Development and alarm threshold evaluation of a side rail integrated sensor technology for the preve

i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 79(2010)

173–180

j o u r n a l h o m e p a g e :w w w.i n t l.e l s e v i e r h e a l t h.c o m /j o u r n a l s /i j m

i

Development and alarm threshold evaluation of a side rail integrated sensor technology for the prevention of falls

Johannes Hilbe ?,Eva Schulc,Barbara Linder,Christa Them

Department of Nursing Science,Austria

a r t i c l e

i n f o Article history:

Received 16April 2009Received in revised form 30November 2009

Accepted 18December 2009

Keywords:Accidental falls Safety Nursing Beds Emergencies Physical restraint Dementia

a b s t r a c t

Objective:Patient falls constitute a serious problem both for the persons fallen and for the institutions involved.Bed-exit alarm systems are used to reduce patient falls.Existing bed-exit alarm systems have several disadvantages depending on the technology used.As in “Evaluation of Bed-Exit Alarms”stated,restless,light weighted,uncooperative,incontinent and confused patients require different systems.

The aim of this work is to present the research and development process of the integrated,universally applicable BUCINATOR bed-exit-alarm system.

Methods:The use-case technique was applied to capture the functional requirements for the development of the new integrated bed-exit alarm system.An experimental study was carried out to collect data regarding preliminary sensitivity and speci?city for alarm set-off.Results:Major identi?ed requirements for an optimized bed-exit alarm system were usability,wide range usage,low costs,hygiene factors,integration into nursing beds and nurse call systems and an adequate alarm/false alarm ratio with early alarm trigger functionality.On the basis of the criteria mentioned above,a sensor system was developed,comprising tubes with an air-?lled passageway attached on the top of side rails.These tubes are coupled via lines to transducers which trigger an alarm when a predetermined level of pressure is reached.Both the preliminary sensitivity (96.0%)and the speci?city (≥95.5%)of the trigger level indicate a satisfactory alarm/false alarm ratio which is now to be evaluated in a clinical trial.

Conclusions:After experimental testing,BUCINATOR shows great potential to be a reliable bed-exit alarm system.In general,bed-exit alarm systems with extended features could play a major role in ambient assisted living technologies.

Limitations:Besides the theoretical evaluation,it will be imperative to perform more tests and to gather more data about the effect on fall rates and resulting injuries.

?2010Elsevier Ireland Ltd.All rights reserved.

1.Introduction

Electronic sensing devices for the prevention of bed-related patient falls are becoming more and more common in health care institutions.These systems are designed to detect

?

Corresponding author at :UMIT -Private University of Health Sciences,Medical Informatics &Technology,Eduard Walln?fer-Zentrum 1,G3Room 055,6060Hall in Tirol,Austria.Tel.:+4305086483862;fax:+4305086483882.

E-mail address:johannes.hilbe@umit.at (J.Hilbe).

patients leaving the bed without assistance and thus are able to avoid dangerous situations.Moore and VanGilder [1]showed in their report that patient falls can be reduced by using intelligent technologies.Besides falls,Jackson and Glea-son [2]showed that with a bed-monitoring system in acute care units,the usage of restraints has decreased by 99%and

1386-5056/$–see front matter ?2010Elsevier Ireland Ltd.All rights reserved.doi:10.1016/j.ijmedinf.2009.12.004

174i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s79(2010)173–180

fewer observers were needed.There are various technolog-ical systems with miscellaneous sensing methods in use, resulting in different advantages and disadvantages.These bed-exit alarm systems can be seen as a substitute for physical restraint methods.

Wilson[3]said that falls are among the most common, yet potentially preventable adverse events experienced by patients in hospitals.Impaired cognitive function and a poor sense of orientation can lead to an increase in falls among those with impaired freedom of movement[4].Results from falling are such serious outcomes as physical and emotional injury,increased dependence,admission to a long-term care facility and poor quality of life.

Fall incidence in acute care hospitals varies between1.1 and9.1/1000patient days[5,6].Approximately half of the nurs-ing home residents fall annually[7].The variability of the ?gures results from different types of institutions and patient populations(presence of fall risk factors)as well as the imple-mentation of fall prevention programs.Therefore,as there are too many unknown factors involved,if benchmarking is used at all,it should be carried out cautiously.Dementia for example is an independent risk factor for falling[8].People with dementia living in nursing homes fall twice as often,as persons with normal cognitive capability.Wandering affects 39%of cognitively impaired nursing home residents and up to70%of community-residing elderly persons with cognitive impairments[9].Therefore,this group of people is in particular affected and could considerably bene?t from bed-exit alarm systems.

