Working mechanisms of imagery rescripting (ImRs) in adult patients with childhood-related PTSD: a pilot study

ABSTRACT

Background: Imagery rescripting (ImRs) has shown to be an effective treatment for posttraumatic stress disorders (PTSD) resulting from childhood-related trauma. The current theory is that the change of meaning of the trauma memory is central to the treatment. Several authors have suggested that the expression of needs, feelings and actions may act as potential healing factors, but little specific research aimed at (in)validating this hypothesis has been done so far.

Objective: In this study we investigated to what extent the expression of inhibited action tendencies and the fulfilling of needs lead to the reduction of PTSD symptoms in clients with early childhood trauma.

Method: Recordings of 249 therapy sessions of 24 ImRs treatments were rated with an observation instrument developed for this purpose, after which the scores were related to pre and posttreatment symptoms, assessed with the Impact of Events Scale-Revised (IES-R).

Results: Scores on the IES-R decreased from pretreatment to posttreatment. The two subscales of the NATS (At-scale and N-scale)significantly predicted the posttreatment scores on the IES-R after controlling for the influence of pretreatment IES-R scores: the better the expression of inhibited action tendencies and the better the fulfilling of needs, the lower the symptoms after treatment.

Conclusions: This pilot study on the underlying mechanisms of ImRs in PTSD treatment has shown that the expression of action tendencies and fulfilling basic needs during ImRs are associated with a decrease in PTSD symptoms after treatment, and that actions and basic needs cannot be viewed separately. Follow-up research could focus on which of the six domains of the Needs and Action tendencies Scale (NATS) has the greatest effect on the reduction of PTSD symptoms. With this information we can further improve the ImRs protocol.

HIGHLIGHTS

• The pilot study of working mechanisms of imaginary rescripting shows that the NATS is a reliable research tool for observing expressed action tendencies and fulfilled needs.

• The better the action tendencies are expressed during treatment and the better the needs are fulfilled, the lower posttreatment symptoms.

• It seems useful if practitioners are specifically trained during the ImRs training in performing actions that lead to the fulfilment of basic needs.

Antecedentes: La reescritura de imágenes (ImRs, por sus siglas en inglés) ha demostrado ser un tratamiento eficaz para los trastornos de estrés postraumático (TEPT) resultantes de un trauma relacionado con la infancia. La teoría actual es que el cambio de significado del recuerdo del trauma es fundamental para el tratamiento. Varios autores han sugerido que la expresión de necesidades, sentimientos y acciones pueden actuar como posibles factores curativos, pero hasta ahora se han realizado pocas investigaciones específicas destinadas a (in)validar esta hipótesis.

Objetivo: En este estudio investigamos hasta qué punto la expresión de tendencias de acción inhibidas y la satisfacción de necesidades conducen a la reducción de los síntomas de TEPT en clientes con trauma infantil temprano.

Método: Las grabaciones de 249 sesiones de terapia de 24 tratamientos ImRs se calificaron con un instrumento de observación desarrollado para este propósito, después de lo cual las puntuaciones se relacionaron con los síntomas previos y posteriores al tratamiento, evaluados con la Escala de Impacto de Eventos Revisada (IES-R, por sus siglas en inglés).

Resultados: Las puntuaciones en el IES-R disminuyeron desde el pretratamiento hasta el postratamiento. Las dos subescalas de la NATS (en escala y en escala N) predijeron significativamente las puntuaciones postratamiento en la IES-R después de controlar la influencia de las puntuaciones de la IES-R previas al tratamiento: cuanto mejor era la expresión de las tendencias de acción inhibida y mejor era la satisfacción de necesidades, menores eran los síntomas después del tratamiento.

Conclusiones: Este estudio piloto sobre los mecanismos subyacentes de las ImRs en el tratamiento del TEPT ha demostrado que la expresión de tendencias de acción y la satisfacción de las necesidades básicas durante las ImR se asocian con una disminución de los síntomas del TEPT después del tratamiento, y que las acciones y las necesidades básicas no pueden verse por separado. La investigación de seguimiento podría centrarse en cuál de los seis dominios de la Escala de Necesidades y Tendencias de Acción (NATS, por sus siglas en inglés) tiene el mayor efecto en la reducción de los síntomas de TEPT. Con esta información podremos mejorar aún más el protocolo ImRs.

