Impact of a Personal Response System on Hospital Use by 
Low-Income African American and White Elders 
Brenda F. McGadney-Douglass
Journal of Health Care for the Poor and Underserved, Volume 12, Number
4, November 2001, pp. 490-503 (Article)
Published by Johns Hopkins University Press
DOI: https://doi.org/10.1353/hpu.2010.0751
For additional information about this article
https://muse.jhu.edu/article/269806/summary
Access provided by University of Ghana (25 Feb 2019 09:57 GMT) 
490        Original paper
IMPACT OF A PERSONAL RESPONSE SYSTEM
ON HOSPITAL USE BY LOW-INCOME
AFRICAN AMERICAN AND WHITE ELDERS
BRENDA E McGADNEY-DOUGLASS, MSW, PhD
University of Ghana
Abstract: In 1994, the American Red Cross initiated a subsidized per-
sonal response system program for low-income elders in the Detroit area
to help these elders maintain their independence at home. Data were col-
lected on 379 low-income, at-risk elderly clients from Detroit and sur-
rounding Wayne County, Michigan. More than 80 percent of the sample
were African American (n = 313), and the remainder were white (n = 66)
elders. Statistically significant changes in number of emergency room vis-
its only occurred among white subscribers, increasing on average by 1.4
annual visits. One year after the installation of the personal response sys-
tem, white users were twice as likely as the African Americans to make an
emergency room visit or to stay overnight in the hospital. Subsidized per-
sonal response system may be an appropriate third-party government-
supported secondary interventionfor home-based low-income and at-risk
elders to prevent premature emergency room use, hospitalizations, and
institutional placements.
Key words: Elders, minority health care, personal response systems,
long-term care, length of stay, hospitalization.
The U.S. elderly population is growing rapidly. Older people (65 years and
older) are expected to make up 20 percent of the entire U.S. population by
year 2030.1 The most dramatic increases are projected to occur between the
years 2020 and 2050. By 2020, the proportion of elderly people who are among
the oldest old (80 years and older) is projected to be 17 percent. Those figures
will increase substantially once "baby boomers" start to become elderly in
2026.1 Minority populations are projected to represent 25 percent of the elderly
population in 2030 compared with 13 percent in 1990. Today, life expectancy is
76 years in the United States; it is expected to reach 83 years by 2050.1
The vast majority of today's elders are living healthier longer. However, a
growing number of older people are living longer with multiple chronic ill-
Received April 20,1999; revised August 18 2000 and June 2001; accepted June 30,2001.
Journal of Health Care for the Poor and Underserved    ·    Vol. 12, No. 4    ·    2001
McGadney-Douglass     491
nesses and disabilities.1 Their needs are putting increasing pressure on com-
munity-based social and health services nationwide. Although elderly Amer-
icans represented 12.6 percent of the population in 1987, Medicare accounted
for 42 percent of acute care hospital admissions.2,3 The average length of stay
(LOS) of elders 65 years and older, according to the annual National Hospital
Discharge Survey of the United States, is 7.2 days for males and 7.5 days for
females.4 The impact on hospitals serving increasing numbers of aging baby
boomers will be significant, with a greater proportion of Medicare expendi-
tures going to hospitals.5
One type of technological assistance to help elders live independently and
avoid premature and/or prolonged hospitalization is a personal response sys-
tem (PRS),3,6"8 a supportive device distributed widely in Western Europe and
North America. On average, subscribers to PRSs are charged an installation
fee of $25 to $30 (U.S. currency) and a monthly fee of $35 to $40. In 1995, the
four million PRS users in the United States and Canada were identified as
mostly female and elderly.9 According to Sherwood and Morris,10 the benefits
of PRSs for elderly users have included reduced anxiety about living alone
and increased ability to Uve independently. PRSs can help elders get immedi-
ate assistance when a serious home-based accident or injury occurs, whereas
delayed attention may result in an emergency room (ER) visit, hospital admis-
sion, prolonged hospital stay, premature institutionalization, or death.3,8
Despite these well-established benefits, the cost of PRSs put them out of
reach of most low-income elders, although this may be changing in some mea-
sure. Occasionally, private insurance companies pay for PRSs on a case-by-
case basis,10 and in some cases, Area Agencies on Aging have made funds
available for hospitals to purchase the equipment for indigent patients. Recent
reviews of support for formal third-party coverage of PRSs suggest a growing
trend to increase the inclusion of secondary services in the homes of frail
adults, particularly the elderly.
