Abstract
Premature aging has been identified as a serious health problem among adults who were formerly homeless and now reside in supportive housing (Salem et al., 2013; Shibusawa & Padgett, 2009). Of the approximately 400,000 formerly homeless adults in the United States, two-thirds are ages 40–64 yr and have chronic mental and physical health problems similar to those in the general population age 70 or older (National Alliance to End Homelessness, 2016). Formerly homeless adults have been shown to have a mortality rate 3–4 times greater than the average adult in the United States; the average life expectancy of an adult with a history of homelessness is approximately 50 yr (Baggett et al., 2013; Nielsen, Hjorthøj, Erlangsen, & Nordentoft, 2011).
Premature aging likely results from years spent living on the street with few opportunities for sanitary bathing and toileting, poor health care and nutrition, and heightened risk of assault and victimization (Brown et al., 2016; Brown, Kiely, Bharel, & Mitchell, 2012). It is estimated that 50%–75% of homeless adults have a history of mental illness, posttraumatic stress disorder, or substance abuse (National Coalition for the Homeless, 2014). Such conditions are believed to contribute to premature aging, chronic health disorders, and cognitive problems even after these adults receive housing and consistent health care (Brown & Steinman, 2013; Sermons & Henry, 2010).
Once homeless adults attain supportive housing, they frequently receive regular health care services, including medical, dental, psychiatric, and case management (Corrigan, Pickett, Kraus, Burks, & Schmidt, 2015; Gilmer, Stefancic, Ettner, Manning, & Tsemberis, 2010). Although housing agencies are commonly responsible for ensuring that residences are in compliance with building safety codes, little information is available about how the health conditions associated with premature aging among formerly homeless people affect home safety and fall and accident risk in client residences. For example, cognitive problems may pose a safety risk when the person is using kitchen appliances and managing medication. Physical health problems may increase fall risk when the person is rising from the toilet or getting into and out of a tub. Disorganization leading to clutter or hoarding may heighten fall risk by reducing clear walking paths in the home environment. Because 40% of hospital admissions among older adults are caused by falls and accidents in the home (DeGrauw, Annest, Stevens, Xu, & Coronado, 2016; de Stampa et al., 2014), it is critical to help prematurely aging, formerly homeless adults stay safe and optimally manage their chronic illnesses in the home environment.
Formerly homeless adults who have prematurely aged as a result of homelessness may not be identified as aging because they have not reached age 65 yr. Although their physical and mental health problems are commonly treated by physicians and psychiatrists once they receive housing (Gilmer et al., 2010; Lebrun‐Harris et al., 2013), they may not receive needed home safety services because they have not been identified as geriatric patients. Physicians and psychiatrists may be unlikely to even consider how this population’s chronic health conditions affect their ability to function safely in the home environment. Primary care providers may also mistakenly believe that housing alone will enhance health and function.
The purpose of this study was to assess the home safety of 25 formerly homeless adults who reside in supportive housing and are age 40 or older. A 2-hr home safety evaluation was conducted for each participant in his or her home environment. To determine the potential risk for falls or accidents, the home safety evaluation was conducted to (1) examine home environment conditions and (2) assess participants’ performance of functional daily life activities—as they are affected by chronic health conditions—such as stepping in and out of the tub and safely using a microwave or stove. This knowledge will serve as the foundation for future services to be developed to enhance home safety in supportive housing for prematurely aging, formerly homeless adults.
Method
Research Design
This was an exploratory study (Stebbins, 2001) designed to identify and better understand the home safety concerns of a group of prematurely aging, formerly homeless adults so that services can be designed to enhance the maintenance of supportive housing. Quantitative and qualitative data were collected by means of a standardized home safety evaluation, the Safety Assessment of Function and the Environment for Rehabilitation (SAFER; Chiu, Oliver, Marshall, & Letts, 2001) and a set of supplemental questions. The Columbia University Medical Center institutional review board and the ethics committees of the two participating housing agencies approved this study, and all participants provided written consent.
