Society & Animals Journal of Human-Animal Studies
Logo - Society and Animals Journal
Volume 1, Number 1

 

Assessment of the Effectiveness of a Pet Facilitated Therapy Program in a Nursing Home Setting

Ira B. Perelle and Diane A. Granville 1

Mercy College

In the past twenty years Pet Facilitated Therapy (PFT) has been used, apparently successfully, with several populations, including nursing home residents. Studies report positive behavior changes as a result of PFT intervention, but little effort has been made to quantify such behavior changes. This study presents the results of a PFT program in a nursing home setting. Results were positive, and were measured using the Patient Social Behavior Scale, designed for this study. Nursing home residents showed an increase of social behaviors from pretest to midpoint and from midpoint to post test, but these behaviors declined four weeks after post test. Although both males and females showed an increase in social behaviors, males' and females' response patterns differed.

The first recorded use of animals in therapy was by William Tuke at the York Retreat, in England, founded in 1792 by the Society of Friends as an alternative to the subhuman conditions of the lunatic asylums of that time (Beck and Katcher, 1983). The basis of treatment was teaching self-control by having the ostensibly weaker creatures dependent on the patients. The York Retreat became a model for the reform of other asylums of the day (Blake, 1980). The next recorded use of animals in therapy is at Bethel, a residential treatment center for epileptics, founded in 1867 in Bielefeld, Germany. From its inception, pets were an integral part of the treatment protocol because of the belief of Bethel's founders that the use of animals was a common-sense approach to treatment. The animal therapy program, still in use, has been expanded to include farm animals and a wild game park (McCulloch, 1983; Bustad and Hines, 1984).

The first use of animals in an organized program in the United Stales appears to have been at the Army Air Corps Convalescent Center in Pauling, New York, in 1944. The Center, run by the American Red Cross, included a working farm and patients were encouraged to interact with the various farm and pet animals. Unfortunately, no records were kept on the effects of the human-animal interactions and the program was discontinued after World War II (McCullough, 1983; Levinson, 1969). No further use of animals in therapy is revealed in the literature until Boris Levinson described his use of animals as adjuncts in his practice of psychology (Levinson, 1962; 1969; 1975). Levinson advocated the use of animals in cases where affection and unconditional acceptance were indicated. He believed, "A pet can provide, in boundless measure, love and unqualified approval. Many elderly and lonely people have discovered that pets satisfy vital emotional needs" (Levinson, 1969, p 368). Levinson described his introduction of animals into the therapeutic milieu as "pet facilitated therapy" (Levinson, 1969), and this has been accepted by most professionals as being appropriately descriptive and accurate.

In the past two decades Pet Facilitated Therapy (PFT) has been used successfully with several populations: coronary patients (Friedman, Katcher, Lynch, and Thomas, 1979), hospitalized psychiatric patients (Corson and Corson, 1981), emotionally disturbed youth (George, 1988), prison inmates (Lee, 1984), and the elderly, both as outpatients (Messent, 1984) and institutionalized ( Brickel, 1979) Human-animal interactions in these programs range from complete responsibility to once-a-week visits (See Bustad, 1990, for an overview). Several studies report positive social behavior changes after introducing an animal into the nursing home environment. Corson and Corson (1978) found a decrease in patients' sense of loneliness and social withdrawal, and an increase in patients' positive interactions to staff when dogs were brought to a nursing home. Brickel (1979), using cats for relatively short periods, found nursing home residents to be more responsive to therapy and to perceive the visits to be pleasurable. The Australian Joint Advisory Committee on Pets in Society conducted a six month study of the interaction of 60 nursing home residents with a dog. Using pre- and post test questionnaires they found positive behavior changes in interest and conversation, and an increase in environment related behaviors (Salmon, Hogarth-Scott, and Lavelle, 1982). Pet Visitation programs, in which volunteers bring companion animals to interact with institutionalized people on a regular schedule, seem to be gaining in popularity (New York Times, 1984), although very few systematic studies of the results of such visits have been reported. This paper reports social behavior change related to a Pet Visitation program, and the relationship between social behavior change and residents' sex, in a nursing home population.

