Falls are a high-priority safety problem in hospitals because of the resulting poor health outcomes and costs [1]. Unlike community-dwelling subjects, inpatients in acute care hospitals are likely to be at a high risk of falls because their disease status changes continuously and rapidly, and various medications and treatment modalities can cause falls [2, 3]. Therefore, falls in acute care hospitals are a major challenge for healthcare providers who prioritize patient safety and require considerable care and attention from healthcare providers [3, 4].

To prevent falls in hospitals, clinicians and researchers are continuing their efforts to identify factors related to falls while managing the known risk factors. Based on systematic reviews and current observational studies, recent clinical practice guidelines have reported that fall history (within the last 6 months), old age (> 70 years), gender (man), disease-related changes such as movement restriction and an increase in gait problems, dementia- and cancer-specific factors, medications such as anticonvulsants, antipsychotics, and sedatives, and external risk factors are significant risk factors for falls [5]. In a previous study comparing 447 fallers and 3,667 non-fallers admitted to an acute care hospital in South Korea, gender (man), low body mass index, defecation problems, gait problems, vision and hearing disturbances, and medications were reported as influencing factors for falls [6]. A systematic review of the risk factors associated with falls in non-ambulatory adults reported that the use of wheelchairs, changes in wheelchair setup and selection, performance of transfers, and impaired seated balance are risk factors for falls [7]. As described above, falls occurring in hospitals are characterized by the contribution of an individual patient’s underlying disease, factors related to changes in the current disease, and risk factors related to the hospital environment.

Recently, patients’ fall risk perception has become a promising new indicator for preventing falls [8]. Fall prevention and clinical management guidelines recommend incorporating patient perspectives as stakeholders in fall prevention and management [9]. Assessing patients’ perceptions of their fall risk can help healthcare professionals provide fall interventions and understand their fall-related behaviors [10]. Thus, fall prevention and management interventions need to be consistently performed, focusing on helping patients actively participate in fall interventions as they adjust to their condition [8]. Falls are accidents that can mostly be prevented, and managing falls with only the effort and attention of healthcare professionals has its limitations [9]. Therefore, patient participation is emphasized because patients are the subjects of fall prevention interventions aimed at decreasing falls in medical institutions [11, 12, 13].

From this perspective, scales for assessing patients’ fall risk perception, such as the falls risk perception questionnaire [14] and the fall risk self-assessment [15], are being developed and implemented. However, the falls risk perception questionnaire was developed for older community-dwelling adults with type 2 diabetes [14], and the fall risk self-assessment was developed for community-dwelling seniors [15]. Therefore, they do not reflect the characteristics of inpatients in acute care hospitals or their environmental characteristics.

Accordingly, we developed a fall risk perception questionnaire (FRPQ) comprising 27 questions regarding patients’ fall risk perceptions in acute care hospitals [16]. However, this questionnaire has limitations in terms of quickly identifying patient perceptions of fall risk in clinical practice. The questionnaire took approximately 7~10 minutes to complete, and only exploratory factor analysis was performed when developing the scale. In an acute care hospital, various treatments are rapidly provided, and the patient’s condition is unpredictable; therefore, the patient’s fall risk can easily change. Therefore, evaluating fall risk repeatedly and rapidly, as assessed by healthcare professionals and perceived by patients, is necessary in acute care hospitals. This study aimed to develop a short version of the fall risk perception questionnaire (Short-FRPQ) to quickly evaluate fall risk perception in patients hospitalized for acute treatments that are easy to repeat.

Falls are considered preventable safety problems, and the first step in their management is to assess the patient’s fall risk. Therefore, an individualized, continuous, and repeatable strategy is required to assess fall risk under changing conditions [27]. In previous studies, nurses’ fall risk assessment was primarily included as a first step in fall prevention [28, 29]. Recently, studies on patient perceptions of fall risk have increased, and it is recommended that such recognition be performed together with nurses’ fall risk assessments [9, 10, 30]. Furthermore, the patient’s condition may change rapidly depending on the treatment administered after admission to an acute care setting, and patients should adapt to the hospital environment. Therefore, we aimed to develop a scale to evaluate patient perceptions of fall risk rapidly and repetitively in acute care hospitals.

