Fearful Dogs in Shelters Can Have A Brighter Future With Evidence-based Interventions
Summary: An interview with two of the authors of the ASPCA’s latest research paper, “Behavioral rehabilitation of extremely fearful dogs: Report on the efficacy of a treatment protocol.”
With a combination of sufficient practical resources and a knowledgeable team of staff and volunteers, shelters can be fertile grounds for research into behavior modification protocols. Being based in a shelter can allow researchers access to a population of dogs who share a living situation and who are being handled by the same people.
The ASPCA has a strong research program and has recently published a study on the rehabilitation of fearful dogs at their shelter. Here’s the study, published in Applied Animal Behaviour Science.1
About the study
The study aimed to evaluate whether a protocol designed by ASPCA was effective in reducing fear-based behaviors in dogs who had been evaluated as having a significant enough problem to make them unlikely to succeed in an adoptive home. There were 441 dogs in the study, and all dogs were given the behavior modification protocol treatment – one group was started on the protocol after their first behavior evaluation, given after four days of acclimation to the shelter, and the other was started after their second behavior evaluations, approximately 25 days after their entry to the shelter. The difference in timing was done to test whether it was the protocol, rather than simply the length of time at the shelter, that was driving any change in behavior.
The protocol approached three areas of behavior modification: handling, interactions with people, and leash walking. In addition, the dogs received positive reinforcement training focused on functional behaviors for riding in a car and habituation to a crate.
“Behavioral Rehabilitation Center (BRC) staff employed several tools to reinforce desired behavior and to achieve desensitization and counterconditioning goals. Social interaction with other dogs, including play, was reserved for behavior modification treatment sessions, where it was used to encourage socially facilitated exploration and interaction with people and to teach dogs to associate exposure to people with pleasurable access to conspecifics. In fact, all treatment sessions were conducted with at least one other dog—a less fearful “helper dog”—until the final stages of the program, when dogs learned to engage in sessions with only people present. Favored in-kennel enrichment items were also incorporated into sessions to stimulate investigatory behavior (e.g., scents planted along walking routes), encourage interactive object play (toys), and create positive associations (e.g., edible chews and food puzzle toys during time in a crate). All meals, which consisted of kibble mixed with canned food and/or high-value treats, were rationed out at the beginning of the day and then provided to each dog during his or her treatment session. (On rest days, dogs received plain kibble in a food puzzle toy.)”
Medication was also used during the dogs’ time in treatment. Every dog received a regiment of fluoxetine and gabapentin when they started behavior modification protocol, under the supervision of a veterinary behaviorist.
The ASPCA website2 lists the following outcomes of the study:
- 86% of the dogs graduated from the program and were made available for adoption.
- Adopter surveys indicated overall satisfaction with adopted dogs.
- Age, sex, breed, and history did not predict behavior grade, improvement, graduation, or adoption.
- Successful treatment required an average of 78 behavior modification sessions over 96 days.
- Non-graduating dogs tended to have more severe fear on intake and less improvement as measured by behavior assessments.
The study suggests that it is possible for dogs to make significant progress on overcoming their fears even while in a shelter environment. Not every shelter is going to be able to dedicate the time, the personnel, or the administrative overhead to testing interventions on the dogs in their care, but every shelter can learn from research on what works and what doesn’t, and adapt the protocols outlined in research to their own circumstances.
We took the opportunity to interview Kristen Collins, MS, ACAAB, vice president of ASPCA Rehabilitation Services (KC), and Emily Patterson-Kane, PhD, director of research at ASPCA (EPK), and took questions from some of our members who work in and with shelters.
Were there any factors about the dogs that predicted whether they would succeed, other than their overall fear level? For example, did the researchers find that dogs who participated and played more with their enrichment along with behavior modification were on a faster trajectory to behavioral rehabilitation and therefore adoption?
EPK: This manuscript focused on high-level indicators of success that applied across the population. In this paper we choose to focus on three indicators: grades in a periodic behavior evaluation (scored a-d as shown in the appendix of the paper), graduation from the program, and adoption (including adopter satisfaction). There is a great deal more information available, such as observations made during treatments, detailed demographics for each animal, and progress in different areas rather than just an overall rating.
KC: It will be fascinating to discover whether we can identify those predictors of success — but from a purely anecdotal perspective, as someone involved in the hands-on work from the beginning, I found myself repeatedly struck by the inaccuracy of my own predictions. Many of the dogs I thought least likely to make significant progress (e.g., older dogs, dogs showing no social behavior initially, more severe fearful behavior) surprised us and did.
