19%
fewer falls where residents were measured every two weeks, versus a 3% rise at control facilities.
9.3%
fewer rehospitalizations at the facilities that screened most, about 17 fewer per 1,000 residents.
$0.6–1.3M
modeled annual cost avoidance per 1,000 residents, before staff time, liability and lost occupancy.
Falls cost the US health system about $50 billion a year, and the organizations that house and treat older adults carry much of it as injury, hospitalization, liability and lost occupancy. Fall-risk assessment in these settings is usually episodic, at admission and quarterly, and relies on clinician impression or brief functional tests. Between assessments, gait can deteriorate for weeks without anyone measuring it.
OneStep measures gait from a standard smartphone carried in the pocket and turns the measurement into a fall-risk level tracked against each resident’s own baseline. This paper brings together outcomes from four deployments, in four different settings and organizations, to ask what happens to falls, rehospitalizations and function when residents’ walking is measured routinely, and what that is worth.
Three findings repeat across the deployments.
These are real-world, pre–post results from supervised deployments with internal comparison groups rather than randomized controls. They show a consistent, dose-dependent association between measurement and outcomes; they do not prove causation.
A skilled nursing and rehabilitation operator deployed OneStep across 29 facilities and left 5 facilities as controls. One year after deployment, we compared falls in each building with its own prior year and grouped buildings by how consistently they measured their residents.
The change in falls tracked the cadence. In the 13 facilities where most residents were measured every two weeks, falls fell by 19%. In the 7 measuring monthly, by 13%. In the 9 facilities where measurement lapsed, falls rose by 4%, almost exactly the 3% rise seen in the control facilities. Across all 29 buildings, including those where use lapsed, falls fell 8–9%.
Change in falls one year after deployment, by measurement cadence
Mean change per group versus each facility’s own prior year. 29 deploying facilities and 5 controls.
Hover or tap to see values. Bi-weekly: most residents measured every two weeks. Monthly: most measured once a month. Lapsed: at least half of residents not measured on a cadence. Control: no deployment.
Five facilities in the same organization deployed later. All five measured bi-weekly from the start, and all five reduced falls in their first year, by between 13% and 25%.
The later-deploying cohort, one year after deployment
Five facilities, all measuring every two weeks. Change in falls versus each facility’s prior year.
Hover or tap to see values. Facilities are shown in order of effect size; each is compared with its own prior year.
The gradient matters more than any single number. A dose–response across four groups, from bi-weekly through monthly to lapsed and control, is harder to explain by selection alone than a simple before-and-after difference. It also matches what we see in the data: decline before a fall typically emerges 10–14 days earlier as a change in gait speed or asymmetry. The organizations in this analysis have since set bi-weekly measurement as the standard for every building.
A family-owned contract therapy provider working across skilled nursing, long-term care, restorative nursing and memory care in 10 states implemented a three-tier model: measurement at admission, passive monitoring by restorative nursing assistants, and same-week reassessment and therapy referral when risk rose. Each facility was compared with its own prior year and grouped by how intensively it screened.
At facilities with higher screening, the share of residents who fell declined from 22.5% to 18.5%. At facilities with low screening it rose, from 16.2% to 18.0%. Relative to that trend, high-screening facilities had 26% fewer falls. At the same higher-screening facilities, rehospitalizations fell from 18.3% to 16.6% of residents, about 17 fewer rehospitalizations per 1,000 residents.
Falls and rehospitalizations by screening intensity
Percentage of residents, prior year versus current year. Each facility compared with its own prior year.
Hover or tap to see values. Falls: share of residents with at least one fall. Rehospitalization: share of residents readmitted to hospital, shown for the higher-screening facilities.
Rehospitalization is the outcome that most directly concerns risk-bearing organizations. Falls are a leading cause of transfer from post-acute care, so fewer falls should produce fewer transfers; the 9.3% reduction is consistent with that mechanism. Skilled nursing facilities already have 2% of Medicare Part A payments at stake under the SNF Value-Based Purchasing program, and at about $18,100 per 30-day readmission it is also the piece of cost avoidance that payers can see on their own ledgers.
Two deployments tracked functional change rather than falls alone. A large US senior-living organization, spanning independent living, assisted living and skilled nursing, introduced OneStep in mid-2025 as a shared platform between ownership and its in-house therapy provider. Over six months, 525 residents were assessed at least twice (4,397 assessments in all). 86.7% maintained or improved functional mobility, twice the improvement observed before OneStep. Among the 253 residents flagged as high or very high risk, an estimated 35 falls were avoided, a 15% reduction the organization describes as a minimum.
A six-month community program run by a social-welfare network enrolled about 500 older adults living at home. Members recorded walks, completed questionnaires and followed a home exercise program in the app; 84% of those with motion-sensor phones were monitored passively. Over six months, 86% maintained function and 9% improved; only 5% declined, compared with 20–25% typically declining per quarter in US senior living.
Functional change over the program
Share of participants who held or improved function, against OneStep’s observed benchmark for senior-living communities without the platform.
Hover or tap to see values. Community program: ~500 members over six months. Senior living: 525 residents over six months. Benchmark: 77.5% maintained or improved per quarter, OneStep’s observed average for senior-living communities without the platform.
Members who exercised completed an average of 24 full exercise days, about 8 per month, against 1–3 for typical digital programs and about 4 for in-person therapy. Improvement was dose-dependent here too: the most active quarter of members improved at ten times the rate of the least active.
Improvement rate by exercise quartile
Share of community-program members whose function improved, by quartile of exercise days. 46 members per quartile.
Hover or tap to see values. Quartiles are ordered from the fewest exercise days to the most.