Nurmi and Lüthje[10]calculated in the year2002the aver-age cost of a fall with944D.Most falls occur in the patient room(60–85%)[11–14]and could be potentially prevented by bed-exit alarm systems.

There are different approaches to prevent falls.Restraints for example are mainly used for the prevention of falls[15]. Evans[16]highlighted in his2002review the potential dan-ger of using physical restraint in acute and residential health care facilities.The predominant justi?cation for using these various restraint methods is maintaining patient safety.Man-aging agitation and aggression,exercising behavioural control, preventing patients from wandering and providing physical support are reasons to apply side rails or other restraint devices.However,while restraint may reduce some hazards, it may also involve other risks.Evans et al.[17]identi?ed increased risk of injury and death to be associated with restraint use.One of the risks mentioned most often is bed-rail entrapment.In1995,the FDA issued a Safety Alert entitled “Entrapment Hazards with Hospital Bed Side Rails”[18]based on102reports of head and body entrapment incidents involv-ing hospital bed side rails(68deaths reported,1990–1995).

This bulletin started a series of research projects deal-ing with the need to reduce or eliminate the use of physical restraints for older people in long-term and acute care set-tings.One part of these projects dealt with new technologies to prevent physical restraints.Nelson et al.[9]stated in2004that technology offers the potential to eliminate or mitigate pre-ventable adverse events that interfere with treatment,delay of rehabilitation,potential impairment,and compromise patient safety.To prevent bed-rail entrapment and falls,new hospital bed designs like height-adjustable low beds,devices to close gaps in legacy beds,and bed side?oor mats were brought on the health care market.

Managing the symptoms of cognitive impairment by addressing agitation,wandering and impulsive behaviour is therefore necessary.But despite initiatives among health care providers to encourage patient autonomy,the use of physical restraints on frail or confused older patients continues to be a common practice in many health care settings[19].It is evi-dent that legislation has a clear impact on the use of restraints in clinical practice[15].Changes in education and legislation promoted the slow acceptance of bed-exit alarm systems in Austria.

Unfortunately,already existing bed-exit alarm systems (e.g.pressure-sensitive?oor mats,pressure-sensitive strips placed on or under mattresses,garment clips,infrared beam detectors)have some striking disadvantages[20].Pressure-sensitive?oor mats are placed alongside the bed.They trigger an alarm when a person applies pressure on the mat,which is too late in case of a bed fall.Pressure-sensitive strips placed on or under the mattress work inverse to the mechanism of the?oor mats.They trigger an alarm if they recognize an absence of weight.Problems occurred with under mattress strips when light-weighted patients in combination with solid mattresses overcharged the sensor system.Garment clips con-sist of adjustable-length cords that connect to a clip attached to the patient’s clothing.Restless patients produced false alarms due to wrapping and pulling at the cord.Other sensor methods are infrared beam detectors applied on both sides of the bed.They also produced false alarms when patients e.g. held their hand into the light beam[21].

Besides the disadvantages mentioned above,there are other problems associated with existing technologies.They are no integral parts of beds or mattresses and in case of need;

a bed-exit alarm system must be taken from the storage room and placed at or in the bed.Furthermore,these systems have to be checked for functionality and correct position from time to time.Integrated bed-exit alarm systems can be found in technologically advanced and therefore more expensive hos-pital beds which are not affordable for full-scale deployment in the geriatric and home care setting.

Beardsley stated in his work that every patient,regardless of the care setting,deserves a safe and comfortable sleeping and bed environment[22].Bed-exit alarm systems are only one step towards coming closer to the statement mentioned above.

The aim of this paper is to show the development process of the BUCINATOR bed-exit alarm system(BUCINATOR:lat. horn blower,Guards in the Roman Empire,who warned about approaching danger situations),which is applied directly at the side rails of nursing beds.

