USF Libraries
USF Digital Collections

Biomechanical evaluation os injury severity associated with patient falls from bed

MISSING IMAGE

Material Information

Title:
Biomechanical evaluation os injury severity associated with patient falls from bed
Physical Description:
Book
Language:
English
Creator:
Bowers, Bonnie E
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla.
Publication Date:

Subjects

Subjects / Keywords:
Height
Floor mats
Bedrails
Elderly
Acceleration
Dissertations, Academic -- Biomedical Engineering -- Masters -- USF   ( lcsh )
Genre:
government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: The incidence of falls in the elderly population is a growing concern in the healthcare industry as associated morbidity is high, particularly morbidity associated with falls from bed. Bedrails were implemented as a device intended to reduce the incidence of falls from bed; however, recent evidence may indicate that bedrails contribute to adverse events including entrapment and entanglement. As such, efforts have been madeto reduce the use of bedrails and implement alternatives including height adjustable beds and floor mats. An instrumented anthropomorphic test dummy was used in the current study to measure the deceleration profiles of the head, thorax, and pelvis upon impact onto a tile surface or floor mat. The height of the fall was varied by using a height adjustable bed, and the impact site was varied by head or feet first falls.The deceleration profiles were used to determine mean maximum values across repeated trials and to calculate injury criteria at the head (HIC), thorax (TIC), and pelvis (PIC). The mean maximum values were further used to estimate the effect of adding bedrails. Injury severity was then predicted from the injury criteria calculated for the head. From this study, the mean maximum values were found to significantly increase with an increase in height regardless of fall direction. As such, the addition of bedrails consequently increased these values. Furthermore, the use of a floor mat significantly reduced the mean maximum values at the head and pelvis during head first falls and at the head and thorax during feet first falls. Injury criteria were also calculated for each body region and found to be significantly increased with an increase in height and decreased with the use of the floor mat.
Thesis:
Thesis (M.S.B.E.)--University of South Florida, 2005.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Bonnie E. Bowers.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 106 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001681104
oclc - 62774433
usfldc doi - E14-SFE0001010
usfldc handle - e14.1010
System ID:
SFS0025331:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Biomechanical Evaluation of Injury Severity Associated with Patient Falls from Bed by Bonnie E. Bowers A thesis submitted in partial fulfillment Of the requirements for the degree of Master of Science in Biomedical Engineering Department of Che mical Engineering College of Engineering University of South Florida Co-Major Professor: William Lee, Ph.D. Co-Major Professor: John Lloyd, Ph.D. Andrea Baptiste, MA Gail Powell-Cope, Ph.D. Date of Approval: April 8, 2005 Keywords: height, fl oor mats, bedrails, elderly, acceleration Copyright 2005, Bonnie E. Bowers

PAGE 2

i Table of Contents List of Tables iii List of Figures v Abstract vii Clinical Observations 1 Introduction 1 Mechanical behavior of physiological tissues 1 Incidence of falls and associated injuries in the elderly population 6 Incidence of falls from bed and associated injuries 7 Risk factors and biomechanical i ssues associated with falls 9 Intrinsic factors 9 Extrinsic factors 15 Environmental hazards 15 Impact surface 17 Gaps in the research 18 Injury Prevention 20 Introduction 20 Bedrails and legal issues 20 Bedrails and adverse events 21 Bedrails and reduction programs 22 Bedrail alternatives 23 Height adjustable beds 24 Gaps in the research 25 Injury Assessment 26 Introduction 26 Head injury criteria 26 Gaps in the research 29 Significance of Current Research 30 Introduction 30 Quantifying the mechanic s of falling from bed 30 Comparing injury prevention methods 31 Assessing injury duri ng a fall from bed 31

PAGE 3

ii Methodology 32 Location 32 Apparatus design and construction 32 Protocol 38 Results 40 Head first falls 40 Acceleration measur ed at the head 40 Acceleration measur ed at the thorax 45 Acceleration measur ed at the pelvis 49 Feet first falls 53 Acceleration measur ed at the head 54 Acceleration measur ed at the thorax 58 Acceleration measur ed at the pelvis 62 Discussion and Interpretation 70 Introduction 70 Head first falls 70 Feet first falls 73 General observations 76 Conclusions and Recommendations 77 References 80 Appendices 85 Appendix A: MatLab Code 86 Appendix B: Protocol 94 Appendix C: SAS Code 95

PAGE 4

iii List of Tables Table 1: Instrinsic factors associated with falling. 13 Table 2: Extrinsic factors asso ciated with falling. 17 Table 3: HIC (15 msec) NHTSA standards. 29 Table 4: Head mean impact decelerations measured with and without a mat during head first falls and calcu lated values based on trend line equations. 41 Table 5: Mean HIC values calculated fo r head deceleration profiles measured during trials with a nd without a mat during head first falls. 43 Table 6: Impact forces cal culated at the head for head first falls with and without a floor mat. 45 Table 7: Thoracic mean im pact decelerations measur ed with and without a mat during head first falls and cal culated values based on trend line equations. 46 Table 8: Mean TIC values calculate d for thoracic deceleration profiles measured during trials with and with out a mat during head first falls. 48 Table 9: Pelvic mean impact decelerati ons measured with and without a mat during head first falls and calcula ted values based on trend line equations. 50 Table 10: Mean PIC values calculated fo r pelvic deceleration profiles measured during trials with a nd without a mat during head first falls. 52 Table 11: Head mean impact decelerations measured with and without a mat during feet first falls and calcula ted values based on trend line equations. 55 Table 12: Mean HIC values calculated fo r head deceleration profiles measured during trials with and without a mat during feet first falls. 57

PAGE 5

iv Table 13: Thoracic mean im pact decelerations measur ed with and without a mat during feet first falls and calculated values based on trend line equations. 59 Table 14: Mean TIC values calculate d for thoracic deceleration profiles measured during trials with and with out a mat during feet first falls. 61 Table 15: Pelvic mean impact decelerati ons measured with and without a mat during feet first falls and calcula ted values based on trend line equations. 63 Table 16: Mean PIC values calculated fo r pelvic deceleration profiles measured during trials with and without a mat during feet first falls. 65 Table 17: Impact forces calcu lated at the pelvis for f eet first falls with and without a floor mat. 67 Table 18: Summary of measured mean ma ximum values and calculated injury criteria. 68

PAGE 6

v List of Figures Figure 1: Compact and spongy bone organization. 3 Figure 2: Compact (cortical ) and spongy (trabecular) bone stress versus strain curves. 5 Figure 3: Viscoelastic res ponse of compact bone. 6 Figure 4: Anisotropic res ponse of compact bone. 6 Figure 5: Bone mass associati on with age and gender. 10 Figure 6: Wayne State Tolerance Curve. 27 Figure 7: Risk of life-threatening brai n injury related to HIC values. 28 Figure 8: 50th percentile male anthropomorphi c test dummy with and without scrubs. 33 Figure 9: Carroll Healthcare ARRO Low Bed used to simula te common heights from which patients fall. 34 Figure 10: Posey floor mat used to simula te mats commonly used in the healthcare environment to cushion falls. 35 Figure 11: Sling designed to standard ize the fall from bed event simulation. 36 Figure 12: Tri-axial accelerom eter aluminum mounting bloc ks for the head, thorax, and pelvis. 37 Figure 13: Head mean impact decelerations plotted with an estimated trend line during head first falls. 42 Figure 14: HIC values plotted with an estimated trend line during head first falls. 44 Figure 15: Impact forces plotted with an es timated trend line during head first falls. 45 Figure 16: Thoracic mean impact decelerations plotted w ith an estimated trend line during head first falls. 47

PAGE 7

vi Figure 17: TIC values plotted with an estimated trend line during head first falls. 49 Figure 18: Pelvic mean imp act decelerations plotted wi th an estimat ed trend line during head first falls. 51 Figure 19: PIC values plotted with an estimat ed trend line during head first falls. 53 Figure 20: Head mean impact decelerations plotted with an estimated trend line during feet first falls. 56 Figure 21: HIC values plotted with an estimated trend line during feet first falls. 58 Figure 22: Thoracic mean impact decelerations plotted with an estimated trend line during feet first falls. 60 Figure 23: TIC values plotted with an estimated trend line during feet first falls. 62 Figure 24: Pelvic mean imp act decelerations plotted wi th an estimat ed trend line during feet first falls. 64 Figure 25: PIC values plotted with an estimat ed trend line during f eet first falls. 66 Figure 26: Impact forces plotte d with an estimated trend line during feet first falls. 67

PAGE 8

vii Biomechanical Evaluation of Injury Severity Associated with Pa tient Falls from Bed Bonnie Bowers ABSTRACT The incidence of falls in the elderly population is a growing concern in the healthcare industry as associated morbidity is high, particularly morbidity associated with falls from bed. Bedrails were implemented as a device intended to reduce the incidence of falls from bed; however, r ecent evidence may indicate th at bedrails contribute to adverse events including entrapment and entang lement. As such, effo rts have been made to reduce the use of bedrails and implement alternatives incl uding height adjustable beds and floor mats. An instrumented anthropom orphic test dummy was used in the current study to measure the deceleration profiles of the head, thorax, a nd pelvis upon impact onto a tile surface or floor mat. The height of the fall was varied by using a height adjustable bed, and the impact site was varied by head or feet first falls. The deceleration profiles were used to determine mean maxi mum values across rep eated trials and to calculate injury criteria at the head (HIC), thorax (TIC), and pelv is (PIC). The mean maximum values were further used to estimat e the effect of addi ng bedrails. Injury severity was then predicted from the in jury criteria calculated for the head. From this study, the mean maximum valu es were found to si gnificantly increase with an increase in height rega rdless of fall direction. As such, the addition of bedrails consequently increased these values. Furthe rmore, the use of a fl oor mat significantly reduced the mean maximum values at the head and pelvis du ring head first falls and at

PAGE 9

viii the head and thorax duri ng feet first falls. Injury crite ria were also calculated for each body region and found to be signi ficantly increased with an increase in height and decreased with the use of the floor mat. The HIC values we re used to predict injury severity and resulted in nearly a 40 percent chance of sustaining a serious brain injury under any condition tested during this study. Based on these results, the recommendation was made to position hospital beds to the lo west available position, place floor mats by the bedside, and remove bedrails to decrease the risk of inju ry as a result of falling from bed.

PAGE 10

1 Clinical Observations Introduction The human body is a complex organism that responds with unique characteristics to applied forces. As such, the human body has certain limitations and exceeding those limitations may result in injur y. For example, as observed in the literature, injury may occurs as a result of over-exe rtion due to falling. As desc ribed by Tinetti, Speechley, and Ginter (1988), a fall occurs when a person “unintentionally [comes] to rest on the ground or at some other lower level, not as a result of a major intrinsic event or overwhelming hazard.” Much research has focused on th e elderly population to identify associated injuries and risk factors because falling is a common occurrence among this population. Lacerations, contusions, fractures, and head injuries are associat ed with falling, in general, and, specifically, falling from be d (Lyons and Oates, 1993; Macgregor, 2000). Risk factors associated with falling are categ orized as intrinsic or extrinsic factors. Common intrinsic factors incl ude age, medication use, a nd decreased mobility; while extrinsic factors includ e stairs, poor lighting, and slippery floors (Baum, Capezuti, and Driscoll, 2002). The characteris tics of impact surface is an ex trinsic factor that has only recently received attention; however, Simpson et al (2004) showed that concrete subfloors increase the risk of hip fractures. This observati on may applicable to future designs of health care facilities. Mechanical behavior of physiological tissues The mechanical behavior of physiologica l tissues provides th e basis of how the human body responds to applied loads. An injury results wh en the applied force exceeds the strength of the tissu e whether hard or soft tissue, su ch as bone or muscle. When determining the strength of a physiological tissue, several factors must be considered including the basic anatomical organization of the tissue, and the rate and direction of the

PAGE 11

2 applied force. The composition, organization, and mechanical behavior of bone will be discussed as an example of th ese universal principles. Bone is composed of organic and inorgani c substances organized to maximize the forces the tissue is able to endure. Calciu m and phosphate minerals embedded in collagen fibers provide rigidity to the bone tissue, while a gr ound substance interspersed throughout the mineralized collagen matrix provides flexibility and resilience. Water is also a key component in th e composition of bone as it surrounds bone cells called osteocytes and binds to glyc osaminoglycans found in th e ground substance. Two distinct types of bone are pr esent in the human body and ar e distinguished by their level of matrix organization (Marieb, 2001). Spongy bone (Figure 1), found predominately in the skull, clavicle, ribs, sternum and the epiphyses of long bones, is characterized by small stin ts of bone or trabeculae. These trabeculae appear to be randomly arranged; however they align themselves along lines of stress to provide th e most support for the bone matr ix. The trabeculae contain lamellae (literally “lit tle plates”) and osteocytes conn ected by canals ca lled canaliculi, which provide nutrients to the bone matrix (Marieb ed, 2001).

PAGE 12

3 Figure 1: Compact and spongy bone organiza tion. Source: Bone and Skeletal Tissues. The structure of compact bone (Figure 1), found in ve rtebrae and long bones, shares some similarities with spongy bone in that they both contain lamellae and canaliculi, but the arrangement of these struct ures differenti ates the two. Compact bone has a much more highly ordere d system of struct ures. The basic st ructural unit of compact bone is the oste on, an elongated cylinder that is or iented parallel to the long axis of the bone. The lamellae of compact bones are organized into concentric layers to form the osteon; adjacent lamellae run in opposi ng directions to best withstand torsional stresses. Lamellae are also found between individual osteons and surrounding the entire bone shaft, just deep to the periosteum; these help fill ga ps between forming osteons and resist torsion of the enti re bone. Nutrients are su pplied to compact bone through haversion canals located in the center of each osteon; thes e canals are connected to the periosteum by Volkmann’s canals. The bon e cells, osteocytes, are located in small Compact bone

PAGE 13

4 cavities within the bone matrix called lacunae, which are connected by canaliculi in a manner similar to spongy bone. Although comp act and spongy bone wi ll predominate in various bones, a structurally intact bone cannot exist without each bone type, as both provide resistance to mechanical stresses. The epiphyses of long bone s, for example, is composed primarily of spongy bone, but comp act bone forms a thin outer shell to increase the overall strength of the bone. Likewise, spongy bone is found in the diaphyses of long bones, alt hough compact bone is the primar y tissue. Figure 1 shows this relationship between compact and s pongy bone of the diaphys is of the femur. (Marieb, 2001). Due to the difference in organization of the lamellae, compact and spongy bone display different biomechanical properties when tested under standardized conditions. Compact bone withstands much higher st ress than spongy bone with values reported between 100 and 150 N/mm2 for compact bone and between 8 and 50 N/mm2 for spongy bone (Nordin and Frankel, 2001). Spongy bone, however, can withsta nd higher percent deformation than compact bone before fa ilure; values range from 2 to 4 percent elongation for spongy bone as compared to 1 to 2 percent for comp act bone (Nordin and Frankel, 2001). Physically, these propertie s mean that compact bone can support much higher forces per unit of area than spongy bone. On th e other hand, spongy bone can store much more energy before fracture occu rs. This behavior is due to the varying density of the two types of bone. Compact bone has a higher dens ity than spongy bone; therefore, it has more materi al to support higher load s. Conversely, spongy bone dissipates more energy effectiv ely through voids between trab eculae. This difference is shown graphically in Figure 2, where the area unde r the curve represents energy absorbed prior to failure (Nordin and Fr ankel, 2001). The occurrence of bone fracture can also be correlated to applied force. The force reporte d in the literature requi red to fracture a skull ranges from 4930 N to 5780 N (Nahum, Gatts, Gadd, and Danforth, 1968); Schneider and Nahum, 1972). Studies have also been condu cted to determine th e force required to fracture the hip. One such st udy reported a force of approx imately 4340 N required to fracture a hip when the soft tissue is present (Ether idge, Beason, Lopez, Alonso, McGwin, and Eberhardt, 2005).

PAGE 14

5 Figure 2: Compact (cortical) and spongy (trabecula r) bone stress versus strain curves. Source: Nordin and Frankel, 2001 These properties, however, are dependent not only on the bone structure but also on the rate and direction of lo ading as these tissues are viscoelastic and anisotropic, meaning they respond differently to forces applied at different rates and directions, respectively. For compact bone, the amount of stress th e tissue is able to endure increases as the rate of impact increases; how ever, the amount of st rain, or deformation per unit area, decreases This means that compact bone can withstand high levels of stress at high loading rate s but only for small amounts of time. Figure 3 graphically illustrates the viscoelastic characteristic of compact bone. The mechanical response of compact bone is also affected by the direction of loading as indicated in Figure 4. As the direction of loading rotates fro m parallel to perpendicular to the long axis, the amount of stress and strain compact bone can endure decreases. The anisotropic nature of compact bone correlates with the orientat ion of the osteons, as they are aligned pa rallel to the long axis of the bone. Likewise the mechanical response of spongy bone displays a similar dependence upon rate and dir ection of loading. The visc oelastic and anisotropic characteristics of bone are impor tant when exploring injury mechanisms of fall events (Nordin and Frankel, 2001).

