An Evidence-Informed Interpretive Structural Modeling Framework of Factors Associated with Posttraumatic Growth in Patients with Gynecologic Malignancies Authors: Meijun Guan, Shaochuan Chen, Fang Wu, Huimin Li Published: 2026 ABSTRACT ObjectiveThe aim of this study was to identify factors influencing posttraumatic growth (PTG) in patients with gynecologic malignancies, construct an interpretive structural modeling (ISM) framework, and examine the logical relationships and hierarchical structure among these factors to provide targets for clinical intervention.MethodsThe system searched eight databases, and through literature screening, data extraction, and quality assessment, relevant studies were included to extract PTG influencing factors and form an initial item pool. Fifteen experts in the field were invited to evaluate, screen, revise, and preliminarily categorize the item pool. The same group of experts was then invited again to make binary judgments on pairwise direct relationships among the factors, based on which an adjacency matrix was constructed. Using MATLAB software, Boolean operations generated a reachability matrix, and hierarchical decomposition identified the level positions and grouping categories of the factors. Finally, an ISM model was developed and verified by expert review.ResultsA total of 32 studies (8 in English and 24 in Chinese) were included, and 42 influencing factors were extracted. After expert consultation and revision, 37 factors were refined. The ISM analysis identified 3 category groups (independent group, associated group, and dependent group) and 4 levels (objective foundation level, psychological intermediary level, social support level, and coping behavior level), forming a hierarchical system of “13 core category factors - 37 operational sub-factors”. The objective foundation level (demographic factors, disease characteristics factors) was positioned as the fundamental underlying factor, belonging to the independent group, and may affect the second-level psychological intermediary level (psychological resources and traits, emotional state, cognitive processing) and the third-level social support level (family function, social support, spiritual support); the psychological intermediary level and the social support level belong to the associated group, and there is an interaction effect between them, which is interpreted as the mediating transformation link; the coping behavior level (coping methods, health behaviors) is at the top level, belonging to the dependent group, and is the most direct influencing factor of PTG.ConclusionThe influencing factors of PTG in patients with gynecological malignant tumors exhibit a four-level hierarchical structure: “objective foundation layer → psychological intermediary layer - social support layer → coping behavior layer”. This framework integrates 32 research evidences and 15 expert consensuses, clarifying the driving-dependent relationships and action paths among various factors. In clinical practice, stratified intervention can be implemented at different levels: prioritize intervention at the coping behavior layer to quickly activate PTG, continuously strengthen psychological and social collaborative intervention to consolidate the effect, and precisely identify high-risk individuals in the objective foundation layer to achieve early warning, thereby providing theoretical references for formulating personalized and precise psychological and social support plans for patients with gynecological malignant tumors. Background Gynecological malignant tumors are a major disease threatening the health of women worldwide. Among these malignancies, cervical cancer, endometrial cancer, and ovarian cancer are the three most prevalent types. Data from the International Agency for Research on Cancer (2020) indicate that the combined global incidence of these three malignancies exceeded 1.3 million new cases, with approximately 670,000 deaths.1 The incidence has continued to increase, with trends toward younger onset and greater malignant potential.2 Notably, the incidence of cervical cancer among women younger than 30 years has increased significantly,3 while the 5-year survival rate for ovarian cancer has remained below 45%.4 Patients with gynecologic malignancies frequently experience persistent psychosocial sequelae associated with altered body image, loss of fertility, and the need to reconstruct social roles.5,6 These challenges substantially increase the risk of anxiety and depression and may even adversely affect survival.5 However, along with these negative emotional experiences, some patients experience positive psychological changes.7 This phenomenon, referred to as posttraumatic growth (PTG), denotes positive psychological transformation following traumatic events such as serious illness and is regarded as an important indicator of positive psychological adaptation.8 PTG may rekindle the will to live in patients, facilitate the adoption of effective coping strategies, and promote health behaviors, thereby improving quality of life and prolonging survival. In recent years, research on PTG in the field of gynecologic oncology has progressed rapidly, encompassing incidence rates, influencing factors, dyadic interaction mechanisms with spouses, and dynamic trajectories, among other aspects.9–13 Studies examining influencing factors have been the most prevalent. The factors examined across the literature vary and primarily include individual characteristics, clinical factors, and psychosocial factors.9,13–19 Despite these advancements, there are still some limitations: Most studies focus on the association between a single factor or a few factors and PTG, lacking in-depth analysis of the directional or hierarchical relationships among the factors; the existing evidence is difficult to integrate into an overall framework that can guide clinical interventions; and it is impossible to identify the key factors driving PTG and the priority intervention targets. To address these limitations, interpretive structural modeling (ISM) was introduced in the present study. ISM is a methodology that uses matrix operations and level decomposition to elucidate the structural relationships among multiple factors within complex systems.20 Compared with methods such as DEMATEL which focus on the assessment of causal strength or AHP which focuses on pairwise importance comparisons, ISM concentrates on the determination of “existence relationships” among factors, without the need to compare factor weights, and is particularly suitable for the construction of exploratory theoretical frameworks.21,22 Its primary advantage lies in the ability to transform seemingly ambiguous and disordered relationships among factors into a clear and intuitive multi-level hierarchical model, thereby clarifying the direct and indirect relationships, driving–dependence relationships, and the overall system architecture among factors.20 Although ISM has been widely applied in engineering, management, and education, its application in healthcare remains at a preliminary stage, and studies specifically involving patients with gynecologic malignancies are particularly limited.21,22 Therefore, this study constituted the first attempt to utilize ISM in patients with gynecologic malignancies, with the aim of systematically analyzing the factors influencing PTG in this population, constructing a hierarchical framework of these factors, and clarifying the intrinsic relationships among PTG-influencing factors in patients with gynecologic malignancies, thereby providing a theoretical basis for the subsequent development of precise intervention strategies. Materials and Methods Inclusion and Exclusion Criteria Inclusion criteria were as follows: (1) study designs: cross-sectional studies, cohort studies, case–control studies, longitudinal studies, and similar designs; (2) study population: patients with gynecologic malignancies confirmed by pathological diagnosis; (3) study content: factors influencing PTG. Exclusion criteria were as follows: (1) publications in languages other than Chinese or English; (2) studies from which key data could not be extracted; (3) conference papers, abstracts, or protocols without reported results; and (4) studies for which the full text was unavailable. Search Strategy A comprehensive and systematic search was conducted across PubMed, the Cochrane Library, Embase, CINAHL, CNKI, CBM, Wanfang, and VIP databases. This search was supplemented by manual searches to identify relevant studies not indexed in these databases. The search period extended from database inception to March 2025. As the initial search yielded a limited number of eligible studies, the search strategy was subsequently expanded. The PubMed search strategy is presented as an example in Figure 1.Figure 1PubMed search strategy.A table showing search strategies using medical and psychological terms. PubMed search strategy. Literature Screening Two trained researchers (CSC and WF) independently and strictly screened the literature according to the inclusion and exclusion criteria. They then cross-checked the screening results (including the literature they excluded and the reasons for exclusion), and if there were any disagreements, the third researcher (GMJ) would make the final decision on whether to include it. A standardized data extraction form was used to extract key study characteristics and findings. Quality Assessment Based on the selected types of literature, the corresponding bias risk assessment tools are adopted. For example, the bias risk assessment tool recommended by the Agency for Healthcare Research and Quality (AHRQ) in the United States23 is used to evaluate the quality of cross-sectional studies. This tool has 11 items, and each item is evaluated using “yes (1 point)”, “no (0 points)”, or “unclear (0 points)”. The total score is the sum of the scores of each item. A score of 0 to 3 indicates low quality, 4 to 7 indicates medium quality, and 8 to 11 indicates high quality. If the study type is a cohort study or a longitudinal study, the Newcastle-Ottawa Scale (NOS)24 is used for quality evaluation. The NOS scale consists of 8 items in 3 dimensions, with a total score ranging from 0 to 9. The quality of the literature is classified based on the total score: 7 to 9 points indicate high quality, 4 to 6 points indicate medium quality, and 0 to 3 points indicate low quality. Literature with a score of 4 or above in both AHRQ and NOS is included for subsequent