Q-methodology was used in this study. It was developed to explore a person’s viewpoints (e.g., feelings, perspectives, and attitudes) on a given topic, systematically identify the similarities and differences between these viewpoints, and probe the reasons behind them (19, 20). Therefore, this methodology is considered suitable for patients’ subjective judgment of the influencing factors of fall self-awareness. This study is not registered as it is not part of a clinical trial. For the specific operation process, see Figure 1.

In this Q-methodology study, participants were purposively recruited from inpatient wards during their hospital stay to capture diverse perspectives. Hospitalized patients aged ≥60 years with a length of stay >24 h were identified by ward staff. Those with severe dementia, communication barriers, or major mental illness were excluded. All enrolled participants provided written informed consent. Data on age, sex, education level, and hospitalization length were collected.

The processes of statement development, data collection and analysis and reporting for this study were conducted in line with published guidelines for Q-methodology studies (19, 21, 22).

Statistical analysis of the data was performed using PQ Method 2.35, PASW 18.0, and KADE 1.2.1 software (24). The data were first entered into PQ Method 2.35 to prevent duplication or omission. Each participant’s ranking of the statements was transformed into an array of numerical data ranging from −5 to +5 according to their placement in the columns of the score sheet. Subsequently, PASW 18.0 was employed for statistical description: measurement data were expressed as mean ± standard deviation, while categorical data were summarized as frequencies, percentages, or rates. The Kaiser–Meyer–Olkin (KMO) measure and Bartlett’s test of sphericity were conducted to verify that the data met the prerequisites for factor analysis. Finally, KADE 1.2.1 was used to perform principal component analysis for factor extraction. The analytical process included the following steps (25): (1) Factor extraction. Principal component analysis was applied to extract factors, retaining those with eigenvalues >1 and a cumulative variance contribution ≥40%. To facilitate interpretation and labeling, the extracted factors were rotated using the varimax method, which enhances the clarity of factor classification. (2) Calculation and classification of factor loadings. Factor loadings reflect the strength of the correlation between participants (P-sample) and the factors. Higher absolute values of factor loadings indicate greater statistical significance and a stronger association between the participants and the corresponding factors. The threshold for a statistically significant factor loading was determined by dividing the multiplier for the desired significance level by the square root of the number of statements in the Q set (multipliers: 3.29 for p < 0.001; 2.58 for p < 0.01; 1.96 for p < 0.05) (22). For a factor to be considered defined, it was required to be loaded by at least two participants. Any participant whose loadings reached significance on two or more factors was considered confounded and therefore excluded from subsequent analyses. (3) Calculation of factor scores. Following factor extraction and rotation, the KADE 1.2.1 software automatically calculated the factor scores for each statement (Q-sort). These scores were then ranked to form a factor array, resulting in an idealized ranking for each distinct viewpoint. Each factor was interpreted using both its factor array and the qualitative data from participants who loaded significantly on that factor. In this context, a factor represents a shared perspective on the elements influencing fall-related self-awareness. This perspective is an abstract construct that encapsulates the common viewpoint shared by participants whose statement rankings were sufficiently similar to load significantly on the same factor.

This study was approved by the Ethics Committee of Tianjin Medical University General Hospital, on 23 January 2025 (approval No.: IRB2025-YX-039-01).

A total of 15 participants aged 68.53 ± 7.836 years were recruited; 73.3% (n = 11) of them were men, as detailed in Table 1. In this study, the KMO measure was 0.797, and Bartlett’s test of sphericity yielded an approximate χ² value of 397.734 (p < 0.001), indicating that the data were suitable for factor analysis. Using the formula multiplier/√n, the significance threshold was calculated as 3.29/√34 ≈ 0.564. Based on the criterion, participants P10 and P12 were removed due to cross-loadings. Consequently, 15 participants were retained and categorized into three distinct factors, as presented in Table 2. Three perspective factors were derived from the Q-sorting results (Table 3). The explained variance by the three perspectives were 50% (eigenvalue = 7.441), 10% (eigenvalue = 1.516), 8% (eigenvalue = 1.205). The total explained variance was 68%.

Figure 3 shows the frequency distribution of the rankings of the 34 statements by the 15 participants. The colors indicate the level of agreement (y-axis) of the participants with the statements (x-axis). A +5 (dark red, agree) indicates that the participant placed the statement in the +5 box on the score sheet, while a −5 (dark green, disagree) represents the opposite.

