Implementing a Clinically Based Fall Prevention Program

Abstract

Introduction. Among people aged 65 and older, falls are the leading cause of both fatal and nonfatal injuries. The burden of falls is expected to increase as the US population ages. The Centers for Disease Control and Prevention (CDC) developed the STEADI (Stopping Elderly Accidents, Deaths, and Injuries) initiative to help primary care providers incorporate fall risk screening, assessment of patients’ modifiable risk factors, and implementation of evidence-based treatment strategies. Methods. In 2010, CDC funded the New York State Department of Health to implement STEADI in primary care sites in selected communities. The Medical Director of United Health Services championed integrating fall prevention into clinical practice and oversaw staff training. Components of STEADI were integrated into the health system’s electronic health record (EHR), and fall risk screening questions were added to the nursing staff’s patient intake forms. Results. In the first 12 months, 14 practices saw 10 702 patients aged 65 and older. Of these, 8457 patients (79.0%) were screened for fall risk and 1534 (18.1%) screened positive. About 52% of positive patients completed the Timed Up and Go gait and balance assessment. Screening declined to 49% in the second 12 months, with 21% of the patients screening positive. Conclusions. Fall prevention can be successfully integrated into primary care when it is supported by a clinical champion, coupled with timely staff training/retraining, incorporated into the EHR, and adapted to fit into the practice workflow.

Falls among older adults (those aged 65 years and older) are the leading cause of both fatal and nonfatal injuries. In 2014, more than 27 000 people aged 65 and older died from falls and an additional 2.8 million older adults were treated in hospital emergency departments for falls.¹ In addition, older adults who fall often experience limited mobility, reduced quality of life, and loss of independence. Many people become afraid of falling and limit their physical activity, leading to muscle weakness and increased fall risk.² The economic impact of falls is substantial, with direct medical costs in 2015 totaling $638 million for fatal and $31 billion for nonfatal falls.³

The burden of falls among older adults is expected to increase because the US population is aging. The number of people 65 and older is projected to rise from from 48 million (15%) in 2015 to 74 million (21%) in 2030.⁴ As a result, reducing falls will become increasingly important to health care providers who treat this vulnerable population.

Research shows that many falls are preventable. The Cochrane Collaboration conducted a meta-analysis of randomized controlled trials that focused on fall interventions. They concluded that, in the clinical setting, assessing and addressing an older person’s fall risk factors, in addition to identifying and treating symptoms of chronic conditions, can reduce falls.⁵ More recently, a systematic review by the US Preventive Services Task Force found that interventions delivered through primary care, including exercise or physical therapy, reduced older adult falls.⁶

Clinical Fall Prevention

An individualized clinical approach for reducing falls among older adults is recommended in the American and British Geriatrics Societies’ clinical practice guidelines.⁷ However, primary care physicians have been slow to put these guidelines into practice.⁸ Many report they do not know how to conduct fall risk assessments or do not have sufficient knowledge about fall prevention.^9,10 Providers also report that they have limited time to spend with each patient.¹¹ To address these implementation barriers, scientists at the Centers for Disease Control and Prevention’s (CDC) Injury Center developed the STEADI (Stopping Elderly Accidents, Deaths, and Injuries) initiative. STEADI, which has been described previously,¹² is based on Wagner’s Chronic Care Model¹³ and on research evidence, and it incorporates input from health care providers.

The resources in STEADI are designed for all members of the health care team. They are intended to help health care providers incorporate fall risk screening, assessment of patients’ modifiable risk factors, and implementation of evidence-based prevention strategies. These fall prevention strategies can be performed in a primary care setting or provided through referrals to specialists and community-based fall prevention programs. In this article, we describe the implementation of STEADI by a large New York State multispecialty group medical practice that included 29 primary care centers, and discuss the implementation process, reach, and lessons learned.

