Technology use by rural and urban oldest old

Abstract

Objectives

Technologies designed to optimally maintain older people as they age in their desired places of residence are proliferating. An important step in designing and deploying such technologies is to determine the current use and familiarity with technology in general among older people. The goal of this study was to determine the extent that community-dwelling elderly at highest risk of losing independence, the oldest old, use common electronic devices found in residential urban or rural settings.

Methods

We surveyed 306 nondemented elderly age 85 or over; 144 were part of a rural aging study, the Klamath Exceptional Aging Project, and 181 were from an urban aging cohort in Portland.

Results

The most frequently used devices were televisions, microwave ovens, and answering machines. Persons with mild cognitive impairment were less likely to use all devices than those with no impairment. Higher socioeconomic status and education were associated with use of more complicated devices. Urban respondents were more likely than rural ones to use most devices.

Conclusion

Technology use by very old community-dwelling elderly is common. There are significant differences in use between rural and urban elderly.

1. Introduction

Use of technology to monitor the safety and well-being of the elderly in their homes can improve quality of life for home-bound elderly and increase the amount of time they can live independently outside of institutions [10]. This technology can improve communication between caregivers or clinicians with home-bound elderly, relieving the strain on caregivers and allowing them more time for their own work or other pursuits [7,16,21,23,25,27,28,35]. Technology can also be used to remind homebound elderly to take medicines or monitor vital signs or laboratory parameters [2,8]. There is evidence that use of technology in the home of elderly persons can be cost-effective, reducing clinic visits and other health costs [6,9]. Studies have shown that elderly persons and their caregivers are receptive, with some concerns about violation of privacy, to the use of technology to monitor the elderly in their homes [5,41]. New devices designed for use by older persons in the community may be easier to use and more effective if these devices are familiar due to every day use of similar devices. The current literature suggests that there is room for improvement in this area [7]. Little is known, however, about the everyday use of existing technology by the elderly.

What is known about technology use by the elderly generally applies to the younger elderly, although those at greatest risk for losing independence are among the rapidly growing octogenarian and older populations. These individuals may be particularly challenged in using technologies because of their higher rates of cognitive and physical impairment [1]. Much of the knowledge about use of technology among senior populations has focused on computers. In a study involving computer use among 352 persons over the age of 60 in Wales, only 79 (22.4%) had used a computer in the previous 12 months [36]. A 2003 Pew Foundation survey found that 17% of rural persons over age 65 accessed the Internet, compared to 22% of urban seniors [4]. This study indicated that rural Americans of all ages were less likely than urban Americans to use the internet, although the difference was partly accounted for by lower income and education levels in rural areas [4]. Elderly persons with disabilities have been shown to be less likely to use computers than their able-bodied peers [12]. Studies indicate that elderly persons are less likely than younger ones to enjoy using computers, but that experience and guidance can improve their usage experience [15], and that better-quality research in how to best utilize technology in care of the elderly is needed [18].

The challenge of using technologies is not limited to devices such as computers, but extends to common household devices such as televisions, telephones and microwave ovens. Use of the latter class of items in particular has not been studied in the oldest populations.

2. Methods

We surveyed 306 elderly persons without dementia to find out about their day-to-day use of common technological devices. All respondents were age 85 or older and were participants in existing community-based aging cohort studies conducted by the Oregon Health and Science University Layton Aging and Alzheimer’s Disease Center. Forty-two percent (128) of the subjects are participants in a rural-based study, the Klamath Exceptional Aging Project (KEAP), which is a joint project of the Layton Center in Portland, Oregon and the Merle West Center for Medical Research in Klamath Falls, Oregon. The remainder of the sample (178; 58%) was an urban-based cohort of healthy aging study subjects drawn from the Portland, Oregon metropolitan area based Layton Center

In both the rural and urban cohorts, research nurses or trained research assistants visit study subjects at their homes at six month intervals to update the medical history and perform physical examinations, laboratory tests and tests of cognitive function. A subset of assessments administered by study nurses as part of their ongoing research were used to stratify study subjects for this study as cognitively intact or with mild cognitive impairment (MCI). These included the Mini-Mental State Examination [17], Clinical Dementia Rating (CDR) scale [20], the Cumulative Illness Rating Scale (CIRS) [24], an Activities of Daily Living (ADL) rating [22] and standard scoring systems for socioeconomic status (SES) [19]. Criteria for the designation of intact cognition were a CDR of 0, a MMSE over 23, and no significant functional impairment. Mild cognitive impairment was defined as a CDR of 0.5 without significant functional impairment.