analysis. In case of disagreement, it is resolved through discussion. Expert Consultation Develop the Expert Inquiry Questionnaire Based on the 42 factors affecting PTG of gynecological malignant tumors extracted from the literature review and the 4 levels preliminarily divided by the research group, an expert inquiry questionnaire was compiled. The questionnaire consists of three parts: the first part is the general information form of experts (age, working years, education level, professional title, position, work unit, etc).; the second part is the expert consultation form for PTG factors, consisting of factor names, definitions, levels, and importance scores of the items, with the importance of the items scored using the Likert 5-point scale (1 = very unimportant, 2 = not important, 3 = average, 4 = important, 5 = very important), and a “modification suggestions” column is set for experts to propose revisions to the factor descriptions, classifications, and level attributions; the third part is the self-assessment form of experts’ familiarity and judgment basis. Select Inquiry Experts 15 experts with relevant knowledge, skills, and background in the research topic were invited. The selection criteria were: ① working in the field of gynecological tumors, psychology, or nursing for 10 years or more; ② having a mid-level or higher professional title; ③ having a master’s degree or above. Screening Criteria for Indicators As long as the following conditions are met, the factors will be included in the consensus:25 the average score of importance is ≥ 4 points (full score 5 points), and the coefficient of variation (CV) is ≤ 0.25. At the same time, the research group conducted group discussions based on the experts’ written opinions to decide whether to retain, merge, or delete factors that do not meet the screening criteria, and recorded the decision basis. Inquiry by Questionnaire The expert inquiry questionnaire was distributed and collected via email. The response period was set at 2 weeks, and timely reminders were made by phone or WeChat. The expert consultation was conducted in 2 rounds. After the completion of the first round of the questionnaire, the research group summarized and analyzed the expert scores and written opinions, and based on the indicator screening criteria, added, deleted, or revised the factors to form the second round of the inquiry questionnaire. The second round process was the same as the first round, and the data were statistically analyzed after the collection. After two rounds of inquiries, the expert opinions tended to be consistent, and the consultation was concluded. Statistical Analysis Statistical analysis was performed using SPSS 22.0 software. The general information of experts was described using frequency, percentage (categorical data) or median (interquartile range) (quantitative data); the degree of experts’ enthusiasm was expressed by the questionnaire response rate; the authority of experts was represented by the authority coefficient (Cr), where Cr = (Ca + Cs) / 2, and when Cr ≥ 0.70, it can be considered that the expert consultation results are reliable; the coordination degree of experts’ opinions was evaluated using the Kendall’s Concordance Coefficient (W), with a value range of 0 to 1, and the larger the value, the higher the coordination degree; the chi-square test was used to determine if the difference was statistically significant (P < 0.05). ISM Construction Establishing the Adjacency Matrix Once again, the same group of 15 experts were invited to determine the direct relationships between each pair of influencing factors. “Direct relationship” is defined as: when a factor changes, it can logically have an impact on another factor without going through any intermediate factors.20 The determination was made using the expert voting method: each expert voted “yes/no” for whether there was a direct relationship between each pair of factors. The consensus threshold was set at ≥ 80% of experts’ unanimous agreement (that is, at least 12 out of 15 experts agreed), and if this threshold was reached, it was recorded as 1 (indicating a direct relationship), otherwise it was recorded as 0 (indicating no direct relationship).20 If the expert voting results for a certain pair of factors did not reach the consensus threshold (that is, the support rate was between 20% and 80%), the research team would conduct a comprehensive assessment based on literature evidence to determine. Based on the determination results, an adjacency matrix of direct binary relationships among the PTG factors was constructed. Informed consent was obtained from all 15 experts prior to their participation in this study. Each expert was provided with detailed information regarding the study purpose, procedures, voluntary nature of participation, confidentiality protections, and the right to withdraw at any time without consequence. Experts were informed that their responses would be used for academic research purposes only and that their personal information would be kept confidential. Written informed consent was obtained from each expert before the commencement of consultation activities. All experts voluntarily agreed to participate in the two rounds of expert consultation for the interpretive structural modeling analysis. Generating the Reachability Matrix Based on the adjacency matrix and the principle of transitive closure (that is, if factor A directly affects factor B, and factor B directly affects factor C, then it can be deduced that factor A indirectly affects factor C through B. The transitivity of this indirect relationship is called “transitive closure”), Boolean operations are carried out using MATLAB software to form an accessibility matrix that reflects all direct and indirect relationships.20 Performing Level Decomposition Hierarchical identification is carried out by calculating the reachable set (the set of all factors that can be reached by a particular factor), the antecedent set (the set of all factors that can reach a particular factor), and their intersection. Taking the “age” factor in the objective base layer as an example: its reachable set includes multiple factors in the psychological intermediary layer and the social support layer, while the antecedent set only includes itself and the demographic factors in the same layer. When the intersection of the reachable set and the antecedent set is equal to the reachable set, this factor is identified as the top level. The factors that have been identified at each level are continuously extracted and the above calculation is repeated until all factors are assigned to the corresponding level. Through this iterative process, the hierarchical position of each factor is determined.20 Constructing the ISM Model The ISM model was constructed based on the results of level decomposition and the reachability relationships among the factors. The model was subsequently compared with the preliminary hierarchical structure established through expert consultation. After clinical considerations were incorporated, the model was further reviewed and refined by the relevant experts, resulting in the final ISM diagram of factors influencing PTG. To further verify the rationality of the model, a MICMAC analysis (Cross-Impact Matrix Multiplication Applied to Classification) was conducted based on the reachability matrix. The driving forces (the ability to influence other factors) and dependencies (the degree of being influenced by other factors) of each factor were calculated, and the factors were classified into independent groups, associated groups, and dependent groups. The results of the MICMAC analysis were consistent with the ISM hierarchical classification results, enhancing the robustness of the model’s classification. Results Literature Search Results A total of 566 records were retrieved through database searches and 3 through manual searching, yielding 569 records in total. Following deduplication using EndNote X9, 236 duplicate records were removed. After titles and abstracts were screened against the inclusion criteria, 284 records were excluded. The remaining 49 records underwent full-text review, from which 32 studies were ultimately included, 8 in English and 24 in Chinese. The literature screening flow diagram is presented in Figure 2, and the basic characteristics of the included studies are summarized in Table 1.7,9–15,17–19,26–46 A total of 42 influencing factors were identified and preliminarily clustered into four hierarchical levels: the objective foundational level, the psychological mediation level, the social support level, and the coping behavior level. Table 1The Characteristics of StudiesNo.StudyCountry (Region)Study DesignCancer TypeTreatment PhaseSample Size (n)Age (Years)PTG levelScaleSamplingAnalytical MethodsInfluencing Factors1Li,202415China (Liaoning Multicenter)Cross-sectional studyCervical cancerPostoperation33452.33±8.0763.94±24.91 Upper-moderatePTGI (Wang Ji’s version)/Multiple linear regression analysis,Mediation effect analysis(17)(23)(36)2Liao,202426China (Guangzhou)Cross-sectional studyGynecological cancer/215/53.22±14.76 ModeratePTGI (Wang Ji’s version)Convenience samplingMediation effect analysis(16)(31)(41)3Liu,202427China (Shandong)Cross-sectional studyCervical cancerPostoperation29048.13 ± 10.3958.24 ± 17.86 ModeratePTGI (Wang Ji’s version)Convenience samplingMediation effect analysis(29)(31)(42)4Zhou,202410China (Jiangsu)Cross-sectional studyOvarian cancerChemotherapy152/59.86±16.52 ModeratePTGI (Wang Ji’s version)Convenience samplingMediation effect analysis(26)(38)(40)5Hao,202328China (Guangxi)Cross-sectional studyCervical cancerConcurrent chemoradiotherapy9651.76±9.5266.33±17.33 ModeratePTGI (Wang Ji’s version)Convenience samplingMultiple linear regression analysis(2)(23)6Zhai,202312China (Xian Multicenter)Cross-sectional studyGynecological cancer/26347.43±5.4356.71±13.85 ModeratePTGI (Wang Ji’s version)/Actor-Partner Interdependence Model (APIM) with Mediation Effects(30)(31)7Hong,202211China (Shanghai)Cross-sectional studyEndometrial cancerPostoperation12051.62±4.85; 51.79±4.65/PTGI (Original version)/Multivariate logistic regression analysis(11)(20)(21)(32)(38)8Li,202217China (Shandong)Cross-sectional studyCervical cancerPostoperation9520-4959.70±9.86 ModeratePTGI (Geng Yaqin’s version)/Multivariate logistic regression analysis(6)(13)(37)9Wang,202229China (Shandong)Cross-sectional studyCervical cancerInitial diagnosis300/60. 57 ± 11. 