-G) identified in Phase 1 are listed, and practice frequency (up to 5 points) identified in Phase 2 are listed. There were 12 types of attributes, and the contribution rate was 49.2%. This study's results show that oncological genetics and genomics nursing can explain half of concept with only 12 attributes. The horizontal axis represents the weight of each attribute. At the top of the figure, the factor load is close to 1, and nursing with a high degree of conceptual explanation is placed here. The factor loadings for each attribute are shown adjacent to the horizontal axis.
The concept developed by qualitative research consists of a noise attribute in which the ideal and reality are mixed to maximize sampling. The content comparison inferred why the “incongruity” occurred, and we examined strategies that make the extended concept the latest, simplest, and easiest-to-understand by meta-inference. This led to the development of a concept consisting of 12 types of attributes, each weighted from 0.47 to 0.79.
An integrated conceptual model of the practical concept of oncological genetic and genomic nursing in Japan was designed to match the actual conditions of clinical practice and expression. In the discussion, the practicality of this concept is described from the perspective of “Does it truly match the actual situation?”, “Is it a meaningful, simple expression?”, and “Was it appropriate to present it as a quantitative and qualitative concept?”
Truly matching the actual situationIn this study, we speculated that the reason why the structure of oncological genetics and genomics nursing is simple in meta-inference is that nurses do not fully understand and practice this type of nursing. This discovery confirmed that the original concept development policy, which required a simple concept, was appropriate. Attributes were weighed for clinical matching. The results were compared with those of a field survey conducted at a US regional hospital (Newcomb et al., 2019). In the United States, one of the lowest practice levels was “family tree construction,” and one of the highest practice levels was “physical assessment.” Replacing the former with the items in this study, Q60 was the closest, and the factor loading was 0.47, the lowest. The latter corresponded to Q87 with a factor loading of 0.78, indicating the highest performance. These results suggest that both the studies showed similar trends. In contrast, item Q24, which showed an association with the item focused on genetic experts and was assessed as being at a middle level in this study, was assessed as being at a low level in the US survey. It is speculated that this is because, in the United States, there is no genetic expertise system specifically for nurses (American Nurses Association, 2016), whereas, in Japan, such a system does exist. Thus, this result may be meaningfully related to the clinical reality.
Meaningful and simple expressionThe developed concept removed 76 noise attributes during the course of the study, resulting in a very simple structure. Such a simple structure carries the risk of treating the concept as ambiguous and unexplained because it is difficult to explain complex concepts. Chinn & Kramer (1995/1997) stated that when expressing an abstract concept, it is necessary to select a representative experience rather than emphasize ambiguity” to solve this problem. Therefore, validity can be verified by confirming that the attributes of a concept represent it. Comparing the attributes that ultimately remained in this study with the subscales of the Genetic Nursing Practice Ability Scale (Susaka et al., 2019), the attributes related to “Factor 1: ethics and attitude” were Q75, Q77, Q79, and Q85. The attributes related to “Factor 2: comprehensive understanding, were Q5 and Q7. The attributes related to “Factor 3: identification” were Q57, Q60, and Q87. The attribute concerning “Factor 4: decision support” was Q16. The attribute concerning “Factor 5: support for daily life” was Q35. The attribute for “Factor 6: collaboration, is Q24, which covers all factors. Therefore, the attributes of this concept may have a certain degree of representation. However, the number of clusters and titles was different from that reported in the United States (American Nurses Association, 2008). Thus, genetic and genomic nursing competency concepts appear everywhere, similar to weeds, and it is difficult to determine whether they are valid.
Appropriate to present the concept with quantitative and qualitative elementsBased on the above discussion, the integration of qualitative and quantitative data is useful for simply expressing truth. In this study, we proved that there is little difference in the degree of conceptual explanation among 88 items that best explain genetic and genomic nursing, and 12 items that provide proper convergence by combining data. The importance of this study is that it suggests that not only the continuous expansion of the concept but also the discovery of the axis of the concept and strengthening its intensity are important.
