Knowledge, attitude and practice toward postoperative upper limb exercise among patients undergoing axillary lymph node dissection for breast cancer Authors: Shengchao Huang, Jinxin Fu, Zhongzeng Liang, Weizhang Chen, Liyan Yu, Pu Qiu, Yuanqi Zhang, Baoying Chen Published: 2025 ABSTRACT AbstractBackgroundThis study examines the knowledge, attitude, and practice (KAP) toward postoperative upper limb exercise among patients undergoing axillary lymph node dissection for breast cancer.MethodsA cross-sectional study was carried out at Guangdong Medical University Affiliated Hospital from July 1, 2023, to May 1, 2024. Demographic information and KAP scores regarding postoperative upper limb exercises were collected using questionnaires.ResultsA total of 479 valid questionnaires were included. Among the respondents, 178 (37.2%) were aged 41–50 years, and 372 (77.7%) learned about postoperative rehabilitation from doctors and nurses. The average knowledge, attitude, and practice scores were 10 [8.00, 10.00] (range 0–10), 24 [24.00, 28.00] (range 6–30), and 36 [32.00, 39.00] (range 9–45), respectively. Multivariate logistic regression showed that having medical insurance (OR = 3.059, 95% CI: [1.063, 8.804], P = 0.038) and having hyperlipidemia (OR = 0.191, 95% CI: [0.039, 0.939], P = 0.042) were independently associated with good knowledge. Additionally, the attitude score (OR = 1.740, 95% CI: [1.275, 2.376], P < 0.001) was independently associated with proactive practice. The structural equation model showed that knowledge directly affected attitude (β = 0.279, P < 0.001), and both knowledge (β = 0.300, P < 0.001) and attitude (β = 0.494, P < 0.001) directly affected practice. Knowledge also indirectly affected practice through attitude (β = 0.138, P < 0.001).ConclusionThe study indicates that breast cancer patients who have undergone axillary lymph node dissection generally possess sufficient knowledge, maintain a positive attitude, and engage in proactive practices regarding postoperative upper limb exercises. It is recommended that healthcare providers continuously enhance patient education programs, particularly emphasizing the importance of medical insurance and managing conditions like hyperlipidemia, to further improve knowledge and practice outcomes in postoperative care.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12889-025-24410-8.Keywords: Knowledge, Attitude, Practice, Upper limb functional exercise, Breast cancer, Patient, Radical axillary lymph node dissection Background The KAP framework provides a structured approach to understanding patient behaviors in postoperative rehabilitation. This model has been widely applied in public health to assess behavioral determinants and guide intervention strategies [1]. However, its application to upper limb exercises in breast cancer patients remains limited, particularly in China, where rehabilitation awareness and adherence levels vary significantly across regions [2]. Given the growing emphasis on patient-centered care, understanding the interplay of knowledge, attitudes, and practices is crucial for designing effective rehabilitation programs tailored to Chinese patients [3]. To prevent cancer from spreading through lymph nodes, axillary lymph node dissection is often performed [4]. However, this treatment can affect the lymphatic and musculoskeletal systems of the torso and upper limbs, causing complications such as skin lymphedema, impaired arm function, and shoulder joint movement disorders [5, 6]. These conditions significantly distress patients and negatively impact their quality of life. The period immediately after breast cancer surgery is an ideal time to start upper limb exercises, which are essential for preventing upper limb dysfunction. According to current rehabilitation guidelines, patients should perform upper limb mobility and resistance exercises at least 2–3 times per day, with each session lasting 15–20 min. The correct execution involves gentle shoulder and elbow movements, progressive resistance training, and lymphatic drainage techniques to minimize lymphedema risk. Patients are encouraged to gradually increase intensity while avoiding excessive strain on the affected limb [7]. Combining aerobic and resistance exercises is particularly beneficial for managing the health of breast cancer survivors, especially considering the limited function often caused by lymphedema and upper limb impairments following surgery [8, 9]. Despite the well-documented safety and benefits of exercise for individuals living with and beyond breast cancer, research consistently shows that physical activity levels remain low among this demographic throughout the continuum of care [10, 11]. Even though there is substantial evidence supporting the integration of exercise into standard supportive cancer care, it has not yet been universally adopted, especially in breast cancer treatment where the most significant evidence exists [12, 13]. Persistent barriers hinder access to exercise resources and programs for those affected by breast cancer [14]. Recent study has identified both physical barriers, such as fatigue, and psychosocial barriers, like a lack of motivation, which significantly impede exercise participation within this population [15, 16]. Previous studies have examined factors influencing adherence to postoperative upper limb exercises among breast cancer patients. For instance, one study validated the Information–Motivation–Behavioral Skills (IMB) model and demonstrated that behavioral skills mediate the relationship between knowledge, motivation, and exercise adherence [17]. Another study found that postoperative education on upper limb use remains inconsistent, with significant variability in the extent of information patients receive regarding exercise, mobility, and lymphedema management [18]. While these studies provide valuable insights, our research uniquely applies the KAP framework to comprehensively assess the determinants of knowledge, attitudes, and practices related to postoperative upper limb