Seediscussions,stats,andauthorprofilesforthispublicationat:http://www.researchgate.net/publication/8504118

Gao,C.&Abeysekera,J.Asystemsperspectiveofslipandfallaccidentsonicyandsnowysurfaces.Ergonomics47,573-598

ARTICLEinERGONOMICS·MAY2004

ImpactFactor:1.56·DOI:10.1080/00140130410081658718·Source:PubMed

CITATIONS

40

READS

68

2AUTHORS,INCLUDING:

ChuansiGao

LundUniversity

62PUBLICATIONS411CITATIONS

SEEPROFILE

Availablefrom:ChuansiGao

Retrievedon:29September2015

A systems perspective of slip and fall accidents on icy and

snowy surfaces

CHUANSI GAO* and JOHN ABEYSEKERA

Division of Industrial Ergonomics, Department of Human Work Sciences, LuleaUniversity of Technology, 971 87Lulea, Sweden

Keywords: Slip and fall; Icy and snowy surfaces; Systems approach; Footwear;Gait biomechanics; Psychophysiological factors; Environment; Prevention.

Current research on slips and falls has mainly focused on floors and/orcontaminated floors. Although icy and snowy surfaces near melting temperatureare more slippery, more important still, slip and falls on icy and snowy surfacesinvolve not only outdoor workers, but also pedestrians and the general public;and occur in cold regions and in the winter season in many parts of the world.However, in comparison with the size of the problem, research work done so farin this area has been limited. The objective of this paper is to present a systemsperspective of slip and fall accidents, with special focus on its occurrence on icyand snowy surfaces. In order to explore the aetiology of slip and fall accidentsfurther, and to provide the basis for prevention, the authors put forward asystems model towards the slips and falls on icy and snowy surfaces based on areview of literature and current knowledge. Various contributing factors aresystematically discussed to highlight the multi-factorial nature of the problem,providing the possibility of a multi-faceted approach to reach systematicprevention. Unresolved issues related to slips and falls on ice and snow are alsoidentified, which necessitate further research.

1. Introduction

Long and dark winter characterize Arctic regions. In Nordic countries, winter lastsfor about six months in a year. Due to the Gulf Stream, the weather is changeable. Itis common to see that roads in winter are either covered with snow, ice, meltingsnow, melting ice or mixed icy and snowy surfaces. These natural climatic factorsresult in the prevalence of slip and fall accidents not only among outdoor workers(forest workers, construction workers, service workers, etc.), but also among thegeneral public and pedestrians in Nordic and other cold regions.

Slipping arises when the coefficient of friction (COF) between footwear andwalkway surface provides insufficient resistance to counteract resultant force. Slipsoccur at either push off (toe off, the rearward slip) or touch down (heel strike,forward slip). The forward slip (heel strike) is more dangerous and frequent(Strandberg and Lanshammar 1981, Leamon and Son 1989, Manning et al. 1991).Whenever humans cannot counterbalance these slips, then loss of balance and fallwill occur, which, in turn, may result in various injuries.

*Author for correspondence. e-mail: Chuansi.Gao@arb.luth.se

Ergonomics ISSN 0014-0139 print/ISSN 1366-5847 online # 2004 Taylor & Francis Ltdhttp://www.tandf.co.uk/journals

DOI: 10.1080/00140130410081658718

ERGONOMICS, 15 APRIL, 2004, VOL. 47, NO. 5, 573 – 598

1.1. Slip and fall accidents on icy surfacesLiterature reveals that many pedestrians have been injured by slips and falls onfrozen roads in cold regions. In Nordic countries, 16% of all accidents at work, athome and during leisure activities were caused by slipping, out of which two thirds ofthe slips occurred on ice or snow (Gronqvist and Hirvonen 1992, 1995). Accordingto Kelkka (1995), about 23 000 pedestrian slips on ice and snow lead to injuries everyyear in Finland, 21 000 of them need treatment in emergency stations, while 2000need longer treatment in hospitals. The economic cost due to slipping accidents intraffic areas is 280 million Finnish Marks (53 million USD approx.) every year (costlevel in 1993).In Sweden, thousands of pedestrians are injured every year, because of slippery

pavements and roadways (Gard and Lundborg 1994). Based on Swedish NationalBoard of Occupational Safety and Health statistics data, slip, trip and fall (STF)accidents were most frequently reported on snow and ice (Andersson and Lagerlof1983), both from male and female workers, 25% (women, 5 45 years), 30%(women, 4=45 years), 13% (men, 5 45 years), 18% (men, 4=45 years)(Kemmlert and Lundholm 2001). Ice and cold related injuries (all categories)accounted for 37% of the total cost of all injuries among the elderly in the trafficenvironment during a 1 year period in Sweden. Half of all injuries were fractures.The ‘cost’ of medical care of these slipping injuries was almost the same as the ‘cost’of all traffic injuries in the area during the same time (Lund 1984, Sjogren andBjorstig 1991, Bjornstig et al. 1997). It is common to see crowded orthopaedic clinicsduring winter particularly in the northern part of Sweden. Expenditure for theseinjuries is high compared to other injuries. The healing process and rehabilitationusually also take a long time, resulting in higher insurance payments to the injuredand compensation claims by the workers.Bentley and Haslam (1996, 1998, and 2001) undertook an analysis of accident data

for 1734 outdoor fall accidents to postal delivery employees occurring over a 2-yearperiod in the UK. The analysis examined the activity of the employee at the time ofthe accident, and the fall initiating event (FIE). The most common FIE was found tobe foot slip, with 46% of slips on the level being ice slips. It was shown that slip,tumble, and fall accidents (STFA) were most common during winter months(November/February), and slipping accidents tended to cluster on single days whereheavy snowfall or ice made conditions particularly hazardous. The analysis showedthat common underfoot conditions related to FIE were snow, ice and uneven ordamaged paving, secondly, the use of unsuitable and worn footwear. More than90% of employees investigated mentioned snow or ice as one of the factors whichmost increased their risk of having a fall accident.In the USA, Leamon and Murphy (1995) revealed that there was significantly

higher numbers of falls in winter months (December/March) through analysis ofworkers’ compensation data of a major insurance company, coinciding withexpected periods of snowfall and ice. An inverse correlation was observed for samelevel falls with temperature and precipitation though further conclusive results arecalled for.

1.2. The state of the artThere has been research work done, dealing with the assessment and measurement ofslip resistance of shoes on indoor dry and contaminated floors (Jung and Schenk1990, Leamon 1992a, Jung and Fischer 1993, Leclercq et al. 1994, Gronqvist 1999).

574 C. Gao and J. Abeysekera

Icy and snowy surfaces near melting temperature are more slippery than floors,involving pedestrians and the general public. However, in comparison with the sizeof the problem, research work put into the prevention of slips and falls on icy andsnowy surfaces has not been paid adequate attention. Gronqvist and Hirvonen(1995) pointed out that there was considerable lack of knowledge about theslipperiness of footwear soles on icy surfaces. Since that time, there still have beenonly a few reports available concerning the slip and fall on icy and snowy surfaces(Bjornstig et al. 1997).

A recent literature review summarized friction mechanisms on warm and cold ice.Various factors are involved in the determination of frictions on ice and snow, suchas water film formation, frictional heating and melting, pressure melting, interfacethermal balance, adhesion, penetration and ploughing, capillary suction, electricalforces, molecular forces and surface energies, etc., which suggests the friction of ice isa complex problem that requires further systematic studies (Chang et al. 2001b).

According to Jung’s framework, using a shoe wearing test with 76 pairs of shoes of13 different types on indoor various floors, it was founded that surface roughness,sole material and state of sole (unworn or worn) could be considered as majorinfluencing factors, which explained 50% of the variance found for the acceptanceangle (ramp test). However, the remaining 50% could not be accounted for (Jung1992). Tisserand (1985) argued that there were many factors contributing to slips,including:

(1) Biomechanical and psychological factors involved in human walking,(2) Factors related to the floor surface,(3) Characteristics of shoes and soles.

1.3. ObjectivesThe objectives of this paper are to present a systematic perspective of slip and fallaccidents, with special focus on its occurrence on icy and snowy surfaces, to highlightthe multi-factorial features, to provide a multi-faceted approach towards prevention,to identify unresolved issues and to explore further research needs related to slipsand falls on icy and snowy surfaces.

2. Systems model of slip and fall accident on icy surfaces

In order to further explore the aetiology of slip and fall accidents, and to provide thebasis for prevention, the authors put forward a systems model of the slips and fallson icy and snowy surfaces based on complete literature survey as shown in figure 1.

The model explains following factors, which contribute to slip and fall accidents:

(1) Footwear (sole) properties including sole material, hardness, roughness,worn/unworn, tread (geometry) design, centre of gravity, anti-slip devices,wearability (weight, height, flexibility, ease of walking, comfort), etc.,

(2) Underfoot surface characteristics, covered with ice, snow, contaminants,anti-slip materials, uneven, ascending/descending slope, etc.,

(3) Footwear (sole)/surface interface—tribological aspect, i.e., coefficient offriction (static, transitional and dynamic),

(4) Human gait biomechanics: muscle strength, postural control, musculoskele-tal function, postural reflex and sway, balance capability, acceleration,

575Slip and fall accidents

deceleration, stride length, step length, heel contact velocity, transition of thewhole body centre of mass, vertical and horizontal forces, required COF,etc.,

(5) Human physiological and psychological aspects, i.e., the so-called intrinsicfactors, including decline in visual, vestibular and proprioceptive functions,ageing, perception of slipperiness, information processing, experience,training, diabetes, drug and alcohol usage, unsafe behaviour (rushing,reading while walking) etc.,

(6) Environment (extrinsic factors): temperature, snowfall, Gulf Stream, lightingcondition, etc.

Since it is a multi-facet issue, Twomey et al. (1995) used artificial neural networkpredictive model for slip resistance as a function of six independent variables. Thefollowing section of this paper will discuss the various contributing factorssystematically with special emphasis on slip and fall accident occurrence on icyand snowy surfaces.

3. Footwear properties

Literature regarding slip resistance of footwear on contaminated floors is available(Manning et al. 1985, Jung 1992, Leclercq et al. 1994, Gronqvist 1995b). However,little research has been done concerning anti-slip properties of footwear (sole) for useon icy and snowy surfaces. Haslam and Bentley (1999) revealed that footwear couldbe inappropriate for delivery work in adverse weather conditions. Provision offootwear with increased grip is recommended. A majority of mail delivery employees

1) Footwear

2) Icy and snowy surfaces3) Footwear (sole)/ice interaction (tribophysics)

4) Human gait biomechanics

5) Human factors (intrinsic factors)

6) Environmental factors(extrinsic factors: cold climate, lighting, etc.)

