Disability and Rehabilitation

ISSN: 0963-8288 (Print) 1464-5165 (Online) Journal homepage: http://www.tandfonline.com/loi/idre20

Effectiveness of a fall-risk reduction programme for inpatient rehabilitation after stroke Nika Goljar, Daniel Globokar, Nataša Puzić, Natalija Kopitar, Maja Vrabič, Matic Ivanovski & Gaj Vidmar To cite this article: Nika Goljar, Daniel Globokar, Nataša Puzić, Natalija Kopitar, Maja Vrabič, Matic Ivanovski & Gaj Vidmar (2016): Effectiveness of a fall-risk reduction programme for inpatient rehabilitation after stroke, Disability and Rehabilitation, DOI: 10.3109/09638288.2015.1107771 To link to this article: http://dx.doi.org/10.3109/09638288.2015.1107771

Published online: 04 Jan 2016.

Submit your article to this journal

Article views: 47

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=idre20 Download by: [University of California, San Diego]

Date: 17 February 2016, At: 08:48

DISABILITY AND REHABILITATION, 2015 http://dx.doi.org/10.3109/09638288.2015.1107771


Effectiveness of a fall-risk reduction programme for inpatient rehabilitation after stroke Nika Goljara,b, Daniel Globokara, Natasˇa Puzic´a, Natalija Kopitara, Maja Vrabicˇa, Matic Ivanovskia and Gaj Vidmara,c

Downloaded by [University of California, San Diego] at 08:48 17 February 2016

a University Rehabilitation Institute, Republic of Slovenia, Ljubljana, Slovenia; bDepartment of Physical and Rehabilitation Medicine, Faculty of Medicine, University of Ljubljana, Ljubljana, Slovenia; cFaculty of Medicine, Institute for Biostatistics and Medical Informatics, University of Ljubljana, Ljubljana, Slovenia



Purpose: To evaluate effectiveness of fall-risk-assessment-based fall prevention for stroke rehabilitation inpatients. Method: A consecutive series of 232 patients admitted for the first time to a subacute stroke-rehabilitation ward during 2010–2011 was studied in detail. The Assessment Sheet for Fall Prediction in Stroke Inpatients (ASFPSI by Nakagawa et al.) was used to assess fall-risk upon admission. Association of ASFPSI score and patient characteristics with actual falls was statistically tested. Yearly incidence of falls per 1000 hospital days (HD) was retrospectively audited for the 2006–2014 period to evaluate effectiveness of fall-risk reduction measures. Results: The observed incidence of falls over the detailed-study-period was 3.0/1000 HD; 39% of the fallers fell during the first week after admission. ASFPSI score was not significantly associated with falls. Longer hospital stay, left body-side affected and non-extreme FIM score (55–101) were associated with higher odds of fall. Introduction of fall-risk reduction measures followed by compulsory fallrisk assessment lead to incidence of falls dropping from 7.1/1000 HD in 2006 to 2.8/1000 HD in 2011 and remaining at that level until 2014. Conclusions: The fall-risk-assessment-based measures appear to have led to decreasing falls risk among post-stroke rehabilitation inpatients classified as being at high risk of falls. The fall prevention programme as a whole was successful. Patients with non-extreme level of functional independence should receive enhanced fall prevention.

Received 15 April 2014 Revised 26 August 2015 Accepted 10 October 2015 Published online 29 December 2015 KEYWORDS

Fall risk; falls; inpatient rehabilitation; prevention; stroke;


 Recognising the fall risk upon the patient’s admission is essential for preventing falls in rehabilitation wards.  Assessing the fall risk is a team tasks and combines information from various sources.  Assessing fall risk in stroke patients using the assessment sheet by Nakagawa et al. immediately upon admission systematically draws attention to the risk of falls in each individual patient.

Introduction Falls are among the more frequent complications during inpatient rehabilitation. It has been estimated that 1–5 of every 10 patients suffer a fall during inpatient rehabilitation.[1–4] The fall risk is particularly elevated in patients after stroke, patients with cognitive impairments and patients with hip fractures.[1,4,5] Fall prevention has become an important part of rehabilitation procedures. Recognising the fall risk in a patient immediately after the admission is paramount because falls often occur within the first week from admission to a rehabilitation ward.[4] Several scales are available for assessing fall risk in rehabilitation inpatients.[6] The Morse scale [7] and the

CONTACT Gaj Vidmar ß 2015 Taylor & Francis

[email protected]

Downton Index [8] have proven validity for the elderly population, but they are not particularly adapted to stroke patients. Similarly, the Fall Assessment Questionnaire scores showed notable correlation with falls during rehabilitation in men after right-hemisphere stroke, but only when combined with an assessment of impulsiveness, which requires a complex assessment procedure that is not particularly appropriate for the clinical setting.[9] There are only a few specific scales for assessing risk for falls in patients after stroke. Olsson et al. [10] reported that the Fall Risk Index for patients undergoing rehabilitation after stroke is associated with the risk for

