Research analysis

1

RUNNING HEAD: Sample Paper

A Sample Research Paper

Modified extensively from:

Stadtlander, L., Giles, M., Sickel, A., Brooks, E., Brown, C., Cormell, M., Ewing, L., Hart, D.,

Koons, D., Olson, C., Parker, P., Semenova, V., & Stoneking, S. (2013). Independent

living oldest-old and their primary health provider: A mixed method examination of the

influence of patient personality characteristics. Journal of Applied Gerontology, 34(7),

906–928. doi: 10.1177/0733464813482182

Independent Living Oldest-Old and Their Primary Health Provider:

A Survey Examining the Influence of Locus of Control

Lee M. Stadtlander

Walden University

2

RUNNING HEAD: Sample Paper

Abstract

This survey study examined 35 healthy, independent living individuals' (over 85 years)

perceptions of their relationship with their primary health provider (PHP) and health practices.

The majority indicated they visited their PHP just for preventative care; the number of PHP visits

per year was significantly lower than reported for individuals over 85 by the CDC, possible

reasons for this finding are provided. A positive relationship between locus of control (LOC) for

the oldest-old was found. Few participants indicated their PHP had discussed normal changes

with aging. This study has deepened understanding of the complexity inherent to the healthy

oldest-old adults' relationship with their PHP. The findings suggest this relationship relates to the

PHP's personal characteristics, the elderly patients' LOC, and the influence of the accompanying

patient escort.

3

RUNNING HEAD: Sample Paper

Independent Living Oldest-Old and Their Primary Health Provider:

A Survey Examining the Influence of Locus of Control

The 2000 U.S. Census (2001) reported 4.2 million people were over the age of 85 (1.5%

of the population); this group has been designated the “the oldest-old” by demographers and is

the most rapidly growing age group. Currently the cost of health service utilization for the oldest-

old averages $22,000 per year compared to $9,000 for individuals 65-74 years old (Krause,

2010). It would be logical to assume individuals on the less healthy side of the spectrum, who

require nursing care or are in facilities (21.9% of this age group, Gist & Hetzel, 2004), bias this

$22,000 mean. However, little is known about the oldest-old who are on the healthy end of the

continuum. How do these individuals interact with their Primary Health Provider (PHP), and

how do their personalities affect the interaction? Participants in the current study, while primarily

seeing physicians, also mentioned physician assistants and nurse practitioners; thus, the more

generic term, Primary Health Provider is used.

The patient-by-treatment-context interactive model of Christensen and Johnson (2002)

provides the framework for the study. The model proposes that the relationship between patient

characteristics (e.g., personality and beliefs) and patient adherence (e.g., preventative care) is

moderated by the treatment context (e.g., PHP characteristics and behavior). The model has been

successfully applied to specific illnesses; for example, renal insufficiency and hemodialysis

(Christensen, Moran, & Ehlers, 1999) and cardiac rehabilitation (Christensen et al., 1999). It

does not appear to have been previously applied in the present context of ongoing preventative

care in older adults. The patient [locus of control (LOC), beliefs, and behavior] and treatment

(PHP characteristics, health behaviors) variables in the current study were derived from the

literature and are consistent with this model.

4

RUNNING HEAD: Sample Paper

The current survey study attempts to clarify PHP interactions and patient personality

characteristics through surveys and interviews examining how independent living oldest-old

perceive their relationships with their PHP and how this affects their health care.

Oldest-Old and Physicians

The relationship between the physician and patient is unique; in a medical encounter,

often involving the meeting of two strangers, the patient reveals very personal topics and an

examination of the patient’s body occurs (Adelman, Greene, & Ory, 2000). How PHPs handle

such encounters can mean the difference between a patient following suggested medical advice

or disregarding it (Adelman et al., 2000).

When considering the oldest-old, one relevant issue is the frequency of physician or PHP

visit. M. Parks (The Centers for Disease Control, [CDC], Division of Heath Care Statistics,

personal communication, June 11, 2012), supplied the information shown in Table 1.

Table 1. Office visits by patient age: United States, 2009

Number of visits per person per

year1 (standard error)

65-74 years 6.7 (0.3)

75-84 years 8.3 (0.5)

85 years and over 7.8 (0.6)

1Visit rates are based on the July 1, 2009 set of estimates of the civilian non-institutional

population of the United States as developed by the Population Division, U.S. Census Bureau.