Figure 1. Systems model of slip and fall accidents on icy surfaces.

576 C. Gao and J. Abeysekera

did not use snow chains because of discomfort on cleared ground and unwillingnessto spend time putting the chains on and off.

3.1. Soling materialsBruce et al. (1986) measured slip resistance of shoes, crampons and chains on icysurfaces (dry ice,7 98C). It was found that all friction values of footwear on ice werelow by comparison with other substrates and were generally lower than thatobtained on an oily steel plate. The highest friction values were generated by thesoftest materials, 0.19 by microcellular polyurethane (PU). The other values indescending order were 0.17 by soft rubber, 0.14 by nitrile rubber, 0.09 by leathersole, 0.08 by old PVC respectively. The COF for crampon was 0.24 – 0.31, whereasthe COF for chain was only 0.11. Therefore, the best traction was provided bycrampons, consisting of steel studs in a rubber strip. Chains attached to the shoesgave very poor traction on smooth ice. The best shoe soling on ice (7 98C) wasdouble density soft PU among those tested. The soft rubber snow boot and a rigidfoam slipper were also slip resistant on icy conditions, but the latter is very flimsy.An old pair of hard PVC slippers was extremely dangerous for use on ice. Bruce et al.(1986) suggest that as a first step, double density microcellular PU soling be used as astandard to reduce injuries on icy surfaces.

Cumulative research over a period of 15 years consistently revealed that a PUsoling is the most slip resistant soling for use on oily and wet floors (Manning et al.1985, Manning and Jones 1994, Gao et al. 2003). Gronqvist (1995b) also revealedthat used PU heels and soles gave a considerably higher coefficient of kinetic frictionon contaminated floors than used heels and soles made of compact nitrile rubber(NR) and of compact styrene rubber (SR) on contaminated floors. However,Gronqvist and Hirvonen (1995) recommended soft heel and sole materials ofthermoplastic rubber (rather than PU) for winter footwear for use on dry ice(7 108C), believing that PU was not safe enough on wet ice (08C). Abeysekera andGao (2001) also demonstrated that footwear with a PU outsole is slippery on wet ice,and even more slippery than footwear with nitrile and natural rubber outsoles. Thestudy of soling materials did not show higher COF of PU on hard ice (Gao et al.2004). A recent study on footwear sole (nitrile rubber, double density polyurethane,thermoplastic polyurethane, styrene rubber, rubber and glass fibre mixture, creperubber and microcellular polyurethane) on the coefficient of friction on melting andhard ice showed that on melting ice (08C), all types of footwear soles had lowestCOF. Hard ice is more slip resistant than lubricated steel plate. The curling footwearwith crepe rubber soling performed best, even better than PU, in terms of slipresistance on hard ice at7 108C. Therefore, crepe rubber is recommended for use onhard ice (Gao et al. 2003).

There seems to be no universal soling material which is slip resistant on differenticy surfaces, e.g., dry and wet ice. From the viewpoint of practice, these researchresults mean that it is challenging to design proper anti-slip footwear for use on alltypes of icy and snowy surfaces in winter, because it is unlikely that one solingmaterial/tread design can accommodate various climatic conditions and subsequentchangeable icy and snowy surfaces. Gronqvist and Hirvonen (1995) pointed out thatnew soling materials and tread combinations must be developed, e.g., very hardmaterial with sharp cleats (scratch formation) in combination with a softer basematerial, and the use of footwear with studded heels and soles during high-risk dayswhen the ambient temperature in close to 08C. However, such soles are not suitable

577Slip and fall accidents

for use on indoor floors. A proper anti-slip material arrangement, fastened on theout sole might be one solution (Noguchi and Saito 1996b). Other slip resistantmaterials and integration designs must be explored to accommodate outdoor andindoor surfaces.

3.2. Sole tread (geometry) patternBesides sole materials, sole (geometry) pattern also affects slip resistance. Differentsurfaces may require different tread designs in order to achieve higher friction.However, there is a paucity of studies on the effect of tread patterns on slipperinessdue to the difficulty of quantifying tread patterns.Gronqvist et al. (1993) compared objective and subjective assessments for

slipperiness of the shoe/floor (glycerine lubricated) interface, the results showedthat a sole with tread removed was more slippery than sole with tread. Leclercq et al.(1994) argued that it was necessary to maximize the contact area of the sole,especially by avoiding the use of ‘micro-tread’ (small bumps) on its surface and byreducing the overall curvature of the sole. Tisserand (1985) showed that soles withasperities had a negative influence on the value of COF because of the small contactarea. However, Leclercq et al. (1994) argued, at the same time, that the ‘tread’ andthe ‘material’ did not act independently on the coefficient of dynamic friction. Theridges of these treads should be as sharp as possible in order to wipe away any liquidpollutants.Strandberg (1985) and Tisserand (1985) stated that it was necessary to provide a

means for the rapid evacuation of any pollutant through channels in the tread. Thepollutant must never be captured between the surface and the shoe, for example by aborder that encloses the channels or by treads in the form of small suction cups. Thisis a rule well known to the manufacturers of automobile tyres.Lloyd and Stevenson (1989) and Stevenson (1997) showed a rounded (bevelled)

heel edge gave a higher slip resistance than a sharp (squared) one, because of theincreased area of contact immediately after heel strike. However, Strandberg (1985)strongly recommended the use of a small effective contact area when walking on aviscous fluid surface at the heel strike moment in order to generate a high pressure.The substantial influence from a smaller area pattern was demonstrated by themultiple increase in COF. Nevertheless, Gronqvist (1995b) revealed that the greaterthe contact angle, i.e., the tangent between the shoe and the contaminated floor, thehigher the risk of slipping seemed to be, and an increase in the pressure did not leadto an increase in the coefficient of friction on smooth lubricated surfaces. On thecontrary, high normal pressures seemed to increase the risk of slipping. On roughfloors, however, the lubricant breakdown is more likely to occur, enabling moreeffective draping, thus, increasing particularly the hysteretic friction. Obviously,results on contact area and pressure are contradictory.Gronqvist (1995b) showed that the sole slabs (PUr-slabs) with a rectangular tread

pattern and with asperities had a slower increase of friction after the first contactwith the steel surface than the sole slabs (NRw-slabs) with a waveform or triangulartread pattern and without asperities. The maximum kinetic friction for the NRw-slabwas reached within 0.1 s after the first contact. However, the maximum kineticfriction for the PUr-slab was not reached until 0.5 s after the first contact, whichnormally is too long a time to prevent a slip from developing into a fall.As well as the heel sole tread pattern, the toe sole plays another role in toe off

(rearward) slip, which is currently a neglected area of research.

578 C. Gao and J. Abeysekera

All the above discussion on tread and COF is based upon floors/lubricated floors.Little literature has been found regarding tread design for use on icy and snowysurfaces. Although the same principles may apply to icy and snowy surfaces, theremay be different requirements.

Gronqvist and Hirvonen (1995) discussed the role the cleat design for use on dryand wet ice as shown below:

(1) Flat cleats with an apparent contact area as large as possible gave the highestfriction readings on dry ice (7 108C),

(2) Sharp cleats combined with very hard heel material gave the highest frictionreadings on wet ice (08C) due to scratch formation.

Jung (1992) observed that there was no correlation between the contact surface areavalues and the slip resistance. Abeysekera and Gao (2001) also showed that therewas no significant correlation between subjective slipperiness ratings, COF andcontact areas.

However, almost all the previous research and measurements did notseparate the anti-slip effects of sole material, hardness, roughness, and thetread design completely. In fact, all the work done before on the frictionaleffect combined factors and whether one or two of them (material, hardness,roughness, or tread design) contributed to it were not known. Furtherresearch is needed.

3.3. Sole hardnessAs discussed by Leclercq et al. (1994), hardness is one of the more widely usedmechanical properties. The measurements on lubricant/floor conditions show that areduction in the hardness of the sole corresponds to an increase in the COF, but theincrease appears to be much too small to be useful in terms of safety. In frozenenvironment, the effect seems to be the same.

Bruce et al. (1986) showed that there was a negative correlation between solehardness and COF on ice (dry ice, 7 98C, r=7 0.876). The poor friction propertieson ice of hard materials have been reported (Kellett 1970 and Perkins 1976, cited byBruce et al. 1986).

Gronqvist and Hirvonen (1995) found a significantly negative correlation betweenkinetic COF and hardness of the heel material and the hardness of the entire solingon dry ice (7 108C). This is consistent with the result by Bruce et al. (1986).However, there was no clear correlation while on wet ice (08C). It will be necessary tofurther examine hardness and friction at a range of temperatures in order to makeclear if hardness of footwear recommended for icy and snowy surfaces is suitable forall temperatures likely to be encountered by potential users. Ahagon et al. (1988)argue that although the friction resistance of rubber on ice is primarily determinedby the viscosity and the thickness of the lubricating fluid layer, further improvementof the friction could be obtained by making rubber more resilient. However, whetherthe expected increase of friction is safe enough at melting point needs furtherresearch. But very hard heels (Shore A4 85) combined with sharp, e.g., conic cleatstended to give the highest friction readings. Such a combination of hardness andcleat-shape was able to scratch the surface of the wet and soft ice, the temperature ofwhich was 08C. On the contrary, the hardest soling materials were the most slipperyones on the harder ice at – 108C. Simply put, in order to increase the slip resistance of

579Slip and fall accidents

footwear on icy surfaces, soft solings are needed for use on hard ice (dry ice),whereas hard solings are best on soft ice (wet ice).It is interesting to notice that a study done by Manning et al. (1985) is based on the

speculation that some animal species may have developed slip resistant feet. Thepolar bear is thought to be adapted to the slippery environment. The footpads werefound to have a rough papillary surface overlying a soft dermis containing a densenetwork of collagen and elastic fibres. These findings support a hypothesis that shoesoling for use on an icy substrate should be soft with a hardness value in the regionof 24 on the Shore A scale (0 – 100). The surface should be covered in conicalprojections having a mean diameter of 1 mm. Further work on the feet of animalspecies could lead to a better understanding of slip resistance and reduce injuries tohumans and livestock. However, the crucial difference between an animal footpadand a human footwear sole is that the former can be self-regenerated, whereas thelatter easily wears out, which has been demonstrated as an important effect on slipresistance (Leclercq et al. 1994, Gronqvist 1995, Hirvonen and Gronqvist 1998, Gaoet al. 2003). Another important difference is that the polar bear is a quadruped,whereas humans are bipedal. Therefore, the gaits cannot be compared. As shown inthe systems model of slip and fall accidents on icy surfaces (figure 1) and thediscussion in section six of this paper, gait biomechanics is one of the importantelements contributing to slips and falls.

