Forensic Science, Medicine and Pathology https://doi.org/10.1007/s12024-018-9964-z

ORIGINAL ARTICLE

Fatal falls involving stairs: an anthropological analysis of skeletal trauma Samantha K. Rowbotham 1

&

Soren Blau 1,2 & Jacqueline Hislop-Jambrich 3 & Victoria Francis 2

Accepted: 13 February 2018 # Springer Science+Business Media, LLC, part of Springer Nature 2018

Abstract The skeletal blunt force trauma resulting from fatal falls involving stairs is complex. There are countless ways an individual may fall when stairs are involved, and thus a variety of ways the skeleton may fracture. Therefore anecdotally, it may be said that there is no specific skeletal trauma characteristic of this fall type. In order to scientifically investigate this anecdotal understanding, this study provides a detailed investigation of the skeletal fracture patterns and morphologies resulting from fatal falls involving stairs. Skeletal trauma was analyzed using the full-body postmortem computed tomography scans of 57 individuals who died from a fall involving stairs. Trauma was examined in the context of the variables that potentially influence how an individual falls (i.e. sex, age, body mass index, number of stairs involved, psychoactive drugs, pre-existing conditions, landing surface and manner of the fall) using logistic regression. Skeletal trauma primarily occurred in the axial skeleton. An analysis of fracture patterns showed the cranial base was less likely to fracture in younger individuals and the cervical vertebrae were more likely to fracture in falls that involved more than half a flight of stairs. A total of 56 fracture morphologies were identified. Of these, diastatic fractures were less likely to occur in older individuals. Findings indicate that there are skeletal fracture patterns and morphologies characteristic of a fatal fall involving stairs. Keywords Forensic anthropology . Forensic pathology . Fatal fall . Fall involving stairs . Skeletal trauma . Blunt force trauma . Postmortem computed tomography . Fracture pattern . Fracture morphology . Autopsy . Bone

Introduction Falls involving stairs are a common type of accidental fall that can result in fatality. This is largely because stairs are an unavoidable object that most individuals encounter on a daily basis, whether in the home or public environment, which require complex biomechanical movements to navigate. Falls that involve stairs are arguably the most complex of all fall

Electronic supplementary material The online version of this article (https://doi.org/10.1007/s12024-018-9964-z) contains supplementary material, which is available to authorized users. * Samantha K. Rowbotham [email protected] 1

Department of Forensic Medicine, Monash University, 65 Kavanagh St, Southbank, Victoria 3006, Australia

2

Victorian Institute of Forensic Medicine, 65 Kavanagh St, Southbank, Victoria 3006, Australia

3

Centre for Medical Research, Toshiba Medical, 12 – 24 Talavera Rd, North Ryde, New South Wales 2113, Australia

types in terms of understanding how the resulting blunt force trauma (BFT) manifests on the skeleton. This complexity is attributed to not only the various types of stairs that may be involved, but also to the myriad of ways an individual may fall on, or from, those stairs. An individual may fall when they slip/stumble down stairs (i.e. descent fall), miss-step and free fall off stairs (i.e. descent fall) or trip up stairs (i.e. ascent fall). As such, falls involving stairs can be documented in three separate categories: ‘tumble’ falls, low free falls, and falls from standing height. Furthermore, there are a myriad of variables to account for with this fall type due to the nature of the stairs involved. These include: human biological characteristics (e.g. age, sex, body mass); human behavioral characteristics (e.g. intoxication, mobility, speed, gait, dual-tasking); stair design and configuration (e.g. height and width, handrail, shape, material); stair maintenance (e.g. broken stairs/rails, loose carpet, objects obscuring stairs) and stair exposure to the weather (e.g. outdoor stairs exposed to rain/snow) [1–4]. Within this context, there is thus an anecdotal assumption that skeletal BFT resulting from this fall mechanism will vary with each fall event and, as such, there is no unique skeletal fracture pattern/morphology associated with a fall involving stairs.

Forensic Sci Med Pathol

Understanding the skeletal trauma that results from a fall involving stairs is important in cases where skeletonized, or differentially preserved remains, exhibit BFT and the mechanism that caused that trauma is unclear or unknown. In such cases, forensic anthropologists and pathologists need to include/exclude the possibility of a fall involving stairs being the mechanism of that trauma. Examples of these situations may include cases where an individual is found deceased in close proximity to a stairwell; an individual is found deceased lying mid-way down a staircase with an object (e.g. baseball bat) nearby; or, as presented in Madea et al.’s [5] case study, cases where an individual is found deceased at the base of a staircase. A comprehensive understanding of the skeletal trauma that results from falls involving stairs, both in terms of the fracture patterns as well as the specific fracture morphologies, will assist practitioners to address questions pertaining to these types of situations where a fall involving stairs is suspected to have contributed to the death. There has been little research which has examined skeletal trauma resulting from falls involving stairs in forensic anthropology contexts [6]. Rather, current understandings of the injuries resulting from these falls are documented from forensic pathology and clinical medicine perspectives. In the context of informing injury prevention in public health and medico-legal contexts, forensic pathology has reported the injuries resulting from falls involving stairs as either Injury Severity Scores (ISS) [7], or general descriptions of anatomical regions that refer to both the hard and soft tissue trauma (e.g. upper extremity) [8, 9]. The injury pattern identified from these studies is that the complete body is susceptible to trauma in a fall that involves stairs, but that there is a particular prevalence for head injuries; including fracturing of all regions of the skull. Whilst these injury patterns provide anthropologists with an overview of the distribution and severity of trauma, they afford very little useable reference material on skeletal trauma detail. Clinical medicine, however, has specifically investigated the skeletal fractures resulting from falls involving stairs in juveniles and adults. In cases involving juveniles, skeletal trauma research has focused on differentiating the BFT resulting from stair falls with non-accidental injury [10]; the BFT that may result from a fall whilst in the arms of a caregiver [11], and identifying the severity and patterning of fractures with falls down stairs [12]. In the first two studies, fracturing was found to predominantly occur in the lower extremities, whilst the third study, which investigated falling down stairs, found fracturing could occur to all of the appendicular skeleton, and even occasionally the skull. In studies investigating both adults [13] and a combination of adults and children [14], fracturing was found to primarily occur in the appendicular skeleton, particularly the bones comprising the ankle and wrist joints, when the fall involving stairs did not prove fatal. Both of these clinical studies provide valuable