KEYWORDS:

• Imagery rescripting
• working mechanisms
• PTSD
• needs
• action tendencies

PALABRAS CLAVE:

• Reescritura de imágenes
• mecanismos de trabajo
• TEPT
• necesidades
• tendencias de acción

1. Background

Imagery Rescripting (ImRs) has shown to be an effective treatment for posttraumatic stress disorders (PTSD) resulting from a variety of index traumas (Arntz et al., 2007, 2013; Grunert et al., 2007; Morina et al., 2017) including PTSD resulting from childhood-related trauma (Raabe et al., 2015). In essence, ImRs involves working with images in order to change meanings and ameliorate distress (Hackmann, 2011). In ImRs, the patient imagines different responses to and outcomes of the original traumatic event and its aftermath. A new script might include someone else entering the scene and bringing safety, or the participant having the power to prevent the trauma, or to take revenge on the perpetrators (Hagenaars & Arntz, 2012). Rescripting can help a patient by offering a fresh perspective on events that happened in the past, eliciting new feelings such as anger instead of fear. Rescripting can also identify unmet needs, such as the need for safety, and fulfil these needs as well, for example because in the rescripted situation the patient (as a child) is being brought to safety. During this process, the patient is also confronted with reality (for example, that some experiences are painful and harmful, such as child abuse), so that a healthy mourning process can ensue (Arntz et al., 2007; Holmes et al., 2007).

Although several studies on the efficacy of ImRs have been conducted (Dibbets & Artnz, 2016; Slofstra et al., 2016), research on the underlying mechanisms of ImRs has just started. According to Arntz and Weertman (1999) and Arntz (2012), ImRs alters the meaning of the original event by changing the original responses to the event and the negative associations of the unconditioned stimulus (US) into more positives ones. In terms of learning theory, this could be conceptualized as ‘US revaluation’, thereby changing the fear network (CS-US associations; Arntz, 2014b). By using imagery, the change in meaning not only takes place on a verbal cognitive level, but also on sensory, emotional and behavioural levels. Although the current theory is that the change in meaning of the trauma memory is central in treatment, it is not clear what exactly causes this change. Arntz (2012) hypothesized that the expression of needs, feelings, and actions, which were inhibited at the time of the trauma (usually for survival reasons), is the most important healing factor. During a traumatic event it is natural that all kinds of needs, emotions, and action tendencies are triggered, but they usually cannot be fully actualized, as this is impossible (e.g. one is immobilized) or too dangerous (e.g. attacking a perpetrator might lead to the perpetrator killing you). In therapy, it is healthy and corrective to imagine emotions to be expressed, actions to be performed and needs to be met (Arntz, 2014b).

For PTSD resulting from childhood-related trauma, the ImRs protocol (Arntz, 2016) consists of twelve sessions, each session is consequently arranged in three phases.

1. Phase 1. The patient is asked to close the eyes and imagine a (traumatic) childhood experience as vividly as possible from the child's perspective. Questions that are asked: ‘Tell me: what do you see, what do you hear, what do you feel, what do you smell?’ ‘Tell me: what is happening?’ ‘What do you feel (emotionally)?’ ‘What do you think?’ Once strong trauma-related emotions are aroused, the final question is asked: ‘What do you need?’ After this Phase 1 ends.

2. Phase 2. The patient enters the image as an adult who witnesses the situation. The adult is prompted to intervene and do whatever is necessary for the child. Phase 2 stops when the patient as an adult is content with the interventions.

3. Phase 3. The patient is asked to imagine the scene again and experience, as a child, the intervention performed in Phase 2. The child is encouraged to express its own needs to the patient as an adult. The rescripting stops when the patient, from the child's perspective, is satisfied with the interventions (Raabe et al., 2015).

An alternative method is that, instead of the patient as an adult, in the first half of the treatment the therapist steps into the image in Phase 2, and intervenes, protects and meets the needs of the child (Boterhoven de Haan et al., 2017). In this way, the therapist acts as an example for the patient, who as yet misses the healthy perspective and strength to intervene (Arntz, 2014a). In the second half of the treatment, the adult patient steps into the image in Phase 2, and rescripts the event from the point of view as an adult. During Phases 2 and 3 in the session, the patient is able to express inhibited action tendencies, including the expression of emotions, and gets unmet needs met (Arntz, 2012). The patient liberates himself from the rigidity and impotency that are linked to the traumatic event. The needs expressed in Phase 3 may differ from those of Phase 2, since the perspective in the third phase is that of the child, which especially may involve need for help, support and comfort (Arntz & Weertman, 1999). Taking action and fulfilling these basic needs of the child can change the meaning of the original event and affect the sense of self. Also, it may help the patient to realize that the original experience was toxic and atypical, and therefore that it is better viewed as an extreme exception to normal life than as a general rule (Hackmann, 2011) or the child's fault. To illustrate, the three phases are elaborated in the case below.

