Reduction in Restraint and Critical Incidents in a Norwegian Residential Treatment Facility for Children Aged 7–13 Following the Implementation of the Neurosequential Model of Therapeutics

Abstract

In child and adolescent inpatient, residential and day-treatment facilities, the use of physical restraints and the occurrence of critical incidents are a significant problem. Restraints may sometimes be necessary if a child exhibits dangerous aggressive behavior, but may also be misused or overused, and have been shown to be preventable in many cases. This study aims to investigate if the implementation of the Neurosequential Model of Therapeutics (NMT) has an effect on the annual number of physical restraint incidents in a residential treatment facility for children with complex mental health disorders. Data before and after NMT implementation were collected from the agency’s restraint records. The results showed that post-NMT implementation, there was a substantial and sustained reduction in restraint incidents, with a Cohen’s d value of 2.03, indicating a very large effect. Limiting restraint use in treating children with complex mental health disorders can foster a safer and more therapeutic environment, with potential improvements in treatment outcomes. This study demonstrated a substantial drop in restraint incidents following the implementation of the Neurosequential Model in a residential facility for children aged 7–13. This reduction is presumably pivotal for children with complex mental health disorders, making treatment less coercive and offering promise for settings in which restraint incidents are a concern.

1. Introduction

In child and adolescent inpatient, residential and day-treatment facilities, the use of physical restraints and the occurrence of critical incidents are a significant problem [1]. Children and adolescents referred to such treatment facilities often exhibit impulsive, aggressive, and explosive behavior that can create a sense of unsafety [2]. Restraints are viewed as a security measure to protect both patients and staff in situations of aggression [3]. Restraints may be necessary if a child exhibits aggressive behavior in an unsafe way but are also noted to have been misused and overused according to the U.S. General Accountability Office (1998) [3] and to be largely preventable [4]. When using the term “restraint” in this article, we refer to ‘…the use of physical contact which is intended to prevent, restrict, or subdue the natural movement of any part of the patient’s body’ [5]. According to Norwegian law, only restraint and the isolated use of short-acting drugs for sedative or anesthetic purposes are allowed as a coercive method for children under the age of 16. Importantly, sedative drugs are rarely used in such contexts. With adults and youth older than 16, one may also use mechanical restraints, i.e., a track bed, or place them behind a locked door. Notably, the law clearly states that for staff to use restraints, a child must be in danger of seriously harming themselves or others or doing serious damage to property [6].

When it comes to the treatment of children, there will always be a fine line and enmeshment between treatment and care. Children and youth in mental health facilities are both in need of treatment, but also of basic care. This means that there, under optimal circumstances, is a smooth transition between setting boundaries and protecting the children and the treatment process. When working with children, there is sometimes a need for a physical intervention, like a “turn and guide”, that is not a physical restraint. A restraint aims to stop a child from harming themselves, others, or property by means of force. When restraint is used, the child loses his/her autonomy, whereas a “turn and guide” is a simple redirection of the child.

The use of restraints in inpatient mental health services for children poses several kinds of potential harm to both the children and the staff involved. In their systematic review, Nielson, Bray, Carter, and Kiernan [7] showed that the use of restraints is not only linked to an increased risk of physical harm to the child, but also to an increased risk of death. The potential psychological harm associated with restraints is less researched. Research by Steckley [8] shows that the use of restraints may provide both physical and psychological containment. However, research by Nyttingnes, Ruud, Norvoll, Rugkasa, and Hanssen-Bauer [9] has showed that youth who have experienced restraints have less confidence and trust in their parents and in the staff. By reducing the use of restraints, patients may benefit more from treatment, as the use of force and restraints is a disruption in the treatment and the therapeutic relationship that milieu therapists try to establish with the children.

The lack of statistics and registration of the use of restraints in children’s mental health facilities also poses a problem for developing clear pathways to prevent the prevalence of restraints. In Norway, as in many other countries, there is no official registration of the occurrence of the use of restraints. As such, it is impossible to know how many children in mental health facilities are exposed to this as part of their treatment experience. However, restraints are common in inpatient treatment facilities, and there are risks associated with both physical safety and unfortunate psychological effects [7].

