This study was a cross-sectional study on nursing students enrolled in three medical schools in Yunnan province using the STROBE guideline for cross-sectional studies. All theoretical courses through online learning of this semester during the epidemic prevention and control from March to July 2020. E-learning methods include the live, record, or sometimes both. It involves the online platform of Rain classroom, Tencent Meeting, and Zoom app. It is essential for implementing online education with computers, virtual reality devices, mobile phones, and personal digital assistants. Institution of higher education has conducted training on the use of the online teaching platform and online collective preparation of instruction to teaching staff. Teachers adjust the curriculum plan in time according to the characteristics of online teaching. The teaching and research section conducts a rational examination and evaluation system for E-learning courses.

Participants were nursing students enrolled in three medical schools in Yunnan Province. Students who met the following criteria were included in this study: (1) currently studying in the school of nursing; (2) participated in E-learning from March to July 2020; (3) participants had no cognitive impairment or obvious language barriers and provided their informed consent to participate in this study. International students were excluded. Inclusion and exclusion criteria were stated in the content of the invitation to participate link.

The questionnaire consisted of three sections: sociodemographic characteristics, learning adaptability, and professional identity.

The sociodemographic characteristics form was designed based on previous studies and the opinion of two nursing specialists, such as age, gender, educational level, residence, household incomes per capita (CNY), duration of classes in a week, preferred teaching method, and network environment.

E-learning adaptability was measured for the undergraduates' learning adaptability questionnaire (ULAQ),and compiled and validated in China by Xu (17). The questionnaire consists of 55 items and two dimensions of learning motivation and learning behavior. Each item was rated on a 5-point Likert scale ranging from 1 (not at all) to 5 (exactly right), such as “1” is poor, “2” is middle and lower, “3” is middle, “4” is middle and upper, and “5” is excellent. Total and two dimensions scores were calculated by taking the average across all items within each dimension. The higher the score, the better the learning adaptability of nursing students. This questionnaire has good reliability and validity. The internal consistency coefficient of total scores and two dimensions were 0.891, 0.789, and 0.864. The correlation coefficient between total scores and two dimensions was 0.783 and 0.893. ULAQ was used to assess the learning adaptability among college students in previous studies (18, 19).

Professional identity was measured using the Professional Identity Questionnaire for Nurse Students (PIQNS), developed and validated by Chinese scholars (20). The questionnaire consists of 17 items, with five dimensions respectively: professional self-image (six items), the benefit of retention and risk of turnover (four items), social comparison and self-reflection (three items),independence of career choice (two items), and social modeling (two items). Each item was rated on a 5-point Likert scale ranging from 1 (not at all) to 5 (exactly right). Total and each dimension scores were calculated by taking the average across all items within each dimension; a higher score indicated a higher level of professional identity. The Cronbach's alpha and the split-half reliability of PIQNS were 0.827 and 0.842, respectively.

As a result of the pandemic outbreak, we conducted a web-based online survey to reduce face-to-face interactions. Therefore, data were collected through a professional online questionnaire platform. After they signed written informed consent, a text message containing the URL of the online survey was sent to all the students. At the start of the online study, we validated our participants by asking them about their major and college. Participant information was electronically encoded for data storage, and the computer of researcher was password-protected to prevent unauthorized access. The system would remind respondents of missing responses before submission, and only full completed questionnaires were allowed to submit. A total of 600 questionnaires were received in this online survey, and 15 invalid questionnaires with incorrect information were excluded. The remaining 585 valid questionnaires were valid, with an effective rate of 97.5%.

All statistical analyses were performed using SPSS Statistics 25.0 (IBM Corp, Armonk, NY, USA). Data were expressed as means ± SD. We used Kolmogorov–Smirnov to assess the normality of the distribution of continuous variables. Descriptive statistics were used for the participant characteristics of E-learning adaptability and professional identity among nursing students. Independent samples t-tests and one-way ANOVA tests were conducted to further analyze the ULAQ scores by sociodemographic variable. Correlation analysis was used to determine the relationship between total dimensions and each dimension of professional identity and E-learning adaptability. The magnitudes of correlations are classified as follows: ≤ 0.25 very low; 0.26 ≤ r ≤ 0.49 low; 0.50 ≤ r ≤ 0.69 moderate; 0.70 ≤ r ≤ 0.89 high; and 0.90 ≤ r <1 very high. Multiple linear regression was used to analyze the influential factors of E-learning adaptability. All tests were two-tailed, with a significance level of p < 0.05.