1.1. Session case Liz

Phase 1. Liz, a 46-year-old woman, describes a situation at the age of five. Her parents argue, where her father takes a gun out of his pocket and point it at her mother's head.

Phase 2. Liz as an adult (‘Big Liz’) steps into the picture, together with six policemen. The policemen manage to overpower father. They handcuff him and on Big Liz's request they put duct tape over his mouth. Then two big policemen hold her father and Big Liz wants two other cops next to her. She stands in front of her father and tells him that he is a bad father and will be taken to prison. Father is led away. The therapist and Big Liz deliberate until Big Liz is satisfied with the actions that have been performed. Then father is chained and put into a cell, with camera surveillance and two large security guards at the door. After this, the therapist stimulates Big Liz to pay attention to Liz as a child (‘Little Liz’). Big Liz sits down with Little Liz and tells her not to be afraid anymore, because her daddy is safely locked up. Little Liz relaxes and Big Liz is satisfied.

Phase 3. Then the therapist asks Liz to describe the situation again from the perspective of Little Liz: about what is happening and about everything Big Liz does. When the therapist asks Little Liz what she still needs, it turns out that Little Liz thinks it's her fault that her parents argued. The therapist encourages Little Liz to ask Big Liz if the fight is her fault. Then the therapist helps Big Liz to explain to Little Liz that the fight is certainly not her fault, that she couldn't help it, that parents are responsible and have to take care of their children. Little Liz seems relieved by this, but it turns out she still needs something. She prefers not to be stay alone at home with her mother, who is lying on the couch under the influence of alcohol. The therapist helps Big Liz to find out where Little Liz would like to go and it turns that she wants to stay with her grandparents. Little Liz is taken to grandpa and grandma. She is satisfied when they tell her she can stay as long as she wants, and when she is sitting at the table with them playing a game of cards.

Although several authors (Arntz, 2012; Arntz et al., 2007; Hagenaars & Arntz, 2012; Holmes et al., 2007) have suggested that the expression of needs, feelings and actions may act as a potential healing factor, little specific research aimed at (in)validating this hypothesis has been done so far. The present study aimed to investigate the relation between, on the one hand, carrying out action tendencies and fulfilling needs for help, support, and comfort during ImRs treatment, and, on the other hand, changes in PTSD symptoms in patients with childhood-related PTSD. We hypothesized a better treatment outcome in terms of PTSD symptoms when needs are fulfilled and action tendencies are expressed during the treatment sessions.

To investigate this hypothesis, we needed an instrument with which it is possible to observe to what extent the needs are met and the action tendencies are expressed during the treatment sessions. To our knowledge, such an instrument didn’t exist at the time of this study. Therefore, an observation instrument, the Needs an Action Tendencies scale (NATS), was developed to assess the fulfilment of needs and expression of action tendencies during the ImRs sessions. To operationalize the needs and action tendencies, we used the emotional basic needs for the human individual, postulated by Young (Young et al., 2003): (1) a safe connection with other people, (2) the need for autonomy, and the development of one's competence and sense of identity, (3) the freedom to express your needs and emotions, (4) spontaneity and play, and (5) clear boundaries.

2. Method

2.1. Study design

This study is part of the international multicentre Randomized Clinical Trial on the effectiveness of ImRs vs eye movement desentization and reprocessing (EMDR) as treatment of childhood trauma-related PTSD in adults (IREM). Participants of the multicentre study were several mental healthcare institutions from different regions of the Netherlands, an institution in Germany and an institution in Australia (Boterhoven de Haan et al., 2017; Boterhoven de Haan et al., 2020). In the Netherlands, the main study was approved by the ‘Ethische Commissie Psychologie’ (ECP) of the University of Maastricht. All participants provided written informed consent. For this additional study, no additional approval was required.