Method

Subjects for this study were 53 residents of a mixed-care nursing home in Westchester County, New York. The subjects were self-selected for inclusion in the program after a brief description of the program was provided by the nursing home staff. Self-selection introduced a potential bias in the outcome of the program, but it is difficult to determine the direction of the bias. Those residents participating in the program were generally less active socially; socially active residents were "too busy', or "didn't have time" to participate. Subjects' ages ranged from 35 years to 95 years, with a mean age of 75.39 years (SD = 11.72 years). Included were 18 males and 35 females. The length of residence ranged from 1 month to 165 months with a mean length of stay of 45.82 months (SD = 40.1 months). All of the subjects were ambulatory to some degree; most required wheelchairs or walkers.

The evaluation instrument for this study was the Patient Social Behavior Scale, designed for this study, and modeled loosely on the survey instrument used by Corson and Corson (1981) in their nursing home studies. The instrument was designed to be administered by the nursing home staff, therefore input was obtained from the nursing home staff to assure the final questionnaire would be acceptable to the staff, and to increase the probability that the questionnaire would be reliably administered by the staff. The questionnaire consisted of 12 questions, each with a five point response scale requiring an evaluation of the residents' various social and self-maintenance behaviors. Evaluation scores ranged from I to 5, with 5 being the most positive score obtainable.

The administration protocol required that the questionnaire (hereinafter called "test") be administered by the nurse most familiar with the resident on the first day of the study (pretest), during the fifth week (midpoint), during the tenth (final) week (post test), and one month after the final week (follow-up). The four tests required for each subject were administered by a single nurse. The nurses involved were trained in administration of the test to assure an acceptably high inter-nurse and intra-nurse reliability. Because of the nature of the research environment, it was not possible for the nurses that administered the rests to be "blind" as to research participants, nor was it possible to have several nurses test a single subject so as to assess inter-nurse reliability. There is no reason to believe, however, that the nurses were biased either for or against this investigation, nor that the nurses completed the tests in any but an unbiased, professional manner.

Resident Status Reports were completed by the nursing home administration to provide the necessary background data for each resident. The Status Reports included residents' age, sex (not always determinable from residents' name), date of admission, diagnosis, and several other variables not used in this study. There is no reason to question the validity of the

Status Report data

The Pet Visitation program at the nursing home was implemented by a team of student volunteers from a local college, under the supervision of two psychologists and one social worker. The volunteers were all enrolled in the college's Veterinary Technology program and had [raining in animal handling and in the procedures involved in the care and lifestyle of the institutionalized elderly prior to the start of the program.

The animals used in the program were selected for their gentleness and cooperation when handled by strangers. The animals consisted of four cats, two small dogs, and one rabbit. Six of the animals were brought to the nursing home each week. The program was conducted for ten weeks, with visits occurring once a week for two hours each.

During the visits, one or two students took an animal to the day room of one of the six floors of the nursing home, with the same students consistently assigned to the same floor. The animals were exchanged among all floors during each visit. Residents assembled in day rooms where they were encouraged by the student volunteers to stroke and/or handle the animals. Student volunteers also talked to the residents about the animals, about the residents' pets (primarily their former pets), and answered any questions asked by the residents.

Results

During the course of this program 18 residents were unable to have all four observations completed because of illness or discharge from the facility. Therefore, complete data sets consisting of pretest, midpoint test, post test, follow-up test, and resident status report were received from 35 residents. As a lest of the overall results of the animal visitation program, the scores of questions on each subject's tests were summed, means and standard deviations of each test were computed, and an Analysis of Variance was performed across tests. The result of this analysis (see Table 1) shows highly significant differences among the means of the tests. To further analyze these differences t -ratios were calculated between the various tests (see Table 2).

Table 1.
Analysis of Variance Report Across Tests

Source

DF

SS

MS

F-Ratio

Probability

Tests

3

3797.714

1265.905

18.23

<0.001

Error

136

9444.972

69.448

 

 

Total

139

13242.690

 

 

 

It can be seen that the mean score increases significantly between the pretest and midpoint, and the midpoint and post test. It can also be seen that the mean score decreases significantly between post test and follow-up, although it remains higher than the pretest mean.