insights into skeletal fracture patterns and morphologies. However, as they detail non-fatal falls, the findings should be used cautiously in medico-legal contexts, as different skeletal fractures may result when the fall proves fatal. The lack of detailed understanding of the skeletal BFT resulting from fatal falls involving stairs means the current anthropological interpretation of skeletal fractures that may result from this fall type, remains anecdotal. To augment this deficit, this study aims to investigate the types of fracture patterns and morphologies, if any, that result from fatal falls involving stairs.

Methodology Individuals who died from falls involving stairs were retrospectively identified using the National Coronial Information System (NCIS) online database for Australia and New Zealand. A stair fall was classified as any event in which a fall occurred involving one or more stairs. In the state of Victoria in Australia, an individual who dies as the result of a fall, which includes any fall involving stairs, is legally required to be reported to the coroner. Upon reporting the death, all details of that individual’s death, as obtained from the forensic pathologist, toxicologist, policeman and coroner, are documented in the NCIS. Simultaneously, the deceased individual is admitted to the Victorian Institute of Forensic Medicine (VIFM) where, on admission, they undergo a routine full-body postmortem computed tomography (PMCT) scan [15]. A retrospective NCIS search for individuals who died between 2005 and 2014 from a fall that involved stairs in the state of Victoria, Australia, identified 102 cases [16]. This sample size was supplemented with an additional 23 cases identified from the VIFM internal Case Management System for cases from 2014 to 2017. The variables known to affect how an individual falls when stairs are involved (e.g. [1, 3, 17]), and thus influence how their skeleton will fracture, were recorded for each case from the aforementioned coronial documentation (Table 1). Of the 125 cases of fatal falls involving stairs, only 57 cases (46%) had all the variables detailed in Table 1 documented and were thus deemed appropriate for use in this study. This sample size reduction was largely due to the absence of postmortem toxicology being undertaken; as many individuals did not die immediately after they experienced the fall. Although including this variable decreased the sample size, it was still considered necessary to account for it given psychoactive drugs have been considered a contributing factor to falls occurring on stairs [8, 19, 20]. Furthermore, no studies investigating falls involving stairs have examined toxicology across their full sample before and so, to date, the role this variable

Forensic Sci Med Pathol Table 1

Variables recorded from the National Coronial Information System (NCIS)

Variable

Categories

Sex

Male ≤ 77 years

Female ≥ 78 years

Underweight (≤ 18.4) Less than half a flight (≤ 8 stairs) Present Present Deformable (soft)

Normal weight (18.5–24.9) More than half a flight (≥ 9 stairs) Absent Absent Non-deformable (hard)

Over/obese weightb (≥ 25)

Accident

Suicide

Homicide

Age a Body mass index (BMI) Number of stairs involved c Psychoactive drugs d Pre-existing conditions e Landing surfacef Manner of the fall a

Undetermined

Division of age was based on the population medium for the purpose of statistical analyses

b

Overweight and obese weights were grouped together for the purpose of statistical analyses

c

In Australia, the standard flight of stairs comprises 18 stairs [18]

d

This includes ethanol

e

These comprise one, or a combination of, the following: physical impairment, mental impairment and/or emotional trauma

f

Deformable (soft) refers to surfaces such as water and snow that can attenuate some of a body’s energy on impact, and non-deformable (hard) refers to surfaces such as concrete, brick, grass or wood that cannot attenuate a body’s energy on impact

plays in contributing towards injury with this particular type of fall, still remains unclear. Skeletal trauma for each of the 57 cases was analyzed using the VIFM PMCT scans. The scans from 2005 to 2009 were acquired using a Toshiba Aquillion 16® (Helical 16 rows, typical tube voltage 120kVp), and the scans from 2009 to 2017 were acquired using a SOMATOM® Definition Flash, Siemens Healthcare (Helical 128 rows, tube voltage 140kVp). Each individual’s Digital Imaging and Communications in Medicine (DICOM) dataset comprised one small field-ofview (FOV) scan of the head and neck at 1 mm contiguous slice thickness, and one large FOV scan for the full-body at 2 mm contiguous slice thickness. The DICOM data was reviewed using the Siemens Healthcare Syngo.via® VB10B multimodality image rendering software. Documentation of fracture patterns and morphologies were undertaken in accordance with criteria outlined in Wedel and Galloway [21], the Scientific Working Group of Forensic Anthropologists [22] and Buikstra and Ubelaker [23]. Fracture patterns were recorded according to the skeletal element and anatomical location traumatised. For recording fracture morphologies, the orthopaedic classfication, or morphological description of the fracture, were documented. Each skeletal element was reviewed for trauma in the axial, coronal and sagittal anatomical planes, as well as volume renders where necessary. Skeletal trauma was recorded as binary (present or absent) for the purpose of statistical analyses. For quality assurance purposes, thirty cases were randomly selected and cross-examined by a forensic pathologist. Skeletal trauma that resulted from poor bone health and or medical intervention was also documented to ensure the trauma resulting from the actual fall event was appropriately differentiated. A spinal Hounsfield Unit (HU) attenuation profile

was recorded for each individual from a 3 – 4cm3 region of the third lumbar vertebral body [24]. This was to ensure the individuals spinal x-ray photon attenuation approximately correlated with their chronological age. A Spearman Rank test was employed to assess this correlation and to highlight any widespread bone-based pathologies. Statistical analyses were investigated using the Statistical Package for the Social Sciences (SPSS) version 23. Descriptive statistics were employed to identify fracture patterns according to the 42 general skeletal regions1 and to detail all fracture morphologies. For the purpose of statistical interrogation, fracture patterns were investigated in the five general anatomical regions (i.e. skull and cervical, thoracic cavity, upper extremities, lumbar and pelvic girdle, and lower extremities); axial and appendicular regions, and the specific sections of the skull/cervical (i.e. facial bones, skull vault, skull base and cervical/hyoid). For fracture morphologies, only those with frequencies of greater than ten were statistically investigated. Statistical analyses of fracture patterns and morphologies were undertaken separately but employed the same approach. Fisher’s exact tests identified the variables that were significantly associated with fracturing at p = ≤ 0.25 [25]. Significant variables were then included in either simple or multiple (with Backward Wald stepwise) logistic regression analyses,