2.Methods

Nelson emphasised that it is critical to involve front-line nurs-ing staff in the testing and in the selection of devices that will be used in their practice[9].Therefore,the research process was structured in a way that allowed nursing staff to partici-pate.In general,the methods can be divided into four phases. Apart from the literature review(Phase I),the researcher

i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s79(2010)173–180175

team invited nursing staff(six registered nurses and six nursing assistants of a gerontological-psychiatric ward)to open interviews in order to gather information regarding the requirements and expectations of a bed-exit alarm system. After12interviews,theoretical saturation was reached and rule-guided qualitative text analysis according to Mayring was conducted[23].The qualitative information available and the results of the literature review were synthesized with the use-case technique and helped to clarify core requirements(Phase II).Thereupon,several prototypes were built until the before-hand determined requirements were reached(Phase III).An experimental design was used to set the alarm threshold (Phase IV).In a?rst step(stage A),the sensor tubes were loaded with different weights(1.2–12kg)in different lengths (5–70cm)and the peak pressures resulting from the loads were measured(manometer:PCE-P05)and registered.These tests were repeated three times and the mean values were calculated.Matlab was used to interpolate the data on a three-dimensional graph.A threshold was established and reviewed in a further experiment.Therefore,in stage B49nursing students and13registered nurses(body weight:R=48–98kg,ˉx=66.4kg,s=12.2kg;age:R=18–60years)were introduced to behave like old persons with motor and cognitive de?cits.The peak pressures resulting from the load on the bed-exit alarm system were measured.A Receiver Operating Characteristic curve(ROC curve)was carried out,ascertaining the best pos-sible alarm threshold(SPSS,Version16.0).

3.Results

Based on the conclusions drawn after a comprehensive liter-ature research(Phase I)and open interviews with registered nurses(Phase II),the side rails of hospital beds were identi-?ed to be the most appropriate place for the collocation of the sensor system which is to be developed.The requirement pro?le which was de?ned on the basis of the Use-Case Method comprises the following factors:

(a)usability/wide range usage;

(b)hygiene factors;

(c)Low cost;

(d)integration into nursing beds;

(e)integration into nurse call systems;

(f)early alarm generation;

(g)excellent alarm/false alarm ratio.

In the?rst development phase(Phase III),industrial sen-sor systems at the side rail were used for experiments.These systems,however,are designed to detect preferably any over-stepping of this zone.Elevators doors,for example,must open as soon as a?nger is located between the two doors and indus-trial machines must switch off as soon as a person is entering the hazard area.For the development of a bed-exit alarm sys-tem,however,researchers had to take into account that false alarms triggered by daily movements which do not necessarily result from abandoning the bed,should be avoided.In contrast to the mechanism of the industrial sensor systems mentioned above,the hand at the edge of the bed or the shaking of the side rail,for instance,should not trigger an alarm for itself.There-fore,it was necessary to develop a new system which allows for a more extended tolerance until an alarm is triggered.

With regard to(a)usability and wide range usage,a univer-sally applicable system was developed which does not require any adjustments due to different body weights or any other individual characteristics of patients.The(b)hygiene factors were also taken into account,selecting adequate material and a?at,moisture-repellent surface with few edges.In order to keep the costs low,great attention was paid to the selection of the components required.Therefore,only industrial standard components were used and it was tried to use as few parts as possible.Thus,the developers succeeded in designing a long-term sturdy and economic system,which was the objective.

(d)Integration into nursing beds means that the system is part of the equipment(side rail).It can be activated if needed without any adjustment work which distinguishes this system from the majority of the existing system.(e)Integration into nurse call system is a factor which is closely connected with usability as all alarms are registered by a central data processing sys-tem which allows for the statistical evaluation of these data afterwards.Furthermore,the integration in existing nurse call systems extends the mobility of the nursing staff compared to a local acoustic or visual alarm.The requirement of an(f)early alarm generation system has been met through the triggering of an alarm already during the abandonment of the bed;thus it is possible,for instance,to?nd a person at risk of falling which is sitting on the bed laterally.