PAGE 15

6 Figure 3: Viscoelastic response of compact bone. Source: Nordin & Frankel, 2001. Figure 4: Anisotropic response of compact bone. Source: Nordin and Frankel, 2001. Incidence of falls and as sociated injuries in the elderly population Fall events affect all those living in a community whether dir ectly as the one who falls or indirectly as a caregiv er, family member, or friend. Studies of self-reported falls indicate that one in three persons over the age of 65 years living in the community will fall at least once annually (T inetti, Speechley, and Ginter, 1988; O’Loughlin et al, 1993). Currently, the elderly populati on includes 12 percent of the total population; however, the

PAGE 16

7 U.S. Census Bureau (2004) projects this num ber to increase to 25 percent of the total population by the year 2030. Not only is the incidence rate high, but according to Cham pion et al (1989), falls account for a substantial portion of injuries in all age gr oups. However, for persons aged 65 years and older, falls account for 40 percent of trauma relate d injuries as compared to 11 percent in younger age groups. Furthermore, eleven percent of all trauma related injuries caused by falls result in death in those aged over 65 y ears (Champion et al, 1989). In a two year study condu cted by Sattin et al (1990) incidence of injury and outcome were recorded for 2,994 injury events resulting fr om falls for a study population of 26,826 persons aged 65 years and older. Common fall-related injuries observed included open wounds, di slocations, sprains/strains, contus ions, and fractures of the skull, spine, upper and lower limbs, and hips. Sattin et al also noted th at women were 1.9-3.1 times more likely to sustain a fracture, other than one of th e skull, as men, presumably due to higher rates of oste oporosis in women compared to men. Of these 2,994 injury events, forty-two percent resulted in hospita lization and 2.2 percent resulted in death prior to or during hospital admission. St erling, O’Connor, and Bona dies (2001) further studied injury rates occurring in persons over 65 years of age as compared to those younger than 65 years and reported an injury ra te seven times higher in the older age group. Interestingly, age was also associated with injury to pa rticular anatomical locations. Those aged over 65 years were prone to injuries to the head/neck, chest, or pelvis/extremity regions, while those aged 65 years and younger were prone to abdomen and skin/soft tissue in juries. Sterling, O’Connor, and Bonadies further showed that injuries in the older study group were more severe than those of the younger study group. Incidence of falls from bed and associated injuries Falls from bed is a specific ty pe of fall that became a focu s of research as early as 1979, as reflected in a study conducted by Walshe and Rosen (1979). In this study, incidence and demographic data were collect ed from incidence reports for a one year period accounting for 86,000 patient bed days. During this time period, 53 reports were recorded. Eighty-three percent of those patie nts who fell from bed were identified as

PAGE 17

8 being over the age of 65 year s, although elderly patients contributed only 22 percent of the total patient population recei ving care. Innes and Turman (1983) further showed that 64 out of 270 falls that occurred over an 11 -month period were identi fied as falls from bed. Twenty-four of these falls from be d resulted in injury including fractures, lacerations, and hematomas. In a similar review of incidenc e reports filed for long-term care facilities, Gurwitz et al (1994) found that 401 falls from bed occurred out of 2,032 total falls reported for the fac ility during a 12-month period. Injury associated with fa lling from bed is not limited to the elderly population as several studies indicate that children also experience such injuries. Lacerations, contusions, fractures, and head trauma are mirrored in the pediatric population (Lyons and Oates, 1993; Macgregor, 2000) However, conflicting inci dence of these injuries is reported in the literature with a range as low as 15 perc ent (Lyons and Oates, 1993) and as high as 52 percent (Macgregor 2000). This wide range of inci dence of injury may simply be the result of vary ing fall environments or may indicate child abuse, as the majority of these falls are not witnessed by individuals other th an the caregiver. In response to these conflicting results, Berto cci et al (2003) proceeded to observe the biomechanics of children fal ling from a bed or couch onto various flooring surfaces by using an instrumented Hybrid II pediatric anthropomorphic test dummy (ATD). Head injury criteria (HIC) for 36 milliseconds (msec) of acceler ation was calculated to compare wood, linoleum, carpet, and pla yground foam flooring su rfaces when a fall occurred from a height of 0.68 meters (m). Axial tension, bending, a nd torsion were also measured in the femur and co mpared for each surface. Bert occi et al showed that playground foam resulted in 660 percent lower mean HIC values th an those calculated for wood. Likewise, playground foam also resu lted in the lowest axial tension values when compared to other flooring surfaces. Howe ver, Bertocci et al also showed that HIC values for all surfaces were not substantial enough to produce the higher incidence of injury as reported in the literature.

PAGE 18

9 Risk factors and biomechanical i ssues associated with falls According to Baum, Capezuti, and Drisco ll (2002), risk f actors for falls are generally categorized into intr insic and extrinsic factors. Intrinsic factor s are based on the patient’s medical condition such as impaired balance, poor physical functioning or medication interaction. On th e other hand, extrinsic factors are those introduced by the environment with which the patie nt interacts, for example, uneven walking surfaces, poor lighting and stairs. Each may contribute to a patient’s risk of fa lling and sustaining an injury. Furthermore, risk of injury increa ses linearly when multiple factors are present (Tinetti, Speechley, and Ginter, 1988). Intrinsic factors Several intrinsic factors have been iden tified by researcher s that increase a person’s risk of falling and su staining an injury. These f actors include various medical conditions, physical limitations due to age, and medication use. As with all mechanical systems, age of the components plays a key role in the performance of the overall system. The same holds true for biologi cal tissues, as bones tend to be come more brit tle with age due to the natural aging proce ss or clinical conditions such as osteoporosis. This change in mechanical properties can be correlated with the trend of decreasing bone mass with age, as shown in Figure 5. A difference in bone mass is also observed between genders and will become clinically relevant with age. Bone mass density is calculated based upon bone mass per unit area, and its association with fracture in cidence in falls among women aged 65 years and older was studied by Nevi tt et al (1993). The researchers concluded that decreased bone mass density at the site of the fall impact significantly increased the risk of sustaining a fractur e at the impact site.

PAGE 19

10 Figure 5: Bone mass association with age and gender. Source: Nordin & Frankel, 2001. As discussed previously, bone tissue ex hibits viscoelastic and anisotropic properties. The importance of these characteristics becomes clinically apparent when fall direction and impact site is considered. Nevitt et al ( 1993) assessed the risk factors associated with fall outcomes among women ag ed 65 years and older. The incidence, circumstances, and outcomes of falls were prospectively studied for 891 women over a 4 year period. Of these, 130 women sustaine d hip fractures, 294 women sustained wrist fractures, and the balance sust ained no fractures associated with falling. Nevitt et al concluded that women who sustained hip fractures were more likely to have fallen sideways or straight down than those who fell and did not sust ain a hip fracture. On the other hand, women who sustained a wrist fr acture were more likely to have fallen backwards than those who did not sustain a wr ist fracture. These conclusions indicate that the direction of the fall significantly in fluenced the site of impact whether on the hip or on the outstretched hand. Al so, the anatomy of the site of impact plays a crucial role in determining risk of fracture. For exampl e, the soft tissue of the buttocks may protect the hip by absorbing energy upon impact of a ba ckwards fall. However, the outstretched hand has little soft tissue to provide protecti on during the same backwards fall and direct impact to the extended wrist is imminent. Janken, Reynolds, and Swiech (1988) conducted a retros pective study of clinical characteristics identifiable on patients’ charts by comparing the charts of 631 patients over 60 years of age. By revi ewing incident reports, 331 pa tients in the sample group were identified as “fallers,” while 300 patien ts were “non-fallers.” “General weakness,

PAGE 20

11 decreased mobility of the lower extremities, sleeple ssness, incontin ence, confusion, depression, and substance abuse” were char acteristics observed to be significantly associated with those patients who fell (Jan ken, Reynolds, and Swei ch, 1988). Tinetti, Speechley, and Ginter (1988) also studied factors associated w ith the risk of falling in persons aged 75 years and olde r. Balance and gait, and se nsory tests were initially conducted, as well as, cognitive ability and medication use. The researchers observed a significant association with the use of specific medicati ons including benzodiazepines, phenothiazines, and antidepressants to fa lling. Cognitive impairment, lower-extremity disability, and palmomental palp reflex were also determined to predispose a person aged 75 years and older to falling. These characteristics are rela ted to the function of the nervous system and palmomental reflex, in particular, is th e contraction of the muscles controlling the movements of the lips and cheeks when the palm of the hand is stroked. To a lesser extent, foot probl ems and gait and balance abnorma lities were also noted to increase the risk of falling. Furthermore, Ti netti, Speechley, and Ginter showed the risk of falling to increase linear ly with the number of risk factors present. In a study conducted by Lord, Clark, and Webster (1991), physiological conditions were measured for 95 residents of a hostel for the aged and analyzed for association with fall events. “Decreased sensation in the lower limbs, de creased visual contra st sensitivity, slow reaction times, muscle weakness and decreased stability” we re significantly associated with falling. Lord, Clark, and Webster also found “contrast sensitivity, proprioception in the lower limbs, ankle dorsiflexion strengt h, reaction time, and sw ay” to distinguish persons who fell multiple times from persons who fell only once during the 1 year study period. Gaebler (1993) confirmed the associati on of blindness or poor vision to persons who fall multiple times in a retrospectiv e study of 50 multiple fa llers matched to 50 single fallers. Medication use and its association with falling was studied by Yip and Cumming (1994). Seventy-one patients ag ed 65 years and older who fe ll at least once during a one year period were compared to 55 patients aged 65 years a nd older who did not fall while residing in a nursing home. Medication use for each patient was recorded daily and analyzed for the 24 hours prior to a fall event. The researchers conclu ded that the use of

PAGE 21

12 antipsychotic medications increas ed the risk of falling in th ose aged 65 years and older. Yip and Cumming also noted that 40 percent of the 201 falls that occurred resulted in minor injuries including bruise s, sprains, and lacerations. However, head injury and fractures were reported in 2 and 4 percen t of the 201 falls, re spectively. Mendelson (1996) also studied various medications and their association with falling in all age groups receiving services in an acute-care hospital. Medication dosage of 253 patients who fell while in the hospital were compared to that of patients, ma tched by age, sex, and service received, who did not fall. Mendelson identified a significant association between certain medications classified as antidepressants, mino r tranquilizers or sedatives, and major tranquilizers and patients who fell. French et al (2004) further studied the association of be nzodiazepine use and dosage to injury. Injury-coded healthcare encounters, totaling 3,139, were reco rded in a veteran’s hospital over a three year period; typical injuries included fractures of the sku ll and extremities, sprains and strains, and contusions. Benzodiazepine use and dosage was identified for each injury and association significance wa s established. French et al determined that a 1 U or valium equivalent increase in dose increased the risk of experiencing an injury by six percent. Furthermore, increasing the exposur e to benzodiazepines by one week increased the risk of injury by four pe rcent. Although this study indi cates an increas ed risk for injury when using benzodiazepines, the injury mechanisms could not be identified due to limitations in the administra tive data. However, many of the injuries reported are commonly associated with falling; therefore, benzodiazepine use should be carefully considered before prescribi ng for individuals identified as at high risk for falls. A comprehensive list of intrinsic factors comm only associated with falls was compiled by Baum, Capezuti, and Driscoll (2002) and is presented in Table 1.

PAGE 22

13 Table 1: Instrinsic factors associated with fa lling. Source: Baum, Capezu ti, & Driscoll, 2002. Chronic Illnesses/Symptoms /Signs Other neurological Functional Cerebellar disorders Inability to move legs or arms independently Un ilateral weakness Shy-Drager syndrome Multiple sclerosis Physical cognitive inability or l ack of Cervical spondylosis knowledge to use assistive device 8th cranial nerve tumor correctly Neurosensory Cardiovascular Postural orthostatic hypotension Aortic Stenosis Congestive heart failure Arrhythmias Impaired hearing Impaired vision: cataracts, glaucoma, macular degeneration, and/or presbyopia Pain, especially of joints Polyneuropathy secondary to diabetes, Anemia e.g. iron deficiency usually peripheral vascular disease, or secondary to Gl blood loss, B12 deficiency, anemia of chronic disease alcoholism Psychiatric Musculoskeletal Arthritis osteo, polymyalgia rheumatica Dementia Depression Foot disorders Post-stroke Osteoporosis Acute Illnesses/Symptoms/Signs Disuse or deconditioning syndrome Functional Osteomalacia History of fracture New-onset of weakness or incapacity in movement of extremities recent, rapid Post-amputation Proximal muscle weakness decline in functional status (IADLs or ADLs) Myopathy Hypovolemia Neuromuscular Low plasma volume Stroke Anemia Parkinson’s disease Venous stasis Huntingdon’s disease Blood loss Severe diarrhea

PAGE 23

14 Table 1: Continued Acute Illnesses/Symptoms/Signs cont. Psychiatric cont Autonomic neuropathies Anxiety Diabetic Hysterical fainting (conversion reaction) Uremic Recent, post-stroke personality change Toxic Musculoskeletal Amyloidosis Fracture (hip, vertebral compression) Neurological Sprain Transient ischemic attacks/recent stroke Respiratory Seizures Tussive syncope (syncope related to Vestibular dysfunction unrelenting cough) Glossopharyngeal neuralgia Pneumonia Cardiovascular Massive pulmonary embolism Postural orthostatic hypotension Pulmonary tamponade Vasovagal response Hypocanpia due to hyperventilation Carotid sinus syncope Hypoxia Vasodepressor syncope (fatigue, hunger, Defecation syncope heat) Acute abdomen cholecystitis, Acute heart failure pancreatitis appendicitis, diverticulitis New-onset arrhythmias Diarrhea Acute myocardial infarction Vomiting Aortic stenosis hypertrophic Blood loss cardiomyopathy Hypo/hyperthyroidism Carotid artery compression Hypogycemia Genitourinary Anemia Post-micturition syncope Hypokalemia Urinary tract infection Dehydration New-onset incontinence Hyponatremia Psychiatric Acidosis Delirium (often indicative of underl ying Hypocapnia (hyperventilation) acute physical illness)

PAGE 24

15 Table 1: Continued Behavioral Symptoms Cancer chemotherapeutic drugs Poor judgment regarding personal safety Quinine drugs Cautiousness due to fear of falling Central nervous system drugs Risk-taking or impulsivity (may be Ototoxic drugs (aspirin) secondary to stroke or impaired Psychotropics cognition) Hypnotics/sedatives Tendency to stand quickly, especially Antidepressants from bed or immediately after a meal Dopamine agonists Effort to remove physical restraint Circulatory drugs Propensity to climb over or around side Diuretics rails Antihypertensives Disinterest or inability to use Vasodilators (nitrates) recommended assistive devices Alpha blockers Vestibulotoxic drugs Beta blockers Aminoglycoside antibiotics Antiarrhythmics Extrinsic factors Environmental hazards According to an article written by Arlene Jech (1992), certain environmental conditions present tripping hazards around th e home. These hazards, seemingly obvious, are often overlooked du e to their temporary nature or ha bitual use. Fo r instance, an extension cord temporarily obs tructing a walkway may not be viewed as a hazard by elderly persons due to years of experience of maneuvering to avoid such circumstances. However, decreased reaction ti me common to persons aged 65 years and older increases the likelihood of failed balance recovery once a fall is initi ated. Other conditions around the home may contribute to falls due to th e changing physicality of elderly persons without a corresponding change in behavior. Decreased visu al acuity requires more adequate lighting in the home, particularly in hallways and around stairs. However, persons habitually using the stairs may feel their surroundings are memorized, and

PAGE 25

16 changes to the environment are not necessar y. Connell and Wolfe (1997) further studied environmental hazards often present in the hom e and the resident’s interaction with his environment on a situational basis. The researchers postulated that while certain environmental geometries are not hazardous per se, hazard potential is introduced when combined with decreasing phys ical ability or inattenti on to surroundings. During the study, eighteen fall or near-fal l incidences were recreated to surmise the ca use of the loss of balance. The researchers concluded that “collisions in the da rk, failing to avoid temporary hazards, preoccupation with temporary conditions, fri ctional variations in foot contact, excessive environmental demands habitual environmental use, [and] inappropriate environmental use” were comm on initiators of falls. Connell and Wolfe emphasized the interaction between the reside nt and the environmen t at the time of the fall. They observed that, though a particular action may be habitual, misjudgment of spatial orientation or preocc upation with temporary circ umstances often lead to a discrepancy in body movements. For instance, a person rising from bed to walk to the bathroom without a light may misjudge proximity of an obs tacle and trip rather than successfully maneuvering around it. Also, a person carrying a box along a familiar path may be preoccupied with the aw kward load rather than paying attention to an obstacle, temporary or permanent. It is these unique si tuations that initiate falls rather than the object itself. However, the presence of th e object presents a potential for hazardous conditions when subjected to the aforementioned circum stances. As such, Baum, Capezuti, and Driscoll (2002) compiled a li st of common environmental conditions that are potentially hazardous and shou ld be assessed for fall prevention; a reprint of this list is presented in Table 2.

PAGE 26

17Table 2: Extrinsic factors associat ed with falling. Source: Baum Capezuti, & Driscoll, 2002. In General Bedroom Poor lighting High bed Slippery floors No hand rail or other transfer enabler Low seating Stairs Unstable furniture No hand rails Shiny floors Worn treads Thick pile carpeting High shelving Stairs not visibly different than adjoining floor Bathroom Stair edge not clearly defined No grab bars Low toilet seats Impact surface The characteristics of the impact surface greatly influence the fall outcome by contributing to the forces involved during the fall event. Researchers have only recently begun to study this association of impact surface and fall outcome. Nevitt et al (1993), as discussed previously, compared circumst ances surrounding the in cidence of falls resulting in no hip fractures a nd falls that resulted in hip fractures among women. Nevitt et al found that women who fell on hard su rfaces, defined as “asphalt, concrete, stone, tile, linoleum, hardwood floors and unpadded carpets,” were more likely to sustain a hip fracture than those who fell on soft surfaces, defined as “g rass, loose dirt, and padded carpets.” Casalena et al (1998) developed a novel floori ng material, known as the Penn State Safety Floor (PSSF), which minimi zes the deflection under normal walking conditions while allowing a maximum defl ection under impact conditions, thereby creating a material with viscoela stic properties. Initial testin g of the PSSF indicated that the design achieved the goal of decreasing th e peak impact force experienced by a hip when measured with an anthr opomorphic mechanical device. Simpson et al (2004) further considered the correl ation between flooring surface and the number of hip fr actures resulting from falls in 35 nursing homes that reported a total of 6,641 falls over a two year period. Using a transducer developed to model an

PAGE 27

18 elderly patient’s hip, Simpson et al measured the impact force produced on carpeted and uncarpeted wooden and concrete sub-floors used in the study environments. Wooden sub-floors resulted in lower im pact forces than those associat ed with concrete sub-floors. These measurements were also correlated to the number of hip fr actures per 100 falls recorded throughout the study pe riod. Consequently, wooden sub-floors resulted in significantly fewer hip fractures per 100 falls when comp ared to concrete sub-floors. However, carpeted floors resulted in 88 percent of the to tal number of falls recorded during the study period, regardless of wooden or concrete sub-floor. This result may indicate a trip hazard associated with carpet or may simply in dicate a higher exposure to carpeted areas in the healthcare f acilities. Gaps in the research Most research focuses on the incidence of falls in those aged 65 years and older because age is a key fact or in determining risk for falls. However, risk factors are not limited to the elderly population. For instan ce, Tsai, Witte, and Radunzel (1998) showed that “history of falling [within the] past six months, generalized weakness, observed difficulty in mobility of lo wer extremities or walking, confusion/disorientation, [and] elimination problems (n octuria, incontinence)” were risk factors common to the elderly population and to patients recei ving care in a psychiatric un it. The researchers also showed an increased risk of falling with an increase in body temper ature and a positive association of falling with certain me dications, including se datives/hypnotics and antidepressants. Similar activ ities surrounding fall events in the elderly population and patients in the psychiatric uni t were also noted to include getting out of bed, moving from a sitting to a standing positi on, and walking to the bathroom Despite the incidence of falls presented by Tsai, Witte, and Radunzel, fe w other studies addre ss the issue of falls in the psychiatric population. Furthermore, although much attenti on has been given to fall patterns, few studies ha ve been conducted to quantify the injury mechanisms of falling from bed. Bertocci et al conducte d a study to quantify th e impact decel eration of children falling from a bed or couch onto various flooring su rfaces, but the researchers did not address the issue of adults falling under simi lar circumstan ces. A basic

PAGE 28

19 understanding of injury mechanisms must be obtained before prevention efforts can be effectively employed. Also, knowledge of th e fall environment, specifically impact surfaces, is lacking as researchers have only just begun to correlate incidence of injury and characteristics of the impact surface. Although Casalena et al have developed a new flooring system that may redu ce the number of injuries a ssociated with falling, the feasibility of implem enting this system into the hea lthcare environmen t has not been addressed.

