Integrating STEADI for Falls Prevention in Outpatient Rehabilitation Clinics: An Outcomes Evaluation Using the RE-AIM Framework

Abstract

Background and Objectives

The RE-AIM (Reach, Effectiveness, Adoption, Implementation, and Maintenance) framework was used to describe the implementation of the Stopping Elderly Accidents, Deaths, and Injuries (STEADI) Initiative (2018–2021) for screening and assessing all older adults ≥65 years for falls risk across 34 outpatient rehabilitation clinics within a large health system.

Research Design and Methods

We described the Implementation process and strategies. Using Electronic Health Records (EHRs), we identified Reach, Adoption, and Maintenance of screenings and physical assessments to identify fall risk among older adults.

Results

STEADI Implementation strategies included health system mandates, EHR revisions, email instructions, educational sessions and resources, clinical leads and champions, and chart audits. Reach: 76.4% (50,023) had a completed screening, and 44.1% screened at risk for falls. Adoption: Clinic-level adoption varied, with most performing screenings. Profession-level adoption was highest for physical therapists (PTs; 94.2% initiated, 80.6% completed) and lowest for speech-language pathologists (SLPs; 79.8% initiated, 55.9% completed). Reach and Adoption of functional outcomes measures (FOM): PTs completed an FOM on 59.5% of at-risk patients, occupational therapists on 11.6%, and SLPs on 7.9%. Maintenance: All measures declined 1%–10% annually between 2018 and 2021.

Discussion and Implications

STEADI screening and FOMs were implemented systemwide in 34 outpatient rehabilitation clinics, reaching over 50,000 older adults. Screening adoption rates varied by clinic. PTs had the highest adoption rate. All adoption rates declined over time. Future research should consider an implementation science approach with input from key partners before implementation to identify barriers and develop strategies to support STEADI in outpatient rehabilitation.

Balance, Implementation science, Injury prevention

Falls are the leading cause of emergency department visits and injuries among older adults (≥65 years), resulting in significant morbidity and mortality (Bergen et al., 2016; Houry et al., 2016). Approximately one in four older adults fall annually, constituting $50 billion in direct medical costs and 60% of all nonfatal injuries among community-dwelling older adults (Florence et al., 2018). Many falls are preventable with evidence-based fall prevention, including screening, assessment, and targeted interventions to decrease risk (Grossman et al., 2018; Hopewell et al., 2020).

The American and British Geriatrics Societies (Panel on Prevention of Falls in Older Persons, American Geriatrics Society and British Geriatrics Society, 2011) and the World Falls Guidelines (Montero-Odasso et al., 2022) recommend that all health care providers screen older adults annually for falls risk and assess and intervene for those at risk. The Centers for Disease Control and Prevention (CDC) developed the Stopping Elderly Accidents, Deaths, and Injuries (STEADI) initiative to promote the uptake of falls prevention by physicians in primary care (Stevens & Phelan, 2013). The STEADI initiative has a plethora of resources for health care providers, older adults, and caregivers, which are not mandatory to use STEADI. For health care providers, it has a 1-hr full-training module, shorter training videos, case studies, posters, handouts for use of STEADI, education for patients, and evaluation and care coordination plans. For older adults and caregivers, it has educational brochures for different fall-risk factors, videos, and real-life scenarios. New content is constantly being added (CDC National Center for Injury Prevention and Control, 2023). Implementation studies of STEADI have been primarily conducted in primary care; however, clinic- and provider-level implementation barriers, including management of numerous comorbidities, reduced one-on-one time with patients, and lack of follow-up, limit uptake, and sustainability (Casey, 2017; Eckstrom et al., 2017; Stevens et al., 2017). Because of these barriers, the CDC called for research on STEADI to develop implementation best practices across different settings (Bergen et al., 2018; Centers for Disease Control and Prevention National Center for Injury Prevention and Control, 2022).

Small studies demonstrate that physical therapists (PTs) use STEADI in community-based falls screenings (Nithman & Vincenzo, 2019; Vincenzo & Patton, 2021). A survey of 425 PTs and PT assistants in the United States revealed that although only 31.1% of PTs were “very familiar” with STEADI, 84.1% (n = 111) of those use STEADI in clinical practice, primarily by choice, even though the majority did not have it embedded in their documentation/workflow (Vincenzo et al., 2022). PTs are qualified to provide falls prevention (Avin et al., 2015) and have more one-on-one time and frequent follow-up than primary care providers (PCPs; Dee & Littenberg, 2019), positioning PTs to incorporate STEADI for falls prevention of all older adults attending outpatient rehabilitation. The American Physical Therapy Association-Geriatrics clinical guidance statement asserts that PTs should screen all older adults for falls risk and provide assessments and interventions within their scope of practice for at-risk older adults (Avin et al., 2015). Yet, no studies to date have investigated PTs implementing STEADI in outpatient rehabilitation.

In December 2017, a large health system in the Pacific Northwest implemented STEADI in 34 outpatient rehabilitation clinics. Using the RE-AIM (Reach, Effectiveness, Adoption, Implementation, and Maintenance) framework (Glasgow et al., 1999), the purpose of this retrospective implementation evaluation was to describe the health system’s implementation process and outcomes of integrating STEADI to screen and assess all older adults for falls risk who attended outpatient rehabilitation between 2018 and 2021. The RE-AIM framework provides systematic guidance on planning and evaluating clinical or community-based projects to support the translation of scientific advances into practice (Glasgow et al., 2019; Holtrop et al., 2021). In addition, the framework is the highest cited framework in aging implementation and dissemination research as of 2023 (Sullivan et al., 2023) and has been used to describe implementation outcomes in studies integrating STEADI for falls prevention in primary care (Casey et al., 2017; Johnston et al., 2023). The findings from this investigation will inform outpatient rehabilitation clinics on the implementation of STEADI as a standard of care for older adults undergoing rehabilitation.