exercises. Moreover, by incorporating structural equation modeling, we quantitatively elucidate the direct and indirect pathways through which knowledge and attitude influence exercise adherence, addressing gaps in previous literature and offering a targeted approach for designing patient-centered rehabilitation programs. The KAP model provides a structured approach to understanding health-related behaviors by examining how knowledge and attitudes shape practices. It is widely applied in public health research to assess behavioral determinants and inform targeted interventions [19–21]. Despite its importance, patient awareness remains insufficient. For example, upper extremity lymphedema, affecting one in five women after breast cancer treatment, significantly contributes to morbidity. Many women report a lack of information about lymphedema and other postoperative side effects [22]. Furthermore, adherence to prescribed postoperative upper limb exercises in breast cancer patients is notably low. Existing studies have primarily focused on general postoperative rehabilitation for breast cancer patients, but few have specifically examined the behavioral determinants of upper limb functional exercise in the Chinese context. This study aims to bridge this gap by applying the KAP framework and structural equation modeling (SEM) to analyze the factors influencing postoperative rehabilitation behaviors [23]. Previous studies have examined factors influencing adherence to postoperative upper limb exercises among breast cancer patients. For instance, one study validated the Information–Motivation–Behavioral Skills (IMB) model and demonstrated that behavioral skills mediate the relationship between knowledge, motivation, and exercise adherence [23]. Another study found that postoperative education on upper limb use remains inconsistent, with significant variability in the extent of information patients receive regarding exercise, mobility, and lymphedema management [18]. While these studies provide valuable insights, our research uniquely applies the KAP framework to comprehensively assess the determinants of knowledge, attitudes, and practices related to postoperative upper limb exercises. Moreover, by incorporating structural equation modeling, we quantitatively elucidate the direct and indirect pathways through which knowledge and attitude influence exercise adherence, addressing gaps in previous literature and offering a targeted approach for designing patient-centered rehabilitation programs. Complications from inadequate postoperative care can significantly affect patients’ quality of life. Therefore, understanding the KAP in this patient group is crucial for developing targeted educational and rehabilitation programs. These programs aim to enhance recovery outcomes, prevent dysfunction, and improve the overall well-being of breast cancer survivors. Identifying knowledge gaps and barriers to effective practice is essential for customizing postoperative care to address the unique needs and challenges of these patients. Currently, there is a lack of research focusing on KAP in this context. Thus, this study investigates the KAP related to postoperative upper limb exercises among breast cancer patients who have undergone radical axillary lymph node dissection. Methods Study design and subjects This cross-sectional study was conducted from July 1, 2023, to May 1, 2024, at Guangdong Medical University Affiliated Hospital, focusing on patients with breast cancer who had undergone radical axillary lymph node dissection. Approval for the study was granted by the Institutional Ethics Committee of Guangdong Medical University Affiliated Hospital (NO: KT2023-065-02), and informed consent was obtained from all participants. Eligible participants were adults definitively diagnosed with breast cancer through pathological biopsy who had received standard surgical treatment according to current guidelines. This included breast-conserving surgery, modified radical mastectomy, or reconstructive surgery with axillary lymph node dissection. Exclusion criteria were: (1) Conditions causing non-surgical limb edema (cardiac/renal/hepatic disease) to isolate ALND-specific outcomes; (2) Cognitive impairment to ensure valid responses; (3) Severe complications preventing safe exercise. This study was conducted in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines to ensure transparency and methodological rigor. Questionnaire introduction The questionnaire used in this study was based on a previous work [24] and initially tested in a small pilot with 31 participants, achieving a reliability coefficient of 0.902. The questionnaire was adapted to assess postoperative upper limb exercise behaviors based on established KAP frameworks in rehabilitation studies. To ensure consistency in assessment, the questionnaire incorporated standardized recommendations for postoperative upper limb exercises. In the pilot study, content validity was reviewed by two rehabilitation specialists and two breast cancer surgeons to ensure alignment with study objectives, and minor revisions were made to improve clarity and relevance. Content validity was assessed by four experts using four-point scales for relevance, clarity, and comprehensiveness. Items with Item Content Validity Index (I-CVI) ≥ 0.78 and Scale Content Validity Index (S-CVI) ≥ 0.90 were retained [25]. The questionnaire was adapted to assess postoperative upper limb exercise behaviors based on established KAP frameworks in rehabilitation studies. The knowledge section focused on patients’ awareness of the benefits, risks, and intensity of exercises, while the attitude section assessed perceptions of rehabilitation importance and willingness to adhere to recommended exercises. To ensure consistency in assessment, the questionnaire incorporated standardized recommendations for postoperative upper limb exercises. The questionnaire was adapted to assess postoperative upper limb exercise behaviors based on established KAP frameworks in rehabilitation studies. The knowledge section