It is apparent that the oldest-old have fewer PHP visits than individuals who are younger.

These trends are similar to Wolinsky, Mosely, and Coe's (1986; Wolinsky, Arnold & Nallapati,

1988) who reported the oldest-old in their sample had fewer physician visits than did younger

individuals. Wolinsky et al. (1988) speculated the decline in physician visits might be due to a

general decline in mobility, making accessing health care more difficult. This is a reasonable

5

RUNNING HEAD: Sample Paper

assumption for the least healthy individuals, but it is not clear if this assumption holds for the

healthier members of this age group.

Patient Satisfaction and the Elderly

There is a large literature on general patient satisfaction (Hertz, 2012; Lee & Kasper,

1998; Sherbourne et al., 1992). Patient satisfaction is an indicator of quality of care; however,

such studies tend to be surveys with little opportunity for the participant to go beyond the

questions and answers listed.

Kong, Camacho, Felman, Anderson, and Balrishnan (2007) reported those over 65 had

higher physician satisfaction scores and were less concerned with waiting times than were the

younger adults. Lee and Kasper (1998) found that relative to 65-69 year olds, people 80-84 were

20-30% less likely to be highly satisfied with their quality of care and physician quality. The

older patients objected to the physician's lack of technical skills and interpersonal manner.

However, Lee and Kasper (1998) relied on archival data and were not able to ask open-ended

questions of their participants. The sample was limited to people under the age of 85, not the

oldest-old. In addition, medical schools have made a recent effort to train physicians in geriatric

care (Croasdale, 2008; Siegler & Capello, 2005), which may affect the current elderly's opinions

of their PHP.

The present study examined a sample of independently living oldest-old from across the

United States using a survey method. There is a lack of research examining certain oldest-old

adults' individual characteristics that are known to be associated with health care and health

providers. Thus, the current study also examined the relationship between LOC to participants'

opinions of their PHP.

6

RUNNING HEAD: Sample Paper

LOC

LOC is derived from Rotter’s (1966) social learning theory. Rotter hypothesized people

who view reinforcements as contingent on their own behavior (internals) are better adjusted than

those who see reinforcements as determined by fate, chance, or powerful others (externals). The

three subscales within the LOC measure are internal control, powerful others, and chance

(Levenson, 1973; Rotter, 1966).

Rennemark, Holst, Fagerstrom, and Halling (2009) found a negative correlation between

physician visits and functional ability, education level, and internal LOC. High scores on the

powerful others and chance subscales are predictors of patients' trust in their physician (Brincks,

Feaster, Burns, & Mitrani, 2010). High internal LOC scores also tend to be correlated with being

proactive in seeking information, making health modifications, and treatment adherence

(Macaden & Clarke, 2010). The present survey study will examine the interrelationships between

LOC and opinions of the oldest-old about their PHP.

Method

Ten doctoral student researchers served as research assistants (RAs) for the study under

the direction of three faculty members [Primary Investigators (PIs)]. RAs were recruited through

student list servs of a large online university. Ten psychology students were chosen as RAs based

upon an interest in gerontology, their ethnicity, and their geographical location to ensure a

nationally representative sample for the study. None of the RAs had previously worked with the

PIs in any related research. All RAs were volunteers and received course credit for their

participation in an online lab course and their work in the study. The RAs were trained on the

design and study protocol. Walden University's Institutional Review Board approved this

project: approval #03-09-11-0784474.

7

RUNNING HEAD: Sample Paper

Participants

Participants were recruited through the RAs’ social network; potential participants were

sent an ad describing the study with a link to SurveyMonkey. The final sample consisted of 35

independent living participants ranging in age from 84 to 96 (M = 88.4 yrs., SD = 3.12). Our

definition of independently living included individuals in assisted living facilities but did not

include nursing home residents. Six males (17.1%) and 29 females (82.9%) participated in the

study; two individuals self-reported as African American (5.7%), and all others were Caucasian.

Twenty-four were widowed (68.6%), seven were married (20.0%), one was divorced (2.9%), and

three were single (8.6%). Four (11.4%) did not finish high school, 12 (34.3%) had just a high

school diploma, 15 (42.9%) had some college or trade school, 3(8.6%) received a bachelor’s

degree, and one (2.9%) had a master’s degree. Sixteen states were represented: Arkansas,

California, Florida, Georgia, Idaho, Kansas, Maryland, New Mexico, New York, North Carolina,

Oregon, Pennsylvania, Texas, Virginia, Washington, and West Virginia.

