Current issues of ACP Journal Club are published in Annals of Internal Medicine


Service dogs improved the lives of persons with ambulatory disabilities

ACP J Club. 1996 Sept-Oct;125:33. doi:10.7326/ACPJC-1996-125-2-033

Source Citation

Allen K, Blascovich J. The value of service dogs for people with severe ambulatory disabilities. A randomized controlled trial. JAMA. 1996 Apr 3;275:1001-6.



To determine the role of service dogs in the lives of persons with ambulatory disabilities.


2-year randomized controlled trial with a split-plot, factorial design.


Community study in New York, Pennsylvania, Massachusetts, and Connecticut, USA.


48 persons (mean age 25 y, 50% men) who had been wheelchair mobile for ≥ 2 years. In addition to ambulatory motor impairment, disabilities included quadriplegia, large muscle atrophy, lack of small muscle coordination, aphasia, and attention and memory disorders.


Participants were matched for age, sex, marital status, race, and disability severity to create 24 pairs; within each pair participants were allocated to an experimental group or to a wait-list control group. Each participant in the experimental group received a service dog 1 month after the study began, and each participant in the control group received a dog 13 months after the study began.

Main outcome measures

Participants were assessed by mailed questionnaires at 0, 6, 12, 18, and 24 months. Standardized instruments were used to assess psychological status (self-esteem, psychological well-being, and internal locus of control), community integration, paid and unpaid assistance, school attendance, part-time employment, marital status, and living arrangements.

Main results

In an analysis comparing the experimental and control groups at 0, 6, and 12 months, positive changes occurred in participants with dogs for all outcomes (P < 0.001 for all comparisons) except for marital status and living arrangements. At 12 months, 18 participants (75%) with dogs were attending school and 14 (58%) had part-time employment; participants without dogs for the first year did not attend school or gain employment. Furthermore, on a biweekly basis, participants with dogs had a mean of 28 paid and 14.8 unpaid assistance hours compared with 84.2 paid and 39.8 unpaid assistance hours among participants without dogs. The participants who received dogs 13 months after the study began showed the same rate of improvement in outcomes as participants who had received their dogs 1 month after the study began. Marital status and living arrangements improved in the 2 groups but did not reach statistical significance.


Psychological and social status improvements occurred in the lives of persons with ambulatory disabilities after receiving a service dog.

Source of funding: Not stated.

For article reprint: Dr. K. Allen, Department of Medicine, Millard Fillmore Hospital, 3 Gates Circle, Buffalo, NY 14209, USA. FAX 716-887-5186.


Allen and Blascovich make a strong case for the use of service dogs to assist the physically disabled in activities of daily living. Although the use of service dogs has been repeatedly advocated, no scientific studies had been done of their effect in this population.

The criteria for pairing a dog with a disabled person in terms of the level of disability were not indicated. The 24 pairs were randomly allocated to treatment groups; however, it was impossible to blind the patients to treatment assignment. These patients, by subsequently completing the questionnaires, judged the treatment outcomes. No indication was made in the study of the main decision point (6, 12, 18, or 24 mo), nor was any predetermination made of the sample size or the primary outcome associated with the main study hypothesis. These may be moot points because statistically significant improvements were obtained for all but 2 outcomes.

Notwithstanding its minor shortcomings, this study shows important benefits after 12 months of pairing a service dog with a physically disabled person. The improvements in psychological status in the experimental group occurred as early as 6 months, peaked at 12 months, and were stable at 18 and 24 months. Equally important, persons in the control group who received their service dogs half way through the study showed similar levels of improvement for all outcomes after 1 year with the service dogs as did those who received their dogs at the start of the study.

The authors are to be congratulated for their contribution to an area of research that has remained problematic and filled with anecdotal reports.

Antoine Helewa, MSc
University of Western OntarioLondon, Ontario, Canada