Michigan, like many states, has stepped up its efforts to encourage the
delivery of long-term care services in the home.11"15 The Michigan Waiver Pro-
gram for the elderly and physically disabled is jointly administered by the
Office of Services to the Aging (in the Department on Aging) and the Medicaid
program.16,17 The most commonly delivered services under the waiver pro-
gram are personal care, home-delivered meals, and PRSs. However, the rigor-
ous screening process in Michigan prohibits or delays many indigent or near-
indigent frail elders from qualifying for subsidized support for a PRS.
In 1988, responding to the health care needs of adults in their communities,
the American Red Cross initiated a Lifeline emergency response service in five
chapters across the country, including the Southeastern Michigan Chapter.
Lifeline is the trade name of a personal response system invented in 1974 by
Boston gerontologist Andrew Dibner. Lifeline served approximately four mil-
lion users in the United States and Canada in 1992.u"13
In 1994, responding to high demand and limited third-party coverage, the
American Red Cross developed the Low-Income Lifeline Program to target
492     Impact of a Personal Response System
elderly subscribers who are both poor and frail to help them maintain an inde-
pendent lifestyle at home.18 The chapter established collaborative relation-
ships and sponsorships with a number of churches, municipalities, health care
organizations, civic groups, and other nonprofit agencies willing to cover all
or some of the costs for needy clients. Other sources of funding included
grants and private donations. Prior to establishing outside funding, the chap-
ter was only able to provide 29 subsidized units, a number that increased
greatly after outside funding was secured.19
Each PRS subscriber has a Lifeline home communicator connected to his or
her telephone and a personal waterproof help button to wear as a bracelet or
neck chain. When the button is pressed, the communicator sends a help-
needed signal to the 24-hour monitoring center. Upon receiving this message,
a trained staff member uses the communicator box in an effort to speak to the
subscriber to find out what kind of assistance is needed. If the subscriber indi-
cates that he or she needs help or is unable to respond, the staff member con-
tacts the subscriber's designated responder (usually a relative or neighbor
who can come to his or her aid quickly). If the responder is not available, the
local police or fire department is called to provide assistance. The responder
must depress a flashing yellow reset bar on the communicator to let the moni-
toring center know that someone has come to the incapacitated person's aid.
Numerous investigators of PRS technology report positive outcomes of its
use, such as reductions in ER visits, reductions in the number of hospital
admissions, and shortened average LOS. These findings are especially strong
when PRS units have been placed in the homes of carefully screened individu-
als.5,12,20"22 Koch20 found that PRS users experienced a 26 percent reduction in
LOS. A 1985 replication of Koch's study conducted in four Boston hospitals
showed that Lifeline technology brought about a 23.2 percent reduction in
LOS and a 6.5 percent decrease in ER admissions.12,20 Cain's 1987 study
reported that Lifeline participation reduced LOS by 69.3 percent.21
Studying the effectiveness of PRSs, Roush and colleagues achieved results
similar to some of those of the present study.5,22 Analyses comparing hospital
utilization before and after subscription to the PRS by 106 subjects in Canada
(with a mean age of 78.6 years, 76.4 percent female) revealed no significant
change in ER visits (0.58 vs. 0.72, Ï• = 0.4) but a significant drop in the mean
number of days of hospitalization (14.37 vs. 5.85, Ï• = O.Ol).22 Roush and
Teasdale later replicated the Canadian study in the United States among community-
based PRS users. During the 1-year follow-up period, both American and
Canadian subscribers using the system experienced a statistically significant
(p = 0.05) decrease in LOS, with a mean reduction of 6.5 days. No significant
change in the number of ER visits occurred.5
These impact studies were for the most part conducted on white, middle-
class, or private pay subscribers, with none of the studies reviewed conducted
on African American, low-income, or subsidized users of PRSs. African
American elders tend to be in poorer health than white elders, and they and
their caregivers tend to have fewer fiscal resources than their white
McGadney-Douglass     493
counterparts,23"26 suggesting an increased need for PRSs, on one hand, and
fewer resources to pay for it, on the other. Because these groups have their own
needs and interests, programs and services often prove effective only when
tailored to fit. ^23,27 Furthermore, given the growing number of urban elders
who are poor but not Medicaid eligible, differences in hospital utilization rates
among PRS users must be studied in a more purposeful way.
Method
For the present study, the sample consisted of everyone (n = 379) who had
subscribed to the low-income PRS program of the Southeastern Michigan
Chapter of the American Red Cross for at least one full year between 1994 and
1997. A comparison group was subdivided along racial lines to include Afri-
can American (n = 313) and white (« = 66) elderly subsidized PRS users. To
ensure that clients identified to receive the Lifeline technology were both low
income and at risk for becoming incapacitated due to home-bound accidents
or premature institutionalization, the American Red Cross collaborated with a
number of agencies that work with the elderly to establish appropriate selec-
tion criteria, resulting in the following list: age, health status, hospital use,
social circumstances, ability to use the equipment, and financial status.