1 cranial vault; cranial base; facial skeleton; cervical vertebrae (including hyoid bone); right clavicle, scapula, humerus, radius, ulna, carpals, metacarpals and phalanges; left clavicle, scapula, humerus, radius, ulna, carpals, metacarpals and phalanges; left ribs; right ribs; sternum; thoracic vertebrae; lumbar vertebrae; sacrum; left os coxa, femur, patella, tibia, fibula, tarsals, metatarsals and phalanges; right os coxa, femur, patella, tibia, fibula, tarsals, metatarsals and phalanges.

Forensic Sci Med Pathol

depending on the number of significant variables, to create predictive models for skeletal trauma at p = ≤ 0.05. The Victorian Institute of Forensic Medicine Research Advisory and Ethics Committees granted approval (EC7/ 2015) for this research.

Table 2 Biological characteristics and circumstances surrounding the falls examined in this study Frequency Number

Percentage (%)

≤ 77 years

28

49.1

≥ 78 years Sex

29

50.9

Male Female

31 26

54.4 45.6

2 18

3.5 31.6

Over/obese weight (≥ 25) Number of stairs involved

37

64.9

Less than half a flight (≤ 8 stairs) More than half a flight (≥ 9 stairs) Psychoactive drugs

39 18

68.4 31.6

Present Absent Pre-existing conditions Present Absent

29 28

50.9 49.1

52 5

91.2 8.8

Landing surface Deformable Non-deformable Manner of the fall

0 57

0 100

Accident Suicide Homicide Undetermined

57 0 0 0

100 0 0 0

Age

Results The circumstances of the 57 fatal falls involving stairs included both mechanical and non-mechanical events that comprised: ‘tumbling’ down stairs (i.e. a series of short falls); free falling; tripping up the stairs (i.e. falling forwards), and slipping/sliding down stairs (i.e. falling backwards). The biological characteristics of the individuals in this study, and the circumstances surrounding their fall events, are detailed in Table 2. The Spearman Rank correlation showed a predicted moderate negative relationship (r = − 0.443, p = 0.001) between chronological age and the spinal HU attenuation profile, indicating chronological age was appropriate to use as a variable. Skeletal trauma occurred in 47 (82.5%) of the falls involving stairs. Of those 47 cases, 27.7% (n = 13) exhibited trauma to a single skeletal region and 72.3% (n = 34) exhibited poly-trauma.

Skeletal fracture patterns The frequency and distribution of skeletal trauma in cases of falls involving stairs showed that the axial skeleton, particularly the skull, was primarily involved (Fig. 1). Statistical interrogation of the five anatomical regions of the skeleton identified that, although there were a number of statistically relevant variables found in the univariate analyses, when contextualized though logistic regression, there was no significant association between fracturing of these skeletal regions and the variables that influence a fall involving stairs. Similarly, fracturing of only the axial skeleton (n = 35), and fracturing of only the appendicular skeleton (n = 3), showed no significant association with the variables. Fracturing of both the axial and appendicular skeleton (n = 9) however, was significantly less likely to occur if an individual was overweight (n = 4) than if an individual was underweight (n = 1) (OR = 0.024, p = 0.042, 95%CI = 0.001–0.880). Given these small sample sizes however, this finding is unlikely to be a true representation of the relationship between skeletal fracturing and BMI. As such, this result has not been considered further. The high frequency of fractures of the skull and cervical vertebrae instigated further statistical interrogation. No significant association was found between the variables and skeletal trauma of the face, despite univariate analyses identifying a significant association (p = 0.027) with sex, nor between the variables and skeletal trauma of the cranial vault. Significant

BMI Underweight (≤ 18.4) Normal weight (18.5–24.9)

associations were, however, identified with the cranial base and cervical vertebrae. Logistic regression showed fracturing of the cranial base was less likely to occur in ‘younger’ (≤ 77 years) individuals than ‘older’ (≥ 78 years) individuals (OR = 0.29, p = 0.027, 95%CI = 0.1–0.87), and that cervical vertebrae fractures were more likely to occur if an individual’s fall involved ≥ 9 stairs compared with ≤ 8 stairs (OR = 5.57, p = 0.016, 95%CI = 1.37–22.65).

Skeletal fracture morphologies A total of 56 fracture morphologies were identified. Each of these fracture morphologies are detailed in the Supplementary Table and select visual examples are provided in Fig. 2 as reference guides. Descriptions and details of these fractures may be found in anthropology [21] and orthopaedic [26] skeletal trauma reference texts. Of the 56 fracture morphologies considered in this study, 29 (52%) occurred only once, 25

Forensic Sci Med Pathol Fig. 1 Schematic illustrating the distribution and frequency of skeletal fractures (by the 42 regions) between falls involving half a flight of stairs or less (≤ 8 stairs) and falls involving more than half a flight of stairs (≥ 9 stairs)

(44%) occurred between two and ten times, and two (4%) occurred more than ten times. Both fracture morphologies that occurred more than ten times (i.e. single linear and diastatic fractures) were located in the skull. Statistical investigation of these two fractures revealed that there was no significant association between linear skull fractures and the variables, despite univariate analyses identifying an association (p = 0.028) between this fracture morphology and BMI. Diastatic fractures however, were significantly less likely to occur in the ‘older’ (≥ 78 years), than ‘younger’ (≤ 77 years), individuals (OR = 0.08, p = 0.018, 95%CI = 0.01–0.65) (Fig. 3).