PAGE 29

20 Injury Prevention Introduction Due to the high incidence of injuries associated with falls, researchers and healthcare providers have developed and impl emented several devices that patients may utilize to prevent injuries if a fall occurs. Falling from bed is a sp ecific circumstance of falling that healthcare provide rs are striving to prevent by first implementing bedrails. However, studies have shown serious adverse ev ents associated with the use of bedrails. As such, world wide efforts ar e directed towards reducing the usage of bedrails by careful patient evaluation and implementation of height adjustable beds, bedside floor mats, and other alternatives. Nevertheless, legal issu es and perceptions of bedrail usage have proven to be a stumbling bloc k to these efforts. Despite these complications, reduction programs have been successful in reducing th e usage of bedrails without increasing the number of injuries associat ed with falling from bed. Bedrails and legal issues Bedrails were introduced in healthcare inst itutions as a device intended to prevent falls from bed. However, the introduction of bedrails into the healthcare environment raises questions of legal liability as disc ussed by Barbee in 1957. Ironically, nurses could be liable for injuries sustai ned by a patient when a fall out of bed occurred whether the bedrails were up or down. Barb ee explained that negligence ca n be argued on the part of the nurse if a hospital or docto r’s order was not followed by ra ising the bedrails and if a nurse failed to professionally judge the requirement of raised bedrails if a standing order was not given. Consequently, Rubens tein et al (1983) assert that the use of bedrails stems from consensus rather than scientific evidence as legal liability a nd malpractice issues became entangled with the h ealthcare industry. Accordi ng to the authors, higher settlements are awarded in cases in which th e hospital or nursing staff fail to produce

PAGE 30

21 evidence of actively attempting to prevent in jury. Since bedrail position can be easily documented, raising the bedrails became st andard practice not only to prevent patient falls but also to prevent claims of negligence on the part of the hospital or nursing staff. Recently, liability issues became further complicated with the inclusion of bedrails in the April, 1992 revision of hospital restraint devices posted by the Health Care Financing Administration (HCFA) (Braun & Ca pezuti, 2000). Bedrails can be classified as restraint devices if a doctor’s order speci fically states the device’s pur pose as a mobili ty restraint; as such, bedrails used as a restraint must be accompanied by a doctor’s order to be legal. Otherwise, bedrails that aid a patient tran sferring into and out of bed or repositioning while in bed are classified as assistive devices; therefore, they do not require a doctor’s order to be implemented (Donius & Rader, 1994). Bedrails and adverse events The decision-making process to use bedrails for a pati ent must consider these legal issues and the effect th is device will have on the pa tient as bedrails reportedly contribute to serious adverse events. Park er and Miles (1997) f ound 74 incidences of patient deaths that were asso ciated with bedrails as repor ted to the US Product Safety Commission between 1993 and 1996 These deaths were clas sified into one of three subgroups by identifying the type of entrapme nt involving bedrails. The majority of the reported incidences (70 percent) were caus ed by the patient being trapped between the mattress and bedrail. Bedrail compression of the patient’s neck also resulted in 18 percent of the incidences reported, while 12 percent were caused by the patient becoming trapped between the bedrail and floor after partially falling or sliding off the bed. Furthermore, 111 cases of bedrail entrapment were identified by Todd, Ruhl, and Gross (1997) in a review of adve rse events reported to the Food and Drug Administration (FDA) for the years 1985 thr ough 1995. Of these 111 entrapme nts, 72 resulted in death with “asphyxiation, strangulat ion, suffocation, cardiac arrest cardiac arrhythmias, or pneumonia” cited as the cause of death. “Fractures, spra ins, soft tissue injuries, and respiratory or circulatory compromise” were also reported in 26 of these entrapments. Due to staff intervention, 13 of these entrap ments had no associated injury. Accordingly,

PAGE 31

22 alerts have been posted by the FDA (FDA, CDRH, 1995) and the Joint Commission on Accreditation of Healthcare Organizati ons (JCAHO) (JCAHO, 2002) regarding the potential safety risk involved when using bedrails. The FDA has also sponsored the organization of the Hospital Bed Safety Workgroup, which is currently striving to establish bedrail and hospital bed standards th at will reduce the inju ry risk posed by these systems. Not only do these events aff ect the patient physically, but they also produce an economic impact as shown by Bradham et al (2003). Incidence and comparison groups were analyzed for acute me dical/surgical care units and nursing home units within a regional Veterans Health Administration ( VHA) network. Over a one year period, 207 medical/surgical admissions, accounting for 236 adve rse incidences, wa s compared to 732 admissions without injury. Likewise, 191 nursing home admissions, accounting for 194 adverse incidences, were co mpared to 194 admissions wit hout injury. By comparing the total number of bed days, procedures, and surgeries, Bradham et al estimated a potential total of $1,858,620 to be saved annually in direct costs if bed-related adverse events were avoided in acute and extended care facilities. Bedrails and re duction programs In response to these report ed adverse events, studies have been conducted to determine if bedrails can be safely remove d without increasing th e number of injuries associated with falling from bed. Accordingl y, Feinsod et al (1997) replaced full-length bedrails with half-length bedrai ls, eliminated them completely, or utilized low beds as an alternative to full-length bedr ails in a long-term care facili ty. Consequently, the injury rate recorded for 118 patients prior to fu ll-length bedrail replacement did not differ significantly after the afor ementioned bedrail alternatives were implemented for 128 patients, nor did any injury requiring hospitalization occur af ter the full-length bedrails were replaced. These findings are mirro red in a study involvi ng a short-term rehabilitation unit. Si et al (1999) indivi dually assessed patients and incrementally removed split rails from beds in a 25-bed unit ov er a one year pe riod accounting for 143

PAGE 32

23 admissions. Incidence of fa ll did not differ significan tly from the study period as compared to the previous year before bedrail usage was reduced. The reduction of bedrail us age has drawn world-wide at tention as Hanger et al (1999) conducted a similar st udy in a New Zealand hospital. Potential hazards of bedrails and available alternatives were presen ted in an educational program in an effort to reduce the use of bedrai ls throughout the 135 bed hospi tal. Six months after implementing the program, the mean number of beds with bedrails was reduced from 40 beds to 18.5 beds. The incidenc e of falls and associated inju ries was also collected and analyzed six months be fore and after the program interv ention. The number of falls did not differ significantly nor did the incidence of injury; however, the morbidity of falls decreased as fewer serious injuries occurred. Fractures, head trauma joint dislocations, lacerations requiring sutures or plastic surgi cal intervention, or hip pain that immobilized the patient were all classified as serious injuries. Likewise, Hoffman et al (2003) implemented a program called BedSAFE to reduce the use of bedrails in three Veterans Health Administration nursing homes. After pa tient assessment, bedra il alternatives were implemented including floor mats, mattress per imeter borders, height adjustable beds, or environmental changes. Overall, a 27 percen t reduction in bedrail use was made over the one year study period of the 60-bed long-term care units. The number of falls from bed also decreased by 11 percent; however, the number of injuries did not differ significantly preand post-BedSAFE. Bedrail alternatives During the bedrail reduction program Be dSAFE (Hoffman et al, 2003), nursing staff used bedside floor mats to prevent inju ries associated with falling from bed. The researchers noted that 89 pe rcent of 126 post-BedSAFE injuri es occurred when the floor mats were not in place. This suggests th at the floor mats di d indeed contribute a protective effect against injuri es resulting from falls out of bed; however, the significance of this result was not analyzed with respect to injury rate. Also, the impact force was not measured during this study, thus the level of protectiveness of the floor mats could only be inferred. Thus far, clinic ians must rely upon manufacturer ’s advertisements of impact

PAGE 33

24 reductions associated with th e use of floor cushions. For in stance, the Posey floor cushion is advertised to provide an 85 percent reduction in impact fo rce, measured in meters per second squared (g), as compared to the im pact generated by a ba seball bat striking a baseball. Despite rigorous testing standard s developed by the United States government, consumers may be misled by subjective interpretations of test results. Height adjustable beds Low-beds or height adju stable beds are a commonly recommended alternative to bedrails; however, increasing th e prevalence of low-back in juries in nursing staff is a concern if patients receive car e while the bed is in the low position. DeLooze et al (1994) conducted a study to measure th e effect of individually ch osen bed heights on peak and time integrated compressive and shear forces on the L5-S1 vertebral joint. Fourteen female and eight male nurses were asked to complete a set of pati ent handling tasks at a standard bed height of 0.715 m and again at a bed height of subject preference. Reflective markers placed on anatomical landmarks were tracked by a motion capture system and used, in conjunction with force pl ate and anthropomorphic data, to calculate joint reaction forces at the L5-S1 joint experienced duri ng the various tasks. The researchers reported a signif icant decrease in peak shear forces and time integrated compressive and shear forces when the subject adjusted the bed to a comfortable working height; therefore, implementa tion of height adjustable be ds was not excluded as a possible alternative to bedrails. Caboor et al (2000) cond ucted a similar study to me asure the effects that individually chosen bed hei ght has on spinal motion, asso ciated muscular activity, and the level of exertion perceived by the nurses du ring patient handling task s. The series of tasks, completed by ten female and eight male nurses, includ ed repositioning a patient in bed and patient transfers between a bed and wheelchair or toilet at a standard bed height of 0.515 m and an individually chosen bed height. Spinal motion was measured by electrogoniometers, while surface electrom yography (EMG) measured the activity of major muscle groups associated with spinal motion. The subjects al so rated the exertion level required to comp lete each task according to a 15 point rating scale ranging from

PAGE 34

25 extremely light to extremely h eavy. Caboor et al showed th at, while the range of spinal motion did not change significantly, the time spen t in the neutral position or safe working zone of the spine was significantly increased with the individually chosen bed height. The authors also showed that according to the EMG data, individual bed height preference did not significantly affect the muscle activity co mpared to that associated with the standard bed height. Similarly, no significant difference was perceived by the subjects with respect to exertion level required at each bed height. Gaps in the research From these studies one may conclude that implementing height adjustable beds is an appropriate alternative to bedrails based on the effect to the nursing staff; however, research has not addressed the effect height adjustable beds will have on patients. Healthcare providers have impl emented this alternative with out conducting the research to prove any benefit of the device for patients. For example, the e ffect of falling from various bed heights ha s not been studied and only assume d that falling from bed at a lower height will give a protec tive effect from injury. Like wise, the question of aiding patient mobility by lowering the bed height has also not been addressed. Floor mats is another form of bedrail alternat ive intended to protect a pati ent from injuries if a fall from bed occurs. These mats have been implemented into the healthcare environment without proof of their protectiv e effect. The performan ce of these devices s hould be compared in objective tests that quantify their impact da mpening capabilities. Also, because these mats must be incorporated into the whole en vironment, other issues must be addressed including tripping hazards, ease of use by nursing staff, and sanitation methods.

PAGE 35

26 Injury Assessment Introduction In the medical community, various in jury assessment scal es are used to communicate injury severity. However, many of these scales are subject to caregiver interpretation and are not appropriate for i ndustrial use when deve loping new products. As a result, the head injury criteria (HIC) was developed as a method of quantifying the mechanical response of the human head to various impact s ituations. This value has, therefore, been used to esta blish standards in th e automotive industry in an effort to promote safety of automobile s. Consequently, the HIC va lue is widely used in the automotive industry but has seen limi ted application in other areas. Head injury criteria Thus far, in the healthcar e industry, assessing injury severity is subject to caregiver interpretation and categ orical scales. One such scal e is the Abbrev iated Injury Scale (AIS) developed by clinicians to assess th e injury severity of a patient. This scale is a categorical scal e from 0 to 6 that allows clinic ians to assess and communicate a patient’s probability of survival. The categories are as follows: 0= No injury 1= Minor injury 2= Moderate injury 3= Serious injury 4= Severe injury 5= Critical injury 6= Unsurvivable injury Categories 0 to 3 are genera lly associated with head and neck pain and mild concussion, while categories 5 a nd 6 are classified as unsurv ivable injuries (Trauma.org, n.d.). Researchers have begun to bridge th e gap between clinical observation and experimental data. This pr ocess began with Lissner (1960) and the introd uction of the Wayne State Tolerance Curve (WSTC) which characterizes the mechanical response of

PAGE 36

27 cadaver heads upon impact. This curve was used to correlate rate and duration of impact with skull fracture and is show n in Figure 6. The curve repres ents the tolerance level of the human head to accel eration and duration of impact; however, this curve does not indicate injury severity. The WSTC wa s further used by Versace (1971) to develop a mathematical predictor of head injury. The He ad Injury Criteria (HIC ) (Equation 1) is an integral calculated for impact accelerat ion up to a 15 ms time period (HIC15)(Mertz, 1994). The maximum value of th is integral is reported as the HIC value for a given impact event and is indirectly correlated with the AIS used by clinic ians by predicting the risk of life-threatening brain in jury as shown in Figure 7. Equation 1 where t= time (ms) a= acceleration (g) Figure 6: Wayne State Tolerance Curve. Source: Versace, 1972.

PAGE 37

28 Figure 7: Risk of life-threatening brain injury related to HIC values. Source: Mertz, 1994. The automobile industry has incorporated calculating the HIC value as a method of predicting injury by using instrumented dummies during crashw orthiness testing of automobiles. Currently, the Hybrid III anthropomorphic test dummy is the most commonly used mannequin. The U.S. Depart ment of Transporta tion has established safety standards based on HIC15 values calculated for various models of the Hybrid III dummy. The most current standards were esta blished in 2002 and are presented in Table 3. Ongoing research to establis h injury criteria for the thorax is being conducted, and the current NHTSA standards are also presented in Table 3. 700

PAGE 38

29Table 3: HIC (15 msec) NHTSA standards. Source: Eppinger, Sun, Kuppa, and Saul, 2000. Head Criteria (HIC15) Chest Acceleration (g) 95th Percentile Male 700 55 50th Percentile Male 700 60 5th Percentile Female 700 60 6 year old Child 700 60 3 year old Child 570 55 1 year old Infant 390 50 Gaps in the research Although widely used in automobile de velopment, the HIC value has limited application to areas outside th is industry. Bertocci et al, as discusse d previously, used this value to predict head in jury associated with falling from bed in children; however, this study is exceptional in that application. At the time of this writing, no other studies utilize the HIC value for this application neither for children nor adults. The HIC value has great potential for allowi ng researchers to objectively assess the effect any given situation will have on the human head. Thus far, this assessment tool has remained virtually untapped with regard to areas ou tside the automobile i ndustry. Furthermore, similar assessment values are being developed for other anatomical lo cations, such as the thorax. However, these valu es are not well established in any industry, much less the medical research community. Th at is not to say that the automobile industry does not measure basic values, such as acceleration and force, for other parts of the body, but these areas of the body have received much less attent ion than the head. Therefore, industry standards are not as rigorously established for th e thorax or pelvis as those of the head. As such, researchers must rely only on these ba sic values to communica te the response of the human body under certain conditions. Fo r instance, falling from bed produces an impact not only to the head but also to th e thorax and pelvis; ye t, researchers can only infer injury based on accel eration measurements.

PAGE 39

30 Significance of Current Research Introduction As discussed previously, information re garding the biomechanics of falling from bed is by no means complete. Currently, most research studies are based on subjective evidence of the protective effects of bedrails and bedrail alternatives, including height adjustable beds and floor mats. These devi ces were implemented into the healthcare environment with the intention of preventi ng injuries caused by fa lling out of bed; however, few studies have been conducted to investigate patient-d evice interactions objectively. Furthermore, the subjective nature of these studies does not allow effective communication between healthcare providers and the manufactures of healthcare devices. The current study proposes to address some of these issu es by first, providing a quantitative measure to descri be the mechanics of falling from bed, second, quantitatively comparing height adjustable beds and floor mats, and third, applying an injury assessment criteria to the specific situation of falling from bed. Quantifying the mechanics of falling from bed The current knowledge base do es not provide any measur e of what happens when a patient falls from bed. There is, howe ver, considerable evidence supporting the incidence of falling from be d and associated injury. Mu ch research has also been conducted to identify risk fa ctors including those specific to a patient and those found in the environment. Unfortun ately, these studies do not communicate effectively the mechanics of a fall from bed event. As su ch, the current study proposes to objectively measure the impact decel eration during a fall fr om bed event. Not only will this measure provide an unbiased descripti on of the moment of injury du ring a fall event but also provide a baseline with which to compare future studies involving the assessment of prevention methods.