A technology use survey was designed to capture the use of common devices by the elderly living in the community. The survey was pilot-tested at a local senior center by twenty volunteers over the age of 80 and revised to improve ease of use and clarity.

The survey asked study participants to quantify how frequently they used eight common household devices (television, microwave, VCR/DVD player, CD player, cassette tape player, answering machine/voicemail, cellular telephone, and home security system/burglar alarm). They were also asked whether they used a computer, how frequently they used a computer, how they accessed computers, and what specific tasks they performed on the computer (Survey available on line). The technology survey was administered to study subjects by study staff during one of their regular visits with study subjects. The survey was administered to each subject between February 1, 2004 and August 31, 2004.

2.1. Statistical analysis

Technology use was viewed as a continuous variable based on a sum of the use of the following technologies: television, microwave, VCR/DVD player, CD player, cassette tape player, answering machine/voicemail, cellular telephone, and home security system/burglar alarm. To determine technology use, we first re-coded the eight technologies into ordinal categories. Daily was assigned “7”, Several Times a Week “5”, Once a Week “3”, Once a Month “2”, Less than once a month “1”, and Never “0”. The sum of uses for all eight devices determined a person’s total Technology Use score, which was assessed as a continuous variable.

Computer, email, and internet use were divided into dichotomous variables, using the following breakdown. Frequent users were designated as those who use computers daily, several times a week, and once a week. Infrequent users were those who used computers once a month, less than once a month, or never.

Living situation (both where and with whom) and education were recoded into dichotomous categorical variables. SES was recoded into a 3-group categorical variable (high/medium/low). The relationship between the six covariates were examined, using standard statistical methods (means, standard deviations, chi-square test for categorical variables, one-way analysis of variance (ANOVA) for continuous measures). The purpose of this step was to thoroughly understand the relationship between all of the predictor variables to each other (for example, “How are Geographic Location and SES correlated?”) before examining their relationship to the outcome variables.

The association between geographic location and technology was determined by assessment of the parameter estimate, 95% confidence interval (95% CI) and p-value from a linear regression analysis. The multivariable model was manually constructed using backward-stepwise regression, with inclusion criteria of p < 0.10. The ordered variables were sequentially fit into the model, starting with the predictor variables of interest (geographical location) and then variables demonstrating the strongest association with the outcome.

The model was stratified by CDR score, as preliminary analysis suggested effect modification by this value.

Covariates included in the model building component are based on background literature and prior findings. They include SES, education level, living situation, and CDR score. Logistic regression was used to assess the relationships between computer, email, and internet use and the geographic variable. All other aspects of the model building process remained the same as in the linear regression technology use analysis. Results of the association are reported in Tables 1 and 2.

Table 1.

Odds ratios for logistic regression models

	CDR = 0.0			CDR = 0.5
	OR	95% CI		OR	95% CI
Computer
Rural Vs Urban	0.34	0.18	0.67	0.992	0.301	2.825
Rural Vs Urban	0.37	0.19	0.75	2.19	0.57	8.43
Education (>High School)	1.77	0.96	3.21	5.24	1.38	19.89
Lives With (Alone/Spouse)	0.62	0.25	1.54	0.63	0.11	3.67
Email
Rural Vs Urban	0.26	0.114	0.58	.	.	.
				.	.	.
Rural Vs Urban	0.29	0.125	0.66	.	.	.
Education (>High School)	1.65	0.85	3.19	.	.	.
Lives With (Alone/Spouse)	0.92	0.33	2.54	.	.	.
Internet
Rural Vs Urban	0.42	0.17	1.00	.	.	.
				.	.	.
Rural Vs Urban	0.48	0.19	1.19	.	.	.
Education (>High School)	2.07	0.95	4.52	.	.	.
Lives With (Alone/Spouse)	0.69	0.23	2.06	.	.	.

