This study involved three stages as illustrated in Fig. 1.
2.1 Phase 1: Identification of dimensions and development of items
In the first stage, through literature review, theoretical analysis and qualitative interview, we constructed the first draft of the core competence evaluation index system of infectious disease specialist nurses which was composed of 6 primary indicators, 17 secondary indicators and 48 tertiary indicators, and adopted the 5-point scoring method. The primary indicators were Nursing Abilities for infectious diseases, Infection Prevention and Control Abilities, Responsiveness to Infectious Diseases, Professional Development Abilities, Communication and Management Abilities and Professionalism and Humanistic Accomplishment. Then, we invited experts in the field of specialized medical treatment and nurses of infectious diseases for Delphi consultation [16]. The inclusion criteria of consultation experts were as follows: (a) has engaged in clinical nursing or medical work of infectious diseases at least 15 years; (b) has intermediate level title or above; (c) has bachelor degree or above; (d) voluntarily participates in the research. Through Delphi expert consultation, the index system was scored and modified, and the judgment coefficient, authority coefficient and familiarity degree of Delphi experts were 0.933, 0.923 and 0.913 respectively. The core competence index system of infectious disease specialist nurses was finally established, which included 6 primary indicators, 16 secondary indicators and 47 tertiary indicators [17].
2.2 Phase 2: Preliminary evaluation and exploration of Infectious Disease Specialist Nurse’s Core Competence Scale
Then, through the panel meeting, we compiled the index system into ‘Infectious Disease Specialist Nurse’s Core Competence’ pretest scale which included 47 items. Before the formal investigation, the preliminary scale was distributed to 5 head nurses to test the level of item comprehension, appropriateness of the font size, survey structure and item length. The evaluation of the pre-test scale by five head nurses had good internal consistency, which was 0.851, indicating that it can be used for formal investigation. In the first round of investigation, 40 head nurses from the infectious diseases department were invited to evaluate the core competence of 516 infectious disease specialist nurses through the pre-test scale. Discrete trend, critical ratio, correlation coefficient, Cronbach's α coefficient and factor analysis were adopted for item analysis. Through item analysis and exploratory factor analysis, we filtered the items and drafted a preliminary scale, which was composed of 5 factors and 36 items. And the scale was basically consistent with the index system of core competence of infectious disease specialist nurses constructed in the previous study.
2.3 Phase 3: Evaluation of reliability and validity of Infectious Disease Specialist Nurse’s Core Competence Scale
In the third stage, we conducted the second round of questionnaire survey. The core competence of 497 infectious disease specialist nurses was evaluated by 42 head nurses with the first draft of Infectious Disease Specialist Nurse’s Core Competence Scale. We took a series of measures including item analysis, reliability test and validity test to filter the scale items and re-explore and verify the structure of the scale. The methods of item analysis were the same as above. Reliability analysis included test-retest reliability, internal consistency and split-half reliability. Validity analysis included content validity and structure validity. In the second round of questionnaire survey, about 10% of the subjects were randomly remarked, and the questionnaire was sent out again 2 weeks later to measure the test-retest reliability. Finally, a scale with high reliability and validity was formed, including 5 dimensions and 34 items.
2.4 Data collection procedure and quality control
Before conducting questionnaire investigation, the research team explained the research purpose and meaning to the head nurses and organized relevant training among them. In the process of evaluation, one-on-one evaluation was adopted, namely, one head nurse just evaluated one infectious disease specialist nurse at a time. The head nurses were responsible for the evaluation of core competence on all specialist nurses in the departments. When the evaluation was over, 10% of the nurses who were tested would be randomly selected and be evaluated again by senior nurses who are experienced in management in the infectious disease department. The senior nurses were also explained with the research purpose and meaning and trained as well. The same evaluation approach was adopted again. The results showed that, the correlation coefficient of head nurse and senior nurse to the certain specialist nurse’s core competence was 0.896 (P < 0.05). This demonstrated that the head nurse’s evaluation on the specialist nurse’s core competence was reliable, with low subjective bias.
The inclusion criteria of head nurses and senior nurses: (a) have engaged in infectious disease nursing for more than 10 years; (b) have nurse in charge title or above; (c) have good communication and expression skills; (d) voluntarily participate in research. The inclusion criteria of infectious disease specialist nurses༚(a) have engaged in infectious disease nursing for more than 5 years; (b) have participated in infectious disease specialist nurse training and got the certificate.
The sample size was determined by the general rule that factor analytic procedure requires a minimum of five respondents per item, but a larger sample is desirable [18, 19]. In our study, ten respondents per item were required to ensure the accuracy of factor analysis. Therefore, during the two rounds of questionnaire investigations,head nurses of infectious diseases department were selected by convenient sampling method to evaluate the core competence of infectious disease specialist nurses.
2.5 Statistical analysis
Data were analyzed by SPSS 23.0 and Mplus 8.3 software.
For item analysis, items were screened with the Classical Test Theory [20] which included discrete trend method, critical ratio method, correlation coefficient method, Cronbach's α coefficient method and factor analysis method. The standard deviation of item scores represented the degree of dispersion. When SD < 0.85, it indicated that the item was not able to distinguish and was to be deleted. The total score of the scale was ranked from high to low, and the relationship between the high-score group (the first 27%) and the low-score group (the last 27%) was analyzed to judge the discrimination of the scale. It was the same to the factor loading. If the total score was less than 0.4, the item needs to be deleted. If Cronbach's α coefficient became larger after deleting the item, it indicated that the item would lower the internal consistency of the scale and should be deleted [21].
Reliability analysis referred to the consistency of the results of repeated measurement of the same object by the same method [22, 23]. For reliability analysis, we used Cronbach's α coefficient, split-half reliability and test-retest reliability. Cronbach's α coefficient was used to evaluate the internal consistency reliability of the scale. The scale was divided into two parts according to the order of oddness and evenness, and the correlation between them was to calculate the split-half reliability. Two weeks later, we would conduct a test-retest on the nurses marked before, and measure the test-retest reliability.
Validity analysis referred to the analysis of the accuracy of the scale [24, 25]. For validity analysis, we conducted content validity analysis and structure analysis. The validity of the content was evaluated by Delphi experts' scores which included the content validity index of the items (I-CVI) and content validity index of the scale (S-CVI). The structure analysis contained exploratory factor analysis and confirmatory factor analysis. Index value standard: The Kaiser-Meyer-Olkin (KMO) > 0.6, χ2/df < 3, Root mean square error approximation (RMSEA) < 0.08, Comparative fit index (CFI) > 0.90, Tucker-Lewis index (TLI) > 0.90, Standard root mean-square residual (SRMR) < 0.80 [26].
2.6 Ethical consideration
Research was approved by the ethics committee of Tangdu Hospital of Fourth Military Military Medical University, China (Number TDLL2019-09-13). Informed consent was obtained from all participants included in the study and they could withdraw from the study at any time for any reason. Moreover, they were assured that the questionnaires would only be used for research.