Table 2.
Means and Standard Errors of each test;
t -ratios and Probabilities Between each Test (n=35)

Test

Mean

Standard Error

t -ratio

Probability

Pretest

39.14

1.623

 

 

 

4.808

<0.001

Midpoint

48.77

1.173

 

 

 

2.904

<0.01

Post test

53.57

1.165

 

 

 

-3.146

<0.01

Follow-up

47.37

1.589

 

 

 

3.623

<0.001

Pretest

 

 

 

 

To investigate any possible differences based on resident's sex in response to the PFT intervention, Analyses of Variance across tests within sex were computed. Results of these analyses are highly significant (see Tables 3 and 4).

 

Table 3.
Analysis of Variance Report Across Tests Within Males

Source

DF

SS

MS

F-Ratio

Probability

Tests

3

2696.275

898.7583

14.83

<0.0001

Error

36

2181.100

60.5861

 

 

Total

39

4877.375

 

 

 

Table 4.
Analysis of Variance Report Across Tests Within Females

Source

DF

SS

MS

F-Ratio

Probability

Tests

3

1926.990

642.331

9.87

<0.0001

Error

96

6245.922

65.062

 

 

Total

99

8172.913

 

 

 

To further investigate these results t -ratios were calculated across tests within sex, and across sex within tests. An interesting pattern emerges when the subjects are divided by sex and the same lest scores are compared (see Tables 5 & 6).

Table 5.
Means and Standard Errors by Sex;
t -ratios and Probabilities Within Tests Across Sex

Test

Males

Females

t -ratio

Probability

Mean

SE

Mean

SE

Pretest

32.90

1.929

41.64

1.939

3.195

<0.01

Midpoint

52.60

1.701

47.24

1.397

2.245

<0.05

Post test

52.90

2.830

53.84

1.378

0.299

NS

Follow-up

43.10

3.523

49.08

1.650

1.537

NS

It is apparent that male and female subjects not only responded to the visiting animals in rather different patterns, they were significantly different with respect to social and self-maintenance behaviors prior to the onset of this study; females scored considerably higher than males in the pretest ( p < 0.01; see Table 5). Although the introduction of visiting animals seemed to have a significant effect on both males and females (both males' and females' pretest to midpoint scores increased at p < 0.001; see Table 6), males' midpoint scores were significantly higher than females' ( p < 0.05; see Table 5). Females' lest scores increased significantly from midpoint to post test ( p < 0.001) while males' scores remained virtually unchanged (Table 6). There was no significant difference between post test scores of males and females (Table 5). By four weeks after cessation of the animals visits both males' and females, scores had fallen significantly (males: p < 0.003; females: p < 0.001; Table 6). Although the decrease in social and self-maintenance behavior score was significant, the behavior score was nevertheless significantly higher than before the study was implemented: Follow-up to pretest comparisons showed follow-up scores to be significantly higher than pretest (Table 6), with no difference between males and females (Table 5).

Table 6.
Means and Standard Errors of Tests by Sex;
t -ratios and Probabilities Between Tests Within Sex

Sex

Test

Mean

SE

t -ratio

Probability

Female

(n=25)

Pretest

41.64

1.939

 

 

 

3.79

<0.001

Midpoint

47.24

1.397

 

 

 

4.81

<0.001

Post test

53.84

1.378

 

 

 

-5.06

<0.001

Follow-up

49.08

1.650

 

 

 

6.19

<0.001

Pretest

 

 

 

 

Male

(n=10)

Pretest

32.90

1.929

 

 

 

8.68

<0.001

Midpoint

52.60

1.701

 

 

 

0.24

NS

Post test

52.90

2.830

 

 

 

-4.04

<0.003

Follow-up

43.10

3.523

 

 

 

2.54

<0.05

Pretest

 

 

 

 