Participants
All participants were residents of two housing agencies that provide supportive housing to homeless adults in New York City. Program directors of each housing agency identified possible participants meeting the inclusion criteria and inquired whether each resident was interested in study participation. Interested residents met with researchers to learn about the study, provide informed consent if they were interested in participating, and set an appointment for a home safety visit. Residents were included in the study if they were currently housed through a housing agency, had a history of homelessness, were age 40–64 yr, were their own legal guardians, and were English speaking. Residents were excluded from study participation if they had a severe behavioral or anger management disorder that made following instructions and cooperation difficult.
Instruments
Safety Assessment of Function and the Environment for Rehabilitation.
The SAFER is a standardized home safety assessment with 97 items that cover 14 domains (living situation, mobility, kitchen, fire hazards, eating, household, dressing, grooming, bathing, medication, communication, wandering, memory aids, and general issues; Chiu et al., 2001). The assessment was designed to identify home safety problems in older adults with physical or cognitive problems and requires approximately 45 min to 1 hr to administer. The assessment is scored using three possible choices: no problem identified, problem identified, and not applicable. Scores can be tallied for both the total assessment and for each domain. Scores are tallied by counting the number of problems identified and then comparing scores with the percentile norms of a reference group of 563 older adults with physical or cognitive problems (found in the SAFER manual). Percentile norms are measures that indicate the value below which a given percentage of group observations fall (Portney & Watkins, 2015).
Reliability was found to be high; internal consistency was found at an α coefficient of .83, p < .05 (Letts & Marshall, 1996); interrater reliability (r ) ranged from .80 to .90, p < .05 (Letts & Marshall, 1996; Letts, Scott, Burtney, Marshall, & McKean, 1998); and test–retest reliability (κ) ranged from .90 to .95, p < .05 (Letts et al., 1998). Content validity, determined by a panel of expert reviewers, was found to be high (Letts & Marshall, 1996). The two SAFER domains of eating and wandering were not assessed in this study because no participants had severe dementia (causing wandering) or dysphagia (eating difficulties).
Supplemental Questions.
All participants in this study were New York City apartment dwellers. Because the SAFER does not address many of the environmental factors found in apartment buildings or provide detailed information about several other factors involving home safety, an additional 177 supplemental questions were developed. The 177 questions covered 17 areas: building entrance, lobby, elevator, stairs, hallway from elevator or stairs to client’s apartment, entryway to apartment, main room, kitchen, bathroom, bedroom, functioning in apartment, medication management, emergency contact and phone use, cleaning, health complaints or concerns, sleep, and general home conditions. The supplemental questions specifically addressed the home environment of apartment dwellers rather than residents of single-family homes. The 177 supplemental questions were reviewed for face validity by two housing agency directors and one occupational therapist who had expertise in home safety.
Procedures
Each participant’s home safety evaluation required approximately 2 hr and was completed in the resident’s home by a research assistant. When requested by the participant, the resident’s case manager was present. The home safety evaluation consisted of administration of the SAFER (Chiu et al., 2001) and the set of 177 supplemental questions. All SAFER and supplemental questions were recorded electronically during the home safety visit.
Data Collection
The research assistants in this study were 11 occupational therapists (the coauthors) who received 5 hr of training in home safety evaluation using the SAFER and supplemental questions; training was provided by the first author. Before home safety visits commenced, the first author and 11 research assistants established interrater reliability with each other using both the SAFER and the supplemental questions with a housing agency resident who did not participate in the study. Interrater reliability was established at an intraclass correlation coefficient of .94, p < .001.
Data Analysis
Because this was an exploratory study, we used descriptive statistics (e.g., frequencies, means, standard deviations) to summarize the data derived from the SAFER and supplemental questions to identify the home safety concerns for this sample (Portney & Watkins, 2015). Narrative participant responses were summarized and categorized to identify trends in the data (Creswell, 2014).