In order to develop an observation instrument, we conducted a web-based Delphi study with 10 ImRs experts. The Delphi technique is a method used to obtain a reliable consensus of a group of experts by a series of intensive questionnaires interspersed with controlled feedback (Fenella et al., 2013). The premise underlying the Delphi method is that the collective opinion of a panel of experts is more valid than the panel members’ individual opinions (Voogt et al., 2016). This Delphi study consisted of two rounds. The degree of agreement between the experts was taken as a starting point for further adaptations.

In the first round the observability and relevance of the five basic needs of Young were reviewed by the experts. A sixth need ‘exculpation’ was added by the researchers based on the experiences of therapists with ImRs in early childhood trauma. Based on the reviews and comments in this round, the first need ‘a safe connection with other people’ was divided in two needs ‘safety ‘and ‘feeling understood and comforted’ and the need ‘clear Boundaries’ was removed because this need was not seen as relevant. We then designed the first draft of the observation instrument, which was presented to the experts in the second round. Their evaluation of this draft eventually resulted in the used instrument, consisting of the six domains: Safety, Feeling understood and comforted, Self-expression, Exculpation, Autonomy and Competence, and Sense of Identity and Spontaneity and Play. In Table 1 short description of needs and action tendencies for each domain is given.

Table 1. Content of NATS domains.

Domain	Needs	Action tendencies
Safety:
• Restraining the perpetrator	Client feels safe; client is no longer afraid of having to return to the old unsafe situation.	Restraining the perpetrator, ensuring client's as a child safety, finding a solution for future safety.
• Client's safety in the short term
• Client's safety in the long term
Feeling understood and comforted:
• Care	Client feels understood, comforted, accepted, and respected.	Comforting, hugging, showing appreciation to the client as a child. Taking care of the client as a child.
• Comfort
• Acceptance
• Respect
Selfexpression:
• Expression of feelings	Client verbalizes own feelings, thoughts, and desires.	Helping the client as a child to put feelings, thoughts and desires into words and validate them.
• Expression of thoughts
• Expression of desires
Exculpation:
• Client is not to blame for the situation.	Client does not feel guilty and responsible for the situation.	Explaining the client as a child that he is not at fault for the situation and bears no responsibility for it. Holding the perpetrator(s) accountable for their actions and their responsibility.
• Client is not responsible for the consequences of the perpetrator's actions.
• Perpetrator is held accountable.
Autonomy, competence and sense of identity:
• Autonomy – having the courage to do things regardless of what others think or expect	Client perceives himself as autonomous and competent.	Client carries out interventions he deems necessary and until he is satisfied with the result.
• Competence – being able to do things, having the belief in one's own capabilities
Spontaneity and play	Client dares to be spontaneous without holding back, to have fun, to do what makes him happy.	Client engages in play and enjoyable activities with the client as a child or allowing the child to play with a trusted adult nearby to supervise.

2.2. Intervention

ImRs treatment consisted of twelve 90-min ImRs sessions, conducted twice a week. Therapists were licensed psychologists, psychiatrists and a nurse specialist who had successfully completed a basic training course in CBT, and a two-day training in ImRs for PTSD related to childhood trauma. Every week the therapists met for one hour of supervision by an ImRs specialist. The treatment was planned within six weeks, but could be extended to eight weeks if necessary. Participants, who no longer experienced PTSD symptoms, based on their score on the primary outcome measure, were allowed to complete treatment earlier than after the allocated twelve sessions. For this additional study clients should have received treatment for at least eight sessions.

2.3. Participants

Inclusion criteria of the IREM study were: age between 18 and 70, PTSD as primary diagnosis as a result of trauma before 16 years of age, and agreement of the participant that this trauma is the focus of treatment. PTSD symptoms had to last longer than three months and were mainly linked to the trauma before 16 years of age. Exclusion criteria were: psychiatric conditions that required specifically other treatment (comorbid psychotic disorder, bipolar disorder, type 1 and alcohol- or drugs dependency; in contrast, a DSM-IV diagnosis of substance abuse was not an exclusion criterion). Other exclusion criteria were: the use of benzodiazepines (patients were motivated to stop benzodiazepine use at least 2 weeks before start of treatment), acute suicide risk, IQ < 80, PTSD due to an experienced trauma within the past 6 months or any PTSD-focused therapy in the past 3 months (Boterhoven de Haan et al., 2020).

For this pilot study, which concerns only the imrs condition, we aimed to involve N = 20 participants.

2.4. Measures

2.4.1. Diagnosis

Psychiatric diagnosis was established with the SCID I screener, followed by the SCID I interview (First et al., 2011).