Procedures

The surveys and open-ended questions were conducted through SurveyMonkey. All

participants completed a consent form at the beginning of the survey. The survey consisted of 24

items from The Levenson Multidimensional Locus of Control scale (Levenson, 1973),

demographic questions, and three open-ended questions asking about their experiences with their

primary care physician. All participants who qualified for the study completed the surveys.

Measures

Participants completed a detailed demographic survey and the Locus of Control measure.

The Levenson Multidimensional Locus of Control scale (Levenson, 1973) is a 24-item

assessment measure, which includes three separate subscales: internal LOC, powerful-others

8

RUNNING HEAD: Sample Paper

LOC, and chance LOC. Respondents were required to indicate the extent to their agreement with

statements on a six-point scale ranging from “strongly agree” to “strongly disagree.” The internal

scale statements include items such as, “When I make plans, I am almost certain to make them

work.” The powerful-others scale statements include items such as, "I feel like what happens in

my life is mostly determined by powerful people." The chance scale statements include items

such as, "When I get what I want, it is usually because I’m lucky." Previously reported internal

consistency (Cronbach alpha) for the subscales is 0.58 for internal, 0.71 for powerful others, and

0.67 for chance (Levenson, 1973). In the current study, the Cronbach alpha was 0.75 for the

internal subscale, 0.79 for powerful others, and 0.82 for chance.

There were two primary research questions for the study. (a) How do the oldest describe

their relationship with their PHP relative to their reported health practices? (b) How do scores on

the LOC interact with PHP relationship descriptions and reported health practices?

Results

The results section is organized through the two research questions. The first research

question asked: How do the oldest-old describe their relationship with their PHP relative to their

reported health practices? Thirty-three (94.3%) individuals indicated they were satisfied with

their PHP, 2 (5.7%) were dissatisfied. An example of a satisfied comment:

"Satisfied. She explained what the results of everything were, what could be done about it

and what couldn’t be done about it" (Jane, age 86).

A dissatisfied comment:

"Dissatisfied. Well, that again had to do with this shoulder thing. And she was going to

send me to a specialist. And I asked her what the point was in sending me to the

specialist. And she said, they probably could do more than she could. And so, on my way

home I got to thinking about it. And as soon as I got to the bottom of the hill I called and

told them I wasn't going to go because there was no point. And she wouldn't talk to me. I

talked to the nurse and I talked to the receptionist, but she wouldn't talk to me. So, I

decided that was it" (Mona, age 90).

9

RUNNING HEAD: Sample Paper

On the demographic survey, 32 (91.4%) indicated they liked the provider. Individuals

without a chronic condition were less likely to indicate their relationship with the PHP was good

or very good (r(34) = -.358, p<.05).

Participants indicated their PHP talked to them from 2 to 30 minutes, with a mean of 15.8

minutes. During the interview, participants described their PHP. Some examples of positive

comments made by participants about their PHP:

"[He gives me] all the time that I need when I go, and as I said, he doesn’t rush me. He

gives me time to think them out" (Lora, age 85).

"He really explains everything good, to me. What’s wrong and what I got to do. I feel

real free talking to him" (Evelyn, age 87).

Some negative comments:

"Her mind just seems to be somewhere else. And I think she's got problems" (Mona, age

90).

"I don’t think he listens to me the way he should" (Terry, age 90, female).

A majority of participants indicated someone went with them to the PHP (n = 21, 60%);

primarily family members escorted them (spouse 14.7%, daughter/daughter-in-law 26.5%,

son/son-in-law 5.9%), 14.7% went with a friend. An escort to the PHP increased the likeability

of the PHP (r(35) = -.347, p<.05), and the PHP was more likely to be described as "kind" (r(35)

= -.516, p<.01).

Only one person reported audio taping the PHP visit. Twenty-five people (71.4%)

reported either family or staff where they live asked them about what happened during the visit.