All of the 379 people enrolled since the program was launched have (or
had) incomes below, at, or minimally above the poverty level. The poverty
line, defined by the 1995 Department of Health and Human Services Poverty
Guidelines, is $7,470 for a single person and $10,030 for a couple. No client had
income greater than the threshold for Medicaid eligibility.
Selected clients were surveyed by American Red Cross staff and/or partner
agency volunteers using a brief four-page quantitative questionnaire. Inter-
views were conducted in the home and self-reported or reported by an infor-
mal caregiver. During the initial visit, agency staff asked the client or the des-
ignated informal caregiver a number of questions to complete the baseline
assessment form. If the client was accepted into the program, a follow-up
assessment was made every 6 months, usually by telephone. Questions were
asked to compare the effectiveness of the technology by demographic group,
health status, activities of daily living (ADL) stabilization patterns, and formal
health care utilization patterns. The latter two measures were collected at
baseline and 1 year after PRS installation.
Demographic measures included race, age, gender, marital status, monthly
income, and household expenses. Income included Social Security, Supple-
mental Security Income (SSI), disability payments, pension, family support,
and savings. Household composition, living situation, and geographical place
of residence were also assessed.
Multiple health problems were assessed by asking clients if they had been
diagnosed with any of nine diseases or debilitating illnesses. These illnesses
included cardiovascular disease (heart attack, stroke, hypertension, renal fail-
ure), a severe crippling or disabling disease (e.g., arthritis, Parkinson's
494     Impact of a Personal Response System
disease, multiple sclerosis), diabetes, severe pulmonary problems (e.g.,
asthma, emphysema), a physical handicap (e.g., paralysis, cerebral palsy),
cancer, blindness or significant loss of sight, deafness or significant hearing
loss, and obesity.
ADL measurements to track clients' abilities to perform certain functions
and to maintain an independent lifestyle were assessed by asking elders
about local transportation, meal preparation, taking medications, bathing/
showering, dressing and undressing, and getting in and out of bed. Clients
were asked, for example, if they could travel using buses, taxi, or their own car
and selected one of three possible responses: (1) unable to travel without spe-
cial arrangements (e.g., ambulance), (2) with some help (e.g., traveling with a
companion), and (3) without help (traveling alone).
Outcome measures included frequency of hospital admissions, total num-
ber of days hospitalized, and total number of ER visits for 1 year prior to the
installation of the PRS and for 1 year following installation of the PRS. Two-
tailed i-tests for independent samples were employed to measure and com-
pare between-group differences by race stratified by designated hospital utili-
zation variables. Two-tailed i-tests for paired samples were employed to test
for statistically significant changes within racial groups.
Results
Descriptive analyses revealed that, regardless of race, this group of sub-
scribers was very homogeneous with respect to several demographic vari-
ables (gender, age, income, and type of residence), health, and ADL stabiliza-
tion patterns. Of the PRS users studied here, 82.6 percent were African
American and 17.4 percent were white (see Table 1). As in other PRS studies,
the majority were female (82.4 percent, π = 330), very old (more than 25 percent
among the oldest old, ranging in age from 85 to 99 years), and widowed. Mari-
tal status was statistically significant by race (p < 0.05), with more African
Americans than whites being widowed. This finding is consistent with other
studies showing premature deaths of African American men due to health
problems associated with social, economic, and environmental conditions.23,26,28
There were no significant differences within the sample in monthly income,
although African American users were more likely than white users to have
less than $500 monthly income. There were significant differences between
the African Americans and whites studied in monthly expenses, with the Afri-
can Americans being more likely than the whites to spend less than $500 per
month (57.7 percent vs. 40.9 percent, Ï• < 0.01). Overall monthly expenses were
just slightly below the level of subscribers' monthly income, for all subscribers
studied.
