Reach Us +447480022765
All submissions of the EM system will be redirected to Online Manuscript Submission System. Authors are requested to submit articles directly to Online Manuscript Submission System of respective journal.

Predictors of Injury Severity for Inpatient Falls in Taiwan: A Retrospective Study

Ching-Mei Chang1*, Cheng-Fan Wen2, Hsien-Feng Lin3

1Department of Nursing, Tungs’ Taichung Metro Harbor Hospital, Taiwan

2Department of Medical Research, Tungs’ Taichung Metro Harbor Hospital, Taiwan

3Department of Family Medicine, China Medical University Hospital, China

Corresponding Author:
Ching-Mei Chang
Department of Nursing
Tungs’ Taichung Metro Harbor Hospital, Taiwan
Tel: 886 (4)2658-1919

Received Date: 30/11/2018; Accepted Date: 04/01/2019; Published Date: 14/01/2019

Visit for more related articles at Research & Reviews: Journal of Nursing and Health Sciences


Background: Falls consist of one of the most frequently reported adverse hospital events. Little is known about the risk of severe injury for inpatients due to falls. The purpose of this study was to determine factors associated with fall injury severity.

Methods: We performed a retrospective study with subjects from a teaching hospital of central Taiwan and collected relevant data from the Adverse Event (AE) reporting system from 2008 to 2013.

Results: A total of 498 patients who had fallen during hospitalization were enrolled in the study. We identified risk factors for severe injuries by logistic multiple regression analysis, medical units (P=0.000, P<0.05), age (P=0.031, P<0.05), lacking a companion (P=0.023, P<0.05), and frequent falls (P=0.023, P<0.05) were significantly correlated with the severity of the falls. The severity of the fall 4-5 times and the severity of the fall more than 6 times are almost the same. Single and recurrent falls resulted in different degrees of injury with a significant correlation between the frequency of falls and the injury severity.

Conclusions: Severe injuries from falls are significantly correlated with ward units, number of falls, and the age. The severity of falls in psychiatric and rehabilitation ward patients was greater than that of patients in acute wards. The injury severity was also correlated with the number of falls. It may be possible to establish a fall prevention model that takes into account specific fall risk factors.


Inpatients; Falls; Injury severity; Retrospective study


Patient safety is an important issue for the World Health Organization. Falls are amongst the most frequently reported adverse hospital events[1]. A fall is operationally defined as “unintentionally coming to lay on the ground, floor or other lower level”[2,3]. The Taiwan Joint Commission on Hospital Accreditation monitors the incidence of patient falls reported from the 513 healthcare organizations, found fall incidence ranging from 0.05% to 0.06%. Regarding hospital fall incidents, the severities of fall injury were no injury (48.1%), had an injury (51.0%). This report injury rate was higher than the USA report fall injury rate 20-30%[4,5]. The fall-related retrospective studies report, some country patients suffered from moderate to serious injuries, and approximately the population is aged 65 years old or older[4-6]. The results of regression analysis showed that the variables include companions, lower limb weakness, and postural hypotension was all positively associated with fall-related injuries[7].

Fall-related injuries include bruises, lacerating wounds, fractures of upper and lower extremities, hip fractures, and intracranial injuries[8,9]. The harm from falls increases the burden of morbidity[10,11], and can result in an added financial burden to patients, families, and the society[8]. For medical institutions, inpatient falls may prolong hospital stays, increase medical staff anxiety, and lead to patient’s complaints and litigations[12,13]. Thus, reducing the number of severe injuries from falls is important.

This study hypothesis the severity of falls with risk factors associated (Eg: companion, the frequency of falls, fall of the situation, medical unit, risk assessment) haven’t the correlation. And single fall and recurrent falls with risk factors aren't the correlation. The main purpose of this study was to characterize the situations leading to inpatient falls and to identify predictive factors associated with fall severity. A secondary aim would be to differentiate the risk factors into those associated with single events and those associated with recurrent falls.


Factors Associated with Inpatient Falls

The severity of inpatient falls has been associated with factors such as the fall being unwitnessed (78%), the fall location (47%), the previous patient activity (27%), the presence of a physical impairment/frailty (9.5%) and the presence of cognitive impairment/confusion (9.2%)[14]. The risk factors for falls are classified into intrinsic and extrinsic factors and include age, physiological disease, lower-extremity weakness, poor hand-grip strength, balance disorders, visual defects, cognitive impairment, and the use of multiple drugs, lighting, defective assistive devices, slippery floors, race, and certain medications[11,15].