Discussion Skeletal trauma resulting from falls involving stairs is arguably one of the most complex and diverse fall types from which to analyze and interpret trauma, in comparison to free falls and falls from standing heights, and this is due to the complex nature of the mechanism. This complexity may be attributed to having to account for both the standard variables that influence a fall, as well as the various ways an individual interacts with the stairs during a fall, and that the fall may involve either falls from standing height, free falling or

tumbling. Subsequently, as has long been anecdotally discussed, there are a myriad of ways an individual’s skeleton may fracture with this fall type. Stairs require complex biomechanical and motor control processes to ascend and descend [4]. When there is a reduction in an individual’s motor control there is an increased chance that the individual will slip, trip or miss-step on the stairs whilst they are descending or ascending and, subsequently, experience a fall [4]. Motor control is often compromised if an individual is elderly, under the influence of psychoactive drugs (e.g. ethanol) and/or if they are suffering health conditions. These three variables have all previously been associated with falls involving stairs [7, 8, 17, 20] and were variables which were all found to be prevalent in this study. The high numbers of elderly individuals in this population may be attributed to the visual, sensory and balance impairments that Startzell et al. [27] has noted places elderly individuals at risk of experiencing a mechanical fall when negotiating stairs. Bone biology also needs to be considered when interpreting skeletal trauma. The decrease in bone density and increase in bone rigidity with age (e.g. osteopenia/osteoporosis), means elderly individuals are more easily susceptible to fracturing [28, 29]. Thus, when the elderly experience a low-energy trauma event, such as a fall involving stairs, the resulting injuries may be more on par with the trauma a young

Forensic Sci Med Pathol

Fig. 2 Examples of the different fracture morphologies identified in this study. These fractures highlight the diversity in skeletal trauma resulting from fatal falls involving stairs. a volume rendered skull (posterior-lateral view) showing a depressed fracture with radiating linear fractures; b axial reconstruction of the cranium showing a depressed fracture of the right maxilla (red arrow); c volume rendered skull (left lateral view) showing a single linear fracture (red arrow); d volume rendered skull (inferior view) showing a hinge fracture (red arrows); e axial reconstruction of a thoracic

vertebra showing a fracture of the transverse process (red arrow); f sagittal reconstruction of the skull and cervical vertebrae showing an odontoid fracture (red arrow); g sagittal reconstruction of the thoracic vertebrae showing a slice fracture of the vertebral body (red arrow); h volume rendered right clavicle (anterior view) showing comminuted fracturing of the lateral aspect (red arrow), i coronal reconstruction of the left patella showing a vertical fracture (red arrow)

individual would experience from a high-energy impact [30]. The presence of psychoactive drugs in an individual’s system, often medications in the elderly (e.g. [31]) and consumption of ethanol in younger individuals [32], is likely to contribute to a fall event occurring as these drugs may impair an individual’s balance, which presents a hazard when navigating stairs. The high prevalence of individuals with poor health in this study is likely attributed to the high number of elderly individuals in this study. As the elderly are predisposed to many pre-existing physical and mental conditions that develop with age, many of the individuals in this study were susceptible to experiencing a healthprecipitated non-mechanical fall event.

No significant association was identified between the variables of psychoactive drugs and pre-existing health conditions and skeletal fracturing in this study. The high prevalence of these variables in this cohort however, suggests they likely contributed, in some capacity, to an individual experiencing the fall event, but perhaps were actually protective of skeletal fracturing (i.e. trauma). The effects of psychoactive drugs, in particular ethanol, have been found by Friedman [33] to be protective against mortality in cases of traumatic injury, and trauma as a direct consequence of syncope, which are typically associated with pre-existing physical health conditions, has been found by Bartoletti et al. [34] to also rarely result in mortality.

Forensic Sci Med Pathol Fig. 3 Examples of diastatic fractures found in this study. These examples highlight the potential morphological variation of this particular fracture type. a volume rendered skull (posterior view) showing bilateral diastatic fractures of the lambdoidal suture (red arrows), b sagittal reconstruction of a skull showing a diastatic fracture of the lambdoidal suture (red arrow), c axial reconstruction of the cranium showing a diastatic fracture of the zygo-temporal suture (red arrow), d volume rendered skull (left posteriorlateral view) showing a diastatic fracture of the left lambdoidal suture (red arrow)

Skeletal fracture patterns The multiple points of impact an individual typically experiences during a fall that involves stairs may be why poly-trauma was so highly prevalent in this, and similar [8, 9], studies. The distribution and frequency of skeletal fractures from this cohort show that, regardless of how many stairs are involved in the fatal fall, trauma is primarily to the axial skeleton and that skeletal trauma is fairly evenly distributed across the number of stairs fallen. This finding suggests that there is no particular difference between fracturing from a fall that was during an ascent (i.e. falls only involving half a flight of stairs or less) and falls that were during a descent (i.e. all the falls involving more than half a flight of stairs, as well as some that were less than half a flight); a contrast to Roys [2] and Templar [35]. The overall fracture pattern of BFT primarily to the skull and thoracic regions, with little fracturing of the pelvis and extremities, is a similar pattern to that found in other investigations of fatal stair fall injuries [8, 19]. This skeletal fracture pattern does not, however, extend to the soft tissue injury pattern; Behera et al. [9] found that the frequency of injuries to the extremities is greater than that to the thoracic region when accounting for soft tissues. Nor does this fracture pattern extend to falls involving stairs that did not result in fatality. Clinical studies by Boele van Hensbroek et al. [14] and Mitchell et al. [13] both identified a high frequency of upper