Method

Sample

De-identified Electronic Health Record (EHR) data were extracted from adults aged 18 years and older who were treated at an outpatient rehabilitation clinic implementing STEADI between January 2018 and December 2021. The University of Arkansas for Medical Sciences Institutional Review Board designated this project exempt (IRB #262879).

Extracted Data

Extracted data included sociodemographics, screenings, functional outcome measures (FOMs), clinician (PT, occupational therapist [OT], speech-language pathologist [SLP]), and clinic-level data. Ages ≥90 years were replaced with 90 for patient anonymity. Extracted FOM data were Timed up and Go (TUG), 30-second chair stand (30CS), Mini-BESTest, Berg Balance Scale, Functional Gait Assessment, Dynamic Gait Index, 6-minute walk, and 10-meter walk (Lusardi et al., 2017). Of note, patients self-reported their race and ethnicity when they are first registered for care in the health system, and that data carry over to all other care within the health system. This is often prior to them receiving rehabilitation services; therefore, the data on race and ethnicity were gathered from the EHR, which was originally obtained by self-report.

The STEADI Stay Independent Questionnaire (SIQ), a 12-item self-report questionnaire for screening fall risk, was collected. A score of ≥4 indicated an increased risk of falling (CDC National Center for Injury Prevention and Control, 2020). The three “Key Questions,” as described by the CDC, were used as an initial screening. The screening was complete if the patient responded “No” to all three questions. If a patient responded “Yes” to any of the three questions, the remaining nine questions were to be completed (Figure 1).

Figure 1.

Adapted Stay Independent Questionnaire.

If an older adult screened at risk for falls, STEADI indicates that they should be administered a balance/gait FOM, suggesting the TUG, 30CS, and/or 4-stage balance test (4SB; CDC National Center for Injury Prevention and Control, 2020). The health system also recommended that therapists administer the Mini-BESTest or any other balance-related assessment. Refer to Supplementary Table 1 for a detailed table of the health system’s STEADI workflow. In our data, a gait or balance FOM was defined as any of the following that were extracted (Lusardi et al., 2017). Test administration and psychometric properties are defined elsewhere (Shirley Ryan AbilityLab Website). Unfortunately, due to limitations in the EHR software and data extraction, we could not obtain data for the 4SB.

Data Reduction

Among 150,287 patients treated between 2018 and 2021, there were 221,248 episodes of care (EOCs). Because older adults are to be screened annually for falls risk, we used each year’s first clinic visit date as the start of an EOC. An EOC that spanned 2 years (start to end dates) was retained in the year that the EOC concluded. Using the first episode only and ≥65 years, the final working sample was 65,399.

Analysis

Measures representing STEADI outcomes were created using four of the five RE-AIM dimensions, as shown and operationalized in Table 1 (Holtrop et al., 2021). Data were analyzed using SAS 9.4 (SAS Institute Inc, 2022). Frequencies and percentages were used to describe categorical data, and means and standard deviations were used for continuous data.

Table 1.

RE-AIM Constructs Definitions and Study Operationalization

RE-AIM Domains	RE-AIM Definition	Study Operationalization
Initial Reach, 2018	The number and proportion of older adults who were screened and assessed for falls in 2018.	Percent of adults ≥65 with initiated screen Percent of adults ≥65 with completed screen Percent of adults ≥65 with screen score ≥4 with gait/balance measure
Maintenance of Reach, 2019–2021	The number and proportion of older adults who were screened and assessed for falls in 2019, 2020, and 2021.	Percent of adults ≥65 with initiated screen Percent of adults ≥65 with completed screen Percent of adults ≥65 with screen score ≥4 with gait/balance measure
Effectiveness	Not identified	Not identified
Initial Adoption, 2018	The number and proportion of clinics and PTs, OTs, and SLPs that screened and assessed older adults for fall risk in 2018.	Clinic-level adoption Range and percent of clinics with initiated screen Range and percent of clinics with completed screen
		Therapist-level adoption Percent of PTs, OTs, and SLPs who initiated screen Percent of PTs, OTs, and SLPs who completed screen Percent of PTs, OTs, and SLPs who completed a balance or gait outcomes measure on an at-risk older adult (screen ≥4)
Maintenance of Adoption, 2019–2021	The number and proportion of clinics and therapists that screened and assessed older adults for fall risk in 2019, 2020, and 2021.	Clinic-level adoption maintenance Range and percent of clinics with initiated screen Range and percent of clinics with completed screen
		Therapist-level adoption maintenance Percent of PTs, OTs, and SLPs who initiated screen Percent of PTs, OTs, and SLPs who completed screen Percent of PTs, OTs, and SLPs who completed a balance or gait functional outcomes measure on an at-risk older adult (screen >4)
Implementation (Process and Strategies)	Implementation process and implementation strategies the health system utilized to integrate the STEADI initiative into outpatient rehabilitation	Process and strategies involved in implementing STEADI initiative. Includes adaptations to STEADI
Maintenance	The extent to which STEADI initiative is sustained and becomes part of routine practices.	Changes in clinic-level and provider-level adoption in 2019, 2020, and 2021. Embedded under maintenance of reach and adoption

Notes: OT = occupational therapist; PT = physical therapist; SLP = speech and language pathologist; STEADI = Stopping Elderly Accidents, Deaths, and Injuries.