Discussion

There is no doubt that the introduction of visiting animals into the nursing home environment during this study did facilitate a dramatic improvement in residents' social and self-maintenance behaviors. The results of statistical tests both across tests and across tests within sex indicate an immediate positive change in behaviors, an increase that was maintained for at least one month after termination of the study. Completely unanticipated was the effect of sex on the outcome measures: the rapid and sizable improvement in behaviors of males, leveling to a plateau, versus the initially slow but steady improvement in behaviors for females. The most probable explanation for this difference is in the differential behavior patterns of men and women under nonintervention conditions. Observation and interviews with the nursing home staff both indicate that female residents interact with each other and with staff at a far greater rate than do males. In fact, many male residents are reclusive in behavior, rarely interacting with anyone. This is confirmed by the extremely low pretest scores of males, 32.90, as compared to a pretest score of 41.64 for females ( t = 3.20; p < 0.01). It is likely that the presence of an animal did facilitate male social interaction with others, possibly reducing inhibitions to interactions, providing positive reinforcement contingent on interaction. Since males rarely socialized prior to the introduction of animals, and since human socializing seems to be a genetically predisposed behavior, months or years of self-imposed isolation quickly gave way to interaction and immediate improvement in various social behaviors. Females, however, interacted among themselves and with staff prior to the introduction of animals, so the animals did not provide the same stimulus for behavior change as they did for males.

Anecdotal accounts of individual behaviors during the study are numerous and interesting in terms of the study. One male subject, a long term resident, had not spoken to anyone since he was admitted to the facility. On the last pet visiting day, for the first time since his admission, he spoke to one of the volunteers (who had brought a camera) requesting a picture of himself with one of the dogs. Several residents who were never seen smiling by staff smiled regularly when interacting with the animals. Eleven residents had to be dressed by others prior to the start of the study; only three had to be dressed by others at the time of the post test. Unfortunately, this number increased to seven at the follow-up. As can be seen in the Tables, test scores of both male and female residents decreased significantly from post test to follow-up, but were still significantly higher than pretest scores. Undoubtedly scores taken later, perhaps six months after post test, would show further decline, although once the inter-resident social activity was initiated it very well may continue because of the reinforcing nature of the activity. The decline at follow-up does, however, indicate the desirability of a resident companion animal.

Animals have been associated with humans for at least fifty thousand years, initially perhaps as scavengers, then as working companions, as domesticated sources of food, and finally as pets and sources of pleasure (Lorenz, 1965). It is reasonable to assume an evolutionary advantage accrued to humans who maintained a beneficial relationship with animals, that is, humans who used animals profitably may have been able to reproduce at a greater rate than those who did not make use of animals. Although genes involved in behavior have yet to be mapped, if there is any genetic component of human behavior, animal association would have been selected for by the differential reproduction rate (Barash, 1977; Trivers, 1985) and modern humans would be genetically predisposed to keep and derive comfort from animals. Indeed, even though vast numbers of urban and suburban dwelling humans have no economic need to keep animals these days, a very large proportion of them do keep one or more companion animals. To a great many people, animals are important members of the family (Beck and Katcher, 1983), and to an increasingly large number of people animals are the only other members of their family. The use of animals as part of a therapeutic team to treat a number of behavioral disorders or eliminate undesirable behaviors is a natural progression in the historical human-animal relationship.

The results of this pet visitation program were, beyond doubt, positive. A note of caution is necessary. This study had the advantage of an institution with a cooperative administration and start: Experience and the literature indicate that most institutions are not at all amenable to the introduction of animals of any type. The volunteers who transported, maintained, and handled the animals were consistent in their commitment to the study. Their assistance is gratefully acknowledged. It is conceivable that volunteers alone, without animals, would have provided the same results, although post study conversations with the subjects indicate this is unlikely. A two-group experiment (volunteers only vs. volunteers with animals) would probably provide an answer.

A second note of caution relative to the animals is also necessary. The animals used in this study were chosen for their gentleness and their benign behavior towards humans. They all appeared to enjoy associating with humans, but they were all under constant observation of the volunteers for any sign of stress or fatigue. If any such signs appeared the animal was immediately retired to a quiet room where it was able to rest and quench its thirst. It is possible that some animal visitation programs do not have the benefit of trained and/or knowledgeable volunteers and, therefore, allow animals to become extremely stressed and uncooperative. This is completely unacceptable for the animal and would probably have a negative effect on the humans whom the animal was visiting. This study is an initial attempt to quantify the results of a Pet Facilitated Therapy intervention, and, as such, a further note of caution related to the design of this study must be stated. Subject self-selection and the absence of a control group always dilute the strength of results, and this study is not an exception. Observations of a control group during the program period, or base line observations for some extended period would have made our results much stronger, but because of the nature of the population and the institution, these measures were not possible. The institution allowed our work with their residents on the condition that all who wanted to participate, but only those who did want to participate would be able to do so. This, of course, precluded a control group. Prior base line observations would have been helpful, but the professional staff were not amenable to the additional work.