Table 2.

Regression models with locale (urban/rural) as primary independent predictor

Variables:	Computer (0/1)			Email (0/1)			Internet (0/1)			Technology Use (Sum)
	Estimate	SE	P-value	Estimate	SE	P-value	Estimate	SE	P-value	Estimate	SE	P-value
CDR = 0.0
Intercept	−0.867	0.17	< 0.0001	−1.38	0.21	< 0.0001	−1.78	0.22	< 0.0001	24.17	0.61	< 0.0001
Locale (Rural)	−0.5361	0.17	0.0017	−0.68	0.21	0.001	−0.43	0.22	0.05	−4.63	1.11	< 0.001
Intercept	−0.67	0.24	0.005	−1.32	0.28	< 0.0001	−1.65	0.30	< 0.0001	21.15	1.73	< 0.0001
Locale (Rural)	−0.49	0.18	0.005	−0.62	0.21	0.0033	−0.37	0.23	0.1139	−3.81	1.17	0.0014
Education (> High School)	0.28	0.15	0.066	0.25	0.17	0.1368	0.37	0.20	0.0671	1.81	1.07	0.09
Lives With (Alone/Spouse)	−0.24	0.23	0.306	−0.04	0.26	0.8654	−0.18	0.28	0.5095	2.08	1.60	0.19
CDR = 0.5
Intercept	−1.5069	0.29	< 0.0001	.	.	.	.	.	.	19.25	1.62	< 0.0001
Locale (Rural)	−0.04	0.29	0.8869	.	.	.	.	.	.	−3.63	2.04	0.0796
Intercept	−1.44	0.45	0.0012	.	.	.	.	.	.	20.67	3.47	< 0.0001
Locale (Rural)	0.39	0.34	0.2543	.	.	.	.	.	.	−0.67	2.40	0.7798
Education (> High School)	0.83	0.34	0.0148	.	.	.	.	.	.	3.16	2.30	1.37
Lives With (Alone/Spouse)	−0.23	0.45	0.6042	.	.	.	.	.	.	−5.02	3.23	0.13

Statistical analyses were performed using SAS version 9.1.¹

3. Results

3.1. Demographic data

Demographic characteristics of the rural and urban study participants are shown in Table 3. The rural (KEAP) sample had more persons with a CDR score of 0.5, which correlates approximately with mild cognitive impairment in a person who is not demented, than the urban sample. The rural sample was also less likely to live independently, had a lower education, was of lower socioeconomic status, and had a higher burden of chronic illness (CIRS score) than the urban group. These factors had to be taken into account in our statistical model, as described above.

Table 3.

Summary of demographic data, comparing rural and urban samples. Data are specific to the time of the visit at which the Tech Survey was administered. Data include information on 306 people over the age of 85, all of whom were not demented. Some fields have missing data for technical reasons, for example MMSE scores cannot be obtained on people who are blind, so the total for the MMSE is less than 306. In this case, an N in parentheses follows the value in the table