92 Upper-moderatePTGI (Wang Ji’s version)/Multivariate stepwise regression analysis(6)(8)(28)10Fu,202130China (Guangxi)Cross-sectional studyEndometrial cancer/168/60.19±15.30 Upper-moderatePTGI (Yu Yongju’s version)/Multiple linear regression analysis(2)(5)(6)(7) (38)11Li,202114China (Shenyang)Cross-sectional studyCervical cancer/27552.56±10.6366.18±11.39 ModeratePTGI (Dong Lu’s version)Convenience samplingMultiple linear regression analysis(1)(11)(16) (25)12Oh,202131KoreaCross-sectional studyOvarian cancerFollow-up148 Aged ≥18 years68.09±20.17 ModeratePTGI(Song’s version)/Multiple linear regression analysis(22)(34)(35)13Song,202113ChinaLongitudinal studyGynecological cancerWithin 1 year of diagnosis40050 (45,55)61.66±12.82 Upper-moderatePTGI (Wang Ji’s version)Convenience samplingMultiple linear regression analysis, Automated Machine Learning,APIM(4)(18)(26) (27)(31)(35)14Xu,202132China (Xuzhou)Cross-sectional studyCervical cancerPostoperation65/62.22±5.12 Lower-moderatePTGI (Dong Lu’s version)/Multivariate logistic regression analysis(1)(2)(16) (28)(32)15Zhou,202118China (Anhui Multicenter)Cross-sectional studyGynecological cancerPostoperation or Chemotherapy34421-7856.5 ModeratePTGI (Wang Ji’s version)Convenience samplingBinary stepwise logistic regression(4)(6)(10) (31)(32)(35)16Cheng,20207China (Shanxi)Cross-sectional studyCervical cancerInitial diagnosis12018-7061.54±17.18 Upper-moderatePTGI (Wang Ji’s version)Convenience samplingMultivariate stepwise hierarchical regression analysis(1)(2)(11) (16)(28)17Xie,202033China (Hunan Multicenter)Cross-sectional studyOvarian cancer/201/58.45±8.38 ModeratePTGI (Ho’s version)Convenience samplingMultiple linear regression analysis(32)(35)18Pei,202034China (Gansu Multicenter)Cross-sectional studyGynecological cancer/31821-7559.25±15.28 Upper-moderatePTGI (Wang Ji’s version)Convenience samplingMultivariate stepwise regression analysis(5)(6)(26) (32)(35)19Qin,202035China (Anhui Multicenter)Cross-sectional studyOvarian cancerChemotherapy26451.42±10.4055.37±16.62 ModeratePTGI (Wang Ji’s version)/Multivariate stepwise regression analysis(2)(6)(7) (31)(32)(35)20Yue,202036China (Shanxi)Cross-sectional studyCervical cancerChemotherapy8044. 85 ± 6. 3353.68±7.36 ModeratePTGI (Unspecified version)//(1)(2)(6) (12)(14)21Qin,201937China (Anhui Multicenter)Cross-sectional studyGynecological cancer/272/56.11±12.77 ModeratePTGI (Original version)Convenience samplingMultiple linear regression analysis(3)(6)(11) (31)(32)22Tian,201938China (Hunan Multicenter)Cross-sectional studyOvarian cancer/20053.76±10.5858.56±8.33 ModeratePTGI (Ho’ version)Random samplingMultivariate stepwise regression analysis(7)(12)(32) (35)23Ma,201839China (Jiangsu)Cross-sectional studyCervical cancerChemotherapy11043.41±5.5157.87±10.27 ModeratePTGI (Wang Ji’s version)Convenience samplingMediation effect analysis(26)(28)24Peng,201840China (Xinjiang)Cross-sectional studyCervical cancerChemotherapy112/58.73±11.09 ModeratePTGI (Ho’s version)/Multivariate stepwise regression analysis(1)(2)(12) (14)(21)(22)(39)25Song,201841China (Jiangxi)Cross-sectional studyCervical cancerPostoperation99/58.36±6.81 ModeratePTGI (Wang Ji’s version)Purposive samplingMultivariate stepwise regression analysis(13)(15)(32)(37)(35)26Ye,20189China (Zhejiang)Cross-sectional studyGynecological cancer/18754.67±7.0353.31±13.75 ModeratePTGI (Wang Ji’s version)Convenience samplingMultivariate stepwise regression analysis(16)(17)(35)27Hill,201742AmericaCross-sectional studyOvarian cancer/6749.54 ±10.5555.27 ± 25.02 ModeratePTGI (Original version)/Multiple mediation analysis, Hierarchical linear regression(26)28Wang,201543China (Tianjin Multicenter)Cross-sectional studyGynecological cancer/169/52.91±16.32 ModeratePTGI (Liu Xiaohong’s version)Convenience samplingMultiple linear regression analysis(5)(19)(32)29Yang,201544China (Jiangxi Multicenter)Cross-sectional studyCervical cancerPostoperation13241.36±6.8160.54±13.68 ModeratePTGI (Ho’s version)Convenience samplingMultiple linear regression analysis(1)(9)(11) (37)30Yu,201445China (Chongqing Multicenter)Cross-sectional studyGynecological cancerPost-treatment9546.0± 3.149.54 ±21.15 Lower-moderatePTGI (Wang Ji’s version)Convenience samplingHierarchical linear regression, Mediation effect analysis(17)(24)(31)31Posluszny,201146America (Pennsylvania)Cross-sectional studyGynecological cancerPostoperation (1 year)34/Advanced-stage: 69 ± 25.3 Upper-moderate; Early-stage: 50.6±27.5 ModeratePTGI (Original version)/Multiple linear regression analysis(2)(11)(22)32Smith,200819America (New Mexico Tumor Registry)Cross-sectional studyCervical cancer/19251± 9.122.3±1.4 Lower-moderatePTGI (Original version)/Multiple linear regression analysis(11)(33)Notes: (1) Age (2) Income (3) Employment status (4) Residence (5) Healthcare payment method (6)Caregiver/spouse education level (7) Marital status (8) Fertility intention (9) Sexual satisfaction (10) Cancer type (11) Pathological stage (12) Chronic disease history (13) Treatment modality (14) Chemotherapy cycles (15) Recurrence status (16) Psychological resilience (17) Self-efficacy (18) Internal locus of control (19) Attachment style (20) SAS (Self-Rating Anxiety Scale) score (21) SDS (Self-Rating Depression Scale) score (22) PTSD (Post-Traumatic Stress Disorder) score (23) Fear of cancer progression (24) Positive emotion (25) Emotional burden (26) Rumination (deliberate/intrusive) (27) Disease subjective severity (28) Family relationship (29) Family resilience (30) Empathy (patient/caregiver) (31)Self-disclosure (patient/caregiver) (32) Social support (33) Spirituality (34) Religious affiliation (35) Coping style (confrontation/avoidance/resignation) (36)Health-related behaviors (37) Postoperative recovery time (38) PSS (Perceived Stress Scale) score (39) Psychological distress score (40) Psychological hardiness (41) Loneliness (42) Mental toughness. ① Age data format: Mean ± standard deviation or median (range), depending on the reporting method of the original literature; ② “/” indicates that this data was not reported in the original literature; ③ The PTG level was measured using the Posttraumatic Growth Inventory (PTGI), and the value is the total score of the scale. The format is mean ± standard deviation or median (interquartile range), depending on the reporting method of the original literature. Figure 2The process flowchart for document screening.PRISMA-style diagram showing how 569 articles were narrowed to 32 included studies.A PRISMA-like selection diagram with stacked boxes connected by arrows, organized by stages along the left side (identification, screening, eligibility, inclusion). It begins with records gathered from multiple bibliographic databases plus a small set found manually, then shows removal of duplicates, title/abstract screening with many exclusions and a full-text review step. A side box lists reasons for excluding some full-text articles. The final box at the bottom indicates the number of studies retained for analysis. Table 1The Characteristics of StudiesNo.StudyCountry (Region)Study DesignCancer TypeTreatment PhaseSample Size (n)Age (Years)PTG levelScaleSamplingAnalytical MethodsInfluencing Factors1Li,202415China (Liaoning Multicenter)Cross-sectional studyCervical cancerPostoperation33452.33±8.0763.94±24.91 Upper-moderatePTGI (Wang Ji’s version)/Multiple linear regression analysis,Mediation effect analysis(17)(23)(36)2Liao,202426China (Guangzhou)Cross-sectional studyGynecological cancer/215/53.22±14.76 ModeratePTGI (Wang Ji’s version)Convenience samplingMediation effect analysis(16)(31)(41)3Liu,202427China (Shandong)Cross-sectional studyCervical cancerPostoperation29048.13 ± 10.3958.24 ± 17.86 ModeratePTGI (Wang Ji’s version)Convenience samplingMediation effect analysis(29)(31)(42)4Zhou,202410China (Jiangsu)Cross-sectional studyOvarian cancerChemotherapy152/59.86±16.52 ModeratePTGI (Wang Ji’s version)Convenience samplingMediation effect analysis(26)(38)(40)5Hao,202328China (Guangxi)Cross-sectional studyCervical cancerConcurrent chemoradiotherapy9651.76±9.5266.33±17.33 ModeratePTGI (Wang Ji’s version)Convenience samplingMultiple linear regression analysis(2)(23)6Zhai,202312China (Xian Multicenter)Cross-sectional studyGynecological cancer/26347.43±5.4356.71±13.85 ModeratePTGI (Wang Ji’s version)/Actor-Partner Interdependence Model (APIM) with Mediation Effects(30)(31)7Hong,202211China (Shanghai)Cross-sectional studyEndometrial cancerPostoperation12051.62±4.85; 51.79±4.65/PTGI (Original version)/Multivariate logistic regression analysis(11)(20)(21)(32)(38)8Li,202217China (Shandong)Cross-sectional studyCervical cancerPostoperation9520-4959.70±9.86 ModeratePTGI (Geng Yaqin’s version)/Multivariate logistic regression analysis(6)(13)(37)9Wang,202229China (Shandong)Cross-sectional studyCervical cancerInitial diagnosis300/60. 57 ± 11. 92 Upper-moderatePTGI (Wang Ji’s version)/Multivariate stepwise regression analysis(6)(8)(28)10Fu,202130China (Guangxi)Cross-sectional studyEndometrial cancer/168/60.19±15.30 Upper-moderatePTGI (Yu Yongju’s version)/Multiple linear regression analysis(2)(5)(6)(7) (38)11Li,202114China (Shenyang)Cross-sectional studyCervical cancer/27552.56±10.6366.18±11.39 ModeratePTGI (Dong Lu’s version)Convenience samplingMultiple linear regression analysis(1)(11)(16) (25)12Oh,202131KoreaCross-sectional studyOvarian cancerFollow-up148 Aged ≥18 years68.09±20.17 ModeratePTGI(Song’s version)/Multiple linear regression analysis(22)(34)(35)13Song,202113ChinaLongitudinal studyGynecological cancerWithin 1 year of diagnosis40050 (45,55)61.66±12.82 Upper-moderatePTGI (Wang Ji’s version)Convenience samplingMultiple linear regression analysis, Automated Machine Learning,APIM(4)(18)(26) (27)(31)(35)14Xu,202132China (Xuzhou)Cross-sectional studyCervical cancerPostoperation65/62.22±5.12 Lower-moderatePTGI (Dong Lu’s version)/Multivariate logistic regression analysis(1)(2)(16) (28)(32)15Zhou,202118China (Anhui Multicenter)Cross-sectional studyGynecological cancerPostoperation or Chemotherapy34421-7856.5 ModeratePTGI (Wang Ji’s version)Convenience samplingBinary stepwise logistic regression(4)(6)(10) (31)(32)(35)16Cheng,20207China (Shanxi)Cross-sectional studyCervical cancerInitial diagnosis12018-7061.54±17.18 Upper-moderatePTGI (Wang Ji’s version)Convenience samplingMultivariate stepwise hierarchical regression analysis(1)(2)(11) (16)(28)17Xie,202033China (Hunan Multicenter)Cross-sectional studyOvarian cancer/201/58.45±8.38 ModeratePTGI (Ho’s version)Convenience samplingMultiple linear regression analysis(32)(35)18Pei,202034China (Gansu Multicenter)Cross-sectional studyGynecological cancer/31821-7559.25±15.28 Upper-moderatePTGI (Wang Ji’s version)Convenience samplingMultivariate stepwise regression analysis(5)(6)(26) (32)(35)19Qin,202035China (Anhui Multicenter)Cross-sectional studyOvarian cancerChemotherapy26451.42±10.4055.37±16.62 ModeratePTGI (Wang Ji’s version)/Multivariate stepwise regression analysis(2)(6)(7) (31)(32)(35)20Yue,202036China (Shanxi)Cross-sectional studyCervical cancerChemotherapy8044. 