In this study, the reliability-measured internal consistency was calculated using McDonald’s omega coefficient. McDonald’s omega coefficients can be used for both unidimensional and multidimensional scales, whereas Cronbach’s alpha has been reported to be suitable for unidimensional scales [31, 32, 33]. As the difference between McDonald’s omega coefficients and Cronbach’s alpha has been reported [32], studies reporting reliability using McDonald’s omega coefficients have increased [34, 35]. Additionally, McDonald’s omega coefficients can help with brief form construction by removing items from a longer scale [31]. Therefore, we decided it would be more appropriate to use McDonald’s omega coefficients for internal consistency in this study.

In this study, McDonald’s omega coefficients of the Short-FRPQ were .90, .86, .70, and .87 for the total, I, II, and III, respectively. Although the interpretation criteria of McDonald’s omega vary depending on the study, the values were mentioned as excellent (.93~.94), strong (.91~.93), and reliable (.84~.90), and a value of .70 or greater is considered acceptable [31]. In this study, McDonald’s omega values for the Short-FRPQ were greater than .70, indicating good reliability. In future research on the development of scales with multidimensional properties, it would be desirable to calculate the internal consistency using McDonald’s omega coefficients.

This study confirmed the construct, convergent, and known-group validity of the Short-FRPQ. In this study, the construct validity of the two models—the original three-factor model and the three-factor model with error correlation—was tested by calculating the goodness-of-fit indices of the measurement models using CFA. The three-factor model with error correlation demonstrated reasonable goodness-of-fit and was selected as the final model for this investigation. The MI offered by AMOS suggested that when fitting the models, one should permit error covariance between the two environmental-related items, “I am at risk of falling because the lighting in the patient room is too low (item 12)” and “I am at risk of falling because the hospital environment is unfamiliar (item 13)”. Justifying the error correlation between pairs of items is acceptable because it frequently denotes perceived repetition in item content or nonrandom errors because of technique impacts [26, 36]. The inclusion of an error correlation between the two items was justified because they reflected the similarities evoked by the hospital’s physical environment in the fall risk set. Therefore, further investigation should be conducted to assess the environmental subdomain, which revealed an error correlation between items 12 and 13.

For convergent validity, we analyzed the correlations among the Short-FRPQ, FRPQ, KFES-I, and MFS. The Short-FRPQ displayed a strong correlation with the FRPQ, and the scale was significantly correlated with the KFES-I and MFS scores. These findings are consistent with the fact that the FRPQ scores correlated with the KFES-I and MFS scores in the original scale development study [16]. Furthermore, the Short-FRPQ I, Short-FRPQ II, and Short-FRPQ III had the highest correlations with FRPQ I, FRPQ II, and FRPQ III, respectively. Therefore, we confirmed the convergent validity of the Short-FRPQ.

For known-group validity using the MFS, a statistically significant difference was observed in the Short-FRPQ scores between the fall and non-fall risk groups. This finding is consistent with the statistically significant difference in FRPQ scores between the fall and non-fall-risk groups in the original scale development study [16]. Therefore, we confirmed the known-group validity of the Short-FRPQ.

The limitations of this study are as follows: First, the original study mainly included patients in the gastroenterology ward. Therefore, we attempted to include participants with a variety of diseases in this study, including orthopedic diseases, in which falls frequently occur [37]. Data were collected from six wards of an acute care hospital. Two-thirds of the participants underwent surgery, and the number of patients receiving medical treatment was relatively small. A previous study reported that a surgical specialty decreases the risk of falls [2]. Therefore, fewer individuals at a high risk of falling may have been included in the study. Further studies that include participants with various diseases and studies that equally include those who received surgical and medical treatment are needed for more well-rounded research. Second, the scale did not derive criterion validity based on fall occurrence because of the low number of falls.

When providing interventions for fall management and prevention, understanding patients’ perceptions and views of falls is important. In addition to the fall risk assessment by healthcare professionals using a scale with established validity, the assessment of a patient’s fall risk perception should be conducted systematically.

In conclusion, the Short-FRPQ presented good reliability and validity. Patient perceptions of fall risk should be assessed using a fall risk recognition scale that can be used rapidly and repeatedly in acute care hospitals. As the risk of falls in patients admitted to an acute care hospital is constantly changing because of rapid changes in their disease, medications, and treatment procedures, healthcare professionals should pay close attention to fall prevention in inpatients. The perception of fall risk is a promising new indicator for preventing falls, and the patients’ perceived fall risk should be considered along with the assessment of fall risk factors. As using a scale for the quick assessment of fall risk perception can potentially help prevent falls in hospital settings, it should be included in fall prevention interventions across all hospital setups. This proactive approach can significantly contribute to enhancing patient care and preventing falls.