A total of 585 nursing students were included in this study, and the average score for age was 20.17 ± 1.38. Participant characteristics are presented in Table 1. Most respondents in this study were women (89.91%) and undergraduate students (52.48%). Approximately 77.09% of respondents live in rural areas, and 63.93% reported a poor network environment. Household incomes per capita (CNY) of respondents mainly were around 1,000–3,000 yuan (57.78%). The duration of classes in a week of most respondents was 15–25 h (38.63%), followed by more than 25 h (33.85%), and <15 h (27.52%). About 70.77% of respondents preferred live as their teaching method in E-learning.

Data shown in Table 2 clearly illustrated the E-learning adaptability among nursing students. Gender (p = 0.146) and household incomes per capita (CNY) (p = 0.159) were no statistical significance with E-learning adaptability. ULAQ score was significantly associated with the educational level of respondents (p < 0.001), residence (p < 0.001), duration of classes in a week (p < 0.001), preferred teaching method (p < 0.001), and network environment (p < 0.001). The highest ULAQ scores were from undergraduate (3.65 ± 0.47), living in urban areas (3.81 ± 0.67), 15–25 h courses per week (3.68 ± 0.41), combination teaching methods (3.76 ± 0.60), and stable network environment (3.58 ± 0.55) among the Chinese nursing students. Post-hoc analyses (Appendix Table 1) result from the comparison showed that 15–25 h courses per week scored better than <15 h (p < 0.001) and more than 25 h (p < 0.001), and <15 h scored better than more than 25 h (p = 0.007). A combination of both scored better than record (p = 0.001) and live (p = 0.002). Compared with the record (3.34 ± 0.57), live (3.40 ± 0.46) had higher ULAQ scores, but the difference was not significant (p = 0.499).

Appendix Table 2 represents the dimensions and total scores of E-learning adaptability and professional identity for nursing students. The average score of E-learning adaptability was 3.41 ± 0.51, including two dimensions of learning motivation (3.34 ± 0.56) and learning behavior (3.42 ± 0.52). The average score of professional identity was 61.60 ± 10.06, it had five dimensions according to the ranking of total scores as follow: (1) professional self-image (22.5 ± 4.03), (2) benefits of retention and risk of turnover (13.54 ± 2.88), (3) social comparison and self-reflection (11.28 ± 2.15), (4) independence of career choice (6.39 ± 1.51), and social modeling (7.85 ± 1.60). There was a positive correlation between total dimensions and each dimension of professional identity and E-learning adaptability (r = 0.316~ 0.505, p < 0.001), which is shown in Table 3.

Multiple linear regression analysis is shown in Table 4. The E-learning adaptability was a dependent variable. The independent variables included the educational level, residence, duration of classes, preferred teaching method, network environment, and professional identity. These were significant predictive factors on the E-learning adaptability of nursing students (F = 85.327, p < 0.001).

With the transformation of teaching methods during the epidemic, the adaptability of online learning of college students has received extensive attention and research. In this study, the E-learning adaptability of nursing students was at a medium level, based on the total average ULAQ score (3.41 ± 0.51). Similar results were observed with Saudi Arabia (21) and Egypt (22) studies, which indicated that about half of the students showed a positive attitude toward E-learning and had some barriers. The effectiveness of e-learning programs in developing countries scores lower than in developed countries among medical students (23). Developed countries may have built and adapted the technological infrastructure for the transition of face-to-face to digital education. In contrast, developing countries are not fully prepared for the shift. The integration of information and communication technology in teaching and learning is still early in the education systems (24). Arthur-Nyarko (25), DePaul (26), and other researchers also reported that unstable network connections and the lack of computers in homes in rural areas are important factors affecting the progress and quality of online learning. The results are in keeping with our findings that nursing students who live in rural areas and have unreliable network connections have lower ULQA scores. At the same time, undergraduate, 15–25 h courses per week and combination of live and record among nursing students have better E-learning adaptability. There were no significant differences in gender and household incomes per capita. Almaiah (27), Singh (28) reported similar results. However, some suggest that affluent family is an essential factor of E-learning feasibility for nursing students, and women showed a more positive response toward E-learning than men in previous studies (29). Therefore, it is necessary to evaluate further and analyze the influential factors of the E-learning adaptability among nursing students. Based on the analysis results, strategies to improve the E-learning adaptability of students will be formulated.