Data	Total Sample N = 306	KEAP, N = 128	URBAN, N = 178	p value for difference between KEAP & URBAN
Age at time of survey, average	90.1 SD 2.9	89.8 SD 2.9	90.3 SD 3.0	0.21
Gender, percent (N) female	190 (62%)	73, (57%)	117, (66%)	0.12
Aged 85–89, N, % of sample	173, (57%)	74, (58%)	99, (56%)	0.77
Aged 90–94, N, % of sample	110, (36%)	46, (36%)	64, (36%)
Aged over 95, N, % of sample	23, (8%)	8, (9%)	15, (8%)
Percent white (Caucasian)	300, (98%)	128, (100%)	172 (97%)	N/A
Assisted Living	8%	12%	5%	NS
Living Independently	280 (92%)	112 (88%)	158 (90%)	0.03
Not Living Independently	33 (11%)	15 (12%)	10 (6%)
Living alone
With Spouse
With Other
Education (yrs)	13.4 (SD 3.0)	12.3 (SD 3.2)	14.2 (SD2.6)	0.000
Educated over 16 years	13%	9%	16%	< 0.0001
Educated 13–16 years	38%	24%	47%
Educated 9–12 years	40%	47%	35%
Educated 1–8 years	9%	20%	2%
High Socioeconomic status	100, (34%)	15, (12%)	85 (48%)	< 0.0001
Middle SES	96, (32%)	43 (36%)	53 (30%)
Low SES	101 (34%)	63 (52%)	38 (22%)
MMSE average score	27.1 (296) SD 2.2	25.8 (126) SD 2.3	27.8 (170) SD 1.9	< 0.0001
CIRS, average score(N)	21.0(218) SD 3.3	22.6(83) SD 3.36	20.0 (135) SD 2.79	< 0.0001
CDR of 0 (not demented), N	71%, 217	55%, 71	82%, 146	< 0.0001
CDR of 0.5 (MCI), N	29 %, 89	45%, 57	18%, 32
CDR Sum Of Boxes	0.48 SD 0.91	0.711 SD 0.97	0.315 SD 0.82	< 0.0001

3.2. Household technology use

Figure 1 illustrates differences in device use between the urban and rural cohorts, as well as between persons with mild impairment and those with no apparent impairment by CDR score. For most devices, persons with a CDR score of 0.5 were less likely to be users than those with a CDR score of 0. Most participants used their televisions ubiquitously, with telephone answering machines and microwaves used frequently, while 25 percent or less of the sample used VCR or DVD players, cell phones, tape players, and burglar alarms. The differences in use between urban and rural cohorts were variable, although urban users were more likely to use answering machines and burglar alarms than their rural counterparts, and rural elderly slightly more likely to use microwaves.

Fig. 1.

Percentage of sample who were frequent users of various devices. For each device, the rural data [R] are on the left and the urban [U] on the right. The persons with CDR=0 (dark bars) are on the left of each rural or urban pair and CDR = 0.5 (light bars) on the right.

After adjustment for educational status, socioeconomic status and living situation, we found, as detailed in Tables 1 and 2, that rural respondents with a CDR = 0.0 have a mean over-all or combined technology use approximately 3.81 (± 1.17) times less than urban subjects (p = 0.001). We found that within a population with a CDR = 0.5 mean technology use is equivalent for rural and urban subjects (p = 0.78).

The effect of gender, CDR score, age, education, socioeconomic status, and rural residency on the likelihood to be a frequent user of a device is detailed in Table 4. Gender generally had little effect on use, while differences due to education, SES, and rural residency were pronounced for some devices. Older respondents were less likely to use all devices.

Table 4.

Details of device use for respondent characteristics. The numbers refer to the percentage in each cell that were frequent users; for example, cell phones were frequently used by 13% of the total sample but 19% of males 9% of females. The p values are Pearson Chi-square values comparing the frequency of use between the various groups; for example, there was a probability of 0.04 that the difference between frequent cell phone use by men and women was not due to chance. The N refers to the number in each group (row) for whom data were available. The total sample was 306 persons. ED: Education. SES: Socioeconomic status