University Rehabilitation Institute, Republic of Slovenia, Linhartova 51, SI-1000 Ljubljana, Slovenia

Downloaded by [University of California, San Diego] at 08:48 17 February 2016



falls, but they suggested that the scale should be modified in order to achieve acceptable predictive accuracy. The Prediction of Falls in Rehabilitation Settings Tool has recently been tested in an acute stroke unit.[11] When this study was designed in 2008, the only clinically useful scale seemed to be the Assessment Sheet for Fall Prediction in Stroke Inpatients (ASFPSI) developed by Nakagawa et al.[12] The ASFPSI and its role in this study are further described in the ‘Methods’ section. We have been striving for a number of years at our Institute to effectively prevent falls of the hospitalised patients. Particular efforts to that end have been made in the stroke rehabilitation ward. The aim of this study was to evaluate the effectiveness of the fall-prevention programme in that ward, which included the application of the ASFPSI to assess the patients’ risk for falls immediately upon admission.

Methods The study included all the patients who were admitted for the first time to the stroke rehabilitation ward of our Institute between 1 September 2010 and 30 September 2011. The study was approved by the Institute’s research ethics committee.

Mini Mental State Examination (MMSE) [16] instead, though with the same cut-off point, because of nearequivalence of the two scales for practical screening purposes [17] and because the MMSE has long been routinely used in our country.[18] The mode of locomotion is scored 2 for using a wheel chair, 1 for walk with walkers and 0 for other. Hence, the total score ranges from 0 to 8. Cut-points had been determined empirically for classifying the score into three fall-risk groups: low (score 0–2), medium (score 3–4) or high (score 5–8).[12] The ASFPSI assessment was performed by the ward physicians (the first three authors), all experienced specialists in physical and rehabilitation medicine (PRM). To assess the patients’ functional status upon admission and upon discharge from inpatient rehabilitation we used the Functional Independence Measure (FIM), which has long been in mandatory routine use at our Institute.[19] Comorbidity was assessed in terms of adjusted severity using the modified Charlson Index (mCI), which has been validated for use in ischaemic stroke outcome studies.[20] The characteristics and consequences of all the falls of the patients during the studied period were recorded as part of the ISO 9001:2008 quality management system of our Institute (Form PMR 014—Record of falls and other injuries). A fall was defined as any accident when the patient involuntarily landed on the floor below knee height.

Data collection and assessment The data on demographic characteristics, type of stroke, side of brain injury, time since stroke and duration of inpatient rehabilitation were collected from medical documentation by the physicians upon admission. After taking the medical history and performing the clinical examination, the risk for falls was assessed using the ASFPSI.[12] Various studies have confirmed that the items of the ASFPSI are valid predictors of falls in stroke inpatients.[6,13–15] High sensitivity and reliability of the scale have been demonstrated.[12] The ASFPSI consists of seven items covering both internal and external factors: central paralysis (which was present in nearly all our patients because of our Institute’s admission criteria), history of previous fall, use of psychotropic medicines, visual impairment, urinary incontinence, cognitive impairment and mode of locomotion. The first five items are scored 1 for presence and 0 for absence. Urinary incontinence is defined as complete inability to control the urethral sphincter muscles (i.e. excluding stress and urge incontinence). The assessment of cognitive impairment is also dichotomised, with presence being defined as the Revised Hasegawa’s Dementia Scale (HDS-R) score of 26. We used the

Fall prevention protocol The following general fall-prevention activities for all admitted patients were introduced in 2009 [21]: improved lightning; demonstration of emergency call device near the bed and in the bathroom; demonstration of bed adjustment mechanism; alerting to the danger of falls and importance of presence of nursing staff while learning to properly and confidently transfer; alerting to slippery floor and the importance of nonslippery footwear; alerting to the danger of using mobile bedside table or unlocked wheelchair as support during transfer; agreement on the voiding plan; and instruction on properly using supports and holders. At the same time, the following individually performed activities were introduced [21]: removal of mobile objects near the bed; collecting information on the patient’s balance and/or coordination disorder; getting information on the patient’s cognitive abilities and memory; informing the patient how psychotropic medication influence state of consciousness; establishing whether the patients is afraid of falling; providing sufficient hydration and nutrition for non-independent patients; verifying adequate size of patient’s clothing; early provision of medical aids to facilitate ambulation;