From the results of this study and similar results reviewed above it appears that animals introduced into an institutional environment do provide The catalyst for positive social behavior change. The animals seem to provide a unique contribution to the institutional environment, one that reduces inhibitions against social contact by males and, to a lesser degree, by females. This can be a result of earlier (perhaps happier) life experiences with animals, or perhaps some component of humans' genetic endowment is programmed for positive association with animals.

Note

1. Correspondence should be sent to Ira Perelle, Department of Psychology, Mercy College, Dobbs Ferry, NY 10522.

References

Barash, D. P. (1977).

Sociobiology and Behavior. New York: Elsevier.

Beck, A. & Katcher, A. (1983).

Between Pets and People. Springfield, IL: Charles C. Thomas.

Blake, D. S. (1980).

On the introduction of pets for the institutionalized aging (Doctoral dissertation, Columbia University Teachers College, 1980). Dissertation Abstracts International, 41, (10) 3732.

Brickel, C. M. (1979).

The therapeutic roles of cat mascots with a hospital- based geriatric population: a staff survey. The Gerontologist, 19, 368-371.

Bustad, L. K. (1990).

Compassion: Our Last Great Hope. Renton, WA: The Delta Society.

Bustad, L. K. & Hines, L. (1984).

Historical perspectives of the human-animal bond. In R. K. Anderson, B. L. Hart, & L. A. Hart (Eds.), The Pet Connection: Its Influence on our Health and Quality of Life (pp. 15-29). Minneapolis, MN: Univ. of Minnesota Press.

Corson, S. & Corson, E. O. (1981).

Companion animals as bonding catalysts in geriatric institutions. In B. Fogel (Ed.), Interactions between people and pets. Springfield, IL: Charles C. Thomas.

Friedman, E., Katcher, A. H., Lynch, J. J., & Thomas, S. A. (1979).

Pet ownership and survival after coronary heart disease. Toronto, Canada: Second Canadian Symposium on Pets and Society.

George, M. H. (1988).

Child therapy and animals. In C. E. Schaefer (Ed.), Innovative interventions in child and adolescent therapy. New York: John Wiley & Sons.

Lee, D. (1984).

Companion Animals in institutions. In P. Arkow (Ed.), Dynamic relations in practice. Animals in the helping professions (pp. 229- 236). Alaceda, CA: Latham Foundation.

Levinson, B. (1962).

The dog as "co-therapist." Mental Hygiene, 46, 59-65.

Levinson, B. (1969).

Pets and old age. Mental Hygiene, 53, 3, 364-368.

Levinson, B. (1975).

Pets and environment. In Anderson, R. S. (Ed.), Pet animals and society. London: Bailliere Tindall, 8-18.

Lorenz, K. (1965).

Evolution and modification of behavior. Chicago: Univ. of Chicago Press.

McCulloch, M. J. (1983).

Animal facilitated therapy: overview and future directions. In A. Katcher & A. Beck (Eds.), New perspectives on our lives with companion animals. Philadelphia: Univ. of Pennsylvania Press.

Messent, P. R. (1984).

Correlates and effects of pet ownership. In R. K. Anderson, B. L. Hart, & L. A. Hart (Eds.), The pet connection: Its influence on our health and quality of life (pp. 331-339). Minneapolis: Univ. of Minnesota Press.

New York Times (1984).

Therapy with pets gains in state. July 22, 1984, p. 38.

Salmon, I. M., Hogarth-Scott, R. S., & Lavelle, R. B. (1982).

A dog in residence, the JACOPIS study. In The Latham Letter. Alaceda, CA: The Latham Foundation.

Trivers, Robert (1985).

Social evolution. Menlo Park, CA: The Benjamin/Cummings Publishing Company, Inc.

 

For abstracts of all issues, including the most current, click Article Abstracts.

To order Society & Animals Journal, go to our secure online ordering page

You can Search the online issues of Society & Animals, as well as the entire Society & Animals Forum (formerly PSYETA) website,
for topics and keywords of your interest:

Google

Search Our Site

 

 
Society&Animals Forum
Violence Link
Animals in the Classroom
Publications
Resources & Educational Material
About
How You Can Help