85 ± 6. 3353.68±7.36 ModeratePTGI (Unspecified version)//(1)(2)(6) (12)(14)21Qin,201937China (Anhui Multicenter)Cross-sectional studyGynecological cancer/272/56.11±12.77 ModeratePTGI (Original version)Convenience samplingMultiple linear regression analysis(3)(6)(11) (31)(32)22Tian,201938China (Hunan Multicenter)Cross-sectional studyOvarian cancer/20053.76±10.5858.56±8.33 ModeratePTGI (Ho’ version)Random samplingMultivariate stepwise regression analysis(7)(12)(32) (35)23Ma,201839China (Jiangsu)Cross-sectional studyCervical cancerChemotherapy11043.41±5.5157.87±10.27 ModeratePTGI (Wang Ji’s version)Convenience samplingMediation effect analysis(26)(28)24Peng,201840China (Xinjiang)Cross-sectional studyCervical cancerChemotherapy112/58.73±11.09 ModeratePTGI (Ho’s version)/Multivariate stepwise regression analysis(1)(2)(12) (14)(21)(22)(39)25Song,201841China (Jiangxi)Cross-sectional studyCervical cancerPostoperation99/58.36±6.81 ModeratePTGI (Wang Ji’s version)Purposive samplingMultivariate stepwise regression analysis(13)(15)(32)(37)(35)26Ye,20189China (Zhejiang)Cross-sectional studyGynecological cancer/18754.67±7.0353.31±13.75 ModeratePTGI (Wang Ji’s version)Convenience samplingMultivariate stepwise regression analysis(16)(17)(35)27Hill,201742AmericaCross-sectional studyOvarian cancer/6749.54 ±10.5555.27 ± 25.02 ModeratePTGI (Original version)/Multiple mediation analysis, Hierarchical linear regression(26)28Wang,201543China (Tianjin Multicenter)Cross-sectional studyGynecological cancer/169/52.91±16.32 ModeratePTGI (Liu Xiaohong’s version)Convenience samplingMultiple linear regression analysis(5)(19)(32)29Yang,201544China (Jiangxi Multicenter)Cross-sectional studyCervical cancerPostoperation13241.36±6.8160.54±13.68 ModeratePTGI (Ho’s version)Convenience samplingMultiple linear regression analysis(1)(9)(11) (37)30Yu,201445China (Chongqing Multicenter)Cross-sectional studyGynecological cancerPost-treatment9546.0± 3.149.54 ±21.15 Lower-moderatePTGI (Wang Ji’s version)Convenience samplingHierarchical linear regression, Mediation effect analysis(17)(24)(31)31Posluszny,201146America (Pennsylvania)Cross-sectional studyGynecological cancerPostoperation (1 year)34/Advanced-stage: 69 ± 25.3 Upper-moderate; Early-stage: 50.6±27.5 ModeratePTGI (Original version)/Multiple linear regression analysis(2)(11)(22)32Smith,200819America (New Mexico Tumor Registry)Cross-sectional studyCervical cancer/19251± 9.122.3±1.4 Lower-moderatePTGI (Original version)/Multiple linear regression analysis(11)(33)Notes: (1) Age (2) Income (3) Employment status (4) Residence (5) Healthcare payment method (6)Caregiver/spouse education level (7) Marital status (8) Fertility intention (9) Sexual satisfaction (10) Cancer type (11) Pathological stage (12) Chronic disease history (13) Treatment modality (14) Chemotherapy cycles (15) Recurrence status (16) Psychological resilience (17) Self-efficacy (18) Internal locus of control (19) Attachment style (20) SAS (Self-Rating Anxiety Scale) score (21) SDS (Self-Rating Depression Scale) score (22) PTSD (Post-Traumatic Stress Disorder) score (23) Fear of cancer progression (24) Positive emotion (25) Emotional burden (26) Rumination (deliberate/intrusive) (27) Disease subjective severity (28) Family relationship (29) Family resilience (30) Empathy (patient/caregiver) (31)Self-disclosure (patient/caregiver) (32) Social support (33) Spirituality (34) Religious affiliation (35) Coping style (confrontation/avoidance/resignation) (36)Health-related behaviors (37) Postoperative recovery time (38) PSS (Perceived Stress Scale) score (39) Psychological distress score (40) Psychological hardiness (41) Loneliness (42) Mental toughness. ① Age data format: Mean ± standard deviation or median (range), depending on the reporting method of the original literature; ② “/” indicates that this data was not reported in the original literature; ③ The PTG level was measured using the Posttraumatic Growth Inventory (PTGI), and the value is the total score of the scale. The format is mean ± standard deviation or median (interquartile range), depending on the reporting method of the original literature. The Characteristics of Studies Notes: (1) Age (2) Income (3) Employment status (4) Residence (5) Healthcare payment method (6)Caregiver/spouse education level (7) Marital status (8) Fertility intention (9) Sexual satisfaction (10) Cancer type (11) Pathological stage (12) Chronic disease history (13) Treatment modality (14) Chemotherapy cycles (15) Recurrence status (16) Psychological resilience (17) Self-efficacy (18) Internal locus of control (19) Attachment style (20) SAS (Self-Rating Anxiety Scale) score (21) SDS (Self-Rating Depression Scale) score (22) PTSD (Post-Traumatic Stress Disorder) score (23) Fear of cancer progression (24) Positive emotion (25) Emotional burden (26) Rumination (deliberate/intrusive) (27) Disease subjective severity (28) Family relationship (29) Family resilience (30) Empathy (patient/caregiver) (31)Self-disclosure (patient/caregiver) (32) Social support (33) Spirituality (34) Religious affiliation (35) Coping style (confrontation/avoidance/resignation) (36)Health-related behaviors (37) Postoperative recovery time (38) PSS (Perceived Stress Scale) score (39) Psychological distress score (40) Psychological hardiness (41) Loneliness (42) Mental toughness. ① Age data format: Mean ± standard deviation or median (range), depending on the reporting method of the original literature; ② “/” indicates that this data was not reported in the original literature; ③ The PTG level was measured using the Posttraumatic Growth Inventory (PTGI), and the value is the total score of the scale. The format is mean ± standard deviation or median (interquartile range), depending on the reporting method of the original literature. The process flowchart for document screening. Quality Assessment Results All included studies were rated as moderate or high quality and were therefore included in the analysis in their entirety. The results are presented in Table 2. Table 2Quality Assessment of StudiesNo.StudyData SourceInclusion/ Exclusion CriteriaTime FrameParticipant ContinuityIsolation of Subjective Indicators from Objective MeasuresQuality AppraisalReason for ExclusionConfounding Factor ControlMissing Data HandlingData CompletenessFollow-Up OutcomesTotal ScoreQuality Rating1Li,2024151111111101110High quality2Liao,2024261111101111110High quality3Liu,202427111111000118High quality4Zhou,202410111101101119High quality5Hao,202328111101000117Medium quality6Zhai,2023121111111011110High quality7Hong,202211111110101119High quality8Li,202217111100101017Medium quality9Wang,202229111100000015Medium quality10Fu,202130111111000118High quality11Li,2021141111111011110High quality12Oh,202131111111100119High quality13Song,202113111111100119High quality14Xu,202132111101000117Medium quality15Zhou,202118111111000017Medium quality16Cheng,20207111111000118High quality17Xie,202033111101000117Medium quality18Pei,202034111111000118High quality19Qin,202035111111100119High quality20Yue,2020361111111011110High quality21Qin,201937111111000118High quality22Tian,201938111111000118High quality23Ma,201839111111000118High quality24Peng,201840111111000118High quality25Song,201841111100001117Medium quality26Ye,20189111111000118High quality27Hill,201742110110101118High quality28Wang,201543111111100119High quality29Yang,201544111111000118High quality30Yu,201445111110001118High quality31Posluszny, 201146110110101118High quality32Smith,200819111110100118High qualityNotes: “1” indicates that the standard is met (scored 1 point), while “0” indicates that the standard is not met or is unclear (scored 0 points). The total score ranges from 0 to 11 points. Quality rating: High quality (total score ≥ 8), Medium quality (total score 4–7), Low quality (total score 0–3). Table 2Quality Assessment of StudiesNo.StudyData SourceInclusion/ Exclusion CriteriaTime FrameParticipant ContinuityIsolation of Subjective Indicators from Objective MeasuresQuality AppraisalReason for ExclusionConfounding Factor ControlMissing Data HandlingData CompletenessFollow-Up OutcomesTotal ScoreQuality Rating1Li,2024151111111101110High quality2Liao,2024261111101111110High quality3Liu,202427111111000118High quality4Zhou,202410111101101119High quality5Hao,202328111101000117Medium quality6Zhai,2023121111111011110High quality7Hong,202211111110101119High quality8Li,202217111100101017Medium quality9Wang,202229111100000015Medium quality10Fu,202130111111000118High quality11Li,2021141111111011110High quality12Oh,202131111111100119High quality13Song,202113111111100119High quality14Xu,202132111101000117Medium quality15Zhou,202118111111000017Medium quality16Cheng,20207111111000118High quality17Xie,202033111101000117Medium quality18Pei,202034111111000118High quality19Qin,202035111111100119High quality20Yue,2020361111111011110High quality21Qin,201937111111000118High quality22Tian,201938111111000118High quality23Ma,201839111111000118High quality24Peng,201840111111000118High quality25Song,201841111100001117Medium quality26Ye,20189111111000118High quality27Hill,201742110110101118High quality28Wang,201543111111100119High quality29Yang,201544111111000118High quality30Yu,201445111110001118High quality31Posluszny, 201146110110101118High quality32Smith,200819111110100118High qualityNotes: “1” indicates that the standard is met (scored 1 point), while “0” indicates that the standard is not met or is unclear (scored 0 points). The total score ranges from 0 to 11 points. Quality rating: High quality (total score ≥ 8), Medium quality (total score 4–7), Low quality (total score 0–3). Quality Assessment of Studies Notes: “1” indicates that the standard is met (scored 1 point), while “0” indicates that the standard is not met or is unclear (scored 0 points). The total score ranges from 0 to 11 points. Quality rating: High quality (total score ≥ 8), Medium quality (total score 4–7), Low quality (total score 0–3). Expert Consultation Results The demographic characteristics of the experts were as follows: age, 39 (35, 42) years; years of experience, 16 (11, 19) years; 