3.1.Illustration of the sensor system

As illustrated in Fig.1,pressure-sensitive rails(purple)?lled with air under atmospheric pressure in the cylindrical cav-ity(radius:3mm)are mounted at the upper edge of the side rail for the BUCINATOR bed-exit alarm system.Material used:Ethylen-Propylen-Terpolymers,EPDM.These rails are connected with the pressure sensor in the electronical unit via pressure-tight cables(green).This construction ensures that there are no electrical tensions in the proximity of the patient.If a patient at risk of falling intends to leave the bed with the side rail in an upright or lowered-down position,the pressure increases in the system through the load and if a cer-tain threshold level is exceeded,an electrical alarm signal is generated.The nursing staff will be informed about the risk situation through the ward’s nurse call system.

An activity diagram in Uni?ed Modelling Language(UML) demonstrates the process of the alarm triggering in Fig.2.If a signal is triggered through an excessive pressure level,this signal can be further processed by different technologies.

3.2.Pressure threshold testing

The(g)alarm/false alarm ratio was tested in two experiments (Phase IV,Stages A and B).

The results of Stage A are demonstrated in Fig.3.The three-dimensional graph clearly shows that it was not possible,not even with the maximum weight used in the experiment,to produce a pressure level of≥35mbar on a little surface(length:≤10cm).

Based on this result,a second experiment(Stage B)with nursing staff was carried out.The patterns of conduct could

176

i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 79(2010)

173–180

Fig.1–Schematic illustration of the BUCINATOR bed-exit alarm system.An alarm can be triggered both in an upright and lowered-down position of the side rail (Source :https://www.sodocs.net/doc/264009353.html, ;altered by Johannes

Hilbe).

Fig.2–UML activity

diagram.

Fig.3–Response characteristics of the BUCINATOR sensor system.

be individually chosen by the test persons.The only rule was to behave within the bounds of possibilities of elderly peo-ple.The scienti?c team divided these patterns of conduct into seven different categories.They recorded the registered pres-sure values of the different scenarios of getting out of the bed on the one hand and the pressure values triggered through touching the sensor system without having the intention to leave the bed on the other hand.In Fig.4,the values reached through the contact with the sensor system (black points)are allocated to the different categories.

i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s79(2010)173–180

177

Fig.4–Results for the optimization of the alarm trigger threshold.

Category I+II:These two categories comprise getting out of the bed with the bedrail either in an upright position(category I)or in a lowered-down position(category II).

Category III:This“having two legs on the bedrail”category, which means sitting on the bed laterally with the bedrail being in an upright position,comprises cases of different qualities which cannot be separated from each other due to the exper-iment design.On the one hand,the test person just put their legs casually on the side rail,on the other hand the sensor sys-tem detected a heavier load(e.g.through the additional weight of the upper part of the body leaning forward)as it is the case right before stepping over the side rail.In general,it is rea-sonable that an alarm is triggered already at this early stage. This applies if not only both legs of the patient at risk of falling are leaning on the bedrail(for comparison:category IV means putting the legs on the side rail from a lying position)but that there is an additional pressure caused by the intention to get out of the bed.

Category IV–VII:These categories involve the pressure val-ues which are generated through the contact with the sensor system as it might be the case during sitting or lying in the bed.

The side rails were put to an upright position for the cate-gories III–VII.

Fig.5demonstrates the ROC curve which was calculated on the basis of the results shown in Fig.4;the category III which cannot be de?nitely allocated,was excluded.(The results incl. category III are described subsequently.)For the area under the ROC curve(AUC),the following value was calculated:0.998 (95%con?dence interval:0.996–1.000[p<.001]).

Table1shows the coordinates of the ROC curve in the area of interest between a sensitivity and speci?city of100%each as well as the potential threshold values.(The threshold levels indicated in the table are average values of two subsequent, structured and observed test values.)

When the pressure value for the optimum alarm trigger threshold level was de?ned to be35mbar,the result was a sensitivity of96.0%(out of99cases in which the patient

leaves

Fig.5–ROC curve.

the bed,an alarm is generated in95cases)and a speci?city of 99.4%(The touching of the sensor system without the inten-tion to leave the bed leads to one false alarm out of178cases).