PAGE 40

31 Comparing injury prevention methods There exists a trend in th e healthcare community th at involves implementing devices intended to prevent in juries without properly asse ssing their effectiveness at achieving that underlying goal. This trend is evidenced with the implementation of bedrails and bedrail alternatives including he ight adjustable beds and floor mats. The current study, however, will provide a measur e to objectively compare the effectiveness of height adjustable beds and floor mats by measuring the impact deceleration during a fall event when these devices are in use. Furthermore, the performance of these devices will be assessed during a clinically relevant environmental arrangement. Consequently, the presence of any physical benefit that may ex ist will be established; therefore, it will allow healthcare providers to make a more informed decision about utilizing these devices. Assessing injury during a fall from bed Injury criteria exist that ar e used to correlate injury severity with a physical measurement; however, these criteria, thus fa r, have not been applied to clinical situations. The Head Injury Criteria (HIC) is an assessment value that correlates acceleration with injury severi ty; however, this criteria has only been applied to the automotive industry. The cu rrent study proposed to a pply HIC to the specific circumstance of falling from bed. As su ch, HIC will provide a correlation between impact deceleration and injury severity. Therefore, clinic ians will be more informed about the potentially hazardous situation of falling from bed. The equation used to calculate HIC values was also applied to thoracic and pelvic acceleration profiles; therefore, injury severity may be inferred once research becomes av ailable regarding the physical limits of the thorax and pelvis.

PAGE 41

32 Methodology Location The current study was conducte d at the Biomechanics Labo ratory of the James A. Hailey Veterans Administration Patient Safety Center. Apparatus design and construction To represent the patient population, a Hybrid III anthropom orphic test dummy (ATD) (manufactured by Denton ATD) was used during this study (See Figure 8). Numerous studies have been c onducted to develop this ATD with physical characteristics similar to those of living persons includi ng anthropomorphic dimens ions, joint range of motion, and response to applied forces (Back aitis and Mertz, 1994). For this study, a 50th percentile male dummy was used, which mean s 50 percent of the to tal population would have anthropomorphic dimensions no larger than those of the ATD. As such, the ATD is designed to weigh 76.3 kg and measure 170.3 cm when standing erect. Since the ATD is not automated, it represents a fully dependen t male patient, a co mmon patient receiving care in the VA healthcare netw ork. To further enhance th e patient simulation, the ATD was clothed in hospital scrubs Although the ATD is designed to meet population data with regard to dimension and response, the AT D skin does not mimic that of human skin with respect to friction coeffi cients. The ATD skin is com posed of vinyl which produces a much higher coefficient of friction than does human skin. The inclusion of scrubs on the ATD decreased the likelihood of skewed data from differi ng friction coefficients as hospital patients are clothed in scrubs or gowns of similar material (See Figure 8).

PAGE 42

33 Figure 8: 50th percentile male anthropomorphic te st dummy with and without scrubs. A Carroll Healthcare ARRO Low Bed was used to simulate the height from which patients frequently fall (See Figure 9). This bed provided an adjustable height range between 33.5 and 97.5 cm, measured from the floor to the top of the uncompressed mattress. Due to the wide variety of beds used in hospitals, the height range from could which patients fall varies greatly. Using th e ARRO Low Bed allowed data collection at the widest height range to encompass as ma ny clinical situations as possible without compromising the data by changi ng beds to accommodate vari ous bed heights. Bedrails were not physically implemented in this st udy because forcing the AT D over the bedrails would introduce additional accel eration that would otherwis e not be present during a gravity driven or passive fall from bed event. Since bedr ails increase the height from which patients fall, the effect of adding bedr ails will be extrapolat ed from data collected at various heights provid ed by the ARRO Low Bed.

PAGE 43

34 Figure 9: Carroll Healthcare ARRO Low Bed used to simulate common heights from which patients fall. A Posey Beveled Floor Cushion, referred to as floor mat or mat for convenience in this paper, was used to represent devices intended to cushion a patie nt if a fall occurs (See Figure 10). The floor ma t measured approximately 183 cm in length, 96.5cm in width, and 2.54 cm in thickne ss. The core of the mat wa s composed of ethylene vinyl acetate (EVA) foam to absorb impact energy, and a vinyl cover was used to provide an easily sterilized surface. This floor mat included a tri-fold design and carrying handle for easy storage and portability. As shown in Figure 10, the edges of the floor mat were beveled to aid wheelcha ir accessibility to the bedside wh en the mat was in use and to decrease the risk of tripping.

PAGE 44

35 Figure 10: Posey floor mat used to simulate mats commonly used in the healthcare environment to cushion falls. To simulate a patient falling from bed, the ATD was raised out of bed using a sling designed for this study (See Figure 11). The sling was attached to a ceiling lift that raised the ATD until gravity caused the ATD to slide out of bed. Th is provided a passive method of simulating a fall from bed, which represents a patie nt falling from bed as a result of position in bed whethe r from misjudgment or loss of balance while attempting to get out of bed. Using the sling not only stan dardized the falling pr ocess, but it also did not increase the level of acceler ation by allowing gravity to initi ate the fall. The positions of the ATD on the sli ng and the sling on the mattress were marked to standardize ATD placement, which was used to control the dir ection of the fall. The ATD was allowed to fall head first or feet first to simulate a more clin ically relevant fall event, as researchers postulate these to be the more common types of falls expe rienced by patients.

PAGE 45

36 Figure 11: Sling designed to standardi ze the fall from bed event simulation. Three 356A02 PCB acceleromet ers, as shown in Figure 12, were used to measure deceleration during an impact ev ent caused from a fall fr om bed. These accelerometers were designed to measure 500 g with an out put of 10 mV/g within a frequency range of 1 to 5000 Hz. The manufacturer calibrated eac h accelerometer individually and certified this calibration. However, because these instruments were interfaced with the data collection software LabVIEW, the calibrati on needed to be verified. This was accomplished using a PCB handheld shaker. The shaker excited each axis of the accelerometer individually at 1 g, and the output was cap tured using LabVIEW. The output ratio of mV/g controlled by LabVIEW wa s set to manufacturer calibration values. Each axis of the accelerom eter was again excited 1 g, and axis calibration was verified. LabVIEW was also used to ce nter the accelerometer output a bout 0 g and verified using the handheld shaker. Other methods of accel erometer calibration also exist and were attempted during the current st udy; however, the shaker meth od was ultimately used. Accelerometers were placed in the hea d, thorax, and pelvis to measure the deceleration at the most criti cal areas of the body. As shown in the literature, incidence of injury to these areas is high and often severe. By coll ecting data at these critical locations, any benefit provid ed by the floor mat or height adjustable bed will be appropriately assessed. Each accelerometer was bolted to an aluminum mounting block that fit inside one of the three cavities (See Figure 12). The use of a mounting stud was

PAGE 46

37 the most appropriate method for securing the accelerometer to the mounting block; however, this method did allow mechanical vibr ation to be introduced into the data. To isolate the true data signal from the mechanical noise introd uced through interactions at the bolt locations of the accelerometer and m ounting block, a digital filter was designed using MatLAB. The data signal with the m echanical noise was analyzed using a Fast Fourier Transform (FFT), wh ich breaks the signal down in to signal frequencies and magnitude. Based on this anal ysis, the true data was f ound below frequencies of 100 Hertz (Hz). This information was then used to build a 4th order Butterworth low pass filter with a cutoff frequency of 150 Hz. Th e MatLAB code used to analyze the data using the FFT and digital filter is available in Appendix A. Figure 12: Tri-axial acce lerometer aluminum mounting blocks for the head, thorax, and pelvis. All trials were also recorded using a video camera. The re cordings provided a visual verification of fall dir ection as well as a basis for de scribing the pro cess of falling from bed. Reviewing the recordings also prov ided possible explana tions for quantitative measurements and statistical analysis. This will be instrumental in furthering clinical understanding of fall from bed events, as few falls are actually observe d by caregivers.

PAGE 47

38 Protocol To standardize the data collection process, a data collection protocol was established and displayed at th e experimenter’s workstation. The data collect ion protocol is presented in Appendix B. The impact deceleration was measured at six different bed heights (33.5, 48, 62.5, 77, 91.5, and 97.5 cm) with and wi thout the floor mat beside the bed. These twelve different configurations were tested with the ATD falling from bed head first and feet first. A power analys is was conducted with = .05 and = .80; the number of trials needed to determine statistical significance wa s 2. During the data collection process the number of trials was increased to 6 to increas e the reliability of the measures as a wide standard deviation was observed. The data collection process began with arranging the bed, floor mat, and fall direction according to one of th e above stated configurations. Once all factors were set to the appropriate conditions, the video camera was set to record the fall event beginning with recording the trial numb er. The ATD was then raised from the bed via the sling, controlled manually by the ceili ng lift activated by the exper imenter. The data collection software LabVIEW was activated just prior to the ATD falling from the bed. Once a fall event was complete, the imp act deceleration was isolated and exported to a comma delimited file for further analys is. All factors were then re turned to initial positions and prepared for further trials. After all trials were comp lete, the data files were c onverted to Microsoft Excel files and analyzed using the MatLAB code pr eviously discussed. The peak deceleration of each trial for the head, thor ax, and pelvis was reported and used to calculate mean maximum values and standard deviations for each test config uration, where mean maximum value equals the mean of the peak deceleration values across similar test conditions. The mean maximum values were also used to calculat impact force for the head during head first falls a nd for the pelvis during feet fi rst falls according to Equation 2. A two-way ANOVA was performed by the statistical software package SAS using the mean and standard deviation calcula tions for each test configuration. The acceleration profile of the h ead acceleration for each trial was furt her analyzed to

PAGE 48

39 calculate the HIC value according to the formula previously discussed in Injury assessment The equation used to calculated HIC values was further applied to acceleration profiles measured for the thorax and pelvis; therefor e, thoracic inju ry criteria (TIC) and pelvic injury criteria (PIC) were computed. A two-way ANOVA was also performed on these values using SAS. The SAS code used to analyze the data in this study can be viewed in Appendix C. Force = mass acceleration Equation 2

PAGE 49

40 Results Head first falls A head first fall event gene rally began with the arm swi nging free of the bedside. As the arm was now no longer in the same plane as th e torso, a rotation about the longitudinal axis began. The head and upper to rso then began to slide from the bed. The events that followed occurred very quickly and almost simultaneously as the video recording revealed. Depending on the height and amount of rotati on that had occurred, the head impacted the ground la terally or anterior ly. The shoulder then impacted the floor, followed closely by the thorax impact. Because the ATD is still rotating throughout the fall event, th e shoulder opposite the bedside where the fall initiated was the shoulder that often impacted first; this was true of heig hts above 48 cm. The pelvis and lower limbs then impacted th e floor to complete the fall ev ent. At the completion of the fall event, the ATD landed in a prone position due to th e 180 degrees of rotation that occurred about the longi tudinal axis. Acceleration measured at the head According to the data coll ection protocol, 72 trials we re conducted and analyzed for head first falls. Of these falls, the mean peak impact decelerati ons measured in the head ranged from 18.60 10.89 g to 70.36 16.52 g when the various heights were measured without a mat and from 6.90 1.41 g to 21.51 7.10 g when measured with a mat. Mean values and standard deviations fo r each test configurati on for head first falls may be viewed in Table 4. As shown in Figure 13, the extreme m easurements of the mean decelerations did not always correspond to the extreme test configuration. For instance, the highest deceleration measured wh en a mat was not in use did not correspond to the extreme height of 97.5 cm rather to 91.5 cm. However, the ANOVA showed a significant increase in the me an impact decelerati ons with an increas e in height (p <

PAGE 50

41 .0001). Likewise, the mean impact decelerati ons measured when a mat was in use were found to be significantly lo wer than those measured wi thout a mat (p < .0001). Furthermore, the ANOVA showed a significant interaction effe ct between the mat and an increase in height (p = .0006). In other wo rds, the mat was more effective at lowering mean impact decelerations measured at the head as height increased. Table 4: Head mean impact decelerations measured with and without a mat during head first falls and calculated values based on trend line equations. Height cm Head g: No Mat (Mean SD) Head g: Mat (Mean SD) 33.5 34.50 15.42 9.22 5.56 48.0 47.69 25.65 12.69 12.74 62.5 18.60 10.89 6.90 1.41 77.0 44.19 15.80 10.70 2.94 91.5 70.36 16.52 12.26 4.19 97.5 64.02 25.33 21.51 7.10 112.5 69.19* 17.63** 115.0 70.47* 17.94** 117.5 71.75* 18.24** 120.0 73.03* 18.55** 122.5 74.31* 18.86** *These values are calculated valu es based on the trend line equation Acceleration = 0.512*(Height) + 11.543. **These values are calculated values based on the trend line equation Acceleration = 0.123*(Height) + 3.792.

PAGE 51

42 Head Deceleration during Head First Falls y = 0.512x + 11.543 R2 = 0.454 y = 0.123x + 3.792 R2 = 0.376 0 20 40 60 80 100 020406080100120140 Height cmDeceleration g Means +-SD No Mat Means +SD Mat No Mat Trend Mat Trend Figure 13: Head mean impact decelerations plotted with an estimated trend line during head first falls. Head injury criteria (HIC ) were also calcula ted for impact decelerations during head first falls measured with and without a mat. Similar to the mean impact decelerations, the extreme HIC values calcu lated for head first falls did not always correspond to the extreme test c onfigurations. As such the HI C value calculated for head first falls measured with no ma t ranged from 13.41 19.44 to 282.68 103.97. Furthermore, the mat appeared to provide some protective effe ct as the HIC values were generally lower when calculate d for trials conducted with a mat, as evidenced in the range of values calculated 1.33 0.48 to 10.01 7.83. ANOVA was also used to analyze the significance of the protec tive effect of the various heights and fl oor mat by comparing the HIC values calculated at th e different test co nditions. There was a significant effect of height on the HIC values. As the height increased, the HIC value significantly increased (p = .0017). Likewise the mat significantly lowe red the HIC value calculated for the same height (p <.0001). The mat was also shown to be more effective as height

PAGE 52

43 increased (p = .0026). Th e trend line equations used to estimate HIC values for heights added by bedrails resulted in a maximum va lue of 325.05 for falls on to the tile surface and 6.81 for falls onto the floor ma t. The HIC values calculated for all test configurations are shown below in Table 5 and graphi cally illustrated in Figure 14. Table 5: Mean HIC values calculated for head acce leration profiles measured during trials with and without a mat during head first falls. Height cm HIC: No Mat (Mean SD) HIC: Mat (Mean SD) 33.5 38.70 30.61 3.35 4.11 48.0 116.59 141.19 9.36 16.16 62.5 13.41 19.44 1.33 0.48 77.0 84.92 76.48 3.21 1.74 91.5 282.68 103.97 3.82 2.08 97.5 260.49 264.30 10.01 7.83 112.5 289.56* 6.51** 115.0 298.43* 6.58** 117.5 307.30* 6.66** 120.0 316.17* 6.73** 122.5 325.05* 6.81** *These values are calculated valu es based on the trend line equation HIC = 3.549*(Height) 109.728. **These values are calculated values based on the trend line equation HIC = 0.030*(Height) + 3.133

PAGE 53

44 HIC during Head First Falls y = 3.549x 109.728 R2 = 0.611 y = 0.030x + 3.133 R2 = 0.043 0 100 200 300 400 500 600 020406080100120140 Height cmHIC Means +SD No Mat Means +SD Mat No Mat Trend Mat Trend Figure 14: HIC values plotted with an es timated trend line during head first falls. Force calculations were also performed for the first impact during head first falls. The head impacted the surface first and momentarily support ed all the body weight of 76.3 kg. As such the force calcu lations were based on this we ight. The force reached a maximum value 5368.16 N at a height of 91.5 cm when impact occurred onto the tile floor. When the floor ma t was used, the maximum va lue was 1641.12 N as shown in Table 6. These values were derived from Equa tion 2; therefore, no statistical analysis was performed. The forces calculated for head first falls with and wi thout a floor mat are graphically represented in Figure 15.

PAGE 54

45Table 6: Impact forces ca lculated at the head for head first falls with and with out a floor mat. Height cm Force N: No Mat Force N: Mat 33.5 2631.99 703.13 48.0 3638.53 968.02 62.5 1419.24 526.79 77.0 3371.51 816.68 91.5 5368.16 935.67 97.5 4884.87 1641.12 Calculated Impact Force for Head First Falls y = 39.097x + 880.733 R2 = 0.454 y = 9.404x + 289.293 R2 = 0.376 0 1000 2000 3000 4000 5000 6000 020406080100120 Height cmForce N No Mat Mat No Mat Trend Mat Trend Figure 15: Impact forces plotted with an es timated trend line dur ing head first falls. Acceleration measured at the thorax Impact deceleration was also measured at the thorax dur ing the 72 head first falls. The mean peak impact decelerations measured in the thorax without a mat ranged from 16.85 6.97 g to 48.50 25.54 g; the thoracic imp act deceleration measured with a mat ranged from 6.61 3.98 g to 46.67 47.78 g. Mean values and standard deviations for

PAGE 55

46 each test configuration for head first fall s may be viewed in Table 7. The impact decelerations measured with a mat were lower than thos e measured without a mat for all heights except 77 cm, as shown in Figure 16. As a result, the presence of the mat was found to have no significant effect upon the im pact decelerations (p = .1639). However, the impact decelerations were found to increase significantly with an increase in height (p = .0052). Table 7: Thoracic mean impact decelerations measured with and without a mat during head first falls and calculated values based on trend line equations. Height cm Thorax g: No Mat (Mean SD) Thorax g: Mat (Mean SD) 33.5 16.85 6.97 6.61 3.98 48.0 28.55 15.79 20.79 19.09 62.5 29.48 20.15 13.48 3.94 77.0 30.82 8.63 46.67 47.78 91.5 36.14 9.55 13.87 2.80 97.5 48.50 25.54 43.11 41.31 112.5 47.11* 39.33** 115.0 47.93* 40.44** 117.5 48.75* 41.55** 120.0 49.56* 42.67** 122.5 50.37* 43.80** *These values are calculated valu es based on the trend line equation Acceleration = 1.148*(Height 0.787). **These values are calculated values based on the trend line equation Acceleration = 0.100*(Height 1.265).