10 experts held senior professional titles and 5 held intermediate titles; educational qualifications included 1 doctoral degree, 14 master’s degrees, and 1 bachelor’s degree. The enthusiasm of experts was measured by the questionnaire response rate. In both rounds of questionnaire distribution, 15 copies were sent out, and 15 valid responses were retrieved, with an effective response rate of 100% and an enthusiasm coefficient of 1, indicating that the experts had a high level of enthusiasm for this research. The expert authority coefficient (Cr) was calculated to be 0.95 in both rounds, indicating a high level of expert authority and the reliability of the results. After the research group analyzed and communicated with the opinions of the experts in the first round, the opinions of the experts in the second round gradually became unified. The coordination coefficients of the experts in the two rounds were 0.58 and 0.72 respectively. Following expert consultation, a total of 20 revisions were made, resulting in a refined set of 37 influencing factors. The specific revisions included the rewording of 11 items, the reclassification of 2 items, the addition of 1 category, and the deletion of 6 items. ISM Construction of PTG Influencing Factors Adjacency Matrix Refer to Table 3. Values of 0 indicate that the row factor has no direct influence on the column factor, whereas values of 1 indicate the presence of a direct influence. Table 3Adjacency Matrix of PTG Influencing FactorsInfluencing Factors12345678910111213141516171819202122232425262728293031323334353637100111001010010000000110010000000100012001110011100111000001100110000001000130101100000000000000011000100100010001401101100000000000000000000000000000005000000000000010000001101010000001000060110101110000100111000000000000010001701110001100000000000000000001000100008000000001000010000001101010010000001090000000000000000000000000000100000000100000000000000000000011001100100100000110000000000010111000000010001000000000120000000000000111000000010001000000000130000000001000111000011010101000000010140000000000000011000011010101000000011150000000000000000100011110101110000011160000000000000000000011110101110000000170000000000000000011011111111111110011180000000000000000101011111111111110011190000000000000000110111011111001110011200000000001000000111011111111111111110210000000001000000111101111111111101111220000000000000000000010001111000000011230000000000000000000011010111000000011240000000000000000000011101111000000011250000000000000000110011110111111111011260000000000000000000011110011001101011270000000000000000000011110101001100011280000000000000000000011110100000000011290000000000000000000011111111011111111300000000000000000000011111111101111111310000000000000000000011111111110111111320000000000000000000011111111111011111330000000000000000110011111111111101111340000000000000000100011111111010110111350000000000000000100011111111010111011360000000000000000000000000000000000001370000000000000000000000000000000000010Notes: Factors were numerically coded 1–37.(1) Age (2) Household per capita monthly income (3) Occupation (4) Usual place of residence (5) Healthcare payment method (6) Education level (7) Marital status (8) Fertility intention (9) Number of children (10) Sexual satisfaction (11) Type of malignancy (12) Pathological stage (13) Comorbidities (14) Treatment modality (15) Ttreatment course (16)Post-treatment recurrence (17) Psychological resilience (18)Self-efficacy (19) Internal locus of control (20) Attachment style (21) SAS (Self-Rating Anxiety Scale) score (22) SDS (Self-Rating Depression Scale) score (23) PTSD (Post-Traumatic Stress Disorder) score (24) Fear of cancer progression (25) Positive emotion (26) Emotional burden (27) Rumination (deliberate/intrusive) (28) Subjective perception of disease severity (29) Relationship with primary caregiver (30) Family resilience (31) Empathy (patient/caregiver) (32)Self-disclosure (patient/caregiver) (33) Social support (34) Spirituality (35) Religious affiliation (36) Coping style (confrontation/avoidance/resignation) (37)Health-promoting behaviors (eg, exercise, nutrition, stress management). In the adjacency matrix, “1” indicates that the row factor has a direct influence on the column factor, while “0” indicates that there is no direct influence between the row factor and the column factor. The diagonal elements (representing the influence of a factor on itself) are all set to 0. Table 3Adjacency Matrix of PTG Influencing FactorsInfluencing Factors12345678910111213141516171819202122232425262728293031323334353637100111001010010000000110010000000100012001110011100111000001100110000001000130101100000000000000011000100100010001401101100000000000000000000000000000005000000000000010000001101010000001000060110101110000100111000000000000010001701110001100000000000000000001000100008000000001000010000001101010010000001090000000000000000000000000000100000000100000000000000000000011001100100100000110000000000010111000000010001000000000120000000000000111000000010001000000000130000000001000111000011010101000000010140000000000000011000011010101000000011150000000000000000100011110101110000011160000000000000000000011110101110000000170000000000000000011011111111111110011180000000000000000101011111111111110011190000000000000000110111011111001110011200000000001000000111011111111111111110210000000001000000111101111111111101111220000000000000000000010001111000000011230000000000000000000011010111000000011240000000000000000000011101111000000011250000000000000000110011110111111111011260000000000000000000011110011001101011270000000000000000000011110101001100011280000000000000000000011110100000000011290000000000000000000011111111011111111300000000000000000000011111111101111111310000000000000000000011111111110111111320000000000000000000011111111111011111330000000000000000110011111111111101111340000000000000000100011111111010110111350000000000000000100011111111010111011360000000000000000000000000000000000001370000000000000000000000000000000000010Notes: Factors were numerically coded 1–37.(1) Age (2) Household per capita monthly income (3) Occupation (4) Usual place of residence (5) Healthcare payment method (6) Education level (7) Marital status (8) Fertility intention (9) Number of children (10) Sexual satisfaction (11) Type of malignancy (12) Pathological stage (13) Comorbidities (14) Treatment modality (15) Ttreatment course (16)Post-treatment recurrence (17) Psychological resilience (18)Self-efficacy (19) Internal locus of control (20) Attachment style (21) SAS (Self-Rating Anxiety Scale) score (22) SDS (Self-Rating Depression Scale) score (23) PTSD (Post-Traumatic Stress Disorder) score (24) Fear of cancer progression (25) Positive emotion (26) Emotional burden (27) Rumination (deliberate/intrusive) (28) Subjective perception of disease severity (29) Relationship with primary caregiver (30) Family resilience (31) Empathy (patient/caregiver) (32)Self-disclosure (patient/caregiver) (33) Social support (34) Spirituality (35) Religious affiliation (36) Coping style (confrontation/avoidance/resignation) (37)Health-promoting behaviors (eg, exercise, nutrition, stress management). In the adjacency matrix, “1” indicates that the row factor has a direct influence on the column factor, while “0” indicates that there is no direct influence between the row factor and the column factor. The diagonal elements (representing the influence of a factor on itself) are all set to 0. Adjacency Matrix of PTG Influencing Factors Notes: Factors were numerically coded 1–37.(1) Age (2) Household per capita monthly income (3) Occupation (4) Usual place of residence (5) Healthcare payment method (6) Education level (7) Marital status (8) Fertility intention (9) Number of children (10) Sexual satisfaction (11) Type of malignancy (12) Pathological stage (13) Comorbidities (14) Treatment modality (15) Ttreatment course (16)Post-treatment recurrence (17) Psychological resilience (18)Self-efficacy (19) Internal locus of control (20) Attachment style (21) SAS (Self-Rating Anxiety Scale) score (22) SDS (Self-Rating Depression Scale) score (23) PTSD (Post-Traumatic Stress Disorder) score (24) Fear of cancer progression (25) Positive emotion (26) Emotional burden (27) Rumination (deliberate/intrusive) (28) Subjective perception of disease severity (29) Relationship with primary caregiver (30) Family resilience (31) Empathy (patient/caregiver) (32)Self-disclosure (patient/caregiver) (33) Social support (34) Spirituality (35) Religious affiliation (36) Coping style (confrontation/avoidance/resignation) (37)Health-promoting behaviors (eg, exercise, nutrition, stress management). In the adjacency matrix, “1” indicates that the row factor has a direct influence on the column factor, while “0” indicates that there is no direct influence between the row factor and the column factor. The diagonal elements (representing the influence of a factor on itself) are all set to 0. Reachability Matrix Refer to Table 4. Values of 0 indicate that no connecting pathway exists between the row and column factors, whereas values of 1 indicate that a connecting pathway exists. Table 4Reachability Matrix of PTG Influencing FactorsInfluencing Factors12345678910111213141516171819202122232425262728293031323334353637111111111110111111111111111111111111112011111111101111111111111111111111111130111111111011111111111111111111111111401111111110111111111111111111111111115000010000101111111111111111111111111160111111111011111111111111111111111111701111111110111111111111111111111111118000000011101111111111111111111111111190000000011011111111111111111111111111100000000001011111111111111111111111111110000000001111111111111111111111111111120000000001011111111111111111111111111130000000001011111111111111111111111111140000000001011111111111111111111111111150000000001011111111111111111111111111160000000001011111111111111111111111111170000000001000000111111111111111111111180000000001000000111111111111111111111190000000001000000111111111111111111111200000000001000000111111111111111111111210000000001000000111111111111111111111220000000001000000111111111111111111111230000000001000000111111111111111111111240000000001000000111111111111111111111250000000001000000111111111111111111111260000000001000000111111111111111111111270000000001000000111111111111111111111280000000001000000111111111111111111111290000000001000000111111111111111111111300000000001000000111111111111111111111310000000001000000111111111111111111111320000000001000000111111111111111111111330000000001000000111111111111111111111340000000001000000111111111111111111111350000000001000000111111111111111111111360000000001000000000000000000000000011370000000001000000000000000000000000011Notes: Factors were numerically coded 1–37.