According to Figure 3, the Q-statements S23 (“The type and number of diseases affects one’s self-awareness of falls”), S10 (“Literacy affects one’s self-awareness of falls”; ranked −4, −5, −4 across factors), and S27 (“Fall Prevention Policy can affect one’s self-awareness of falls”; ranked −2, −3, −2 across factors) showed the highest level of consensus. This consistency suggests that disease type and number are factors they consider to be associated with self-awareness of falls, while education and fall prevention policies are two factors that are overlooked by patients. The values in parentheses represent the Q-sort scores of each statement across the three factors.

Based on the results of principal component analysis (Table 3), a total of three factors were extracted in this study. The corresponding Q sample sizes are: 7, 2, 6.

Seven participants were mapped to Factor 1, comprising five males and two females. The corresponding subjects were P2, P3, P4, P6, P7, P8, P9. As shown in Table 1, these patients had a mean age of 72.85 ± 7.471 years and had a shorter duration of hospitalization compared to those in Factor 2 and Factor 3. Two patients with previous falls were categorized in this group. Since these patients primarily based their judgment of fall awareness on concerns related to the fall itself, we named this perspective the Fall-Derivative Type. For example, the q-statement S15 “The level of fall knowledge will affect one’s self-awareness of falls” (Z-score, Q-sort: 1.66^**, 5) and S24 “The unpredictability of falls can affect one’s self-awareness of falls” (1.37^**, 4), suggesting that hospitalized older patients perceive fall-related content as a major influencing factor, see Supplementary Table S2.

Two participants were mapped to Factor 2, including one male and one female. The mean age of these patients was 62.50 ± 3.536 years. A notable feature is that they were both cancer patients, had the longest hospitalization time, and lived in rural areas. They all considered their own risk of falling out of concern for their families and children, so they were named Family-Oriented Type. Among them, S31 “Economic conditions can affect one’s self-awareness of falls” (2.03^**, 5) and S21. “The number of children can affect one’s self-awareness of falls” (1.56^**, 4) are the most important influencing factors.

In addition to the uniform negative factors, the main statements from which participants in this group dissented, differing from their peers, were: “The degree of attention paid to fall prevention will affect one’s self-awareness of falls” (−0.94^**, −3); “The level of fall knowledge will affect one’s self-awareness of falls” (−1.25^**, −4). The low ranking of these statements indicates that they do not value the impact of the fall itself.

Six participants were mapped to Factor 3, of which five were male and one was female. The corresponding subjects are P1, P5, P11, P14, P15, P17. The patients in this group had a mean age of 65.50 ± 7.120 years, a relatively long hospital stay, a high level of education, and no fall experience. They are more concerned about their condition and assess their risk of falls based on their physical health. Hence the name Healthy-State Type. Such patients believe that the judgment of whether they are at risk of falling is based on their current physical condition, stating that “changes in their condition will affect self-awareness of falls” (1.84^**, 5) and “Physical fitness can affect one’s self-awareness of falls” (1.74^*, 4) are the main influencing factor expressed by this type of patient.

Patients classified under the “Fall-Derivative” profile perceive a direct connection between their self-awareness of falls and fall-related factors, primarily including the unpredictability of falls and their level of knowledge about fall prevention. This profile represents a considerable proportion of the study sample and is characterized mainly by advanced age and a history of previous falls.

Research indicates that a history of falls is closely associated with the level of knowledge about falls. Patients who have experienced a fall tend to possess a better understanding of fall-related information than those without such a history. This may be because older adults who have experienced a fall often prioritize fall prevention in their daily decision-making, pay more active attention to potential environmental risk factors, and are consequently more motivated to proactively learn about fall-related knowledge (36, 37). At the same time, a history of falls influences older patients’ attitudes toward falling. Those who have never experienced a fall often perceive it as an accidental event, believing that preventive measures are unnecessary before a fall occurs and that taking early action would be futile. Consequently, they may underestimate the need for advance prevention or vigilance (38). For this group, health education needs to emphasize the preventable nature of falls, enhance patients’ understanding, and thereby promote improved self-awareness of falls.

Patients categorized under the “Family-Oriented” profile demonstrated a relative neglect of personal fall risk, focusing instead on the potential family distress and caregiving burden that a fall might cause. In this study, this profile was characterized by a longer duration of illness and significantly longer hospital stays compared to other types. Their perspective on factors influencing self-awareness of falls showed a distinct family-oriented tendency, which aligns with previous research (39).

