This study was conducted between October 2021 and June 2022 in two breast surgery wards (A and B) of a tertiary hospital in Zhejiang Province, China. We recruited patients with breast cancer undergoing their first course of postoperative chemotherapy. Using a nonrandomized, quasi-experimental design, patients from Ward A were assigned to the control group, and those from Ward B were assigned to the intervention group after obtaining informed consent. The nonrandomized design was adopted due to practical considerations: the digital intervention was implemented as part of an online case management system. Patients within the same ward frequently interacted online, making allocation concealment and prevention of between-group contamination unfeasible [17].

Inclusion criteria were as follows: (1) age of ≥18 years, (2) pathologically confirmed diagnosis of breast cancer, (3) ability to read and write in Chinese, (4) access to a smartphone (either the patient or primary caregiver), and (5) voluntary participation with signed informed consent. Exclusion criteria were as follows: (1) patients unaware of their condition, (2) patients with physical disabilities affecting participation, or (3) patients with a history of psychological disorders or psychiatric illnesses.

The sample size was determined using repeated-measures ANOVA for quantitative data, with the following formula:

A two-sided significance level of α=.05 and a statistical power of 80% were used. The sample size calculation was based on the primary outcome, namely the CD-RISC total score. Based on published postintervention psychological resilience scores measured using the CD-RISC [18] (control group: mean 54.55, SD 8.95; intervention group: mean 60.21, SD 9.50), the calculated required sample size was approximately 34 participants per group. Considering a potential 10%‐20% dropout rate, we ultimately enrolled a total of 80 participants, with 40 participants in each group.

A researcher (QZ) blinded to group allocation performed data collection at three predetermined time points: baseline (T0, on discharge day following the first postoperative chemotherapy cycle), 3-month follow-up (T1), and 6-month follow-up (T2).

Statistical analyses were conducted using Statistical Package for the Social Sciences 26.0 (SPSS 26.0; IBM Corp). Continuous variables were presented as mean and SD or median and interquartile range according to the normality of distribution, whereas categorical variables were described as frequencies and percentages. Between-group comparisons were performed using independent-samples t tests for continuous variables and chi-square tests for categorical variables. Longitudinal changes in psychological resilience, anxiety, and depression, social support, self-efficacy, and quality of life across the three time points were analyzed using repeated-measures ANOVA. Cohen d was calculated using the pooled SD to quantify the magnitude of between-group differences at each time point.

This study was approved by the hospital’s Ethics Committee (Approval No.: IRB-2020‐408). Written informed consent was obtained from all participants before data collection. Participant data were anonymized and kept confidential, and no identifiable personal information is reported in this manuscript. Participants did not receive financial compensation for participation. The study was prospectively registered with the Chinese Clinical Trial Registry (Registration No.: ChiCTR2100047251).

Initially, 80 eligible patients with breast cancer during their postoperative chemotherapy intervals were enrolled. Ultimately, 71 participants completed the study (intervention group: n=36; control group: n=35). The attrition details were as follows. At the T1 assessment, 1 participant in the intervention group withdrew due to loss of interest, and 3 participants in the control group withdrew (1 due to nonresponse and 2 due to loss of interest). At the T2 assessment, 3 participants in the intervention group withdrew (1 due to nonresponse, 1 due to loss of interest, and 1 due to hospital transfer), while 2 participants in the control group withdrew (1 due to nonresponse and 1 due to treatment discontinuation). The details of participant attrition are presented in Figure 2.

There were no statistically significant differences between the two groups in baseline sociodemographic or clinical characteristics (P>.05; Table 1).

The total scores and subscale scores of resilience, anxiety, depression, social support, self-efficacy, and quality of life at three time points (T0, T1, and T2) for both groups followed a normal distribution but violated the sphericity assumption (P<.05). Thus, the Greenhouse-Geisser correction was applied. Cohen d values for the between-group comparisons at each time point are presented in Tables 2-6.