Results

Implementation Process and Implementation Strategies

The health system and the balance and falls clinical lead therapist implemented STEADI using several processes and strategies (Figure 2).

Figure 2.

STEADI Implementation and Maintenance Phase 1–3 (October 2016–2022). Implementation of STEADI was continuous from December 2017 through 2022 even when programmatic work to support implementation was paused.

Phase I pre-implementation

STEADI and the PT’s role in STEADI were introduced to a newly formed, voluntary group of PTs, known as fall champions, in October 2016, prior to the EHR build and clinic implementation of STEADI tools and workflows. Fall champions served the health system’s rehabilitation clinics as a site-specific resource for falls prevention and as a reference for the Balance and Falls Clinical Advancement Lead (CAP), a PT/staff member who guided fall prevention practice for all rehabilitation clinics across the state. Fall champions attended biannual meetings/trainings with the CAP and trained clinicians in their respective clinics on falls prevention. Tools introduced at the October 2016 inaugural fall champions meeting included:

• SIQ (already in use by Providence’s PCPs);

• new patient gait and balance intake form;

• EHR note template;

• patient education smart phrase (EHR tool that allows an established set of text to populate with words or phrases in the documentation);

• therapist education on evidence-based programs (Otago, Stepping On, Tai Chi);

• exercise dosage for falls prevention;

• goal of preventing falls versus addressing falls reactively.

Approximately 16 clinics were involved in the pre-implementation phase, during which the champions were educated about STEADI and educated the clinics and clinicians. However, go live/use of STEAI did not occur until December 2017.

Phase I implementation

Once implementation went live in December 2017, and the initial STEADI falls screening workflow started in the EHR in December 2017, all existing clinics (n = 30) implemented STEADI. In the following years, two newly opened clinics implemented STEADI in 2019 and two additional clinics in 2020, for a total of 34 clinics implementing STEADI. The clinics varied in size from having 2 to 42 therapists including permanent clinicians and those who worked between multiple clinics. The clinics also varied in their focus, from sport-specific and worker rehabilitation clinics to general orthopedics, neurological rehabilitation, or clinics that served a combination of patient populations. The percentage of older adults served across clinics varied from 7.8% in worker rehabilitation to 68.8% of the patient population in a rural neurological rehabilitation-focused clinic.

The STEADI “go live” was announced through email communication to clinical and front office staff. The email included (1) purpose of the change: to align with the Providence PCP falls screening tool and broaden rehabilitation fall-risk screening to be preventive; (2) who should receive the SIQ: ambulatory patients ≥65 years presenting to PT, OT, or SLP for an evaluation; (3) where and how to enter the SIQ into the EHR; and (4) the screening process (delineated above). See Supplementary Table 1 for workflow.

During the initial implementation phase (December 2017 to June 2018), STEADI workflows and EHR tools were developed and tested, with updates to correct errors and optimize utility. After the initial implementation, additional clinician training on the use of STEADI for developing further assessment; a PT plan of care; clinical mentorship; patient and clinician resources for multifactorial risk factors and assessment; development of training materials for new staff occurred. Fall champions also trained clinical staff during regular staff meetings in June 2018.

Phase 2 implementation—pre-pandemic maintenance

From July 2018 to December 2019, STEADI was expanded to new clinics and new hires; updates were completed to workflows within the EHR, billing, or insurance; interventions regarding home safety and referral to evidence-based fall prevention classes in the community (Tai Chi and Strong for Life) were added; and new clinical and patient resource documents were developed. These updates were delivered via quarterly newsletters and biannual meetings with the fall champions who then educated clinicians about the resources. Clinicians also taught free, public community fall-risk reduction classes incorporating STEADI’s framework to older adults and received feedback on their use of STEADI in the clinic via peer-reviewed quarterly chart audits or 1:1 supervisor: clinician quarterly meetings.

Phase 3 implementation—maintenance during COVID-19 pandemic

In January 2020, the focus shifted to optimization of interventions and assessment, with a view toward multidisciplinary care. OTs and SLPs were added to the fall champion group, and in February 2020, a multidisciplinary steering group met to develop and recommend resources for clinicians for patients with cognitive impairment and to select and integrate an appropriate FOM for patients at risk for falls into their plan of care. None of these objectives were achieved due to the start of the coronavirus disease 2019 (COVID-19) pandemic in March 2020.

From mid-March to May 2020, all clinics closed to in-person care due to the COVID-19 pandemic, seeing a limited population of patients (telehealth, post-op, and acute neurological injuries). From March 2020 to April 2021, implementation continued as usual but additional programmatic support for implementation paused. Three newsletter updates were shared by the CAP lead during this period. By April 2021, ~40% of the fall champion team and 20% of the steering committee had changed clinics, changed work status, or left the organization. Therefore, program time was dedicated to redeveloping the fall champion team and updating them on the current fall-risk workflows and resources. An internal survey of all rehabilitation clinicians was conducted in October 2021 to redefine the priority of needs for fall-risk screening, assessment, and intervention. Clinicians defined three priorities: (1) how and when SLPs and OTs should refer to PTs or PCPs when STEADI scores were positive, (2) optimizing inconsistencies in the FOM documentation in EHR, and (3) refreshing the workflow to support staff regarding when and to whom to provide SIQ. A falls steering committee meeting was held in October 2021, where implementation strategies were discussed for these priorities. Priority 3 was achieved immediately whereas the other two priorities were still in process at the end of data collection.