Studies indicate that such patients often strive to maintain a higher level of self-awareness of falls in order to reduce the caregiving burden on their children (8, 9). This tendency stems from two main concerns: first, falls could impose additional medical expenses on their children, increasing the financial burden on the family; second, in single-child families, the absence of sibling support may exacerbate the psychological and emotional pressure on the child. As a result, these patients are more inclined to consider family welfare in their behavior. Unlike the other two profiles, however, these patients tended to overlook the direct impact of falls themselves on self-awareness. Further interviews revealed a cognitive bias: although they self-rated their knowledge of fall prevention as adequate, their actual preventive measures remained limited to basic physical protection, with clear deficiencies in understanding professional aspects such as medication side effects and disease-specific risks. This cognitive limitation reflects a tendency among long-term hospitalized patients to develop “prevention fatigue” and exhibit cognitive rigidity regarding fall prevention (13).

Consequently, interventions targeting this group should aim to reshape these cognitive patterns. Implementing a family-collaborative education model that emphasizes knowledge of medication and disease-related risks, coupled with establishing a family supervision mechanism, could help ensure the implementation of preventive measures. In addition, healthcare providers should pay particular attention to the influence of patients’ financial circumstances on protective behaviors, ensuring that intervention plans balance professionalism and feasibility. Such a multifaceted approach could more effectively enhance fall self-awareness in this patient profile.

Patients categorized under the “Healthy State” profile were characterized by relatively short hospital stays. Their self-awareness of falls was predominantly influenced by their own health status: when they perceived their health as poor, their attention to falls increased. This result is consistent with studies by Ong et al. (40), which indicates that older persons with objectively poor health may have increased self-awareness of falls, despite having decreased physical function. This heightened awareness may stem from a more realistic appraisal of their functional limitations, prompting greater vigilance.

Conversely, patients in good health may lower their alertness to fall risks due to overconfidence in their own abilities. However, interviews revealed that some older patients had an inadequate perception of their own health status, manifested as overconfidence, a lack of self-protection awareness, or underestimation of the severity of fall risks (41). This may be related to their shorter hospitalization periods. Moreover, the health status of hospitalized older adults can change with disease progression. For such patients, healthcare providers could introduce simple, user-friendly self-assessment tools during fall prevention education, encourage the development of habitual self-health monitoring, and provide timely personalized health guidance. These strategies represent promising interventions for enhancing fall self-awareness in this patient profile.

A key strength of this study lies in its inclusive design. The use of magnetic boards to facilitate the sorting task helped mitigate common participation barriers related to vision or manual dexterity, thereby optimizing the engagement experience and data quality for older patients. Furthermore, this research applies Q-methodology to systematically identify, from the patients’ own subjective perspective, a typology of factors influencing fall self-awareness, thereby providing a novel evidence base for developing targeted health education interventions in clinical practice. In order to facilitate the clinical application of patient typology, a rapid screening method suitable for busy hospital environments can be developed in the future. Healthcare professionals can briefly assess the patient’s perspective through short, targeted questions that align with these three characteristics.

This study has the following shortcomings: first, the enrolled older inpatients were all recruited from a single large tertiary hospital in Tianjin, and the sample was not multi-centered. Consequently, the relatively homogeneous background of the participants may limit the richness and diversity of the perspectives collected. The employed methodology, Q-methodology, is inherently designed to capture individual subjective viewpoints and to “discover” distinct perspectives rather than to ensure their generalizability to a broader population. The small sample size further restricts the generalizability of the findings. Future research should aim to include a larger and more diverse sample. Second, regarding the development of the Q-statements, although they were grounded in a comprehensive literature review and semi-structured interviews, the perspectives of inpatients from other provinces and cities were not incorporated. Therefore, the coverage of potential viewpoints may not be fully comprehensive. Third, during the Q-sorting process, despite the researcher spending considerable time explaining the procedure to participants, the number of statements to be sorted was somewhat large, placing high demands on them. This may have, to some extent, induced fatigue or frustration in participants, potentially leading to biases in the resulting sorts. Finally, this study had a higher proportion of male participants in the sample due to the large number of male patients who actually met the inclusion criteria during the study collection period. Given that gender may influence self-awareness of falls, this imbalance should be considered a potential limitation, as it could affect the generalizability of the findings to female older inpatients. Future studies should aim to include more gender-balanced samples to further validate and extend these findings.