No statistically significant differences were observed in CD-RISC total or subscale scores between the two groups at baseline (T0). The time effect was statistically significant for the total resilience score and all subscale scores in both groups (P<.05). The between-group effect was not statistically significant for the total resilience score or any subscale scores (P>.05). A statistically significant time×group interaction effect was found for the total resilience score as well as the tenacity and optimism subscales (P<.05; Table 2). As shown in Figure 3, the total resilience score of the intervention group exhibited an upward trend over time, whereas the control group showed an initial decline followed by an increase. For the CD-RISC total score, the between-group effect size was moderate at T1 (Cohen d=0.51) and small at T2 (Cohen d=0.25).

The results showed no statistical differences in the total HADS scores or the individual anxiety and depression subscale scores at T0 between the two groups. The time effect, group effect, and group×time interaction effect for both the total HADS scores and the individual anxiety and depression subscale scores were all statistically significant (P<.05). Details are presented in Table 3. As shown in Figure 4, the intervention group exhibited a declining trend in total HADS scores over time, whereas the control group showed an initial increase followed by a subsequent decrease. For the total HADS score, the between-group effect sizes were large at both T1 (Cohen d=−2.70) and T2 (Cohen d=−1.60), indicating lower overall anxiety and depression scores in the intervention group.

The results showed no statistically significant differences in SSRS and subscale scores between the two groups at T0. The time effect was statistically significant for the total SSRS score, as well as for subjective and objective support subscale scores (P<.05). However, neither the group effect nor the group×time interaction effect reached statistical significance for the total SSRS score or any subscale scores (P>.05; Table 4). As shown in Figure 5, the intervention group exhibited a progressive increase in social support level over time, whereas the control group showed an initial rise followed by a decline. For the total SSRS score, the between-group effect sizes remained small and relatively stable at T0 (Cohen d=−0.41), T1 (Cohen d=−0.40), and T2 (Cohen d=−0.39).

As shown in Table 5, the analysis revealed no statistically significant between-group difference in BCSSS scores at baseline (T0). The time effect and group×time interaction effect for the BCSSS scores in both groups were statistically significant (P<.05). In contrast, the group effect for BCSSS scores was not statistically significant (P>.05). As shown in Figure 6, the intervention group exhibited a progressive increase in BCSSS scores over time, whereas the control group showed an initial decline followed by subsequent improvement. For the BCSSS score, the between-group effect size was 0.73 at T1 and decreased to 0.31 at T2.

The results showed that there were no statistically significant differences in FACT-B scores or dimension scores between the two groups at T0. The time effect was statistically significant for the FACT-B scores and dimension scores (except for the additional concerns dimension) in both groups (P<.05). However, there were no statistically significant differences in the group effect for the FACT-B scores and dimension scores between the two groups (P>.05). A significant group×time interaction effect was observed only in the additional concerns dimension (P<.05). Details are presented in Table 6. As shown in Figure 7, the total FACT-B scores of both groups exhibited an upward trend over time. For the total FACT-B score, the between-group effect sizes remained small at T0 (Cohen d=−0.04), T1 (Cohen d=0.18), and T2 (Cohen d=0.28).

This study evaluated a digital mindfulness intervention developed based on the RM-BC model, which was designed for home-based recovery of postoperative patients with breast cancer during chemotherapy intervals. The findings showed distinct patterns across the measured outcomes. The intervention group demonstrated a more favorable recovery trajectory in psychological resilience and self-efficacy, with significant between-group differences observed at the early follow-up but not maintained at the 6-month follow-up. Anxiety and depression showed more sustained improvement, with lower HADS scores in the intervention group at both follow-up points. In contrast, no clear statistically significant between-group differences were observed in social support or quality of life, although the direction of change was generally favorable. These findings suggest that the intervention may be particularly useful in buffering early psychological stress and supporting emotional adjustment during the vulnerable period of home recovery between chemotherapy cycles.