Most of the PRS users lived alone in Detroit in single-family dwellings. In
keeping with eldercare literature on extended-family makeup among African
Americans,21,23,29"32 African Americans were significantly more likely than
whites to have two or more persons living in the household ( Ï• < 0.01). In
McGadney-Douglass     495
TABLEl
DEMOGRAPHICS FOR LOW-INCOME PERSONAL
RESPONSE SYSTEM SUBSCRIBERS (IN PERCENTAGES)
BLACKS WHITES SIGNIFICANCE
(n = 313) (n = 66) (t-TEST)
Gender
Male 13.1 86.9
Female 12.1 87.9
Age
60-64 5.2 3.2
65-69 11.5 11.3
70-74 14.1 19.4
75-79 21.0 16.1
80-84 22.6 22.6
85-99 25.6 27.4
Mean 78.2 78.3
Marital status 2.01*
Married 8.0 9.1
Widowed 74.1 62.1
Separated/divorced 16.0 27.0
Never married 1.9 1.5
Household 1.45*
Alone 83.1 89.4
Two or more 16.9 10.6
Residence -2.90*
Detroit 91.3 66.7
Outside of city 8.7 33.3
Housing
Single 48.9 50.0
Multiple 17.6 15.2
Apartment 9.3 10.6
Subsidized 24.3 24.2
Income (monthly)
<$500 46.5 33.3
$501-$700 20.5 21.2
$701-$1,000 27.9 31.8
> $1,000 5.1 13.6
Expenses (monthly) -2.74*
<$500 57.7 40.9
$501-$700 11.2 13.6
$701-$1,000 25.3 28.8
> $1,000 5.8 16.7
*p < 0.05. **p < 0.01. ***p < 0.001.
keeping with the current racial makeup of most large U.S. cities and migration
patterns to the suburbs, more than 91.3 percent of the African American PRS
users lived in the city of Detroit compared with 66.7 percent of white users (p <
0.001).33 Approximately half of the subscribers served by the program lived in
single-family houses, while one-quarter lived in subsidized "senior" housing.
496     Impact of a Personal Response System
This reflects the fact that multiple smgle-family dwellings were built to show-
case and meet the needs of a thriving and successful automobile industry in
Detroit's heyday as "Motor City"; Detroit boasts a greater proportion of single-
family homes than any other city in the country.
Most of the clients receiving Lifeline services reported having multiple
health problems (see Table 2). Consistent with other studies,34 heart disease
was the most prevalent for both groups, with 80 percent reporting cardiovas-
cular disease. More African American PRS users than white users reported
that they had a crippling disease (p < 0.05), were diabetic (p < 0.05), or were
blind (p < 0.001). In contrast, more whites than African Americans reported
that they suffered from pulmonary illnesses (p < 0.01) and obesity (p < 0.01).
There were no statistically significant differences between the races regarding
cardiovascular disease, physical handicap, cancer, or deafness. The higher
incidence of blindness among African American elders may be due to a higher
rate of diabetes and possible delay in receiving treatment. The higher inci-
dence of pulmonary disease among whites may be due to smoking. Findings
from comparative studies indicate that African Americans are more likely to
be in poorer health than whites.1,7,26,29,32
A comparison of elders' self-report of ADL at both baseline and a follow-up
assessment taken 1 year after installation of the PRS indicates favorable ADL
stabilization patterns for both African Americans and whites (see Table 3),
findings compatible with other studies.7,24,35,36 Mean scores indicate that while
these elders are frail, since installation of the service, there was no significant
decline in functional status or disability or level of independence. The vast
majority were able to complete ADLs (traveling, meal preparation, taking
medications, bathing/showering, dressing/undressing, getting in/out of
bed) with some help, indicating that most of these frail elders were able to
maintain an independent lifestyle at home. These findings are consistent with
recent studies of disability among the elderly showing that 1.2 million fewer
elderly Americans were disabled in 1994 than had been projected,1,37 trends
that may be associated with increased levels of exercise and improvements in
medicine, rehabilitation, and public health, as well as to the use of PRSs.
Within-group analyses of hospital use, including a pre- and a
postassessment of the PRS, measured by admissions, mean number of annual
ER visits, and LOS, revealed mixed results (see Table 4). Among African
American users, ER visits stayed constant at 1.2 visits yearly, but among white
users, there was a significant increase from 1 ER visit per year to 1.4 visits (p <
0.01). An increase in ER use is not consistent with most PRS studies.12,20 As
expected, there were statistically significant declines in admissions and LOS
for both groups: among African Americans, there was a 49 percent reduction
in reported hospital admissions (p < 0.001) and 4.5 fewer days hospitalized (p <
0.001). Similarly, among whites, there was a 37 percent reduction in admis-
sions (p < 0.001) and 3.6 fewer days hospitalized (p < 0.01). These declines are
consistent with other PRS studies.5,12,20
McGadney-Douglass     497
TABLE 2
SELF-REPORTED MORBIDITY FOR LOW-INCOME PERSONAL
RESPONSE SYSTEM SUBSCRIBERS (IN PERCENTAGES)
BLACKS WHITES SIGNIFICANCE
(τι = 313) (τι = 66) (t-TEST)
Cardiovascular 80.8 80.3 ns
Crippling diseases 44.4 31.8 -1.48»
Diabetes 28.8 22.7 1.04*
Pulmonary 20.4 30.3 -1.61**
Physical handicap 17.3 21.2 ns
Cancer 16.9 19.7 ns
Blindness 14.4 6.1 2.28***
Deafness 11.2 15.2 ns
Obesity 5.8 10.6 -1.20**
*p < 0.05. **p < 0.01. ***p < 0.001.