and lower extremity fractures in individuals who survived. These studies found that fractures most commonly occurred in the bones comprising the wrists and ankles, with very few fractures [14], or even no fractures [13], occurring in the axial skeleton. The completely contrasting fracture pattern between nonfatal falls involving stairs [13, 14] and the skeletal trauma identified in this study, highlights that there is actually a distinct skeletal fracture pattern associated with fatal falls involving stairs. Whilst the fractures presented in this study are in no way a confirmation that these specific fractures caused, or directly contributed to, the death (i.e. some of these deaths resulted from medical events such as cardiac arrest and were completely unrelated to skeletal injury), this study suggests that fatal injuries are primarily associated with trauma to the axial region. This finding has been noted in other low-energy fatal falls [36] and supports Rowbotham and Blau’s [6] discussion that, because the axial skeleton houses the bodies’ vital organs, damage to this region may be associated with damage to the body’s vital organs and thus fatality. The disparity in the frequency of fractures between fatal and non-fatal cases and the upper extremities, may be associated with injuries resulting from falls onto the outstretched hand/s (FOOSH). The high prevalence of FOOSH related trauma in the non-fatal cases [13, 14], compared with the very minimal (n = 4) prevalence of FOOSH in the fatal cases in this

Forensic Sci Med Pathol

study, suggest that many individuals who died did not attempt to, or were unable to, arrest their fall. Consequently, the brain and vital organs in the torso would have been more exposed to the force on impact. Possible scenarios for the lack of FOOSH may include non-mechanical fall events (e.g. influenced by pre-existing health conditions during the fall, (e.g. [37]), or slow reflex times in the elderly [38]. The disparity in the frequency of lower extremity fractures between non-fatal and these fatal cases is potentially attributed to the way an individual lands or moves. Luetters et al. [39] found 25% of foot fractures in individuals over 45 years were attributed to falling whilst ascending or descending stairs, with individuals tripping on the stairs or twisting/turning their foot. There are a variety of situations that may result in these movements, and most of these falls would be fairly minor and thus unlikely to prove fatal. The frequency and distribution of fractures in this study suggest there was an association between axial skeletal trauma and fatal falls involving stairs, however this pattern was not statistically significant. Although a similar fracture pattern of primarily axial trauma has been identified in other investigations of fatal falls involving stairs [9, 19], these other studies have reported only descriptive or univariate statistical analyses and do not support this finding with detailed contextual statistical investigations. Thus, despite the fairly distinct pattern of axial skeletal trauma, this association with falls involving stairs is not significant and so confirms, at present, the current anecdotal understanding that there is no unique skeletal fracture pattern involving the full body with this fall type. The skull has been identified as the skeletal region most commonly fractured from falls involving stairs [6]. Presence of fractures to all aspects of the skull suggest the impacts may have come from any direction and that there was no unique position the head was in when impacted during these falls. The high frequency of skull vault fractures, and in the context of fracture morphologies such as that illustrated in Fig. 2a, suggests the hat brim line (HBL) rule, proposed by Kremer et al. [40] and Kremer and Sauvageau [41], is not a methodology that is applicable to cases of fatal falls involving stairs. The HBL refers to an area of the skull located between two lines that run parallel to the Frankfurt horizontal plane; the superior line passes through the glabella and the inferior line runs through the external auditory meatus [40, 41]. The rule suggests that fractures resulting from falls are located within the HBL, whilst fractures resulting from blows may be found above, but also within, the HBL; thus the rule is nondiscriminatory for fractures resulting from falls [40, 41]. The location of skull fractures in this study involved all aspects of the skull, that is both within and above the HBL; a finding that has also previously been reported by Bux et al. [19], Behera et al. [9] and Preuß et al. [8]. As these fractures are not exclusively located within the HBL, this finding further supports Kremer and Sauvageau [41] and Kremer et al.’s [40]

conclusion that the location of fractures resulting from falls relative to the HBL is non-discriminatory. All fracturing of the facial skeleton occurred in the middle third of the face (see Supplementary Table). Fractures in this particular region are characteristic of a direct impact and Roccia et al. [42] has found that this region fractures in 65.9% of cases regarding falls specifically involving stairs. This indicates individuals in this study were not landing on, and thus impacting, their chin (i.e. lower third of the face) or forehead (i.e. upper third of the face). Rather, in falls where stairs are involved, individuals most commonly land directly onto the center of their face. The finding that fracturing of the cranial base was less likely to occur in the ‘younger’ (≤ 77 years) half of the population may be attributed to the individual’s landing position and/or bone biology. In this study many of the ‘older’ individuals fell backwards whilst ascending stairs. Biomechanically, free falls such as these have been found by Freeman et al. [43] to likely result in ‘inverted’ landing positions. The nature of the inverted landing position in these cases means the primary impact is to the poster aspect of the head, and this has specifically been associated with basal skull fractures [44]. With regards to bone biology, Lillie et al.’s [45] study found an association between cortical thinning of the occipital bone and increasing age, particularly in females, and that this bone had one of the highest percentages of cortical thinning compared to the other cranial bones. This association suggests the cranial base is more susceptible to fracturing than other aspects of the skull as an individual becomes older, and thus is possibly why skull base fractures were less likely to occur in the ‘younger’ (≤ 77 years) individuals rather than the ‘older’ (≥ 78 years) individuals in this study. The finding that the cervical vertebrae were more likely to fracture if the fall involved more than half a flight of stairs may be attributed to biomechanics. When ≥ 9 stairs are involved, an individual is more likely to land in the inverted position. When impacting the ground/step in this position, an individual’s head becomes stationary while their body’s inertia continues, causing the torso and extremities to continue moving, which causes additional force loading to the neck in the forms of hyperextension, hyperflexion and/or axial compression of the cervical vertebrae [46]. Fractures of the cervical vertebrae have been associated with these trauma mechanisms in other types of fatal low-energy falls in the elderly [36], but have not previously been associated with falls involving stairs. To date, studies of fatal falls involving stairs have found cervical vertebrae fractures to be more rare occurrences than head trauma [6], although if any of the vertebral spine has been found fractured, it has typically been in relation to the cervical [8, 9].