Selecting the First STEADI Site

In 2010, CDC funded the New York State Department of Health to implement STEADI in selected communities.¹⁴ The state health department chose Broome County, with a population of 200 600 located in New York State’s Southern Tier, because of their high number of reported older adult falls. The Broome County Health Department (BCHD) in turn approached the United Health Services (UHS) Medical Group and proposed that they act as the pilot implementation site for STEADI. UHS is a regional not-for-profit health care system serving the Greater Binghamton region in New York State. It includes more than 500 physicians and 60 locations, containing 29 primary care sites, throughout Broome, Chenango, Delaware, and Tioga Counties. The New York Department of Health also investigated local resources and developed connections to 3 community fall prevention programs: tai chi,¹⁵ Stepping On,¹⁶ and In Balance, a program developed by UHS.

CDC recommended enlisting a champion to effect change in clinical practice.¹⁷ Dr Frank Floyd, Associate Medical Director at UHS Medical Group and a practicing internal medicine physician, became the champion for STEADI. With his support and through the collaborative efforts of UHS Primary Care, the BCHD and the Broome County Office for Aging, STEADI was launched at the first site in 2012. This served as the pilot site before rolling out STEADI to other practices.

Integrating STEADI Into the Electronic Health Record and Clinic Workflow

Key steps in making fall prevention part of the usual practice for the pilot site were integrating the components of STEADI into the existing clinical workflow and modifying the electronic health record (EHR). These efforts were supported by a team of 2 physicians, 2 nurse practitioners, and 8 clinical office staff, including supervisors, who identified workflow changes to streamline the process and ensure that the clinical staff could incorporate the STEADI components into a regular office visit. The team identified those UHS staff members who would implement the STEADI work flow and worked with the internal information technology (IT) department to incorporate the STEADI components into the EHR. To help ensure that the new STEADI components would be seen and used, the fall risk screening questions were added to the nursing staff’s existing patient intake forms and automatically appeared as part of the recurring intake questions for all patients 65 and older.

STEADI Training

The STEADI training at all sites consisted of a group session open to all clinical staff and supervisors. Dr Floyd began by personalizing the issue—asking how many had an older adult family member or loved one who had died or been seriously injured in a fall. He then presented data on the high incidence of fall injuries nationally and locally. This was followed by a presentation by a nurse administrator on how to conduct the falls screening and the process for notifying patients found to be at increased risk. Finally, Dr Floyd and a representative of the health department provided information about the resources available for at-risk patients and how clinicians could access these resources. Most sites required 1 session although 1 site required 2 sessions. A representative from the UHS Nursing Administration followed up with the sites several times after the screening started to answer questions and review any issues that arose.

Screening by Nurses

Using the EHR module, nurses were instructed to screen all patients aged 65 and older for fall risk by asking whether the person (1) had experienced 2 or more falls in the past 12 months, (2) had 1 fall in the past 12 months with an injury, (3) had 1 fall in the past 12 months and gait or balance problems, (4) had any gait or balance problems, and/or (5) was presenting with an acute fall? If the patient screened positive (ie, answered yes to any screening question), he or she was considered at increased risk of falling.

Patients Who Screened Positive

For a patient who screened positive, a nurse or other clinical staff would evaluate the patient’s gait and balance using the Timed Up and Go (TUG) test.¹⁸ The TUG test begins by having the patient stand up from a chair, walk to a line 10 feet away, turn around, walk back to the chair, and sit down. A person aged 65 to 84 who took 12 or more seconds, or a person aged 85 or older who took 15 or more seconds, “failed” the TUG and was considered at high risk of falling. The TUG results were recorded in the EHR.

STEADI includes educational materials for older adults and their caregivers. For patients who screened positive, the nurse provided Check for Safety, a Home Safety Checklist, and What You Can Do to Prevent Falls, informational brochures (available at www.cdc.gov/steadi). The nurse also checked the boxes in the EHR indicating that the patient was given these materials.

Finally, for patients who screened positive, the nurse would place a paper Fall Prevention Referral Form on the computer keyboard for the physician to complete during the patient’s exam. This form listed a number of fall prevention strategies that the provider might propose, such as a medication review and management, vitamin D supplements, or referral to physical therapy, occupational therapy, or a community fall prevention program.