PAGE 56

47 Thorax Deceleration during Head First Falls y = 0.100x1.265R2 = 0.478 y = 1.148x0.786R2 = 0.866 0 20 40 60 80 100 020406080100120140 Height cmDeceleration g Means +SD No Mat Means +SD Mat Mat Trend No Mat Trend Figure 16: Thoracic mean impact decelerations plotte d with an estimated tre nd line during head first falls. Injury criteria were also calculated for impact decel erations measured at the thorax during head first falls The thoracic in jury criteria (TIC) ranged from 23.45 31.61 to 220.85 274.74 when calculated for imp acts onto the tile surface. The TIC calculated for impacts onto the floor mat ranged from 1.58 2.04 to 525.12 824.83. Interestingly, the extreme TIC values calculated for impacts onto the tile surface corresponded to the extreme bed heights; how ever, the TIC values calculated for impacts onto the floor mat did no t follow a similar pattern. Imp acts occurring from a height of 77 cm onto a floor mat resulted in the highest TIC values, whereas impacts occurring from a height of 33.5 cm onto a floor mat resulted in the lowest TI C values. As shown in Table 8, the TIC values did not consistently decrease with the use of the fl oor mat, as expected, nor was there a consistent incr ease due to height. As su ch, the TIC valu es were not significantly increased with an increase in height (p = .1180) or signifi cantly decreased with the use of the fl oor mat (p = .4286). Trend line equations were used to estimate TIC values for heights added by bedrails a nd resulted in a maximum value of 259.83 for

PAGE 57

48 falls onto the tile surface and 236.45 for falls on to the floor mat. Figure 17 illustrates these values graphically. Table 8: Mean TIC values calculated for thoracic a cceleration profiles measured during trials with and without a mat during head first falls. Height cm TIC: No Mat (Mean SD) TIC: Mat (Mean SD) 33.5 23.45 31.61 1.58 2.04 48.0 60.24 58.58 68.51 139.95 62.5 89.89 152.01 6.37 5.92 77.0 37.49 5.26 525.12 824.83 91.5 102.09 32.50 8.23 4.26 97.5 220.85 274.74 247.28 350.52 112.5 203.37* 301.14** 115.0 216.22* 322.79** 117.5 229.87* 345.48** 120.0 244.39* 369.23** 122.5 259.83* 394.08** *These values are calculated valu es based on the trend line equation TIC = 12.920*e (0.0245*Height). **These values are calculated values based on the trend line equation TIC = 0.0001*(Height 3.1586).

PAGE 58

49 TIC during Head First Falls y = 0.0001x3.1586R2 = 0.3223 y = 12.920e0.025xR2 = 0.599 0 200 400 600 800 1000 1200 1400 020406080100120140 Height cmTIC Means +SD No Mat Means +SD Mat Mat Trend No Mat Trend Figure 17: TIC values plotted with an es timated trend line during head first falls. Accelerations measured at the pelvis During the head first falls, impact deceleration was also measured at the pelvis. When measured without a mat, the pelvic impact deceleration ranged from 11.46 7.54 g to 20.74 3.85 g. The pelvic impact deceler ation measured with a mat ranged from 8.48 4.03 g to 20.52 10.99 g. Table 9 includes mean s and standard deviations for all test configurations, and Figure 18 illustrates thes e values. The mean impact decelerations were significantly higher with an increase in height (p = .039 7). Likewise, the use of the mat significantly lowered the mean impac t decelerations p = .0224). However, the effectiveness of the ma t did not change with a change in height.

PAGE 59

50Table 9: Pelvic mean impact decelerations measured with and without a mat during head first falls and calculated values based on trend line equations. Height cm Pelvis g: No Mat (Mean SD) Pelvis g: Mat (Mean SD) 33.5 14.29 8.25 8.48 4.03 48.0 11.46 7.54 13.97 7.45 62.5 20.74 3.85 11.31 6.82 77.0 19.82 10.67 12.44 4.16 91.5 18.06 5.91 20.52 10.99 97.5 20.63 7.33 16.32 9.94 112.5 21.31* 18.78** 115.0 21.45* 19.03** 117.5 21.60* 19.28** 120.0 21.74* 19.52** 122.5 21.88* 19.77** *These values are calculated valu es based on the trend line equation Acceleration = 6.754*Ln (Height) 10.592. **These values are calculated values based on the trend line equation Acceleration = 1.073*(Height 0.606).

PAGE 60

51 Pelvis Deceleration during Head First Falls y = 6.754Ln(x) 10.592 R2 = 0.527 y = 1.073x0.606R2 = 0.664 0 5 10 15 20 25 30 35 020406080100120140 Height cmDeceleration g Means +SD No Mat Means +SD Mat No Mat Trend Mat Trend Figure 18: Pelvic mean impact decelerations plotted with an estimated trend line during head first falls. Pelvic injury criteria (PIC) were also calculated for impact decelerations measured during head first fa lls. The PIC values calculate d for impacts onto the tile surface ranged from 7.52 8.46 to 21.69 13.78, as shown in Table 10. The PIC values calculated for impacts onto the floor mat ranged from 2.14 1.74 to 17.05 13.56. Similar to the mean maximum values, th e highest PIC value calculated did not correspond to the extreme trial condition of 97.5 cm with no fl oor mat, rather to 77 cm with no floor mat. As such, th e PIC values did not significantly increase with an increase in height (p = .2245). Furthermore, imp acts onto the floor mat did not result in significantly lower PIC values than those onto the tile floor (p = .0930). A maximum value of 21.25 was est imated for falls onto th e tile surface from heig hts added by bedrails and 23.03 for falls onto the floor mat when a trend line was fitt ed to the measured data. Figure 19 graphically illustrates the PIC values calculated for head first falls.

PAGE 61

52Table 10: Mean PIC values calculat ed for pelvic deceleration profil es measured during trials with and without a mat during head first falls. Height cm PIC: No Mat (Mean SD) PIC: Mat (Mean SD) 33.5 11.21 14.41 2.14 1.74 48.0 7.52 8.46 10.98 15.07 62.5 17.10 10.22 12.70 12.01 77.0 21.69 13.78 9.08 5.76 91.5 16.76 12.48 17.05 13.56 97.5 17.90 9.56 11.92 16.58 112.5 20.48* 20.37** 115.0 20.68* 21.03** 117.5 20.88* 21.69** 120.0 21.07* 22.36** 122.5 21.25* 23.03** *These values are calculated valu es based on the trend line equation PIC = 9.082*Ln(Height) 22.414. **These values are calculated values based on the trend line equation PIC = 0.023 *(Height 1.440).

PAGE 62

53 PIC during Head First Falls y = 9.082Ln(x) 22.414 R2 = 0.533 y = 0.023x1.440R2 = 0.645 0 5 10 15 20 25 30 35 40 020406080100120140 Height cmPIC Means +SD No Mat Means +SD Mat No Mat Trend Mat Trend Figure 19: PIC values plotted with an es timated trend line during head first falls. Feet first falls As the description indicates feet first falls were initiated with the feet sliding from the bed. At heights a bove 48 cm, the lower limb s slide from the bed in a nearly straight position. However, the feet a nd knees initiated a “crumple” effect at lower heights as these areas were in contact with the floor fo r long periods of time before torso impact occurred. Similar to head first falls, ro tation occurred about the longitudinal axis; however, the dummy completed 180 degrees of rotation before impacting the floor. Consequently, when the pelvis im pacted the floor, im pact occurred either laterally at the hip farther from the bedside where the fall initiated or on the entir e posterior portion of the pelvis. The rotation contin ued as the impact events occu rred, and the thorax impacted fully in the posterior position regardless of height. The fall terminated with the head impacting the floor in the pos terior position. As such, th e dummy landed in a supine position at the end of the fall event. Furthermore, the du mmy was observed to come to rest at greater distan ces away from the bed as height increased. With an increase in

PAGE 63

54 height, the ATD’s final positi on on the floor moved passed the top of the bed and laterally away from th e side of the bed. Acceleration measured at the head Although 72 trials were conducted with th e fall direction as feet first, impact deceleration at the head wa s only measured for 54 of those trials; accelerometer frequency limitations prevented data collec tion above heights of 62.5 cm when measured without a mat. Since the accelerometer had a defined frequency range, any frequency measured by the accelerometer exceeding that range caused th e instrument to shut down to prevent damage to the el ectronics. The mechanical vi brations between the mounting block and accelerometer may have caused the excessiv e frequency readings. Consequently, the impact decelerations measured without a mat ranged from 74.13 58.41 g to 152.47 46.12 g. Those measured with a mat included all heights and ranged from 8.48 6.66 g to 91.58 47.24 g. Table 11 displa ys all mean and standard deviation values for head impact deceler ation measured during feet firs t falls. The data measured at heights 33.5, 48, and 62.5 cm were th e only heights included in the ANOVA. Nonetheless, a change in height was s hown to increase the impact deceleration significantly (p = .0004). Furthermore, the mat significantly lowered the impact deceleration (p < .0001); however, this effect was not dependent upon a change in height, as shown in Figure 20.

PAGE 64

55Table 11: Head mean impact decelerations measu red with and without a mat during feet first falls and calculated values based on trend line equations. Height cm Head g: No Mat (Mean SD) Head g: Mat (Mean SD) 33.5 74.13 58.41 8.48 6.66 48.0 152.47 46.12 41.51 22.40 62.5 131.81 31.07 75.93 26.38 77.0 187.86* 91.58 47.24 91.5 222.71* 66.41 38.37 97.5 237.11* 54.52 27.73 112.5 273.06* 118.03** 115.0 279.04* 122.56** 117.5 285.02* 127.16** 120.0 291.00* 131.83** 122.5 296.98* 136.57** *These values are calculated valu es based on the trend line equation Acceleration = 2.589*(Height 0.986). **These values are calculated values based on the trend line equation Acceleration = 0.036* (Height 1.714).

PAGE 65

56 Head Deceleration during Feet First Falls y = 0.036x1.714R2 = 0.658 y = 2.589x0.986R2 = 0.655 0 50 100 150 200 250 300 020406080100120140 Height cmDeceleration g Means +SD No Mat Means +SD Mat Mat Trend No Mat Trend Figure 20: Head mean impact decelerations plotted with an estimated trend line during feet first falls. Head injury criteria (HIC) were also calculated for feet firs t falls for the trials with available data. Mean HIC values calculated for trials conducted without a mat ranged from 486.51 880.08 to 1234.63 945.72. Trials conduct ed with a mat resulted in lower mean HIC values and ranged from 3.96 5.37 to 374.35 389.04. All means and standard deviations of calcul ated HIC values are presented in Table 12 and illustrated in Figure 21. ANOVA was also used to determine if changing the height or removing the mat had a significant effect upon HI C values. From this analys is, a change in height was determined to have no signi ficant effect upon the HIC valu es (p = .2136). However, removing the mat significan tly increased the HIC valu es (p = .0006). Trend line equations were used to estimat e the HIC values for heights a dded by bedrails and resulted in a maximum value of 1468.59 for falls onto the tile surface and 737.35 for falls onto the floor mat.

PAGE 66

57Table 12: Mean HIC values calculated for head deceleration profiles measured during trials with and without a mat during feet first falls. Height cm HIC: No Mat (Mean SD) HIC: Mat (Mean SD) 33.5 486.51 880.08 3.96 5.37 48.0 1234.63 945.72 73.92 65.03 62.5 697.79 321.44 212.26 151.19 77.0 1063.05* 374.35 389.04 91.5 1198.69* 196.82 204.70 97.5 1252.87* 133.75 100.85 112.5 1384.08* 557.23** 115.0 1405.41* 599.00** 117.5 1426.61* 642.91** 120.0 1447.66* 689.00** 122.5 1468.59* 737.35** *These values are calculated valu es based on the trend line equation HIC = 51.708*(Height 0.696). **These values are calculated values based on the trend line equation HIC = 0.0001*(Height 3.289).

PAGE 67

58 HIC Values during Feet First Falls y = 51.708x0.696R2 = 0.215 y = 0.0001x3.2889R2 = 0.6832 0 500 1000 1500 2000 2500 020406080100120140 Height cmHIC Means +SD No Mat Means +SD Mat No Mat Trend Mat Trend Figure 21: HIC values plotted with an esti mated trend line during feet first falls. Acceleration measured at the thorax Due to accelerometer limitations, data was collected for all heights except 97.5 cm. For the remaining heights, the mean im pact decelerations meas ured at the thorax during feet first falls ra nged from 8.47 5.16 g to 95.12 43.13 g without a mat and from 3.29 0.61 g to 58.25 46.01 g with a mat. All means and standard deviations are presented in Table 13 and graphically illustrate d in Figure 22. Analysis of the available data showed a significant increa se in impact deceler ation with an increas e in height (p < .0001). Also, mean impact decelerations sign ificantly increased when the mat was not used (p < .0001). The mean impact d ecelerations measured during mat usage was influenced by a change in hei ght, as an increase in height in creased the effect of the mat (p = .0103).

PAGE 68

59Table 13: Thoracic mean impact decelerations me asured with and without a mat during feet first falls and calculated values based on trend line equations. Height cm Thorax g: No Mat (Mean SD) Thorax g: Mat (Mean SD) 33.5 8.47 5.16 3.29 0.61 48.0 17.80 6.04 7.15 1.58 62.5 43.06 16.00 20.26 10.17 77.0 50.23 13.64 38.30 28.26 91.5 95.12 43.13 38.06 16.24 97.5 105.54* 58.25 46.01 112.5 148.02* 99.54** 115.0 155.91* 105.61** 117.5 164.04* 111.90** 120.0 172.41* 118.43** 122.5 181.02* 125.18** *These values are calculated valu es based on the trend line equation Acceleration = y = 0.002*(Height 2.364). **These values are calculated values based on the trend line equation Acceleration = 0.0003*(Height 2.2916).

PAGE 69

60 Thorax Deceleration during Feet First Falls y = 0.002x2.364R2 = 0.981 y = 0.0003x2.6916R2 = 0.9735 0 30 60 90 120 150 180 020406080100120140 Height cmDeceleration g Means +SD No Mat Means +SD Mat No Mat Trend Mat Trend Figure 22: Thoracic mean impact decelerations plotte d with an estimated trend line during feet first falls. Injury criteria were cal culated from the thoracic accel eration profiles measured for feet first falls. These values are shown for impacts wi th and without a floor mat in Table 14 and illustrated in Figure 23. The thoracic injury criteria (TIC) reached a maximum of 1103.40 733.32 at a height of 91.5 cm a nd a minimum of 1.45 1.16 when measured without a fl oor mat. The maximum TIC value calculated for impacts onto the floor mat was 3 79.02 456.90 and a minimum of 0.17 0.05. These values increased significantly with an increase in height (p < .0001) and decr eased significantly with the use of a floo r mat (p = .0007). Furthermore, th e mat more effectively decreased the TIC values at high er heights (p = .0002). Trend lines were also used to project TIC values that may result with the addition of be drails. These values reached a maximum of 8053.09 without a floor mat and 1995.77 wh en a floor mat was utilized.

PAGE 70

61Table 14: Mean TIC values calculated for thoracic deceleration profiles meas ured during trials with and without a mat during feet first falls. Height cm TIC: No Mat (Mean SD) TIC: Mat (Mean SD) 33.5 1.45 1.16 0.17 0.05 48.0 13.69 8.84 1.38 0.76 62.5 150.50 151.35 39.54 30.86 77.0 298.17 368.83 141.69 194.83 91.5 1103.40 733.32 111.20 131.19 97.5 1776.46* 379.02 456.90 112.5 4582.18* 1082.58** 115.0 5300.02* 1267.72** 117.5 6111.15* 1479.49** 120.0 7025.33* 1721.03** 122.5 8053.09* 1995.77** *These values are calculated valu es based on the trend line equation TIC = 1E-10*(Height 6.6216). **These values are calculated values based on the trend line equation TIC = 2E-12*(Height .7.183).

PAGE 71

62 TIC during Feet First Falls y = 1E-10x6.6216R2 = 0.9889 y = 2E-12x7.183R2 = 0.9512 0 1000 2000 3000 4000 5000 6000 7000 8000 020406080100120140 Height cmTIC Means +SD No Mat Means +SD Mat No Mat Trend Mat Trend Figure 23: TIC values plotted with an es timated trend line during feet first falls. Acceleration measured at the pelvis Although data collected at the head and thorax was li mited by the accelerometers, impact deceleration was measur ed at the pelvis for all heights. The mean impact decelerations measured without a mat ranged from 4.84 1.49 g to 36.97 21.52 g, while those measured with a mat ranged fro m 5.13 3.13 g to 24.53 8.02 g. Table 15 shows means and standard deviations measured for al l trials conducted with and without a mat. Changing the height si gnificantly increased the mean impact dece lerations (p < .0001); however, the presence of the mat had no si gnificant effect on the mean impact decelerations (p = .0589), although Figure 24 shows the mean impact decelerations measured with a mat to be lower th an those measured without a mat.

PAGE 72

63Table 15: Pelvic mean impact decelerations measu red with and without a mat during feet first falls and calculated values based on trend line equations. Height cm Pelvis g: No Mat (Mean SD) Pelvis g: Mat (Mean SD) 33.5 4.84 1.49 9.97 8.90 48.0 8.18 3.21 5.13 3.13 62.5 8.88 3.55 8.05 3.57 77.0 23.46 10.75 18.17 7.41 91.5 32.28 23.19 19.28 4.52 97.5 36.97 21.52 24.53 8.02 112.5 62.72* 25.81** 115.0 68.09* 26.47** 117.5 73.93* 27.13** 120.0 80.27* 27.79** 122.5 87.15* 28.44** *These values are calculated valu es based on the trend line equation Acceleration = 1.549 e (0.033*Height). **These values are calculated values based on the trend line equation Acceleration = 0.263*Height – 3.774.

PAGE 73

64 Pelvis Deceleration during Feet First Falls y = 1.549e0.033xR2 = 0.960 y = 0.263x 3.774 R2 = 0.754 0 20 40 60 80 100 020406080100120140 Height cmDecleration g Means +SD No Mat Means +SD Mat No Mat Trend Mat Trend Figure 24: Pelvic mean impact decelerations plotted with an estimated trend line during feet first falls. Pelvic injury criteri a were calculated also from the me asured acceleration profiles. The PIC values ranged fr om 0.43 0.24 to 54.46 54.98 when calculated for falls onto the tile surface and from 1.03 1.66 to 24.30 12.38 when calculated for falls onto the floor mat. Statistical analysis of these valu es showed the PIC values neither to decrease significantly with the use of a floor mat (p = .0930) or increase signifi cantly with an increase in height (p = .2245) Trend line equations were also used to estimate PIC values at heights added by bedrails. Th ese values reached a maximum of 146.87 without a floor mat and 28.02 with a floor mat. A gr aphical representation of the PIC values calculated for impacts onto the tile surface and floor mat can be viewed in Figure 25.