Sample

Demographics and screenings are in Table 2. Therapists initiated >65,000 outpatient therapy EOCs for older adults between 2018 and 2021. Most older adults had one EOC per year (84.9%) and were seen by a PT (88.1%). Most older adults self-identified as Caucasian (86.4%) and female (63.3%). Eighty-four percent were seen in an urban clinic (Am I Rural? Tool—Rural Health Information Hub, 2023). Between 2018 and 2021, 50,023 older adults had completed SIQ screenings, of which an average of 44.1% were at risk for falls.

Table 2.

STEADI Implementation: 2018–2021 Aged 65 and Older

Measure	4 Years Total N (%)	2018 N (%)	2019 N (%)	2020 N (%)	2021 N (%)
Demographics of all older adults seen who received therapya
Sex	N = 65,399	N = 17,232	N = 18,308	N = 14,164	N = 15,695
Female	41,424 (63.3)	11,0574 (64.2)	11,609 (63.4)	8,782 (62.0)	9,976 (63.6)
Male	23,975 (36.7)	6,175 (35.8)	6,699 (36.6)	5,382 (38.0)	5,719 (36.4)
Race/ethnicity	N = 65,375	N = 17,231	N = 18,304	N = 14,154	N = 15,686
Caucasian	56,500 (86.4)	14,892 (86.4)	15,780 (86.2)	12,311 (87.0)	13,517 (86.2)
Hispanic/Latino	1,655 (2.5)	413 (2.4)	467 (2.6)	348  (2.5)	427 (2.7)
Asian	2,133 (3.3)	593 (3.4)	595 (3.3)	416 (2.9)	529 (3.4)
African American	958 (1.5)	252 (1.5)	277 (1.5)	211 (1.5)	218 (1.4)
Other	1,271 (1.9)	325 (1.9)	379 (2.1)	272 (1.9)	295 (1.9)
Unknown	1,068 (1.6)	272 (1.6)	305 (1.7)	237 (1.7)	254 (1.6)
Refused	1,372 (2.1)	374 (2.2)	385 (2.1)	278 (2.0)	335 (2.1)
AIAN	278 (0.4)	78 (0.5)	74 (0.4)	52 (0.4)	74 (0.5)
NHPI	140 (0.2)	32 (0.2)	42 (0.2)	29 (0.2)	37 (0.2)
Older adults who had an initiated or completed STEADI screenings
Initiated screeningsb	59,071 (90.3)	16,094 (93.4)	16,467 (89.9)	12,534 (88.5)	13,716 (87.3)
Completed screeningsc,d	50,023 (76.4)	13,564 (78.3)	14,269 (77.9)	10,602 (74.9)	11,570 (73.7)
Caucasian	43,365 (76.8)	11,772 (79.1)	12,339 (78.2)	9,271 (75.3)	9,983 (73.9)
Hispanic/Latino	1,204 (72.8)	306 (74.1)	352 (75.4)	243 (69.8)	303 (71.0)
Asian	1,699 (79.7)	498 (84.0)	476 (80.0)	329 (79.1)	396 (74.9)
African American	677 (70.7)	182 (72.2)	202 (72.9)	146 (69.2)	147 (67.4)
Other	917 (72.2)	235 (72.3)	286 (75.5)	188 (69.1)	208 (70.5)
Unknown	814 (76.2)	212 (77.9)	224 (73.4)	178 (75.1)	200 (78.7)
AIAN	198 (71.2)	54 (69.2)	53 (71.6)	32 (61.5)	59 (79.7)
NHPI	111 (79.3)	23 (71.9)	37 (88.1)	21 (72.4)	30 (81.1)
At risk of falls*	22,041 (44.1)	6,012 (44.3)	6,258 (43.9)	4,550 (42.9)	5,221 (45.1)

Notes: Percentage based on valid screenings. STEADI completed screening score >3/14. AIAN = American Indian/Alaska Native; NHPI = Native Hawaiian or Pacific Islander.

^aThe first episode was determined by clinic visit date, per year, which allows patients to appear in the data set more than once if they visited multiple years since older adults are expected to be screened annually for falls. Episodes that spanned across 2 years (start to end dates) were retained in the year they were concluded.

^bA STEADI initiated screening was determined by completion of one or more of the first questions on the Stay Independent Questionnaire (SIQ; Q1, Q1A, Q1B).

^cA STEADI completed screening was designated by meeting one of the following two criteria:

-All three SIQ screener responses were complete and their sum was 0, with no missing values.

-All three SIQ screener responses were complete, their sum was greater than 0, and all 12 SIQ questions had a nonmissing response.

^dSTEADI completed screening by race and ethnicity were calculated using the number of older adults who self-identified as a specific race or ethnicity who had a completed screen divided by the total sample of older adults who attended therapy in that year who identified as a specific race or ethnicity (e.g., for NHPI 4-year total; 111 received a completed screening/140 who attended therapy in the 4 years = 79.3%).

Reach

Refer to Table 3 for results.

Table 3.