Physician’s Medical Assessment

The second phase of a STEADI evaluation was a medical assessment by a physician based on the patient’s screening results, TUG test score, and medical conditions. In the initial pilot site, Dr Floyd made it a practice to ask his patients who screened positive why they thought they might be at risk. He also observed how the patient walked. Based on the patient’s symptoms and his observations, Dr Floyd conducted a medical assessment and developed a plan of care that included falls interventions such as adjusting medications, recommending vitamin D, and making a referral to physical therapy or to a community fall prevention program.

On completing the patient’s exam and addressing their fall risk issues, the physician developed and documented a plan of care and checked off the recommended interventions on the paper Fall Prevention Referral Form. The physician also checked the “fall prevention referral completed” box on the EHR.

On the first day that the screening process was rolled out, a patient was identified as being at increased risk of falls although he had no noticeable issues. When I asked the patient, he said he was having a problem with his balance. I watched the patient walk and noticed that he shuffled a bit, did not swing his arms much. He also had bradykinesia (eg, decreased facial expression and blinking). Although he had no tremor, I realized this patient had very early onset Parkinson’s disease. I put him on medication and his balance improved, thereby reducing his fall risk. Through falls screening, I’ve now identified an estimated 15 to 20 patients with early Parkinson’s in my practice who are now benefiting from appropriate treatment.

(Dr Frank Floyd)

A care coordinator later scanned the referral forms and added them into the patient’s chart. The care coordinator also helped with scheduling appointments, supplied educational materials, and provided patient follow-up. Although the STEADI algorithm recommends that high-risk patients be followed up within 30 days, the care coordinators’ work load did not allow this activity to be sustained.

Implementation Roll Out

Approximately 7 months after implementing STEADI at the pilot site, Dr Floyd began to roll it out, initially at the 13 family practice and internal medicine sites in Broome County, and then at other primary care practices across UHS. Twelve sites were added between April and June 2013 and 6 additional sites from December 2013 to December 2014 (Figure 1). This process was simplified by already having incorporated STEADI into the EHR that was shared across sites.

Figure 1.

Roll out of STEADI older adult fall prevention at 19 primary care practices from August 2012 through December 2014.

Disseminating STEADI required training individual practices on a site-by-site basis. Dr Floyd and 2 lead nurses from the pilot site, often accompanied by a representative from the BCHD, went to each site, met with physicians and nurses at each practice, and discussed the rationale and mechanics of STEADI. While Dr Floyd focused on the physicians, the pilot site nurses worked directly with the nurses at the new site.

The majority of clinicians were receptive to falls screening but it helped that STEADI was being implemented successfully at my pilot site. That provided instant credibility. It also helped make the statistics about falls personal. At every site, I asked about their experiences with family members who had fallen.

Some clinicians balked about the extra work. At one site, a doctor simply refused to do any screening. In response, the site manager started a competition. The prize was lunch for the first physician and nurse to achieve a 70% screening rate. This internal competition was very successful and motivated the reluctant doctor to begin screening patients. This doctor later told me he was “blown away” by how well his patients liked tai chi!

(Dr Frank Floyd)

Dr Floyd began each training session by describing his personal experience with STEADI and demonstrating how he had streamlined the program to fit into his office workflow. The fact that he had successfully implemented it in his own practice gave the physicians confidence. He emphasized a team-based approach to involve everyone in the practice in reducing falls. This approach enabled Dr Floyd to obtain the cooperation of most of the primary care providers.

Reach of STEADI

Practices had varying levels of success in implementing STEADI. Table 1 compares the number and percentages of patients during the first two 12-month periods who were screened and, among those who screened positive, the number who completed the TUG test. Frequencies are based on output from the UHS EHR fall prevention module. Although implementation start dates differed, data for each practice is reported for 24 successive months. As of November 2015, 14 of the 19 participating practices had at least 24 months of data and were included in the table. In the first 12 months, these practices saw a total of 10 702 unique older patients; 8457 patients (79.0%) were screened for fall risk and 1534 (18.1%) screened positive.

Table 1.

