A convenience sampling method was employed to conduct an online survey between November and December 2024 involving 263 oncology nurses from six tertiary hospitals in Guangdong Province. The inclusion criteria were: possession of a nurse practice license, a minimum of one year of oncology nursing experience, and informed consent with voluntary participation in the study. The exclusion criteria were: trainee nurses, and nurses who were on leave or attending off-site training and thus not on duty during the survey period. Based on the sample size calculation for multivariate analysis—which requires a sample size at least 5 to 10 times the number of independent variables, plus 20% to account for invalid responses—the calculated sample size ranged from 114 to 228 participants. Ultimately, 263 valid cases were included in this study. Ethical approval was obtained from the hospital ethics committee (B2024-127-01).

2.2.1. General Information Questionnaire

Following a review of relevant literature, the researchers independently developed the design, which incorporates variables such as gender, age, marital status, educational background, professional title, years of work experience, job level, whether the institution regularly organizes psychological counseling lectures or activities, the degree of harmony in doctor-nurse relationships within the department, and whether the individual has experienced workplace violence, among others.

2.2.2. Self-Compassion Scale

The scale was developed by international scholars Neff et al. [21], and subsequently adapted for use in Chinese populations by Chen Jian et al. [22]. It comprises six dimensions: self-kindness, common humanity, and mindfulness (scored positively), as well as self-judgment, isolation, and over-identification (scored reversely), totaling 26 items. Utilizing a five-point Likert scale, responses are scored from 1 (“very inconsistent”) to 5 (“very consistent”), yielding a total score range of 26 to 130. Higher scores indicate greater levels of self-compassion in the subjects. The Cronbach’s α coefficient for the original scale is 0.84, while the coefficient in the present study was 0.896.

2.2.3. Acceptance and Action Questionnaire Second Edition

The scale was originally developed by international researchers, including Bond [23], and was later adapted for use in China by scholars such as Cao Jing [24]. This unidimensional scale comprises 7 items in total. It utilizes a 7-point Likert scoring system, with responses ranging from “never” (1 point) to “always” (7 points), yielding a total possible score between 7 and 49. Higher scores indicate lower psychological flexibility among respondents. The scale demonstrated a Cronbach’s α coefficient of 0.88. In the present study, the Cronbach’s α coefficient was 0.95.

2.2.4. Perceived Organizational Support Scale

The perceived organizational support scale was originally developed by Chinese scholar Zhixia Chen et al. [25], and subsequently revised by Hongmei Zuo et al. [26] to accommodate the specific characteristics of the nursing profession. This scale comprises two dimensions—emotional support and instrumental support—with a total of 13 items. It employs a 5-point Likert scale, assigning scores from 1 (“strongly disagree”) to 5 (“strongly agree”), yielding a total score range of 13 to 65. Higher scores indicate a greater level of perceived organizational support among the respondents. The Cronbach’s α coefficient for this scale is 0.934, while in the present study, the Cronbach’s α coefficient is 0.977.

Following approval from the hospital nursing department, an electronic questionnaire was developed using the Wenjuanwang platform. Each item was designated as mandatory. To minimize invalid responses and enhance data quality, a lie-detection question was integrated into the survey. This involved including a positively-worded item (e.g., “I am generally kind to myself”) and later presenting its reverse-phrased counterpart (e.g., “I am often unkind to myself”) to check for response consistency, a method commonly used in psychological surveys. Restrictions included limiting each IP address to a single submission, requiring completion of all items prior to submission, and enforcing a minimum response time of 3 minutes. Investigators outlined the survey’s purpose, significance, and confidentiality protocols within a WeChat group, after which oncology nurses independently completed the questionnaire. During the review phase, two researchers examined the internal logic of responses and eliminated invalid questionnaires. Of the 267 questionnaires collected, 263 were deemed valid after removing those with inconsistent or repetitive responses, yielding an effective response rate of 98.5%.

Data analysis was conducted using SPSS version 27.0. General information and scale scores were described with frequency and mean ± standard deviation. Comparisons between groups employed the independent samples t test or one-way analysis of variance. Correlations between scales were assessed using Pearson correlation analysis. Variables with P < 0.05 in the univariate analysis were subsequently included in the multiple linear regression analysis. A result of P < 0.05 was considered statistically significant.