Downloaded by [University of California, San Diego] at 08:48 17 February 2016


alerting the patient to the danger of losing balance when picking up objects from the floor; and reminding the patient’s family members of carrying our fall prevention activities. When compulsory fall-risk assessment was introduced in the autumn of 2010, the following measures were implemented for the patients belonging to the high-risk group according to the ASFPSI: setting up a discrete high-risk mark that constantly reminded the staff of the patient’s risk; implementing special safety measures as agreed with the physician (bed fences, wheelchair tables, belts preventing uncontrolled standing up); intensified balance training in physiotherapy, emphasis on proper standing up and transfer in occupational therapy and physiotherapy; speech therapist’s assistance during explanations of fall risk to patients with impaired communication; assistance of psychologist when explaining the importance of bed fences and supervised transfer; consultations of the staff with the patient’s family members; accompanying the patient to and from therapy premises; hourly inspections by nursing staff; placing the patient in a room close to the nursing staff room; and alerting the patient’s visitors to the fall prevention measures. Every morning the registered nurse would talk to nurse technicians about the necessary measures for the patients at high risk of fall and report on those patients to the physician. Later the same morning the registered nurse would also present those cases to the other members of the rehabilitation team beside the physician and herself (i.e. the physiotherapist, occupational therapist, psychologist, social worker and speech therapist). The resulting fall prevention protocol is summarised in Figure 1. In addition to the reporting according to the quality management standards, reporting every fall to the rehabilitation team members the next morning was introduced in the second half of 2011, so that the introduction of measures for preventing a repeated fall could be discussed at the team meeting.

Statistical analyses Descriptive statistics (men, median, SD and range for numeric variables, proportions for categorical variables) were calculated first. For comparing the characteristics of the patients who fell and those who did not fall during their rehabilitation, univariate statistical tests (Fisher’s exact test—FET—for dichotomous risk factors, chisquared test of linear association for ASFPSI risk group, exact Mann–Whitney test—EMWT—for numeric risk factors) and bivariate association measures (Spearman’s rho and Somers’ d for pairs of ordinal or numeric variables) were used. We included all the statistically


significant predictors from the univariate analyses and the ASFPSI risk group in the multiple logistic regression model for classifying patients as either fallers or nonfallers. Statistical analyses were conducted using the IBM SPSS Statistics 20 software (IBM Corp., Armonk, NY, 2011). In addition, a retrospective audit of the yearly incidence of falls per 1000 hospital days (HD) in the studied hospital ward from 2006 to 2014 was conducted. Exact confidence interval (CI) for Poisson rate was computed for each yearly incidence of falls. The CI was computed even though we had complete data for our hospital because the observed values can be thought of as estimates for a broader population. The trend in the yearly rate was tested using linear regression of logtransformed rates.[22] Average yearly incidence rate was compared between the pre-intervention and the postintervention period using permutation t-test (which is recommended for analogous single-case research designs [23]).

Results During the study period, 232 patients were admitted for the first time to the stroke rehabilitation ward. Among them, 93 (40.1%) were women and 139 (59.9%) were men; 147 (63.4%) had ischaemic stroke, 53 (22.8%) had brain haemorrhage, 14 (6.0%) subarachnoid haemorrhage and 18 (7.8%) had consequences of other brain disease (mainly brain tumour). The mean age of the patients was 58 years (SD 14, median 59, range 19–91 years). The mean length of hospital stay was 50 days (SD 14, median 53, range 3–88 days) for a total of 11 663 HD. The average time from disease onset to admission to inpatient rehabilitation was 167 days (SD 133, median 138, range 10–955 days). The mean total FIM score was 77 upon admission (SD 26, median 78, range 20–126) and 86 upon discharge (SD 25, median 94, range 20– 126). The mean ASFPSI score was 4.1 (SD 1.3, median 4.0, range 2–7). In total, 35 falls were recorded during the study period. Among the 232 patients, 28 fell; 22 (78.6%) fell once, 5 (17.9%) twice and 1 (3.6%) thrice. The incidence of falls during the study period was 3.0/1000 HD (95% CI 0.6–8.8). Nineteen patients did not suffer any injury during the fall, one patient suffered a split wound on the head (which was successfully treated at the ward) and eight patients suffered bruises or minor abrasions. The mean time from admission to fall was 21 days (SD 20, median 16, range 1–80 days). Within the first week from admission, 11 patients fell (39% of the 28 who fell) and six patients (21%) fell within the first 3 days.

Downloaded by [University of California, San Diego] at 08:48 17 February 2016



Figure 1. Summary of the fall prevention protocol including compulsory fall risk assessment introduced in the autumn of 2010.