Injury Severity of Inpatient Falls

Falls can occur either one time or multiple times in one patient and the risk factors associated with them can differ. Recurrent falls are associated with higher mortality[15,16] and elderly patients, those who live alone, or those with unstable gait are prone to recurrent falls[17-19]. Studies have indicated that patients over the age of 60 have a higher incidence of falls[20,21]. A significant correlation has been found between the numbers of falls experienced by a patient and his or her age or the presence/absence of a companion at the time of the fall[21]. Most patients’ fall-injuries comprise skin tears and bruising[22]. In 4–6% of cases falls are severe and include fractures, subdural hematomas, bleeding and even death[22]. More than 50% of injury-related hospitalizations occur among people 65 years or older. Additionally, falls cause more than 95% of all hip fractures in the elderly[23]. Even though the literature includes studies of factors that lead to patients’ falls, the factors associated with the severity of those falls have not yet been identified. With this study, we tried to answer this question in order to provide information necessary for planning strategies to reduce the severity of fall-related injuries.


Study Design and Setting of the Study

This is a retrospective study conducted in a private teaching hospital in central Taiwan. At the time, the hospital had on average 580–650 inpatients daily from internal medicine, surgery, gynecology and obstetrics, podiatrists, ICUs, rehabilitation and psychiatry wards. The average hospital stay at the time was 5 to 6 days.

Population and Sample

Subject’s histories were collected from the Adverse Event (AE) reporting system from January 1, 2008, to December 31, 2013. We included all inpatients with falls reported by the AE reporting system and excluded outpatients.

Study Definition and Data Collection

The AE reporting system is an electronic system that allows clinicians, health-care managers to record and manage adverse clinical incidents that occur in a hospital setting. The system on falls includes the basic information for the patient (gender, age, medical record number, bed number) and the information about the fall (high-risk group, health status, medications, environmental factors, location, situation, presence or absence of a companion, severity of the fall, and the medical unit). The severity of falls can be divided into none injuries; minor injuries, for example, minor skin tears, bruising; moderate injuries, for example, soft tissue injuries; major injuries, for example, head injury, fracture; and death. Minor injuries require little or no treatment, moderate injuries require minor medical or nursing treatment or observation, and major injuries need medical treatment and consultation.

Statistical Analysis

Data were manually coded and entered into a database for analysis with the SPSS software version 19 (IBM SPSS Inc., Chicago Illinois). The descriptive statistical analysis is presented as percentages, means and standard deviations. And, the inferential statistical analysis included Pearson’s correlation and Bonferroni correction and Logistic multiple regression analysis. A p-value less than 0.05 was considered statistically significant.

Ethical Considerations

This study adhered to all relevant research ethical criteria and the institutional review board of our hospital approved the study (TTMHH104025).


General Condition of Patients with Falls

The cases of 498 patients who had fallen during hospitalization between 2008-2013 years were used in the study. Patients with companions had more falls than those without one (62.7% vs. 37.3%, respectively). The majority of patients fell only once (94%) and the severity of most falls was minor (38.8% were none injuries and 38.2 % had only minor injuries). Most falls (38% occurred in patients 65 years or older; 32.1% occurred in patients 41-64 years). The acute medical units were the most frequent units with falls in the hospital (82.9%) (Table 1).

Item Category Number Percentage
  No 186 37.3
Yes 312 62.7
Frequency of falls      
  One time 468 94
Twice-thrice times 12 2.4
Four-five times 8 1.6
Six and more 10 2
Severity of injury      
  none injuries 193 38.8
Minor injuries 190 38.2
Moderate injuries 96 19.3
Major injuries 19 3.8
  0-18 years 79 15.9
19-40 years 70 14
41-64 years 160 32.1
>65 years 189 38
Medical unit      
  Psychiatrists 58 11.6
Rehabilitation 27 5.4
Pediatrics 49 9.8
Obstetrics and Gynecology 14 2.8
  Medical 130 26.1
Surgical 180 36.2
Intensive care unit 7 1.4
Others 33 6.7
Risk assessment      
  Low risk 281 56.4
High risk 217 43.6

Table 1 Characteristics of patients with falls, N=498.