	N	Television	Micro- wave	VCR	TAPE	Answering Machine	Cell Phone	Burglar Alarm	Com- puter
Total	306	97%	80%	19%	22%	57%	13%	12%	18%
MALE	116	97%	78%	17%	22%	56%	19%	11%	21%
FEMALE	190	96%	83%	22%	22%	57%	9%	13%	17%
p value		0.42	0.54	0.30	0.88	0.73	0.04	0.73	0.62
CDR 0	217	98%	85%	21%	25%	65%	14%	13%	24%
CDR 0.5	89	93%	71%	14%	16%	36%	10%	9%	5%
p value		0.03	0.02	0.17	0.21	0.00	0.47	0.50	0.00
AGE 85-89	173	97%	87%	23%	25%	64%	15%	13%	23%
AGE 90-94	110	96%	74%	15%	17%	51%	11%	12%	14%
AGE>94	23	96%	70%	13%	22%	30%	9%	9%	9%
p value		0.90	0.036	0.16	0.53	0.00	0.69	0.774	0.05
ED ELEM	28	100%	86%	16%	21%	29%	7%	4%	0%
ED HS	121	96%	83%	16%	21%	50%	10%	9%	13%
ED COLL	114	96%	78%	23%	23%	62%	17%	17%	25%
ED GRAD	41	98%	78%	22%	27%	78%	17%	15%	27%
p value		0.84	0.64	0.5	0.78	0.00	0.18	0.55	0.00
SES LOW	101	95%	82%	14	20%	42%	8%	6%	9%
SES MID	96	98%	83%	22	22%	46%	11%	15%	17%
SES HI	100	97%	75%	24	25%	82%	20%	17%	30%
p value		0.64	0.43	0.18	0.76	0.00	0.08	0.23	0.00
RURAL	128	95%	88%	15%	23%	35%	9%	2%	5%
URBAN	178	98%	75%	22%	21%	72%	16%	20%	28%
p value		0.24	0.01	0.125	0.72	0.00	0.04	0.00	0.00

Table 5 shows how many respondents had particular devices. It also indicates the number who had devices and did or did not use them and the number who did not have a device who did or did not desire access to it. For example, of the 85 of 306 (28%) respondents who had a cell phone, 18 (21%) never used them, while of the 221 who did not have a cell phone, 44 (22%) indicated they would like to have one, with 177 (78%) indicating no interest.

Table 5.

Summary of device ownership. The numbers are N, of the total sample of 306

	Have	Don’t have	Have and use	Have but never use	Don’t have and don’t want	Don’t have, but would use	Total
Television	301	5	299	2	4	1	306
Microwave	283	23	267	16	20	3	306
VCR or DVD	205	101	159	46	83	18	306
CD Players	153	153	122	31	124	29	306
Tape Players	213	93	173	40	84	9	306
Answer Machine	201	105	189	12	94	11	306
Cell Phone	85	221	67	18	177	44	306
Burglar Alarm	67	239	47	20	199	40	306
Computer	75	231	60	15	195	36	306

3.3. Computer use

110 (36%) respondents indicated they had ever used a computer, and 100 (33%) had one in their home. 56 (18%) indicated that they used a computer at least once a week; of these only 6 were rural residents. 38 (18%) of the 206 respondents who did not have a computer in their home indicated they would like to have one. 22 persons reported that they had used a computer at a library. The uses for computers reported by the 56 frequent users included email (55 users), Internet use (33 users), getting news (25 users), getting health information (19 users), playing games (34 users), tracking expenses (12 users), shopping (10 users), and other uses (25 users).

After adjusting for covariables, rural respondents with CDR = 0.0, were significantly less likely to use a computer than an urban subject (odds ratio 0.37 (0.19, 0.75). For the population with a CDR = 0.5, the odds of using a computer for a rural subject are 2.19 (0.57, 8.43) greater than the odds for using a computer for an urban subject. This finding is not statistically significant (p = 0.25).

With respect to what they used a computer for, and after controlling for covariates (age, sex, education, socioeconomic status), there was no difference (p = 0.14) in use of the Internet between elderly rural and urban subjects. Although there was no difference in Internet use between rural and urban elderly respondents with a CDR = 0.0, rural subjects were less likely than urban ones to use email (odds ratio: 0.29, (0.13, 0.66); p = 0.003), after controlling for other covariates in the model.

4. Discussion

We found in this survey of the oldest old that people over the age of 85 are most likely to use devices that have been available the longest: televisions, microwave ovens, and answering machines. The more recently developed or deployed technologies such as cell phones and personal computers are the least used. The frequent use of televisions and microwave ovens may also be a reflection of their perceived utility, as described by Melenhorst [32]. Televisions provide a connection with what is going on in the world as well as a source of entertainment and may be particularly important for persons living alone, which is more common with advancing age. Similarly, microwave ovens allow easy preparation of convenient pre-packaged foods appropriate for a person living alone with disabilities affecting movement, endurance and vision.