Participants were asked on the demographic survey about various health-related

practices. Twenty-nine percent indicated occasionally or frequently drinking alcohol; males were

more likely to be in these categories (p< .05). The majority (91%) indicated they do not use

tobacco. Seventy-seven percent reported they occasionally or frequently exercised. Eighty-six

10

RUNNING HEAD: Sample Paper

percent indicated they read for pleasure, the majority of whom were women (p < .05). Twenty-

six percent of the participants indicated they never or seldom gambled, while 97.1% ate meat on

a regular basis. The combination of tobacco use, drinking alcohol, and gambling tended to be

correlated (p< .05).

The second research question examined: How do scores on LOC interact with PHP

relationship descriptions and reported health practices?

The LOC scale had three subscales: internal (M = 35.9, SD = 9.1, range = 12-48),

powerful others (M = 16.8, SD = 10.6, range = 0-39), and chance (M = 21.3, SD = 11.4, range =

0-45). Higher chance score was related to a higher powerful others score [r(35) = .722, p < .001].

A high internal LOC score was correlated with a high resiliency score [r(34) = .364, p < .05].

Having an escort to the PHP was related to a higher powerful others score [r(32) = .397, p < .05].

A difference was present for internal LOC and gender: no males were in the low range [r(34) = -

.364, p< .05]. Individuals with a high internal LOC were more likely to ask for a referral from

their PHP than those with a low score [r(34) = .431, p< .01]. Individuals with a low belief in

powerful others were more likely to state they exercise regularly [r(34) = -.315, p< .05], as were

individuals with a low belief in chance [r(34) = -.513, p< .01].

Discussion

The first research question asked: How do the oldest-old describe their relationship with

their PHP relative to their reported health practices? Lee and Kasper (1998) reported relative to

65-69 year olds, people 80-84 were 20-30% less likely to be highly satisfied with their quality of

care and physician quality. The older individuals particularly had issues with the physicians’

technical skills and interpersonal manner.

11

RUNNING HEAD: Sample Paper

Contrary to Lee and Kasper (1998), the majority of the participants in the current study

indicated they were satisfied with and liked their PHP. Two participants indicated they were

dissatisfied; these individuals indicated they felt the PHP did not pay sufficient attention to their

concerns, which is in accordance with Lee and Kasper's (1998) findings. Both of the dissatisfied

individuals were over the age of 90; while the number is insufficient to draw any conclusions, it

suggests that additional research should examine the topic further, particularly with the oldest

individuals.

Males were most likely to report drinking alcohol regularly. The majority (91%) of

participants indicated they do not use any form of tobacco (consistent with Cherry et al., 2011),

and they read for pleasure (85.7%). Reading has been associated with a reduction in the

development of dementia (Tesky, Thiel, Banzer, & Pantel, 2011). Virtually all participants

(97.1%) indicated they eat meat regularly; although meat has been a traditional dietary staple for

this age group, its use is related to an increase in cardiovascular risk (American Heart

Association, 2012; Polychronopoulos et al., 2010). The combination of tobacco use, drinking

alcohol, and gambling was correlated; thus, PHPs should be aware these practices occur

concurrently, particularly with males.

The results suggest it would be advantageous for PHPs to recommend that someone

accompany the oldest-old on visits. Individuals who had escorts tended to indicate they liked

their PHP and considered them more kind than people who went alone. There are a number of

possible explanations for this finding; it may be that escorts encourage patients to change PHPs

if they do not approve of them. On the other hand, having someone else approve of the PHP may

cause a halo effect (Greenwald & Banaji, 1995), improving the oldest-olds' perception of the

12

RUNNING HEAD: Sample Paper

PHP. An alternative is those who like their PHP may be more motivated to get treatment, and

thus more likely to secure an escort. Additional research is needed in this area.

Only five (31.4%) participants indicated their PHP had discussed normal changes with

aging with them. The participants expressed interest in a number of topics related to aging. It

appears it would be advantageous for PHPs to have brochures or booklets available on aging

topics for their patients.

The second research question examines: How do scores on LOC interact with PHP

relationship descriptions and reported health practices?

Previous research has reported that LOC is associated with patient-provider relationships.

Brincks et al. (2010) and Rennemark et al. (2009) found a negative correlation between

physician visits and functional ability, education level, and internal LOC. Individuals with a high

internal LOC score also tend to be proactive in seeking information, making health

modifications, and treatment adherence (Macaden & Clarke, 2010).