2.4.2. Treatment mechanism measure

The Needs and Action Tendencies Scale (NATS) consists of 6 domains: Safety, Feeling understood and comforted, Self-expression, Exculpation, Autonomy, Competence and sense of identity and Spontaneity and play.

Each domain contains a Needs Subscale (N-scale) and an Action tendencies Subscale (At-scale) and is scored on a three-point rating scale, ranging from 0 (‘need not met’ respectively ‘action tendency not expressed’), to 1 (‘need is a little fulfilled’ respectively ‘actions, but insufficient results’) to 2 (‘need completely fulfilled’ respectively ‘action tendency fully expressed’). A score of 1 on the At-scale thus means that an action has been performed on the domain in question, but that this action has had insufficient results: to fully express the action tendency, an additional action should have been performed. For example, with a score of 1 on the domain ‘Exculpation’ (see Table 2) only the perpetrator is addressed, but to the client as a child it is not explained that it is not to blame for the situation. We computed both a total score, combining the two subscales into one score as well as scores on each subscale. The two subscales of the NATS ranged 0–12 per session. To compute the total score on each subscale the scores on each of the 11 sessions were added, resulting in a range of 0-132. To compute the combined total score we added the scores of each subscale, resulting in a range of 0–24 per session and 0–264 for all 11 sessions.

Table 2. Scoring instruction for the domain Exculpation (domain 4) from the NATS.

Scoring instruction:
Need	Action tendency
0. Client still feels guilty/responsible	0. The therapist or client does not explain to the client as a child about guilt and responsibility and does not hold the perpetrator(s) accountable for their actions and their responsibility.
1. Client feels a little less guilty/responsible	1. The therapist or client explains that the situation is not the fault of the client as a child, but that others are responsible for the situation. Or the perpetrator(s) are held accountable for their actions and their responsibility.
2. Client no longer feels guilty / responsible	2. The therapist or client explains that the situation is not the fault of the client as a child, but that others are responsible for the situation. And the perpetrator(s) are held accountable for their actions and their responsibility.

Since, in this study, the therapist enters the image during the first six sessions in Phase 2 to intervene and the client fully experiences this phase from the perspective of the child, it was decided to only score the therapist's performed action tendencies on the At-scale. Possible action tendencies the client performed as a child were not included in the score, so if only the client performed actions in Phase 2, it was scored as 0. For the last six sessions, in which the client intervenes as an adult, the therapist's actions were not scored in Phase 2.

To investigate the reliability of the NATS two assessors watched the same video recordings of 10 randomly selected ImRs sessions (up to 90 min) of different clients and scored these independently with the NATS. On the basis of the observed differences in scoring a manual was developed and a standard score was determined for these 10 sessions. This standard score was used in the training of future assessors, where after scoring of the same 10 sessions the scores on the NATS were compared with the standard score.

To determine agreement with the standard scores and inter-rater reliability, a third assessor also scored the 10 ImRs sessions with the NATS.

Cohen's κ was computed to determine the interrater-reliability, which showed excellent agreement between the two assessments (κ = 0.83, (95% BI, 0.30–0.89), p < .001). To measure the internal consistency we calculated Cronbach’s α of the 24 clients and their 11 sessions (n = 249; some clients finished earlier the treatment). The reliability analysis of the 24 clients showed that both the N-scale (Cronbach's α = 0.68) and the At-scale (Cronbach's α = 0.70) had moderate internal consistency.

2.4.3. Primary outcome measure

As a primary outcome instrument we used the Impact of Events Scale-Revised (IES-R; Creamer et al., 2003; Weiss & Marmar, 1997). The IES-R is a 22-item self-report questionnaire measuring symptomatic response (i.e. subscales Intrusions, Avoidance and Hyperarousal) over the last 7 days, to a specific trauma event. The questionnaire scored with 0 (‘not at all), (1 ‘a little bit’), 2 (‘moderately’), 3 (‘quite a bit’) and 4 (‘extremely’). The lower the score on the IES-R, the fewer PTSD symptoms a participant has. The IES-R was filled out twice: once for the ‘Index trauma’ and again for ‘all other traumas’ as identified on the Life Events checklist for the DSM-5 (LEC-5; Weathers et al., 2013). The index trauma is the event or series of events of the same type experienced before 16 years of age and ‘all other traumas’ are any other trauma experienced across the lifetime. We used the accumulated total score of the two subscales, because during the treatment the focus was on the index trauma (the index trauma had to be treated in the first six sessions), but other traumas were also treated, as a trauma could be treated for a maximum of 2 sessions, after which the treatment had to continue to focus on the next trauma. The score range is 0–176 for each session.