The total score for the self-compassion scale among oncology nurses was (85.82 ± 12.93), with a mean item score of (3.30 ± 0.50). Given that the theoretical score range of the SCS is 26 to 130, with a midpoint of 78, the observed mean score falls above the median, indicating a moderately high level of self-compassion. This score is also notably higher than those reported in studies involving psychiatric nurses [27], and emergency department nurses [20]. The psychological flexibility scale yielded a total score of (24.55 ± 9.85). The perceived organizational support scale had a total score of (46.68 ± 12.12). Details are presented in Table 1.

Table 1. Self-compassion scores of oncology nurses (n = 263).

A group comparison of self-compassion scores was conducted based on the general demographic characteristics of oncology nurses. The results revealed statistically significant differences in self-compassion scores among oncology nurses with different marital statuses and physician-nurse relationships within the department (P < 0.05), as presented in Table 2.

Table 2. Univariate analysis of self-compassion among oncology nurses (score, x ¯ ± s ).

The total self-compassion score of oncology nurses was designated as the dependent variable, while variables exhibiting statistically significant differences in univariate analysis—along with the total scores of the perceived organizational support and psychological flexibility scales—were selected as independent variables for multiple linear regression analysis. Variable assignments are presented in Table 3. The results indicate that marital status, perceived organizational support, and psychological flexibility are significant predictors of self-compassion among oncology nurses (all P < 0.05), jointly accounting for 39.5% of the total variance. More detailed results can be found in Table 4.

Table 3. Independent variable assignment table.

Table 4. Multiple linear regression analysis of factors influencing self-compassion among oncology nurses.

Note: F = 43.853,R² = 0.405; adjusted R² = 0.395; P < 0.001.

The findings of this study indicate that oncology nurses exhibited a self-compassion score of 85.82 ± 12.93, which exceeds both the scale’s median score of 78 and the self-compassion levels reported among psychiatric nurses [27], emergency department nurses [20], and midwives [18]. This outcome may be attributable to the educational background of the study participants, as 93.9% held a bachelor’s degree or higher. Within the context of higher education, mental health instruction is more prevalent, facilitating greater access to relevant psychological knowledge and resources such as counseling and self-care courses, which foster attitudes of self-kindness. Consequently, oncology nurses with higher educational attainment benefit from cognitive advantages conferred by diverse educational experiences, equipping them with robust critical thinking and problem-solving skills. In their clinical practice, these nurses are able to rationally appraise occupational stress, promptly recognize their own emotional states, actively regulate negative thoughts, and efficiently manage adverse emotions—an essential capacity in an environment characterized by frequent exposure to grief and death. Furthermore, all participants in this study were recruited from tertiary hospitals, where oncology nurses typically have access to more comprehensive training resources, well-developed mental health support systems, and broader social support networks [28]. These institutions place a greater emphasis on the psychological well-being of oncology nurses, routinely offering mental health courses or mindfulness meditation training to enhance nurses’ self-compassion, enabling them to respond to challenges and setbacks more rationally and adapt more effectively. Based on these findings, it is recommended that administrators provide oncology nurses with relevant continuing education, resources, tools, training, and support, and regularly organize in-hospital psychological group activities [5] to promote continuous learning and further enhance self-compassion levels among oncology nurses.

Further investigation revealed that the mindfulness dimension received the highest score. Mindfulness refers to being aware of the present moment with clarity and balance, neither neglecting nor becoming overly entangled in adverse aspects of oneself or life. It represents self-compassion at the attentional level [4]. Oncology nurses are routinely exposed to life-and-death scenarios and must maintain composure and focus to perform complex medical procedures. These occupational demands facilitate greater present-moment awareness and non-judgmental emotional acceptance, which in turn help to reduce emotional exhaustion and enable nurses to support and encourage themselves during periods of low mood or setbacks. These findings suggest that nursing managers in oncology specialty hospitals should regularly implement psychological support programs, with particular emphasis on emotional regulation for nurses. Strategies such as narrative reflection or peer support can provide nurses with channels for emotional expression and contribute to a psychologically safe cultural environment. Consequently, oncology nurses will be able to maintain self-kindness and more effectively cope with emotional challenges, thereby further enhancing their self-compassion.