TABLE 3
ACTIVITIES OF DAILY LIVING (ADL): COMPARISON OF PRE- AND
POST-PERSONAL RESPONSE SYSTEM INSTALLATION (MEAN)
ONE YEAR
BASELINE AFTER SIGNIFICANCE
ADL ASSESSMENT INSTALLATION (t-TEST)
Blacks (n = 313)
Travel 1.89 1.90 ns
Meals 1.89 1.94 ns
Medications 2.29 2.14 ns
Bath/shower 2.00 2.23 ns
Dress/undress 2.22 2.13 ns
Get in/out of bed 2.32 2.17 ns
Whites (n = 66)
Travel 1.89 1.93 ns
Meals 1.95 1.97 ns
Medications 2.33 2.19 ns
Bath/shower 2.01 2.25 ns
Dress/undress 2.30 2.17 ns
Get in/out of bed 2.32 2.23 ns
Note: Mean values: 1 = unable to do without special arrangements; 2 = can do with some help; and
3 = can do without help.
Between-group differences exist on all hospital use variables measured 1
year after PRS installation (see Table 5). White users were twice as likely to
make visits to the ER (p < 0.001), be admitted to the hospital (p < 0.001), and
stay in the hospital longer (p < 0.001).
498     Impact of a Personal Response System
TABLE 4
USE OF HOSPITAL SERVICE: WITHIN-GROUP PRE-
AND POST-PERSONAL RESPONSE SYSTEM
INSTALLATION COMPARISONS (MEAN)
ONEYEAR            ONEYEAR           t-VALUEFOR
BEFORE                 AFTER                 CHANGE
INSTALLAnON      INSTALLATION          (t-TESTS)*
Emergency room visits (η)
Blacks 1.20 1.23 ns
Whites 1.02 2.41 -3.06**
Admissions (η)
Blacks 0.72 0.23 14.35***
Whites 0.69 0.37 3.92***
Overnight stays (n)
Blacks 6.02 1.24 8.70***
Whites 6.04 2.42 3.06**
*p < 0.05. **p < 0.01. ***p < 0.001.
TABLE 5
USE OF HOSPITAL SERVICE: BETWEEN-GROUP COMPARISONS
(MEAN)
ONE YEAR AFTER INSTAUATION
EMERGENCY STAYS
ROOM ADMISSIONS (NUMBER
VISHS (τι) (YES) OFDAYS)
Blacks 1.23 0.23 1.24
Whites 2.41 0.37 2.42
Significance: f-test -1.2* -0.13* -1.2*
*p < 0.001.
Discussion
This study shows that the greatest benefits to subsidized PRS users was a
reduction in hospital use—specifically, a reduction in admissions and in the
number of inpatient days. Emergency room visits by African Americans,
which averaged one per year, remained constant while those by whites signifi-
cantly increased from one to almost two and a half annually.
A variety factors may explain the increase in ER use by white elders, includ-
ing race, gender, fear of asking for help, lack of social support, or lack of alter-
natives to emergency rooms for urgent care for the poor in southeastern Mich-
igan. It is not known how patients in this sample were transported to the ER
McGadney-Douglass     499
(ambulance or private transport) or how they paid for their visits. White
elders in the metropolitan Detroit PRS sample had somewhat more disposable
income, and this may have contributed to their making more ER visits. Alter-
natively, fear and isolation among some white elders may be associated with
greater rates of ER use. Frail white elders, often widows, living alone in single-
family homes may be part of a small minority of white residents in predomi-
nantly African American neighborhoods.33 Some research suggests that ethnic
and racial characteristics of neighborhood environments may influence the
social ties of residents.38"40 For example, Thompson and Krause41 suggest that
anticipated social support is lower among elders who live in deteriorated
neighborhoods than among older adults who Uve in well-maintained neigh-
borhoods, particularly among alder adults who live alone. Other research
suggests that white elders may be less likely to identify a neighbor as a PRS
respondent because they fear becoming victimized by crime or because they
do not know any neighbor well enough to ask.42"44
The availability of family caregivers may be another key factor associated
with increased number of visits to the ER. Specifically, white elders might
have named as their respondent a family member who lives outside the city
more frequently than African American elders did. Such a respondent is more
likely to be delayed than someone who lives nearby and, furthermore, may be
put off by urban conditions such as limited street parking, physical quality of
structures, abandoned buildings, and fear of crime and, for one of these rea-
sons, fail to respond promptly to calls for help from the elder, thus necessitat-
ing a visit to the ER.44"47
While family and fictive kin (nonblood relatives such as neighbors) may
have been able to attend to the needs of African American elders earlier than
whites, another reason African American elders may have visited the ER less
frequently is that a fear of not returning home caused African American elders
and their families to avoid hospital visits whenever possible. Research by Gur-
ley et al.18 has shown high rates of hospitalization with no return home among
this population.