Attributing quantitative qualities to the concept was a unique attempt in the history of the development of nursing science showing that “the concept is expressed in language” (Walker & Avant, 2005/2008). When we write a concept in Japanese, it becomes “gai-nen.” “gai” is a Kanji character that represents the stick used to flatten the raised grain when pouring it into the trough and implies leveling the difference in quantity (Association for the Study of Science, 1995). Therefore, it is reasonable to assume that an essential concept includes both qualitative and quantitative meanings.
Contribution to Mixed Methods Research Meta-inference understood from the resultsTwo inferences were drawn from this study. First, the nursing that uses molecular-targeted drugs has different characteristics than the other types of nursing. Antineoplastic agents have been shown to be a clinical application of pharmacogenetics in that they act directly on cancer cells, but not on normal cells, and suppress the growth of these cells (Jenkins, 2005). Therefore, nurses are required to manage drug treatment after understanding somatic gene mutations, which can be considered a new form of genetic and genomic nursing. The reason it showed different characteristics compared to other types of nursing is that it is already used as a standard treatment in Japan (Japan Breast Cancer Society, 2018); thus, it is a common measure for use by nurses responsible for drug distribution. In this respect, it has good heterogeneity and can act as a breakthrough for improving genetic and genomic nursing abilities.
Second, there was no significant difference between the basic and advanced nursing groups. The cluster showing the basic responsibilities of nursing reflects the abilities of nurses who have completed basic nursing education, and the cluster showing the incorporation of blood relatives into preventive elements is a new ability not covered in the existing basic nursing education (Moriya et al, 2019). It was predicted that the reason the two were not differentiated was that they were practiced only at a low level. In Japanese nursing education, students’ lack of ability to apply their knowledge of anatomical physiology to clinical reasoning is regarded as a problem (Ministry of Health, Labor and Welfare, 2022). Therefore, even if the topic of human genetics is taught in basic nursing education, low clinical reasoning ability appears to have made it difficult to grasp this concept. Both results support the assertion that the concept can be simply presented at the current practical level of genetic and genomic nursing. In the future, when the practice of genetic and genomic nursing becomes more active and various characteristics appear, a new concept is likely to be required.
Comparison with monomethodologyThis study was challenging because it was a concept that included the term “nursing.” Nursing is one of the most abstract constructs in the structural hierarchy of modern nursing knowledge (Fawcett, 1993/2008). This indicates the difficulty of concept development: choosing abstract language to avoid complexity or using concrete language to prevent the risk of losing sight of the essence of the concept in question. In this study, the latter was selected because this concept was a nursing practice unfamiliar to clinical nurses. To avoid further complexity, we reduced the noise by comparing the qualitative and quantitative data. A hybrid model (Schwartz-Barcott & Kim, 2000) is a concept development model that aims to eliminate the dilemmas commonly encountered in clinical concept development. This method has been used in clinical and traditional concept development, for example, “the nursing presence” (Mohammadipour et al., 2017). In this method, the concept is defined using literature in Phase 1, clinical fieldwork in Phase 2, and an amalgam of the two in Phase 3, which is a methodological approach. As one of the authors, Schwartz-Barcott et al. (2002) reported, regarding the limitation of this method, the academic basis is weak in Phase 1, and the time and resource consumption are substantial in Phase 2 fieldwork. By contrast, in this study, academic vulnerability may have been avoided by incorporating fieldwork into Phase 1. Moreover, because Phase 2 involved a simultaneous survey, it is possible to reduce the cost by shortening the research process. In the traditional method, Phase 1 has an increased risk of becoming vulnerable in areas where there is limited literature, such as in the case of genetic and genomic nursing. Therefore, we recommend using the method developed in this study for concept development in understudied clinical fields.
Suggestions for utilizing the resultsIn the field of health science, including nursing, there are concepts that everyone understands empirically, such as height and weight, as well as highly abstract concepts, such as physical and mental health. As the behavioral patterns and communications in nursing practice addressed in this study lie in the middle of the continuity of these empirical and abstract concepts (Chinn & Kramer, 1995/1997), scales are usually used to measure them. Therefore, the method used in this study may be effective for the concept that is the target of scale development in nursing science. Thus, the results of this study are recommended for the development of a scale for clinical nurses to self-assess their ability to practice genetic and genomic nursing.