focused on patients’ awareness of the benefits, risks, and intensity of exercises, while the attitude section assessed perceptions of rehabilitation importance and willingness to adhere to recommended exercises. According to current rehabilitation guidelines, patients should perform upper limb mobility and resistance exercises at least 2–3 times per day, with each session lasting 15–20 min. The correct execution involves gentle shoulder and elbow movements, progressive resistance training, and lymphatic drainage techniques to minimize lymphedema risk [7]. Patients are encouraged to gradually increase intensity while avoiding excessive strain on the affected limb. The practice section evaluated adherence to prescribed exercises and engagement with healthcare professionals for guidance. In the pilot study, content validity was reviewed by two rehabilitation specialists and two breast cancer surgeons to ensure alignment with study objectives, and minor revisions were made to improve clarity and relevance. The final questionnaire, written in Chinese, covered four key areas (see Appendix): basic demographic data (including gender, age, education, and marital status) and three dimensions assessing knowledge (10 items), attitudes (6 items), and practices (9 items) related to postoperative care. In the knowledge section, responses were scored as correct (1 point) or incorrect (0 points), with a total possible score ranging from 0 to 10. The attitude and practice dimensions used a five-point Likert scale, ranging from very positive (5 points) to very negative (1 point), with possible scores ranging from 6 to 30 for attitudes and 9 to 45 for practices. Sufficient knowledge, positive attitude, and proactive practice was determined by scores exceeding 80% of the possible highest total score [26]. Questionnaire distribution and quality control Participants were recruited from the hospital’s breast cancer and breast surgery clinics, focusing on those who had undergone radical axillary lymph node dissection and were willing and able to participate. The questionnaire was distributed in two ways: in paper form at the clinics and electronically via email or text message. To ensure high-quality data, the research team implemented a quality control process, including setting submission deadlines, following up with patients who delayed responses, verifying completed questionnaires for accuracy, and analyzing the data to extract insights on patients’ knowledge, attitudes, and practices regarding postoperative upper limb exercises. Based on these insights, targeted recommendations for improvement were formulated. Sample size calculation Sample size was calculated using the formula for cross-sectional studies: n = Z²α/2 p(1-p)/δ², where Zα/2 = 1.96 when α = 0.05, the assumed degree of variability of p = 0.5 maximizes the required sample size, and δ is admissible error (which was 5% here). The theoretical sample size was 480 which includes an extra 20% to allow for subjects lost during the study. Statistical analysis Data analysis was performed using SPSS 22.0 (IBM, Armonk, NY, USA), R 4.3.2, and Stata 18.0 (Stata Corporation, College Station, TX, USA). Continuous variables were described using mean ± standard deviation (SD)/ Median (25%,75%). Categorical variables were presented as n (%). Normality of continuous data was assessed, with the t-test applied to normally distributed data and the Wilcoxon Mann-Whitney test to non-normally distributed data for comparisons between two groups. For analyses involving three or more groups, ANOVA was used for continuous variables with normal distribution and equal variances, and the Kruskal-Wallis test was employed for data not normally distributed. Pearson correlation analysis was used to assess the correlations between knowledge, attitude, and practice scores. Univariate and multivariate logistic regression were performed to explore the risk factors associated with knowledge, attitude, and practice. Univariate variables with P < 0.05 were included in the multivariate regression. Structural equation modeling (SEM) was used to analyze the mediating role of attitude between knowledge and practice within the KAP theoretical framework. Model fit was evaluated using root mean square error of approximation (RMSEA), incremental fit index (IFI), Tucker-Lewis index (TLI), and comparative fit index (CFI). Two-sided p < 0.05 was considered statistically significant in this study. Results Of 509 initial participants, 30 (5.9%) were excluded: 8 for edema-related comorbidities, 3 for cognitive impairment, and 19 for other reasons. This left 479 valid samples, with an effective rate of 94.11%. Among the participants, 178 (37.2%) were aged 41–50 years, 421 (87.9%) were married, 278 (58.0%) lived in urban areas in the last year, 157 (32.8%) had a college education or higher, 206 (43.0%) had a family monthly income of 2000–5000 Yuan, 240 (50.1%) had undergone radical axillary lymph node dissection 12 months or more ago, 370 (77.2%) had undergone rehabilitation training, and 372 (77.7%) learned about postoperative rehabilitation from doctors and nurses. The mean knowledge, attitude, and practice scores were 10 [8.00, 10.00] (possible range: 0–10), 24 [24.00, 28.00] (possible range: 6–30), and 36 [32.00, 39.00] (possible range: 9–45), respectively. Analysis of demographic characteristics found that participants’ knowledge, attitude, and practice scores varied significantly based on whether they performed rehabilitation training (P < 0.001, P = 0.004, P < 0.001) and their sources of information about postoperative rehabilitation (P < 0.001, P = 0.018, P = 0.004). For attitude scores, differences were observed among patients of different ages (P = 0.028), places of residence (P = 0.031), education levels (P < 0.001), family monthly incomes (P = 0.001), and types of medical insurance (P = 0.001). Differences in practice scores were found among patients from different places of residence (P = 0.037). The top 80% of scores for the knowledge, attitude, and practice dimensions were