A Chi2test showed a highly signi?cant result(Chi2=255,659, df=1,p<.001).After including the category III as“contact with

Table1–Subarea of the coordinates of the ROC curve. Threshold level[mbar]Sensitivity1Speci?city

29.50 1.0000.073

30.500.9900.056

31.500.9700.034

32.500.9700.022

34.500.9600.006

36.500.9390.000

178i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s79(2010)173–180

the sensor system without leaving the bed”,the result for optimum alarm trigger threshold level of34mbar was a sensi-tivity of96.0%as well as a speci?city of95.5%.Now,43further contacts with the sensor systems were integrated in the cal-culations.For the area under the ROC curve(AUC),a value of 0.991with a con?dence interval of95%at0.984–0.997(p<.001) was calculated.

4.Discussion

Despite the fact that there are various bed-exit alarm systems on the health care market,there is only limited evidence that already existing bed-exit alarm systems have the capacity to prevent falls.Nevertheless,the authors of the Clearinghouse Guideline synthesis in the USA have come to the conclusion that with a technology like bed-exit alarm systems,a fall reduction can be expected(evidence stage IV,expert opinion) [24].The literature review showed only one article describing fall outcomes regarding bed-exit alarm systems.This report stated a signi?cant reduction of patient falls[1].Following the incorporation of bed-integrated exit alarm systems(Yuma Regional Medical Center,Hillrom VersaCare A.I.R.TM)into the fall prevention program,there was an overall decrease of patient falls from an average of5.85falls per1000patient days to2.81falls.Unfortunately,limitations of this report are various.There were no data about the patient popula-tion,diagnosis and other possible interventions.The bed-exit alarm systems were constantly applied for high risk patients and from7pm until7am for all other patients,regardless of their risk level.In Europe,the emphasise on forensic care is lower due to cultural differences and therefore bed-exit alarm systems for every patient at night are dif?cult to imag-ine.Thus,the effectiveness of the bed-exit alarm system for the reduction of falls is not suf?ciently proven by this study report.

Although BUCINATOR has not been tested yet in a hospi-tal or nursing home setting,the results indicate that it can be a reliable tool for fall prevention.The design of BUCINA-TOR was planned to?t the needs of both the nursing staff and the patients.It is expected that the new system can be used for all described patient characteristics regardless of the prob-lem they are suffering from.Restless,uncooperative,confused patients activate an alarm only if they put enough weight on an area of the side rail that is large enough and therefore the alarm is prede?ned in the event of an exit.Light-weighted patients do not make a signi?cant difference in response char-acteristics of the system,since only a small part of the body weight counts to trigger an alarm(e.g.15cm4.8kg).Although the system is primarily targeting bed exits and resulting falls, the researchers believe that the use of restraints like?xations and raised side rails are becoming less common.A decrease of restraints as reported by Jackson and Gleason[2]is based on the effect that there is another additional reliable system the nursing staff can rely on.So if risen side rails are used less often,serious events regarding side rail entrapments occur less often.But if side rails have to be raised for some reason, dangerous patient movements may be detected by BUCINA-TOR’s directly integrated sensor before a serious situation occurs.

Compared to the pressure-sensitive strips on mattresses, the hygiene factors like contamination through incontinence do not constitute a problem,as the sensor is not placed in a contamination-sensitive area.Furthermore,standardised wipe disinfection can easily be performed.

Another,probably most important,point is cost.Nurs-ing homes with limited?nancial resources are only going to use bed-exit alarm systems if they can economically argue the advantages of these systems.The researcher team was addressing these needs,by using cheap robust industrial components.Calculating costs for a system like BUCINATOR depends largely on the number of units produced.Material costs are around20D without assembling,marketing and other escorting activities like of?ce costs,margins,etc.By comparing this?gure with the costs of a fall(944D)[10]it is quite astonishing that bed-exits alarm systems are still so rarely used.

Unfortunately,the diffusion of innovations like bed-exit alarm systems is not only interfered by costs but also by pro-prietary designs.Integration into nurse call systems is seen as mandatory to keep the technological over?ow for nursing staff as small as possible.There is a need for standards which allow for the integration into nurse call systems,without adapting the plugs to?t into proprietary sockets of the different nurse call systems,which incurs further costs.

Kohn et al.[25]wrote in2000that“health care is decades behind other industries in terms of creating safer systems”. The topic of bed-exit alarm systems is a good example how the health sector and the health care industry are dealing with this subject.Tideiksaar[26]argued in2006that bed-related falls are not a coincidence,rather they are preventable,because a great number of accidents can be foreseen and therefore prevented, e.g.with bed-exit alarm systems.