RE-AIM Constructs and Outcomes

RE-AIM Domains	RE-AIM Definition	Study Operationalization	Results N = total eligible older adults of variable of interest, n (%) = sample (percent) of eligible with variable of interest
Initial Reach	The number and proportion of older adults who were screened and assessed for falls in 2018	Percent of adults ≥65 with initiated screen	Older adults with initiated screen (N = 17,232)
			16,094 (93.4)
		Percent of adults ≥65 with completed screen	Older adults with completed screen (N = 17,232)
			13,564 (78.3)
		Percent of adults ≥65 with screen score ≥4 with gait/balance measure	Embedded under adoption
Maintenance of Reach	The number and proportion of older adults who were screened and assessed for falls in 2019, 2020, and 2021	Percent of adults ≥65 with initiated screen	Older adults with initiated screen
			2019 (N = 18,308)	2020 (N = 14,164)	2021 (N = 15,695)
			16,467 (89.9)	12,534 (88.5)	13,716 (87.3)
		Percent of adults ≥65 with completed screen	Older adults with completed screen
			14,269 (77.9)	10,602 (74.9)	11,570 (73.7)
		Percent of adults ≥65 with screen score ≥4 with gait/balance measure	Embedded under adoption
Effectiveness	Not identified	Not identified	Not identified
Initial Adoption	The number and proportion of clinics and PTs, OTs, and SLPs that screened and assessed older adults for fall risk in 2018	Clinic-level adoption (N varies by clinic)*
		Range and percent of clinics with initiated screen	Clinic initiated screen—Range
			n = 0–2,114 % = 0–98.8
		Range and percent of clinics with completed screen	Clinic completed screen—Range
			n = 0–1,659 % = 0–92.7
		Therapist-level adoption
		Percent of PTs, OTs, and SLPs who initiated screen	Therapist-level initiated screen
			PT (N = 15,108)	OT (N = 1,369)	SLP (N = 635)
			14,234 (94.2)	1,255 (91.7)	507 (79.8)
		Percent of PTs, OTs, and SLPs who completed screen	Therapist-level completed screen
			PT	OT	SLP
			12,174 (80.6)	955 (69.8)	355 (55.9)
		Percent of PTs, OTs, and SLPs who completed a balance or gait outcomes measure on an at-risk older adult (screen ≥4)	Therapist-level completed balance/gait measure
			PT (N = 5,568)	OT (N = 257)	SLP (N = 151)
			3,219 (59.5)	29 (11.6)	12 (7.9)
Maintenance of Adoption	The number and proportion of clinics and therapists that screened and assessed older adults for fall risk in 2019, 2020, and 2021	Clinic-level adoption maintenance (N varies by clinic)a
			Clinic initiated screen—Range
		Range and percent of clinics with initiated screen	2019	2020	2021
			n = 0–2,173 % = 0–99.1	n = 0–1,410 % = 0–100	n = 1–1,525 % = 6.7–97.8
		Range and percent of clinics with completed screen	2019	2020	2021
			n = 0–1,766 % = 0–94.4	n = 0–1,165 % = 0–96.6	n = 1,265 % = 0–94.1
		Therapist-level adoption maintenance
		Percent of PTs, OTs, and SLPs who initiated screen	Therapist-level initiated screen
			2019	2020	2021
			PT (N = 16,040) 14.784  (92.2)	PT (N = 12,327) 11,088 (89.9)	PT (N = 13,600) 12,129 (89.2)
			OT (N = 1,465) 1,307 (89.2)	OT (N = 1,152) 915 (79.4)	OT (N = 1,369) 1,075 (78.5)
			SLP (N = 648) 507 (78.2)	SLP (N = 492) 363 (73.8)	SLP (N = 607) 424 (69.9)
		Percent of PTs, OTs, and SLPs who completed screen	Therapist-level completed screen (same N in corresponding cells above)
			2019	2020	2021
			PT = 12,875 (80.3)	PT = 9,576 (77.7)	PT = 10,361 (76.2)
			OT = 943 (64.4)	OT = 654 (56.8)	OT = 829 (60.6)
			SLP = 334 (51.5)	SLP = 252 (51.2)	SLP = 302 (49.8)
		Percent of PTs, OTs, and SLPs who completed a balance or gait functional outcomes measure on an at-risk older adult (screen >4)	Therapist-level completed balance/gait measure
			2019	2020	2021
			PT (N = 5,785) 3,219 (59.5)	PT (N = 4,231) 2,298 (54.3)	PT (N = 4,786) 2,563 (53.6)
			OT (N = 257) 29 (11.6)	OT (N = 172) 19 (11)	OT (N = 257) 21 (8.2)
			SLP (N = 150) 2 (1.3)	SLP (N = 86) 2 (2.3)	SLP (N = 141) 5 (3.5)
Implementation (Process and Strategies)	Implementation process and implementation strategies the health system utilized to integrate the STEADI initiative into outpatient rehabilitation	– Process and strategies involved in implementing STEADI initiative. Includes adaptations to STEADI – Chosen and directed by the health system	Mandate  EHR revisions  Email with instructions  Educational sessions  Educational materials and resources  Clinical lead  Clinic champions  Chart audits  Newsletters
Maintenance	The extent to which STEADI initiative is sustained and becomes part of routine practices	Changes in clinic-level and provider-level adoption in 2019, 2020, 2021.	Embedded under maintenance of reach and maintenance of adoption

Notes: OT = occupational therapist; PT = physical therapist; SLP = speech and language pathologist. Percentages may not total 100 due to rounding. Balance/Function Functional outcomes measure = Timed up and go, 30-second chair stand, miniBest Test, Berg Balance Scale, Functional Gait Assessment, Dynamic Gait Index, 6-minute walk, and 10-meter walk.

^aSupplementary Table 3 provides detailed descriptive statistics by each clinic.