The repeated-measures ANOVA for CD-RISC total scores revealed a nonsignificant effect of group (F_1,69=0.751, P=.389), indicating that the overall mean level of psychological resilience across the three time points did not differ significantly between the intervention and control groups. However, a statistically significant group×time interaction effect was observed (F_1.408,97.123=7.595, P=.003), demonstrating that the trajectories of psychological resilience differed significantly between the two groups. As shown in Figure 3, the control group exhibited an initial decline in resilience at T1 followed by a rebound at T2. This pattern may reflect the natural restoration process of psychological resilience after the trauma of disease and treatment. In marked contrast, the intervention group did not exhibit such an initial decline. Instead, its resilience scores showed continuous improvement from baseline onward. Moreover, compared with the control group, the intervention group demonstrated a significantly faster rate of recovery in psychological resilience after the intervention (Figure 3). Specifically, at T1, the CD-RISC total score in the intervention group (mean 67.58, SD 11.41) was already higher than that in the control group (mean 62.09, SD 10.18); at T2, the intervention group (mean 68.75, SD 9.34) continued to show higher scores than the control group (mean 66.43, SD 9.14). These differences suggest that while the RM-BC model-based digital intervention did not yield a sustained absolute advantage in overall mean scores, it played a crucial protective role in preventing the commonly observed decline in psychological resilience during the early transitional phase of postoperative chemotherapy and in accelerating the trajectory of resilience recovery. At T1, the between-group difference in CD-RISC total scores corresponded to a moderate effect size (Cohen d=0.51). At T2, the effect size decreased to a small magnitude (Cohen d=0.25), which was consistent with the reduced between-group difference observed at the 6-month follow-up. However, we did not identify a validated minimal clinically important difference for the CD-RISC in postoperative patients with breast cancer during chemotherapy intervals. Therefore, although the short-term difference may be of potential clinical relevance, its clinical meaningfulness should be interpreted cautiously.

The mechanisms through which the digital intervention promoted recovery in psychological resilience may mainly involve two interrelated pathways. First, the intervention guided and encouraged patients to engage in regular mindfulness practice at home, including yoga, meditation, and breathing exercises. Sustained mindfulness practice may help patients internalize mindfulness as a personal coping strategy, thereby improving emotion regulation, alleviating physical discomfort, and managing intrusive thoughts [32,33]. These effects may, in turn, reduce the “risk factors” described in the RM-BC model [19]. Second, patients shared their experiences of mindfulness practice, symptom management, and rehabilitation through the interactive module of the digital platform. This process not only provided emotional validation and social support for participants who shared their stories but also allowed others to derive hope and motivation from their peers’ positive experiences. Such online interaction may have strengthened the “protective factors” in the RM-BC model, including social support, self-efficacy, and hope. The combined effect of reducing risk factors and enhancing protective factors may have helped offset the decline in psychological resilience that could otherwise be triggered by early stressors during the postoperative chemotherapy interval, while also supporting a steady improvement in resilience among patients in the intervention group.

In summary, the resilience trajectory observed in the control group appears to reflect the natural pattern of psychological recovery after trauma, whereas the sustained linear improvement observed in the intervention group highlights the potential value of the digital mindfulness intervention during the early recovery phase of the postoperative chemotherapy interval. Its benefit may not lie merely in raising the upper limit of resilience, but rather in buffering the impact of early stressors through timely intervention. In this way, the intervention may help preserve the continuity and stability of psychological recovery and protect patients from stress-related disruptions during this vulnerable period [34].

This study found that HADS scores in the intervention group were lower than those in the control group at both T1 (mean 11.78, SD 1.48 vs mean 15.86, SD 1.54) and T2 (mean 9.69, SD 1.51 vs mean 12.03, SD 1.42). This suggests that the digital intervention based on the RM-BC model may help alleviate anxiety and depression among patients with breast cancer during the postoperative chemotherapy interval. This finding is generally consistent with the results reported by Salman et al [35]. Figure 4 further shows that the longitudinal trends in HADS scores differed between the two groups. In the control group, anxiety/depression scores increased at T1 and then partially decreased at T2, although they remained higher than those in the intervention group. In contrast, HADS scores in the intervention group showed a continuous downward trend throughout the follow-up period. Notably, the between-group difference in the anxiety subscale score at T2 did not reach statistical significance. This may be related to the gradual psychological adaptation of patients in the control group over time and the partial relief of treatment-related stress, suggesting that the between-group advantage of the digital intervention in reducing anxiety symptoms may have diminished during the later follow-up period.

Overall, the decreasing trend in HADS scores was opposite to the increasing trend in CD-RISC scores, suggesting a possible association between the alleviation of anxiety and depression and the improvement of psychological resilience. In other words, the digital intervention may have promoted resilience recovery not only by enhancing patients’ positive psychological resources but also by reducing negative emotions such as anxiety and depression, thereby mitigating the interference of early stressors with the psychological recovery process. This finding further supports the interpretation of resilience recovery trajectories from the perspective of negative emotional outcomes.