Falls Screening and Assessments Among Patients Aged ≥65 Treated in 14 UHS Primary Care Practices^a.

	Patients Seen in the First 12 Months			Patients Seen in the Second 12 Months
	N	%	% Range	N	%	% Range
Total patients seen	10 702	100.0	—	11 678	100.0	—
Screening
Screened for fall risk	8457	79.0	50.6-97.4	5739	49.0	18.3-76.9
Screened positive for fall risk	1534	18.1	6.1-30.2	1227	21.4	14.3-46.8
Assessment among those who screened positive
TUG performed	791	51.6	7.1-84.8	514	41.9	4.8-70.3
Passed TUG	432	54.6	0.0-81.0	274	53.3	0.0-73.0
Failed TUGb	359	45.4	19.0-100.0	240	46.7	27.0-100.0
Unable to perform TUG	385	25.1	3.5-58.0	411	33.5	1.1-76.0
Missing TUG results	358	23.3	4.9-80.0	302	24.6	4.3-79.3

Abbreviations: UHS, United Health Services Medical Group; TUG, Timed Up and Go test; STEADI, Stopping Elderly Accidents, Deaths, and Injuries.

^aAlthough 19 practices implemented STEADI, only 14 practices had 24 months of data.

^bA person aged 65 to 84 who took 12 or more seconds or a person aged 85 or older who took 15 or more seconds failed the TUG.

Of the 1534 patients who screened positive, 791 (51.6%) completed a TUG test and 359 (45.4% of test takers) failed the TUG. However, staff members did not administer the TUG test to 385 patients (25.1%) whom the staff believed could not complete the TUG due to mobility problems. Finally, no TUG results were recorded for 358 (23.3%) of patients who failed the initial screen; we assume these patients never received the TUG and their TUG results are reported as missing.

Challenges With Roll Out

Certain STEADI components proved problematic during the roll out. Although gait and balance problems are well established fall risk factors, Dr Floyd suggested that not all providers were convinced that the TUG was useful. In addition, administering the TUG did not fit well into the clinical workflow due to space and time constraints. Clinicians also felt that routine checks for orthostatic blood pressure would cause unacceptable delays so these checks were limited to patients who reported dizziness or had undetermined causes of gait or balance difficulties.

Providers also had concerns about doing additional potentially uncompensated work. Dr Floyd demonstrated that for patients who screened positive, the additional evaluations could increase the level of their evaluation and management coding. The greater the complexity of the visit meant that a more highly reimbursed code could be used, as long as there was proper documentation.

Sustainability

Use of STEADI tended to decrease over time. As Table 1 shows, the proportion of older patients screened in the second 12-month period declined from 79.0% to 49.0%, while the proportion who screened positive increased slightly, from 18.1% to 21.4%.

Based on these observations, Dr Floyd and the implementation team recognized that sustaining the program required ongoing performance monitoring of each practice site and clinician. He found that sharing performance results with each site helped sustain interest and emphasized the importance of using STEADI to promote fall prevention. In the face of staff turnover and decreasing screening levels, new staff training and refresher training sessions were needed on an ongoing basis. The impact of these sessions can be inferred by comparing screening rates for the 19 practices over 3 years (Table 2). In year 1, the annual screening rates ranged from 50.6% to 97.4%. In year 2, annual screening rates declined among 16 of the 19 practices. However, rates improved for 12 of 14 practices that had data for a third year.

Table 2.

Falls Screening Rates Over 3 Years for 19 Practices.