Some 65,000 children are treated in mental health facilities in Norway, most of whom are in outpatient clinics, while about 4% are in residential treatment [10]. Children who are patients in residential treatment facilities present with a complex picture of mental health disorders, and as such, are more prone to aggressive and acting-out behaviors, behaviors that may be perceived as a risk to other patients and staff. As Hambrick et al. [1] points out, the threshold for the use of restraints is lowered when the staff is overwhelmed or frightened by the behavior of the patients. To reduce the number of restraints, significant efforts have been made, and strategies that both focus on strong leadership coupled with staff training and preventive interventions show the most promising outcomes [2,11]. Moreover, research by Hambrick et al. [1] showed a significant reduction in the use of restraints and critical incidents following the implementation of the Neurosequential Model of Therapeutics—NMT [1]—in residential and day-treatment facilities for children and youth. And as Paterson, Bradley, Stark, Sadler, Leadbetter, and Allen [12] highlight, there is a need to look into systematically gathered evidence to enhance safety and quality in health and social care.

The Neurosequential Model of Therapeutics (NMT) is a relatively recent approach to clinical problem solving for treating children and adolescents with complicated developmental challenges and/or mental health disorders, especially those who have experienced trauma and adversity. Contrary to traditional therapeutic models, NMT is not defined by a particular therapeutic approach or intervention, but rather is a neurodevelopmentally informed, biologically respectful framework created to assist the clinician in comprehending the therapeutic, relational, and developmental needs of the patient and planning treatment interventions accordingly, increasing the focus on individualized treatment interventions catered to the patient’s needs [13]. The treatment is planned by utilizing a transdiagnostic approach rather than targeting a single clinical problem or a patient’s functioning across multiple domains [14,15].

In the context of a residential mental health facility for children in Norway, we wanted to investigate if the implementation of NMT would impact the number of incidents of restraints. The patients at this residential treatment center were children aged 7–13, with an average treatment length of 18–24 months. The staff consisted of trained milieu therapists with a three-year Bachelor’s degree in child welfare education. The patients are placed in one of three different units, each consisting of eight patients and ten milieu therapists, and led by an experienced child psychologist responsible for each patient’s treatment plan. The center is closed on weekends and holidays, so all patients also have a home base. The children typically spend two nights per week at the center; all other days, they are returned to their home base before nighttime. All patients attend the onsite school, which operated exclusively for the patients at the center. Every day at the center has the same structure. The mornings are spent at the treatment units, after which the patients attend school until lunch, which is served at the units, and return to school for the afternoon sessions. The remainder of the day is spent at the treatment center with milieu staff. As this is a treatment facility, the competence of the staff, the employment of a child psychologist, etc., are not necessarily the same as in child care institutions.

Aim of the Study

The aim of this study is to answer the following research question: Is there a reduction in the annual number of physical restraint incidents after the implementation of the Neurosequential model at the facility as compared to previous years?

2. Materials and Methods

2.1. Theoretical Basis of the Program for Prevention and Management of Critical Behavior

For many years prior to implementation of NMT, the institution has consistently emphasized the prevention, mitigation, and compassionate management of aggressive and acting out behavior while safeguarding the well-being of the children involved. The specialized program for this purpose is referred to as “Managing Challenging Behavior—MCB”. This program includes the training of milieu staff in understanding the children’s needs and their behavior and strategies for preventing and mitigating aggressive behavior. Ultimately, this program also addresses the delicate task of applying restraint in the gentlest manner possible. The program is deliberately referring to “challenging behavior” and not to violence or aggressive behavior as one wants to encompass a wider array of behaviors that may disrupt treatment interactions with a relational focus, acknowledging that staff might find both types of behavior challenging. This includes acting-out behavior, such as spitting, hitting, kicking, property destruction, and name calling, as well as internalizing behavior, such as withdrawing from contact, the rejection of relationships, dissociation, and daydreaming.