Mindfulness-based interventions emphasize purposeful and nonjudgmental awareness of present-moment experiences, thereby cultivating attentional regulation and emotional adjustment. In this study, with the guidance of meditation audio materials provided through the mini-program, patients learned to anchor their attention to breathing during seated meditation, breathing exercises, and yoga practice. They were encouraged to observe changes in breathing rhythm, muscle tension, and bodily sensations, while noticing chemotherapy-related emotional fluctuations with a less reactive attitude. Meanwhile, mindfulness body scanning helped patients extend this nonjudgmental awareness to whole-body experiences, enabling them to consciously identify, accept, and adaptively respond to areas of tension and discomfort during yoga postures coordinated with breathing and movement [36]. Such mind-body integrative practice may help patients better cope with physical discomfort and psychological distress during the postoperative chemotherapy interval and approach treatment-related challenges in a calmer and more proactive manner.

In addition, regular practice of the mindfulness course may not only cultivate present-moment awareness, acceptance, and emotional regulation, but also provide a structured, low-intensity, home-based form of physical activity through mindful yoga [37]. Therefore, the intervention may have exerted its effects through dual pathways involving psychological regulation and physical activity. On the one hand, mindfulness practice may help reduce rumination, emotional reactivity, and threat perception; on the other hand, yoga and breathing exercises may promote physical relaxation, relieve muscle tension, and improve patients’ tolerance of bodily discomfort. These synergistic physiological and psychological effects may be an important reason for the sustained improvement in anxiety and depression observed in the intervention group.

The results showed that there were no statistically significant differences in SSRS scores between the two groups at any time point, suggesting that the digital intervention had a relatively limited effect on improving social support among patients with breast cancer during the postoperative chemotherapy interval. Compared with individual-level psychological variables such as anxiety, depression, and self-efficacy, the formation and enhancement of social support may depend more heavily on long-term, stable, and high-quality interpersonal networks. Although short-term digital interaction can provide some informational support, experience sharing, and emotional companionship, it may be insufficient to significantly change patients’ overall social support status within a relatively short follow-up period.

This finding may also be related to the inherent limitations of online interaction. Social support includes not only informational and experiential support but also sustained emotional responsiveness, relational trust, and real-life companionship. For patients with breast cancer during the postoperative chemotherapy interval, social support may come not only from peer communication but also from family caregiving, communication with healthcare professionals, and continuous companionship in daily life. Therefore, relying solely on an online interactive module may be insufficient to substantially change the overall level of social support reflected by SSRS scores within a short period.

Nevertheless, social support remains an important protective factor in the RM-BC model and an important foundation for improving psychological resilience, self-efficacy, and quality of life. In this study, peer experience sharing and emotional support provided through the interactive module may have helped alleviate patients’ disease-related isolation to some extent and provided them with rehabilitation references and psychological encouragement. However, such support may have been more situational and supplementary in nature, and may not have been sufficient to translate into significant improvements in social support scale scores.

Future studies could further optimize the interactive design of digital platforms by incorporating nurse-led thematic discussions, peer group communication, regular online group activities, or a combination of online and face-to-face support. These strategies may enhance the continuity, emotional depth, and intensity of interaction. By improving the quality and continuity of online interaction, digital interventions may more effectively promote patients’ social support.

Repeated-measures ANOVA of BCSSS scores showed a significant time effect (F_{1.782,122.981}=40.84, P<.001), but the main effect of group was not statistically significant (F_1,69=1.24, P=.27). This indicates that self-efficacy changed over time in both groups, but the overall mean self-efficacy scores across the three time points did not differ significantly between the intervention and control groups. However, the single-time-point between-group comparison at 3 months post intervention (T1) showed that the self-efficacy score in the intervention group was significantly higher than that in the control group (mean 42.36, SD 3.59 vs mean 39.23, SD 4.90; P=.003). This finding suggests that the effect of the digital intervention on self-efficacy may have been most evident in the early postintervention period. By the 6-month follow-up, as patients in the control group gradually adapted to the treatment and recovery process, the between-group difference may have weakened, resulting in a nonsignificant overall group effect.