used as cut-off values to divide the groups. The number of participants above the cut-off values were 351 (73.3%) for knowledge, 209 (43.6%) for attitude, and 183 (38.2%) for practice (Table 1). Table 1Patient baseline and KAP scoresN(%)Knowledge P Attitude P Practice P Total score 479 (100.0)10 [8.00, 10.00]24 [24.00, 28.00]36 [32.00, 39.00] Age 0.2080.0280.924≤ 40 years old100 (20.9)10.00 [7.00, 10.00]26.00 [24.00, 29.00]35.00 [31.00, 41.00]41–50 years old178 (37.2)10.00 [9.00, 10.00]24.00 [24.00, 27.00]35.50 [32.00, 39.00]50–60 years old163 (34.0)10.00 [9.00, 10.00]24.00 [24.00, 28.00]36.00 [32.00, 39.00]> 60 years old38 (7.9)10.00 [7.25, 10.00]24.00 [23.25, 25.75]35.00 [32.25, 37.75] Marital status 0.1550.3870.445Unmarried14 (2.9)9.50 [6.25, 10.00]26.00 [24.25, 28.75]33.00 [26.25, 39.50]Married421 (87.9)10.00 [8.00, 10.00]24.00 [24.00, 28.00]36.00 [32.00, 39.00]Divorced/widowed44 (9.2)10.00 [7.00, 10.00]24.00 [24.00, 27.50]35.00 [31.75, 39.00] Place of residence in the past year 0.0750.0310.037Rural area170 (35.5)10.00 [8.00, 10.00]24.00 [24.00, 26.00]35.00 [31.00, 37.00]City278 (58.0)10.00 [8.00, 10.00]24.00 [24.00, 29.00]36.00 [33.00, 40.00]Suburbs31 (6.5)10.00 [9.50, 10.00]24.00 [24.00, 27.50]36.00 [30.50, 41.50] Education 0.952< 0.0010.158Elementary school and below47 (9.8)10.00 [9.00, 10.00]24.00 [24.00, 25.00]35.00 [31.50, 37.00]Junior high school177 (37.0)10.00 [8.00, 10.00]24.00 [24.00, 26.00]35.00 [32.00, 39.00]High school/technical secondary school98 (20.5)10.00 [8.25, 10.00]24.00 [24.00, 26.00]36.00 [31.25, 38.00]College degree and above157 (32.8)10.00 [8.00, 10.00]26.00 [24.00, 30.00]36.00 [32.00, 41.00]Family monthly income, yuan0.3040.0010.188<2000136 (28.4)10.00 [8.00, 10.00]24.00 [24.00, 25.25]35.50 [31.00, 37.25]2000–5000206 (43.0)10.00 [8.00, 10.00]24.00 [24.00, 27.00]35.00 [32.00, 39.00]5000–10,00098 (20.5)10.00 [9.00, 10.00]25.50 [24.00, 29.00]36.00 [33.00, 41.00]> 10,00039 (8.1)10.00 [7.50, 10.00]26.00 [24.00, 30.00]38.00 [33.50, 41.00] Medical insurance type 0.1420.0010.076No health insurance26 (5.4)9.00 [7.00, 10.00]24.00 [24.00, 25.00]35.50 [29.25, 36.75]Have medical insurance402 (83.9)10.00 [8.00, 10.00]24.00 [24.00, 28.00]36.00 [32.00, 39.00]Have medical insurance and other commercial insurance49 (10.2)10.00 [9.00, 10.00]27.00 [24.00, 30.00]37.00 [35.00, 42.00]No medical insurance, but other commercial insurance2 (0.4)6.50 [4.75, 8.25]23.50 [23.25, 23.75]33.00 [31.50, 34.50]After radical axillary lymph node dissection, month0.3550.1150.052≤ 156 (11.7)9.00 [8.00, 10.00]25.00 [24.00, 28.00]36.00 [32.00, 41.00]2–373 (15.2)10.00 [8.00, 10.00]24.00 [24.00, 29.00]37.00 [33.00, 42.00]3–12110 (23.0)10.00 [8.00, 10.00]24.00 [24.00, 26.00]35.00 [32.00, 38.75]> 12240 (50.1)10.00 [8.00, 10.00]24.00 [24.00, 28.00]35.50 [32.00, 38.00] Rehabilitation training < 0.0010.004< 0.001Yes370 (77.2)10.00 [9.00, 10.00]24.00 [24.00, 28.00]36.00 [33.00, 40.00]No109 (22.8)9.00 [5.00, 10.00]24.00 [24.00, 26.00]32.00 [28.00, 36.00] Ways to learn about postoperative rehabilitation trainning < 0.0010.0180.004Doctors and nurses372 (77.7)10.00 [9.00, 10.00]24.00 [24.00, 28.00]36.00 [32.00, 40.00]Fellow patients32 (6.7)7.00 [5.75, 9.00]24.00 [23.00, 26.25]35.00 [29.00, 37.00]Internet29 (6.1)10.00 [7.00, 10.00]24.00 [24.00, 29.00]36.00 [31.00, 40.00]Family members5 (1.0)10.00 [9.00, 10.00]24.00 [24.00, 24.00]36.00 [36.00, 37.00]Television2 (0.4)5.00 [3.50, 6.50]25.50 [24.75, 26.25]32.50 [32.25, 32.75]Other sources25 (5.2)9.00 [8.00, 10.00]24.00 [24.00, 27.00]33.00 [29.00, 36.00]Unware of postoperative rehabilitation training14 (2.9)2.50 [0.25, 3.75]24.00 [22.50, 24.00]32.00 [26.25, 35.00] Patient baseline and KAP scores In the realm of knowledge, three questions showed notably low affirmative responses: “Do you understand the sequelae of axillary lymph node dissection for breast cancer, including lymphedema in the upper limbs, local scar tissue hyperplasia, nerve damage, etc.” (K2) had a response rate of 78.7%, “Are you familiar with the intensity of postoperative rehabilitation upper limb exercises, typically requiring exercise 1–3 times daily, 15–20 minutes each session” (K9) scored 79.3%, and “Do you know how to manage postoperative pain and discomfort, where pain is considered normal and upper limb exercises are recommended” (K6) garnered 82.2% (Table 2). Table 2Responses to knowledge dimensionsKnowledgeYesNo1. Do you know the upper limb functional exercises after surgery?442 (91.5%)41 (8.5%)2. Do you know what the sequelae of axillary lymph node dissection for breast cancer include (lymphoma in the upper limbs, local scar tissue hyperplasia, nerve damage, etc.)380 (78.7%)103 (21.3%)3. Do you know that upper limb functional exercises are required after axillary lymph node dissection for breast cancer?439 (90.9%)44 (9.1%)4. Do you know why upper limb functional exercises should be performed after breast cancer surgery (functional upper limb exercises on the affected side can restore upper limb function)422 (87.4%)61 (12.6%)5. Do you know how soon after surgery you can perform functional rehabilitation exercises and the content of the exercises? (The first stage is about 1–3 days after the operation. First, you need to do finger exercises, or wrist and elbow joint exercises)424 (87.8%)59 (12.2%)6. Do you know how to deal with pain and other uncomfortable symptoms after surgery (pain is normal, upper limb functional exercise is recommended)397 (82.2%)86 (17.8%)7. Do you know the significance of upper limb functional exercise after breast cancer surgery (due to the dissection of axillary lymph glands, lymphatic drainage is blocked and the upper limbs are edema, so upper limb functional exercise after surgery is very important. Exercise should start early after surgery and wait until surgery After scar formation, exercise will be less effective)413 (85.5%)70 (14.5%)8. Do you know the method of postoperative rehabilitation function exercise (you can move your fingers and wrists within 24 h after the operation, and you can move your upper limbs 1–3 days after the operation, or you can have others assist you in elbow flexion and arm extension exercises, and gradually Transition to the shoulder joint. 