Table 1 lists the characteristics of the patients who fell and those who did not fall during the study period. There was no notable association between falling and either the ASFPSI score (p ¼ 0.445 from EMWT) or ASFPSI risk group (p ¼ 0.674 from FET). The lack of association between ASFPSI scores and falls was confirmed with additional analyses: chi-squared test of linear association showed no statistically significant trend in the proportion of falls with increasing ASFPSI risk group (p ¼ 0.255); there was no statistically significant ordinal correlation between the ASFPSI score and the number of falls (rho ¼ 0.049, p ¼ 0.457); and there was no statistically significant ordinal correlation between the ASFPSI risk group and the number of falls (d ¼ 0.058, p ¼ 0.289). We hypothesised that either a very low FIM score (indicating a high degree of functional dependence) or a very high FIM score (indicating a high degree of functional independence) would be associated with lower fall risk with the respect to non-extreme FIM score, so we dichotomised the FIM score based on the observed interquartile range (Q1¼55, Q3¼101). The univariate analysis (Table 1) confirmed that conjecture (p ¼ 0.008 from FET) and estimated the odds of falling to be about 3.6 times higher for the patients whose FIM score at admission was in the central 50% of the distribution. Falls were also found to be statistically

significantly associated with the left body-side being affected as compared with the right side or both body sides (p ¼ 0.042 from FET, estimated odds ratio 2.3). As expected, falls were also statistically significantly associated with the length of hospital stay (p ¼ 0.046 from EMWT). There was no statistically significant association of falls with the patient’s gender (p ¼ 1.000 from FET), stroke type (p ¼ 0.532 from FET), psychotropic medication (p ¼ 0.835 from FET), modified Charlton Index (p ¼ 0.169 from FET), age (p ¼ 0.580 from EMWT) or time from disease onset (p ¼ 0.747 from EMWT). Multiple logistic regression confirmed the abovementioned findings (Table 2). The model was statistically significantly better than the null model (likelihood ratio test: p ¼ 0.001) and it did not statistically significantly misfit the data (Hosmer–Lemeshow test: p ¼ 0.681). Nagelkerke pseudo-R2 value for the model was 0.16. The estimated adjusted odds ratio for falling was about 2.8 for left body-side being affected vs. right or both body-sides, and about 2.9 for FIM score at admission between 55 and 101 vs. below 55 or above 101. The retrospective audit (Figure 2) revealed that the yearly incidence of falls decreased from 7.1/1000 HD (95% CI 3.5–14.4) in 2006 (29) through 4.2/1000 HD (95% CI 1.6–10.2) in 2009 [24] to 2.8/1000 HD (95% CI 0.6–7.2) in 2011, and remained stable until 2014.[25–27] Hence,



Table 1. Characteristics of those who fell and those who did not fall among the patients admitted to inpatient rehabilitation after stroke for the first time—univariate statistical comparisons.

Downloaded by [University of California, San Diego] at 08:48 17 February 2016

Characteristic Gender Male Female Stroke type Ischaemic Other Affected body-side Right or both Left Psychotropic drug No Yes mCI 0–1 2 ASFPSI Median (IDR) ASFPSI risk group low or medium High Age (years) Median (IDR) Length of stay (days) Median (IDR) Time from disease onset (days) Median (IDR) FIM at admission 54 or 102 55–101

Non-fallers (n ¼ 204)

Fallers (n ¼ 28)


OR (95% CI)

122 (87.8%) 82 (88.2%)

17 (12.2%) 11 (11.8%)


0.96 (0.43–2.16)

131 (89.1%) 73 (85.9%)

16 (10.9%) 12 (14.1%)


1.35 (0.60–3.00)

123 (91.8%) 81 (82.7%)

11 (8.2%) 17 (17.3%)


2.35 (1.05–5.27)

130 (88.4%) 74 (87.1%)

17 (11.6%) 11 (12.9%)


1.14 (0.51–2.56)

186 (91.2%) 18 (8.9%)

23 (82.2%) 5 (17.8%)


2.25 (0.76–6.62)

4.0 (2.0–6.0)

4.0 (2.9–6.0)


1.11 (0.85–1.45)

134 (88.7%) 70 (86.4%)

17 (11.3%) 11 (13.6%)



1.24 (0.55–2.79)

59.0 (39.0–76.0)

61.0 (38.0–76.2)


1.0.1 (0.98–1.04)

53.0 (25.0–67.0)

53.0 (32.0–86.2)


1.03 (1.01–1.06)

135.0 (37.5–352.0)

145.0 (36.9–327.0)


1.00 (0.99–1.01)

101 (94.4%) 103 (82.4%)

6 (5.6%) 22 (17.6%)



3.60 (1.40–9.24)

OR—Unadjusted odds ratio for falling (categorical variables: bottom vs. top category; numeric variables: per unit increase); CI—confidence interval; mCI— modified Charlson Index; ASFPSI—Assessment Sheet for Fall Prediction in Stroke Inpatients; FIM—Functional Independence Measure; FET—Fisher’s exact test; EMWT—exact Mann–Whitney test; IDR—interdecile range.

the reduction in the incidence of falls was by about twothirds over the 7-year period (2006–2012), during which systematic fall prevention was introduced in the fourth year (i.e. at the beginning of 2009) and compulsory fall risk assessment was introduced at the end of the fifth year (i.e. in the autumn of 2010). The decreasing trend was consistent and statistically significant (p50.001 for testing the null hypothesis of zero slope of the logtransformed yearly rates). Another simple statistical demonstration of the programme’s effectiveness is that the average yearly incidence rate of falls fell from 6.7 in the pre-intervention period (2006–2008) to 2.8 in the post-intervention period (2011–2014; p ¼ 0.029 from permutation t-test).