Were dogs assigned specific handlers? If so, was there a difference in success rates and data within certain individual handlers that dogs responded better toward and were less fearful with over time?
KC: Because of staffing reasons and because we wanted the dogs to associate people, not just one person, with good things (food, access to other dogs, enrichment items) as quickly as possible, we made the decision not to assign dogs to a single behavior rehabilitation specialist early on. We continue to see success with a few different specialists working each dog from the start. Very rarely, if a dog isn’t making much progress (or isn’t making progress at all), we’ll decide to limit interactions to a single specialist for a brief period and then go back to incorporating exposure to others as soon as we see a reduction in fear. Exploring differences in those dogs’ responses to treatment wasn’t a research goal of ours — and the situation was so uncommon that we wouldn’t have enough data to analyze anyway.
Were there any side effects to starting all dogs on psychopharmaceuticals?
KC: Happily, the [study] dogs haven’t shown any noticeable signs of experiencing medication side effects, no.
Why did you choose to include breed data? If you had access to more reliable breed information (such as through genetic testing) do you think it would have made a difference?
EPK: It seemed likely that many readers would be interested in breed data to understand more about the dogs in the program. For this particular data it was important to demonstrate that they were a diverse group and that breed was not strongly associated with the findings we reported here. This finding supports the current approach of focusing on the individual dog’s behavior and not breed or age for decisions such as intake criteria and treatment plans. That said, information such as breed and age are generally not precisely known for many dogs in the shelter population, and this analysis was at a high level. More subtle associations may be revealed on future analysis and assist with further refining the program.
There is no plan currently to carry out genetic testing. However, if we find evidence suggesting this might benefit the program and the dogs, and an appropriate partner can be found, it would be a possible avenue for further investigation — not only as another indicator for breed or type, but other potential markers of biological contributors to trauma, resilience, and rehabilitation.
Were the results for the program dogs in the post-adoption satisfaction survey significantly different from adopters of non-program dogs?
EPK: Our follow-up surveys for dogs from different programs are not directly comparable and so it is not easy to judge. However, there are indicators that BRC graduate dogs are more likely to be reported as still showing some fear behavior, but owner satisfaction is even higher than found for other populations. So overall, it appears that these dogs are very successful as companion animals.
How might shelters with less money or other pertinent resources (like space or staffing) implement a program like the one the paper describes?
KC: The short, frank answer is that many organizations can’t implement a comprehensive program like ours — one that requires significant resources they simply don’t have. We conducted the pilot in a “regular” animal shelter, not in a purpose-built facility designed to reduce stress and facilitate recovery. That’s excellent news because we can say the program worked well for most animals without the benefit of a strictly controlled, consistently calm environment — but you do need some relatively quiet indoor and outdoor treatment spaces to perform behavior modification protocols. Those areas should also allow you to expose undersocialized dogs to sights, sounds, and smells they’ll experience in pet homes. Capacity for care and extending length of stay for treatment pose challenges, too, for virtually all animal welfare organizations. Finally, to implement an intensive behavior rehabilitation program for severe problems, you need well-qualified shelter animal behavior experts — and we know those are scarce.
Having said that, I think the results of our study suggest, in general, that fearful shelter dogs can often recover if treated. It’s certainly possible to scale a program like ours, shaping it to fit the resources available. For any organization interested in developing or expanding a shelter behavior program, I recommend first outlining those resources so you can then build the structure of your program appropriately, making sure you have all the foundational pieces in place to set staff and animals up for success. Some organizations only have the resources to support behavioral health through strategies like using low-stress handling, providing enrichment, and modifying animals’ individual housing environments. Some connect with animal behavior experts in their communities or develop specialized volunteer programs that focus on training and basic behavior modification. Some may be able to treat a limited number of behavior problems, like mild to moderate fear, but don’t have the ability to take on more complex, difficult cases. Some can tackle more severe problems, including those that have a good prognosis but take time to resolve, but may only have the capacity to treat a small number of those cases at a time. Despite resource differences, all organizations can implement core behavior best practices and establish a staff culture that emphasizes the importance of both physical and psychological health. That alone, in my opinion, is a win worth celebrating and one that will reap big benefits for shelter animals.
Could you list a few key elements of the core behavior best practices for staff and volunteers?
KC: In my opinion, the most critical core shelter behavior best practices include low-stress handling training for all staff, the provision of in-kennel enrichment, and housing choices/modifications that minimize stress (e.g., cardboard box hiding spots for fearful cats), teaching staff to use objective language to communicate behavior observations, and the inclusion of psychological health as a consideration during pathway planning for each animal.