3.4. Wear and tearThe slip resistance of a shoe is not a stable property. It varies during the course of itsuse. Leclercq (1999) argues that the slip resistance of footwear is a transient as well asa versatile quality. The slip resistance may decrease or increase depending on thesoling materials and the amount of wear (Manning et al. 1985, Leclercq et al. 1994,Gronqvist and Hirvonen 1995, Hirvonen and Gronqvist 1998, Gao et al. 2003).Manning et al. (1985) tested boots worn on oily surfaces for a period of more than

1 year. The results showed that, in general, the friction fluctuated over the course ofwearing. For PU material, improvement in friction was recorded after a short periodof wear. Manning suggests that to deliberately scuff new shoes on a very roughsurface will increase the friction based upon the fact that scarification by metal swarfon the floor increases friction readings for both NR and PU soles.Leclercq et al. (1994) argued that the friction coefficient of a shoe increases as soon

as it was worn, and the relative increase of the friction coefficient of expandedpolyurethane soles was greater than that noted for soles made of compactedelastomers. Leclercq (1999) claimed that slightly worn PU was generally preferred,when new, this material did not provide a very high level of slip resistance.Gronqvist (1995b) recommended footwear must be discarded before the tread

pattern was worn out. According to the assessment of new and used (from 2 – 3months to over 8 months) footwear with three types of materials (compact nitrilerubber-NR, compact styrene rubber-SR, and microcellular polyurethane-PU), theshoe heels made of PU and SR were less slip resistant when new than used, especiallyfor PU soles. This result is consistent with those obtained by Manning et al. (1985).The slightly used heels (2 – 3 months usage) had the best slip resistance properties,probably due to the optimum combination of sufficient tread pattern profile depthand sufficient surface roughness. However, there was a gradual decrease in the slipresistance of shoes after 4 months of use. However, the new and used sole heels madeof NR seemed to remain the same level of coefficient of kinetic friction on

580 C. Gao and J. Abeysekera

contaminated floors. Gronqvist (1995b) argued that one reason could be themicroporosity of the material, which was revealed after only some wear.Microporosity reduces the hydrodynamic pressure at asperity peaks, thus enablingthe elastic pressure of the heel material to exceed this and break the fluid film toincrease the slip resistance. The relatively soft heel and sole of PU material may alsohave better deformation and damping properties than harder materials (SR, NR). Ahigher damping effect may particularly increase the coefficient of friction on therough plastic floor, whereas the microporosity effect probably dominates on thesmooth steel floor.

Hirvonen and Gronqvist (1998) also observed that PU soling materials increasethe friction by a mean of 32% after being abraded by testing them on a roughconcrete surface and steel/sand combination. The friction decreased by a mean of42% with styrene rubber, and both decrement and increment were observed withnitrile rubber soling after abrasion.

However, there is little research reported on the effect of sole wear and tear on icyand snowy surfaces on the frictional properties. The latest research showed thatartificially abrasive wear of nine pairs of soles improves slip resistance on hard ice(7 108C), but not on melting ice (Gao et al. 2003).

3.5. Asperity (roughness)Roughness and wear and tear are interrelated. PU is the most slip resistant soling onoily and wet floors because it becomes very rough when the surface skin is abradedand this roughness appears to be an important contributor to the slip resistance(Manning and Jones 1994). Hirvonen and Gronqvist (1998) also argued that theimproved friction performance of PU soling was most likely caused by changes in theroughness of the soling, as well as by increased porosity of the cellular structurewhich was broken through the action of wear.

Jung (1992) found an important positive correlation between the size-specificroughness characteristic Rz (average peak-to-valley height) and the acceptanceangle a (in ramp test on oily steel floors). However, Tisserand (1985) showed thatsoles with asperities had a negative influence on the value of COF, which alwaysincreases when the asperities are removed (smooth sole) by modifying the moulds(not by abrasion of the soles). Although Manning et al. (1991) showed that themean of all COF results on the seven different surfaces including dry and wet icewas significantly related to (positive correlation) roughness (Rtm, i.e. Rz) of solingmaterials of thirteen pairs of discarded working footwear, the walking traction testmethod (Rig 3) did not specifically reveal the effect of footwear roughness on COFon dry and wet ice. A recent study by Manning and Jones (2001) has demonstratedthat there is no effect of five footwear solings with graded roughness (4.4 – 19.12Rtm) on COF on ice (7 38C, on an ice skating rink) spread with water using thewalking traction method (Rig 3), although the effect was demonstrated on wet andoily floor surfaces.

The roles of both footwear and floor surface roughness in the measurement ofslipperiness, as well as various techniques, parameters and instruments for roughnessmeasurements are reviewed by Chang et al. (2001c). It is concluded that surfaceroughness on shoe and floor surface affect slipperiness significantly. The slipresistance measurement should be based on an understanding of the nature offrictional and wear phenomena and surface analysis techniques for characterizingboth the surfaces of the shoe and floor and their interactions. It is argued that surface

581Slip and fall accidents

roughness measurement may provide an objective alternative to overcoming thelimitations of friction measurements.

3.6. Other properties of footwearAbeysekera and Khan (1998) argued that to prevent falling after a slip, the balancingproperties of the shoe should be improved. They argued if the centre of gravity(COG) of a shoe was situated in-line with or very close to the centre of gravity of thewearer’s foot, it might improve the wearer’s balancing ability. In order to test thisassumption, Abeysekera and Gao (2001) carried out a pilot study of footwear COGon the slipperiness on ice. But no correlation between COG and slipperiness wasobserved. However, further research is needed.A questionnaire survey by Bergqvist and Abeysekera (1994) revealed that when

designing safety shoes for use in cold climates, the following ergonomics aspectsshould be considered: fit, thermal comfort, protection from work hazard, low weightand anti-slip. Gao and Abeysekera (2002) argue that, in addition to thermalinsulation, slip resistance must be another functionality requirement of footwear foruse on icy and snowy surfaces. The slip resistance, thermal insulation, andwearability must be harmoniously integrated in winter footwear design. Otherproperties should not compromise slip resistance property.Additional footwear attachments (anti-slip devices) are other means to prevent

slipping on ice and snow, especially for the elderly (Gard and Lundborg 2000). Butthey may be heavy, bulky, and inconvenient, resulting in poor usability.Halogenation of the sole can either increase or decrease slip resistance, but it

seems not to have a significant influence (Leclercq et al. 1994). No correlationbetween slip resistance and the flexibility of outsole could be found (Jung 1992).

4. Icy and snowy surfaces

As indicated previously, there have been a number of methods, equipment, andstandards (including ISO, CEN, ASTM, BSI) concerning slip resistance on lubricant(water, oil) contaminated floors (Manning et al. 1985, Jung and Schenk 1990, Jung1992, Leamon 1992b, Jung and Fischer 1993, Gronqvist 1995b). However, there islittle research, and no standards available governing the measurement of slipperinessbetween icy and snowy surfaces and footwear, even though icy surfaces (near meltingpoint) are more slippery than lubricated floors and the population associated withslips on ice and snow is much more than that on contaminated floors. Therefore,there are desperate needs to explore the slipperiness of footwear on icy and snowysurfaces.Traditionally, there have been measures used to prevent slip and fall and traffic

accidents on frozen roads such as de-icing chemicals (salting, anti-freeze mixtures)and road heating (Kobayashi et al. 1997). Asphalt mixtures containing reclaimedtyre rubber particles (AMRP) has been practiced in Japan (Taniguchi et al. 1997).Spreading boiler slag, coke cinders, sand and crushed stone (gravel), etc. on ice orsnow is also commonly seen. But there is little research on appropriate anti-slipfootwear and its interaction with the icy and snowy surfaces.De Koning et al. (1992) showed that the minimum value of coefficient of friction

was found at an ice surface temperature of between 7 68C and 7 98C (mean7 7.48C) during skating. The corresponding mean values of the coefficients offriction for the straight and curve skating were 0.0046 and 0.0059 respectively. Theice surface temperature when at7 7.48C seems the best for skating, but the worst for

582 C. Gao and J. Abeysekera

walking as far as the slip and fall accidents are concerned. However, precautionsshould be taken while adopting these results measured at an interface between steelblade and ice.

Ice is not always slippery (Petrenko 1994), the ice friction coefficient canassume both very small value (m5 0.01), at high temperatures (7 18C) and highvelocities (3 ms7 1), or very large value (m=0.6) at low temperatures (7 408C)and low velocities (0.01 cms7 1). Roberts and Richardson (1981) measured ahemisphere of rubber friction on polished ice over the temperature range from7 32 to 7 18C, the results showed that the friction coefficient on cold ice washigh, however, there was a sharp fall at temperatures above – 108C. The fall wasdramatic, 20 times or more, at 7 18C it was only about 0.02. The latest studycompared the friction on hard ice, melting ice and lubricated steel plate. Thefriction is highest on hard ice (7 108C), followed by lubricated steel plate.Melting ice is most slippery in terms of friction (Gao et al. 2003). In general, theproperties of ice, e.g., temperature, structure and hardness, as well as thethickness of the water layer, seem to determine the friction during a slip to agreater extent. The loss of adhesion and friction on ice near its melting point isdetermined more by the properties of the ice than by the properties of the rubber(Gnorich and Grosch 1975, Roberts 1981).

The American Society for Testing and Materials (ASTM 2001) recommended thatthe static COF value for a slip resistant polish coated surface should be 0.5. TheCEN/TC 161/WG3 (2003) removed the specifications of the COF requirement.According to the slipperiness classification by Gronqvist et al. (1989), 0.2 was theminimum requirement for slip resistance on contaminated floors. At present, there isno standard for COF on icy and snowy surfaces.

In Sweden, house-owners are commonly responsible for anti-skid management onsidewalks (pavements). In public areas, the local authority is responsible. A lot morepedestrians choose to stay inside during slippery weather (Gard and Lundborg1994). Local bodies spend large amounts of money on anti-slip materials and spreadthe materials during winter. But the anti-slip effect of different materials has still notbeen adequately researched.

Spreading sand (180 gm7 2) and gravel (F4 – 8, 150 gm7 2) could greatly increasethe friction on wet ice. Sand was tested as the best anti-slip material spread on icyand snowy surfaces (Abeysekera and Khan 1998, Abeysekera and Gao 2001).Kobayashi et al. (1996) also showed that static COF on ice sprinkled with sandincreased more than 0.5, which was quite effective in increasing walking safety forpedestrians. Spreading sand on level and inclined frozen roads in winter has becomea routine practice in Lulea, Sweden and Hokkaido, Japan respectively, and in othernorthern parts of the world.