Results
Twenty-five residents of supportive housing (17 men, 8 women) participated in this study. Fourteen (56%) participants resided in single, scattered-site apartments located throughout the New York City area. Eleven (44%) participants resided in housing agency–owned apartment buildings that were renovated for the purpose of housing homeless adults. Ages ranged from 48 to 64 yr (mean [M] = 57.64, standard deviation [SD] = 6.63). Participants’ race and ethnicity were as follows: African-American (n = 13; 52%), Hispanic (n = 7; 28%), and White (n = 5; 20%). Participants had completed high school (n = 8; 32%), some high school (n = 7; 28%), general equivalency diploma (n = 5; 20%), some college (n = 3; 12%), or college (n = 2; 8%).
The most commonly reported physical health problems (Table 1) were coronary artery disease (n = 15; 60%), arthritis (n = 13; 52%), and diabetes (n = 13; 52%). Mental health problems (see Table 1) included anxiety (n = 11; 44%), depression (n = 9; 36%), and schizophrenia (n = 6; 24%). Fifteen (60%) of the 25 participants used a mobility device, including straight canes (n = 9; 36%), wheeled walkers (n = 4; 16%), a manual wheelchair (n = 1; 4%), and a power wheelchair (n = 1; 4%).
Participant Physical and Mental Health Problems (N = 25)
SAFER Scores
One goal of our study was to understand whether our sample had experienced premature aging. Compared with a reference group of 563 older adult clients (data available in the SAFER manual; Chiu et al., 2001) with a mean age of 78 and a primary diagnosis of dementia (27%), stroke (8%), depression (5%), frailty (4%), or hip fracture (4%), 24 (96%) of our participants fell at or above the 95th percentile—meaning that they had more safety concerns than 95% of the SAFER reference group. Eleven participants (44%) fell between the 95th and 99th percentiles; 4 participants (16%) fell at the 99th percentile; and 7 participants (28%) fell above the 99th percentile. Only 1 (4%) participant fell below the 95th percentile (Table 2).
SAFER Scores and Percentiles
Note. SAFER = Safety Assessment of Function and the Environment for Rehabilitation.
Building Entrance
Despite the high number of participants who required mobility devices (n = 15; 60%), 18 (72%) lived in a building that had steps leading to the front door (M = 4.66 steps, SD = 4.77; range = 1–20). Sixteen (64%) lived in buildings that did not have a ramp for people having difficulty negotiating steps. Fourteen (56%) lived in residences with a front door that was difficult to open and would close quickly on residents, posing a fall risk, particularly for those with mobility devices. Although most of the front entrances were adequately lit, nine (36%) were not and caused difficulty for residents with low vision.
Lobbies and Elevators
The most common concern in building lobbies was the lack of nonskid flooring. Fourteen (56%) of the residences had a lobby with tile flooring that was slippery, chipped, and not level. No lobby had a large runway mat with rubber backing and beveled rubber edges to reduce fall risk. Twelve (48%) residences had elevators; residents in the two nonelevator buildings had significant mobility challenges (i.e., knee replacement, lower extremity arthritis, and severe asthma) that, without an elevator, impaired community access. Additional elevator concerns included insufficient space for a wheelchair (n = 3; 12%), insufficient lighting for people with low vision (n = 5; 20%), no handrails to assist residents while standing (n = 6; 24%), no contrast between floor buttons and the elevator panel (n = 3; 12%), and an elevator opening that was not level with the floor (n = 4; 16%).
Building Stairs and Hallways
Although most participants were not physically able to navigate building stairs because of health problems, 12 (48%) participants reported that they sometimes had to walk stairs or could not go in or out of the building because of elevator operating failure. Building stairs were commonly in poor condition, chipped, made of slippery material, and not level (n = 16; 64%). These conditions were made more hazardous in 12 (48%) residences in which stair lighting was inadequate. These same problems were often found in building hallways (n = 14; 56%). Twelve (48%) participants additionally reported that they had difficulty walking the length of the hallway because of health problems.