The current study was consistent with these findings (although the correlation with

education was not significant). A high internal LOC was correlated with a willingness to ask for

a referral from the PHP, evidence of assertiveness. A high internal LOC was also correlated with

visiting the PHP for preventative care, consistent with Macaden and Clarke (2010). Participants

with a high belief in powerful others were more likely to have an escort to the PHP. Perhaps

these individuals believe the escort would mitigate the power of the PHP or would be an

advocate if needed.

Individuals with a low belief in powerful others were more likely to state they exercise

regularly, consistent with previous research (Grant-Savela, 2009). This finding suggests internal

beliefs may be the influencing factor rather than an external source such as the media or PHP

13

RUNNING HEAD: Sample Paper

recommendation. Individuals with high scores on the chance subscale were also less likely to

indicate they would confront a PHP when they were unhappy with their care. These individuals

appear to believe confrontation would not be effective, but rather they must cope with the status

quo.

The patient-by-treatment-context interactive model of Christensen and Johnson (2002)

was found to be consistent with the current study. It led to the prediction that there would be

consistent preventative care (adherence) when the patient has higher internal LOC, which was

found.

Further, it was also predicted that there would be less preventative care (lack of

adherence) when the patient had lower internal LOC (more external), which was found. It was

expected that higher powerful others and chance scores would be indicative of less trust in the

PHP, these scores were correlated with less exercise and that they would not confront the PHP.

Additional research is indicated which specifically targets oldest-old individuals in

minority groups, in order to examine the relationship of the model to their care. In the current

study, the participants were on the healthier end of the spectrum for their age group. It would be

interesting to examine the model's reliability with less healthy members of the age group.

It is important to consider the generalizability of the current study. This study examined

relatively healthy, independently living, oldest-old adults recruited through online students'

social networks; thus, the results are specific to this group and may not be generalizable to the

general population of oldest-old. This sample is only generalizable to the extent other individuals

share the sample’s key demographics. A modified collaborative sampling technique was used,

which is commonly used to recruit hard-to-reach populations (Knight et al., 2009); therefore, it is

not known the extent to which the current sample reflects the population. This sampling method

14

RUNNING HEAD: Sample Paper

was chosen specifically because the researchers wanted to reach minority participants, yet it still

resulted in only a small number of minorities in the sample. This experience highlights the need

for more attention to the process of recruiting minority oldest-old for research projects.

This survey study's results suggest the relationship between the oldest-old and their PHP

is more complex than has been previously reported. The PHP relationship with the oldest-old

appears to relate to not only the PHP's personal qualities but also personality characteristics and

beliefs of the elderly patients and even the influence of the accompanying escort of the patient.

15

RUNNING HEAD: Sample Paper

References

Adelman, R. D., Greene, M. G., & Ory, M. G. (2000). Communication between older patients

and their physicians. Clinics in Geriatric Medicine, 16(1), 1–24.

American Heart Association. (2012). Meat, poultry and fish. Retrieved May 6, 2012, from

http://www.heart.org/HEARTORG/GettingHealthy/NutritionCenter/Meat-Poultry-and-

Fish_UCM_306002_Article.jsp.

Brincks, A., Feaster, D., Burns, M., & Mitrani, V. (2010). The influence of health locus of

control on the patient-provider relationship. Psychology, Health & Medicine, 15(6), 720–

728.

Cherry, K. E., Walker, E. J., Brown, J. S., Volaufova, J., LaMotte, L. R., Welsh, D. A., Su, L. J.,

Jazwinski, S. M., Ellis, R., Wood, R. H., & Frisard, M. I. (2011). Social engagement and

health in younger, older, and oldest-old adults in the Louisiana Healthy Aging Study.

Journal of Applied Gerontology, 32(1), 51–75.

Christensen, A. J., & Johnson, J. A. (2002). Patient adherence with medical treatment regimens:

An interactive approach. Current Directions in Psychological Science, 11(3), 94–97.

Christensen, A. J., Moran, P. J., & Ehlers, S. E. (1999). Prediction of future dialysis regimen

adherence: A longitudinal test of the patient by treatment interactive model. Paper

presented at the annual meeting of the Society of Behavioral Medicine. San Diego.

Christensen, A. J., Edwards, D. L., Moran, P. J., Burke, R., Lounsbury, P., & Gordon, E. I.

(1999). Cognitive distortion and functional impairment in patients undergoing cardiac

rehabilitation. Cognitive Therapy and Research, 23(2), 159–168.