Reach First Year

In 2018, the first year after implementation, 93.4% (n = 16,094) of older adults had an initiated fall-risk screening and 78.3% (n = 13,564) had a completed screening, of which 44.3% were identified as at risk for falls and 56.2% (n = 3,354) of older adults who screened at risk for falls had a gait/balance FOM. Due to low sample sizes among self-reported non-Caucasian older adults attending rehabilitation, by frequency only, older self-reported American Indian/Alaska Natives who attended rehabilitation in 2018 had the lowest completed screenings (69.2%, n = 54/78), and self-reported Asians had the highest completed screenings (84%, n = 498/593).

Maintenance of Reach After First Year

Across all years, ~12%–15% more older adults had initiated screenings compared to completed screenings. Older adults who had an initiated screening decreased the most (3.5%) from 2018 to 2019 and ~1% per year afterward. Older adults who had a completed screening decreased the most (3%) from 2019 to 2020 and then decreased by ~1% in subsequent years. Reach of screenings by race and ethnicity was described using descriptive statistics. Due to the low sample sizes among self-reported non-Caucasian races and ethnicities, we were unable to analyze differences in screenings. Completed screenings were lowest for older self-reported American Indian/Alaska Natives who attended rehabilitation in 2019 (71.6%, n = 53/74) and 2020 (61.5%, n = 32/52) but lowest in 2021 for self-reported African Americans (67.4%, n = 147/218). Although self-reported American Indian/Alaska Native older adults had the lowest screening rates in 2019 and 2020, they had the second highest screening rate in 2021. These numbers should be interpreted with caution considering the small sample sizes among racial and ethnic minorities.

Between 2018 and 2021, 11,512 (52.8%) older adults who screened at risk for falls had completed a gait/balance FOM by any therapy professional. The greatest reach by all professions combined was achieved in 2018 (56.2%, n = 3,354). Reach steadily declined from 52.4% (n = 3,250) in 2019, 51.7% in 2020 (2,319), and 49.9% (n = 2,589) in 2021.

Adoption

Clinic-Level Adoption

Clinic-level adoption in the first year

Initiated fall-risk screenings across clinics the first year after implementation ranged from 0% to 98.8% (Supplementary Table 3). All but three of the clinics (27/30, 90%) that were open in 2018 had initiated screenings with over 80% of patients. Completed screening rates ranged from 0% to 93%, with the majority (24/30, 80%) of clinics completing screenings on >70% of older adults. All clinics were lower in completed screenings compared to their initiated screenings. Clinics with low numbers of older adult patients had low rates of initiated and completed screenings. One clinic had initiated fall-risk screening on 83% of older adults, and 0% had completed screenings (Clinic 1, Supplementary Table 3).

Maintenance of screenings at the clinic level

Maintenance of initiated and completed screenings varied by clinic. There did not appear to be trends in any year or across any years by individual clinic, although there was a trend in an overall decline in screenings by the end of 2021. By 2021, 29/34 (85%) clinics initiated screenings on >80% of older adults, and 22/34 (65%) clinics completed screenings for >70% of older adults.

Therapist-Level Adoption

Therapist-level adoption in the first year

Screening and FOM rates were identified at the therapy profession level (Supplementary Table 4). In 2018, PTs initiated (n = 14,234, 94.2%) and completed (n = 12,174, 80.6%) the highest rate of screenings, followed by OTs (n = 1,255, 91.7% initiated, and 69.8% completed) and SLPs (n = 507, 79.8% initiated, and 55.9% completed). PTs completed a gait/balance FOM on 59.5% of at-risk older adults, OTs completed them on 11.6%, and SLPs completed them on 7.9%.

Maintenance of screenings and FOMs at the therapist level after the first year

Initiated and completed screening rates declined ~1%–2% annually for PTs and ~1–5% for SLPs annually. OTs declined ~2% in initiated screenings and ~5% in completed screenings from 2018 to 2019 but showed a large decline of 10% for initiated and ~8% for completed screenings from 2019 to 2020. OTs were the only therapy profession who increased screening rates in any subsequent year, with completed screenings increasing ~4% from 2020 to 2021.

Across all therapy professions, baseline FOM completion rates were highest in 2018 and declined by 2021. PT’s baseline FOM completion was maintained in 2019 but declined by ~5% in 2020 and ~1% in 2021. OTs completed baseline FOMs at 11.6% in both 2018 and 2019, and rates decreased by <1% in 2020 and ~3% in 2021. SLP screening rates were lower in all years after 2018, completing baseline FOMs on two patients in 2019 (1.3%) and 2020 (2.3%) and five (3.5%) in 2021.

The specific FOMs completed were identified at the therapist profession level. The TUG had the highest completion of any gait/balance FOM across all professions (Supplementary Table 5) followed by the 30CST and were the only gait/balance FOMs completed by OTs and SLPs, which also align with STEADI recommendations (Centers for Disease Control and Prevention National Center for Injury Prevention and Control, 2020). The Mini-BESTest was the third highest administered FOM among PTs, which aligns with the health system’s adapted STEADI to administer any of the three recommended STEADI tests or the Mini-BESTest. PTs were the only therapists who completed other gait or balance FOMs besides those recommended within the STEADI initiative, and OTs completed non-gait/balance FOMs on ~8%–12% of at-risk older adults annually.

Discussion

Our study is the first to describe the implementation, reach, adoption, and maintenance of the STEADI initiative between 2018 and 2021 in outpatient rehabilitation clinics associated with a large health system. We found that reach and adoption rates of screenings and FOMs were highest in the first year after implementation and tended to decline over time. Initiated fall-risk screenings were greater than completed screenings by both clinic and therapist types. PTs had the highest screening and FOM completion rate followed by OTs and SLPs. The TUG and 30CST were the most frequently administered FOMs across all therapy professions.