Skeletal fracture morphologies The variation in skeletal fracture morphologies resulting from these falls involving stairs highlight the many ways these falls

Forensic Sci Med Pathol

may occur, the different forces involved, and the various directions of impact the body may be subjected to. Some general observations on these fracture morphologies (refer to the Supplementary Table) are discussed below. Skull fractures are diverse and range from low-energy single impact linear fractures of the vault/base, through to high-energy ring fractures that result from compression of the spine onto the skull base [47]. The severity and diversity of these fracture morphologies illustrate the variety of positions the head, relative to the body and the stairs, may be in which may lead to fracturing. Types of vertebral fractures were also diverse and, biomechanically, indicate there are a variety of landing positions associated with these falls that include: hyperextension, hyperflexion, axial compression and direct blows to the posterior aspect [21]. Thoracic cavity fracture morphologies illustrate that all aspects of the rib cage, including the sternum, are susceptible to fracturing. These fractures were primarily the result of the chest directly (i.e. transverse fractures) [48] impacting an object (e.g. stair, banister, wall); suggesting this morphology may be more characteristic of the ‘tumble’ fall down stairs. Upper extremity fracture morphologies indicate that trauma of the shoulder was typically from a direct impact and that FOOSH cases were rare; findings which support the descriptive fracture pattern identified. These fractures may possibly be attributed to falls/slips backwards on stairs, or the ‘tumble’ motion where the shoulder region may be impacted more often than the actual arm/wrist. Pelvic and lower extremity fracture morphologies indicate that falls with impacts to the lateral aspect of the hip occurred only occasionally. This is in contrast to fractures of the pelvis and femur in cases of other low-energy falls [36, 49]. Fracture morphologies identified in this study also indicated both low- and high-energy forces occur in falls involving stairs, and that both may occur in any region of the skeleton. High-energy fractures included depressed with radiating linear and concentric fractures, comminuted clavicle fractures and vertebral body slice fractures. Some of the high-energy cranial fractures specifically (e.g. ring and hinge fractures), occurred in the comparatively younger individuals (i.e. those in their mid-fifties); suggesting fracturing actually resulted from a high-energy force, and not because an individual was elderly and their bone rigidity may have caused high-energy fractures to result from low-energy forces [30]. Despite the fact that the falls in this study were fatal, the presence of low-energy fracture morphologies (e.g. Colles fractures and vertebral transverse process fractures) highlight that the majority of the fractures resulting from falls involving stairs would not be directly associated with the cause of fatality (refer to Supplementary Table). Similarly, Wyatt et al. [7] found only 7.8% of their fatal falls down stairs exhibited trauma that was considered ‘incompatible’ with life. Of the 56 fracture morphologies identified in this study, diastatic skull fractures were the only type found to have a

significant association with fatal falls involving stairs. Diastatic fractures, that is linear fractures which radiate though the sutures of the cranium to widen or separate the cranial bones [50], may result from both low- and high-energy forces (see Fig. 3). Sharkey et al.’s [51] biomechanical simulation study investigating the forces required to fracture the skull, found that diastatic fractures were always the first type of fracture to result from an impact as they occurred under the least amount of force (starting at ≤ 3, 340 N), and thus their frequency would increase as the newtons increased. Subsequently, diastatic fractures may occur from both lowand high-energy forces and, as such, it is a fracture morphology that is likely to be prevalent in a fall type such as this, where the fall may involve either low- or high-energy. The presence of diastatic fractures in an adult skull has been found to occur in association with falls, and implies an individual has experienced a severe head injury [52]. The significant association of this fracture being less likely to occur in the older individuals (≥ 78 years) may be attributed to the obliteration of skull sutures with increased age. As a diastatic fracture is only possible if a suture is present, and as ectocranial sutures obliterate as an individual’s age increases, diastatic fractures may thus be more commonly associated with younger individuals who still have sutures present. The obliteration of ectocranial sutures has long been considered an indicator of age estimation for adult individuals, even though it is a highly controversial method [53, 54] that should be applied cautiously. The general consensus on the chronological ages at which ectocranial skull suture obliteration should have occurred are, based on the work of Meindl and Lovejoy [55] and more recently Donato et al. [54], between the late 60s to 80 years old. As the ‘younger’ individuals in this study are referring to individuals who were 77 years or younger, it is probable that these ‘younger’ individuals still had cranial sutures present, which means it was possible for diastases to occur. Whilst sutures in all of the ‘older’ individuals (≥ 78 years) would most probably have been completely obliterated, or close to, and so diastases may have not physically been possible.

Limitations The greatest limitation to this study was the small sample size (n = 57). The reduced sample size was deemed necessary in order to account for all the variables that are known to influence falls involving stairs from the coronial data, and thus provide the most accurate and contextual investigation of skeletal trauma possible. An additional limitation to using this coronial data was the inability to account for how each individual actually fell (i.e. their entire fall event) and obtain details such as the design of the stairs. Furthermore, a number of the PMCT scans were incomplete, as the elbows or feet were outside the CT scan FOV. This was due to the position the body was in when scanned and/or because the individual was

Forensic Sci Med Pathol

too tall or wide to sufficiently fit within the CT FOV. Thus it is acknowledged that some extremity skeletal fractures may not have been accounted for.

Research Training Program Scholarship who funded her doctoral research, for which this paper was a component.

Funding Australian Government Research Training Program Scholarship.

