Photo: Nancy Marie Davis | Flickr / Creative Commons

[image: sepia-tone print of a clenched fist, with superimposed scratched lines.]

Maxfield Sparrow


A little over two years ago, Crystal Garrett wrote an article for Thinking Person’s Guide to Autism about the long-term traumatic effects on her Autistic son of the restraints and seclusion used against him at school. Garrett chose to end her career as a journalist to stay at home and school Zachary herself. Garrett wrote,

“We know a restraint and seclusion free environment is realistic. Virginia-based Grafton Integrated Health Network, an organization that works with children and adults with autism and co-occurring psychiatric diagnoses, went restraint and seclusion free ten years ago. Since then, their client and staff injury rate has dramatically gone down, while employee satisfaction has increased. They are now teaching their system, Ukeru, to others across the country, in order to create a trauma-informed environment for addressing aggressive behavior.”

Garrett also wrote about the importance of more people learning about the Ukeru model:

“Perhaps if more educational bodies knew about this approach, my 60-pound six-year-old wouldn’t have ever been placed in the back of the patrol car that hot summer day. We also wouldn’t have to travel the 30-mile stretch of highway for trauma-based therapy each week.”

Over a Decade of Research into Alternatives to Restraint and Seclusion

I couldn’t help thinking about Crystal and Zachary Garrett today when I learned that an official study of Grafton’s Winchester, Virginia, Ukeru program was published in Advances in Neurodevelopmental Disorder in August of 2018.

The study’s authors are Jason H. Craig (a BCBA at Grafton School) and Kimberly L. Sanders, who has filled many roles at Grafton and is currently the Chief Outcomes Officer. In another article earlier this year, Sanders writes about the Ukeru vision as “comfort versus control,” and writes that it’s more than just how people are trained to approach alternatives to restraint and seclusion but a “core operating principle” informing the entire philosophy of the organization. She links to a capsule explanation of The Grafton Method, the seven-point framework for addressing issues of restraint and seclusion with the committed intention to minimize or eliminate restraint entirely:

  • Leadership. Decide the organization’s mission, envision a path between the present and the future, plan actions and keep everyone focused on the goals.
  • Communications. Communicate the vision to staff. Set up communication practices within the organization that will keep everyone informed and on point.
  • Training. Teach everyone the best practices to use instead of restraint and seclusion.
  • Measurement. Keep track of what’s working and what isn’t. Measuring and record-keeping are what keep education scientific rather than intuitive, and documentation is crucial for repeatable results.
  • Debriefing. When an unwanted outcome arises, it’s important to establish clear communication to discuss what happened and how to achieve more desirable results next time.
  • Alternative solutions. At Grafton, brainstorming about alternative solutions was how Ukeru was developed.
  • Therapeutic planning. Don’t leave a student/client’s progress to chance: 
    1. Assess a student’s strengths and weaknesses
    2. Develop strategies around a student’s strengths, encouraging academic growth in an environment of safety and security for the student
    3. Train staff in techniques and documentation methods
    4. Monitor progress
    5. Modify what’s happening, based on progress toward goals (as well as adjusting goals when they no longer suit a client’s preferences and needs).

The study Sanders and Craig published is important because it represents peer-reviewed scientific evidence that eliminating restraints and seclusion not only have the direct benefit of treating clients with dignity and respect, rather than traumatizing them, but the Ukeru program has documented evidence that eliminating restraint saves on workplace injuries caused by client responses—which also saves organizations millions of dollars.

A copy of the Grafton study needs to be in every parent’s toolkit when arguing for the well-being of their children in the education system. Sanders and Craig have produced reputable scientific evidence that serves as valuable munitions for every parent’s IEP battle against a school that uses restraints and seclusion.

In a recent Scientific American article, Alycia Halladay writes about hyped-up autism studies and how harmful they are to Autistic people, saying,

“The mainstream news media need to consider a more measured and responsible approach to covering autism research. This should include very careful vetting of which studies are reported. Not all scientific journals are equal in their scientific rigor or review policies, so just because a study is ‘published’ does not mean it necessarily has scientific significance.” 