(1) Age (2) Household per capita monthly income (3) Occupation (4) Usual place of residence (5) Healthcare payment method (6) Education level (7) Marital status (8) Fertility intention (9) Number of children (10) Sexual satisfaction (11) Type of malignancy (12) Pathological stage (13) Comorbidities (14) Treatment modality (15) Ttreatment course (16)Post-treatment recurrence (17) Psychological resilience (18)Self-efficacy (19) Internal locus of control (20) Attachment style (21) SAS (Self-Rating Anxiety Scale) score (22) SDS (Self-Rating Depression Scale) score (23) PTSD (Post-Traumatic Stress Disorder) score (24) Fear of cancer progression (25) Positive emotion (26) Emotional burden (27) Rumination (deliberate/intrusive) (28) Subjective perception of disease severity (29) Relationship with primary caregiver (30) Family resilience (31) Empathy (patient/caregiver) (32)Self-disclosure (patient/caregiver) (33) Social support (34) Spirituality (35) Religious affiliation (36) Coping style (confrontation/avoidance/resignation) (37)Health-promoting behaviors (eg, exercise, nutrition, stress management). In the accessibility matrix, “1” indicates that there is a direct or indirect path (accessibility) from the row factor to the column factor, while “0” indicates that there is no accessibility path between the row factor and the column factor. The diagonal elements (from a factor to itself) are set to 1. Table 4Reachability Matrix of PTG Influencing FactorsInfluencing Factors12345678910111213141516171819202122232425262728293031323334353637111111111110111111111111111111111111112011111111101111111111111111111111111130111111111011111111111111111111111111401111111110111111111111111111111111115000010000101111111111111111111111111160111111111011111111111111111111111111701111111110111111111111111111111111118000000011101111111111111111111111111190000000011011111111111111111111111111100000000001011111111111111111111111111110000000001111111111111111111111111111120000000001011111111111111111111111111130000000001011111111111111111111111111140000000001011111111111111111111111111150000000001011111111111111111111111111160000000001011111111111111111111111111170000000001000000111111111111111111111180000000001000000111111111111111111111190000000001000000111111111111111111111200000000001000000111111111111111111111210000000001000000111111111111111111111220000000001000000111111111111111111111230000000001000000111111111111111111111240000000001000000111111111111111111111250000000001000000111111111111111111111260000000001000000111111111111111111111270000000001000000111111111111111111111280000000001000000111111111111111111111290000000001000000111111111111111111111300000000001000000111111111111111111111310000000001000000111111111111111111111320000000001000000111111111111111111111330000000001000000111111111111111111111340000000001000000111111111111111111111350000000001000000111111111111111111111360000000001000000000000000000000000011370000000001000000000000000000000000011Notes: Factors were numerically coded 1–37.(1) Age (2) Household per capita monthly income (3) Occupation (4) Usual place of residence (5) Healthcare payment method (6) Education level (7) Marital status (8) Fertility intention (9) Number of children (10) Sexual satisfaction (11) Type of malignancy (12) Pathological stage (13) Comorbidities (14) Treatment modality (15) Ttreatment course (16)Post-treatment recurrence (17) Psychological resilience (18)Self-efficacy (19) Internal locus of control (20) Attachment style (21) SAS (Self-Rating Anxiety Scale) score (22) SDS (Self-Rating Depression Scale) score (23) PTSD (Post-Traumatic Stress Disorder) score (24) Fear of cancer progression (25) Positive emotion (26) Emotional burden (27) Rumination (deliberate/intrusive) (28) Subjective perception of disease severity (29) Relationship with primary caregiver (30) Family resilience (31) Empathy (patient/caregiver) (32)Self-disclosure (patient/caregiver) (33) Social support (34) Spirituality (35) Religious affiliation (36) Coping style (confrontation/avoidance/resignation) (37)Health-promoting behaviors (eg, exercise, nutrition, stress management). In the accessibility matrix, “1” indicates that there is a direct or indirect path (accessibility) from the row factor to the column factor, while “0” indicates that there is no accessibility path between the row factor and the column factor. The diagonal elements (from a factor to itself) are set to 1. Reachability Matrix of PTG Influencing Factors Notes: Factors were numerically coded 1–37.(1) Age (2) Household per capita monthly income (3) Occupation (4) Usual place of residence (5) Healthcare payment method (6) Education level (7) Marital status (8) Fertility intention (9) Number of children (10) Sexual satisfaction (11) Type of malignancy (12) Pathological stage (13) Comorbidities (14) Treatment modality (15) Ttreatment course (16)Post-treatment recurrence (17) Psychological resilience (18)Self-efficacy (19) Internal locus of control (20) Attachment style (21) SAS (Self-Rating Anxiety Scale) score (22) SDS (Self-Rating Depression Scale) score (23) PTSD (Post-Traumatic Stress Disorder) score (24) Fear of cancer progression (25) Positive emotion (26) Emotional burden (27) Rumination (deliberate/intrusive) (28) Subjective perception of disease severity (29) Relationship with primary caregiver (30) Family resilience (31) Empathy (patient/caregiver) (32)Self-disclosure (patient/caregiver) (33) Social support (34) Spirituality (35) Religious affiliation (36) Coping style (confrontation/avoidance/resignation) (37)Health-promoting behaviors (eg, exercise, nutrition, stress management). In the accessibility matrix, “1” indicates that there is a direct or indirect path (accessibility) from the row factor to the column factor, while “0” indicates that there is no accessibility path between the row factor and the column factor. The diagonal elements (from a factor to itself) are set to 1. Level Decomposition Results Three category groups were identified: independent group, linkage group, and dependent group, along with four hierarchical levels: objective foundational level, psychological mediation level, social support level, and coping behavior level. ISM Model Construction From the above results, the ISM diagram of factors influencing PTG was constructed, as presented in Figure 3. The objective foundational level constituted the bottom-level fundamental factors. Within this level, demographic factors partially influenced disease characteristics, and the individual factors within each of these two elements mutually influenced one another. The objective foundational level belonged to the independent group and drove the second-level psychological mediation level and the third-level social support level. Within the second level, factors such as psychological resources and traits, emotional state, and cognitive processing mutually influenced one another and interacted bidirectionally with the third-level social support level. Within the third level, factors such as family functioning, social support, and spiritual support mutually influenced one another. The second and third levels both belonged to the linkage group and collectively facilitated the fourth-level coping behavior level. Within the fourth level, coping style and health behaviors constituted the top-level direct factors, mutually influencing one another and belonging to the dependent group. Through these hierarchical relationships, PTG in patients with gynecologic malignancies was influenced in a bottom-up manner.Figure 3Interpretive structural modeling of factors influencing posttraumatic growth in gynecological malignancy patients.Layered flowchart linking demographics, psychology, support and coping/health actions via arrows.Black-and-white conceptual flowchart with boxed levels connected by arrows and a legend for dashed (partial), solid (full) and double-headed (two-way) influence. Bottom Objective Foundation Level contains Demographic Factors (socioeconomic factors 1 to 5; sociocultural factors 6 to 9) and Disease Characteristics (severity 11 to 13; treatment features 14 to 15; prognostic risk 16). Above it, a dashed-outline Psychological Mediation Level includes psychological resources/traits (10, 17 to 20), emotional state (21 to 26) and cognitive processing (27 to 28) with bidirectional links. Next, a dashed-outline Social Support Level shows family functioning (29 to 32), social support (33) and spiritual support (34 to 35), also linked bidirectionally. At the top, Coping Behavior Level contains coping styles (36) and health behaviors (37) connected by a two-way arrow. On the right, dotted connectors label three groupings: Independent cluster (foundation), Interrelated cluster (psychological and social support) and Dependent cluster (coping/health behaviors). Interpretive structural modeling of factors influencing posttraumatic growth in gynecological malignancy patients. Secondary Expert Review The constructed ISM model was submitted to the expert panel for re-evaluation. The experts unanimously considered that “sexual satisfaction” should be reassigned from the sociocultural factors within the objective foundational level to psychological resources and traits within the psychological mediation level, while all other stratifications remained unchanged. Discussion Necessity and Significance of Constructing a Hierarchical Framework of Factors Influencing PTG To address this gap, ISM was introduced into PTG research involving patients with gynecologic malignancies. A hierarchical system of PTG influencing factors was successfully constructed, which included 3 categories (independent group, associated group, dependent group) and 4 levels (objective foundation level, psychological intermediary level, social support level, coping behavior level), covering a total of 13 core category factors and 37 operational sub-factors. Compared with prior studies, the hierarchical model constructed in this study presents several advantages. First, comprehensive integration of influencing factors was achieved. The model incorporated objective foundational factors such as patient age and disease stage, psychological mediation factors such as anxiety, depression, and psychological resilience, social support factors such as family functioning and spiritual support, and coping behavior