With the implementation of NMT, core concepts from this model were included in the MCB program. This means the MCB program was altered to encompass core concepts from the Neurosequential model. Because of this, the MCB program before and after implementation of NMT differed. This was to enhance the staff’s understanding of the emotional, social, and behavioral problems that are common in children with complex mental health problems and to shift the focus from “what is wrong with you” to “what has happened to you” [1]. This is an important shift to avoid co-dysregulation, which increases the risk of escalation and in turn leads to increased risk of restraint incidents. As Hambrick et al. [1] states in their article showing a significant reduction in use of restraints after implementation of NMT, “Training in NMT emphasizes aspects of relational ‘contagion’ that will dramatically influence interpersonal interactions with dysregulated children and youth in both positive and negative ways. A focus on creating proactive regulating interactions and environments helps staff move away from managing each behavior reactively. Such individual changes in staff practices parallel organizational shifts in policies and program elements related to the NMT. This results in improved conditions for both the children and the staff”.

2.2. Procedures

The patients at the residential treatment center were all referred from outpatient mental health clinics in Eastern Norway. All patients are between 7 and 13 years of age. They present a complex array of diagnoses and symptoms, and all have tried and failed outpatient treatment. The treatment center is privately owned by a non-profit organization and funded by the regional public healthcare organization. The implementation of NMT was chosen by the leadership of the treatment facility as a promising approach to improve treatment processes and outcomes. This was not motivated by the issue of reducing restraints per se. The restraints records which are the basis for looking into changes in use of restraints are mandatory in order to document the use of restraints according to Norwegian law for treatment facilities such as this. The study was classified as mandated quality assurance in accordance with the Norwegian Specialist Health Services Act and approved by the Data Protection Officer at Østbytunet on 3 December 2019. All caretakers gave their written informed consent before any program evaluation data were collected about their child. The children were too young to give consent according to Norwegian law, and since the restraint records are mandatory according to the law, no consent was sought from the children.

2.3. Participants

The participants were a total of 80 children aged 7–13 years (mean age of 10.6 years), composed of 5% girls and 95% boys, receiving long-term residential treatment. Data were collected between 2009 and 2023 and apply to all children enrolled in that timespan. All patients that at some point were subjected to physical restraint were noted in a coercion record. During the period between 2015 and 2020, the residential center began implementing the Neurosequential Model of Therapeutics (NMT) across the organization. Implementation of NMT began in 2015, and most of the staff were trained in and implementing NMT (e.g., communicating with other staff about client care using NMT frameworks, treatment setting within the NMT framework, etc.) beginning in 2016. The reason for the unbalanced girl-to-boy ratio in this sample is that the vast majority of patients subjected to restraints are boys. This is both because there are more boys referred to the facility for showing extensively disruptive behavior, which is a core reason for referral to the facility, and because more boys show disruptive behavior assessed to put them or others in danger while undergoing treatment, with restraints being used on them one or more times during the treatment period.

2.4. Therapists and Training

Since 2009, the front-line staff have undergone a course program in the MCB within their first months after starting work. Every two years, there was a “refresher” course. The MCB program before NMT implementation consisted of training in understanding “the curve of activation”, i.e., being able to read children for signs of overactivation and take measures to decrease overactivation, as well as factors in their milieu that may increase their level of activation, like sensory signals, broken furniture, etc. The MCB program pre-NMT focused on how to understand the triggering factors and the function of these in activated situations. It did not take a holistic approach to the whole functioning of and treatment process for each child.

When the implementation of the NMT began, most front-line staff were highly experienced therapists who had received training in relationally oriented milieu therapy. All units were supervised by a lead psychologist who was trained in the Neurosequential Model of Therapeutics and who supervised the milieu therapist in the NMT every morning during the week. The progress of each child throughout the treatment and the staff’s adherence to the model were evaluated once a week. The content of the MBC was changed so that it focused on the core concepts of NMT. The front-line staff go through mandatory MBC training when they are first hired. The training consists of both a 3 h training session on core concepts and a 2 h training session on how to restrain children in the least invasive manner possible. Every week, the front-line staff also have a 1 h training session on MBC, both on core concepts and the use of physical restraint. After NMT implementation, there was also a cross-unit training session and discussion approximately every two weeks. Whenever a restraint has been used, the incidence is evaluated according to the following question: “is there anything we could have done differently to avoid using restraint?”