Apartment Entryway
Entryways to apartments were often cluttered with shopping and laundry carts, bikes, shoes, boxes, and chairs that hindered entrance (n = 12; 48%). Some entryway thresholds were greater than 1 in. (n = 5; 20%) posing a fall risk; doorways and entryways were not wide enough to accommodate a wheelchair (n = 13; 52%); flooring materials were slippery (n = 18; 72%); light switches were not accessible from a wheelchair position (n = 18; 72%); lighting was inadequate (n = 14; 56%); and front doors were difficult to open, stuck, or had broken locks (n = 6; 24%).
Main Room
In the majority of apartments (n = 14; 56%), the main room was severely cluttered without a clear walking path. Clutter consisted of too many furniture pieces and furniture that was too large for the space, boxes, laundry, bikes, electrical cords, and trash. Clutter was found on the floor, and wall-mounted storage systems were not present. Electrical cords posed safety hazards because they ran from one room to another, were situated directly within walking paths, were located near radiators, or were plugged into overloaded outlets (n = 8; 32%). Main room seating was commonly low, difficult to rise from and sit down on, and unstable and unsafe (n = 17; 68%). Ten participants (40%) reported that they could not use their mobility device in the main room because of clutter and used furniture to enhance their balance as they ambulated. The apartments of 12 participants (48%) were so cluttered that they could not access windows. Fall risk was heightened by inadequate lighting (n = 13; 52%) and apartment floors that were slippery, chipped, and not level (n = 13; 52%). In 14 (56%) apartments, main room light switches were not accessible from a wheelchair position.
Kitchen and Meal Preparation
Clutter was also problematic in a majority of kitchens and left little room to walk or prepare food (n = 18; 72%). Eight participants (32%) had difficulty standing at the sink and stove because of health concerns; however, none of the apartments had been modified for wheelchair use at the sink or had adequate floor space for a stool. Eleven (44%) participants could not access below-counter cabinets, and 8 (32%) could not access above-counter cabinets because of physical disabilities; yet, none had been provided with reachers or accessible, wall-mounted kitchen storage systems.
Stoves posed a heightened safety hazard for a majority of participants. One participant had to light burners with a match; for others, only some burners worked (n = 11; 44%). Participants reported that they used the stove to heat the apartment (n = 2; 8%) or as a nightlight (n = 2; 8%). In a majority of apartments (n = 16; 64%), burner controls were faded, worn away, and unreadable. Although most participants (n = 20; 80%) stated that they did not use a timer for meal preparation, 8 (32%) reported past grease fires and burned food, and 10 (40%) had potholders that were severely burned. Fifty percent of participants reported that they did not use potholders or oven mitts; instead, they stated that they used towels or old shirts, which could easily catch fire. Nine participants (36%) also had flammable items located too close to the stove (e.g., curtains, towels, potholders, food containers).
Sink faucet temperature control labeling was similarly worn away and unreadable in a majority of apartments (n = 19; 76%), and participants reported that they often scalded themselves when washing dishes. Many apartments (n = 18; 72%) did not have adequate light, particularly above food preparation areas, and light switches were sometimes inaccessible (e.g., behind the refrigerator). Fall risk was heightened by slippery floors in poor condition and old spills that had not been cleaned (n = 17; 68%). Seven (28%) participants had refrigerators with spoiled food, raising questions about their safety in meal preparation.
Bathroom, Bathing, and Toileting
Bathrooms posed the most urgent safety concerns. Despite the high number of participants (n = 15; 60%) with balance and lower extremity problems, only 6 bathrooms (24%) had grab bars installed in the shower or tub area and around the toilet. Participants reported that they often used the toilet paper holder, towel rack, and sink for assistance in the absence of grab bars. Although 18 participants (72%) had difficulty standing in the shower and bending to bathe their lower extremities, only 4 (16%) had a shower bench, and 3 (12%) had a handheld shower head.