Discussion Fall prevention is very important in the rehabilitation environment. The consequences of a fall for the patient can be severe, including traumatic injuries and reduced functional ability, fear from falling again and therefore reduced activity.[13] These consequences can negatively affect the rehabilitation process and the rehabilitation outcomes, hospitalisation can be prolonged and the costs of care can soar.[28] Randomised controlled studies have shown that introducing consorted activities

for preventing falls in rehabilitation wards can reduce the number of falls by 30–60%.[29–31] In view of these facts, our study aimed at evaluating the use of the ASFPSI upon the stroke patient’s admission to the rehabilitation ward. Fall risk should namely be performed within the first 24 h upon admission to inpatient rehabilitation.[28] The observed incidence of falls in our stroke rehabilitation ward during the study period (3 falls per 1000 HD) was relatively small compared with other rehabilitation hospitals treating stroke patients in the subacute phase, where it was reported to be between 5 and 18 falls per 1000 HD.[4,32,33] Moreover, none of our patients sustained a severe injury such as fracture or haemorrhage during the study period. We believe that the key to our successful reduction of the incidence of falls and subsequent successful maintenance of the low falls rate lie in the team approach, i.e. the active cooperation of nurses, physicians, physiotherapists, occupational therapists, psychologists, social workers and speech therapists. Although the mean time from admission to fall was 3 weeks (and the median time was over 2 weeks), about two-fifths of the falls occurred within the first week after admission, and one-fifth during the first 3 days. We observed no association of the ASFPSI score (either raw



Table 2. Summary of the multiple logistic regression model for predicting whether a patient fell during first inpatient rehabilitation after stroke. Predictor Affected body side (left vs. right or both) ASFPSI risk group (high vs. low or medium) Length of stay (days) FIM at admission (55–101 vs. 54 or 102)

b (SE) 1.032 0.806 0.034 1.071


(0.439) (0.778) (0.014) (0.498)

0.019 0.301 0.020 0.031

OR (95% CI) 2.81 2.24 1.03 2.92

(1.19–6.62) (0.49–10.29) (1.01–1.06) (1.10–7.74)

Downloaded by [University of California, San Diego] at 08:48 17 February 2016

OR—estimated adjusted odds ratio for falling; CI—confidence interval; ASFPSI—Assessment Sheet for Fall Prediction in Stroke Inpatients; FIM—Functional Independence Measure.

Figure 2. Yearly incidence of falls from 2006 to 2014 in the studied stroke rehabilitation ward. Detailed data were studied for 2010 and 2011; systematic fall prevention was introduced at the beginning of the year 2009; compulsory fall risk assessment was introduced in the autumn of 2010.

or categorised into risk groups) with falling (either dichotomised or analysed as the number of falls of a patient). It is therefore plausible that the process of fall risk assessment and classification of patients into fall risk groups upon admission itself represented a key step towards fall prevention. This finding of the prospective detailed study is corroborated by the retrospective audit because the largest reduction in the yearly incidence rate of falls was observed over the years 2009 and 2010, when systematic fall prevention was first implemented and compulsory fall risk assessment subsequently introduced. Other studies have found that staff attention is a better potential predictor of falls than assessment scale scores [28,34,35], but the design of our study does not permit a judgement on this issue (as further discussed below). Assessing fall risk is particularly important in patients aged over 65 and in patients with cognitive deficits.[28] In our study, we observed an increased risk for falls among the patients with non-extreme FIM score, i.e. neither high functional dependence nor high functional independence. This is understandable because the patients with very low FIM scores virtually

cannot—and in the vast majority of cases therefore do not—perform activities of daily living without assistance of the staff, while those with very high FIM scores exercise good motor and cognitive control. Falls were also found to be associated with left body-side being affected, which is more often associated with difficulties in dealing with space and unilateral neglect.[36,37] A recent review highlighted a relationship between falls in patients with stroke and impaired balance, visuospatial hemineglect and impaired performance of the activities of daily living.[38] The roles of cognitive function, incontinence, visual field deficits and stroke type appeared to be less certain, and age, gender, stroke location and impaired vison or hearing were not identified as related to fall risk.[38] The latest retrospective study of falls in post-stroke inpatients defined seven different types of falls, whereby six of the fall types were related to the high dependency needs of poststroke patients, visuospatial difficulties, or delirium, and one fall group was associated with higher functioning at the time of falls and better clinical outcome.[39] The association of falls with the length of hospital stay was expected. It can be primarily attributed to longer