Research by Creamer et al. (2003) shows that the internal consistency of the IES-R is excellent (Cronbach's α = 0.97). This also applies to test-retest reliability (r = 0.89–0.94). In the current study we calculated the Cronbach’s α of the 24 clients over session 2 and posttreatment and found similar results. The reliability analysis of the 24 clients showed that for the index trauma (Cronbach's α = 0.97) and for all other traumas (Cronbach's α = 0.96) had excellent internal consistency.

2.5. Procedure

Inclusion of clients proceeded according to the research protocol of the main study (Boterhoven de Haan et al., 2020). The IES-R was filled out by the client at the start of each therapy session. All sessions were video recorded. Because session 1 was an introduction session, we used the video recordings of session 2–12. For this pilot study four trained assessors scored 249 video recordings of 24 clients with the NATS. The video recordings of the various Dutch locations were collected personally by one of the trained assessors and transported via a secure external hard disk. After that, the video recordings were stored in a central database in accordance with the General Regulation Data Protection (AVG) on a European server. Only the assessors of this study had access to the recordings.

2.6. Statistical analyses

Scores on the NATS served as independent variables and were based on observations of 11 sessions. If a session score was missing, for example by early completers, then the mean of the non-missing values from that participant was used. The mean was only calculated when at least eight sessions were present. The mean score over these sessions was calculated and multiplied by 11 to calculate the total scores. We used the total scores on the subscale fulfilled needs (N-scale) and the total scores on the subscale expressed action tendencies on the six domains (At-scale) as well as a combined total score consisting of the sum of the N-scale and the A-scale.

The distribution of the NATS was analysed by calculating the minimum and maximum scores for both subscales, the mean and standard deviation. For the variables N-scale (W(24) = 0.97, p = .57) and At-scale (W(24) = 0.94, p = .19) the assumption of normality was met according to the Shapiro–Wilk test. To quantify the relationship between the two subscales of the NATS we calculated the Pearson correlation between the total score on the N-scale and the total score on the At-scale. Since higher scores on the IES-R at the start of the treatment could implicate more room for change for patients, we analysed the correlations between the scores on the N-scale and the At-scale on the one hand and pretreatment and posttreatment scores on the IES-R on the other, with the Pearson correlation coefficient.

We then performed a hierarchical multiple linear regression analysis with the scores on the At-scale and N-scale as independent variables to analyse the predictive value of PTSD symptoms after controlling for the influence of pretreatment PTSD symptoms. The assumption for normal distribution of the residuals turned out to be violated. However, since transforming the data had insufficient effect, it was decided to leave the data untransformed. This means that the results cannot be generalized to the general population.

3. Results

A total of 74 participants were included in the ImRs condition in the international multicentre randomized clinical trial. At our request four out of five Dutch treatment centres provided video material of 39 therapies. Of three ImRs treatments the sessions did not comply with the ImRs protocol, of three ImRs treatments we received the video material too late, of two ImRs treatments we received less than eight sessions, and in one ImRs treatment, the recordings of the therapy sessions were too short; all in all, we were left with material from 30 therapies. In this study we scored the therapies of 24 participants with the NATS, which amounted to 249 videos. The other six therapies were not included because of the amount of time it took to score all the individual videos.

The 24 participants in this pilotstudy completed the treatment and filled in the posttreatment measurement. The treatment sample consisted of four men and 20 women. The mean age of the participants was 38.54 years (SD = 10.88), within a range of 19–55 years. 20 participants had completed a tertiary education, ranging from vocational to academic education; three participants had no further education after secondary school and one participant only had attended primary school. The ethnic background of the participants, 54.2% (n = 13) came from the Netherlands, 8.3% (n = 2) from another European country, 25.0% (n = 6) from Morocco, 4.2% (n = 1) from India and 4.2% (n = 1) from Mexico. The duration of PTSD symptoms ranged from 7 months to 53 years, with a mean of 20.64 years (SD = 16.45).