4–7 days after surgery, mainly shoulder joint exercises. After 7 days, you can gradually raise the upper limbs and do wall climbing exercises)435 (90.1%)48 (9.9%)9. Do you know the intensity of postoperative rehabilitation functional exercises (generally required to exercise 1–3 times a day, 15–20 min each time)383 (79.3%)100 (20.7%)10. Do you know the precautions for postoperative functional rehabilitation exercises (the range should not be too large, the movements should not be too violent, and it should be done step by step)456 (94.4%)27 (5.6%) Responses to knowledge dimensions Attitudinal responses indicated a generally positive outlook among patients. Notably, 37.5% strongly agreed with the importance of upper limb exercises for the affected limb post-surgery (A1), while 65.4% agreed on the significance of actively engaging in the doctor’s upper limb functional exercise regimen and timely communication with healthcare professionals for limb rehabilitation (A6). Of interest, 9.7% remained neutral on whether the current limb function assessment scale adequately reflected their limb rehabilitation (A3), with 2.7% indicating otherwise (Supplementary Table 1). Regarding practices, 65.2% consistently followed medical instructions and attended regular checkups (P7), while 47.8% consistently prioritized a healthy lifestyle and protected the affected arm (P1). Conversely, 23.8% occasionally engaged in patient communication for mutual encouragement (P6), 12.6% occasionally sought guidance from healthcare professionals when unsure about upper limb exercises (P4), and 6% neglected to routinely check for skin injuries or swelling on the affected limb (Supplementary Table 2). Multivariate logistic regression showed that having medical insurance (OR = 3.059, 95% CI: [1.063, 8.804], P = 0.038) and having hyperlipidemia (OR = 0.191, 95% CI: [0.039, 0.939], P = 0.042) were independently associated with knowledge. These findings align with prior research indicating that individuals with medical insurance are more likely to access healthcare resources and receive postoperative education, leading to improved knowledge levels. Conversely, the negative association between hyperlipidemia and knowledge may reflect that patients with multiple comorbidities prioritize other health concerns over rehabilitation education, as suggested by previous studies on chronic disease management (Table 3). No factor was independently associated with attitude (Table 4). However, attitude score (OR = 1.740, 95% CI: [1.275, 2.376], P < 0.001) was independently associated with proactive practice (Table 5). Table 3Analysis of factors influencing good knowledgeUnivariate analysisMultivariate analysisOR (95%CI) P OR (95%CI) P Age ≤ 40 years oldRef.41–50 years old1.026 (0.268,3.645)0.96950–60 years old0.529 (0.126,2.058)0.364> 60 years old0.513 (0.081,3.373)0.472 Marital status UnmarriedRef.Married1.023 (0.046,11.243)0.986Divorced/widowed1.250 (0.043,23.526)0.880 Place of residence in the past year Rural areaRefCity1.738 (0.650,4.679)0.270Suburbs// Education Elementary school and belowRef.Junior high school1.500 (0.429,5.867)0.530High school/technical secondary school1.067 (0.225,4.719)0.932College degree and above1.556 (0.244,10.816)0.640Family monthly income, yuan<2000Ref.2000–50000.656 (0.192,2.153)0.4905000–10,0002.333 (0.542,12.355)0.275> 10,0000.438 (0.091,2.029)0.288 Medical insurance type No health insuranceRef.Ref.Have medical insurance2.643 (0.967,7.377)0.0593.059 (1.063,8.804) 0.038 Have medical insurance and other commercial insurance0.857 (0.032,23.263)0.9160.765 (0.043,13.655)0.855After radical axillary lymph node dissection, month≤ 1Ref.2–31.714 (0.220,16.091)0.6083–121.029 (0.169,5.622)0.974> 120.714 (0.138,3.010)0.659 Rehabilitation training Yes2.222 (0.790,6.272)0.128NoRef. Hypertension NoRef.Yes0.611 (0.124,3.320)0.541 Hyperlipidemia NoRef.Ref.Yes0.245 (0.047,1.092)0.0700.191 (0.039,0.939) 0.042 Diabetes NoRef.Yes0.960 (0.088,21.257)0.974Bold values indicate P < 0.05 Analysis of factors influencing good knowledge Bold values indicate P < 0.05 Table 4Analysis of factors influencing positive attitudeUnivariate analysisMultivariate analysisOR (95% CI) P OR (95% CI) P Knowledge 1.288 (1.042,1.594) 0.020 1.218 (0.938,1.582)0.138 Age ≤ 40 years oldRef.Ref.41–50 years old0.600 (0.173,1.943)0.4020.544 (0.110,2.689)0.45550–60 years old0.364 (0.091,1.350)0.1380.355 (0.070,1.804)0.212> 60 years old0.200 (0.023,1.226)0.0990.147 (0.015,1.412)0.097 Marital status UnmarriedRef.Married/0.990Divorced/widowed/0.991 Place of residence in the past year Rural areaRef.Ref.City4.444 (1.676,12.585) 0.004 2.850 (0.802,10.130)0.106Suburbs1.111 (0.048,13.134)0.9350.721 (0.045,11.461)0.817 Education Elementary school and belowRef.Junior high school0.300 (0.061,1.400)0.125High school/technical secondary school1.500 (0.255,9.346)0.654College degree and above2.875 (0.648,13.131)0.161Family monthly income, yuan<2000Ref.Ref.2000–50002.031 (0.644,6.706)0.2321.675 (0.388,7.221)0.4895000–10,0005.250 (1.454,21.593) 0.015 2.389 (0.391,14.585)0.345> 10,0001.875 (0.409,8.786)0.4141.322 (0.163,10.717)0.794 Medical insurance type No health insuranceRef.Ref.Have medical insurance3.253 (1.228,9.110) 0.020 1.126 (0.251,5.051)0.876Have medical insurance and other commercial insurance/0.993/0.992After radical axillary lymph node dissection, month≤ 1Ref.2–32.333 (0.384,16.285)0.3653–120.700 (0.141,3.419)0.656> 121.105 (0.268,4.557)0.888 Rehabilitation training Ref.Ref.Yes3.214 (1.161,9.631) 0.029 2.929 (0.790,10.859)0.108No Hypertension NoRef.Yes0.669 (0.124,3.247)0.615 Hyperlipidemia NoRef.Yes1.619 (0.368,8.390)0.531 Diabetes NoRef.Yes1.895 (0.174,41.813)0.608Bold values indicate P < 0.05 Analysis of factors influencing positive attitude Bold values indicate P < 0.05 Table 5Analysis of factors influencing proactive practiceUnivariate analysisMultivariate analysisOR (95%CI) P OR (95%CI) P Knowledge 1.704 (1.171,2.481) 0.005 1.341 (0.877,2.050)0.176 Attitude 1.843 (1.438,2.363) < 0.001 1.740 (1.275,2.376) < 0.001 Age ≤ 40 years oldRef.Ref.41–50 years old0.520 (0.158,1.653)0.2700.458 (0.077,2.730)0.39150–60 years old0.582 (0.154,2.122)0.4150.990 (0.145,6.775)0.991> 60 years old0.133 (0.006,1.004)0.0880.121 (0.004,4.040)0.238 Marital status UnmarriedRef.Married1.350 (0.123,29.900)0.810Divorced/widowed2.667 (0.169,75.896)0.497 Place of residence in the past year Rural areaRef.City2.126 (0.812,5.852)0.132Suburbs1.111 (0.048,13.134)0.935 Education Elementary school and belowRef.Junior high school0.700 (0.140,4.007)0.668High school/technical secondary school3.500 (0.581,25.383)0.185College degree and above2.833 (0.655,15.102)0.181Family monthly income, yuan<2000Ref.Ref.2000–50001.286 (0.385,4.414)0.6831.152 (0.175,7.584)0.8835000–10,0004.952 (1.390,19.711) 0.017 3.108 (0.303,31.830)0.339> 10,0000.980 (0.172,4.779)0.9800.715 (0.044,11.668)0.814Medical insurance type:No health insuranceRef.Ref.Have medical insurance2.827 (1.040,8.365) 0.048 0.374 (0.047,3.002)0.355Have medical insurance and other commercial insurance/0.993/0.992After radical axillary lymph node dissection, month≤ 1Ref.2–33.500 (0.581,25.383)0.1853–121.273 (0.244,7.619)0.778> 121.725 (0.412,8.909)0.474 Rehabilitation training Yes3.279 (1.120,11.112) 0.039 2.684 (0.480,15.006)0.261NoRef.Ref. Hypertension NoRef.Yes1.060 (0.197,5.160)0.942 Hyperlipidemia NoRef.Yes0.828 (0.160,3.648)0.806 Diabetes NoRef.Yes2.933 (0.269,64.799)0.388Bold values indicate P < 0.05 Analysis of factors influencing proactive practice Bold values indicate P < 0.05 The residual connection method was used to adjust the SEM model to achieve a good fit (RMSEA value: 0.062, SRMR value: 0.054, TLI value: 0.904, and CFI value: 0.914) (Table 6). Analysis of direct and indirect effects showed that knowledge directly affected attitude (β = 0.279, P < 0.001), knowledge (β = 0.300, P < 0.001) and attitude (β = 0.494, P < 0.001) directly affected practice, and knowledge indirectly affected practice through attitude (β = 0.138, P < 0.001) (Table 7 and Fig. 1). Table 6Model fit resultsIndicatorsReferenceResultsRMSEA< 0.08 Good0.062SRMR< 0.08 Good0.054TLI> 0.8 Good0.904CFI> 0.8 Good0.914 Model fit results Table 7SEM resultsModel pathsTotal effectsDirect EffectIndirect effectβ (95% CI) P β (95% CI) P β (95% CI) P AsumKsum0.279 (0.184,0.374)< 0.0010.279 (0.184,0.374)< 0.001PsumAsum0.494 (0.415,0.574)< 0.0010.494 (0.415,0.574)< 0.001Ksum0.438 (0.354,0.522)< 0.0010.300 (0.215,0.385)< 0.0010.138 (0.087,0.189)< 0.001 SEM results SEM for KAP. This SEM illustrates the direct and indirect effects of knowledge on practice through attitude. The observed variables (k1–k10 for knowledge, a1–a6 for attitude, and p1–p9 for practice) are represented as rectangles, while the latent constructs (knowledge, attitude, and practice) are shown as ellipses. Path coefficients are displayed along the connecting arrows, indicating the strength of relationships between variables. The direct effects are represented by solid arrows, with knowledge influencing both attitude (β = 0.28) and practice (β = 0.30), and attitude exerting a stronger direct effect on practice (β = 0.49). The residual errors (ε) reflect unexplained variance for each observed variable Discussion The study reveals that breast cancer patients who have undergone radical axillary lymph node dissection generally have adequate knowledge, a positive attitude, and proactive practice regarding postoperative upper limb exercises. To further enhance patient outcomes, healthcare providers should continue to emphasize education on postoperative exercises during consultations and consider integrating more personalized rehabilitation programs based on individual patient profiles, such as medical insurance status and comorbid conditions like hyperlipidemia. The results from this study on postoperative upper limb exercises among breast cancer patients post-axillary lymph node dissection show significant relationships within demographic categories, supported by findings from both significant differences and multivariate logistic regression analyses. For example, attitudes and practices varied across age groups, with younger patients (≤ 40 years) demonstrating higher attitudes towards rehabilitation. This is supported by a higher likelihood of proactive practices observed through logistic regression. This may reflect a generational difference in accessing and valuing postoperative information, as suggested by previous studies, which reported that younger patients often feel more empowered to engage in self-care due to higher health literacy [27, 28]. The role of residential setting also showed significant differences, with urban residents displaying more positive attitudes and better practices compared to rural or suburban counterparts. Logistic regression supports this, showing that urban residents have a higher likelihood of engaging in proactive rehabilitation practices. This disparity is likely due to better access to healthcare facilities and specialized providers in urban areas [29]. Similar findings have been reported in international studies. A study indicated that urban breast cancer survivors had higher participation in rehabilitation programs due to better healthcare access [30]. Another study highlighted that structured rehabilitation programs improved adherence to upper limb exercises, emphasizing the importance of standardized protocols [31]. However, in low-resource settings, limited access to physiotherapy services remains a challenge, suggesting that mobile health interventions could be a viable alternative for enhancing rehabilitation adherence [32]. These comparisons underscore the need for localized strategies that address healthcare disparities while incorporating effective global models. Education emerged as a pivotal factor, with those holding a college degree or higher demonstrating significantly more favorable attitudes. Additionally, demographic factors such as age, place of residence, and medical insurance status influenced knowledge, attitude, and