The lack of well-designed integrated bed-exit alarm sys-tems in professional care settings like hospitals and nursing homes constitutes a challenge for the health care industry.A new market can be established by integrating bed-exit alarm systems into mattresses,side rails or other parts of a nursing bed.

Future use-cases are seen in the?eld of ambient assisted living technologies.The home care setting could greatly bene?t from bed-exit alarm systems with some extended communication features.Bed-exit alarm systems, for instance,can provide information if the person affected has left the bed.If activated,this information can be routed via text messaging(SMS)to any cell phone and inform car-ing relatives about the status change of the affected family member.A reliable bed-exit alarm system can provide con?-dence concerning the safety of their relatives and will make life easier and less strained for caring family members.So, socializing and other easing activities can recharge the quite often empty batteries and help relatives in ful?lling their role as the responsible caregiver.It is expected that technologies like BUCINATOR can be a component to relieve strain on rela-tives and might postpone a potential nursing home admission.

Further adjustments and extensions of functionality are conceivable.Ambulant services could bene?t from technolo-gies like BUCINATOR.A software application could provide additional safety for the caregivers and the persons affected.

A simple traf?c light system displayed on a computer monitor

i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s79(2010)173–180179

Summary points

What was known before the study:

?Multifactorial intervention programs can prevent falls and reduce the rate of injury for cognitively normal

older people.

?Cognitive impairment is a major risk factor for falls and is associated with increased risk of severe fall related

fractures.

?Most falls in nursing homes occur while the resident is getting in or out of bed.

?Physical restraints are not effective in preventing falls.?There are various bed-exit alarm systems with some advantages and disadvantages in use.

?Fall prevention strategies have not been tested by con-trolled trials in patients with cognitive impairment and

dementia who fall.

What the study has added to the body of knowledge:

?A requirement pro?le for bed-exit alarm systems has been generated.

?An uncomplex,universally applicable bed-exit alarm architecture has been developed and validated.?Sensitivity and the speci?city of the trigger level indicate a satisfactory alarm/false alarm ratio in the

prestudy.

?Bed-exit alarm systems can be a part of ambient assisted living technologies.

?Further consistently designed studies,particularly randomized controlled trails are required to show the

effect on falls of bed-exit alarm systems.

with information about mean bed times could give basic infor-mation on changes in bed rest times.If an affected patient stays longer in bed than a predetermined time,a light can change the status from green into red and automatically inform on potential danger.

Of course,all new technologies carry both risks and bene?ts.In particular,providers of ambient assisted living applications should take privacy and ethical aspects into account.

Further research is needed to show the effect of bed-exit alarm systems on fall rates.Research should also focus on the potential bene?t of these systems in the home care area and other use-cases.

Acknowledgements

The BUCINATOR Project was supported by formerly HITT-Health Information Technologies Tirol(now CEMIT-Center of Excellence in Medicine and IT).

Author contributions:The development and evaluation study was conducted by Johannes Hilbe as part of a Ph.D.study.He developed the initial framework which was critically reviewed and modi?ed by the other authors.The?rst draft of the paper was prepared by Johannes Hilbe and the other authors all con-tributed to developing the?nal text.

r e f e r e n c e s

[1]M.Moore,C.VanGilder,High Yields Result from a Pressure

Ulcer Prevention Program,Hill-Rom Services,2006.

[2]L.Jackson,J.Gleason,Proactive Falls Management—breaks

the fall cycle.Patient safety special,Nursing Management

635(2004),37,38.

[3] E.Wilson,Preventing patient falls.AACN clinical issues,

advanced practice in acute and critical care,Preventing

Negative Outcomes of Acute Illness in Adults9(1)(1998)

100–108.

[4] E.Fonad,T.Wahlin,B.Winblad,A.Emami,H.Sandmark,

Falls and fall risk among nursing home residents,Journal of Clinical Nursing171(2008)126–134.

[5] C.Schaubmayr,T.Praxmarer,W.Günter,Erforschung der

Ursachen von Patientenstürzen im Krankenhaus,

PrInterNet,2005,p.3.

[6]https://www.sodocs.net/doc/264009353.html,mbert,J.Wood,I.Kowanko,S.Simon,L.Thomas,P.