Predicted Factors Associated with the Severity of Falls

We identified medical units (P=0.000, P≦0.05) and age (P=0.031, P≦0.05) as independent factors associated with the severity of falls. Also, if children were excluded from the analysis, the presence of a companion was associated with the severity of the fall (P=0.023, P≦0.05). According to our results, the specific medical unit (P=0.000) was the most important factor associated with the severity of falls followed by the fall frequency (P=0.023, P≦0.05), the presence of a companion (P=0.023, P≦0.05) and the patient’s age (P=0.036, P≦0.05 ) (Tables 2 and 3).

Item Category Mean SD P value
Companion       0.004*
  No 2.02 0.870  
Yes 1.80 0.826  
Frequency of falls       0.005*
  First time 1.86 0.847  
Twice-thrice times 1.92 0.739  
Four-five times 2.13 0.835  
Six and more 2.80 0.422  
Fall of situation       0.069
  Sleeping 2.00 0.707  
Trips to the toilet 1.75 0.889  
Walking 2.03 0.770  
Moved to/from bed 2.00 0.862  
Bath/wash 1.75 0.676  
Siting 1.92 0.717  
Picked up 1.76 0.970  
Use a walker 2.28 1.018  
Other factors 1.78 0.796  
Medical unit     0.000*
  Psychiatrists 2.47 0.731  
Rehabilitation 2.63 0.792  
Pediatrics 1.69 0.508  
Obstetrics and Gynecology 1.43 0.514  
Medical 1.85 0.846  
Surgical 1.65 0.801  
Intensive care unit 1.43 0.787  
Outpatient 2.19 0.981  
Risk assessment     0.075
  Low risk 1.82 0.822  
High risk 1.96 0.878  

Table 2 Correlation between falls injuries severity and related factorsN=498. Note:≦0.05.

Items Coefficient T value P value
Medical units 1.613 12.026 0.000*
Psychiatrists 0.616 4.500 0.000*
Companion 0.036 0.438 0.023*
Age 0.003 2.169 0.031*
Risk factor of falls 0.052 0.698 0.485
frequency of falls 0.036 0.438 0.023*

Table 3 Logistic multiple regression analysis of fall severity. Note: ≦0.05.

Further analysis revealed that the surgery ward had the highest fall percentage (180/498, 36.1%) of the acute wards. And, that this fall percentage (36.1%) was greater than the fall rate of the medical ward (130/498, 26.1%), the pediatric ward (49/498, 9.8%), the obstetrics and gynecology ward (14/498, 2.8%), and the intensive care unit (7/498, 1.4%) (Table 1). Additionally, even though the acute wards had the higher number of falling inpatients, the psychiatric and rehabilitation wards had the highest percentage of severe falls (both with P=0.000, P≦0.05) (Table 3).

Factors Associated with Frequency of Falls

Our results also showed that recurrent falls caused more major injuries. Four or five times of falls caused the severity of injury similar to six times of falls (Table 4). The correlation between the frequency of falls and the severity of falls was shown significant statistically (P=0.009, P≦0.05). Patients with recurrent fall most had moderate injuries (14/30 patients; 46.7%), while those with a single fall experienced mostly none injuries (187/468 patients, 40.0%) and minor injuries (180/468 patients, 38.5%). Variables such as the presence of a companion (P=0.001, P≦0.05), the severity of injury (P=0.001, P≦0.05), the risk factors of fall (P=0.009, P≦0.05), the medical units (wards) (P=0.000, P≦0.05), and the patients’ ages (P=0.001, P≦0.05) had associated with single or recurrent falls. Patients with companions had a higher percentage of single falls (302/468, 64.5%) than those without a companion (166/468, 35.5%). However, patients without a companion had more recurrent falls (10/30, 33.3%) than those with a companion (20/30, 66.7%).