Although the findings of this study are intriguing, several limitations sug-
gest that they should be interpreted with caution. Limiting the study to a sin-
gle location minimizes generalizability of these results to comparable popula-
tions of urban, elderly PRS users. This may be particularly true because the
location chosen was the Detroit metropolitan area, which has significant de
facto racial segregation in housing and residence patterns. These concerns can
be addressed by focusing future research on hospital use over time by a ran-
domly selected matched sample of private pay and subsidized elders from a
highly integrated urban community. Such a design might allow one to identify
other sources of variation that might have predicted observed changes on the
dependent variables (e.g., admissions, number of days hospitalized, ER vis-
its). Other sources of variance that might have predicted changes in outcomes
include such confounding variables as demographics (age, race, monthly
income, and expenses), living arrangements, insurance status, health, frailty,
500     Impact of a Personal Response System
social support (primary and/or secondary caregivers, distance from elder,
home health care), and level of isolation based on a sociability leisure scale (lei-
sure activities such as TV watching, reading, crafts, telephone use, attendance
at religious services, outside visits, and shopping). The question to be
answered is, "What contributes to differences in admissions, ER visits, and
LOS for African American and white elderly users of PRSs?"
The strength of the present study lies in the fact that the sample includes
comparable groups of previously understudied, urban, low-income African
American and white elderly PRS users. The African American and white sub-
groups are almost equally matched on demographics, health, functional sta-
tus, and hospital use at baseline. Also, the analyses consisted of intergroup
and intragroup comparisons. Intragroup comparisons are particularly impor-
tant because there is substantial diversity within racial groups. The heteroge-
neity with regard to hospital use needs to be recognized and tested further. For
both groups, although elders are frail, ADL measures indicate that there was
no significant decline in functional status, indicating that these elders
remained stable and independent at home after the installation of the PRS.
While elders and their caregivers will simply celebrate this finding, formal
health care providers and policy makers should take note of them as an indica-
tion that a secondary intervention such as a PRS is often successful in helping
to maintain the independence of elders at home.
Economic constraint and denial of need are obstacles to PRS use. All elders
are susceptible to becoming incapacitated in their homes, particularly the old-
est old, who are more likely to have high rates of incapacitation, leading to pre-
mature institutionalization or death.18 The rate of such incidents increases dra-
matically with age. In 1996, people turning 65 could expect 12.2 additional
years of full function and 5.3 years of dysfunction characterized by acute or
chronic illness.1 Mittlemark and his colleagues reported that among elders
with no ADL limitations, 34 percent of African Americans and 26 percent of
whites reported having unmet needs for one or more formal community or
home services.48 Thus, a comprehensive model for long-term care should con-
sider the distribution of PRSs as an effective health promotion and primary
prevention strategy for all elders.
Personal response system use would be more widespread if it were subsi-
dized. The current method of payment for long-term care services for all
elders regardless of health status is abarrier to primary preventive, restorative
care, and health maintenance. Publicly funded health care programs rarely
pay for personal response systems. Medicare does not cover them at all, and
the criteria for coverage under the Medicaid Waiver Program by some states
are often extremely restricted,14 promoting tertiary intervention at a time
when frail elders are most vulnerable for costly hospitalization,
institutionalization, or death.
Given the extraordinary cost of hospital and emergency care, it is extremely
important that the use of PRSs is associated with elders' ability to live inde-
pendently. Because Medicare is primarily an acute care system, it has not
McGadney-Douglass     501
historically supported preventive geriatric care. Personal response systems
are a long-term care service and should be part of any comprehensive long-
term care effort. Until they are, third-party coverage under Medicaid is vital
for elders living near, at, or below the poverty level.
In conclusion, the growth and acceptance of PRSs are being fueled both by
changes in the population and the need to reduce health care costs. Because
home care is preferred by individuals, personal response systems can benefit
elders by helping them maintain their sense of security and independence in
the community.49 At the macro level, PRSs may contribute to reduced acute-
care hospitalization and attendant costs.50
Acknowledgments
The author wishes to acknowledge the support of her late colleague and friend Lucille H.