5. Footwear (sole)/ice interaction—mechanism of friction

Slip occurs at the interface, i.e., footwear (sole)/underfoot surface. The tribologicalphenomena that occur at the interface are very complex (Leclercq 1999). It isgenerally accepted that friction is caused by adhesion, plastic and elastic deformationof surfaces (hysteresis). If a low viscosity fluid as lubricant is present, the tworubbing surfaces are more or less separated. This separation leads to a largereduction in friction.

On the contrary to prevent slips and falls, in skating and skiing, many suggestionsare made for achieving lower frictions with a thin film of liquid water between the icy

583Slip and fall accidents

and snowy surfaces and the skate or ski. It is, therefore, expected that if waterformation is reduced, the friction can be increased.Theories about the presence of water between the surfaces are focused on the

‘liquid-like’ properties of the ice surface, the formation of water by pressure meltingand by melting due to frictional heating and surface melting (De Koning et al. 1992,Wettlaufer and Dash 2000). De Koning et al. (1992) argued that it was impossible tosay which mechanism caused the low friction on ice. The latest review on frictionmechanisms of the footwear-ice interface showed that the properties of the interfacelayer in ice and snow friction are still poorly understood (Chang et al. 2001b).Fall accident analysis by Strandberg (1985) indicated that most slipping accidents

occurred on surfaces covered with snow, ice, grease or liquid, and argued that slip-resistance of lubricated conditions underfoot was dependent on at least threedifferent processes.

(1) The squeeze-film process, when the normal force between shoe and flooringdisplaces a lubricating fluid,

(2) The development of a so-called hyperesis component of friction force whensufficient fluid has been displaced to allow draping of the sliding shoeelastomer on the floor asperities,

(3) The development of a so-called adhesion component of friction force due tomolecular bonding between those parts of the shoe and flooring surfaceswhich are in true contact, i.e., where the interfacial fluid has been completelyremoved.

Gronqvist (1999) elaborated a friction model for slipping, which took into accountof the drainage capacity of the shoe floor contact surface (squeeze film processes), thedraping of the shoe bottom above the asperities of the floor surface (deformationand damping), and finally the true molecular contact (adhesion and wear) betweenthe interacting surfaces (traction).On warm ice (near the melting point), the friction seems to be governed

particularly by ice flow and melting (Roberts and Richardson 1981). Ionic impuritiesin ice lower its melting point, thereby forming liquid brine at the surface, which hasan essential lowering effect on adhesional friction. The thickness of the water layer isinversely proportional to friction (Chang et al. 2001b). Warm ice is also sensitive tohigh pressure effects, like at heel strike in normal gait. At high pressure points ice willeither flow to relieve the pressure or melt. Both phenomena tend to lower thecoefficient of friction when the temperature of ice is warmer than – 108C (Gronqvist1999). On cold ice, at very cold temperatures, the ice surface is dry as no liquid film isformed. Under these conditions, ice behaves like any ordinary dry surface, whereadhesional friction plays a major role. Other mechanisms on ice include asperity,ploughing effect, capillary suction, electrical forces, pressure melting, imbalance ofsurface energy, etc. (Makkonen 1994, Chang et al. 2001b).

5.1. Mechanical COF measurementThere are a number of methods and related mechanical equipment for measuring thecoefficient of friction on floors or contaminated floors. Strandberg (1985) reportedthat about 70 types of slip resistance testers could be found in literature, and theydiffered substantially in their design. Lin et al. (1995) also argued that at least 70different meters had already been cited in the literature. However, no method or

584 C. Gao and J. Abeysekera

apparatus has achieved universal acceptance for slip resistance measurements (Lin etal. 1995, English 1996). Most of the testing devices are not mechanically capable ofsimulating the human gait parameters deemed crucial during actual slippage (Lin etal. 1995). Emphasis has been placed on the use of one reading from one instrumentto predict the suitability of a floor for all situations. According to some researchers,additional efforts are still required in pursuing more realistic COF measurementsthat represent the actual kinematics during critical gait phases (Gronqvist et al.1993).

Two main problems of measuring COF are the test validity and reliability.According to Jung and Fischer (1993), the inter-laboratory results do not confirmthe validity of the ISO test method. According to Leclercq et al. (1995) and Leclercq(1999), biomechanical studies on slipping that were performed in the laboratoryrevealed a high level of intra- and inter-subject variability in the data recorded at theshoe/surface interface. This variability constitutes one of the biggest obstacles toprogress in this area. Since the methods in use are not identical, the results areoccasionally contradictory from one laboratory to the next. Despite a large numberof comparative studies and standardization work, there is still no consensus on aparticular measurement technique at the current time.

More recently, Chang et al. (2001a) reviewed 23 friction (static, dynamic andtransitional) measurement devices including field and laboratory based methods andconcluded that the devices appear to be generally valid and reliable. However, thevalidity of the devices still could be improved to reflect actual human slippingconditions observed in biomechanical studies.

5.2. Other biomechanical and subjective slipperiness measurements5.2.1. Ramp test: A ramp test can be used (Wilson and Perkins 1985, Jung 1992, deLange and Gronqvist 1997, Redfern et al. 1998) measuring the angle as the indicatorof the slipperiness. De Lange and Gronqvist (1997) conducted comparativemeasurements between mechanical and biomechanical slip tests, the latter consistedof three subjective evaluation methods, i.e., the inclined plane (ramp) and twohorizontal slipping experiments.

5.2.2. Paired comparison (Thurstone binary choice method): Tisserand (1985)utilized subjective evaluation experiment (paired comparison), this allowed subjectssimultaneously to wear a different shoe on each foot, and it was believed to be easilyable to compare the two types of shoe. A classification could be obtained by takingseveral different types of shoe and comparing each type with all the others, two at atime. The results were correlated with dynamic COF, but not related to static COF.

Myung et al. (1993) also used a subjective method assessing floor slipperiness. Thedifferent feature compared to other subjective assessments was that they used aforced choice method, i.e., each subject walked down the floor with left foot on onesurface and the right foot on another surface of a different floor material. After thesubjects had completed each walk, they indicated which of the two surfaces (left orright) felt more slippery. However, the gait and sensation differences between leftand right foot may be a factor influencing subjective feelings unless the experimentaldesign is completely randomized.

5.2.3. Traction test: Manning et al. (1990, 1991) and Manning and Jones (1994,2001) used a walking traction test rig (fixed and mobile) to measure COF, with

585Slip and fall accidents

subject walking backwards on the heels and forwards on a slippery surface, pullingagainst a set of springs anchored to a wall, until the feet slipped. The subject wassupported by a fall-arrest harness, and the load cell was positioned between the beltand the springs. The load cell measures the maximum force on the spring prior to theslip. The COF is calculated by dividing the force on the springs by the standardizedbody weights. Kobayashi et al. (1996) used similar method to measure footwearCOF.

5.2.4. Rating scales (perception of slipperiness): Swensen et al. (1992) used graphicrating scale for measuring slipperiness of climbing and coating surfaces. Gronqvist etal. (1993) and Gao and Abeysekera (2002) used a 5-point rating scale to assess theshoe/floor and shoe/ice interface slipperiness respectively. Cohen and Cohen (1994a,b) utilized a 13-point rating scale to measure walking surface slipperiness. All resultsshowed there was significant correlation between subjective ratings and COFmeasurements. Chiou et al. (2000) applied a perceived sense of slip (PSOS) scale tomeasure static slipperiness while doing different standing tasks and reportedsignificant associations between PSOS and the objective postural instability variables.Gard and Lundborg (2000) used rating scales for perceived walking safety andbalance when wearing anti-skid devices attached to shoes for use on icy surface.

5.2.5. Biomechanical method: Biomechanically-oriented approaches were sum-marised by Redfern et al. (2001). Previous research has focused on kinematics,kinetics (normal and shear forces) and required friction. Hanson et al. (1999) usedthe relationship between available friction and required friction to predict slips andfalls (logistic model). Recent study suggested that friction demand as measured byrequired coefficient of friction (RCOF) may not be a totally deterministic predictorof actual fall events. Heel contact velocity (nhc) and the transition of the whole bodycentre of mass (COM) can be analysed to evaluate slip initiation. They may not betotally deterministic predictors of actual fall events either (Lockhart et al. 2003).Little research has been done on the role of lower extremity muscles in maintaininggait balance on slippery surface and in recovering balance after slipping.In addition, Noguchi (2000) uses vector locus and a danger coefficient of slip

(DCS) to express and diagnose slipperiness in human gait. Hirvonen et al. (1994) andNoguchi (1996a) identify slips by measuring body acceleration while walking.The Liberty Mutual Research Centre for Safety and Health hosted an

international symposium on ‘The Measurement of Slipperiness’ in the year 2000with nearly 30 participants from nine countries. Out of 14 participants responded tothe survey, 12 (86%) did not feel that slipperiness measurement was adequate.Experts could not reach a clear consensus on an ideal measurement approach toassess slipperiness (Chang et al. 2003).

5.2. COF measurement method and equipment on iceAlthough there are a number of methods and equipment for measuring COF on dryfloors and contaminated/lubricated floors (Strandberg 1985, Lin et al. 1995, Leclercq1999, Chang et al. 2001a), there are very few methods available for measuring COFon icy surfaces. Tisserand (1985) pointed out that for testing purpose, ice was adifficult and unpredictable surface. Manning et al. (1991) observed the difficulty ofmeasuring friction indices on ice because the friction was very low and the surfacewas highly irregular. Gronqvist and Hirvonen (1995) developed a friction

586 C. Gao and J. Abeysekera

measurement apparatus equipment with a temperature controlled cooling system forice formation on force plate which could be used for measuring COF of footwearsole/ice interaction in laboratory. However, the measurement set-up is placed in theroom temperature environment where the icy surface is formed. Footwear is notconditioned in cold environment before measurement. It would be more realistic ifthe equipment and footwear were placed in a cold chamber. Further studies areneeded in this regard.

In order to measure the skate-to-ice friction, De Koning et al. (1992) used ameasuring system, which could simultaneously measure friction while skating. Thesystem measured the real time friction with considering not only the mechanicalfactor, but also biomechanical and psychophysiological factors incorporated in realskating. The measuring system consisted of a pair of instrumented skates and aportable data acquisition microcomputer, in which, transducers for both normalforce (vertical) and frictional force (horizontal) were built between the shoe and theblade of the skate. The principle may be applied to measuring friction betweencommon footwear and icy surfaces in real human walking.