Severity of falls frequency of falls Mean SD P value 95% CI
Low High
first time 2-3 times 0.156 0.280 1.000 -0.59 0.90
4-5 times -0.303 0.297 1.000 -1.09 0.48
Above sixth -0.996* 0.254 0.001* -1.67 -0.32
2-3 times   4-5 time First time -0.156 0.280 1.000 -0.90 0.59
4-5 times -0.458 0.404 1.000 -1.53 0.61
Above sixth -1.152* 0.374 0.013* -2.14 -0.16
First time 0.303 0.297 1.000 -0.48 1.09
2-3 times 0.458 0.404 1.000 -0.61 1.53
Above sixth 0.693 0.386 0.440 -1.72 0.33
Above sixth First time 0.996* 0.254 0.001* 0.32 1.67
2-3 times 1.152* 0.374 0.013* 0.16 2.14
4-5 times 0.693 0.386 0.440 -0.33 1.72

Table 4 Severity and frequency of falls. Note: ≦0.05.

Furthermore, the health condition was the most important factor associated with single or recurrent falls. The surgery ward had the highest percentage of single falls (176/468, 37.6%) and the psychiatric ward had the highest percentage of recurrent falls (21/30, 70.0%). Patients 65 years or older had the highest percentage of single fall events (184/468, 39.3%). And, those with ages ranging from 41 to 64 years had the highest percentage of recurrent falls (15/30, 50.0%). Single falls occurred more frequently while going to the toilet (155/468, 33.1%) whereas recurrent falls were associated with moving to/from the bed (9/30, 30.0%) (Table 5).

Factors   Frequency of falls P value
single fall
n     %
recurrent falls
n     %
n    %
Companion No 166 35.5 20 66.7 186 37.3 0.001*
Yes 302 64.5 10 33.3 312 62.7  
Situation Sleeping 14 3.0 3 10.0 17 3.4 0.063
Going to the toilet 155 33.1 6 20.0 161 32.3  
Ambulating 57 12.2 8 26.7 65 13.1  
Move to/from the bed 105 22.4 9 30.0 114 22.9  
Bath/wash 23 4.9 1 3.3 24 4.8  
Siting 24 5.1 0 0.0 24 4.8  
Take someone 17 3.6 0 0.0 17 3.4  
Use a walker 18 3.8 0 0.0 18 3.6  
Others 55 11.8 3 10.0 58 11.6  
Severity of injury None 187 40.0 6 20.0 193 38.8 0.001*
Minor 180 38.5 10 33.3 190 38.2  
Moderate 82 17.5 14 46.7 96 19.3  
Major 19 4.1 0 0.0 19 3.8  
Risk factors of fall Health cond. 255 54.5 20 66.7 275 55.2 0.009*
Drug reaction 26 5.6 5 16.7 31 6.2  
Environment 77 16.5 4 13.3 81 16.3  
Others 110 23.5 1 3.3 111 22.3  
Medical units OB/GYN 14 3.0 0 0.0 14 2.8 0.000*
Pediatric 48 10.3 1 3.3 49 9.8  
Medicine 128 27.4 2 6.7 130 26.1  
Surgery 176 37.6 4 13.3 180 36.2  
ICU 7 1.5 0 0.0 7 1.4  
Psychiatrists 37 7.9 21 70.0 58 11.6  
Rehabilitation 25 5.3 2 6.7 27 5.4  
Others 33 7.0 0 0.0 33  6.7  
Age 0-18yrs 78 16.7 1 3.3 79 15.9 0.001*
  19-40yrs 61 13.0 9 30.0 70 14.0  
41-64yrs 145 31.0 15 50.0 160 32.1  
>65yrs 184 39.3 5 16.7 189 38.0  

Table 5 The analysis of factors associated with frequency of falls, N=498.


The present study showed that inpatients 65 years and older comprised the majority of those falling. This agrees with previous reports[12,16,21]. The most frequent situations associated with hospital falls in this study were going to the toilet, moving to and from the bed, and ambulating. This too is similar to other reports[14,21]. Among the acute hospital wards, the surgery ward was the location with most falls. We found that falls in the surgery ward occurred in similar situations (moving to and from a bed, during toilet or bathroom trips) and that most falls occurred during the night and morning. Our results also show that patients with companions suffered more falls than those without one (62.7% vs. 37.3%, respectively). However, the reasons for that included situations in which the companion was not available (companion was asleep, was gone, was careless or unable to help, or the patient had refused help). Other studies have found similar results[14,21].