Advances in Neurodevelopmental Disorders is a reputable journal, and the Grafton study is crucial to developing appropriate methods for schools to interact with Autistic students, but I haven’t seen much, if any, journalistic hype for this study. I guess there’s nothing sensationalistic and headline-grabbing about NOT harming students.

Why the Grafton Study is Useful and Important

The Grafton study provides clear data useful when debating against the many people who insist that there is no way to get around using restraint against “violent people” because there is no other way to control dangerous behavior. These people often add statements about such harsh treatment such as saying that it is done “for the client’s own good” adding that people will have better future outcomes if we control their behavior now, and shape them into something socially acceptable.

As an Autistic advocate/activist myself, I also get told that I can’t understand the needs and benefits for Autistic people who “aren’t like [me]” because I’m allegedly only “mildly affected” and can’t understand the needs of “serious autism.” And if you’ve ever read anything else I’ve written, I don’t need to tell you how frustrating and infuriating that kind of talk is to me.

Humor me while I unpack those representative statements a little: There aren’t Autistic people who “aren’t like me” in a real sense, because we’re all different (being, you know, human beings) and no one on this planet is actually “like me” but the Autistic people on this planet are a heck of a lot more like me than the non-autistic people. Yes, I’m even talking about “those people”—you can fill in that blank however you choose. I have more in common with any Autistic person, even those who can do things I can’t, even those who can’t do things I can.

“Mildly affected” and “serious autism” are just euphemisms for functioning labels, and we Autistic folks have written so much already against that misunderstanding of autism. Let me just skip past that this time to tell you that I have a personal history of being restrained, secluded, and institutionalized. I am not speaking from some philosophical stance or set of political beliefs when I say that treating Autistic people with restraint and seclusion make us worse, not better. I am speaking from personal experience. Now, thanks to the Grafton study, we can all speak from a scientific basis as well when we talk about the ways restraint and seclusion make things worse for everyone, not just for the Autistic people these methods are used on.

In the introduction/literature review section of  their paper, Craig and Sanders talk about the 2006 policy paper issued by the National Council for Behavioral Health that said restraint and seclusion should be methods of last resort, used only after all other options have been exhausted. They also cite a 2012 paper from the American Journal of Orthopsychiatry that found restraint and seclusion to be “traumatizing and dangerous to both the children and the staff involved in each incident.” Some agencies have been successful in reducing restraint and seclusion while others struggle with the mandate.

Next, Craig and Sanders talk about the Trauma-informed approach (TIA) model. Instead of focusing on controlling someone’s behavior, TIA focuses on establishing a relationship between the student/patient and a trusted adult caregiver. When the student feels emotionally and physically safe, they will be more invested in accepting help and considering new ideas about their lives. This safe space allows the student to shift from a position focused on bare survival into a position where growth and learning can take place. Grafton’s comfort vs. control perspective grew out of TIA practices. Comfort vs. control aims to change interactions between staff and student such that the risks of traumatizing or re-traumatizing the student are significantly minimized, using the following techniques/philosophies:

  1. Response blocking. The simplest example is putting a pillow between someone’s face and fist to stop them from hitting their face and instead hit the pillow. Response blocking is controversial among some people. Most parents support response blocking because it directly prevents a student from injuring themselves. Response blocking in an ABA context can be either a punishment or an extinction, depending on how the client responds to being blocked.
  2. An understanding of behavioral intent and client needs. In other words, not just deciding what a client should or should not be allowed to do, but rather working to understand why a client chooses a particular action, feedback, etc. What unmet needs are driving the client’s behavior? “Behavior” does not happen in a vacuum; behavior is communication and/or an attempt to meet unmet needs.
  3. Developing creative solutions as alternatives to restraint and seclusion. How can using comforting techniques help clients calm and self-regulate? How can staff take care to not use control tactics on behavior that is disruptive and appears “out of control”?

These three points created a mindset shift among staff, allowing workers to:

  • Reassure clients.
  • Ask questions of clients instead of jumping to inaccurate conclusions and assumptions.
  • Embrace flexibility
  • Do not strive to always keep the upper hand / hold an imbalance of power / maintain control over situations and people.
  • Treat others with kindness and respect.

Sanders and Craig write,

“The belief was that many situations in which a restraint or seclusion was used could be better resolved by a non-coercive, caring intervention from a person focused on peaceful conflict resolution who was willing and able to spend time with the upset or angry individual.” 