Other safety concerns included slippery tile flooring (n = 23; 92%); absence of nonskid mats by the sink and inside and outside of the tub or shower (n = 20; 80%); lack of space to accommodate a wheelchair (n = 18; 72%); light switches that could not be accessed from a wheelchair position (n = 15; 60%); inadequate light in the bathroom, particularly in the tub or shower area (n = 16; 64%); temperature controls that could not be easily read or adjusted to prevent scalding (n = 17; 68%); sink faucets that could not be reached from a wheelchair position (n = 21; 84%); and above- and below-sink cabinets that could not be accessed because of physical disability (n = 13; 52%). Only 4 bathrooms (16%) had a toilet that was in compliance with the Americans With Disabilities Act of 1990 (Pub. L. 101-336); heights ranged from 17 to 19 in. All other toilet seats were 13–15 in. high. In 9 bathrooms (36%), the toilet seat was not securely fastened to the toilet, increasing fall risk.
Bedroom, Bed Area, and Dressing
Five participants resided in a studio apartment, in which the main room was also the bedroom area; 20 participants lived in a one-bedroom apartment. A primary concern in the bed area was clutter. Seventeen (68%) participants had clutter on and around the bed, which made it difficult for them to safely access the bed. Several participants had difficulty getting in and out of bed because of physical health problems but did not have a safety hand rail (n = 10; 40%). The majority of participants (n = 21; 84%) dressed while seated on the bed, and 16 (64%) had difficulty maintaining their balance while dressing their lower extremities. No client had a sock aid or long-handled shoe horn. Because of clutter, 19 participants (76%) could not access bedroom windows. In nine bedrooms (36%), windows were inoperable. Fall risk was also heightened by the presence of electrical cords located in walking paths (n = 11; 44%); slippery, chipped, and nonlevel floors (n = 12; 48%); and door thresholds that were higher than 1 in. (n = 5; 20%).
Medication Management and Emergency Contact Information
Ten (40%) participants either used a medication management system or received assistance from case managers. Fifteen (60%) reported that they relied on their memory; of these 15, 7 (47%) stated that they had difficulty remembering to take, refill, and store medications. All but 2 participants could identify appropriate emergency contact persons and numbers; however, 9 participants (36%) stated that they kept emergency phone numbers only in their cell phone, without a backup hard copy.
Cleaning
Most participants (n = 13; 52%) possessed appropriate and sufficient cleaning supplies. Cleaning and doing laundry, however, were identified as major sources of difficulty, and 15 participants (60%) reported being unable to engage in these activities when desired because of physical disabilities. No participants had been provided with adaptive cleaning equipment or energy conservation strategies.
Sleep
We asked participants about sleep habits because sleep can affect health and daily function (Auyeung et al., 2013; Garms-Homolovà, Flick, & Röhnsch, 2010). Twenty participants (80%) reported difficulty falling asleep, maintaining sleep throughout the night, and obtaining 7 to 8 hr of sleep per night. These participants also reported feeling tired the next day because of poor sleep and believed that poor sleep adversely affected their memory and daily function.
General Safety Information
Only 2 participants (8%) wore an emergency call button, and all but 3 apartments (12%) had round doorknobs instead of lever-style ones (round doorknobs can aggravate arthritis and are more difficult to operate). Fourteen participants (56%) smoked in their apartments, and none of the apartments had an automatic nightlight in each room to reduce fall risk.
Discussion
Although the 25 participants in this study had a mean age of 57.64, the amount and severity of their health concerns more closely resembled those of adults age 65 and older—a finding that supports the literature suggesting that homelessness highly correlates with premature aging (Salem et al., 2013; Sorrell, 2016). The participants’ SAFER scores also reflect their premature aging. That 24 of the 25 participants fell at or above the 95th percentile rank of the SAFER reference group—a reference group with a mean age of 78 and severe chronic illnesses—indicates that our sample had more home safety concerns as a result of illness than adults approximately 20 yr older.
The high number of home safety concerns affecting function that we found in residents’ apartments indicates that this group was not identified as needing geriatric health services to ensure home safety. Agency staff members—who were public health administrators, social workers, and case managers—may not have understood the correlation between homelessness and premature aging and may have inadvertently not considered the need for health services beyond primary medical care. Staff members may also have been unaware of services such as occupational therapy and the profession’s role in home safety and aging. Yet, all participants were under the care of physicians who could have referred participants for occupational therapy services but were likely unaware of the participants’ home safety risks.