4.2.1. Marital Status

The findings of this study indicate that unmarried oncology nurses exhibit significantly higher levels of self-compassion compared to their married counterparts. Married oncology nurses frequently assume greater family responsibilities, such as caring for children and elderly family members, which unquestionably heightens both the frequency and intensity of work-family conflicts [29]. When faced with urgent tasks or unexpected situations at work, married nurses may be compelled to concurrently manage additional family obligations, resulting in markedly increased psychological stress. In contrast, unmarried oncology nurses are more likely to closely associate their sense of self-worth with their professional roles [30]. Without the distractions arising from marital roles, they are generally able to concentrate more fully on their professional performance and are willing to invest considerable energy in their work. This heightened focus allows them to actively and accurately define their work roles, enabling more attentive and flexible responses to emergencies. Accordingly, nursing administrators should offer greater understanding, support, and assistance to married as well as divorced or widowed nurses in both their professional and personal lives, ensuring that each nurse truly feels cared for and valued, thereby enhancing their satisfaction and subjective well-being. It is essential to provide effective guidance to oncology nurses in adapting and balancing the relationship between life and work, supporting them to wholeheartedly engage in a positive and active state within oncology nursing.

4.2.2. Psychological Flexibility

The findings of this study indicate a negative correlation between psychological flexibility and self-compassion levels. As the AAQ-II is scored such that higher scores indicate lower psychological flexibility, this result means that oncology nurses with lower psychological flexibility tend to have lower self-compassion levels. Research by Sun Xiangzhi [31]-[33] and others demonstrates that individuals with greater psychological flexibility are able to adaptively shift their cognition and attitudes toward adverse events in the workplace, thereby facilitating more effective adaptation to their work environment. For oncology nurses, a high degree of psychological flexibility signifies advanced emotional regulation skills. When confronted with negative emotions, these nurses are able to employ positive coping strategies to manage emotional distress, reduce tendencies toward self-criticism and self-denial, and thereby enhance their self-compassion levels [34]. Based on these findings, nursing managers should place significant emphasis on cultivating and enhancing the psychological flexibility of oncology nurses in clinical practice. This can be accomplished by implementing a series of comprehensive and targeted psychological interventions, such as providing specialized mental health training programs that include practical skills in emotion management and stress coping; offering individual psychological counseling services to create a private and personalized support space for nurses; and organizing group psychological counseling activities to promote communication and mutual assistance, as well as to strengthen team cohesion and psychological resilience. Through these diverse intervention strategies, nurses’ psychological adaptation abilities in complex work settings can be substantially enhanced, ultimately achieving a meaningful improvement in their self-compassion levels.

4.2.3. Perceived Organizational Support

The findings of this study indicate a positive correlation between perceived organizational support and self-compassion levels; oncology nurses with higher perceived organizational support scores exhibit higher self-compassion scores. Organizational support theory emphasizes that an organization’s care and appreciation for its employees are critical factors in motivating staff retention and effort, and also serve as significant indicators for assessing the degree of organizational support provided to employees [35]. Previous research has demonstrated [14][36] that the level of perceived organizational support exerts a significant influence on nurses’ professional attitudes, emotions, and job performance. When oncology nurses genuinely experience organizational care and support, enjoy favorable remuneration and benefits, and encounter clear and smooth career development pathways, they are more likely to engage in their work with better physical and mental well-being, exhibit increased motivation, and possess greater confidence in achieving their personal goals. When confronted with problems, they are able to employ positive attribution styles, thereby effectively reducing psychological distress. Within the nursing organizational environment, factors such as organizational policies and culture are often continuously reflected through the role behaviors of nursing managers. Therefore, it is necessary for nursing managers to strengthen their understanding of organizational support theory, while also fully addressing the organizational support needs of oncology nurses. By fostering a harmonious and inclusive work environment, and providing oncology nurses with enhanced training opportunities and avenues for promotion, their sense of organizational support can be increased, thereby improving their self-compassion levels.

4.2.4. Note on the Doctor-Nurse Relationship Factor

It is noteworthy that while the univariate analysis indicated a significant association between the doctor-nurse relationship and self-compassion, this variable was not a significant predictor in the multiple linear regression model. This suggests that its observed effect in the univariate analysis may be accounted for by shared variance with other, stronger predictors in the model, such as perceived organizational support and psychological flexibility. The organizational climate, reflected in perceived support, might encompass aspects of interprofessional collaboration, thereby attenuating the unique contribution of the doctor-nurse relationship when these other factors are considered simultaneously. This is a common phenomenon in multivariate analysis where correlated independent variables are included.