The frequency of use of common electronic devices among very old persons in our study dropped off rapidly with the relative complexity of the device. While almost all respondents watched television regularly, only 18% used a computer with any frequency. We also noted a significant rural-urban digital divide, with only 6 of 56 frequent computer users living in a rural area. Higher socioeconomic status and education were also associated with use of more complex devices such as cell phones, DVD players, or computers. The finding that for many devices, elderly persons who did not have some devices were interested in acquiring them indicates that, as discussed in our introduction, there is interest in use of technology among the elderly. However, it is certainly possible that elderly persons, like all of us, might report a potential interest in a technology that, if they did have it, would not necessarily translate into finding it useful or using it regularly.

Although persons with a CDR score of 0.5 are not considered demented by definition, we did find that mild cognitive impairment, defined as a CDR of 0.5, was consistently associated with lower likelihood of use of devices, but not with no use. Other studies have found that persons with mild impairment do use some devices and there is a need for devices designed specifically for their use [23]. It is likely that persons who were more frankly impaired would have even lower rates of device use; 14 persons with dementia (CDR = 1.0, data not shown) were surveyed in this study as well and there was a trend to less technology use, though even those individuals used some devices.

Prior studies have found that technology use in general declines with age, although older persons are often receptive to technology use if they can find a direct benefit and if it is designed in a way that makes it practical for them to use. In a study of American and Dutch adults comparing email users to non-users, older persons were more likely to use a technology for which they perceived a direct benefit, regardless of the difficulty involved in use. In this study, older persons found cell phones as more beneficial than email [31,32]. Cody et al., offered a four-month training program on Internet use to elderly residents of assisted living and independent living facilities (average age was 80). They found that increased Internet use was associated with a positive attitude towards aging, a low initial level of computer anxiety and high computer efficacy. They recommend tailoring computer instruction to individuals based on their clinical and cognitive characteristics. Elderly persons who were more connected in their community (less isolated) used a wider range of Internet functions and used e-mail more [11]. One factor limiting computer use among older persons can be declining vision and fine-motor coordination, so that devices such as mice may need design modification [38].

Ethical considerations are important in the use of technology in the homes of elderly patients and their caregivers. It is important to consider whether obtaining informed consent and considering whether any loss of privacy due to technology use is offset by a gain in independence and increased function [3,13, 26,30]. One survey showed that caregivers appreciated the benefits of technology but felt that it could potentially be dehumanizing compared to more personal contact [34]. The development of robots to monitor elderly in their homes has great potential but raises many of these issues [39]. Medico-legal aspects of home technology have also been addresssed [40].

5. Limitations

Our survey was designed by our investigators to learn about the use of common household devices by the elderly. The study is limited in that our survey was not validated in multiple samples. However, we have had a chance since this initial survey to return and repeat the survey among many of our study subjects and have found that their responses were not changed (data not shown).

We were not able to find other surveys of technology use by persons over the age of 85, but our study is consistent with published research regarding use of electronic devices by persons over the age of 65 [14,29]. The finding that socioeconomic status and mild cognitive impairment affects the use of devices by the elderly suggests the importance of keeping costs low and designing devices that are easy to use. Research indicates that designs that allow persons with mild memory or visual impairment to use devices, such as built in instruction screens that are cued by context may be cost-effective [33,37]. Once elderly persons gain access to a device and become more familiar with its use, they perceive it as more useful [37]. The gap between rural and urban use of technology is particularly important since many services that are less available in geographically remote regions, such as patient education or clinical data about individual patients could be accessed through computers. The fact that computers as classically conceived and televisions are merging in function provides an opportunity to make these new hybrid devices and services they may provide more widely used regardless of location.

In conclusion, we found that very elderly persons in both a rural and urban sample were frequent users of technology, and that use was predicted by education, socioeconomic status, age, and urban or rural residence. This information is useful for persons designing technology for use in the homes of elderly persons.