Regarding implementation, in the implementation science literature, there is guidance on identifying barriers and facilitators to implementation and selecting and tailoring discrete strategies to support implementation (Bauer & Kirchner, 2020). The health system did not follow a particular implementation framework or process, but it utilized multiple processes and implementation strategies to support the uptake of the STEADI initiative, which had a positive impact. Prior studies support their chosen implementation strategies in increasing the adoption of falls prevention (Vandervelde et al., 2023). Specific to STEADI in primary care, EHR revisions to support workflow, audit and feedback, champions, and education improved uptake of STEADI (Casey, 2017; Eckstrom et al., 2017; Johnston et al., 2023; Stevens et al., 2017). Mandates are also successful to support the implementation of new initiatives (Powell et al., 2015; Sperber et al., 2019) but have not been investigated with STEADI implementation. Considering the varying rates of reach and adoption and waning rates over time, it would be beneficial to use an implementation science approach (Bauer & Kirchner, 2020) to collaborate with key partners by co-developing implementation strategies specific to identified barriers and facilitators for the implementation of the STEADI initiative to support uptake, maintenance, and fidelity (Wensing et al., 2020).

Reach of screenings was highest in the first year after implementation. Because only ambulatory older adults were to be screened, a screening rate of 100% would be unlikely due to some older adults being nonambulatory, which is beyond the scope of this study. STEADI screening rates within the first year after implementation in various primary care settings (Eckstrom et al., 2017) have been reported as low as 48% (Coe et al., 2017) and as high as 79% (Stevens et al., 2017). The percentage of older adults screened for falls in our study in the first year was comparable to the high end (~78%) of other studies. Although the percentage of older adults reached by screenings declined annually in our study, the lowest was ~74% of older adults screened in 2021. Only Stevens et al. (2017) reported maintenance of annual screenings after the first year of implementation, which declined in primary care from 79% to 49% (Stevens et al., 2017). They also found that the percent of older adults who screened at risk for falls and received a gait/balance FOM was ~52% in the first year, which declined by 10% in the second year, compared to Coe et al. (2017) who reported only ~30% of older adults completed an FOM the first year of implementation (Coe et al., 2017). FOM adoption and reach were higher in our study in the first year and subsequent years. Over 56% of at-risk older adults received a gait/balance FOM throughout the 4 years. These results suggest that outpatient rehabilitation may be a more favorable setting to implement STEADI compared to primary care, which may be due to therapists having longer visits with patients as well as more frequent visits than PCPs (Dee & Littenberg, 2019), but future studies in different health systems and outpatient rehabilitation clinics are warranted.

Considering demographics, the majority of older adults who attended outpatient rehabilitation during the time of implementation self-reported as Caucasian and female. Reach of screenings by self-reported race and ethnicity in any year ranged between a low of ~62% for American Indian/Alaska Native in 2020 and a high of ~88% for Native Hawaiian or Pacific Islanders in 2019. However, due to small sample sizes among self-reported non-Caucasian races and ethnicities, we were unable to analyze differences in screening rates by race and ethnicity. It is also possible that there could have been language and/or cultural barriers with administering the SIQ in different populations (Gustavson et al., 2022), which may impact reach. For example, two studies revised and adapted the SIQ for use with Thai (Loonlawong et al., 2019) and Portuguese (Monteiro et al., 2023) older adults, which required a multistep process for language and cultural translation and testing for reliability and validity.

An average of 44% of patients with a completed screening were at risk for falls. These fall-risk rates are two times higher than studies implementing STEADI in primary care, where between 18% and 22% of older adults screened at risk for falls based on SIQ scores (Eckstrom et al., 2017; Stevens et al., 2017). The higher rate of fall risk may be explained by the results of Moreland et al.’s (2018) National Health and Aging Trends Study of characteristics of older adults’ rehabilitation utilization (Moreland et al., 2018). They found that older adults who received fall-related rehabilitation compared to no rehabilitation had higher odds of falls, hip fractures, and mobility disability compared to older adults who did not receive rehabilitation. Additionally, even older adults who received nonfall-related rehabilitation had higher odds of falls, greater mobility disability, and more comorbidities compared to older adults who did not receive rehabilitation (Moreland et al., 2018). These data support that falls screening and management should be an integral part of outpatient rehabilitation regardless of if an older adult is attending for falls-related rehabilitation.

Regarding adoption, screening rates varied greatly by clinic, which is consistent with STEADI implementation data in primary care (Eckstrom et al., 2016; Stevens et al., 2017). Although the majority of clinics had high screening rates, some had very low screenings or high initiated yet low completed screenings. This indicates there may be a gap in therapists’ knowledge of what constitutes a completed SIQ fall-risk screening. Screening rates were also consistently higher for PTs compared to OTs and SLPs. Vincenzo and colleagues conducted a survey and found that PTs who reported that they were very familiar with STEADI also reported the highest use of STEADI (Vincenzo et al., 2022). Education may be beneficial to address gaps in screening fidelity considering that completed screening rates were lower than initiated screenings by both clinic and by therapist and varied greatly by clinic (Shaw et al., 2020). Future STEADI implementation efforts and studies should consider explicitly identifying and measuring fidelity to screening as delineated within their implementation process.