There was a positive correlation between E-learning adaptability and professional identity, and the higher the professional identity, the higher the E-learning adaptability. In the regression model, professional identity is an essential predictor of E-learning adaptability among the Chinese nursing students during COVID-19. Existing research showed that professional identity and learning burnout were negatively correlated (30). More than 60% of Chinese college students are often unclear about their future career goals and lack active academic participation (31, 32). Students often face difficulties due to the lack of a good learning attitude, such as the lack of self-control and indiscipline when self-isolated at home (33). Therefore, by helping the formation of professional identity in nursing students, learning adaptability will possibly be more developed in them. Professional identity is considered to construct and deconstruct through nursing education continuously (34, 35). It is suggested that colleges and universities pay attention to career guidance of nursing students, improve the online career planning curriculum, and guide nursing students to set up the correct values of career choice in the teaching management process. Additionally, teachers should help students to set up the learning goals of professional courses and complete online learning courses in a planned way. On the other hand, it strengthens the cultivation of professional identity among nursing students. Through guiding students to understand the value of nursing work in the learning process, stimulating nursing students demands for academic knowledge, and thus significantly promote the E-learning adaptability of nursing students.

Previous studies have shown similar findings that the higher the educational level, the higher the E-learning adaptability of nursing students. Lu (36) shows that the higher the educational level, the better the online learning self-efficacy and deep learning level. Undergraduate and junior college nursing students have different self-learning abilities, and different coping styles in solving significant learning problems might be one of the reasons causing the difference in E-learning adaptability. Similarly, Zhao (37) reported that postgraduates have better overall online learning adaptability than undergraduates due to stronger independent innovation thinking and more precise knowledge needs. Another study found that the learning burnout of junior college students is higher than undergraduate students (35). Therefore, it is necessary to fully evaluate the current situation of the problems and needs encountered by different educational background groups in the progress of online learning and to apply online tests intellectually to teach students according to their aptitude.

We observed that the residence and network environment were essential predictors of E-learning adaptability. Compared with living in the city, limited internet access, such as an unstable network environment and a lack of computers in homes in rural areas, affected the E-learning progress. Students from cities have better availability of Wi-Fi routers and dedicated rooms at home and better proficiency in computer and internet usage than those in towns and villages, which determines the feasibility of e-learning (38). Previous research found that inadequate access to technology, studying materials, and computers can leave students marginalized and anxious, affecting online learning (39). At the same time, students are more satisfied with the teaching method of combination of live and record, which reduces the problem of poor network connection to a certain extent and helps students to after-school review (40). It is recommended that improving the Information and Communication Technology infrastructure is warranted in the long run. Meanwhile, using and developing the offline curriculum resources is essential in enhancing E-learning adaptability. Colleges and universities should assess the difficulties in E-learning from rural nursing students so that alternate study plans and strategies can be worked out in advance.

It was of great interest to discover that duration of classes can be used as one of the significant predictors in E-learning adaptability. Similar findings to those in a previous study found that most of the students prefer the duration of each class to be capped at a maximum of 45 min, which could be attributed to the maximum attention span, a student can have (41). Previous studies have shown that health issues are more common in those students who have to attend classes for >4 h a day and have long courses. It may result from prolonged exposure to digital gadgets, which is associated with various ailments, such as digital eye strain and cervical spondylosis (42). Therefore, it is crucial that colleges and universities set aside guidelines for online teachings, such as the number of classes per day, length, and break between classes. Time-to-effort optimization and inclusion of practices, such as adopting creative and blended learning will make E-learning more effective.