Minimum and maximum scores, the mean and standard deviation of the two subscales of the NATS, and the total score that combines the two subscales and the IES-R for the index trauma and all other traumas combined for pretreatment and posttreatment can be found in Table 3, who showed a wide distribution in scores between the total score and subscales of the NATS and the on the IES-R. The scores on the N-scale ranged from 17.0 to 97.0 and on the At-scale from 47.0 to 112.0. This means much variability among treatments in fulfilled needs and expressed actions.

Table 3. Minimum and maximum scores, mean and standard deviation of the two subscales, total score of the NATS, and the IES-R for the Index Trauma and All Other Traumas for Pretreatment and Posttreatment (n = 24).

	Minimum	Maximum	Mean	Std. Deviation
Total score N-scale	17.0	97.0	65.6	21.1
Total score At-scale	47.0	112.0	71.7	15.5
Combined total score	67.0	209.0	137.3	34.1
IES-R Pretreatment trauma combined	76.0	153.0	108.9	21.6
IES-R Posttreatment trauma combined	9.0	150.0	56.5	41.2

Table 3 shows a wide distribution in pretreatment and posttreatment of PTSD symptoms. On average there is a decrease of the mean scores on the IES-R, but when we zoom in, it turns out that in one participant the PTSD symptoms had increased with 38 points on the IES-R, three participants showed hardly any a change (+2.7 to −2) and in 20 participants there is a decrease in IES-R scores (−6 to −117).

The correlation coefficients between the score on the N-scale and the score on the At-scale are described in Table 4. There was a significant positive correlation between the N-scale and the score on the At-scale (r = 0.73, p < .01). This means that the more the action tendencies were expressed, the more the needs of a patient were fulfilled, and vice versa.

Table 4. Correlations between the Needs Subscale, Action tendencies Subscale and the IES-R for the Index Trauma and All Other Traumas without Index Trauma for Pretreatment and Posttreatment (n = 24).

Total score	NATS		IES-R Pretreatment		IES-R Posttreatment
	N-scale	At-scale	Index trauma	All other traumas	Index trauma	All other traumas
N-scale	1.00	.73**	.16	−.18	−.50*	−.34
At-scale	.73**	1.00	−.01	−.21	−.44*	−.29

* Correlation is significant at the 0.05 level (2-tailed). ** Correlation is significant at the 0.01 level (1-tailed).

Table 4 further shows that the correlations between the two subscales of the NATS with the pretreatment scores on the IES-R were not statistically significant, indicating that NATS scores were independent of the severity of the symptoms at the start of the treatment. Table 4 also shows that posttreatment scores on the IES-R for index trauma were statistically significant correlated with the scores on the N-scale (r = −0.50, p < .05), and with the scores on the At-scale (r = −0.44, p < .02); correlations of the scores on the IES-R for other traumas with both subscales of the NATS were not statistically significant. These findings indicate that the more the action tendencies are expressed and the more the needs are met, the lower the PTSD Symptoms on the index trauma after treatment.

A hierarchical multiple regression was used to assess the ability of two subscales of the NATS (At-scale and N-scale) to predict posttreatment IES-R scores after controlling for the influence of pretreatment IES-R scores. Preliminary analyses were conducted to ensure no violation of the assumptions of normality, linearity, multicollinearity and homoscedasticity. The N-scale and At-scale were highly correlated, r = .725, which could indicate multicollinearity. Tolerance was not smaller than .10 (.47) and VIF value was below 10 (2.1). Pretreatment IES-R scores were entered at Step 1, explaining 35.8% of the variance in posttreatment IES-R scores. After entry of the At-scale and N-scale at step 2 the total variance explained by the model as a whole was 59.4%, F(3, 18) = 8.778, p < .001. The two subscales explained an additional 23.6% of the variance in posttreatment IES-R after controlling for pretreatment IES-R scores, R_{squared change} = .236, F_change (2, 18) = 5.24, p = .016. In the final model, the At-scale semipartial correlation value (sr = −.30. p = .061) and N-scale (sr = −.045. p = .766) individually did not significantly predict the scores on the IES-R. This means that the overall model, incorporating both subscales of the NATS, significantly predicted approximately a quarter of the variance in posttreatment IES-R scores. It's important to note that the individual contributions of the NATSs were not statistically significant. However, the Action Tendencies scale was marginally significant.

4. Discussion

The aim of this study was to investigate the extent to which fulfilling basic needs and performing action tendencies in ImRs treatment predicts PTSD symptomatology after treatment.