How does the author see the program translating into post-adoption life for the dogs? What kind of support did the dogs get after they had been adopted?
KC: Of course, all dogs are individuals, and some are more adaptable and resilient than others. We expect program graduates to regress a little at first (and our data show that), but new adopters should still be able to handle and manage the dogs during their initial adjustment period. The grads should be able to do things like walk on leash, tolerate handling, and learn housetraining. Within days to weeks, we expect grads to settle in and start exhibiting the kinds of behavior we saw while they were in treatment: playing, seeking proximity to people, enjoying petting, exploring, etc. Because we can only expose the dogs to so many experiences during their time with us, we know they’ll encounter novelty in the home. We expect them to react fearfully at first to new things, but we also expect them to recover quickly and, overall, experience not just an acceptable but a good quality of life. Using decision-making tools like our Quality of Life Assessment Tool and Adoptability Guidelines (Appendix 3 in the research paper) helps us make outcome calls with an eye to placing animals who will be able to adjust to new homes and thrive as pets. It’s critical for us to collect follow-up data, research aside, so we can continually assess those outcome decisions.
We place a small percentage of dogs directly from the BRC in North Carolina, but most are placed with partner shelters for adoption. When we transfer grads to shelter partners for placement, we send materials written for shelter staff as well as other materials for adopters, aiming to set the grads and their caregivers up for success. We also provide troubleshooting support, if needed, via phone and email.
If you had no resource limitations, what would follow-up support look like?
It’s tempting to imagine providing robust follow-up engagement and support for dogs and adopters, given no resource limitations. However, if we achieve our treatment goals at the BRC, extensive follow-up support really shouldn’t be necessary in most cases. We aim to “graduate” dogs when we feel confident that they can have a good quality of life in adoptive homes and that they can function as pets (go on walks, tolerate handling/restraint, etc.). Although they may regress a bit initially when they land in their new homes, we expect that they’ll recover relatively quickly and continue to progress as they settle in and explore the world. Having said that, we do offer consultation for adopters who want guidance as their dogs acclimate to their new lives as companion animals — and if we had no resource limitations, it would be awesome to provide that in person, in the home, rather than via phone or email.
Do you have any additional information about the dogs now, after a few months?
KC: We collected the formal follow-up data we used for our study at six weeks post-placement, but we’ve also received some interesting informal feedback from dogs’ adopters. For example, some dogs acclimate to life in homes more slowly that we thought they would or have long-term challenges with something specific, like riding in a car or handling at a vet clinic. Often, however, we learn that dogs are doing exactly what we envisioned: becoming gradually more and more playful and living the lives we hoped they could after treatment. There’s nothing like hearing about dogs settling in and enjoying normal “dog stuff,” like snuggling on couches, going to the beach, hiking, and romping at home with favorite toys and friends. Long-term follow-up like that, considering our dogs’ beginnings, is incredibly gratifying.
How was the writing process for you? Had any of the authors published before?
EPK: The authors have different levels of direct experience with developing the program, the hands-on work, and publishing in peer-reviewed journals. They developed the manuscript as a team to make sure all of the necessary skills and experience were available. The data used were collected to support treatment of the dogs and assess the program for internal processes. Once the data was identified as a potentially important source of generalized knowledge, the staff with direct knowledge (overseeing or delivering the rehabilitation services) recruited me to extract the data and work with them to create the manuscript.
KC: Dr. Pam Reid, Dr. Kat Miller, and I collaborated to design the study. I oversaw the development of our detailed program structure and behavior modification protocols, as well as their implementation and data collection. I worked closely with Emily P-K to edit and write portions of the manuscript. All authors listed reviewed the manuscript drafts at multiple stages and provided input and editing suggestions.
Do you have plans to research anything else on this topic?
EPK: There are two planned future papers. One of these will look at a much wider range of variables looking for predictors of successful graduation from the program. The current paper is documenting that this course of treatment does significantly improve the behavior of the entire population. When the question is “what makes some dogs respond better to the treatment than others?” — many more factors will be considered, as well as the interactions between them.
We have a very active research program dedicated to investigating a range of behavior concerns that impact shelter dogs and cats. In addition to studying our own animals to elevate our ability to improve their quality of life and develop more effective treatment interventions, we collaborate with partner shelters to collect data on a greater variety of animals and we provide grants to other researchers studying topics in shelter animal behavior.
If you have any questions or comments about this research, please let us know in the comments below!