Predicted Factors for the Fall Severity

The medical unit was a stronger predictor of the fall severity than the patient’s age. For instance, patients from psychiatric or rehabilitation wards experienced the more severe falls injuries. This is likely due to the gait and stance impairments in elderly patients in those wards, due to medication and physical weakness[24]. It has been reported that patients with arthritis or psychiatric history are more prone to falls due to their prescribed medications[15]. In addition, excluding children, recurrent falls were the main factor associated with the fall severity. Therefore, psychiatric or rehabilitation wards and inpatients above 65 should be targeted by plans to reduce the risk of falls and consequently reduce fall severity.

Correlation between the Fall Severity and Frequency of Falls

The present results showed that the severity of fall injury was correlated with recurrent falls (P=0.009). Compared with recurrent falls, single falls often caused none injury or minor injury. However, patients with recurrent falls had a moderate degree of injury. In addition, there was a significant correlation between the number of falls and the severity of falling events occurring more than 3 times causing similar higher injury severities (Table 4). This is consistent with the findings that mortality is correlated with recurrent falls[15,16] and shows why it is important to prevent recurrent fall for the patient safety. We further analyzed factors associated with either single or recurrent falls. We found that patients older than 65 years fell more often than the rest, but those with ages ranging from 41–65 were the ones having most recurrent falls. This was different from other reports showing older patients had the higher recurrent falls percentage[18]. The discrepancy may be due to the different study populations. In our study, patients with recurrent falls were mostly from the psychiatric and rehabilitation wards. These patients have more physical health and drug factors and the prevention of falls tends to one way education, so it is easier to recurrent the fall. Accordingly, patients with unstable gait or under certain medications have been found to be more prone to recurrent falls[25,26]. The single fall often occurs in the situation of having a companion. Recurrent falls often occur in the unaccompanied situation, mainly because a single fall is related to the negligence of the companion or lack fall prevention, but the recurrence falls are related of the inadequate of the companion. Presumably, patients with companions get help more often after their single fall and prevent recurrent falls. Single falls occurred more frequently when patients were going to the toilet, whereas recurrent falls were associated with moving to from the bed, and with wheelchair transfers (probably due to improper use of patient aids). We found that patients with dependence on patient aids, those with unstable gait, or without a companion had a higher risk of falls. This is consistent with the finding that patients living alone, with unstable gait, or who use a walker suffer more recurrent falls[17,18]. Therefore, the factors associated with single falls and those associated with recurrent falls are quite different. To prevent fall injuries from recurrent falls, attention should be paid to inpatients using patient aids, those with unstable gait, and those without a companion.


According to our research, severe injuries from falls are significantly correlated with ward units, number of falls, and the patient’s age. The severity of falls in psychiatric and rehabilitation ward patients was greater than that of patients in acute wards. The injury severity was also correlated with the number of falls. Recurrent falls caused more severe injuries, and those suffering more than 3 falls experienced similarly severe injuries. Patients 65 years or older had the highest percentage of single falls, and those between 41 and 64 years had the highest percentage of recurrent falls.

Relevance to Clinical Practice

Based on our study we suggest several preventive measurements.

1) Targeting specific inpatient populations in fall prevention models (psychiatric and rehabilitation wards patients need an improved fall prevention plan).

2) Paying attention to high-risk individual patients such as those older than 65 years, those without a companion, those requiring patient mobility aids, and those with unstable gait.

3) Preventing recurrent falls by evaluating the cause of the first fall and educating the patient in fall prevention practices.

It is important to acknowledge the problem of inpatients falling and to assess the repercussions on the functional capacity of the patients. Multicomponent interventions are more effective in preventing falls in clinical practice and include setting up a fall prevention protocol implemented in the institution and using signage or bed/chair alarm systems. Based on our finding that psychiatric patients may have fallen due to sedation and inherent psychomotor changes, we suggest periodic medication effect reviews or increased nursing supervision. Attention should be given to the hospital environment to reduce the risk of falls related to physical structures and furnishings, including in the patient's bedroom and bathroom.

Research Limitations

This study was limited that it is a retrospective analysis and from the Adverse Event (AE) reporting system from a single teaching hospital, so the data may be biased and limits its inference potential. In addition, this study not collects detail on the number of medications the subjects are on, the future study could deep explore. A longitudinal study may help better understand predictive factors to help prevent future falls.