Manning et al. (1991) applied a walking traction test to various slippery surfacesincluding oil-lubricated, water-lubricated, dry and wet ice. Gard and Lundborg(1994) conducted outdoor trials with elderly subjects walking on icy surfaces inwinter to test the effect of anti-skid devices attached to footwear, and walkingbalance based on the subjective perceptions. Gao and Abeysekera (2002) conductedoutdoor walking trials on icy surfaces by using 5-point rating scales, directobservation and videotaping to measure the slipperiness of the footwear and the icysurfaces, and slipping and falling events. Further research is needed to develop morereliable and valid method and equipment to measure the COF on icy surfaces.

6. Human gait biomechanics

During human gait cycle, there are swing and stance phases for each foot. The swingphase can be divided into a period of acceleration, mid swing and a period ofdeceleration. Whilst, the stance phase is divided into heel strike, mid stance and toeoff. There are two basic foot movements at the ankle joints during these two phases,i.e., dorsiflexion and plantar flexion. Some researchers categorize the gait into fourdistinct phases: (1) heel strike (touch down, landing); (2) stance phase (stationary);(3) toe off (push off, take off); (4) swing phase (Lin et al. 1995).

A recent literature review by Redfern et al. (2001) showed that the forceinteraction between the shoe and floor are probably the most critical biomechnicalparameters in slips and falls. If the shear forces generated during a particular stepexceed the frictional capability of the shoe/floor interface, then a slip is inevitable.Since the shear forces are highest near the heel contact (first peak in the forwarddirection) and push off (second peak in the rearward direction) phases, these are thepoints where slips most often occur (Redfern and Dipasquale 1997).

According to Kirtley (1999) concerning centre of mass (COM) and base of support(BOS) for stability, the COM must remain within the BOS. However, in gait it passesoutside the BOS twice every gait cycle—it passes in front of the trailing toe in earlyswing, and is behind the leading heel in late swing. These two periods in each gaitcycle during which there is no support, i.e., the COM is outside the BOS and is infree fall, they, each lasts for at least 100 ms, are particularly dangerous times, whenfalls might occur, so they must be of significance to slip and fall accidents on icy andsnowy surfaces. These two periods are near the toe off and heel strike periods. It

587Slip and fall accidents

coincides with the widely accepted findings, i.e., the slip occurs most frequentlyduring heel strike and toe off periods during human gait cycle (Strandberg andLanshammar 1981, Davis 1983, Leamon and Son 1989, Manning et al. 1991,Gronqvist 1999). It implies that precautions and anti-slip measures such as enforcingCOF in heel sole and fore sole have to be taken to prevent slips during heel strike andtoe off. So far, almost all the mechanical friction test equipment and method cannotsimulate the COM and BOS reciprocal changes during gait cycle, the applicability ofthe measurement, when applied to the prevention of human slips and falls can,therefore, be questioned.Based on ground reaction forces and pressure distribution at out sole during gait,

Noguchi and Saito (1996b) suggested a proper arrangement for fastened anti-slipglass fibre compound on rubber sole at heel and fore sole parts for use on icy andsnowy surfaces by the disabled. The effect of the reinforcement on frictional propertyon icy and snowy surfaces by normal pedestrians needs further exploring.Leamon (1992b) pointed out that there were a number of characteristics of human

gait, which might be very significant for slipping and falling. Load carrying may alsointerfere with the normal technique of balancing, which involves rotation around thehips and arm swings which are used to maintain the position of the centre of thegravity over the appropriate foot. Davis (1983) observed that the relative posturestability is reduced by load holding roughly linearly with the magnitude and heightof the weight by using the anterior-posterior and lateral sway angles. Such factorsneed to be further systematically investigated. Myung and Smith (1997) have carriedout the research about the effect of load carrying and floor contaminants on slip andfall parameters. The results showed that an abnormal gait pattern, short stridelength, was seen on oily floors or with heavy load carriage because subjects adjustedtheir stride length for a better stance.Experiments have indicated that people manipulate gait when walking on slippery

surfaces. It is the common experience that people can, and do, walk on ice and otherslippery surfaces. When a slippery condition (e.g., icy walkway) is perceived, eitherby visual and/or tactile cues, a person’s walking gait is adjusted accordingly. Thelength of the stride is shortened, which consequently produces low foot velocities andsmaller foot shear forces, as the body’s centre of gravity is better maintained over thebalance zone (Swensen et al. 1992). The heel strike and toe off phases are alsosignificantly reduced to diminish the likelihood of slipping. The risk of slipping andfalling seems to depend to a great extent on a person’s perception of the potentialslipperiness of the actual conditions (Gronqvist 1999, Gao and Abeysekera 2002).James (1983) also illustrated that it was necessary to shorten the stride on a slipperyfloor where the coefficient of friction is very low. Leamon (1992b) claimed thatproprioceptive recognition, visual anticipatory evaluation of the slipperiness of thesurface were the prerequisites of postural adjustment.But there has been little biomechanical research reported on human gait change on

icy surfaces.Strandberg and Lanshammar (1981) argued that a valid test of COF should

consider the following variables from crucial gait phases: (1) the angle of the foot tothe floor, (2) the point of application of the contact force between the shoe and thefloor, (3) the vertical force, (4) the sliding velocity, (5) the time scale.For measuring slip resistance, Gronqvist (1999) put forward a biomechanical

parameter model, arguing that any friction test instrument should meet thebiomechanical criteria. The measurement parameters and their ranges should reflect

588 C. Gao and J. Abeysekera

gait biomechanics and tribophysics of actual slipping incidents. Recent review onfriction measurement device parameters indicated (Chang et al. 2001a):

(1) normal force build-up rate should be at least 10 kNs7 1 for whole-shoedevice,

(2) normal pressure should be between 200 and 1,000 kPa,(3) sliding velocity should be between zero and 1.0 ms7 1,(4) maximum time of contact prior to and during the coefficient of friction

computation should be 600 ms.

In a real human gait cycle, stride length, step length, and acceleration may also playroles in determining the slip resistance, which are not considered in current COFmeasurement methods. Myung and Smith (1997) argued in measuring slipperinesswith a slip tester, that the set-up value for heel velocity has to be classified differentlyfor dry and oily floors. On icy surfaces, it may have different human gait parameters.

Noguchi and Saito (1996a) used accelerations of foot and body centre movementsas parameters to evaluate winter shoe anti-slip device worn on indoor floors andoutdoor frozen roads. The results of acceleration recordings showed that there werenegative accelerations indicating rearward slip at toe off phase on frozen roads.

In addition, walking speed, e.g., in hurry, turning, and straight walking, willinfluence slips and falls on icy and snowy surfaces. It is common that slips and fallsoccur more frequently while walking in haste, in turning and on up and down slopes.More research is necessary on the effects of these factors on human gait on ice andsnow.

7. Human factors (intrinsic factors)

The role of intrinsic factors in fall mechanisms was discussed by Tideiksaar (1990)and Gauchard et al. (2001), which include physiological and pathological aging,experience, attention, fatigue, etc. Davis (1983) and Tisserand (1985) use a humanmental model for slipperiness to explain the psychophysiological factors involved inwalking. Walkers have a constant image in a part of their memory, a model of thefriction limits of the surface on which they are walking. Tisserand (1985) agues it isimportant to establish that any fall caused by slipping and any non-falling skid(except for microslip) is in fact the result of a discrepancy between the model andreality, i.e., a failure of the evaluation system. Tisserand points out that the risk ofslipping lies more in the gradient of the friction coefficient of the surface than in itsabsolute constant value, and suggests further research on the mental model, whichmonitors slipperiness.

English (1994) observed, from video recordings of walking trials, how the testsubjects were able to apply muscular force to counter incipient slips so as to arrestthem. There were many micro slips observed, but no actual falls occurred due topostural reflexes and muscular force applied. English made an important point thatpossibly the major determinant of slips and falls is a person’s knowledge of thesurface. People are perfectly capable of walking on ice if they know about it and howto carry out complex movements on the deliberately slippery surfaces found inbowling centres and on dance floors. As discussed previously in this review, the COFof skating is about 0.0046, in such a very slippery situation, skaters can still manageto balance without falls. The difference compared to pedestrians is that they are welltrained and experienced athletes and dancers, and they perform in a relatively short

589Slip and fall accidents

period of time on the highly slippery surfaces. This strongly suggests thatpsychophysiological factors, balance capacity and gait biomechanics play veryimportant roles in preventing slips and falls on icy surfaces.

7.1. AgeingHuman physiological functions change with age. English (1994) notes that theslowing of reflexes and increased skeletal fragility which occurs with age makescorrective postural strategies less effective during slip and intensifies the injuriesassociated with nonfatal falls in older people, and argues that elderly pedestrians areless able to recover their equilibrium in an incipient slip than their youngercounterparts. The paramount aspect of gait dynamics in tribometry is perhaps thatthe most vulnerable population is less able to take required corrective action in thebeginning phases of heel slip.Bjornstig et al. (1997) reported that injury rate caused by slipping on ice or snow

was highest among the elderly, particularly elderly women. According to Tideiksaar(1990), age-related changes (declines) in visual, vestibular, proprioceptive, andmusculoskeletal function were identified as the intrinsic causes. Intrinsic abnorm-alities of the gait cycle (Parkinsonism and peripheral neuropathy) and decreasedvision due to cataracts, in combinations with environmental conditions (that is, lowfrictional resistant or irregular ground surfaces) were found to be the leading causeof slips and trips in the elderly.Andres and Hemeon-Heyer (1988) showed, from a standing balance test, that

the older subjects used their leg muscles for longer time periods and at higheractivation levels than the young subjects. Laughton et al. (2003) demonstratedsimilar results. The implication is that fatigue may occur sooner in an older workerrequired for stand and/or walk. Tanaka et al. (1999) postulated that the base ofsupport area of the older adults is smaller than that of young adults, indicatingpoor balance ability.The ability to walk safely and preserve balance in the event of a slip is dependent

on co-ordination of the visual, proprioceptive and musculoskeletal systems. All thosefunctions deteriorate with age. It may be necessary to suggest a higher frictionrequirement of surface and footwear for the aged population.

7.2. Visual functionsThe visual field is an important physiological parameter involved in gait regulation.Lin et al. (1995) argued that if a slippery condition was not detected within a person’seffective visual field, usually 3 – 5 m ahead, the likelihood of a fall accidentsignificantly increases. Warning signs or markings with high contrast were effectivemethods to direct a person’s gaze downward, so that the effective visual field couldbe readjusted and an accident could be avoided.Cohen and Cohen (1994a, b) demonstrated that tactile cues were most sensitive to

physical measurement of static COF. The potential for an accident could be createddue to a misjudgement of the slipperiness based on initial visual sensing and thelimited time available to make immediate adjustment in gait to accommodate for thehazardous conditions. This is why an unexpected slippery spot (such as water, oildrop, etc.) easily causes slips.Other visual functions, such as visual acuity, contrast sensitivity, dark and bright

adaptation, depth perception, etc. will all deteriorate with age, which will no doubtincrease the likelihood of slips and falls.