In terms of longitudinal trends, Figure 6 shows that self-efficacy scores in the intervention group continued to increase after receiving the digital intervention, whereas the control group showed a pattern of initial decline followed by recovery. This trajectory was broadly consistent with the pattern of psychological resilience observed in this study, suggesting that the digital intervention may have provided support during the early postoperative chemotherapy interval and helped patients build confidence in disease management, symptom coping, and rehabilitation more rapidly. In other words, the value of the intervention may not necessarily lie in producing a long-term and stable absolute between-group advantage, but rather in buffering the early decline in self-efficacy and facilitating the earlier restoration of active coping capacity.

According to Bandura’s self-efficacy theory, self-efficacy is shaped mainly by mastery experiences, vicarious experiences, verbal persuasion, and physiological and emotional states [38]. In this study, the mindfulness course delivered through the WeChat mini-program provided patients with continuous and practical self-management exercises. Through regular breathing exercises, meditation, and yoga practice, patients may have gradually accumulated mastery experiences, such as “I can actively regulate my own state” and “I can cope with treatment-related discomfort.” These sustained practice experiences may have strengthened patients’ confidence in their self-management abilities.

At the same time, peer sharing in the interactive module may have provided patients with vicarious experiences and emotional encouragement. By reading or participating in the sharing of rehabilitation experiences from peers in similar situations, patients could observe how others persisted in practice, managed symptoms, and gradually recovered, thereby enhancing their own confidence and expectations for recovery. In addition, the check-in reminders and practice records in the platform may have reinforced patients’ sense of participation and accomplishment, allowing them to visually track their continued practice. These factors may jointly explain why self-efficacy was significantly higher in the intervention group than in the control group at T1.

This study found that FACT-B scores did not differ significantly between the two groups at T0, T1, or T2, indicating that the digital intervention did not produce a clear between-group difference in quality of life. However, the direction of the mean scores changed over time. At baseline, the control group had slightly higher FACT-B scores than the intervention group, whereas at both postintervention follow-up points, T1 and T2, the intervention group had higher mean FACT-B scores than the control group. Because these differences did not reach statistical significance, this trend should be interpreted cautiously. Nevertheless, it may suggest that the digital intervention based on the RM-BC model has a potential positive effect on quality of life among patients with breast cancer during the postoperative chemotherapy interval.

Quality of life reflects patients’ comprehensive subjective evaluation of their physical, psychological, and social functioning [39]. For patients with breast cancer receiving chemotherapy, quality of life is influenced by both treatment-related factors and psychosocial factors, including anxiety, depression, psychological resilience, and social support [40,41]. Therefore, improvement in quality of life is unlikely to be achieved rapidly through a single intervention component alone; rather, it may require sustained and multidimensional psychological, behavioral, and social support.

Considering the overall pattern of outcomes in this study, the intervention group showed relatively favorable trends in psychological resilience, anxiety and depression, and self-efficacy, which may have provided a foundation for improved quality of life. First, at the psychological level, the continuous decrease in HADS scores in the intervention group suggests that anxiety and depression were alleviated. The reduction of negative emotions may improve patients’ subjective experience of the disease and treatment process, thereby potentially enhancing emotional functioning and overall quality of life. Second, in terms of positive psychological resources, improvements in psychological resilience and self-efficacy may have enhanced patients’ ability to cope with treatment stress, symptom fluctuations, and rehabilitation challenges.

In terms of intervention content, the WeChat mini-program provided support at both psychological and physical levels. Mindful breathing, meditation, and yoga practice may help relieve physical tension, improve fatigue, and enhance patients’ tolerance of bodily discomfort. At the same time, mindfulness practice may help patients regulate emotional responses and reduce excessive attention to symptoms and disease-related uncertainty. The interactive module provided a degree of social connection and rehabilitation reference through peer experience sharing and emotional support. Although SSRS scores did not show significant between-group differences in this study, online interaction may still have served as a supplementary source of support, helping patients obtain experiential references and psychological encouragement.