practice. Younger patients exhibited more positive attitudes and greater adherence to recommended upper limb exercises, potentially due to higher health literacy and better access to digital health resources. Urban residents demonstrated more proactive rehabilitation behaviors compared to rural counterparts, likely reflecting differences in healthcare accessibility and availability of rehabilitation facilities [33, 34]. Moreover, medical insurance was associated with higher knowledge scores, suggesting that access to healthcare services and financial security play a role in patient education and adherence to postoperative care. This correlation is reinforced by regression analysis, which associates higher education with a greater likelihood of engaging in beneficial practices. Such associations align with previous findings, indicating that education enhances an individual’s capacity to understand and implement health-related advice [35, 36]. Interestingly, the study’s multivariate analysis highlighted that having medical insurance was positively associated with knowledge, emphasizing the role of socioeconomic factors in patient education and engagement. Those with insurance access may have better opportunities for receiving comprehensive postoperative care and information [37, 38]. However, no distinct factors were independently associated with attitudes, suggesting that attitudes might be influenced by a complex interplay of various non-measurable psychosocial factors. Furthermore, the SEM and logistic regression analyses together clarify the interconnectedness of knowledge, attitudes, and practices. Knowledge significantly influenced attitudes and directly, as well as indirectly, affected practices through attitudes. These findings confirm that knowledge not only helps patients understand the necessity of rehabilitation but also builds a positive attitude, which in turn fosters adherence to recommended practices [39]. Hospitals and policymakers should consider implementing standardized rehabilitation education protocols, ensuring that all patients receive comprehensive information on postoperative upper limb exercises during hospital stays and follow-up visits. Additionally, insurance providers could play a role in incentivizing rehabilitation participation by covering physiotherapy consultations or structured exercise programs, thereby improving adherence to rehabilitation recommendations. The association between knowledge, attitude, and practice in health behaviors is well supported by behavior change theories, such as the Information-Motivation-Behavioral Skills (IMB) model and the Theory of Planned Behavior (TPB) [40, 41]. These frameworks suggest that knowledge alone is insufficient to drive behavior change and that motivation (attitude) plays a crucial mediating role in influencing practice. Our findings align with these theoretical models, as attitude was found to mediate the effect of knowledge on postoperative upper limb exercise adherence. The survey results showed significant gaps in knowledge regarding the intensity of postoperative upper limb exercises and specific sequelae of axillary lymph node dissection. Specifically, three questions had notably low response rates: awareness of potential complications following axillary lymph node dissection, understanding of the recommended intensity and frequency of upper limb exercises, and knowledge of pain management strategies. These gaps indicate that while general awareness of rehabilitation exists, detailed understanding of specific aspects remains insufficient, warranting further educational emphasis. This lack of understanding may compromise the effectiveness of rehabilitation efforts. Literature indicates that comprehensive patient education can significantly improve both the initiation and adherence to prescribed upper limb exercises [42]. To address these shortcomings, targeted educational interventions could be developed. Utilizing popular Chinese social media platforms like WeChat and Douyin to disseminate easy-to-understand, engaging content about the necessity and specifics of rehabilitation could reach a wide audience. Additionally, creating visually appealing infographics and short video tutorials that detail exercise intensity and address common post-surgical complications could enhance understanding and retention of crucial information [43, 44]. While most patients recognize the importance of exercising the affected limb and seek information from various sources, a smaller subset remains neutral or unsure about the efficacy of their limb function evaluation forms and postoperative guidance. This suggests a need for more personalized and persuasive communication from healthcare providers to reinforce the benefits of active participation in rehabilitation programs. To address the observed knowledge gaps, particularly regarding exercise intensity and the management of postoperative complications, targeted educational interventions should be developed. Incorporating digital tools such as mobile apps, video tutorials, and interactive platforms can enhance patient understanding and engagement. WeChat-based rehabilitation support groups, which have been successfully used in other chronic disease management programs in China, could be leveraged to provide ongoing guidance and peer support [45, 46]. Additionally, structured workshops led by physiotherapists and oncologists should be included in routine postoperative care to reinforce key rehabilitation concepts. A systematic review highlighted the effectiveness of patient education programs in improving survivors’ outcomes [47]. Recommendations include conducting regular workshops and Q&A sessions with medical professionals, as well as leveraging patient testimonials. These testimonials can be shared through WeChat groups or live-streamed on various platforms to address concerns and motivate patients through peer and expert testimonies [48, 49]. Practices vary widely among patients, with some consistently engaging in recommended behaviors while others only