I feel like that sentence highlights the biggest reasons why we are still wrestling with the demon of restraint and seclusion, even 14 years after a strong governmental policy recommendation against it: “willing and able to spend time.” So many workers are under such constraints of time and money that TIA feels overwhelming. Staff need to be educated about the time, money, and (most importantly!) human safety and well-being that will be preserved through implementing a TIA approach instead of restraint and seclusion.

How the Grafton Experiment was Conducted

Sanders and Craig examined the entire Grafton system, including over 750 employees and 3244 clients, male and female, from age 6 to age 22, with intellectual and developmental disabilities presenting at a wide variety and levels of support needs. At the beginning of the study, when Grafton began making concerted efforts to reduce restraint and seclusion, any and all of these 3244 clients were potential candidates for restraint and/or seclusion, if they were deemed to be a risk to themselves or others. The restraint reduction program began in 2004 and the year before the program there were 6646 incidents of restraint or seclusion. In other words, a client would be restrained or secluded on average more than twice per year.

The study authors also laid groundwork statistics representing the pre-intervention levels of measurable outcomes, particularly:

  • Rate of incidents of restraint and seclusion.
  • Rate of restraint-induced staff injuries (Why they did not count rate of restraint-induced client injuries is beyond me.)
  • Client-induced staff injuries.
  • Lost time and modified duty for workplace injuries.
  • Lost money due to all of the above causes.
  • Client outcomes / goal mastery

The results of the study came from a combination of carefully conducted interviews, and the raw data related to the above focused outcomes.

Research Outcomes

The researchers found that the Grafton Method reduced instances of restraint by 99%.

They reduced instances of seclusion by 100%.

Restraint-induced injuries to staff were reduced in the same time period by 97%.

Client-induced injuries to staff were reduced by 64%.

(I should note that client-induced injuries is one of the biggest reasons I hear cited for the need for staff at facilities to be able to restrain clients: fear of injury to staff. This result shows that staff were actually being injured MORE by the clients when there was restraint and seclusion and far LESS after restraint and seclusion were removed.)

There was an 81% reduction in lost days and modified schedules after reducing restraint and seclusion.

Lost expenses decreased by 75% overall. Between money lost to sick/injured days, worker’s compensation, employee turnover, and training, in the 12 years covered by the study, Grafton saved approximately $16.4 million (an average savings of $1.37 million per year) by restricting/eliminating restraint and seclusion practices.

As for client goal mastery, when the study started in 2005 the client goal mastery rate was 34%. By 2008 the client goal mastery was 66%. By 2018, that rate was 80%. Overall, this represents a 133% increase in the ability of the students to learn, grow, and change in positive ways desired by both the students and the staff.


Who (except us Autistics and many of our parents!) expected so much positive gain from simply choosing to treat clients with respect and kindness? Craig and Sanders advise other organizations to look at these results and choose to eliminate restraints and seclusion as well. They wrote, “Organizations need to learn to be more responsible and accountable, especially when it comes to the treatment to vulnerable individuals.”

It is my fervent hope that this study from Craig and Sanders will be widely cited and shared in the research community as well as the community of parents of school-aged Autistic children. We all need to get the word out about these results and get other researchers and organizations on board with the TIA approach to helping students feel more comfortable and less manipulated, controlled, bullied, and traumatized.

Everything in this study seems obvious to us, of course. Whether you’re an Autistic person, the parent of an Autistic person, a compassionate teacher looking for better ways to mentor your students, an activist or policy maker—it should seem obvious that life is better for everyone when we aren’t tackling vulnerable people to the ground, punishing, and isolating them for the “crime” of being different and struggling with self-regulation, sensory onslaught, and the countless other daily battles Autistic people have with our environments.

But these things are not obvious to everyone. A study that demonstrates significant savings in money will sway some organizations. A study showing less injury to staff will sway others. This study demonstrates both and more. Add it to your collection and use it as evidence in your fight to make the world a better place for Autistic children, the Autistic adults they will grow into, and everyone whose life touches or is touched by the lives of our precious Autistic children who deserve so much better than our society has been giving them.