PAGE 74

65Table 16: Mean PIC values calculated for pelvic deceleration profil es measured during trials with and without a mat during feet first falls. Height cm PIC: No Mat (Mean SD) PIC: Mat (Mean SD) 33.5 0.43 0.24 5.68 8.32 48.0 5.14 9.29 1.03 1.66 62.5 2.74 2.52 4.17 5.85 77.0 20.77 16.79 16.65 22.00 91.5 53.77 40.98 15.48 8.89 97.5 54.46 54.98 24.30 12.38 112.5 100.97* 24.91** 115.0 111.22* 25.69** 117.5 122.26* 26.47** 120.0 134.13* 27.24** 122.5 146.87* 28.02** *These values are calculated valu es based on the trend line equation PIC = 1E-07*(Height 4.4007). **These values are calculated values based on the trend line equation PIC = (0.310*Height) – 9.972.

PAGE 75

66 PIC during Feet First Falls y = 0.310x 9.971 R2 = 0.745 y = 1E-07x4.4007R2 = 0.9041 0 40 80 120 160 200 020406080100120140 Height cmPIC Means +SD No Mat Means +SD Mat Mat Trend No Mat Trend Figure 25: PIC values plotted with an es timated trend line during feet first falls. During feet first falls, the pelvis impact ed the surface before the thorax or head; therefore, force calculations we re performed for the pelvis not the head. Equation 2 was used to calculate force based on the mean maximum accelerati ons measured for feet first falls at the pelvis and the to tal body weight minus the weight of the lower extremities. When calculated for impacts onto the tile surface, the ma ximum force was 1974.55 N at a height of 97.5 cm, as shown in Table 17. The maximum force decreased to 1310.10 N at a height of 97.5 cm when the floor mat was us ed. As mentioned previously, no statistical analysis was performed on these calcula tions. See Figure 26 for a graphical representation for these calculations. A summa ry of all measured mean maximum values and calculated injury criteria for head and feet first fall s is presented in Table 17.

PAGE 76

67Table 17: Impact forces calculated at the pelvis fo r feet first falls with and without a floor mat. Height cm Force N: No Mat Force N: Mat 33.5 258.73 532.60 48.0 437.06 273.73 62.5 474.54 430.06 77.0 1252.80 970.41 91.5 1724.29 1029.64 97.5 1974.55 1310.10 Calculated Impact Force for Feet First Fallsy = 14.039x 201.563 R2 = 0.754 y = 82.708e0.033xR2 = 0.960 0 250 500 750 1000 1250 1500 1750 2000 020406080100120 Height cmForce N No Mat Mat Mat Trend No Mat Figure 26: Impact forces plotted with an esti mated trend line during feet first falls.

PAGE 77

68 Table 18: Summary of measured mean maximum values and calculated injury criteria. Head First Falls Head Means SD (g) Injury Criteria Height (cm) No Mat Mat No Mat Mat 33.5 34.50 15.42 9.22 5.56 38.70 30.61 3.35 4.11 48 47.69 25.65 12.69 12.74 116.59 141.19 9.36 16.16 62.5 18.60 10.89 6.90 1.41 13.41 19.44 1.33 0.48 77 44.19 15.80 10.70 2.94 84.92 76.48 3.21 1.74 91.5 70.36 16.52 12.26 4.19 282.68 103.97 3.82 2.08 97.5 64.02 25.33 21.51 7.10 260.49 264.30 10.01 7.83 Thorax Means SD (g) Injury Criteria Height (cm) No Mat Mat No Mat Mat 33.5 16.85 6.97 6.61 3.98 23.45 31.61 1.58 2.04 48 28.55 15.79 20.79 19.09 60.24 58.58 68.51 139.95 62.5 29.48 20.15 13.48 3.94 89.89 152.01 6.37 5.92 77 30.82 8.63 46.67 47.78 37.49 5.26 525.12 824.83 91.5 36.14 9.55 13.87 2.80 102.09 32.50 8.23 4.26 97.5 48.50 25.54 43.11 41.31 220.85 274.74 247.28 350.52 Pelvis Means SD (g) Injury Criteria Height (cm) No Mat Mat No Mat Mat 33.5 14.29 8.25 8.48 4.03 11.21 14.41 2.14 1.74 48 11.46 7.54 13.97 7.45 7.52 8.46 10.98 15.07 62.5 20.74 3.85 11.31 6.82 17.10 10.22 12.70 12.01 77 19.82 10.67 12.44 4.16 21.69 13.78 9.08 5.76 91.5 18.06 5.91 20.52 10.99 16.76 12.48 17.05 13.56 97.5 20.63 7.33 16.32 9.94 17.90 9.56 11.92 16.58 Feet First Falls Head Means SD (g) Injury Criteria Height (cm) No Mat Mat No Mat Mat 33.5 74.13 58.41 8.48 6.66 486.51 880.08 3.96 5.37 48 152.47 46.12 41.51 22.40 1234.63 945.72 73.92 65.03 62.5 131.81 31.07 75.93 26.38 697.79 321.44 212.26 151.19 77 N/A 91.58 47.24 N/A 374.35 389.04 91.5 N/A 66.41 38.37 N/A 196.82 204.70 97.5 N/A 54.52 27.73 N/A 133.75 100.85 Thorax Means SD (g) Injury Criteria Height (cm) No Mat Mat No Mat Mat 33.5 8.47 5.16 3.29 0.61 1.45 1.16 0.17 0.05 48 17.80 6.04 7.15 1.58 13.69 8.84 1.38 0.76 62.5 43.06 16.00 20.26 10.17 150.50 151.35 39.54 30.86 77 50.23 13.64 38.30 28.26 298.17 368.83 141.69 194.83 91.5 95.12 43.13 38.06 16.24 1103.40 733.32 111.20 131.19 97.5 N/A 58.25 46.01 N/A 379.02 456.90

PAGE 78

69 Table 18: Continued Feet First Falls Continued Pelvis Means SD (g) Injury Criteria Height (cm) No Mat Mat No Mat Mat 33.5 4.84 1.49 9.97 8.90 0.43 0.24 5.68 8.32 48 8.18 3.21 5.13 3.13 5.14 9.29 1.03 1.66 62.5 8.88 3.55 8.05 3.57 2.74 2.52 4.17 5.85 77 23.46 10.75 18.17 7.41 20.77 16.79 16.65 22.00 91.5 32.28 23.19 19.28 4.52 53.77 40.98 15.48 8.89 97.5 36.97 21.52 24.53 8.02 54.46 54.98 24.30 12.38

PAGE 79

70 Discussion and Interpretation Introduction During a fall from bed event investigated in this study, the resulting mean maximum values and injury criteria recorded at the head, thorax, and pelvis were dependent upon the height from which the fall occurred and th e presence or absence of a floor mat. Statistically these factors affect ed each body region differ ently as the direction of impact changed from head first to feet first falls. The varia tions reported in the statistical results can be explained by th e amount of bounce or rebound off the impact surface and the presence of bending between body regions. Head first falls The order of impact was observed visually to be head, thorax, and pelvis for head first falls and confirmed by th e acceleration profiles recorded in LabView. The mean maximum values recorded at th e head, thorax, and pelvis a ll increased significantly as height increased, as expected. However, the mat did not significantly decrease the mean maximum values at all body re gions. As discussed previous ly, the ATD rotated about the longitudinal axis during a fall from bed event cau sing the thorax to impact the shoulder rather than the anterior or pos terior portion of the region. Th e lateral impact of the thorax decreased the amount of surface area available to support the weight; therefore, the mat was unable to adequately cushion the fa ll and decrease the mean maximum values measured at the thorax. Furt hermore, the tile surface defo rmed less during a fall event than did the mat; ther efore, the ATD bounced or rebounded off the tile duri ng a fall. By decreasing the amount of bounce the mat increased the amount of ti me each body region was in contact with the su rface to reduce significantly the mean maximum values recorded at the head and pelvis.

PAGE 80

71 The injury criteria calculated for the hea d, pelvis, and thorax were also dependent upon the amount of bounce and be nd that occurred during a fa ll event. As the mean maximum values measured at the head were significantly affect ed by height and the presence of the mat, the HIC values were, likew ise, expectedly affect ed by height and the presence of the mat. However, the thorax and pelvis did not produce similar results. The injury criteria calculated for the thorax or pelvis were not af fected by height as were the mean maximum values. The TIC value is calculated based on the integral of acceleration with limits up to a 15 msec time period. For the TIC values to rema in unaffected by the significant increase in mean maximum values there must be a corr esponding decrease in sustained acceleration. During im pact the head and pelvis were allowed to pivot with respect to the thorax, thus pus hing the thorax into the impact surface. As this occurred, the thorax experienced limited rebound off the surface; therefore, th e thorax d ecelerated more quickly with the sustaine d contact with the surface. Li kewise, the PIC values were not significantly incr eased as height increased for the same reason; the pelvis was allowed to pivot with respect to the thorax and re main in contact with the surface. The knees further pushed the pelvis into contact with the surface. Th e observed increase in mean maximum values measured at the pelvis was balanced by the corresponding decrease in sustaine d acceleration caused by the sust ained surface contact; hence, no height effect was observed to increase significantly the PI C values. Furthermore, the presence of the mat did not significantly d ecrease the PIC values, because the mat allowed the pelvis to continue to acceler ate into the surface thereby increasing the sustained acceleration. The resulting mean maximum values and injury criteria recorded for head first falls were compared with injury preventi on standards often used in the automotive industry. As discussed previously, the head injury criterion (HIC) was developed to provide a value to co rrelate accelerat ion with injury severity. According to these standards, HIC values must not exceed 700 wh en calculated over limits up to a 15 ms time period. The HIC valu es calculated for head first falls onto the tile surface reached a maximum of 282.68 103.97 at 91.5 cm. When the height increased to 122.5 cm to account for bedrails, the HIC value increased to 325.05. As shown in Figure 7, these HIC

PAGE 81

72 values correlated with a less than one percen t chance of experiencing a serious injury as a result of falling out of bed, wher e serious is defined as an AIS score greater than or equal to 3. The use of a mat signifi cantly reduced the HIC values; th erefore, the risk of injury to the head was further reduced. Although Lyons and Oates (1993) and Macgregor (2000) reported head trauma to be specifically associated with falling from bed, these reports were documented inciden ces for children, not adults. As the ATD is designed to mimic a 50th percentile male, direct appl ication of these head trau ma reports is not valid. Skull fractures have, however, been associat ed with falls in general but the specific circumstances causing those inju ries are not reflected in th e literature. The literature supports this finding as severe brain injury specifically re sulting from falling out of bed has not been documented in adults, even t hough head trauma resul ting from falling from bed was documented in children and skull fractures have been recorded as being associated with falls (Lyons and Oates, 1993; Sattin et al, 1990). However, the forces calculated during head first falls may indicate a risk of sustaining a skull fracture as the literature reports a range of values inclusive of the for ces calculated in this study (Nahum, Gatts, Gadd, and Danforth, 1968; Schneider and Nahum, 1972). Sterling, O’Conner, and Bonadies (2001) fu rther noted neck injuries to be associated with falls; however, neither the amount of bending nor acce leration was measured at the neck during this study, although both were visu ally observed to occur subseq uent to head impact. Currently, no injury criteri on and industry standard exists specific to the thorax or pelvis. However, the automotive industry ha s set forth a limit of 60 g for chest (thoracic) acceleration. The mean maximum values measur ed at the thorax dur ing head first falls reached a maximum of 48.50 25.54 g onto th e tile surface and 46.67 47.78 g onto the floor mat. These values did not exceed the industry standa rd; however, no inference can be made with regard to inju ry severity either from these values or the calculated TIC values. Likewise, the pelvic in jury criteria calculated in this study cannot, currently, infer injury severity, but as more research becomes available thes e values may prove useful in assessing injury severity associated with falling from bed. Although no inference can be made concerning thoracic or pelvic injury based upon measured or calculated values, visual observation of impact site may indicate a higher risk of injury. Nevitt et al (1993)

PAGE 82

73 concluded that fall direction affected the location of impa ct and injury to specific anatomical regions. During head first falls rotation occurred about the longitudinal axis and resulted in lateral impaction upon the s houlder and pelvis. As such, the risk of clavical or scapula and hip fractures may be increased as these areas have decreased soft tissue to cushion such an impact. Furthermore, dislocations have been documented to be associated with falls, hence the lateral impacts observed during head first falls may account for this associat ion (Sattin et al 1990). Feet first falls Changing the fall direction from head first to feet firs t consequently changed the impact order to pelvis, thorax, and head wh ich was observed visually and confirmed with the acceleration prof iles recorded in LabView. As with head first falls, the mean maximum values were dependent upon the amount of bounce off the surface. The mean maximum values measured at th e head, thorax, and pelvis were all significantly increased with increasing height as expected. However, the mat did not significantly reduce mean maximum values for all body regi ons or for the same body regions as head first falls. This apparent discrepancy re sulted from the differing fall mechanics and orientations. During feet first falls, the ATD completed nearly 360 degrees of rotation about the longitudinal axis before impacting the surface. As such, the thorax impacted a greater percentage of the posterior surface than with the lateral impact observed in head first falls. This increase in impact surface area allowed the mat to support more of the thoracic weight directly; th erefore, the mean maximum values were significantly decreased by the mat. Alt hough the mean maximum values me asured at the pelvis during head first falls were significa ntly decreased by the mat, the prior impact of the feet and legs during feet first falls ab sorbed much of the impact that would otherwise have been supported by the pelvis to create a “crumple” effect. The dependence on height a nd mat of the thor acic injury criteria and dependence on the presence of the mat of the head injury criteria was ex pected as the mean maximum values were found to be si milarly affected. However, the HIC values were not significantly increased with an increase in height because the head was able to bend and

PAGE 83

74 remain in contact with the impact surface. This sustai ned contact resulted in a corresponding decrease in sust ained acceleration; therefore, the significant increase observed with the mean maximum values wa s balanced by the decrease in sustained acceleration. As with head first falls, the pe lvis injury criteria cal culated for feet first falls was not dependent upon height or mat even though the mean maximum values were found to be so. However, as discussed previously, the injury criteria can remain unaffected by a significant increase in mean maxi mum values by a corresponding decrease in sustained accelerati on. The mat allowed the pelvis to continue to accelerate into the surface; therefore, the acceleration was sustained lo nger than that onto the tile surface. As a result, the mat did not significantly decrease the PIC values calculated for feet first falls. Similar to head first falls the mean maximum values and injury criteria calculated during feet first falls were co mpared to standards utilized by the automotive industry. The head injury criteria (HIC) calculated for feet first falls was limited by th e inability to obtain acceleration profiles for heights greater than 62.5 cm when a floor mat was not in use. This limitation resulted in an incomplete data set for th e higher heights. However, a trend line equation fitted to existing data wa s used to estimate HIC values at those heights. The HIC values calculated for accel eration profiles meas ured without a mat reached a maximum of 1234.63 945.72 at a hei ght of 48 cm. As shown in Figure 7, this value results in approximately a 25 percen t chance of experiencing an injury with an AIS score greater than or equal to 3 under these conditions. Furthe rmore, the projected data for bed heights up to 97.5 cm increase d to 1252.87, which indicates approximately a 25 percent chance of serious inju ry. With the inclusion of he ight added by bedrails, this risk increases to a 40 percen t chance of a serious injury resulting from falling from bed and impacting a tile surface. On the othe r hand, the HIC values reached a maximum of 374.35 389.04 at a height of 77.0 cm when the floor mat was in use. This value indicated a less than one perc ent chance of serious injury when compared to Figure 7. Moreover, the risk of injury does not increase with an increase in height added by bedrails, as the projected HIC values only increase to 250.70. Th is relatively high risk of serious injury associated with falling out of bed feet first without a mat is not supported

PAGE 84

75 by the literature as brain injury has not been specifically a ssociated with falls from bed. However, these results may in dicate that patients do not commonly fall from bed feet first. Due to accelerometer limitation, accelerati on profiles could not be measured at the thorax for a bed height of 97.5 cm. As such, the injury criteria calcul ated for the acceleration profiles measured at the thorax for heights up to 91.5 cm resulted in a maximum TIC value of 1103.40 733.32. Thoracic injury criteria for subsequent bed and bedrail heights were estimated by fit ting a trend line to the existing data; the estimated TIC values increa sed to 8053.09. Although literatur e currently does not exist to correlate these values with injury severity, an exponentia l increase in TIC values is observed with an increase in height. Assu ming the calculated TIC values follow a similar trend as the HIC value, this increase in TIC value would indicate an increase in risk of serious injury to the thorax. On the other hand, the impact dur ing feet first falls occurred on the posterior portion of the thorax; therefore, a disloca tion or fracture is not as likely as with the lateral impact obser ved in head first falls. Similar to head first falls, the mean maximum values measured at the thorax were compared to standards set forth by the automotive industry. The values measured dur ing falls onto the tile surface reached a maximum of 95.12 43.13 g and 58.25 46.01 g dur ing falls onto the floor mat. Clearly, these values exceed the automotive in dustry standard of a 60 g limit. Although the PIC values calcula ted for feet first falls in creased significantly with height to reach a maximum of 54.46 54.98 when calculated for impacts onto the tile surface, a correlation with inju ry severity cannot be determined. Furthermore, the pelvis impacted the posterior portion du ring feet first falls; therefor e, falling from bed feet first would probably not result in a hip fracture rather a wrist fracture according to Nevitt et al (1993). This is supported by th e literature as the force re quired to fracture a hip is reported to be approximately 4340 N, and the forces calculated in the current study were below this fracture threshold (Etheri dge, Beason, Lopez, Alonso, McGwin, and Eberhardt, 2005). Th e literature has also documented fr actures of the extremities to be specifically associated with fa lling and fractures to be gene rally associated with falling from bed (Sterling, O’Conner, and Bonadies, 2001; Innes and Turman, 1983). As the

PAGE 85

76 feet impact the surface first and dampen the acceleration m easured at the pelvis, the incidence of extremity fractur es may possibly be the result of just such a fall. General observations Throughout the data collecti on process, several observ ations were made with regard to fall mechanics and possible resulting injury. For instance, the material around the knee and shoulder impact sites became torn over time and may support the incidence of lacerations documented in the literatur e (Lyons and Oates, 1993; Macgregor, 2000). Furthermore, inspection of th e mat revealed permanent deformation at the head and thorax impact sites. This information may be useful in determin ing mat placement and design. As discussed previous ly, the ATD fell from bed in a specific manner to impact particular body regions in a sp ecific orientation. However, adding bedrails to a bedside may produce very different fall mechanisms as the bedrail may provide an additional pivot point about which the ATD can rotate. Particularly during feet first falls, the ATD may complete more degrees of rotation and imp act the pelvis and thorax laterally rather than on the posterior portion as observed duri ng this study. Additionally, the literature documents the incidence of bedrail entanglemen t; hence falling from be d with bedrails in place may increase the incide nce of extremity fractures a nd dislocations by allowing entanglement to occur (FDA, CDRH, 1995; JCAHO, 2002). There has been some discussion concer ning the applicability of the measures determined in this study to human subjects as the ATD has different biomechanical properties with respect to the vinyl skin. Th e properties of the vinyl allowed the ATD to rebound off the impact surface. However, as the human body displays viscoelastic properties during applied forces, one can presume that some re silience is present in the human skin. Furthermore, the effect of rebound on the meas ured and calculated values would result in liberal estimates of these calculations; therefore, the calculations and measures reported in this study would apply to the more conservative results theoretically expected in human subjects.