We hypothesized better treatment outcomes in terms of PTSD symptomatology, when needs are met and action tendencies expressed during treatment sessions, for clients with PTSD due to childhood trauma. To this end, we developed a new observation tool, the NATS. Our results show that the inter-rater reliability of the NATS is good, but also that the internal consistency of both subscales is moderate. This should be taken into account when interpreting the results.

Results show that about a quarter of the variance in posttreatment PTSD symptoms can be explained with the NATS. However, it cannot yet be firmly established that this is due to one of the two subscales of the NATS, since the regression coefficients of both subscales are not statistically significant. This implies that when considering the NATS as a whole, it provides valuable predictive information about posttreatment IES-R scores, suggesting that clients’ needs and action tendencies, when taken together, play a role in posttreatment distress levels. At the same time, both scales correlate around −0.50 with the posttreatment score on the IES-R for the index trauma. This indicates that the more action tendencies are expressed and the more the needs are met the lower the PTSD Symptoms on the index trauma after treatment.

Although there was no multicollinearity between the At-scale and the N-scale, there was a relatively high correlation between them (>0.70). It is quite possible that the subscales explain an overlapping part of the variance of the posttreatment score on the IES-R, so that each of them was not predictive in the model. This would mean that these two subscales cannot be seen separately. Obviously, this suggests that without taking action, it becomes difficult to fulfil basic needs. Once in a while it happened that a client did not perform any action tendencies, while in these sessions the therapist was active, making it possible to have a score of 1 or 2 on the N-scale. However, this was very rare. Most importantly, what we saw was that in that situation the therapist encouraged the client to perform certain actions, without taking over the process, and the client then performed as the therapist said. This resulted in a scored action tendency. These observations and the close correlation between need fulfilment and action tendencies in ImRs give rise to the idea that action tendencies are a condition for need fulfilment. This could be an interesting topic in a future study.

One main weakness in the study is the lack of other variables examined as possibly related to a reduction in PTSD symptoms. Previous research shows that a number of factors might be important, and it would be preferable to examine all potential variables that can explain the variance. For example, a variable that we have not examined is the influence of the behaviour of the therapist. The question could be to what extent it is important that the therapist performs certain actions instead of the client and whether this differs for clients with high or low PTSD scores.

Another important limitation of this pilot study is of course the limited sample of n = 24 clients. It would be of great help if a labour-intensive study like this could be replicated, for example also including administration of the IES-R before and after the session. This makes it possible to gain more insight into cause and effect. Performing action tendencies and fulfilling basic needs may be conditional for the reduction of PTSD symptoms, but the reverse is also possible: perhaps the reduction in PTSD symptoms makes it easier for the client to perform action tendencies and fulfil basic needs.

The findings from this study make us curious about which actions are particularly important in phases 2 and 3. A possible hypothesis is that actions and fulfilled needs are not responsible for the reduction of PTSD symptoms in all domains. For example, is this decrease mainly due to bringing the child to safety? Or due to the exculpation by the adult? Follow-up research could focus on which of the six domains of the NATS has the greatest effect on the reduction of PTSD symptoms. With this information we can further improve the ImRs protocol.

This pilot study on the underlying mechanisms of ImRs in PTSD treatment has shown that the expression of action tendencies and fulfilling basic needs during ImRs offers predictive value for posttreatment distress, and that actions and basic needs cannot be viewed separately. Further investigation is needed to understand the specific contributions of each subscale and their individual predictive power.

A study as this is very labour-intensive, but at the same time very important to better understand the underlying mechanisms of a treatment such as ImRs. In practice, it can encourage therapists not only to ask during the treatment whether the rescripting itself was sufficient according to the client, but also to check, for example, whether sufficient actions have been taken and whether the basic needs have been met.

Acknowledgements

We would like to thank Arnoud Arntz, principal investigator for the IREM study, Mariel Meewisse, principal investigator for the GGZ-NHN site and the research assistants Mulan Koopmans, Shekib Haidari and Myrza Bossenbroek for their contribution to this study. This article is a revised version of an article in the Dutch magazine Gedragstherapie. https://www.tijdschriftgedragstherapie.nl/inhoud/tijdschrift_artikel/TG-2022-0-5/Werkingsmechanismen-van-imaginaire-rescripting-ImRs-bij-volwassen-clienten-met-PTSS-vanuit-de-kindertijd. The editors and the publisher give permission for publication in EJPT.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Data is available upon request by contacting [email protected].