590 C. Gao and J. Abeysekera

7.3. Musculoskeletal systemThe reduction in muscle strength due to ageing inhibits certain body movement andrestricts co-ordinated response and balance ability, which can cause increased slipsand falls. The study on the proactive control of gait have shown that proximal (hip/trunk) muscles are the primary contributors to balance control, while studies onreactive balance control in slips accruing at heel strike have shown that the distal andintermediate (leg and thigh) muscles are the key muscles to overcome the balancethreats (Tang et al. 1998).

7.4. Proprioceptive systemThe kinaesthetic system consists of sensors in the muscles, tendons, and joints, whichsense the relative positions and movements of the limbs and of body parts, togetherwith vestibular system in the ear and visual feedback to maintain stable posture andbalance. This is the system that keeps us upright when we stumble over a rock. Theposition sense that allows us such control is provided primarily via the afferent inputof two sensory receptors located in skeletal muscles, the muscle spindle and Golgitendon organs (GTO). Muscle contraction is controlled by alpha and gamma motorneurons (Athabasca University 2003).

7.5. Balance abilityHuman balance ability during standing (static balance) and walking (dynamicbalance) is dependent on feedback from the vestibular, visual and proprioceptivesystems. Normal postural sway of humans is an important response to thedisplacement change of the body’s centre of gravity. It helps to erect body andmaintain upright stability through exciting postural reflex. However, with thedeterioration of visual, musculoskeletal, proprioceptive functions, the posture swaywill be beyond limits acting to preserve balance, resulting in postural unbalance andfalls. Age-related disturbance or sway has been attributed to visual, vestibular andproprioceptive dysfunction. Pyykko et al. (1990) argues that vestibular inflex governs65% of the body sway during sudden perturbation, whilst 35% is accounted for byvisual and proprioceptive inflex.

Postural control is a complex sensorimotor function requiring the centralintegration in the brainstem of information from multiple sensory afferences (visual,vestibular, and somatosensory) leading to the selection and execution of specificmotor response. This balance function applies to the gaze stabilization by vestibulo-ocular reflex and to the posture stabilization by the vestibulospinal reflex (Gauchardet al. 2001). A perturbation or an alteration of the postural regulation atinformation, integration or motor levels can induce equilibrium dysfunction.

Lord et al. (1996) argued that increased variability of stride to stride time was arisk factor associated with falling in older people. Because of the frequent changes ofstride rhythm, postural sway increases correspondingly and the body needs tofrequently adjust posture through immediate postural reflex. However, due to age-induced changes in the nervous system, the reaction time of reflex is lengthened,resulting in the risk of falls. Latencies for corrective reflex responses for healthyyoung men in the recovery from tripping have been reported to be 60 – 140 ms. Dueto the defective co-activation of the functional stretch reflexes, the very elderly relymostly on slower (latency 120 – 200 ms) visual control of balance, which contributeto an increased risk of falling (Pyykko et al. 1990). However, vision may be the onlysensory mode allowing a person to predict the potential slipperiness of a surface

591Slip and fall accidents

before stepping onto it (Gronqvist 1999). Several rating scales for assessing balancehave been summarised by Gronqvist et al. (2001). Tisserand (1985) suggests moreresearch work on psycho-physiological equilibrium on slippery surfaces is needed.

7.6. ExperienceThe subjects’ experience is an important intrinsic factor, by allowing them to adaptthemselves to the environment. An experienced subject on a slippery surface canadapt his/her gait to move with success on this type of ground (Gauchard et al.2001). The study of strategies for dynamic stability during locomotion on a slipperysurface showed that prior experience with slip perturbations allows subsequentmodification and knowledge of the surface condition results in proactive adjustmentsto safely transverse the slippery surface (Marigold and Patla 2002). Recent researchshowed that slips and falls on ice and snow reduce with increased experience of livingin winter climate (Gao and Abyesekera 2003).

8. Environmental factors (extrinsic factors)

Extrinsic factors involved in fall mechanisms are summarised by Gauchard et al.(2001), including organizational and environmental factors. Environmental factors,such as temperature, humidity, snowfall, Gulf Stream, lighting conditions, etc. willaffect the whole walking system. For example, in Scandinavian regions, thechangeable climate due to the Gulf Stream in winter season leads to a variety of roadsurface conditions, including snow, melting snow, ice, melting ice and the mixture ofthese types. According to Bentley and Haslam (1998), slip, tumble, and fall accident(STFA) were most common during winter months, and slipping accidents tended tocluster on single days where heavy snowfall or ice made conditions particularlyhazardous. De Koning et al. (1992) showed that different ice surface temperatureaffected the COF. Secondly, cold environment will influence footwear (sole) slipresistant properties, which is currently a neglected area of research. Thirdly, thethermal environment also impacts human neuromuscular system (e.g., bodycooling), which, in turn, affects human gait on icy and snowy surface.Furthermore, in Arctic regions, the winter is long and dark. Short and insufficient

daylight results in lack of visual cues for postural adjustment, which might be animportant contributing factor to slips and falls on icy and snowy surface in theregion. As a result, much more preventive counter-slip measures should be taken inthe relatively dark environment than in other parts of the world.

9. Systematic prevention of slip and fall accidents on icy and snowy

surfaces

Gronqvist (1995a), Abeysekera and Gao (2001) and Gronqvist et al. (2001) discussedprimary and secondary risk factors. A systematic analysis (figure 1) of the slips andfalls on icy and snowy surfaces provides us the possibility of a multi-facetedapproach to systematic prevention.

(1) Winter footwear: the use of specifically designed proper anti-slip footwearand anti-skid device at certain circumstances, taking into consideration solematerial, tread design, sole hardness, roughness, wear and tear, COG,wearability (Gao and Abeysekera 2002), flexibility, durability, etc.

(2) Icy and snowy surfaces: effective snow clearing, anti-slip materials spreading(sand, gravel, etc.), which is feasible for main roads. Kelkka (1995) and

592 C. Gao and J. Abeysekera

Kobayashi et al. (1997) indicated that in some northern countries, goodpreventive results were reached by warming the busiest walkways electrically.

(3) Footwear/ice interaction: validation, standardization of COF measurementmethod and equipment at footwear/ice interface.

(4) Human factors and biomechanics: provision of walk aids, gait and balancerehabilitation and training, slip risk warnings (improving the awareness) forspecial population (e.g., the elderly); ‘padding’ of older women, who are themost common victims of slips and falls on icy roads (Bjornstig et al. 1997).

(5) Environment: improvement of visual and lighting conditions, provision ofweather information service (Penttinen et al. 1998), etc.

(6) Management: measures such as prioritizing winter road maintenance, e.g.,paying more attention to busiest and slope road, use of road slipperinesswarning signs, registration systems for pedestrian slipping and fallingincidents and accidents should be improved in order to get more detailedinformation for statistics analysis and aetiology determination.

(7) Training and behaviour approach: safe behaviour and practices, to avoidshortcuts, rushing and lax attitudes while walking and working on icy andsnowy surfaces.

10. Unresolved issues and further research needs

A recent international symposium on ‘The measurement of slipperiness’suggested a number of crucial research needs in surface aspects, human sensoryand biomechanical aspects, systems modelling, preventive validity, and standar-dization (Chang et al. 2003). However, those needs do not specifically involveslipperiness on ice and snow. The following unresolved issues are identified andmore related to slips and falls on ice and snow, which necessitate furtherresearch.

(1) Most researches and measurements done before concerning the COF offootwear did not separate the anti-slip effect of various footwear properties,such as sole material, hardness (e.g., hardening in cold environment),roughness and tread design, wear and tear completely. In fact, the frictionaleffect was the combined one, whether one, two or more variables contributedto it were not quite conclusive. Further research is necessary to differentiateand quantify the effects.

(2) Since most slips and falls happen during heel strike and toe off phases,therefore, the sole tread pattern near the heel and toe part plays an importantrole in forward and rearward slip. It is postulated that enhancement of COFof those two parts should be more effective in slip prevention, especially onice and snow.

(3) The relationship between footwear/surface contact area, pressure and COF isa debatable area. Further research work is needed.

(4) Most of the research on sole material and tread design was based uponfloors/lubricated floors. Little research has been done on material and treaddesign for use on icy and snowy surfaces. Although the same principles mayapply, there may be different requirements in order to improve COF.According to the above discussion, current soling materials seem to be notsafe enough on ice, especially on melting ice, therefore, more slip resistant

593Slip and fall accidents

material and sole design should be developed further, e.g., fasten anti-slipmaterials at certain parts of sole.

(5) There is little research, and no standard available governing the slipperinessmeasurement on icy and snowy surfaces, even though ice surface near meltingpoint is more slippery than lubricated floors, and the population associatedwith slip on ice is much more than that on contaminated indoor floors.Therefore, there are desperate needs to explore the interaction mechanismsbetween icy and snowy surface and footwear, to set up relevant standard andto develop appropriate footwear.

(6) The anti-slip effect of spreading different anti-slip materials on icy and snowysurfaces has still not been adequately researched.

(7) It is common that slips and falls occur more frequently while walking inhaste, in turning and in going up and down slopes. Human balance abilityand gait biomechanics on icy slope need attention.

(8) Environmental factors influence both footwear and icy surface slip resistantproperties. The cold thermal environment also affects human neuromuscularand balance system (i.e., cooling effect), which, in turn, affects human gait onicy and snowy surface, which warrants further study.

ReferencesABEYSEKERA, J. and GAO, C. 2001, The identification of factors in the systematic evaluation of

slip prevention on icy surfaces, International Journal of Industrial Ergonomics, 28, 303 –313.

ABEYSEKERA, J. and KHAN, Z. 1998, Slipping and falling accidents on icy surfaces: a case studyfrom northern Sweden, in I. Holmer and K. Kuklane (eds), Problems with Cold(Stockholm: The National Institute for Working Life), 201 – 204.

AHAGON, A., KOBAYASHI, T. and MISAWA, M. 1988, Friction on ice, Rubber and Chemistry andTechnology, 61, 14 – 35.

ANDERSSON, R. and LAGERLOF, E. 1983, Accident data in the new Swedish information systemon occupational injuries, Ergonomics, 26, 33 – 42.