Our study’s findings suggest that geriatric services such as occupational therapy have a significant role in supportive housing services for those who are formerly homeless. Moreover, occupational therapy and related geriatric professions must advocate for their role by making legislators, administrators, and housing staff members aware of both the link between homelessness and premature aging and the ways in which their services can decrease fall and accident risk in the home environment. Advocacy must extend to primary care physicians, who frequently provide the initial health care services received by formerly homeless adults. Occupational therapists are increasingly advocating for participation in primary care practice (Lamb & Metzler, 2014; Metzler, Hartmann, & Lowenthal, 2012; Muir, 2012), and the gap in health services for prematurely aging, formerly homeless people explicates the need for therapists to become members of the primary care team for this population.
Moreover, as Oakes (2015) suggested, occupational therapists need to help architects better design residences for the aging general population, including people who are prematurely aging and formerly homeless. One difference in safety risk that we found in our sample related to whether residences were located at scattered sites or in specific buildings designed to house formerly homeless adults. The common areas (e.g., building entrance, lobby, hallways, elevator) of the housing agency–owned apartment buildings, which housed 11 participants, had fewer safety issues than the scattered-site residences. The common areas of the agency-owned buildings were more adequately lit, less cluttered, and often better able to accommodate wheelchairs and mobility devices. The safety of these common areas implied that the building designer had some knowledge of general housing safety issues and used these when constructing or renovating the building.
Although general safety concerns may have been addressed in building common areas, the home safety concerns specific to an aging population had not been addressed in resident apartments. In these areas, only a health professional such as an occupational therapist would have knowledge of needed modifications, because of the practitioner’s understanding of how chronic illness affects residents’ safe functioning in the home environment. The safety concerns that were found in the resident apartments of the agency-owned buildings were equivalent to those found in scattered-site housing, likely because no aging specialist, such as an occupational therapist, had been consulted during either the apartments’ initial design or renovation or a specific resident’s occupancy.
Limitations
Limitations of this study included a small sample that cannot be generalized to the larger population of prematurely aging, formerly homeless adults. A further limitation is related to geographical region. Because this study examined supportive housing in New York City, the findings may not be generalizable to other geographical regions, particularly nonurban areas, because the cost of living and smaller apartment size in urban areas may be a causative factor in the high amount of clutter and hoarding that was observed. Additionally, all participants resided in an apartment, which may pose unique safety concerns that are different from those in a single-family house.
Future Research
Future research must consist of studies with larger sample sizes and studies comparing home safety risks across various geographical areas and residence types. Once it is clear that a large population of prematurely aging, formerly homeless people exhibit home safety concerns that place them at risk for falls and accidents in the home, future studies should examine how well occupational therapy home safety modifications are able to reduce such events as falls, accidents, 911 calls, and emergency department visits. Such information is critically needed to support the effectiveness of occupational therapy interventions and ensure occupational therapy’s role in the primary care of this population.
Implications for Occupational Therapy Practice
This study has the following implications for occupational therapy practice:
Homelessness prematurely ages people, and a subgroup of formerly homeless people who have prematurely aged are not receiving geriatric services because they are not identified as elderly.
Prematurely aging, formerly homeless people have significant physical, mental health, and cognitive problems that have an impact on their ability to function safely in their supportive living residences.
Occupational therapy practitioners must increase the awareness of their role with this population and become one of the primary care specialists who evaluate for and administer client services.
Conclusion
This study provides further evidence that chronic homelessness prematurely ages people and places them at risk for home safety events, including falls and injuries. Prematurely aged, formerly homeless adults are not commonly identified as needing geriatric services and, as a result, may encounter environmental home conditions that increase the risk of accidents, hospitalizations, and mortality. It is critical that the services of aging specialists, such as occupational therapy practitioners, be provided to help this population maintain and safely function in supported housing.
Footnotes
Acknowledgments
We thank Kim Baltich for her assistance in recruitment and data collection.