Table 1 Characteristics of included studies

Donald 2000 [66]

RCT

N = 54; Female = 44; Age = 83; Admitted for rehabilitation in a hospital, UK.

Hospital duty ‘Flotex® 200’ carpet vs. latex vinyl square tile; sub-floors NR

Incidence of falls; Injuries partially reported; Satisfaction of cleaning.

Drahota 2013 [14]

Cluster RCT

N = 448; Female = 353; Age = 81; NHS patients in geriatric wards across 8 hospitals, UK. Floors with a ramp test slip rating of ‘R9’ (slippery when wet).

Sports floor (8.3 mm Tarkett Omnisports Excel) vs. 2 mm vinyl (3 sites) or 2 mm thermoplastic tiles (1 site); Concrete sub-floors

Injurious falls rate; Injury severity; Fall rate; Adverse events; No. of fallers and falls.

Mackey 2019 [16]

RCT

N = 357; Female = 151; Age = 81.5; Residents living in study rooms of a care home, Canada.

Novel shock-absorbing floor (2.54 cm SmartCells) with 2 mm hospital-grade vinyl vs. 2.54 cm plywood with 2 mm hospital-grade vinyl; Concrete sub-floors

Serious fall-related injury; Minor fall-related injury; any fall-related injury; Falls; Fractures.

Gustavsson 2018 [15]

Prospective cohort

N = 114; Female = 80; Age = 85; Residents of a care home, Sweden.

Novel shock-absorbing floor (12 mm Kradal) vs. standard vinyl/lino/ceramic; Concrete sub-floors

Injury rate per fall; Falls per 1000 PBD; No. of falls with injury; Injury severity

Hanger 2017 [57]

Prospective cohort

N (bedroom fallers) = 178; Female = 112; Age = 83; Admitted to a geriatric ward with a focus on medical and rehabilitation needs in a hospital, NZ.

Novel shock-absorbing floors (12 mm Kradal & 25 mm SmartCells), and a sports floor (8 mm Tarkett Omnisport Excel) vs. 3–4 mm vinyl; Concrete sub-floors

Fall rate per 1000 PBD; Fall-related injury rate per 1000 PBD; Injury severity; Injury type

Hanger 2020 [68]

Controlled cohort study

N (injuries) = 163; Female = NR; Age = NR; Study included any staff injury occurring on a study ward where mechanism of injury might be related to flooring in 4 older persons’ wards in a hospital, NZ.

Novel shock-absorbing floors (12 mm Kradal & 25 mm SmartCells), and a sports floor (8 mm Tarkett Omnisport Excel) vs. standard vinyl; Concrete sub-floors

Staff injuries.

Harris 2017 [65]

Prospective cohort

N = 36; Female = 30%; Age > 50% were 60–79 years; Assigned to one of six rooms in a telemetry unit for heart patients in a hospital; Country not clear.

Carpet tile (tufted loop with thermoplastic composite polymer backing vs. vinyl composition tile; Sub-floors NR

Preferences / satisfaction; “risk of falling” assessment ratings; behavioural mapping

Healey 1994 [12]

Retrospective cohort

N (falls) = 213; Female = 68; Age (median) = 85; Study involved a random sample of 225 accident forms from a care of the elderly unit over a four year period in a hospital, UK.

Carpet (varied but all single fibres rather than looped, with thin underlay) vs. vinyl; sub-floors NR

Fall related injury

Knoefel 2013 [64]

Retrospective/ Prospective cohort

N (falls) = 167; Female = 78; Age = 74; All documented falls on novel flooring and every 3rd fall on regular flooring.in a care home, USA.

Novel shock-absorbing flooring (SmartCells) vs. “regular floor”; sub-floors NR

No. of falls with injury; Type of injury; No. of fractures

Simpson 2004 [13]

Prospective cohort

N (falls) = 6641; Female = NR; Age = NR; Residents living in 34 care homes for older people, UK.

Carpet (with concrete or wooden sub-floor); Wooden sub-floor (with or without carpet); Uncarpeted (with concrete or wooden sub-floor); Concrete sub-floor (with or without carpet).

No. of falls per room; fractures per 100 falls; No. of hip fractures

Wahlström 2012 [67]

Controlled before-after study

N = 153; Female = 153; Age = 46; All nursing assistants; Males excluded post-hoc due to low response rate; 2 geriatric care centres, Sweden.

1.5 mm homogenous polyvinyl chloride covering with 2.5 mm foam backing (4 mm total) vs. 2 mm homogenous polyvinyl chloride covering; Concurrent control: 2.5 mm linoleum

Pain ratings in lower back, hips, knees, and feet at 6 weeks, 1 and 2 years. Adverse events not measured.

Warren 2013 [63]

Interrupted time series

N = 4641; Female = 2694; Age = 81; Admitted to the geriatric ward in a hospital (sub-acute), NZ.

5 mm carpet (tiles with loop pile) vs. 5 mm vinyl; concrete sub-floors.