2.5. Statistical Analysis

For analyzing changes in the use of restraint across the study period, we applied ordinary least squares regression analyses to identify both overall growth trajectories and individual trajectories for the different treatment units. We examined the explanatory power and fit of linear, quadratic, and cubic models at both overall and unit levels. To test the significance and magnitude changes in use of restraint before and after implementation of NMT, we then conducted an independent samples t-test comparing overall numbers from the years prior to NMT implementation compared to corresponding numbers in the years after (2009–2015 vs. 2016–2023). For establishing the magnitude of change, the effect size (Cohen’s d) was calculated by dividing estimated overall change in the use of restraints pre- and post-NMT by corresponding pooled standard deviations. Cohen’s [16] standards for evaluating the magnitude of effects were used, i.e., small effects were classified as d = 0.2–0.5, medium effects were classified as d = 0.5–0.8, and large effects were classified as d ≥ 0.8. All statistical analyses were conducted with IBM SPSS, version 28.0.

3. Results

3.1. Changes in the Use of Restraint Throughout the Study Period

The results of the ordinary least squares regression analyses showed that both the overall use of restraint and the use of restraint at each unit were systematically reduced over time. In all cases, a cubic model fitted the data best; thus, the use of restraint appeared to be relatively stable until 2015, before being substantially reduced in the following years and then stabilizing at a new lower level. Figure 1 depicts the observed scores and estimated developmental trajectories overall (Panel A) and at the unit levels (Panel B).

3.2. Testing the Difference in the Use of Restraint Before and After NMT Implementation

Looking more specifically at the change in the mean levels of the use of restraint before and after NMT implementation, the independent samples t-test directly demonstrated substantial reductions in the years after the NMT was implemented. The results of this analysis are presented in

Table 1. Results of independent samples t-test comparing use of restraint pre- and post-NMT.

Treatment	N (years)	Mean	SD	t	df	p	d
Pre-NMT	7	41.71	13.31
NMT	8	19.13	8.85	−3.92	13	0.002	2.03

Note: N (years) = the number of years with data reported pre- and post-NMT implementation, SD = Standard deviation, t = t-value for the difference in scores, df = degrees of freedom, p = p-value, d = Cohen’s d.

. The analyses demonstrated statistically significant and substantial reductions, with the mean changing from an average of 41.71 restraints per year at the center pre-NMT to an average of 19.12 restraints per year with the NMT, i.e., a reduction of more than 50%. The effects size of this difference was 2.03, indicating a very large effect associated with NMT implementation.

4. Discussion

The aim of this study was to answer the following research question: Is there an effect on the annual number of restraints after the implementation of NMT?

4.1. Summary of Main Findings

Reducing restraint in treatment of children with complex mental health disorders is of vital importance, as restraint can cause trauma. The results indicate that the implementation of the NMT reduced the use of restraints in a residential treatment facility for children in Norway. We note that the children admitted to the treatment facility before and after the implementation of the NMT did not differ in their levels of intake severity, i.e., data from the CBCL demonstrated no significant differences in symptom levels on admittance when comparing patients pre- and post-NMT implementation (with a mean of 72.06 vs. 73.36 on the CBCL in the two groups, respectively). For further details, see also [14]. There were differences regarding the baseline and outcome points at the three different units. This was mainly caused by some children being restrained more than others.

In 2020, there was a big drop in the number of restraints. This coincides with the COVID-19 pandemic, during which the institution due to disease prevention changed its practice to arena flexible treatment and thus reduced some triggers for acting-out behavior, as children were together less frequently and there was more emphasis on helping families cope with the impact of the societal implications of the COVID-19 lock-down, etc.

This study has demonstrated a significant and highly substantial reduction in the number of restraints during and after the period in which the NMT was implemented as compared to before. As noted, the impact of implementing the NMT on the reduction in the use of coercive measures yielded an effect size (Cohen’s d) of 2.03, constituting what is commonly operationalized as a very large effect. Even though no causal inference is possible based on our naturalistic data, the results do indicate that the implementation of the NMT may have had a powerful and positive influence on the reduction in the use of restraints, as this was the only substantial change made in the treatment process during this period.