Downloaded by [University of California, San Diego] at 08:48 17 February 2016


exposure to the risk of falling, but it could also be related to wider extent of impairment and/or fear of movement. We did not observe any notable association of falls with gender, stroke type, psychotropic medication, comorbidity (as assessed by the modified Charlton Index), age or time from disease onset. Fall risk in the subacute stroke rehabilitation setting therefore seems to be related to clinical complexity and to a moderate level of disability, which is in line with previous epidemiological evidence.[1,15] Even though the number of studies demonstrating the advantages of fall prevention is rising, the implementation of preventive measures is still relatively limited.[28,40] Published guidelines on what exactly the measures should be and how they should be introduced into clinical practise are hard to find. Furthermore, different fall-prevention strategies may be required for different patient subgroups.[39] In our case, following the analysis of falls in the stroke rehabilitation ward in 2006,[41] monitoring of patients under elevated risk of falling was increased especially in the afternoon, especially during changing the seat and walking, and medication was prescribed more carefully in order to avoid sedation, disorientation and postural hypotension. By paying more attention to balance training, teaching the patients to move safely within the limits of their abilities, and to checking the patients’ comprehension of the instructions, the patients’ awareness of their abilities was improved and they were encouraged to call for help when needed. The bulk of the efforts were directed towards the nursing staff, but—as already emphasised— the team approach was essential. Instituting the fall prevention programme provided the nurses with a vehicle for discussing fall risk with the team, and the team approach was also applied to informing the family members. The efficiency of the initial measures soon became observable but the key reduction in the yearly incidence of falls occurred after systematic fall risk prevention had been introduced in 2009. In addition to the reduction in the incidence rate of falls, the proportion of patients who fall more than once during inpatient rehabilitation has also been reduced as nearly a half of the fallers fell more than once in 2006 [41] whereas only a fifth did so during our detailedstudy period (i.e. from September 2010 to September 2011).

Limitations The first potential limitation to our study is that the reliability and validity of the ASFPSI scale had not been tested in our institute before the implementation of the assessment. It could be argued that validation was


necessary despite the fact that the original study was also performed on stroke rehabilitation inpatients [12] because our population is culturally different, and despite the fact that the scale’s items refer to basic and culturally independent aspects of functioning. The second limitation is that inter-rater reliability was not formally assessed. We doubt that disagreement could be expected among highly trained and experienced PRM specialists regarding the ASFPSI items, and informal agreement checks never indicated a case of disagreement, but we intend to conduct an inter-rater agreement study in the future. In addition, our results cannot rule out the explanation that the lack of difference in fall incidence between the patients identified as under high risk (and thus receiving additional fall prevention measures) and those classified as under low or medium risk is due to the scale not being suitable for our population, even though such explanation seem highly implausible. As already stressed, at the time when our study was designed the ASFPSI was the instrument of choice for fall-risk assessment in post-stroke rehabilitation inpatients. The recently validated Stroke Assessment of Fall Risk assesses seven stroke-specific risk factors identified from the published literature and clinical audits, which comprise four impairments (impulsivity, hemi-neglect, static and dynamic sitting balance) and three functional limitations (lowest score on three FIM items: transfers, problem solving and memory).[42] The majority of those factors are incorporated in the ASFPSI, so the choice of the scale does not constitute a major limitation of our study, though the more recent scale should be preferred in future studies and for future fall-risk prevention programmes. Regarding the audit, a potential limitation is that the study design does not permit a firm judgement on the effect of the assessment-based measures alone. It seems obvious from Figure 2 that the introduction of systematic fall prevention explains the observed reduction of falls incidence rate from46.5 per 1000 HD to54.0, and the introduction of assessment-based additional measures explains further reduction to the stable level of 2.8. However, these two effects cannot be discerned in the absence of a control group where the programme and/or only the assessment were absent (but that would have been neither ethically acceptable nor practically possible), or a comparable hospital without the programme and/or the assessment (but none exists because our Institute is the only tertiary rehabilitation hospital in the country). Furthermore, in our view, separating the two is an academic dilemma rather than a practical clinical issue because the combined effect is certain, large and lasting.