Conclusion Anecdotally, it has long been considered that there is no specific skeletal fracture pattern or morphology characteristic of a fatal fall that involves stairs, due to these fall events being so varied in terms of how the individual falls, and the nature of the stairs involved in the fall. Results of this study show that, whilst the axial skeleton is the region most frequently fractured in falls involving stairs, almost all areas of the skeleton are susceptible to fracturing. Although the fracture pattern of axial trauma was not statistically significant, fracture patterning identified that the cranial base was less likely to fracture in the younger individuals, and that the cervical vertebrae were more likely to fracture if more than half a flight of stairs was involved in the fall. An analysis of fracture morphologies (n = 56) identified that a wide range of low- and high-energy fractures are possible, and that diastatic fractures of the skull are more likely to occur in ‘younger’ (≤ 77 years) individuals. This detailed and contextual examination of skeletal trauma confirms the current anecdotal understanding that there is no specific skeletal fracture pattern/morphology characteristic of all falls involving stairs. Rather, there are a number of fracture patterns, and one fracture morphology, that are associated with this fall type, and these should be taken into consideration when examining remains with BFT.

Compliance with ethical standards Conflict of Interest The authors declare that they have no conflict of interest.

References 1.

2. 3. 4.

5. 6. 7. 8.

9.

10.

Key points 1. The skeletal BFT that results from falls involving stairs is complex due to the different types of falls (e.g. ‘tumble’ falls, low free falls and falls from standing height) that may potentially be involved. 2. There are a number of possible fractures characteristic of falls involving stairs. 3. The axial skeleton, particularly the skull, was primarily fractured in these falls. 4. The cranial base and cervical vertebrae were significantly associated with falls involving stairs. 5. There were 56 types of fractures identified in this study. Of these, diastatic fractures were characteristic of falls involving stairs.

11.

12.

13.

14.

15.

16. Acknowledgements The authors wish to thank the Victorian Institute of Forensic Medicine for providing access to the PMCT data; the National Coronial Information System for their assistance in attaining the relevant cases, and the manuscript reviewers for their constructive comments. Author SKR would like to acknowledge the Australian Government

17.

Scott A. Falls on stairways - literature review. Report No. HSL/ 2005/10. 2005. http://www.hse.gov.uk/research/hsl_pdf/2005/ hsl0510.pdf. Accessed 15 May 2017. Roys MS. Serious stair injuries can be prevented by improved stair design. Appl Ergon. 2001;32:135–9. Jackson PL, Cohen HH. An in-depth investigation of 40 stairway accidents and the stair safety literature. J Saf Res. 1995;26:151–9. Jacobs JV. A review of stairway falls and stair negotiation: lessons learned and future needs to reduce injury. Gait Posture. 2016;49: 159–67. Madea B, Dettmeyer R, Musshoff F. Fall downstairs: accident, homicide or natural death? Forensic Sci Med Pathol. 2008;4:122–8. Rowbotham SK, Blau S. Skeletal fractures resulting from fatal falls: a review of the literature. Forensic Sci Int. 2016;266:582.e1–e15. Wyatt JP, Beard D, Busuttil A. Fatal falls down stairs. Injury. 1999;30:31–4. Preuß J, Padosch SA, Dettmeyer R, Driever F, Lignitz E, Madea B. Injuries in fatal cases of falls downstairs. Forensic Sci Int. 2004;141:121–6. Behera C, Rautji R, Dogra TD. Patterns of injury seen in deaths from accidental falls down a staircase: a study from South Delhi. Med Sci Law. 2009;49:127–31. Pierce MC, Bertocci GE, Janosky JE, Aguel F, Deemer E, Moreland M, et al. Femur fractures resulting from stair falls among children: an injury plausibility model. Pediatrics. 2005;115:1712– 22. Pennock AT, Gantsoudes GD, Forbes JL, Asaro AM, Mubarak SJ. Stair falls: caregiver's "missed step" as a source of childhood fractures. J Child Orthop. 2014;8:77–81. Docherty E, Hassan A, Burke D. Things that go bump... bump... bump: an analysis of injuries from falling down stairs in children based at Sheffield Children's Hospital. Emerg Med J. 2010;27:207– 8. Mitchell SE, Aitken SA, Court-Brown CM. The epidemiology of fractures caused by falls down stairs. ISRN Epidemiology. 2013; https://doi.org/10.5402/2013/370340. Boele van Hensbroek P, Mulder S, Luitse JSK, van Ooijen MR, Goslings JC. Staircase falls: high-risk groups and injury characteristics in 464 patients. Injury. 2009;40:884–9. O'Donnell C, Rotman A, Collett S, Woodford N. Current status of routine post-mortem CT in Melbourne, Australia. Forensic Sci Med Pathol. 2007;3:226–32. Rowbotham SK, Blau S. The circumstances and characteristics of fatal falls in Victoria, Australia: a descriptive study. Aus J Forensic Sci. 2017;49:403–20. Ozanne-Smith J, Guy J, Kelly M, Clapperton A. The relationship between slips, trips and falls and the design and construction of buildings. Monash University Accident Research Centre; 2008.

Forensic Sci Med Pathol 18.

19. 20. 21. 22.

23.

24.

25. 26.

27.

28.

29.

30. 31.

32.

33. 34.

35. 36.