Davis, PhD, RN, FAAN, who read earlier drafts of this manuscript; Fay Flowers, RN, Director,
Department of Health and Community Outreach, Southeastern Michigan, Chapter of the Ameri-
can Red Cross, 100 Mack Avenue, Detroit, MI; and Patricia Penyman, Lifeline Program Manager
at the Southeastern Michigan Chapter of the American Red Cross.
REFERENCES
1. American Association for World Health (AAWH). World health day 1999: Healthy aging,
healthy living—start now! Washington, DC: U.S. Department of Health and Human Services /
Centers for Disease Control and Prevention, Pan American Health Organization, World
Health Organization, 1999.
2. National Center for Health Statistics. Health data on older Americans: National discharge sur-
vey: Series 3, No. 27. Hyattsville, MD: National Center for Health Statistics, 1992. (DHHS Pub.
No. [NCHS] 92-4001,2)
3. U.S. Department of Health and Human Services, Public Health Service. Vital and health statis-
tics: health data on older Americans. Washington, DC: Government Printing Office, 1993.
(DHHS Pub. No. [PHS] 93-1411,3)
4.  Graves EJ, Gilum BS. 1994 summary: National hospital discharge survey. Advance Data
1996;178(1):1-12.
5. Roush RE, Teasdale TA. Reduce hospitalization rates of two sets of community-residing older
adults after use of a personal response system. J Applied Gerontol 1997;16(3):355-66.
6. Waldo DR, Sonnefeld ST, McKusick DR, et al. Health expenditures by age group, 1977 and
1987. Health Care Finance Rev 1989;10(2):137.
7. Jackson SA, Mittelmark MB. Unmet needs for formal home and community services among
African American and white older adults: The Forsyth County Aging Study. J Applied
Gerontol 1997;16(3):298-316.
8. McGadney BF. Southeast Michigan chapter of the American Red Cross: Emergency response
program for low-income African American elders. Paper presented at the annual meeting of
the Michigan Academy, Grand Rapids, ML Mar 12,1999.
9.  Fisk MJ. A comparison of personal response systems in Canada and the U.K. J Telemed
Telecare 1995;l(3):145-56.
10. Sherwood S, Morris JN. A study of the effects of an emergency alarm and response system for
the aged: A final report of grant no. HS01788. Springfield, VA: National Center for Health Ser-
vices Research, 1980. (NTIS No. PB 82-137242)
11. Montgomery C. Personal response systems in the U.S. Home Health Care Serv Q 1992;13(3-
4):201-22.
502     Impact of a Personal Response System
12. Dibner AS. Personal response systems: An international report of a new home care service.
Binghamton, NY: Haworth, 1992.
13. Maddox GL. Foreword. Personal response systems: An international report of a new home
care service. Home Health Care Serv Q 1992;13(3-4):11.
14. Benson WF. Public financing for personal response systems: a federal viewpoint. Home Health
Serv Q 1992;13(34):223-28.
15. Lipson DJ, Birnbaum M, Wall S, et al. Health policy for low-income people in Michigan. Wash-
ington, DC: State Reports, The Urban Institute, 1997.
16. Leutz W. Policy choices for Medicaid and Medicare waivers. Gerontol 1999 Feb;39(l):86-93.
17. Vladeck BC. Overall perspectives. In: Wiener JM, Clauser SB, Kennell DL, eds. Persons with
disabilities: Issues in health care financing and service delivery. Washington DC: Brookings
Institute, 1995.
18. Gurley RJ, Lum N, Sande M, et al. Persons found in their homes helpless or dead. New Engl J
Med 1996 Jun 27;334(26):1710-16.
19. Perryman PK, manager of Lifeline. Southeastern Michigan Chapter of the American Red
Cross, Detroit, MI, 1999. Personal correspondence.
20. Koch WJ. Emergency response system assists in discharge planning. Dimens Health Serv 1984
Nov;61(ll):30-31.
21. Cain BA. Effects of a lifeline program on hospitalization. Master's thesis, California State Uni-
versity, Long Beach, 1987.
22. Roush RE, Teasdale TA, Murphy JN, et al. Impact of a personal emergency response system on
hospital utilization by community-residing users. South Med J 1995 Sep;88(9):917-22.
23. McGadney BF, Madison A. Use of formal community services by family caregivers of African
American elders. Michigan Academician 1997;29(2):199-220.
24. Chatters LM, Taylor RJ, Jackson JS. Size and composition of the informal helper networks of
elderly African Americans. J Gerontol 1985 Sep;40(5):605-14.
25. Davis LH, McGadney BF. Self-care practices of black elders. In: Barresi CM, Stull DE, eds. Eth-
nic elderly and long-term care. New York: Springer, 1993.