This study has two advantages over conventional scale development methods. First, it shows the number of clusters in a qualitative study to clarify the minimum criterion for the number of items to be deleted. Generally, it is impossible to remove more items than the number of clusters derived from qualitative research. Second, it helps avoid biased item selection. When a cluster that researchers do not anticipate appears, it is possible to infer the reason by using meta-inference. In other words, one can expect to reduce the risk of accidentally discarding items that are difficult to interpret. In addition, this approach is recommended for use when developing a scale that requires precise scores and considering intervention methods that require prioritization. This concept is expected to be used in the development of self-assessment scales for competencies and educational programs.
Limitations of the studyIn the process of obtaining these results, we did not attempt all the methods of concept development in nursing science in advance. In particular, this study did not discuss the antecedent or consequential requirements of the concept proposed by Walker and Avant (2005/2008). There are also research design limitations regarding the data collection periods and sampling. The literature collected through qualitative analysis had the highest percentage of commentaries because of the small number of previous studies, which affected the quality of the review. Although the interview data confirmed theoretical saturation in the facility, other factors may explain the concept, as the data were only collected from one facility. In addition, the samples of the qualitative and quantitative surveys were the same at only one facility and did not match completely. The target of the quantitative analysis was only 137 individuals because of the limited interest of the community in this field, which was inadequate for the 88-item survey. Another reason for the insufficient sample size may be that participants were unfamiliar with the online survey. In particular, the survey format, which cannot be answered without reading the QR code, may not be friendly to the target audience. This may be resolved by allowing the choice of bearing self-administered questionnaires and online surveys depending on the subject’s preference. In addition, although the sample age and educational background were similar to the employment rate in recent years (Ministry of Health, Labour, and Welfare, 2014; Japanese Nursing Association, 2017), cooperation between the northernmost and southernmost regions was not obtained, and the recovery rate was 10.5%. In other words, attribute cut-offs and weighting may be overly dependent on the target facility as a research environment. Moreover, in this study, the space of the concept that is not expressed may be increased by eliminating the attributes. However, because concepts have the property of not being fully expressed regardless of the number of attributes (Shoemaker et al., 2004), this is also a concept development feature that uses language.
We developed a concept of clinical competency in oncological genetics and genomics nursing based on the characteristics of intensity in actual situations in Japan. Using meta-inference, the structural incongruity obtained from these qualitative and quantitative data may be due to a mixture of noise attributes resulting from infrequent practice. Therefore, the concept developed by qualitative research needs to cut off the noise attribute, which is a mixture of the ideal and reality, and reconstruct it to have the characteristic of intensity by weighting. This led to the integration of a concept model consisting of 12 types of attributes, each of which was weighted from 0.47 to 0.79. Finally, it was confirmed that it has the characteristics of “principles that are true to the clinical situation, up to date, and have a simple structure.” Mixed methods have helped to develop the concept of “clinical competency in oncological genetics and genomics nursing,” which combines the characteristics of intensity.
AcknowledgmentsWe thank Nanae Kamogawa, Naoko Harada, Kei Takeshita, Yuko Ohnuki, Kazumi Takahashi, Mayuri Ohkami, and Mayako Terao for the data analysis. All the authors are sincerely grateful to the nurses who participated in the interviews and surveys.
IRB approvalThe investigation conforms with the principles outlined in the Declaration of Helsinki. This study was carried out with the approval of the Institutional Review Board for Clinical Research, Tokai University (No. 17R214. Approved on December 14, 2017; No. 18R226. Approved on January 16, 2019).
Author ContributionsHM provided research ideas, design, collection, analysis, interpretation, and integration of qualitative and quantitative data and full manuscript preparation. YY oversaw the research design and provided technical advice and article improvement for the overall research process and analysis. NY, HY, TU, and SI contributed to the study design, were responsible for some of the data collection and analysis and interpretation, and improved the manuscript. All authors have read and approved the final manuscript. The authors declared no potential conflicts of interest concerning the research and publication of this manuscript.
Declaration of Conflicting InterestsThe authors declare that they have no conflict of interest.
FundingThe authors disclosed receipt of the following financial support for the research, authorship, and publication of this manuscript. This work was supported by JSPS KAKENHI Grant Number JP17K18124 and JP20K10722.