Butterworth,T.Mckinlay,A.Tolfts,A.Zoerner,Falls in

hospitals,in:Best Practice;Evidence Based Practice

Information Sheets for Health Professionals,vol.2,Joanna

Briggs Institute,1998,pp.1–6.

[7]L.Rubenstein,K.Josephson,Falls in the nursing home,

Annals of Internal Medicine121(1994)442–451.

[8] C.VanDoorn,A.Gruber-Baldini,S.Zimmerman,J.Hebel,C.

Port,Dementia as a risk factor for falls and fall injuries

among nursing home residents,Journal of the American

Geriatrics Society51(2003)1213–1218.

[9] A.Nelson,G.Powell-Cope,D.Gavin-Dreschnack,P.Quigley,

T.Bulat,A.Baptiste,S.Applegarth,Y.Friedman,Technology to promote safe mobility in the elderly,Nursing Clinics of

North America393(2004)649–671.

[10]I.Nurmi,P.Lüthje,Incidence and costs of falls and fall

injuries among elderly in institutional care,Scandinavian

Journal of Primary Health Care20(2002)118–122.

[11] E.Hitcho,M.Krauss,S.Birge,W.Dunagan,I.Claiborne,S.

Fischer,P.Johnson,E.Costantinou,V.Fraser,Characteristics and circumstances of falls in a hospital setting.A

prospective analysis,JGIM7(19)(2004)732–739.

[12] C.Becker,U.Lindemann,E.Kapfer,B.Eichner,M.Hausner,T.

Nikolaus,Verminderung von sturzbedingten Verletzungen

bei Alten-und P?egeheimbewohnern,in Mobilit?t und

Mobilit?tsst?rungen von Heimbewohnern,D.B.d.U.

Modellvorhaben,Ulm(2001).

[13]T.Krause,Sturzfolgen bei geriatrischen

Krankenhauspatienten,P?ege118(2005)39–42.

[14]W.VonRenteln-Kruse,T.Krause,Sturzereignisse station?rer

geriatrischer Patienten–Ergebnisse einer3-j?hrigen

prospektiven Erfassung,Zeitschrift für Gerontologie und

Geriatrie137(2004)9–14.

[15]J.Hamers,A.Huizing,Why do we use physical restraints in

the elderly?Zeitschrift für Gerontologie und Geriatrie38(1) (2005)19–25.

[16] D.Evans,J.Wood,https://www.sodocs.net/doc/264009353.html,mbert,M.FitzGerald,Physical

Restraint in Acute and Residential Care:A Systematic

Review,The Joanna Briggs Institute,2002.

[17] D.Evans,J.Wood,https://www.sodocs.net/doc/264009353.html,mbert,Patient injury and physical

restraint devices:a systematic review,Journal of Advanced

Nursing413(2003)274–282.

[18]FDA,Food Drug Administration,Safety Alert:Entrapment

Hazards with Hospital Bed Side Rails,1995.

180i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s79(2010)173–180

[19]P.Cheung,B.Yam,Patient autonomy in physical restraint,

International Journal of Older People Nursing in Association with Journal of Clinical Nursing14(3a)(2005)34–40.

[20]ECRI,National Institute for Clinical Excellence,Bed-exit

alarms,a component(but only a component)of fall

prevention,Health Devices33(5)(2004)157–168.

[21]ECRI,National Institute for Clinical Excellence,Evaluation

bed-exit alarms,Health Devices33(9)(2004)325–329. [22] D.Beardsley,M.Bias,G.Burger,Clinical Guidance for the

Assessment and Implementation of Bed Rails in Hospitals, Long-term Care Facilities,and Home Care Settings,H.B.S.

Workgroup,2003.[23]P.Mayring,Qualitative Inhaltsanalyse.Grundlagen und

Techniken,Beltz Verlag,Weinheim,Basel,2007.

[24]ECRI,National Institute for Clinical Excellence,Prevention of

Falls in the Elderly,American Geriatric Society,British

Geriatrics Society and American Academy of Orthopaedic

Surgeons,National Guideline Clearinghouse,2008,p.29.

[25]L.Kohn,J.Corrigan,M.Donaldson,To Err Is Human:Building

a Safer Health System,The National Academies Press,2000,

p.71.

[26]R.Tideiksaar,Bed Safety in Long-term Care,2006.