Australian Building Codes Board. Building code of Australia class 1 and class 10 buildings. In: National construction code -volume 2. 2016. http://www.abcb.gov.au/Resources/Publications/NCC/NCC2016-Volume-Two. Accessed 20 June 2017. Bux R, Parzeller M, Bratzke H. Causes and circumstances of fatal falls downstairs. Forensic Sci Int. 2007;171:122–6. Ragg M, Hwang S, Steinhart B. Analysis of serious injuries caused by stairway falls. Emerg Med. 2000;12:45–9. Wedel VL, Galloway A. Broken bones: anthropological analysis of blunt force trauma. 2nd ed. Charles C. Thomas: Springfield; 2014. Scientific Working Group of Forensic Anthropologists. Trauma analysis. 2011. http://swganth.startlogic.com. Accessed 15 Sept 2014. Buikstra J, Ubelaker D. Standards for data collection from human skeletal remains: proceedings of a seminar at the Field Museum of Natural History. Arkansas Archaeological Survey: Arkansas; 1994. Celenk C, Celenk P. Bone density measurement using computed tomography. In: Saba L, editor. Computed tomography - clinical applications. Croatia: InTech; 2012. p. 123–36. Mickey RM, Greenland S. The impact of confounder selection criteria on effect estimation. Am J Epidemiol. 1989;129:125–37. Court-Brown CM, Heckman JD, McQueen MM, Ricci WM, Tornetta P, McKee MD. Rockwood and Green's fractures in adults. 8th ed. Philadelphia: Wolters Kluwer Health; 2015. Startzell JK, Alfred Owens D, Mulfinger LM, Cavanagh PR. Stair negotiation in older people: a review. J Am Geriatr Soc. 2000;48: 567–80. Kieser J. Biomechanics of bone and bony trauma. In: Kieser J, Taylor M, Carr D, editors. Forensic biomechanics. Chichester: Wiley-Blackwell; 2012. p. 36–69. Schuit SCE, van der Klift M, Weel AEAM, de Laet CEDH, Burger H, Seeman E, et al. Fracture incidence and association with bone mineral density in elderly men and women: the Rotterdam study. Bone. 2004;34:195–202. Switzer JA, Gammon SR. High-energy skeletal trauma in the elderly. J Bone Joint Surg Am. 2012;94A:2195–204. Woolcott JC, Richardson KJ, Wiens MO, Patel B, Marin J, Khan KM, et al. Meta-analysis of the impact of 9 medication classes on falls in elderly persons. Arch Intern Med. 2009;169:1952–60. Kool B, Ameratunga S, Jackson R. The role of alcohol in unintentional falls among young and middle-aged adults: a systematic review of epidemiological studies. Inj Prev. 2009;15:341–7. Friedman LS. Dose-response relationship between in-hospital mortality and alcohol following acute injury. Alcohol. 2012;46:769–75. Bartoletti A, Fabiani P, Bagnoli L, Cappelletti C, Cappellini M, Nappini G, et al. Physical injuries caused by a transient loss of consciousness: main clinical characteristics of patients and diagnostic contribution of carotid sinus massage. Eur Heart J. 2008;29: 618–24. Templar J. The staircase: studies of hazards, falls and safer design. Massachusetts: MIT Press Ltd; 1992. Rowbotham SK, Blau S, Hislop-Jambrich J, Francis V. Skeletal trauma resulting from fatal low (≤ 3 m) free falls: an analysis of fracture patterns and morphologies. J Forensic Sci. 2017; https:// doi.org/10.1111/1556-4029.13701.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47. 48. 49.

50. 51.

52. 53.

54.

55.

Yamamoto T, Takasu K, Emoto Y, Shikata N, Matoba R. Case report of death from falling: did heart tumor cause syncope? Int J Legal Med. 2012;126:633–6. Robinovitch SN, Normandin SC, Stotz P, Maurer JD. Time requirement for young and elderly women to move into a position for breaking a fall with outstretched hands. J Gerontol A Biol Sci Med Sci. 2005;60A:1553–7. Luetters CM, Keegan THM, Sidney S, Quesenberry CP, Prill M, Sternfeld B, et al. Risk factors for foot fracture among individuals aged 45 years and older. Osteoporos Int. 2004;15:957–63. Kremer C, Racette S, Dionne CA, Sauvageau A. Discrimination of falls and blows in blunt head trauma: systematic study of the hat brim line rule in relation to skull fractures. J Forensic Sci. 2008;53: 716–9. Kremer C, Sauvageau A. Discrimination of falls and blows in blunt head trauma: assessment of predictability through combined criteria. J Forensic Sci. 2009;54:923–6. Roccia F, Boffano P, Bianchi FA, Zavattero E. Maxillofacial fractures due to falls: does fall modality determine the pattern of injury? J Oral Maxillofac Res. 2014;5:e5. Freeman MD, Eriksson A, Leith W. Head and neck injury patterns in fatal falls: epidemiologic and biomechanical considerations. J Forensic Legal Med. 2014;21:64–70. McElhaney JH, Hopper RH, Nightingale RW, Myers BS. Mechanisms of basilar skull fracture. J Neurotrauma. 1995;12: 669–78. Lillie EM, Urban JE, Lynch SK, Weaver AA, Stitzel JD. Evaluation of skull cortical thickness changes with age and sex from computed tomography scans. J Bone Miner Res. 2016;31:299–307. Nightingale RW, McElhaney JH, Richardson WJ, Best TM, Myers BS. Experimental impact injury to the cervical spine: relating motion of the head and the mechanism of injury. J Bone Joint Surg Am. 1996;78A:412–21. Voight GE, Sköld G. Ring fractures of the base of the skull. J Trauma. 1974;14:494–505. DiMaio VJ, DiMaio D. Forensic pathology. 2nd ed. Boca Raton: CRC Press; 2001. Bakke HK, Dehli T, Wisborg T. Fatal injury caused by low-energy trauma – a 10-year rural cohort. Acta Anaesthesiol Scand. 2014;58: 726–32. White TD, Black M, Folkens PA. Human osteology. 3rd ed. Amsterdam: Academic Press, Elsevier; 2012. Sharkey EJ, Cassidy M, Brady J, Gilchrist MD, NicDaeid N. Investigation of the force associated with the formation of lacerations and skull fractures. Int J Legal Med. 2012;126:835–44. Grossart KWM, Samuel E. Traumatic diastasis of cranial sutures. Clin Radiol. 1961;12:164–70. Hershkovitz I, Latimer B, Dutour O, Jellema LM, Wish-Baratz S, Rothschild C, et al. Why do we fail in aging the skull from the sagittal suture? Am J Phys Anthropol. 1997;103:393–9. Donato L, Cipolloni L, Ozonoff A, di Luca A. A preliminary study of the relationship between obliteration of cranial sutures and age at time of death. Biol Syst. 2016; https://doi.org/10.4172/2329-6577. 1000163. Meindl RS, Lovejoy OC. Ectocranial suture closure: a revised method for the determination of skeletal age at death based on the lateral-anterior sutures. Am J Phys Anthropol. 1985;68:57–66.