16

RUNNING HEAD: Sample Paper

Croasdale, M. (2008). Gearing up for a graying generation: Training more doctors in geriatrics

skills. American Medical Association. Retrieved August 31, 2010, from http://www.ama-

assn.org/amednews/2008/06/09/prsa0609.htm#relatedcontent

Gist, Y. J., & Hetzel, L. I. (2004). We the People: Aging in the United States. (U.S. Census

Bureau Publication No. CENSR-19). Washington, DC: U.S. Government Printing Office.

Grant-Savela, S. D. (2009). Active living among older residents of a rural naturally occurring

retirement community. Journal of Applied Gerontology, 29(5), 531–553.

Greenwald, A., & Banaji, M. R. (1995). Implicit social cognition: Attitudes, self-esteem, and

stereotypes. Psychological Review, 102(1), 4–27.

Hertz, B. (2012). Doctor on solid footing by addressing health need. Medical Economics, 89(1),

28–29.

Knight, G. P., Roosa, M. W., & Umaña-Taylor, A. J. (2009). Studying ethnic minority and

economically disadvantaged populations: Methodological challenges and best practices.

Washington, DC: APA.

Kong, M. C., Camacho, F. T., Felman, S. R., Anderson, R. T., & Balrishnan, R. (2007).

Correlates of patient satisfaction with physician visit: Differences between elderly and

non-elderly survey respondents. Health and Quality of Life Outcomes, 5, 62–67.

Krause, N. (2010). Close companions at church, health, and health care use in late life. Journal

of Aging & Health, 22(4), 434–453

Lee, Y., & Kasper, J. D. (1998). Assessment of medical care by elderly people: General

satisfaction and physician quality. Health Services Research, 32(6), 741.

Levenson, H. (1973). Multidimensional locus of control in psychiatric patients. Journal of

Consulting and Clinical Psychology, 41(3), 397–404.

17

RUNNING HEAD: Sample Paper

Macaden, L., & Clarke, C. (2010). The influence of locus of control on risk perception in older

South Asian people with Type 2 diabetes in the UK. Journal of Nursing & Healthcare of

Chronic Illnesses, 2(2), 144–152.

Polychronopoulos, E., Pounis, G., Bountziouka, V., Zeimbekis, A., Tsiligianni, I., Qira, B., & ...

Panagiotakos, D. (2010). Dietary meat fats and burden of cardiovascular disease risk

factors, in the elderly: A report from the MEDIS study. Lipids in Health & Disease, 9,

30–35.

Rennemark, M., Holst, G., Fagerstrom, C., & Halling, A. (2009). Factors related to frequent

usage of the primary healthcare services in old age: Findings from the Swedish National

Study on Aging and Care. Health & Social Care in the Community, 17(3), 304–311.

doi:10.1111/j.1365-2524.2008.00829.x

Rotter, J. B. (1966). Generalized expectancies for internal versus external control of

reinforcement. Psychological Monographs: General and Applied, 80(1), 1–28.

Sherbourne, C. D., Hays, R. D., Ordway, L., DiMatteo, M. R., & Kravitz, R. L. (1992).

Antecedents of adherence to medical recommendations: Results from the medical

outcomes study. Journal of Behavioral Medicine, 15(5), 447–468.

Siegler, E. L., & Capello, C. F. (2005). Creating a teaching geriatric service: Ten important

lessons. Journal of the American Geriatrics Society, 53(2), 327–330.

Tesky, V. A., Thiel, C., Banzer, W., & Pantel, J. (2011). Effects of a group program to increase

cognitive performance through cognitively stimulating leisure activities in healthy older

subjects: The AKTIVA study. Geropsych: The Journal of Gerontopsychology and

Geriatric Psychiatry, 24(2), 83–92.

18

RUNNING HEAD: Sample Paper

U.S. Census. (2001). The 65 Years and Over Population: 2000. Retrieved July 26, 2007, from

http://www.census.gov/prod/2001pubs/c2kbr01-10.pdf

Wolinsky, F. D., Arnold, C. L., & Nallapati, I. V. (1988). Explaining the declining rate of

physician utilization among the oldest old. Medical Care, 26(6), 544–553.

Wolinsky, F. D., Mosely, R. R., & Coe, R. M. (1986). A cohort analysis of the use of health

services by elderly Americans. Journal of Health and Social Behavior, 27, 209–219.