26. McGadney BF. Stressors and social supports as predictors of burden for black and white care-
givers of elders with dementia. Unpublished dissertation, University of Chicago, 1992.
27. McGadney BF. Personal response systems for low-income elders in Southeastern Michigan.
Paper presented at the annual meeting of the Gerontological Society, Philadelphia, Nov 17,
1998.
28. Stone R, Cafferata GL, Sangl J. Caregivers of the frail elderly: a national profile. Gerontol 1987
Oct;27(5):616-26.
29. Wykle M, Kaskel B. Increasing the longevity of minority older adults through improved health
status. In: Gerontological Society of America, eds. Minority elders: five goals toward building
a public policy base. Washington, DC: Gerontological Society of America, 1994.
30. Taylor RJ. The extended family as a source of support to elderly African Americans. Gerontol
1985 0ct;25(5):488-95.
31. Taylor RJ. Aging and supportive relationships among black Americans. In: Jackson JS, ed. The
black American elderly. New York: Springer, 1988.
32. Harper MS. Elderly issues in the African-American community. In: Braithwaite RL, Taylor SE,
eds. Health issues in the black community. San Francisco: Jossey-Bass, 1992.
33. Herron J. Afterculture: Detroit and the humiliation of history. Detroit, MI: Wayne State Uni-
versity Press, 1993.
34. Soldo BJ, Hurd MD, Rodgers WL, et al. Asset and health dynamics among the oldest old: An
overview of the AHEAD study. J Gerontology Series B, Psychol Sei Soc Sei 1997 May;52 Spec
No:l-20.
35. McCaslin R, Calvert WR. Social indicators in black and white: some ethnic considerations in
delivery of service to the older adults. J Gerontol 1975 Jan;30(l):60-66.
36. Stump TE, Clark DO, Johnson RJ, et al. The structure of health status among Hispanic, African
American, and white older adults. J Gerontology Series B, Psychol Sei Soc Sei 1997 May;52
Spec No:49-60.
McGadney-Douglass     503
37. Rowe JW, Kahn RL. Successful aging. New York: Pantheon, 1998.
38. Kennedy LW, Silverman RA. Significant others and fear of crime among the elderly. Int J Aging
Develop 1984-1985;20(4):241-56.
39. Krause N. Neighborhood deterioration and social isolation in later life. Int J Aging Develop
1993;36(l):9-38.
40. Lawton MP. Environment and aging. Monterey, CA: Brooks/Cole, 1980.
41. Thompson EE, Krause N. Living alone and neighborhood characteristics as predictors of social
support in late life. J Gerontology Series B Psychol Sei Soc Sei 1998 Nov;53(6):S354-64.
42. Akers RL, LaGreca AJ, Sellers C, et al. Fear of crime and victimization among the elderly in dif-
ferent types of communities. Criminol 1987;25(4):487-505.
43. Kleinman PH, David DS. Victimization and perception of crime in a ghetto community. Crimi-
nology 1973;ll(3):307-43.
44. Ward RA, LaGory M, Sherman SR. Fear of crime among the elderly as person/environment
interaction. Soc Q 1986;27(3):327-41.
45. Coulton C, Frost AK. Use of social and health services by the elderly. J Health Soc Behav 1982
Dec;23(4):330-39.
46. Lewis DA, Salem G. Fear of crime: Incivility and the production of a social problem. New
Brunswick, NJ: Transaction Books, 1986.
47. McGadney BF, Starrels M, Neal M. Dynamics of work/family stress: Black and white care-
givers of elders. Paper presented at the annual meeting of the Gerontological Society of Amer-
ica, Atlanta, GA, Nov 20,1995.
48. Mittlemark MB, Shiferaw B, Anderson R, et al. Predictors of service utilization and unmet
needs for services in frail older adults: The Forsyth County Aging Study. Paper presented at
the North Carolina Consortium on Aging, Fall Symposium, Winston-Salem, NC, 1993.
49. McGadney-Douglass BF. The health care needs of the black elderly: from well to frail. In:
McGadney-Douglass BF, Logan SL, Freeman EM, eds. Health care and empowerment practice
in the black community: Empowerment, knowledge, skills and collectivism. Binghamton, NY:
Haworth, 2000.
50. Dibner AS. Effects of personal emergency response service on hospital use. Boston: Lifeline
Systems Inc., 1985.
DR. McGADNEY-DOUGLASS is visiting scholar at the University of Ghana, Legon Campus, Accra,
Ghana, West Africa. She completed the present paper while associate professor in the School of Social Work
at Wayne State University in Detroit, Michigan.