ANDRES, R. and HEMEON-HEYER, S. 1988, Slip, trip, and fall risks for the older worker:implications from balance sheets, Proceedings of the Human Factors Association ofCanada 21st Annual Conference, 14 – 16 September 1988, Edmonton, Alberta, 141 – 143.

ASTM D2047-99. 2001, Standard test method for static coefficient of friction of polish-coatedfloor surfaces as measured by the James machine, Annual Book of ASTM Standards,15.04 (West Conshohocken, PA: ASTM).

Athabasca University. 2003, Muscle contraction (online). Athabasca University. Available at:http://psych.athabascau.ca/html/Psych402/Biotutorials/31/part1.html (Accessed 19Sept 2003).

BENTLEY, T. A. and HASLAM, R. A. 1996, Outdoor falls in postal delivery employees: Asystematic analysis of in-house accident data, in A. F. Ozok and G. Salvendy (eds),Advances in Applied Ergonomics, (West Lafayette, IN: USA Publishing Corporation),204 – 207.

BENTLEY, T. A. and HASLAM, R. A. 1998, Slip, trip and fall accidents occurring during thedelivery of mail, Ergonomics, 41, 1859 – 1872.

BENTLEY, T. A. and HASLAM, R. A. 2001, Identification of risk factors and countermeasures forslip, trip and fall accidents during the delivery of mail, Applied Ergonomics, 32, 127 –134.

BERGQVIST, K. and ABEYSEKERA, J. 1994, Ergonomics aspects of safety shoes worn in a coldclimate, Proceedings of the 3rd Pan-Pacific Conference on Occupational Ergonomics(Ergonomics for Quality Life), 13 – 17 November 1994, Seoul, Korea, 590 – 594.

594 C. Gao and J. Abeysekera

BJORNSTIG, U., BJORNSTIG, J. and DAHLGREN, A. 1997, Slipping on ice and snow—Elderlywomen and young men are typical victims, Accident Analysis and Prevention, 29, 211 –215.

BRUCE, M., JONES, C. and MANNING, D. P. 1986, Slip resistance on icy surfaces of shoes,crampons and chains—A new machine, Journal of Occupational Accidents, 7, 273 – 283.

CEN/TC 161/WG 3 (N52). 2003, prENV 13287, Personal Protective Equipment—Footwear—Slip Resistance.

CHANG, W. R., COURTNEY, T. K., GRONQVIST, R. and REDFERN, M. 2003, Measuringslipperiness—discussion on the state of the art and future research, in W. R. Chang,T. K.Courtney, R. Gronqvist and M. S. Redfern (eds), Measuring Slipperiness: HumanLocomotion and Surface Factors (London: Taylor & Francis), 165 – 171.

CHANG, W. R., GRONQVIST, R., LECLERCQ, S., BRUNGRABER, R. J., MATTKE, U., STRANDBERG, L.,THORPE, S. C., MYUNG, R., MAKKONEN, L. and COURTNEY, T. K. 2001a, The role offriction in the measurement of slipperiness, Part 2: Survey of friction measurementdevices, Ergonomics, 44, 1233 – 1261.

CHANG, W. R., GRONQVIST, R., LECLERCQ, S., MYUNG, R., MAKKONEN, L., STRANDBERG, L.,BRUNGRABER, R. J., MATTKE, U. and THORPE, S. C. 2001b, The role of friction in themeasurement of slipperiness, Part 1: Friction mechanisms and definition of testconditions, Ergonomics, 44, 1217 – 1232.

CHANG, W. R., KIM, I. J., MANNING, D. P. and BUNTERNGCHIT, Y. 2001c, The role of surfaceroughness in the measurement of slipperiness, Ergonomics, 44, 1200 – 1216.

CHIOU, S., BHATTACHARYA, A. and SUCCOP, P. A. 2000, Evaluation of workers’ perceived senseof slip and effect of prior knowledge of slipperiness during task performance on slipperysurfaces, American Industrial Hygiene Association Journal, 61, 492 – 500.

COHEN, H. H. and COHEN, D. M. 1994a, Perceptions of walking surface slipperiness underrealistic conditions utilizing a slipperiness rating scale, Journal of Safety Research, 25,27 – 31.

COHEN, H. H. and COHEN, D. M. 1994b, Psychophysical assessment of the perceivedslipperiness of floor tile surfaces in laboratory setting, Journal of Safety Research, 25,19 – 26.

DAVIS, P. R. 1983, Human factors contributing to slips, trip and falls, Ergonomics, 26, 51 – 59.DE KONING, J. J., DE GROOT, G. and SCHENAU, J. V. I. 1992, Ice friction during speed skating,

Journal of Biomechanics, 25, 565 – 571.DE LANGE, A. and GRONQVIST, R. 1997, Bridging the gap between mechanical and

biomechanical slip test methods, in, P. Seppala, T. Luopajarvi, C.-H. Nygard and M.Mattila (eds), Proceedings of the 13th Triennial Congress of the International ErgonomicsAssociation (Vol. 3), (Helsinki: Finnish Institute of Occupational Health), 368 – 370.

ENGLISH, W. 1994, The validation of slipmeters, in S. A. Robertson (ed), ContemporaryErgonomics 1994, (London: Taylor & Francis), 347 – 352.

ENGLISH, W. 1996, Continuing progress toward an ergonomically sound shoe traction tester, inA. Mital (ed), Advances in Occupational Ergonomics and Safety I (vol.2), (InternationalSociety for Ergonomics and Safety), 919 – 924.

GAO, C., ABEYSEKERA, J., HIRVONEN, M. and ASCHAN, C. 2003, The effect of footwear soleabrasion on the coefficient of friction on melting and hard ice, International Journal ofIndustrial Ergonomics, 31, 323 – 330.

GAO, C., ABEYSEKERA, J., HIRVONEN, M. and GRONQVIST, R. 2004, Slip resistant properties offootwear on ice, Ergonomics, in press.

GAO, C. and ABEYSEKERA, J. 2002, The assessment of the integration of slip resistance, thermalinsulation and wearability of footwear on icy surfaces, Safety Science, 40, 613 – 624.

GAO, C. and ABEYSEKERA, J. 2003, Slips and falls on ice and snow in relation to experience inwinter climate and winter sport, Safety Science, in press.

GARD, G. and LUNDBORG, G. 1994, Pedestrians transportation problems on different slipperysurfaces during winter and the functionality of different anti-skid devices, Proceedings ofInternational Conference on Developing and Commercial Utilisation of Technologies inPolar Regions (Polar-Tech), Sweden.

GARD, G. and LUNDBORG, G. 2000, Pedestrians on slippery surfaces during winter – methods todescribe the problems and practical tests of anti-skid devices, Accident Analysis andPrevention, 32, 455 – 460.

595Slip and fall accidents

GAUCHARD, G., CHAU, N., MUR, J. M. and PERRIN, P. 2001, Falls and working individuals: roleof extrinsic and intrinsic factors, Ergonomics, 44, 1330 – 1339.

GNORICH, W. and GROSCH, K. A. 1975, The friction of polymers on ice, Rubber Chemistry andTechnology, 48, 527 – 537.

GRONQVIST, R. 1995a, A Dynamic Method for Assessing Pedestrian Slip Resistance, DoctoralDissertation, Finnish Institute of Occupational Health, Helsinki, Finland.

GRONQVIST, R. 1995b, Mechanisms of friction and assessment of slip resistance of new andused footwear soles on contaminated floors, Ergonomics, 38, 224 – 241.

GRONQVIST, R. 1999, Slips and Falls, in S. Kumar (ed), Biomechanics in Ergonomics, (London:Taylor & Francis), 351 – 375.

GRONQVIST, R. and HIRVONEN, M. 1992, Slip resistance of winter shoes on icy surfaces,Proceedings of The Fourth Scandinavian Symposium on Protective Clothing AgainstChemicals and Other Health Hazards, (Helsinki: NOKOBETEF), 123 – 129.

GRONQVIST, R. and HIRVONEN, M. 1995, Slipperiness of footwear and mechanisms of walkingfriction on icy surfaces, International Journal of Industrial Ergonomics, 16, 191 – 200.

GRONQVIST, R., ABEYSEKERA, J., GARD, G., HSIANG, S. M., LEAMON, T. B., NEWMAN, D. J.,GIELO-PERCZAK, K., LOCKHART, T. E. and PAI, C. Y.-C. 2001, Human-centredapproaches in slipperiness measurement, Ergonomics, 44, 1167 – 1199.

GRONQVIST, R., HIRVONEN, M. and TUUSA, A. 1993, Slipperiness of the shoe-floor interface:comparison of objective and subjective assessments, Applied Ergonomics, 24, 258 – 262.

GRONQVIST, R., ROINE, J., JARVINEN, E. and KORHONEN, E. 1989, An apparatus and a methodfor determining the slip resistance of shoes and floor by simulation of human footmotions, Ergonomics, 32, 979 – 995.

HANSON, J. P., REDFERN, M. S. and MAZUMDAR, M. 1999, Predicting slips and falls consideringrequired and available friction, Ergonomics, 42, 1619 – 1633.

HASLAM, R. A. and BENTLEY, T. A. 1999, Follow-up investigations of slip, trip and fallaccidents among postal delivery workers, Safety Science, 32, 33 – 47.

HIRVONEN, M. and GRONQVIST, R. 1998, Influence of wear on footwear solings in slip resistanceassessment, Proceedings of International Fall Protection Symposium & Exhibition,September 15 – 18, 1998, Germany, (Stadthalle Wuppertal: Bau-BerufsgenossenschaftRheinland und Westfalen Zentrum fur Sicherheitstechnik), 108 – 109.

HIRVONEN, M., LESKINEN, T., GRONQVIST, R. and SAARIO, J. 1994, Detection of near accidents bymeasurements of horizontal acceleration of the trunk, International Journal of IndustrialErgonomics, 14, 307 – 314.

JAMES, D. I. 1983, Rubbers and plastics in shoes and flooring: the importance of kineticfriction, Ergonomics, 26, 83 – 99.

JUNG, K. 1992, Factors influencing anti-slip properties of footwear, Proceedings of The FourthScandinavian Symposium on Protective Clothing Against Chemicals and Other HealthHazards, (Helsinki: NOKOBETEF), 112 – 122.

JUNG, K. and FISCHER, A. 1993, An ISO test method for determining slip resistance offootwear, determination of its precision, Safety Science, 16, 115 – 127.

JUNG, K. and SCHENK, H. 1990, An international comparison of test method for determiningthe slip resistance of shoes, Journal of Occupational Accidents, 13, 271 – 290.