4.2. Clinical Implications

Reducing the use of restraints can contribute to a safer and more therapeutic treatment environment, leading to improved treatment outcomes for children. Children may experience fewer potentially traumatic or triggering events, thus gaining better access to therapeutic interventions. The substantial changes before and after NMT implementation indicate that this approach can be highly beneficial for the institution and the patients it treats. However, it is important to remember that other factors may also influence treatment outcomes.

Clearly, reducing the number of restraint incidents is a goal in and of itself when implementing a new treatment and intervention model. This is because experiencing restraint and physical intervention is likely to be counterproductive to the treatment children receive. By reducing the use of restraint, patients may benefit more from treatment, as the use of force and restraints constitutes a disruption in their treatment and the therapeutic relationship that milieu therapists try to establish with them. The power differential concept within the NMT is important for understanding the reduction in restraint incidents. It emphasizes the de facto power imbalance that exists between children and adult milieu therapists. This awareness is vital in avoiding the use of force based on the adult’s physical size or perceived ability to exert power. It also addresses how the power imbalance itself can be intimidating and trigger negative, aggressive behavior from the child.

4.3. Study Strengths and Limitations

The sample size in this study consisted of 80 children aged 7–13 receiving treatment in a Norwegian institution for complex mental disorders. The sample size is an important factor to consider when interpreting the study’s findings. A larger sample size typically provides more representative results, but it is also essential to assess whether this size is sufficient to detect the effects of NMT implementation.

The method of data collection involved a review of registered restraint incidents from 2009 to 2023. The data were collected from the institution’s own restraint records. This approach is reliable and valid if there are standardized procedures for registration and documentation and if the staff follows them. However, there might be room for reporting errors or the underreporting of incidents.

The data were collected from the restraint records of the institution both before and during the implementation of the NMT. The so-called “Hawthorne effect” [17], a change in the occurrence of the use of restraints, may be an effect of the implementation process, with the increased focus on the professional understanding and handling of the treatment process possibly being one reason for the reduction in the use of restraints. However, the reason for implementing the NMT was not to reduce restraints per se, but to see if an increase in the effect of the treatment process was possible. A reduction in restraint was therefore not the focus during the implementation process for the miliu staff.

One major limitation of the study is the absence of randomization and a controlled design. The absence of a randomized design and a control group makes it challenging to determine whether the observed reduction in restraint incidents can be solely attributed to the implementation of the NMT or if other external factors, e.g., the increased focus on the treatment process, could have contributed to the decrease in critical incidents. Without this kind of design, it is also impossible to establish a causal relationship between the implementation and delivery of the NMT and the documented reduction in restraints in a scientifically convincing way.

4.4. Recommendations for Future Research

Future research should consider the possibility of using a randomized, controlled design and to clarify whether other factors besides the implementation of the NMT are contributing to the reduction in restraint incidents. Future research should also assess the accuracy of restraint incident reporting, ensuring that it complies with legal regulations and current guidelines to minimize potential reporting biases. Future researchers should also look into other age groups, as the participants in this study were in the age range of 7–13 years. Older children or youth may be subjected to more force when restrained due to their body size but may also be prone to a higher level of reflection due to their cognitive maturity and therefore may be less subjected to restraint in the first place. The impact of the Neurosequential Model on restraints in other age groups is therefore in need of investigation. It is also necessary to look into if there are differences between girls and boys, across ages, or in regards to other demographics. Finally, future studies should investigate whether treatment effectiveness improves when the number of restraint incidents decreases.

5. Conclusions

This study demonstrates a statistically significant and very large reduction in the use of restraints following the implementation of the NMT in a residential treatment facility for children aged 7–13. Reducing the use of restraints is crucial in the treatment of children with complex mental disorders, as this group of patients is particularly vulnerable. In conclusion, the NMT appears to hold significant promise in terms of making the experience of treatment less coercive for children in treatment settings in which the use of restraints is a factor.