Cognitive deterioration upon admission is an important risk indicator of falls in the respiratory ward

However, this tool has not been adapted according to the specific characteristics of each ward, although many studies have been made on falls in chronic disease wards and in patients with cerebral infarction. In addition, no fall risk assessment tool to predict falls in hospitalized patients from each department has been statistically verified to predict the accuracy, suggesting that patients at risk of falls may not be accurately grasped. As the cause of falls differs in each ward, dealing with each department is desirable. The fall rate in respiratory wards is known to be high [3,4], but the cause remains unclear. Therefore, this study aimed to analyse the risk of falls in respiratory wards.


Introduction
Japan has become a super-aging society, and the proportion of elderly inpatients is continuously increasing. As a result, falls are the most common type of medical accident, and its incidence is increasing each year [1]. Prevention of falls is an important issue for medical safety because falls not only require new treatment but also lengthen hospital stays and increase medical expenses [2].
Effective intervention to prevent falls involves screening for falls and preselecting high-risk patients upon admission. The usefulness of a fall risk assessment tool has been suggested and is used in several facilities in japan.
However, this tool has not been adapted according to the specific characteristics of each ward, although many studies have been made on falls in chronic disease wards and in patients with cerebral infarction. In addition, no fall risk assessment tool to predict falls in hospitalized patients from each department has been statistically verified to predict the accuracy, suggesting that patients at risk of falls may not be accurately grasped. As the cause of falls differs in each ward, dealing with each department is desirable. The fall rate in respiratory wards is known to be high [3,4], but the cause remains unclear. Therefore, this study aimed to analyse the risk of falls in respiratory wards.

Patients and methods
This study examined 361 patients (including multiple hospitalizations) who were admitted to our hospital's respiratory ward for at least one week from April 1, 2018, to March 31, 2019. Patients in other departments and with insufficient data were excluded.
The survey items were as follows: falls, age at admission, gender, main illness, complications, devices (oxygen administration, infusion, thoracic drain tube, tracheal cannula), body mass index (BMI), hemoglobin (Hb), Alb, fall assessment sheet score upon admission, presence of rehabilitation, discharge destination, and Functional Independence Measure (FIM). However, FIM was measured only in rehabilitation patients (N=216).
FIM was performed using a 7-point scale, with 13 items on motor subscale and 5 items on cognition subscale for daily life activity independence. The FIM is one of the most commonly used in rehabilitation programmes in japan. The FIM total score (T-FIM), the FIM motor score (M-FIM), and FIM cognitive score (C-FIM) were compared a fall group to a non-fall group. This measurement was performed by a physiotherapist/occupational therapist after admission and immediately before discharge. Patients were divided into a fall group (N=18) who fell and a non-fall group (N=343) who did not fall. Pearson's X 2 test and Mann-Whitney U-test were performed to determine the occurrence of falls.

Results
Total fall rates in our wards were 1.80 falls per 1,000 patient days. In the fall group, the length of hospital stays, the number of rehabilitation introductions, and the fall assessment sheet score were significantly high, but the difference in BMI, Alb, and Hb was not significant (Table  1 and 1 A). Table 2 shows the examination results of the main diseases and the environment. Lung cancer was the most common in-patient disease, followed by community-acquired pneumonia and aspiration pneumonia. No between-group differences were observed based on disease variability. Majority of lung cancers were advanced stage, and 42 patients had brain metastases. Best supportive care comprises 34% of treatment strategies. No significant differences were found in the following comorbidities: cerebrovascular disease, neurological disease, epilepsy, dementia, and mental illness. The turning point was that the hospital was frequently transferred in the fall group. As regards the environmental factors, 94% of patients received intravenous infusions, and all patients who fell down received infusions. The fall group had no thoracic drain tube and tracheal cannula inserted.
The fall assessment sheet score is divided into a low-risk group, (0-5 points), medium-risk group (6-10 points), and high-risk group (11-16 points), with fall rates of 1.9%, 3.4%, and 9.4%, respectively. The fall score was significantly higher in the fall group than that in the non-fall group (Table1 and 1A), and the difference in the distribution according to the risk level was found. In the examination according to each item, only the cognitive ability was significantly lower in the fall that that in the non-fall group (Table1), whereas other items showed no significant differences (Table 3).
Except for patients who underwent rehabilitation intervention, FIM-based studies showed that no difference was found M-FIM upon admission. Only C-FIM was significantly lower in the fall group with a significant difference (Table 4). On discharge, exercise improvement was observed. When comparing before and after hospitalization results, improvement in C-FIM was poor. M-FIM improvement was also poor in the fall group (