PAGE 86

77 Conclusions and Recommendations Even though the HIC values calculated in this study re sulted in approximately a 40 percent chance of sustaining a serious brain injury as a consequence of falling from bed, the use of a mat signifi cantly reduced this relatively high risk of injury. Furthermore, the mat provided a protective eff ect for the pelvis duri ng head first falls and for the thorax during feet first falls. As such, a floor mat should be used in the healthcare environment to prevent injuries associated w ith falling from bed. Ho wever, the floor mat used in this study did not provi de appropriate floor coverage to ensure impact onto the mat rather than onto the tile surface. Du ring data collection, the mat was repositioned several times as the pelvis of the ATD tended to impact farther past the foot of the bed with an increase in height during head first fa lls. Similarly, the h ead of the ATD tended to impact farther past the head of the bed during f eet first falls. Howe ver, in a clinical situation the bed may be positioned against a wa ll; therefore, the length of the mat would not need to be lengthened in that direction. To increase the prevention potential of the floor mat, it should be lengthe ned approximately 30 cm to co ver a total length of 213 cm to extend past the foot of the bed. The ATD also impacted the surf ace farther away from the bedside with an increase in height; th erefore the mat was repositioned several times during both head first and feet first falls to account for this movement. For the best clinical performance, the width of th e floor mat should also be increased by approximately 15 cm to increase to a total width of 111.5 cm. Howeve r, clinicians have posed concerns regarding tripping hazards and sanitation methods associated with the use of floor mats. A floor mat with a beveled edge should be used in the clinic al setting to reduce the risk of introduci ng a trip hazard, and a mat with a plastic or other easily cleaned surface should be used to reduce the risk of spreadin g infection. If the use of a floor mat is not a feasible op tion for a particular patient or facility, other protective devices are available such as hip protectors. However, patients often do not use them

PAGE 87

78 consistently as they have been reported to be uncomfortable to wear and are not aesthetically pleasing to the pa tient. As such, the use of a floor mat is the less invasive injury prevention device when compared to hip protectors. Regardless of fall direction, the mean maximum values measured at the head, thorax, and pelvis were all de termined to increase significantly with increasing height. To prevent the most injuries resu lting from falling from bed, the bed should be positioned to the lowest height available wh ile the patient is left unatte nded. The literature supports repositioning the bed to a height comfortable for caregivers to reduce the risk of low back injuries experienced by the caregiver (DeLooze et al, 1994 ; Caboor et al, 2000). By utilizing an adjustable bed in the clinical environment, inju ries can be prevented for both the patient and th e caregiver. Bedrails were not physically implemente d in this study; however, mean maximum values and injury criteria were extrapolated from estimated trend lines of the data collected during this study. These values reflected a cons istent increase in mean maximum values and injury criteria for all test conditions and body regions. When the entrapment and entanglement issues presented in the literature are taken into account, the removal of bedrails from the clinical e nvironment should be solemnly considered. Furthermore, the literature has documented fa lls from bed even when the bedrails were in place; therefore, the bene fit of utilizing bedrails must come into question. The results of this study clearly support removing bedrails simply based on the ef fect the increase in height has on the mean maximum values and injury criteria. In conclusion, the results of this study have important clinical applications as injury prevention is paramount to mainta ining a quality healthcare environment. According to the results of this study, the ideal environment for preventing injuries resulting from falling out of bed include positioning the be d to the lowest available height, placing a floor mat be side the bed, and removing bedr ails. Historically, injury prevention devices, such as bedrails or physi cal restraints, were implemented in part to show an active participation in injury prevention to the le gal system. By implementing the aforementioned changes in the clinical environment, th e patient and caregiver could

PAGE 88

79 be assured that the highest quality care is being provided as each are easily documented and pose little negative impact on the daily function of prov iding care to patients.

PAGE 89

80 References Backaitis, S.H, & Mertz, H.J. (1994), Hybrid III: The first hum an-like crash test dummy Warrendale, Pennsylvania: Society of Automotive Engineers, Inc. Barbee, G.C. (1957). More a bout bedrails and the nurse. The American Journal of Nursing, 57, 1441-1442. Baum, T., Capezuti, E., & Dris coll, G.(2002). Falls. In Cott er, V.T. & Strumpf, N.E. (Eds.), Advanced Practice Nursing with Ol der Adults: Clinical Guidelines (pp.245-259). New York: McGraw Hill. Bertocci, G.E., Pierce M.C., Deemer, E., Aguel, F., Janosky, J.E., & Vogeley, E. (2003).Using test dummy experiments to investigate pediatric injury risk in simulated short-distance falls. Archives of Pediat rics and Adolescent Medicine 157, 480-486. Bones and Skeletal Tissues. In E. N. Marieb (Ed.), Human Anatomy & Physiology (pp. 172-197). New York: Benjamin Cummings. Bradham, D.D., MacClellan, L.R., South, B ., Tate, M., Powell-Cope, G., Luther, S., et al.(2003). Hospital bed-re lated adverse events, Part II : Direct costs to a VHA healthcare network. Journal of Healthcare Safety 1(3), 24-30. Braun, J., & Capezuit, E. (2000). Siderail use and legal liabil ity in Illinois nursing homes. Illinois Bar Journal vol(), pp-pp. Caboor, D.E., Verlinden, M.O., Zinzen, E ., van Roy, P., van Riel, M.P., & Clarys, J.P. (2000). Implications of an adjustable bed hei ght during standard nursing tasks on spinal motion, perceive d exertion and muscular activity. Ergonomics 43 (10), 1771-1780. Capezuit, E., Talerico, K.A., Cochran, I., Be cker, H., Strumpf, N., & Evans, L.(1999). Individualized interventions to prevent bed-related falls and reduce siderail use. Journal of Geronotological Nursing vol, 26-34. Casalena, J.A., Badre-Alam, A., Ovaert, T. C., Cavanagh, P.R., & Streit, D.A. (1998). The Penn State safety floor: Part II-R eduction of fall-related peak impact forces on the femur. Journal of Biomechanical Enigneering 120, 527-532.

PAGE 90

81 Champion, H.R., Copes, W.S., Buyer, D. Flanagan, M.E., Bain, L., & Sacco, W.J. (1989). Major trauma in geriatric patients. American Journal of Public Health 79, 1278-1282. Copes, W.S., Sacco, W.J., Ch ampion, H.R., & Bain, L.W. (n.d.). Progress in characterizing anatomic injury. In Proceedings of the 33rd Annual Meeting of the Association for the Advanc ement of Automotive Medicine USA, 205-218. Retrieved November 29, 2004, from http ://www.trauma.org/scores/ais.html de Looze, M.P., Zinzen, E., Caboor, D., Heyblom, P., van Bree E., van Roy, P., et al. (1994). Effect of individua lly chosen bed-height ad justments on the low-back stress of nurses. Scandinavian Journal of Environmental Health 20, 427-434. Donius, M., & Rader, J. (1994) Use of siderails: Rethinking a standard of practice. Journal of Gerontological Nursing vol(), 23-27. Eppinger, R., Sun, E., Kuppa, S., & Saul, R. Reports to National Highway Safety Administration. (2000). Supplemen t: Development of improv ed injury criteria for the assessment of advanced automotive restraint system s-II. Retrieved March 23, 2005 from http://www-nrd.nhsta.dot.gov/pdf/n rd-11/airbags/finalrule_all.pdf Etheridge, B.S, Beason, D.P., Lopez, R.R., Al onso, J.E., McGwin, G., & Eberhardt, A.W. (2005). Effects of trochanteric soft tissues and bone de nsity on fracture of the female pelvis in experimental side impacts. Annals of Biomed ical Engineering 33, 248-254. Feinsod, F.M., Moore, M., & Levenson, S.A. (1997). Eliminating full-length bed side rails from long-term care facilities. Nursing Home Medicine 5(8), 257-263. French, D.D., Campbell, R., Spehar, A., & Angaran, D.M. (2004). Benzodiazepines and injury: A risk adjusted model. Pharmacoepidemiology and Drug Safety. Gaebler, S. (1993). Predicting which patient will fall again…and again. Journal of Advanced Nursing 18, 1895-1902. Gurwitz, J.H., Sanchez-Cro ss, M.T., Eckler, M.A., & Matulis, J. (1994). The epidemiology of adverse and unexpected events in the long-term care setting. The Journal of the Amer ican Geriatrics Society 42, 33-38. Hanger, H.C., Ball, M.C., &W ood, L.A. (1999). An analysis of falls in the hospital: Can we do without bedrails? The Journal of the Amer ican Geriatrics Society 47, 539-531.

PAGE 91

82 Hoffman, S.B., Powell-Cope, G., MacClella n, L., & Bero, K. (2003). BEDSAFE: A bed safety project fo r frail older adults. Journal of Gerontological Nursing vol, 34-42. Hospital Bed Safety Workgroup. (2000, October). A guide to bed safety Retrieved July 9,2004, from http: // www.fda.gov/cdrh/beds/ Innes, E.M., & Turman, W.G. (1983 ). Evaluation of patient falls. QRB vol, 30-35. Janken, J.K., Reynolds, B.A., Swiech, K. ( 1988). Patient falls in the acute care setting: Identifying risk factors. Nursing Research 35(4), 216-219. Joint Commission on Accreditation of Hea lthcare Organizations. (2002, September 6). Sentinel Event Alert. Retrieved May 14, 2004, from http://www.jcaho.org/ about+us/news+letters/sentin el+event+alert/print/sea_27.htm Lyons, T.J., & Oates, R.K. (1993) Falling out of bed: A rela tively benign occurrence. Pediatrics 92, 125-127. Macgregor, D.M. (2000). Injuries associated with falls from beds. Injury Prevention 6, 291-292. Mendelson, W.B. (1996). The use of sedative/hypnotic me dication and its correlation with falling down in the hospital. Sleep 19(9), 698-701. Mertz, H. (1994). Anthropomorphic test devi ces. In Backaitis, S.H.& Mertz, H. (Eds.), Hybrid III: The first hum an-like crash test dummy (pp387-405). Warrendale, PA: Society of Automotive Engineers, Inc. Nahum, A.M., Gatts, J.D., Gadd, C.W., & Danfor th, J.P. (1968). Impact tolerance of the skull and face, Proceedings of the 12th STAPP Conference, SAE 680785. National Center for Injury Prevention and Control. (n.d.). Falls and hip fractures among older adults. Retrieved April 13, 2004, from http://www.cdc.gov/nci pc/factsheets/falls.htm Nevitt, M.C., Cummings, S.R., & Study of Osteoporotic Fractures Research Group. (1993). Type of fall and risk of hip and wrist fractures: The study of osteoporotic fractures. The Journal of the Amer ican Geriatrics Society 41, 1226-1234. Nordin, M. & Frankel, V.H. (2001). Basic Biomechanics of the Musculoskeletal System (3rd ed.). New York: Lippinco tt Williams & Wilkins.

PAGE 92

83 O’Loughlin, J.L., Robitaille, Y., Boivin, J., & Suissa, S. (199 3). Incidence of and risk factors for falls and injurious falls among the community-dwelling elderly. American Journal of Epidemiology 137, 342-354. Parker, K., & Miles, S.H. ( 1997). Deaths caused by bedrails. The Journal of the American Geriatrics Society 45, 797-802. Santora, T.A., Schinco, M.A, & Trooskin, S. Z. (1994). Management of trauma in the elderly patient. Surgical Clinics of North America, 74, 163-186. Sattin R.W. et al. (1990). Th e incidence of fall injury ev ents among the elderly in a defined population. American Journal of Epidemiology, 131(6), 1028-1037. Schneider, D.C., & Nahum, A.M. (1972). Imp act studies of facial bones and skull, Proceedings of the 16th STAPP Conference, SAE 720965. Si M., Neufeld, R.R., & Dunbar, J. (1999). Removal of bedrails on a short-term nursing home rehabilitation unit. The Gerontologist 39(5), 611-614. Simpson, A.H.R.W., Lamb, S., Roberts, P.J ., Gardner, T.N., & Grimley Evans, J. (2004). Does the type of flooring affect the risk of hip fracture? Age and Ageing 33, 242-246. Sterling, D.A., O’Connor, J.A., and Bonadies J. (2001). Geriat ric falls: Injury severity is high and disproportionate to mechanism. The Journal of TRAUMA Injury, Infection, and Critical Care 50, 116-119. Tinetti, M.E., Speechley, M., Ginter, S.F. 1988. Risk factors for falls among elderly persons living in the community. New England Journal of Medicine 319, 1701-1707. Todd, J.F., Ruhl, C., & Gross, T.P. (1997). Injury and death associated with hospital bedside-rails: Reports to the US Food and Drug Administration from 1985 to 1995. The American Journal of Public Health 87 (10), 1675-1677. Trauma.Org (n.d.). Abbreviated injury s cale. Retrieved November 29, 2004 from http://www.trauma.org/scores/ais.html U.S. Food and Drug Administration. ( 1995, August 23). FDA Safety alert: Entrapment hazards with hospital bed side rails. Retrieved July 9, 2004, from http://www.fda.gov/cd rh/bedrails.html Versace, J. (1972). A review of the severity index, Proceedings of the 15th Stapp Car Crash Conference, SAE 710881, Society of Automotive Engineers. 771-796.

PAGE 93

84 Walshe, A. & Rosen, H. (1979). A study of patient falls from bed. Journal of Nusing Adminstration

PAGE 94

85 Appendices

PAGE 95

86 Appendix A: MatLab Code %create empty matrices to store values mone= []; mtwo=[]; mthree=[]; mfour=[]; %loop trial numbers to reduce processing time for i = [204 205 206]; %read excel files, time, x,y,and z axes for head, pelvis, and thorax T= XLSread (sprintf( 'TRIAL_%g' ,i),sprintf( 'TRIAL_%g' ,i),'m5:m8000'); Accelxh = xlsread (sprintf( 'TRIAL_%g' ,i),sprintf( 'TRIAL_%g' ,i), 'd5:d8000'); Accelyh = xlsread (sprintf( 'TRIAL_%g' ,i),sprintf( 'TRIAL_%g' ,i), 'g5:g8000'); Accelzh = xlsread (sprintf( 'TRIAL_%g' ,i),sprintf( 'TRIAL_%g' ,i), 'j5:j8000'); Accelxp = xlsread (sprintf( 'TRIAL_%g' ,i),sprintf( 'TRIAL_%g' ,i), 'f5:f8000'); Accelyp = xlsread (sprintf( 'TRIAL_%g' ,i),sprintf( 'TRIAL_%g' ,i), 'i5:i8000'); Accelzp = xlsread (sprintf( 'TRIAL_%g' ,i),sprintf( 'TRIAL_%g' ,i), 'l5:l8000'); Accelxt = xlsread (sprintf( 'TRIAL_%g' ,i),sprintf( 'TRIAL_%g' ,i), 'e5:e8000'); Accelyt = xlsread (sprintf( 'TRIAL_%g' ,i),sprintf( 'TRIAL_%g' ,i), 'h5:h8000'); Accelzt = xlsread (sprintf( 'TRIAL_%g' ,i),sprintf( 'TRIAL_%g' ,i), 'k5:k8000'); %fft and filter head, pelvis, thorax individual axes freq = 1000; points = 5000; trans_head_x = fft(Accelxh,points); trans_head_x(1)= []; head_x = trans_head_x.*conj(trans_head_x); fhead_x = freq/points*(0:(points/2)-1); trans_head_y = fft(Accelyh,points); trans_head_y(1)= []; head_y = trans_head_y.*conj(trans_head_y); fhead_y = freq/points*(0:(points/2)-1); trans_head_z = fft(Accelzh,points); trans_head_z(1)= []; head_z = trans_head_z.*conj(trans_head_z); fhead_z = freq/points*(0:(points/2)-1);

PAGE 96

87 Appendix A: Continued trans_pelvis_x = fft(Accelxp,points); trans_pelvis_x(1)= []; pelvis_x = trans_pelvis_x.*conj(trans_pelvis_x); fpelvis_x= freq/points*(0:(points/2)-1); trans_pelvis_y = fft(Accelyp,points); trans_pelvis_y(1)= []; pelvis_y = trans_pelvis_y.*conj(trans_pelvis_y); fpelvis_y= freq/points*(0:(points/2)-1); trans_pelvis_z = fft(Accelzp,points); trans_pelvis_z(1)= []; pelvis_z = trans_pelvis_z.*conj(trans_pelvis_z); fpelvis_z= freq/points*(0:(points/2)-1); trans_thorax_x = fft(Accelxt,points); trans_thorax_x(1)= []; thorax_x = trans_thorax_x.*conj(trans_thorax_x); fthorax_x= freq/points*(0:(points/2)-1); trans_thorax_y = fft(Accelyt,points); trans_thorax_y(1)= []; thorax_y = trans_thorax_y.*conj(trans_thorax_y); fthorax_y= freq/points*(0:(points/2)-1); trans_thorax_z = fft(Accelzt,points); trans_thorax_z(1)= []; thorax_z = trans_thorax_z.*conj(trans_thorax_z); fthorax_z= freq/points*(0:(points/2)-1); %filter individual axes of head, pelvis, and thorax using fft %frequencies Ts=0.001; Ws=1/Ts; Wn=Ws/2; n=4; newW=150/500; [b,a]=butter (n,newW); [H,k]=freqz(b,a); q=k*Wn/(2*pi); r=abs(H); headfilt_x = filtfilt (b,a,Accelxh); headfilt_y = filtfilt (b,a,Accelyh); headfilt_z = filtfilt (b,a,Accelzh); pelvisfilt_x = filtfilt (b,a,Accelxp); pelvisfilt_y = filtfilt (b,a,Accelyp); pelvisfilt_z = filtfilt (b,a,Accelzp); thoraxfilt_x = filtfilt (b,a,Accelxt); thoraxfilt_y = filtfilt (b,a,Accelyt); thoraxfilt_z = filtfilt (b,a,Accelzt);