occasionally follow their prescribed rehabilitation routines. Notably, a significant number of patients infrequently check for skin damage and swelling on the affected side, and a concerning percentage seldom consult medical staff when questions arise. This lack of consistent practice could be improved by establishing more robust support systems within hospital settings and online. Implementing a digital tracking and reminder system through a mobile app could help patients maintain a regular schedule of exercises and check-ups. Additionally, integrating interactive online platforms where patients can ask questions and receive timely responses from medical staff could ensure better adherence to medical advice. Collaborations with popular health influencers on platforms like WeChat to promote the benefits of consistent practice in postoperative recovery could also enhance patient engagement and compliance [50, 51]. This study has several limitations that should be considered. First, as it was conducted at a single institution, the findings may not be generalizable to broader populations with varying healthcare access and rehabilitation resources. Future studies should employ multi-center or population-based sampling to enhance external validity. Second, the reliance on self-reported data introduces potential recall and social desirability biases, which may lead to overestimations of adherence to upper limb exercises. Integrating objective measures, such as wearable activity trackers or direct clinical assessments, could improve data accuracy. Third, the cross-sectional design limits causal inferences, capturing only associations at a single time point. Longitudinal studies are needed to assess how knowledge, attitudes, and practices evolve postoperatively and to establish causal relationships. Fourth, while our questionnaire underwent content validity review by experts and showed good reliability (Cronbach’s alpha of 0.902), more comprehensive validity testing could further strengthen future research. Future studies should consider more extensive psychometric validation, including factor analysis to confirm the construct validity of the KAP domains and test-retest reliability assessment. Additionally, incorporating qualitative methods such as cognitive interviews with patients could help refine questionnaire items to better capture the nuanced experiences of breast cancer survivors engaging in postoperative upper limb exercises. Additionally, notable knowledge gaps were identified, particularly regarding the intensity and specifics of postoperative exercises, as reflected by lower response rates for certain items. This highlights the need for refining educational content and improving patient communication strategies to enhance comprehension and adherence. Lastly, digital health interventions, including mobile applications and social media-based support groups, present a promising avenue for improving patient education and engagement [52]. Prior research has demonstrated their effectiveness in promoting health behaviors, suggesting that integrating interactive content and real-time feedback into rehabilitation programs could enhance adherence. Future studies should explore the feasibility and impact of these digital strategies in optimizing postoperative rehabilitation. Conclusion In conclusion, while breast cancer patients who have undergone radical axillary lymph node dissection generally exhibit adequate knowledge and a positive attitude, their engagement in postoperative upper limb exercises varies. Structural equation modeling revealed that knowledge directly influences attitude and, along with attitude, significantly impacts practice. Multivariate analysis further identified medical insurance as a positive factor for knowledge, whereas hyperlipidemia was negatively associated. Attitude was found to be the strongest predictor of proactive practice. Given the significant role healthcare providers play in patient education, it is recommended to enhance the training and communication skills of doctors and nurses to further improve patient outcomes. Additionally, incorporating tailored educational programs on postoperative exercises into routine care could maximize the therapeutic benefits for these patients. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 2 Acknowledgements None. Abbreviation Knowledge, attitude, and practice Author contributions Shengchao Huang and Jinxin Fu composed the manuscripts; Yuanqi Zhang and Baoying Huang carried out the studies, Zhongzeng Liang and Weizhang Chen participated in collecting data, and drafted the manuscript. Liyan Yu performed the statistical analysis and participated in its design. Pu Qiu participated in acquisition, analysis, or interpretation of data and draft the manuscript. All authors read and approved the final manuscript. Funding 1. Affiliated Hospital of Guangdong Medical University “Clinical Medicine +” CnTech Co-operation Project (NO: CLP2021B002). 2. Affiliated Hospital of Guangdong Medical University Clinical Research Program (NO: LCYJ2021B003). Data availability All data generated or analysed during this study are included in this published article. Declarations Ethical approval and consent to participate All procedures were performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments. This study was approved by the Institutional Ethics Committee of Guangdong Medical University Affiliated Hospital (NO: KT2023-065-02), and informed consent was obtained from all participants. The study was carried out in accordance with the applicable guidelines and regulations. Consent for publication Not Applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Shengchao Huang and Jinxin Fu are co-first author. Contributor Information Yuanqi Zhang, Email: zhangyuanqi@gdmu.edu.cn. Baoying Chen, Email: doctor81@163.com. References Associated Data Supplementary Materials Supplementary Material 1 Supplementary Material 2 Data Availability Statement All data generated or analysed during this study are included in this published article.