factors into a unified analytical framework. Second, the hierarchical attributes and transmission relationships of each factor were clearly delineated. The objective foundational level was identified as the bottom-level driving factor, and the core pathway through which it acts on the coping behavior level was clarified through chain transmission via the psychological mediation level and social support level, revealing that the formation of PTG might be the result of the combined effect of multiple factors. Third, the hierarchical stratification explicitly identified priority nodes for intervention. Fourth, the distinctive advantages of ISM in analyzing hierarchical relationships among multiple factors provide a novel methodological reference for mechanistic research on complex psychological phenomena such as PTG. It should be noted that previous studies on PTG have widely employed structural equation modeling (SEM) or path analysis to test the mediating effects among factors.10,13,15,26,47 Unlike these methods that are oriented towards statistical verification, the unique value of ISM lies mainly in hierarchical decomposition (ie, revealing which factors are deep “rooting” factors and which are surface “direct” factors) and category grouping (ie, using MICMAC analysis to divide factors into independent groups, associated groups, and dependent groups).20 These two methods complement each other in methodology - ISM provides structured hypotheses for the conceptual framework, while SEM can be used for the subsequent statistical verification of this framework. Moreover, in recent years, studies have applied advanced computational methods such as deep learning to the prediction of gynecological cancer risks,48 suggesting that future research can explore the combination of the ISM conceptual framework with computational modeling methods to achieve the entire process analysis from “factor identification” to “risk prediction”. Analysis of PTG Influencing Factors Based on ISM Hierarchy Objective Foundational Level The objective foundational level, representing the bottom-level driving factor within the ISM hierarchical framework of this study, belongs to the independent group and encompasses two major dimensions: demographic factors and disease characteristics. Within this level, demographic factors are further subdivided into socioeconomic factors (age, monthly household income per capita, occupation, place of permanent residence, and mode of medical expense payment) and sociocultural factors (educational level, marital status, reproductive intention, and number of children). Disease characteristics include severity (malignancy type, pathological staging, and comorbidities), treatment characteristics (treatment modality and number of treatment cycles), and prognostic risk (post-treatment recurrence status). Within the demographic dimension, some studies have indicated that younger patients with cervical cancer, owing to lower psychological resilience and greater family responsibility pressures, are more likely to experience disease-related fear, which significantly suppresses PTG levels.49 However, a cross-sectional study involving 771 patients with gynecologic cancers reported that younger age was associated with higher PTG levels.6 Although these findings may appear contradictory, they are consistent with the results of the present study, as the cited studies examined only the independent effect of age. However, the ISM model of this study suggests that age may exert an indirect effect on the psychological intermediary layer through the interaction with disease characteristics such as pathological stage and treatment method. Therefore, the relationship between age and PTG may not be a simple linear one. Within the disease characteristics dimension, existing evidence indicates that patients with advanced disease, owing to higher recurrence risk and prolonged treatment duration, are more likely to experience uncertainty about the future, thereby necessitating targeted psychological interventions to enhance PTG.11 This study further revealed through the ISM model that disease characteristics such as pathological stage and recurrence risk may influence psychological mediating variables such as anxiety and psychological resilience through direct or indirect pathways. This provides potential target references for clinical intervention. The ISM model demonstrates bidirectional influences among the sub-elements within both demographic factors and disease characteristics. Empirical evidence has indicated that among patients with gynecologic malignancies, lower-income patients experience reduced treatment adherence mediated by financial toxicity, which ultimately increases the risk of recurrence.50 Intervention programs should therefore address influencing factors through coordinated management rather than targeting any single factor in isolation. Many factors at this level are inherently immutable or difficult to modify in the short term (eg, age, income); however, through the hierarchical framework constructed in this study, it can assist in identifying and screening out high-risk groups, enabling focused attention and early intervention, thereby potentially alleviating the inhibitory factors of PTG at the root level. Psychological Mediation Level and Social Support Level The psychological mediation level and the social support level, as key components in the formation of PTG in patients with gynecologic malignancies, belong to the linkage group. Both levels are jointly driven by the objective foundational level, exhibit significant bidirectional interactions with each other, and ultimately act on the coping behavior level. The psychological mediation level focuses on internal psychological processes in patients and comprises three elements: psychological resources and traits, emotional state, and cognitive processing among which significant interactive relationships exist. Psychological resources and traits constitute the core foundation of the psychological mediation level and encompass sexual satisfaction, psychological resilience, self-efficacy, internal health locus of control, and attachment style, among others. These traits directly determine the psychological reserve capacity of patients for coping with disease-related trauma. Evidence indicates that patients with high psychological resilience demonstrate higher PTG levels through active emotional regulation and the mobilization of social resources when coping with adversity.26 Emotional state encompasses anxiety, depression, post-traumatic stress disorder, fear of disease progression, positive affect, and emotional burden. These emotional states are regulated by psychological resources and traits while simultaneously exerting a direct influence on the direction of cognitive processing. The ISM model in this study demonstrates that negative emotional states such as anxiety, depression, and fear of disease progression directly impede PTG formation, whereas positive affect exerts a facilitating effect; these findings are consistent with those reported by Gao et al51 One notable finding suggested by this model is that the inhibitory effect of negative emotions on PTG may have a hierarchical dependence - when the negative factors at the objective foundation level (such as recurrence, low income) are relatively significant, their inhibitory effect may be amplified; while when the psychological resource traits (such as high psychological resilience) are abundant, it may weaken the transmission intensity of negative emotions. This hypothesis needs to be further verified through empirical research. Cognitive processing encompasses ruminative thinking and subjective perception of disease severity and serves as the final pathway through which the psychological mediation level acts on the coping behavior level. The regulatory direction of ruminative thinking and subjective perception of disease severity directly determines the developmental trajectory of PTG: intrusive rumination reinforces the transmission of negative emotions and suppresses PTG, whereas deliberate rumination facilitates positive cognitive restructuring and promotes PTG development.52 Findings from a longitudinal study further indicate that illness uncertainty influences cognitive processing through psychological resilience.47 The social support level, as an important external component in PTG formation, together with the psychological mediation level constitutes a transmission bridge between the objective foundational level and the coping behavior level. From the ISM level decomposition results, this level can be subdivided into three dimensions: family functioning, social support, and spiritual support, each of which functions independently while contributing synergistically to a supportive network. The primary mechanism through which this level exerts its effects involves dynamic interaction with the psychological mediation level, thereby enabling more precise regulation of PTG, a finding consistent with the perspective reported by Zhang.53 Family functioning constitutes the core of the social support level and encompasses elements such as the relationship with the primary caregiver, family resilience, empathy, and self-disclosure. Evidence indicates that patients with high family cohesion are more likely to exhibit sustained high-growth trajectories.52 Multiple studies have reported that self-disclosure can significantly enhance PTG levels by alleviating emotional distress, strengthening intimate relationships, and modifying the cognitive appraisal of stressful events.54,55 These findings are consistent with the transmission pathways identified in the ISM model of this study and provide more precise mechanistic support for family-based intervention approaches in clinical practice. Within the social support dimension, assistance from fellow patients as well as close relatives and friends can reinforce the function of the psychological mediation level by facilitating self-disclosure. High levels of interpersonal interaction support can attenuate the inhibitory effect of fertility concerns on PTG among patients with cancer of reproductive age.56 The value of medical support within social support lies in its capacity to enhance patients’ understanding of the disease, reduce illness uncertainty, and alleviate negative emotions. It also facilitates cognitive restructuring to promote PTG through health education and promotion, diagnostic and therapeutic communication, and related approaches.47 