Downloaded by [University of California, San Diego] at 08:48 17 February 2016

Conclusion Recognising fall risk upon the patient’s admission plays an important part in the efforts for preventing falls in rehabilitation wards. Assessing the fall risk is a team tasks and combines information provided by the patients and their relatives, medical examination, observation and assessment by the nursing staff and information provided by other rehabilitation team members in addition to the formal assessment sheet. The results of our study indicate that assessing fall risk in subacute stroke rehabilitation inpatients using the assessment sheet by Nakagawa et al. immediately upon admission is sensible primarily because of systematically drawing attention to the risk of falls in each individual patient. Another recommendation arising from the results is that patients with non-extreme level of functional independence should receive enhanced fall prevention. The comprehensive fall prevention programme including general measures for all patients and additional measures for patients identified as being at high risk was successful because the incidence of falls in our rehabilitation word for stroke patients dropped by one-third. The key to the success of our programme most probably lies in the team approach, i.e. coordinated efforts of nurses, physicians, physiotherapists, occupational therapists, psychologists, social workers and speech therapists.

Declaration of interest The authors report no conflict of interest. They received no funding for the study.

References 1. Nyberg L, Gustafson Y. Patient falls in stroke rehabilitation. A challenge to rehabilitation strategies. Stroke. 1995;26: 838–842. 2. Sze KH, Wong E, Leung HY, et al. Falls among Chinese stroke patients during rehabilitation. Arch Phys Med Rehabil. 2001;82:1219–1225. 3. Teasell R, McRea M, Foley N, et al. The incidence and consequences of falls in stroke patients during inpatient rehabilitation: factors associated with high risk. Arch Phys Med Rehabil. 2002;83:329–333. 4. Vassalo M, Sharma JC, Briggs RS, et al. Characteristics of early fallers on elderly patient rehabilitation wards. . Age Ageing. 2003;32:338–342. 5. Eriksson S, Gustafson Y, Ludlin-Olsson L. Characteristics associated with falls in patients with dementia in a psychogeriatric ward. Aging Clin Exp Res. 2007;19:97–103. 6. Perell KL, Nelson A, Goldman RL, et al. Fall risk assessment measures: an analytic review. J Gerontol Med Sci. 2001;56A:761–766. 7. Morse JM. Enhancing the safety of hospitalization by reducing patient falls. Am J Infect Control. 2002;30: 376–380.

8. Nyberg L, Gustafson Y. Using the Downton Index to predict those prone to falls in stroke rehabilitation. Stroke. 1996;27:1821–1824. 9. Rapport LJ, Webster JS, Flemming KL, et al. Predictors of falls among right-hemisphere stroke patients in the rehabilitation setting. Arch Phys Med Rehabil. 1993;74:621–626. 10. Olsson E, Lofgren B, Gustafson Y, et al. Validation of a fall risk index in stroke rehabilitation. J Stroke Cerebrovasc Dis. 2005;14:23–28. 11. Nystro¨m A, Hellstro¨m K. Fall risk six weeks from onset of stroke and the ability of the Prediction of Falls in Rehabilitation Settings Tool and motor function to predict falls. Clin Rehabil. 2013;27:473–479. 12. Nakagawa Y, Sannomiya K, Kinoshita M, et al. Development of an assessment sheet for fall prediction in stroke inpatients in convalescent rehabilitation wards in Japan. Environ Health Prev Med. 2008;13:138–147. 13. Mayo NE, Korner-Bitensky N, Levy AR. Risk factors for fractures due to falls. Arch Phys Med Rehabil. 1993;74:917–921. 14. Nyberg L, Gustafson Y. Fall prediction index for patients in stroke rehabilitation. Stroke. 1997;28:716–721. 15. Saverino A, Benevolo E, Ottonello M, et al. Falls in a rehabilitation setting: functional independence and fall risk. Eura Medicophys. 2006;42:179–184. 16. Folstein MF, Folstein SE, McHugh PR. Mini-mental state. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975;12:189–198. 17. Kim KW, Lee DY, JHoo JH, et al. Diagnostic accuracy of mini-mental status examination and revised Hasegawa dementia scale for Alzheimer’s disease. Dement Geriatr Cogn Disord. 2005;19:324–330. 18. Vodusˇek DB. Higher nervous activity. Med Razgl. 1992;31:369–400. 19. Vidmar G, Burger H, Marincˇek Cˇ, et al. Analysis of data on assessment with the Functional Independent Measure at the Institute for Rehabilitation, Republic of Slovenia. Inf Med Slov. 2008;13:21–32. 20. Goldstein LB, Samsa GP, Matchar DB, et al. Charlson index comorbidity adjustment for ischemic stroke outcome studies. Stroke. 2004;35:1941–1945. 21. Kopitar N, Goljar N, Mlinaricˇ Lesˇnik V. Fall prevention activities for inpatients at stroke rehabilitation setting and their impact on the incidence of falls. Obzor Zdrav Neg. 2014;48:206–214. 22. Rosenberg D. Methods for analyzing trend data. In: Handler A, Rosenberg D, Monahan C, et al., editors. Analytic methods in maternal and child health. Washington: Division of Science, Education, and Analysis, Health Resources and Services Administration, Maternal and Child Health Bureau; 1998, p. 191–224. 23. Dugard P, File P, Todman J. Single-case and small-n experimental designs: a practical guide to randomization tests, 2nd ed. New York: Routlege; 2011. 24. Univerzitetni rehabilitacijski insˇtitut Republike Slovenije— Socˇa. Strokovno porocˇilo za leto 2011 za JZZ [Annual report for 2011 for public health institutions], p. 20. www.ir-rs.si/f/docs/Informacija_javnega_znacaja/ StrokovnoPorocilo_URISoca_2011.pdf. 25. Univerzitetni rehabilitacijski insˇtitut Republike Slovenije— Socˇa. Strokovno porocˇilo za leto 2012 za JZZ [Annual report for 2012 for public health institutions], p. 24.