Discussion
In recent years, inpatients have become increasingly aging in Japan. In this study, 85% of inpatients are elderly people aged ≥ 65 years, and the number of inpatients highly at high for falling is increasing [4].
As measures to prevent falls at our hospital, nurses evaluate the fall risk of patients using a fall assessment sheet in Japan, whereas the fall risk assessment of patients under rehabilitation intervention is performed in the rehabilitation department using FIM. If the risk of falling is over the intermediate risk, information is shared at the conference and measures are taken to improve the breathing state and device arrangement. When falls occur, medical professionals submit an incident report and future countermeasures should be considered. The fall rate in acute care hospitals varies among different reports but been reported from 1.38 to 4.3 % falls per 1,000 patient days [3,5,6]. Sato et al. reported that the fall rate in respiratory medicine and rheumatology wards was higher than the overall fall rate [6]. However, in our ward, the fall rate was 1.80 % falls per 1,000 patient days, and our measures were considered effective.
In this study, a difference was found only in the length of hospital stay and the place of discharge according to the patient background. Dunne et al. reported that falls in acute care hospitals make discharges 2.4 times more difficult [7]. In this study, no differences were found in terminal patients, background disease/blood data, comorbidities, and motor function. Patients were considered difficult to discharge to their respective homes and had prolonged hospital stay because of the fall. Motor function improvement in the fall group is poor even under rehabilitation intervention and is important to prevent falls (Table 5).
In japan, the fall assessment sheet is commonly used, and results of this study also showed a significant difference between the fall and nonfall groups. With this, the risk for falls can be sufficiently evaluated even in the respiratory ward. However, when comparing the items in detail, only the difference in cognitive dysfunction was significant based on the results in the fall group, which was the same as in FIM. The cognitive function deterioration could lead to falls due to misjudgement [8].
In addition, environmental characteristics in the respiratory ward include several devices connected to the oxygen tube, the infusion line, and the patient. In this study, 94% of inpatients were treated with intravenous therapy, and approximately half of them received nasal or mask oxygen. Patients receiving oxygen are reported to be at increased risk of falling [9]. In an environment where the risk of falls is high, lack of cognitive ability to grasp changes in the surrounding environment and physical changes seem to have been the most important factors that cause falls.
In addition, as shown in Table 5, improvement of exercise items can be expected by improving the general condition and rehabilitation accompanying disease remission; however, cognitive function improvement is poor. This suggests that cognition function evaluation upon admission is important to assess the risk for falls in the future.

Limitations
The limitations of this study are as follows: the analysis period is only 1 year and further accumulation of the number of falls is necessary. In this study, the evaluation was limited to simple items such as assessment sheets, and therefore, examinations using cognitive function tests such as Hasegawa's formula should have also been used. In addition, FIM was measured only for rehabilitation-introduced patients. However, a bias often occurs in patients undergoing rehabilitation because they are often introduced in patients with reduced ADL. In addition, fall assessment and FIM tend to be similar, and we analysed them to be reliable in this study. All cases should be investigated in the future. A fall assessment sheet specifically made for respiratory wards should also be considered with emphasis on cognitive skills.

Conclusion
During respiratory hospitalization, patients used many devices, making them highly at risk of falling. As the value of the fall assessment sheet (FIM) decreases, the fall rate increases. In addition, cognitive deterioration during hospitalization is the most important cause of falls, and therefore, these patients need further measures to prevent falls.