We translated the observed changes into cost avoidance per 1,000 residents per year using published unit costs: $7,500 per fall as a conservative blended cost across falls with and without medical treatment, $14,000 per injurious fall, and $18,100 per Medicare 30-day readmission. The estimates are illustrative and deliberately conservative: they exclude emergency-department visits without admission, and they omit the indirect costs operators carry for incident response, documentation, liability, regulatory exposure and lost occupancy.
Modeled cost avoidance per 1,000 residents per year
Observed change in event rates multiplied by published unit costs.
| Source of avoidance | Basis | Events avoided | Unit cost | Avoidance |
|---|---|---|---|---|
| Falls, higher-screening facilities | 22.5% → 18.5% of residents falling | 40 | $7,500 | $300,000 |
| Falls, bi-weekly measurement | −19% applied to a 22.5% baseline | 43 | $7,500 | $320,000 |
| Falls, senior living | 35 avoided / 525 residents / 6 months | 133 | $7,500 | $1,000,000 |
| Falls, at injurious-fall cost | Higher-screening reduction at injurious-fall cost | 40 | $14,000 | $560,000 |
| Rehospitalizations | 18.3% → 16.6% of residents | 17 | $18,100 | $308,000 |
| Combined range | Falls (either skilled-nursing site) plus rehospitalizations, up to senior living plus rehospitalizations | $0.6–1.3M |
Illustrative; excludes staff time, liability, lost occupancy and quality-program payments. For the bi-weekly row, where falls were reported as a percentage change, the 22.5% baseline observed at the contract-therapy site was applied. The figures are consistent with OneStep’s standing model of $400,000–700,000 from a 25% fall reduction and $250,000–500,000 from a 10% readmission reduction per 1,000 monitored patients.
Value also shows up outside the model. In the senior-living deployment, automated documentation saved about 1,650 minutes per location, and remote therapeutic monitoring generated $4,206 in revenue in six months.
The reductions in the most consistently measured groups (19% and 17.8% against their own prior years; 26% relative to low-screening facilities) are close to the 23% reduction in fall rate that exercise programs achieve in randomized trials. That is notable because measurement is not itself an intervention. Its effect runs through what it triggers: earlier referral to therapy, reassessment, medication and environmental review, and sustained engagement with exercise. Three priorities follow.
Fewer falls means fewer injuries, fewer hospital transfers, and an objective record that the fall program works.
Decline is caught between screenings, residents return to skilled therapy before the fall, and documentation is grounded in objective data.
Each avoided fall and readmission is cost the ACO, plan or episode does not carry, and rehospitalization is a measure CMS and payers already track.
Every two weeks. In the skilled-nursing operator, facilities measuring bi-weekly cut falls by 19%, monthly by 13%, and facilities where measurement lapsed saw falls rise 4%, in step with controls. Decline before a fall typically emerges 10–14 days earlier as a change in gait speed or asymmetry, which a quarterly screen cannot see. The organizations in this analysis have since set bi-weekly measurement as the standard for every building.
No. Measurement works through what it triggers. When a resident’s risk rises, the care team is alerted and the resident is referred into or back to skilled therapy, reassessed the same week, or managed through medication review or environmental changes. Stable residents follow a digital wellness and home-exercise program. The reductions observed are similar in size to those exercise programs achieve in randomized trials.
At the contract-therapy facilities that screened most, the share of residents rehospitalized fell from 18.3% to 16.6%, a 9.3% relative reduction and about 17 fewer rehospitalizations per 1,000 residents. At roughly $18,100 per Medicare 30-day readmission, that is about $308,000 per 1,000 residents per year, and it is the component of cost avoidance that payers and CMS already track.
These are observational, pre–post analyses of operational data, so they cannot establish causation. Facilities were not randomized and intensity groups were self-selected: a building that measures every two weeks may differ in leadership and staffing from one where measurement lapsed. What the data do show is a consistent, dose-dependent association across four very different settings, which is what a causal effect would look like. A prospective controlled study is the next step.
This is a retrospective, multi-site analysis of operational outcome data from four organizations that deployed OneStep between 2024 and 2026. Each organization compared outcomes after deployment with its own prior period, and three of the four provided an internal comparison: facilities that did not deploy, facilities where measurement lapsed, or participants grouped by measurement or exercise intensity.
Deployments included in the analysis
| Setting | Scale | Comparison | Period | Outcomes |
|---|---|---|---|---|
| Skilled nursing and rehabilitation operator | 29 facilities + 5 controls | Prior year; controls; measurement cadence | 1 year | Falls per facility |
| Contract therapy across SNF, LTC, restorative and memory care | Network in 10 states | Prior year; screening intensity | 1 year | % residents falling; % rehospitalized |
| Senior living (IL, AL, SNF) | 525 residents; 4,397 assessments | Pre-deployment function; expected fall incidence | 6 months | Function; falls avoided; documentation time |
| Community-dwelling older adults | ~500 members; 374 onboarded | Senior-living benchmark; exercise quartiles | 6 months | Function; adherence |
Falls were taken from each organization’s incident reporting. Functional change was classified as improved, maintained or declined from the resident’s gait parameters and Walk Score over the program; the 77.5% benchmark is OneStep’s observed average for senior-living communities without the platform rather than an external norm. Secular trends and changes in fall reporting after deployment cannot be excluded, although the control facilities, which rose by 3%, argue against a general downward trend. The four sites differ in setting, population and outcome definition, so results are presented side by side rather than pooled.
The consistency and dose dependence of these findings justify a prospective, controlled evaluation: facilities within a single operator randomized to bi-weekly measurement or usual screening, with falls, injurious falls, hospital transfers and 30-day readmissions as pre-specified outcomes.
Suggested citation: OneStep Research Team. Longitudinal outcomes of OneStep in senior living and post-acute care. OneStep White Paper; October 2026. The full paper, including per-site methods, is available at onestep.co/resources.