KELKKA, M. 1995, Pedestrian slipping accidents and their prevention, Report, HighwayEngineering, Helsinki University of Technology.

KEMMLERT, K. and LUNDHOLM, L. 2001, Slips, trips and falls in different work groups withreference to age and a prevention perspective, Applied Ergonomics, 32, 149 – 153.

KIRTLEY, C. 1999, Teach-in’99: Centre of Mass and Base of Support (online). Available at:http://guardian.curtin.edu.au/cga/teach-in/bos (Accessed 22 Sept 2003).

KOBAYASHI, S., YOSHIDA, N. and YAMADA, T. 1997, Experimental studies on the prevention ofroad freezing using electric heat, in M. Izumi, T. Nakamura and R. C. Sack (ed) SnowEngineering: Recent Advances, (Rotterdam: Balkema), 375 – 378.

KOBAYASHI, T., NOHGUCHI, Y., KAWASHIMA, K. and IKARASHI, T. 1996, Slip tests between thesurfaces of snow/ice and some kinds of shoes, Proceedings of the 3rd InternationalConference on Snow Engineering, Japan.

LAUGHTON, C. A., SLAVIN, M., KATDARE, K., NOLAN, L., BEAN, J. F., KERRIGAN, D. C., PHILLIPS,E., LIPSITZ, L. A. and COLLINS, J. J. 2003, Aging, muscle activity, and balance control:physiologic changes associated with balance impairment, Gait and Posture, in press.

596 C. Gao and J. Abeysekera

LEAMON, T. B. 1992a, The reduction of slip and fall injuries: Part I – Guidelines for thepractitioner, International Journal of Industrial Ergonomics, 10, 23 – 27.

LEAMON, T. B. 1992b, The reduction of slip and fall injuries: Part II – The scientific basis(knowledge base) for the guide, International Journal of Industrial Ergonomics, 10, 29 –34.

LEAMON, T. B. and MURPHY, P. L. 1995, Occupational slips and falls: more than a trivialproblem, Ergonomics, 38, 487 – 498.

LEAMON, T. B. and SON, D. 1989, The natural history of a microslip, in A. Mital (ed), Advancesin Industrial Ergonomics and Safety I, (London: Taylor & Francis), 633 – 638.

LECLERCQ, S. 1999, The prevention of slipping accidents: a review and discussion of workrelated to the methodology of measuring slip resistance, Safety Science, 31, 95 – 125.

LECLERCQ, S., TISSERAND, M. and SAULNIER, H. 1995, Assessment of slipping resistance offootwear and floor surfaces: Influence of manufacture and utilization of the products,Ergonomics, 38, 209 – 219.

LECLERCQ, S., TISSERANT, M. and SAULNIER, H. 1994, Slip resistant footwear: A means for theprevention of slipping, in F. Aghazadeh (ed), Advances in Industrial Ergonomics andSafety VI, (London: Taylor & Francis), 329 – 337.

LIN, L. J., CHIOU, F. T. and COHEN, H. H. 1995, Slip and fall accident prevention: a review ofresearch, practice, and regulations, Journal of Safety Research, 26, 203 – 212.

LLOYD, D. G. and STEVENSON, M. G. 1989, Measurement of slip resistance of shoes on floorsurfaces, Part 2: Effect of a bevelled heel, Journal of Occupational Health Safety –Australia and New Zealand, 5, 229 – 235.

LOCKHART, T. E., WOLDSTAD, J. C. and SMITH, J. L. 2003, Effects of age-related gait changes onthe biomechanics of slips and falls, Ergonomics, 46, 1136 – 1160.

LORD, S. R., LLOYD, D. G. and LI, S. K. 1996, Sensor-motor function, gait patterns and falls incommunity-dwelling women, Age and Ageing, 25, 292 – 299.

LUND, J. 1984, Accident falls at work, in the home and during leisure activities, Journal ofOccupational Accidents, 6, 181 – 193.

MAKKONEN, L. 1994, Application of a new friction theory to ice and snow, Annals ofGlaciology, 19, 155 – 157.

MANNING, D. P. and JONES, C. 2001, The effect of roughness, floor polish, water, oil and ice onunderfoot friction: current safety footwear solings are less slip resistant thanmicrocellular polyurethane, Applied Ergonomics, 32, 185 – 196.

MANNING, D. P., COOPER, J. E., STIRLING, I., JONES, C. M., BRUCE, M. and MCCAUSLAND, P. C.1985, Studies on the footpads of the polar bear (Ursus maritimus) and their possiblerelevance to accident prevention, Journal of Hand Surgery, 10, 303 – 307.

MANNING, D. P., JONES, C. and BRUCE, M. 1991, A method of ranking the grip of industrialfootwear on water wet, oily and icy surfaces, Safety Science, 14, 1 – 12.

MANNING, D. P., JONES, C., BRUCE, M. 1990, Proof of shoe slip-resistance by a walking tractiontest, Journal of Occupational Accidents, 12, 255 – 270.

MANNING, D., JONES, C. and BRUCE, M. 1985, Boots for oily surfaces, Ergonomics, 28, 1011 –1019.

MANNING, D. P. and JONES, C. 1994, The superior slip-resistance footwear soling compoundT66/103, Safety Science, 18, 45 – 60.

MARIGOLD, D. S. and PATLA, A. E. 2002, Strategies for dynamic stability during locomotion ona slippery surface: effects of prior experience and knowledge, Journal of Neurophysiol-ogy, 88, 339 – 353.

MYUNG, R. and SMITH, J. L. 1997, The effect of load carrying and floor contaminants on slipand fall parameters, Ergonomics, 40, 235 – 246.

MYUNG, R., SMITH, J. L. and LEAMON, T. B. 1993, Subjective assessment of floor slipperiness,International Journal of Industrial Ergonomics, 11, 313 – 319.

NOGUCHI, T. and SAITO, J. 1996a, Change of acceleration during gait, Proceedings of the 36thConference of Hokkaido Mechanic Society, Japan, 39 – 40.

NOGUCHI, T. and SAITO, J. 1996b, Proper arrangement system fastened anti-slip device onwinter shoes for the disabled – basic ways of thinking and trial application, Proceedingsof the 36th Conference of Hokkaido Mechanic Society, Japan, 41 – 42.

597Slip and fall accidents

NOGUCHI, T. 2000, Method of expressing slipperiness in gait, Proceedings of the NationalOccupational Injury Research Symposium (NOIRS 2000), Pittsburgh, Pennsylvania,USA, October 17 – 19.

PENTTINEN, M., BYGARD, M., HARJULA, V. and ESKELINEN, M. 1998, Pedestrians slippingaccidents, detecting and forecasting of hazardous road conditions – Informationcampaign on the Helsinki region, Research Notes, Technical Research Centre (VTT) ofFinland.

PETRENKO, V. F. 1994, The effect of static electric fields on ice friction, Journal of AppliedPhysics, 76, 1216 – 1219.

PYYKKO, I., JANTTI, P. and AALTO, H. 1990, Postural control in elderly subjects, Age and Ageing,19, 215 – 221.

REDFERN, M. S. and DIPASQUALE, J. D. 1997, Biomechanics of descending ramps, Gait andPosture, 6, 119 – 125.

REDFERN, M. S., CHAM, R., GIELO-PERCZAK, K., GRONQVIST, R., HIRVONEN, M., LANSHAMMAR,H., MARPET, M., PAI, C. Y.-C. and POWERS, C. 2001, Biomechanics of slips, Ergonomics,44, 1138 – 1166.

REDFERN, M. S., HANSON, J. P. and MAZUMDAR, M. 1998, Comparing required and availablefriction during walking to prevent falls, in S. Kumar (ed) Advances in OccupationalErgonomics and Safety, (Amsterdam: IOS Press), 92 – 96.

ROBERTS, A. D. and RICHARDSON, J. C. 1981, Interface study of rubber-ice friction, Wear, 67,55 – 69.

ROBERTS, A. D. 1981, Rubber-ice adhesion and friction, Journal of Adhesion, 13, 77 – 86.SJOGREN, H. and BJORNSTIG, U. 1991, Injuries to the elderly in the traffic environment, Accident

Analysis and Prevention, 23, 77 – 86.STEVENSON, M. G. 1997, Evaluation of the slip resistance of six types of women’s safety shoe

using a newly developed testing machine, Journal of Occupational Health Safety –Australia and New Zealand, 13, 175 – 182.

STRANDBERG, L. 1985, The effect of conditions underfoot on falling and overexertion accidents,Ergonomics, 28, 131 – 147.

STRANDBERG, L. and LANSHAMMAR, H. 1981, The dynamics of slipping accidents, Journal ofOccupational Accidents, 3, 153 – 162.

SWENSEN, E. E., PURSWELL, J. L., SCHLEGEL, R. E. and STAMEVICH, R. C. 1992, Coefficient offriction and subjective assessment of slippery work surfaces, Human Factors, 34, 67 – 77.

TANAKA, T., TAKEDA, H., IZUMI, T., INO, S. and IFUKUBE, T. 1999, Effects on the location of thecentre of gravity and the foot pressure contribution to standing balance associated withageing, Ergonomics, 42, 997 – 1010.

TANG, P. F., WOOLLACOTT, M. H. and CHONG, R. K. Y. 1998, Control of reactive balanceadjustments in perturbed human walking: roles of proximal and distal postural muscleactivity, Experimental Brain Research, 119, 141 – 152.

TANIGUCHI, T., INABA, Y., MURAI, S. and NISHIZAWA, T. 1997, Mechanism of ice debonding onan asphalt mixture containing rubber, in M. Izumi, T. Nakamura and R. C. Sack (ed)Snow Engineering: Recent Advances, (Rotterdam: Balkema), 379 – 386.

TIDEIKSAAR, R. 1990, The biomechanical and environmental characteristics of slips, stumbles,and falls in the elderly, in B. E. Gray (ed), Slips, Stumbles, and Falls: PedestrianFootwear and Surfaces, (Philadelphia, PA: ASTM), 17 – 27.

TISSERAND, M. 1985, Progress in the prevention of falls caused by slipping, Ergonomics, 28,1027 – 1042.

TWOMEY, J. M., SMITH, A. E. and REDFERN, M. S. 1995, A predictive model for slip resistanceusing artificial neural networks, IIE Transactions, 27, 374 – 381.

WETTLAUFER, J. S. and DASH, J. G. 2000, Melting below zero, Scientific American, 282(2), 34 –37.

WILSON, M. P. and PERKINS, P. J. 1985, Evaluation of a slip-resistance test for shoes,Ergonomics, 28, 1081 – 1085.

598 C. Gao and J. Abeysekera