It should be noted that improvement in quality of life is usually a comprehensive and gradual process, and a longer period of sustained intervention may be required before significant changes can be detected in scale scores. In this study, FACT-B scores in the intervention group were higher than those in the control group at both T1 and T2, but the differences were not statistically significant. This suggests that the digital intervention may have potential benefits for quality of life, but its effect intensity, intervention duration, or sample size may still have been insufficient to generate clear between-group differences. Future studies may consider extending the intervention period, optimizing the quality of support provided through the interactive module, and further strengthening comprehensive intervention components targeting symptom management, social support, and self-efficacy, so as to more fully evaluate the long-term effects of digital interventions on quality of life.

This study has several limitations. First, this was a single-center quasi-experimental study. Participants were allocated at the ward level rather than through individual randomization. This approach may have reduced contamination among patients within the same ward. However, it could not fully eliminate selection bias, clustering effects, or differences in ward environments and nursing practices. Therefore, the strength of causal inference remains limited. The generalizability of the findings should also be interpreted with caution. Future studies should adopt a multicenter randomized controlled design. Statistical analyses should also account for potential clustering effects.

Second, the sample size was estimated based on psychological resilience as the primary outcome. However, the study may still have had limited power to detect small between-group differences in secondary outcomes, such as social support and quality of life. Therefore, nonsignificant findings with favorable trends should be interpreted cautiously. In addition, some intervention effects appeared to diminish at the 6-month follow-up. This suggests that the digital intervention may be more effective in buffering early stress during home recovery between postoperative chemotherapy cycles. However, the durability of these effects requires further investigation. Future studies could increase the sample size and extend the follow-up period. Booster sessions or sustained support modules may also be incorporated to maintain intervention effects.

Finally, the implementation process and underlying mechanisms of the intervention were not fully evaluated. The digital platform provided mindfulness practice, interactive sharing, check-in reminders, and practice records. However, this study did not examine patients’ actual usage frequency, practice duration, or interactive engagement. It also did not explore how these engagement indicators were associated with changes in outcomes. Therefore, the relative contribution of each intervention component remains unclear. In addition, the proposed mechanism of reducing “risk factors” and enhancing “protective factors” was mainly based on the RM-BC model and observed outcome trends. It was not directly tested through mediation or path analysis. Future research should incorporate objective backend usage data from the platform. This would help examine intervention adherence, dose–response relationships, and mechanisms of action.

The findings of this study support the potential use of the RM-BC model–based digital mindfulness intervention in continuing care for patients with breast cancer. This intervention may provide psychological support during home recovery between postoperative chemotherapy cycles.

During follow-up, nurses can assess psychological resilience, anxiety and depression, and self-efficacy as part of routine care. These assessments may help identify patients who need additional psychological support. Based on patients’ emotional status, physical condition, and functional recovery, nurses can provide individualized guidance on mindfulness practice and self-management.

In clinical practice, digital platforms such as WeChat mini-programs can be used to deliver structured mindfulness courses. They can also remind patients to practice regularly and provide opportunities for peer experience sharing. Interactive modules may offer supplementary emotional support during home-based recovery.

Compared with traditional one-time health education, this approach can extend nursing support beyond the hospital. It may help patients receive continuous, accessible, and low-burden psychological support between chemotherapy cycles. Future practice could integrate digital mindfulness interventions with symptom management, rehabilitation exercise guidance, and routine follow-up care. This may help develop a more comprehensive support model for postoperative rehabilitation in patients with breast cancer.

This study evaluated a digital mindfulness intervention based on the RM-BC model, which was specially designed for the home-based recovery of postoperative patients with breast cancer during chemotherapy intervals.

The findings suggest that this intervention can help alleviate anxiety and depression. It also supported the early recovery of psychological resilience and self-efficacy. However, no clear statistically significant advantages were observed in social support or quality of life. The direction of change in these outcomes was generally favorable, but these findings should be interpreted with caution.

Overall, digital mindfulness interventions can help reduce the time, location, and physical burden associated with traditional face-to-face interventions. They may also serve as a useful supplement to psychological support and continuing care for patients with breast cancer during home-based recovery. Future studies with more rigorous designs are needed to verify the long-term effects and underlying mechanisms of this intervention and to inform the development of a comprehensive support model for postoperative rehabilitation in patients with breast cancer.