PAGE 97

88 Appendix A: Continued %calculate rms values for head, pelvis, thorax rmshead = sqrt((headfilt_x.*headfi lt_x) + (headfilt_y.*headfilt_y) + (headfilt_z.*headfilt_z)); rmspelvis= sqrt((pelvisfilt_x.*pelvisfilt_x) +(pelvisfilt_y.*pelvisfilt_y)+( pelvisfilt_z.*pe lvisfilt_z)); rmsthorax= sqrt((thoraxfilt_x.*thoraxfilt_x) +(thoraxfilt_y.*thorax filt_y)+(thoraxfilt _z.*thoraxfilt_z)) %plot raw and filtered rms data figure(' Name' ,sprintf( 'Raw and Filtered Data for Trial_%g' ,i)); subplot(2,1,1); plot(T,Accelxh, T, Accelyh, T, Accelzh); legend ( 'x' 'y' 'z' ); title ( 'unfiltered data' ); subplot(2,1,2); plot (T,rmspelvis, T, rm shead, T, rmsthorax); legend ( 'pelvis' 'head' 'thorax' ); title ( 'filtered data' ); %find max value for eff ect size calculation headmax = max(rmshead); pelvismax = max(rmspelvis); thoraxmax = max(rmsthorax); %create matrix to displa y Trial # and max value mone = [mone; i headmax pelvismax thoraxmax]; %calculate hic values endpoint = length(rm spelvis) 150; for h=[1:endpoint]; two=1+h; headint=trapz(T(h:tw o),rmshead(h:two)); headhic_1=(((headint) .*(1/((two-h)/1000)) )^ 2.5).*((two-h)/1000); three=2+h; headint=trapz(T(h:three),rmshead(h:three)); headhic_2=(((headint).*(1/((thr ee-h)/1000)) )^2.5).*((three-h)/1000); four=3+h; headint=trapz(T(h:four ),rmshead(h:four)); headhic_3=(((headint).*(1/((fou r-h)/1000)) )^2.5).*((four-h)/1000);

PAGE 98

89 Appendix A : Continued five=4+h; headint=trapz(T(h:five),rmshead(h:five)); headhic_4=(((headint).*(1/((five -h)/1000)) )^2.5). *((five-h)/1000); six=5+h; headint=trapz(T(h:six),rmshead(h:six)); headhic_5=(((headint) .*(1/((six-h)/1000)) )^2.5).*((six-h)/1000); seven=6+h; headint=trapz(T(h:seven),rmshead(h:seven)); headhic_6=(((headint) .*(1/((seven-h)/1000)) )^ 2.5).*((seven-h)/1000); eight=7+h; headint=trapz(T(h:eight ),rmshead(h:eight)); headhic_7=(((headint) .*(1/((eight-h)/1000)) )^ 2.5).*((eight-h)/1000); nine=8+h; headint=trapz(T(h:nine),rmshead(h:nine)); headhic_8=(((headint) .*(1/((nine-h)/1000)) )^ 2.5).*((nin e-h)/1000); ten=9+h; headint=trapz(T(h:ten),rmshead(h:ten)); headhic_9=(((headint).*(1/((tenh)/1000)) )^2.5).*((ten-h)/1000); eleven=10+h; headint=trapz(T(h:eleven ),rmshead(h:eleven)); headhic_10=(((headint).*(1/((eleven-h)/1000)) )^2.5).*((eleven-h)/1000); twelve=11+h; headint=trapz(T(h:twelve),rmshead(h:twelve)); headhic_11=(((headint).*(1/((twelve-h)/1000)) )^2.5).*((twelve-h)/1000); thirteen=12+h; headint=trapz(T(h:thirteen ),rmshead(h:thirteen)); headhic_12=(((headint).*(1/((thirteenh)/1000)) )^2.5).*((thirteen-h)/1000); fourteen=13+h; headint=trapz(T(h:fourteen ),rmshead(h:fourteen)); headhic_13=(((headint).*(1/((fourteen-h)/1000)) )^2.5).*((fourteen-h)/1000); fifteen=14+h; headint=trapz(T(h:fiftee n),rmshead(h:fifteen)); headhic_14=(((headint).*(1/((fifteen-h)/1000)) )^2.5).*((fifteen-h)/1000);

PAGE 99

90 Appendix A: Continued sixteen=15+h; headint=trapz(T(h:sixteen),rmshead(h:sixteen)); headhic_15=(((headint).*(1/((sixteen-h)/1000)) )^2.5).*((sixteen-h)/1000); mtwo=[mtwo; i h headhic_15]; end %calculate tic values for h=[1:endpoint]; two=1+h; thoraxint=trapz(T(h:two),rmsthorax(h:two)); thoraxtic_1=(((thoraxint).*(1/(( two-h)/1000)) )^2.5).*((two-h)/1000); three=2+h; thoraxint=trapz(T(h:three),rmsthorax(h:three)); thoraxtic_2=(((thoraxint).*(1/((th ree-h)/1000)) )^2.5) .*((three-h)/1000); four=3+h; thoraxint=trapz(T(h:f our),rmsthorax(h:four)); thoraxtic_3=(((thoraxint).*(1/(( four-h)/1000)) )^2.5).*((four-h)/1000); five=4+h; thoraxint=trapz(T(h:five),rmsthorax(h:five)); thoraxtic_4=(((thoraxint).*(1/((fi ve-h)/1000)) )^2.5).*((five-h)/1000); six=5+h; thoraxint=trapz(T(h:six),rmsthorax(h:six)); thoraxtic_5=(((thoraxint).*(1/(( six-h)/1000)) )^2.5) .*((six-h)/1000); seven=6+h; thoraxint=trapz(T(h:seven),rmsthorax(h:seven)); thoraxtic_6=(((thoraxint).*(1/((seve n-h)/1000)) )^2.5).*((seven-h)/1000); eight=7+h; thoraxint=trapz(T(h:eight),rmsthorax(h:eight)); thoraxtic_7=(((thoraxint).*(1/((ei ght-h)/1000)) )^2.5).*((eight-h)/1000); nine=8+h; thoraxint=trapz(T(h:nine),rmsthorax(h:nine)); thoraxtic_8=(((thoraxin t).*(1/((nine-h)/1000)) )^ 2.5).*((nine-h)/1000);

PAGE 100

91 Appendix A: Continued ten=9+h; thoraxint=trapz(T(h:ten),rmsthorax(h:ten)); thoraxtic_9=(((thoraxint).*(1/(( ten-h)/1000)) )^2.5).*((ten-h)/1000); eleven=10+h; thoraxint=trapz(T(h:eleven),rmsthorax(h:eleven)); thoraxtic_10=(((thoraxin t).*(1/((eleven-h)/1000)) )^2.5).*((eleven-h)/1000); twelve=11+h; thoraxint=trapz(T(h:twelve),rmsthorax(h:twelve)); thoraxtic_11=(((thoraxin t).*(1/((twelve-h)/1000)) )^2.5).*((twelve-h)/1000); thirteen=12+h; thoraxint=trapz(T(h:thirteen),rmsthorax(h:thirteen)); thoraxtic_12=(((thor axint).*(1/((thirteen-h)/1000)) )^2.5).*((thirt een-h)/1000); fourteen=13+h; thoraxint=trapz(T(h:fourteen ),rmsthorax(h:fourteen)); thoraxtic_13=(((thor axint).*(1/((fourteen-h)/1000)) )^2.5).*((fourteen-h)/1000); fifteen=14+h; thoraxint=trapz(T(h:fifteen),rmsthorax(h:fifteen)); thoraxtic_14=(((thor axint).*(1/((fifteen-h)/1000)) )^2.5).*((fifteen-h)/1000); sixteen=15+h; thoraxint=trapz(T(h:sixteen),rmsthorax(h:sixteen)); thoraxtic_15=(((thor axint).*(1/((sixteen-h)/1000)) )^2.5).*((sixteen-h)/1000); mthree=[mthree; i h thoraxtic_15]; end %calculate pic values for h=[1:endpoint]; two=1+h; pelvisint=trapz(T(h:two),rmspelvis(h:two)); pelvispic_1=(((pelvisint).*(1/((two-h)/1000)) )^2.5).*((two-h)/1000); three=2+h; pelvisint=trapz(T(h:three),rmspelvis(h:three)); pelvispic_2=(((pelvisint).*(1/((three-h)/1000)) )^2.5).*((three-h)/1000);

PAGE 101

92 Appendix A: Continued four=3+h; pelvisint=trapz(T(h:fou r),rmspelvis(h:four)); pelvispic_3=(((pelvisint).*(1/((four-h)/1000)) )^2.5).*((four-h)/1000); five=4+h; pelvisint=trapz(T(h:five),rmspelvis(h:five)); pelvispic_4=(((pelvisint).*(1/((fiv e-h)/1000)) )^2.5).*((five-h)/1000); six=5+h; pelvisint=trapz(T(h:six),rmspelvis(h:six)); pelvispic_5=(((pelvisint).*(1/((six-h)/1000)) )^2.5).*((six-h)/1000); seven=6+h; pelvisint=trapz(T(h:seven),rmspelvis(h:seven)); pelvispic_6=(((pelvisint).*(1/((seven-h)/1000)) )^2.5).*((seven-h)/1000); eight=7+h; pelvisint=trapz(T(h:eight),rmspelvis(h:eight)); pelvispic_7=(((pelvisint).*(1/((eight-h)/1000)) )^2.5).*((eight-h)/1000); nine=8+h; pelvisint=trapz(T(h:nine),rmspelvis(h:nine)); pelvispic_8=(((pelvisint).*(1/((nine-h)/1000)) )^2.5).*((nine-h)/1000); ten=9+h; pelvisint=trapz(T(h:ten),rmspelvis(h:ten)); pelvispic_9=(((pelvisint).*(1/((ten-h)/1000)) )^2.5).*((ten-h)/1000); eleven=10+h; pelvisint=trapz(T(h:eleven ),rmspelvis(h:eleven)); pelvispic_10=(((pelvisint).*(1/((eleven-h)/1000)) )^2.5).*((eleven-h)/1000); twelve=11+h; pelvisint=trapz(T(h:twelve),rmspelvis(h:twelve)); pelvispic_11=(((pelvisint).*(1/((twelve-h)/1000)) )^2.5).*((twelve-h)/1000); thirteen=12+h; pelvisint=trapz(T(h:thirteen),rmspelvis(h:thirteen)); pelvispic_12=(((pelvisint).*(1/((thirtee n-h)/1000)) )^2.5).*((thirteen-h)/1000); fourteen=13+h; pelvisint=trapz(T(h:fourteen ),rmspelvis(h:fourteen)); pelvispic_13=(((pelvisint).*(1/((fourteen-h)/1000)) )^2.5).*((fourteen-h)/1000);

PAGE 102

93 Appendix A: Continued fifteen=14+h; pelvisint=trapz(T(h:fifteen ),rmspelvis(h:fifteen)); pelvispic_14=(((pelvisint).*(1/((fift een-h)/1000)) )^2.5).*((fifteen-h)/1000); sixteen=15+h; pelvisint=trapz(T(h:sixteen),rmspelvis(h:sixteen)); pelvispic_15=(((pelvisint).*(1/((sixteen-h)/1000)) )^2.5).*((sixteen-h)/1000); mfour=[mfour; i h pelvispic_15]; end end csvwrite ( 'maxvalue.csv' ,mone); csvwrite ( 'hicvalue.csv' ,mtwo); csvwrite( 'ticvalue.csv' ,mthree); csvwrite( 'picvalue.csv' ,mfour);

PAGE 103

94 Appendix B: Protocol 1. Turn on computer 2. Open LabView VI: “FP + Load Cells ” a. Set max cutoff frequency to 500 Hz b. Set min cutoff frequency to 1 Hz c. Set sample frequency to 1 000 Hz d. Set filter to None 3. Turn on SCUXI 4. Turn power to on position for accelerometer power supply 1,2,3 5. Connect each cable to correct positio n on project box for accelerometer 1,2,3 a. Head accelerometer-Accel erometer 1 (SN 35568) b. Thorax accelerometer-Accel erometer 2 (SN 37400) c. Pelvis accelerometer-Accel erometer 3 (SN 37401) 6. Position bed to correct height (measured in cm from ma ttress top with no ATD on surface) 7. Lock bed brakes 8. Position sling on bed according to marks i ndicated on mattress for feet or head first drop 9. Position ATD on the sling according to marks indicated for feet or head first drop 10. Position video camera to view impact surface 11. Record Trial number with video camera 12. Raise sling using ceiling lift a nd coordinate data acquisiti on with height of sling relative to the bed height. 13. Export impact event to file 14. Repeat steps 8-13 until all tr ials are complete, stopping periodically to detangle accelerometer cables

PAGE 104

95 Appendix C: SAS Code libname look 'e:';run; ********************************************************************** Head first analysis *****************************************************************; proc glm data = headfirst_max; class height mat; model head = height mat height*mat; title 'Fixed ANOVA for Head First Head Acceleration'; run; proc glm data = headfirst_max; class height mat; model pelvis = height mat; title 'Fixed ANOVA for Head First Pelvis Acceleration'; run; proc glm data = headfirst_max; class height mat; model thorax = height mat; title 'Fixed ANOVA for Head First Thorax Acceleration'; run; ********************************************************************** Feet first analysis *****************************************************************; data feetfirst_max2; set feetfirst_max; if height = 33.5 then h_2 = 1; else if height = 48 then h_2 = 2; else if height = 62.5 then h_2 = 3; else h_2 = .; if head = then delete; run; proc glm data = feetfirst_max2; class h_2 mat; model head = h_2 mat; title 'Fixed ANOVA for Feet First Head Acceleration'; run; proc glm data = feetfirst_max; class height mat; model pelvis = height mat; title 'Fixed ANOVA for Feet First Pelvis Acceleration'; run;

PAGE 105

96 Appendix C: Continued data feetfirst_max3; set feetfirst_max; if height = 33.5 then h_2 = 1; else if height = 48 then h_2 = 2; else if height = 62.5 then h_2 = 3; else if height = 77 then h_2 = 4; else if height = 91.5 then h_2 = 5; else h_2 = .; if thorax = then delete; run; proc glm data = Feetfirst_max3; class h_2 mat; model thorax = h_2 mat h_2*mat; title 'Fixed ANOVA for Feet First Thorax Acceleration'; run; **************************************************** HIC analysis ***********************************************; data feet_2; set feet; if height >= 77 then delete; run; proc glm data = Feet_2; class height mat; model hic = height mat; title 'Fixed ANOVA for F eet First HIC Values'; run; ***************************************************** Head first HPT ***********************************************; proc glm data = HeadFirst_HPT; class height mat_no; model hic = height mat_no height*mat_no ; title 'Fixed ANOVA for He ad First HIC Values'; run;

PAGE 106

97 Appendix C: Continued proc glm data = HeadFirst_HPT; class height mat_no; model pic = height mat_no; title 'Fixed ANOVA for He ad First PIC Values'; run; proc glm data = HeadFirst_HPT; class height mat_no; model tic = height mat_no; title 'Fixed ANOVA for Head First TIC Values'; run; ***************************************************** Feet first HPT ***********************************************; data FeetFirst_HPT2; set FeetFirst_HPT; if height >= 77 then delete; run; proc glm data = FeetFirst_HPT2; class height mat; model hic = height mat; title 'Fixed ANOVA for fe et First HIC Values'; run; proc glm data = FeetFirst_HPT; class height mat; model pic = height mat; title 'Fixed ANOVA for fe et First PIC Values'; run; data FeetFirst_HPT3; set FeetFirst_HPT; if height >= 97.5 then delete; run; proc glm data = FeetFirst_HPT3; class height mat; model tic = height mat height*mat; title 'Fixed ANOVA for F eet First TIC Values'; run;


xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001681104
003 fts
005 20060215071239.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 060104s2005 flu sbm s000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001010
035
(OCoLC)62774433
SFE0001010
040
FHM
c FHM
049
FHMM
090
R856 (Online)
1 100
Bowers, Bonnie E.
0 245
Biomechanical evaluation os injury severity associated with patient falls from bed
h [electronic resource] /
by Bonnie E. Bowers.
260
[Tampa, Fla.] :
b University of South Florida,
2005.
502
Thesis (M.S.B.E.)--University of South Florida, 2005.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 106 pages.
520
ABSTRACT: The incidence of falls in the elderly population is a growing concern in the healthcare industry as associated morbidity is high, particularly morbidity associated with falls from bed. Bedrails were implemented as a device intended to reduce the incidence of falls from bed; however, recent evidence may indicate that bedrails contribute to adverse events including entrapment and entanglement. As such, efforts have been madeto reduce the use of bedrails and implement alternatives including height adjustable beds and floor mats. An instrumented anthropomorphic test dummy was used in the current study to measure the deceleration profiles of the head, thorax, and pelvis upon impact onto a tile surface or floor mat. The height of the fall was varied by using a height adjustable bed, and the impact site was varied by head or feet first falls.The deceleration profiles were used to determine mean maximum values across repeated trials and to calculate injury criteria at the head (HIC), thorax (TIC), and pelvis (PIC). The mean maximum values were further used to estimate the effect of adding bedrails. Injury severity was then predicted from the injury criteria calculated for the head. From this study, the mean maximum values were found to significantly increase with an increase in height regardless of fall direction. As such, the addition of bedrails consequently increased these values. Furthermore, the use of a floor mat significantly reduced the mean maximum values at the head and pelvis during head first falls and at the head and thorax during feet first falls. Injury criteria were also calculated for each body region and found to be significantly increased with an increase in height and decreased with the use of the floor mat.
590
Adviser: William Lee.
Co-adviser: John Lloyd
653
Height.
Floor mats.
Bedrails.
Elderly.
Acceleration.
690
Dissertations, Academic
z USF
x Biomedical Engineering
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.1010