Falls rate per 1000 PBD; fall related injuries; No. of fractures

Drahota 2011 [69]

Exploring perceptions via semi-structured face-to-face interviews; Thematic content analysis

N = 12 patients (11 female), 8 visitors, 77 hospital staff (67 female)^a; Study included patients/visitors in the ‘study area’, orientated to person/time/place, and staff with experience of working in the study area in 8 elderly care hospital wards, UK.

Tarkett Omnisports Excel (8.3 mm sports floor); vinyl.

The problem of falls; Protecting patients with floors; Environmental comfort; Push and pull challenges; Walking and mobilising; Cleaning and maintenance; The novelty factor; Adapting to a compliant floor; Installation.

Gustavsson 2017 [70]

Exploratory study of shared experiences; Two focus groups; Qualitative content analysis.

N = 8 enrolled nurses; Female = 8; Age range = 40–60; Enrolled nurses with 12 months experience of impact absorbent flooring in a nursing home, Sweden.

Kradal (12 mm closed cell tiles).

The problem of falls; Protecting patients with floors; Environmental comfort; Push and pull challenges; The novelty factor; Adapting to a compliant floor.

Gustavsson 2018 [71]

Grounded Theory study using in-depth semi-structured individual interviews.

N = 8 residents; Female = 6; Age range = 74–94; Study included residents with sufficient cognitive ability to participate in an interview, who lived in residential care with compliant flooring for at least 3 months; 2 nursing homes, Sweden.

Kradal (12 mm closed cell tiles).

The problem of falls; Protecting patients with floors.

Lachance 2018 [72]

Exploring perceptions via in-depth, semi-structured face-to-face interviews, analysed via a thematic framework method.

N = 18 senior managers; Female = 15; Age range = 37–66; In a senior management role at a Long-Term Care site in the locality; Involved in clinical and operational aspects, including implementing fall injury prevention interventions; 16 nursing homes, Canada.

Purpose-designed compliant flooring (a padded layer, generally found beneath vinyl or carpet).

Protecting patients with floors; Environmental comfort; Push and pull challenges; Walking and mobilising; Cleaning and maintenance; The novelty factor; Adapting to a compliant floor; Installation; Costs and funding.

Rigby 2012 [73]

Exploring experiences via guided tours and conversations lasting 1–6 h at each site. Extensive note-taking post visit of observations and conversations. Analysis method not described.

Hosts (N) = 14 nurse managers or ward sisters; 1 palliative care specialist nurse; 1 medical consultant. Plus other ‘interested staff’. Host staff all ‘older’ females; Host staff provided a guided tour, other ‘interested staff’ also joined in discussions; 7 hospices and 9 care homes in England & Australia.

Carpets and vinyl.

Push and pull challenges; Adapting to a compliant floor.

Lange 2012 [75]

Cost utility analysis; Perspective: Societal; Model time horizon: 1 year; Life of floor: 20 years; Discount rate: 3%; Currency: 2011 SEK; Model: decision tree.

59 nursing home residents, Sweden

Kradal vs. linoleum

Costs: purchase, installation, medical costs associated with hip fracture and death, healthcare consumption; QALY loss due to hip fracture/death. ICER.

Latimer 2013 [78]

Cost utility analysis; Perspective: NHS and personal social service; Model time horizon: 15 years; Life of floor: 15 years; Discount rate: 3.5%; Currency: 2009/10 GBP; Model: decision tree. Measurement and valuation via EQ-5D supplemented by assumptions.

452 older adult hospital in-patients, UK

8.3 mm Tarkett Omnisports Excel vs. 2 mm vinyl / 2 mm thermoplastic tiles on concrete

Costs: installation, hospitalisation, falls of different severities, 3 month post-discharge resource use (hospital admissions, outpatient/healthcare visits, place of residence), mortality. QALYs associated with different types of falls. ICER.

Njogu 2008 [77]

Cost utility analysis; Perspective: NR; Model time horizon: 40 years; Life of floor: 40 years; Discount rate: NR; Currency: $ (assuming NZ $, price date NR); Model: decision tree.

Simulated care home residents, NZ

Kradal^b vs. traditional floor

Costs: additional purchase costs (not installation), hip fracture (inpatient and rehabilitation costs), cost of head injury and other fracture reported but not used in analysis. QALY loss due to hip fracture. ICER.

Ryen 2015 [76]

Cost utility analysis; Perspective: Societal; Model time horizon: 10 years; Life of floor: 20 years; Discount rate: 3%; Currency: SEK (price date NR); Model: Markov state.

Simulated care home residents, Sweden

Kradal vs. “standard” floor

Costs: installation, hip fracture (in- and out-patient and general practitioner costs, rehabilitation/physical therapy, transport), added life years. QALY weights for healthy and hip fracture states. ICER.

Zacker 1998 [74]

Cost-effectiveness and cost-benefit analysis; Perspective: Societal; Model time horizon: 40 years; Life of floor: 20 years; Discount rate: 5%; Currency: 1995 USD; Model: decision tree implicit.

Simulated high risk care home residents, USA

> 25 mm dual stiffness underlay vs. standard concrete floor

Costs: manufacture, installation, replacement, resident screening; Benefits: direct medical costs avoided, indirect morbidity avoided, indirect mortality avoided as a result of preventing hip fracture.