Downloaded by [University of California, San Diego] at 08:48 17 February 2016





www.ir-rs.si/f/docs/Informacija_javnega_znacaja/Strokovn oPorocilo_URISoca_2012.pdf. Univerzitetni rehabilitacijski insˇtitut Republike Slovenije— Socˇa. Strokovno porocˇilo za leto 2013 za JZZ [Annual report for 2013 for public health institutions], p. 24. www.ir-rs.si/f/docs/Informacija_javnega_znacaja/Strokovn oPorocilo_URISoca_2013.pdf. Univerzitetni rehabilitacijski insˇtitut Republike Slovenije— Socˇa. Strokovno porocˇilo za leto 2014 za JZZ [Annual report for 2014 for public health institutions], p. 24. www.ir-rs.si/f/docs/Informacija_javnega_znacaja/ StrokovnoPorocilo_URISoca_2014.pdf. Aberg AC, Lundin-Olsson L, Rosendahl E. Implementation of evidence-based prevention of falls in rehabilitation units: a staff’s interactive approach. J Rehabil Med. 2009; 41:1034–1040. Haines TP, Bennell KL, Osborne RH, et al. Effectiveness of targeted falls prevention programme in subacute hospital setting: randomised controlled trial. BMJ. 2004;328:336. Healy F, Monro A, Cockram A, et al. Using targeted risk factor reduction to prevent falls in older in-patients: a randomised controlled trial. Age Ageing. 2004;33:390–395. Stanvall M, Olofsson B, Lundstrom M, et al. A multidisciplinary, multifactorial intervention program reduces postoperative falls and injuries after femoral neck fracture. Osteoporosis Int. 2007;18:167–175. Mayo NE, Korner Bitensky N, Becker R, et al. Predicting falls among patients in a rehabilitation hospital. Am J Phys Med Rehabil. 1998;68:139–146.


33. Czernuszensko A, Czlonkowska A. Risk factors for falls in stroke patients during inpatient rehabilitation. Clin Rehabil. 2009;23:176–188. 34. Oliver D. Falls risk prediction tools for hospital inpatients. Time to Put Them to Bed? Age Agening. 2008;37:248–250. 35. Meyer G, Kopke S, Haastert B, et al. Comparison of a fall risk assessment tool with nurses’ judgement alone: a cluster-randomised controlled trial. Age Ageing. 2009;38:417–423. 36. Stone SP, Wilson B, Wroot A, et al. The assessment of visuo-spatial neglect after acute stroke. J Neuro Neurosurg Psychiatry. 1991;54:345–350. 37. Vallar G. Spatial hemineglect in humans. Trends Cogn Sci. 1998;2:87–95. 38. Campbell GB, Matthews JT. An integrative review of factors associated with falls during post-stroke rehabilitation. J Nurs Scholarsh. 2010;42:395–404. 39. Hanger HC, Wills KL, Wilkinson T. Classification of falls in stroke rehabilitation-not all falls are the same. . Clin Rehabil. 2014;28:183–195. 40. Cambpell AJ, Robertson MC. Implementation of multifactorial interventions for fall and fracture prevention. Age Ageing. 2006;35:ii60–ii64. 41. Marolt M, Goljar N. Falls in stroke patients during inpatient rehabilitation. Rehabilitation (Ljubljana). 2008;7:17–22. 42. Breisinger TP, Skidmore ER, Niyonkuru C, et al. The stroke assessment of fall risk (SAFR): predicitve validity in inpatient stroke rehabilitation. Clin Rehabil. 2014;28: 1218–1224.

Effectiveness of a fall-risk reduction programme for inpatient rehabilitation after stroke.

To evaluate effectiveness of fall-risk-assessment-based fall prevention for stroke rehabilitation inpatients...
566B Sizes 0 Downloads 4 Views