Practice	Year 1		Year 2		Year 3
	Months of Data	% Screened	Months of Data	% Screened	Months of Data	% Screened
1	12	86.9	12	76.9	12.0	71.5
2	12	94.7	12	69.5	7.8	77.9
3	12	94.7	12	18.3	6.8	63.0
4	12	89.2	12	68.9	6.6	74.0
5	12	94.9	12	45.8	6.3	70.0
6	12	92.3	12	23.1	6.3	70.0
7	12	91.9	12	60.2	6.1	60.3
8	12	50.6	12	46.1	6.1	67.2
9	12	88.8	12	42.2	6.1	55.4
10	12	89.5	12	61.1	5.9	40.2
11	12	97.4	12	35.2	5.8	73.3
12	12	90.6	12	30.0	5.4	64.0
13	12	77.0	12	26.7	5.3	59.2
14	12	73.8	12	61.8	2.2	64.5
15	12	80.7	11.4	61.1	—	N/A
16	12	86.5	11.2	76.2	—	N/A
17	12	51.6	7.4	64.7	—	N/A
18	12	74.5	7.1	82.3	—	N/A
19	12	53.9	1.5	55.1	—	N/A

The team was able to incentivize participation through a pay-for-performance (P4P) program in partnership with major insurers. P4P is an umbrella term for initiatives aimed at improving the quality, efficiency, and overall value of health care. P4P provides financial incentives to hospitals, physicians, and other health care providers to carry out improvements and achieve optimal patient outcomes. In June of 2015—subsequent to the first 2 years of implementation at the original 14 sites—Excellus Blue Cross/Blue Shield incorporated fall screening rates into a P4P program. The most recent screening target was 72%. This program provided substantial reimbursement to practices that achieved the target screening rate and boosted administrative and clinical support for ongoing screenings.

Discussion

This article reports findings from the integration, implementation, and roll out of STEADI in a large New York health system. Prior to implementing STEADI, UHS had no routine fall risk screening or fall prevention activities. Beginning in 2013, 19 primary care practices initiated a clinical fall prevention program; as of November 2015, 14 sites had collected over 2 years of data. In the first 12 months, these 14 practices screened nearly 80% of their older adult patients for fall risk, although screening in the second year declined to about 50%. However, data from the first half of the third year indicated that, with refresher training sessions, screening rates were improving.

There were a number of critical lessons learned regarding implementation. Most clinicians readily saw the need for the program and expressed a genuine interest in improving their patients’ health and safety. Although they were receptive to adopting STEADI, many were concerned that it might not be practical to implement in their practice setting. Starting small was key. UHS began in a single facility led by a clinical champion and 2 champion nurses. This allowed the team to optimize the implementation process before expanding to additional sites. Furthermore, the clinicians’ limited time combined with the complex medical needs of older patients required streamlining the STEADI activities to fit into the clinical workflow. Integrating the fall prevention tools into existing sections of the EHR supported modifications to the clinical workflow, and a shared EHR across sites allowed for expanding beyond the initial site. In addition, a training manual would have helped standardize training across sites.

Even with a streamlined process and EHR integration, there were challenges. Patients with positive screening results did not always receive appropriate risk assessments, resulting in lost opportunities to implement fall prevention strategies. For example, high-risk patients, those who used an assistive device or appeared physically incapable of performing the TUG, did not consistently receive the TUG test, so gait and balance problems may have been overlooked. The EHR check-box indicating that the older adult was “unable to do TUG due to immobility” was checked for 25% of positive patients, suggesting a mobility issue, while 23% were completely missing a TUG result. Finally, the number of patient care managers was limited, so follow-up usually occurred when the patient returned, which could be months after the initial screening visit.

Last, while some STEADI components were entered directly into the UHS EHR fall prevention module, making them relatively easy to track and report, the Fall Prevention Referral Form was scanned into the patient’s chart. A review of these data would require manual chart reviews, which was beyond this scope of this study. A manual review of these data that includes details on the follow-up care received by patients who were identified as at increased risk of falling and their health outcomes is underway and will be described in a subsequent publication.

Conclusions

STEADI can be integrated into primary care settings when it is adapted to fit into the practice workflow and when training is customized for each practice. Regular retraining is needed to maintain the program and counteract personnel turnover. Clinical champions are helpful for promoting and maintaining implementation within individual settings and expanding and supporting implementation across multiple settings. However, improving case management and implementing strategies that promote and support patient adherence to evidence-based strategies is crucial to successfully reducing falls. Finally, improving and strengthening incentives can support STEADI implementation and substantially improve the health and well-being of older patients.