The ISM model in this study establishes that the psychological mediation level and the social support level can not only act directly on the coping behavior level but can also indirectly guide positive behaviors through their mutual interaction, thereby influencing PTG. With respect to the regulatory effect of the social support level on the psychological mediation level, the core function of social support is to activate the positive functioning of the psychological mediation level, with different support dimensions exerting differential effects on this level. Family support tends to regulate emotional state and psychological resources and traits; interpersonal support primarily facilitates self-disclosure; and medical support focuses on cognitive processing.47,52,57 This characteristic of precise targeted regulation indicates that clinical interventions should align appropriate support resources with the specific vulnerabilities within the psychological mediation level in patients. In the reciprocal interaction between these two levels, psychological state and the level of psychological resources in patients directly determine the extent to which social support is absorbed. Patients with high psychological resilience are more likely to perceive and use social support resources; self-efficacy levels influence receptiveness to support-related information; and cognitive processing style affects the efficiency of social support utilization.58–60 It is evident that these two levels jointly constitute the core mediating components of the model, serving a bridging role and representing critical points for intervention. This finding has important implications for clinical intervention. Reinforcement of social support alone, without activation of the psychological mediation level, may result in inefficient utilization of support resources. Conversely, exclusive emphasis on psychological regulation without the establishment of a support network may hinder the long-term stability of PTG due to insufficient external support. Based on the structural relationship of this model, psychological and social collaborative intervention may be a direction worth exploring. Future research can investigate the effectiveness of the psychological and social collaborative intervention model, such as using specific methods like mindfulness therapy, acceptance and commitment therapy (ACT), narrative care, couple communication intervention, and peer support, to test its effectiveness and applicability in improving PTG in patients with gynecological malignancies. It should be noted that during the ISM construction process, the expert panel unanimously considered that sexual satisfaction should be reclassified from the objective foundational level to the domain of psychological resources and traits within the psychological mediation level. This adjustment may be based on the following clinical logic: treatment of gynecologic malignancies is commonly accompanied by damage to reproductive organs and declining sex hormone levels, which directly impair female sexual function. The maintenance of sexual satisfaction can, on one hand, enhance self-identity and, on the other hand, function as a key psychological factor influencing the quality of partner support, enhancing the psychological resilience of the patients, reinforcing intimate relationship bonds, and promoting greater willingness to accept partner support and to seek assistance such as psychosexual counseling.61,62 This reclassification is therefore clinically justified. However, it is also necessary to note that this reclassification decision itself reflects the subjectivity in expert judgment - different experts may have different perceptions of the hierarchical level to which a particular factor belongs, which suggests that future research can verify the rationality of this classification through empirical data at the patient level. Coping Behavior Level The coping behavior level occupies the top tier of the ISM model, belongs to the dependent group, and represents the ultimate behavioral manifestation of PTG in patients with gynecologic malignancies. It encompasses coping style (coping strategies) and health behaviors (health-promoting lifestyle) and represents the most direct level for intervention, and the behavioral choices at this level directly determine the occurrence and developmental trajectory of PTG. Based on the structural relationship of this model, the behavioral layer should be identified as the most direct influencing factor of PTG (that is, there are no other intermediary levels between this layer of factors and PTG in the model). This structural relationship is consistent with the existing empirical research evidence: Zhao reported that improvement in PTG trajectories among patients with cervical cancer undergoing postoperative chemotherapy was accompanied by a shift in coping style from avoidance to confrontation.52 Korean scholars have proposed that coping style is a key factor influencing PTG in patients with ovarian cancer, with positive reappraisal representing the predominant approach.31 Another large-scale study involving 1221 cancer survivors further confirmed that active coping was significantly positively correlated with the PTG level.63 Health-promoting behaviors serve as an important safeguard for maintaining PTG. Detail-oriented behavioral nursing care based on the health belief model can significantly enhance PTG levels by reinforcing health-promoting behaviors in patients, and patients with cervical cancer demonstrate better adaptive capacity through lifestyle modification.15,64 Furthermore, this model indicates that the coping behavior layer is not independent but is indirectly influenced by the psychological intermediary layer and the social support layer. Therefore, simple behavioral persuasion may have limited effectiveness. It is necessary to comprehensively promote effective coping and healthy behaviors by considering the patient’s internal psychological state and the external support environment. Future research can be based on the hierarchical structure of this model and design intervention plans for different levels (such as cognitive-behavioral training for the coping behavior layer, mindfulness intervention for the psychological intermediary layer, and couple-based intervention for the social support layer, etc). The mutual influence between coping style and health behavior factors within the coping behavior level provides clear pathways for precision intervention in clinical practice. Patients should be guided by nursing staff to reframe their circumstances by drawing on positive psychological and cognitive resources, reconstruct their illness experience, make timely physical and psychological adjustments, and respond positively to trauma. Self-efficacy should be strengthened through cognitive-behavioral interventions. A supportive environment for emotional expression should be established through narrative nursing care, couple communication training, and similar approaches. Health behaviors should be reinforced through the development of targeted intervention programs and phased feedback, and behavioral maintenance should be supported through peer support groups and family support. In summary, the integration of multiple resources to strengthen positive coping and health behaviors in patients represents an effective pathway for promoting PTG.31,63 Limitations It should be noted that as a structural modeling method, the core of ISM lies in revealing the structural relationships (drive-dependence relationships) among factors rather than statistical causal relationships. The ISM model constructed in this study is based on literature evidence and expert consensus and belongs to a hypothesis-generating framework. Its hierarchical structure and action pathways need to be further verified through empirical studies on patient cohorts (such as structural equation models, longitudinal tracking studies, etc). Additionally, the sample size of the experts is relatively small, and the expert opinions may still be influenced by the cognitive preferences of individual experts, which to some extent limits the robustness of the results. Moreover, the expert-led relationship may involve subjectivity and confirmation bias. Finally, the literature included in this study is mainly cross-sectional and focuses on the Chinese population. The extrapolation of the research results should be treated with caution. These limitations do not affect the theoretical value of this study as a conceptual framework, but when translating the model conclusions into clinical practice, individualized judgments based on specific patient situations should be made. Conclusion This study innovatively introduced the ISM into the research on the influencing factors of PTG in patients with gynecological malignant tumors. It systematically identified 13 core category factors and 37 operational sub-factors, and constructed a four-level hierarchical system of “objective foundation layer → psychological intermediary layer - social support layer → coping behavior layer”. It proposed a conceptual framework of the multi-factor synergy of PTG. In this framework, the demographic and disease characteristics elements of the objective foundation layer were positioned as the bottom-level factors. Although they are not easy to change, they can provide references for high-risk population screening and clinical resource allocation. The structural relationships of coping methods and health behaviors in the coping behavior layer based on the model were positioned as the most direct influencing factors of PTG, and intervention attempts could be made through health behavior promotion methods within a short period of time. The psychological intermediary layer and social support layer, as the core hubs, played a mediating and transforming role in the model, and there might be interactive influences between them. Further empirical research is needed to verify these. Based on this model, nursing practice can consider the following directions: prioritize the coping behavior layer and attempt to activate PTG through problem-oriented coping training and health behavior reinforcement methods; continuously explore psychological and social collaborative intervention, matching corresponding support resources for the weak links in the psychological intermediary layer; precisely identify high-risk populations in the objective foundation layer to achieve early warning and stratified management.