We also analyzed the adoption of gait/balance FOM for older adults who were screened at risk for falls at the therapist level. On average, PTs completed gait/balance FOMs on >54% of all at-risk older adults across implementation years, OTs completed FOMs <12%, and SLPs completed FOMs <8% of the time. The TUG and 30CS were the most frequently completed gait or balance FOMs for all therapy professions and the only gait/balance FOMs completed by OTs and SLP, which aligns with STEADI (CDC National Center for Injury Prevention and Control, 2020). The Mini-BESTest was the third highest administered FOM among PTs, which aligns with the health system’s adapted STEADI to administer any of the three recommended STEADI tests, the Mini-BESTest (which includes the TUG), or another gait/balance FOM. There may be multiple reasons that PTs administered approximately six times the rate of gait/balance FOMs compared to OTs and SLPs. At a high level, OTs focus more on activities of daily living and occupational and leisure participation, SLPs focus on speech, swallowing, and cognition, and PTs focus on gait, mobility, strength, and balance (Peranich et al., 2010). The FOMs in STEADI assess mobility, gait, and balance, which is most aligned with PT. Additionally, if an older adult is impaired in any of those areas, STEADI algorithm and the World Guidelines for Falls Prevention and Management recommend referral to physical therapy, further supporting that PTs specialize in balance, gait, and mobility (CDC National Center for Injury Prevention and Control, 2020; Montero-Odasso et al., 2022).

We would be remiss to not mention the potential effect of the pandemic on the implementation outcomes. Clinics closed for a period of time, staffing decreased, and workflow was reprioritized. Fewer older adults were seen in the clinic in 2020, although fall-risk rates were similar to prior years for those who were screened. There appeared to be no trends in reach or adoption by clinic or therapist before and after 2020. Also, although SLPs and OTs were added as fall champions in 2020, these additions did not appear to have a large impact on corresponding profession-level adoption rates of screenings or FOMs, which may be impacted by the timing of this initiative. Hence, there was not a clear impact of the pandemic on STEADI implementation outcomes, which may be because STEADI was already embedded in the workflow and part of usual patient care.

There are limitations to our study. Although our study is the first to identify implementation outcomes of STEADI in outpatient rehabilitation, we did not identify if implementing STEADI in outpatient rehabilitation impacts an older adult’s modifiable fall-risk factors (Montero-Odasso et al., 2022). It is important that future studies analyze the adoption of and fidelity to multifactorial fall-risk assessment and fall-risk interventions, which we were unable to ascertain in our retrospective postimplementation evaluation due to limitations in the EHR workflow and data extraction. These have been investigated, albeit in a few studies, in primary care implementation of STEADI. Eckstrom et al. (2017) found that 64% of patients who screened at risk for falls received STEADI interventions in their implementation study in a large primary care clinic (Eckstrom et al., 2017). Johnston et al. (2018) found that older adults screened and treated for falls risk who had a fall care plan were 0.6 times less likely to be hospitalized for a fall-related issue compared to those without a care plan (Johnston, 2018). Thus, appropriate interventions to address modifiable fall-risk factors are imperative to successful fall prevention (Montero-Odasso et al., 2022). Unfortunately, we were not able to capture interventions in our study due to limitations in the EHR, which is a limitation that needs to be addressed in future studies. Other limitations of our study include that it was only conducted in one health system, and there were limitations in the EHR data extraction capabilities which limited the available data and some of the analyses. It would be beneficial to conduct a full implementation study in the future to align EHR revisions with data extraction capabilities to maximize outcome analyses.

Conclusion

STEADI screening and gait/balance assessment were implemented in 34 outpatient rehabilitation clinics between December 2017 and December 2021 and reached over 50,000 older adults. Adoption rates for screenings varied by clinic but declined annually from the first to the fourth-year postimplementation. Adoption rates of screenings and FOMs were highest for PTs and lowest for SLPs across all years. Although all adoption rates declined over time, PTs continued to have the highest rates of screenings and gait/balance FOM completion in 2021. Future research should consider an implementation trial with input from key partners to select and tailor implementation strategies to address barriers to integrating STEADI into outpatient rehabilitation.

Funding

This project was supported by the National Institutes of Health (NIH, K76AG074920 to J. L. Vincenzo), the University of Arkansas for Medical Sciences (UAMS) Translational Research Institute, grants KL2 TR003108 (J. L. Vincenzo) and UL1 TR003107 (J. L. Vincenzo, G. M. Curran; UAMS), through the National Center for Advancing Translational Sciences of the National Institutes of Health. This project was also supported by the Center on Health Services Training and Research (CoHSTAR), funded by the Foundation for Physical Therapy Research in partnership with the American Physical Therapy Association and Clin-STAR Coordinating Center, Award Number U24AG065204 from the National Institute on Aging (to J.L. Vincenzo). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

Conflict of Interest

None.

Data Availability

The data are available upon request to the corresponding author and subject to institutional review board approval and appropriate agreement from the health system to share the data. This retrospective analysis of a clinical implementation in a health care system was not a clinical trial and was, therefore, not registered.

Acknowledgments

We would like to acknowledge all of the clinicians and staff at the health system who were involved in implementing STEADI. We would also like to acknowledge Erin Gloster for her assistance with proofreading and formatting.

Author Contributions

J. L. Vincenzo—conceptualization, formal analysis, investigation, methodology, writing—original draft. J. Caulley—conceptualization, methodology, writing—original draft. A. J. Scott—formal analysis, writing—original draft. B. S. Wilson—data curation, methodology, formal analysis. M. Wingood—formal analysis, methodology, writing—review and editing. G. M. Curran—methodology, formal analysis, writing—review and editing.