A Review of Empirical Studies Investigating Narrative, Emotion and Meaning-Making Modes and Client Process Markers in Psychotherapy

Despite the importance of narrative, emotional and meaning-making processes in psychotherapy, there has been no review of studies using the main instruments developed to address these processes. The objective is to review the studies about client narrative and narrative-emotional processes in psychotherapy that used the Narrative Process Coding System or the Narrative-Emotion Process Coding System (1.0 and 2.0). To identify the studies, we searched The Book Collection, PsycINFO, PsycARTICLES, PsycBOOKS, PEP Archive, Psychology and Behavioral Sciences Collection, Academic Search Complete and the Web of Knowledge databases. We found 27 empirical studies using one of the three coding systems. The studies applied the Narrative Process Coding System and the Narrative-Emotion Process Coding System to different therapeutic modalities and patients with various clinical disorders. In some studies, early, middle and late phases of therapy were compared, while other studies conducted intensive case analyses of Narrative Process Coding System and Narrative-Emotion Process Coding System patterns comparing recovered vs unchanged clients. The review supports the importance to look for the contribution of narrative, emotion, meaning-making patterns or narrative-emotion markers, to treatment outcomes and encourages the application of these instruments in process-outcome research in psychotherapy.


Introduction
The narrative of lived stories has been intrinsic to human beings and their relationships (Angus 2012;Gonçalves and Gonçalves 2007). The importance of personal narratives to our existence is consensual among diverse authors from philosophy, social sciences and psychology (McAdams 2008). Given its importance to human interaction, some authors were interested in narrative expression on therapist-client discourse (Gonçalves 1995;Neimeyer and Levitt 2000).
The critical assumption of the Narrative Process theory of therapy is that therapists and clients work together to form a coherent self and a meaningful client macronarrative ). According to this theory, all forms of psychotherapy with good therapeutic results involve the articulation, elaboration and transformation of clients' macronarratives. In the 90s, researchers published the first studies on narrative processes in psychotherapy with the development of the Narrative Process Coding System, intentionally designed to identify extended sequences of therapist-client dialogue (Angus 2012;Angus and Hardtke 1994;Angus et al. 1999). These studies already focused on the interconnection of narrative storytelling, emotional expression and meaning-making processes. Since 2014, with the development of the Narrative-Emotion Process Coding System, other articles appeared that focused on increasing levels of narrative-emotion integration in psychotherapy ). This integration of narrative and emotion is mainly due to the relevance of scientific evidence that shows that emotional expression and narrative expression, acting alone, are not enough for good therapeutic outcomes (Boritz et al. 2011). However, despite the importance of narrative and emotional processes in psychotherapy, a review of the Narrative Process Coding System (Angus et al. 1992(Angus et al. , 1996 and the Narrative-Emotion Process Coding System (Angus Narrative-Emotion Process Lab 2015; Angus et al. 2017) has not been published yet. Two previous publications (Angus 2012;Angus et al. 2017) focus on the multi-methodological steps to the creation of Narrative-Emotion Process Coding System (Versions 1.0 and 2.0). In this study, we will briefly describe the development of Narrative Process Coding System and Narrative-Emotion Process Coding System and then focus on their empirical research findings. To our knowledge, there is no other coding system that explicitly addresses the essential interrelationship of narratives (autobiographical memory descriptions) and emotional/ experiential processes in therapy session dialogues. It is also the only coding system that explains how narrative and emotion processes are the basis for enhanced self-reflectivity, new meaning-making and self-narrative change.

The Development of Coding Systems
The Narrative Process Coding System is an observer-based empirical measure used by independent raters that enables researchers to identify narrative process modes occurring in therapy session transcripts throughout therapist-client' dialogue, regardless of therapeutic modality . It was drawn from a dialectical constructivist view of therapeutic change and psychotherapeutic discourse. It enables researchers to capture the micronarrative and macronarrative change processes in psychotherapy in terms of one of three process modes which lead to self-change in psychotherapy. These three narrative processes emerged from the recognition of similarities in the discourse used in the coconstruction of the clients' macronarratives within the therapy hour, in different modalities.
Both clients and therapists engage in these three different narrative modes: (1) External Narrative mode includes the disclosure of personal stories/autobiographical memories (micronarrative). (2) Internal Narrative mode includes descriptions of bodily felt feelings and emotions, the expression and articulation of affections. (3) Reflexive Narrative mode or meaning-making processes draws into the coupling of both storytelling and emotion processes, and it results in new life themes and self-understandings. These three categories made the Narrative Process Coding System a comprehensive category method of therapeutic interaction, allows the researchers to identify interactional units that contain the verbal interchanges between therapists and clients and understand what happened (External Narrative mode), how it feels (Internal Narrative mode) and what it means (Reflexive Narrative mode). All three narrative types had a critical function to fulfil in the coconstruction of the clients' selfnarrative change .
Over the years, the researchers have established good levels of interrater agreement for the measure' application (Cohen's Kappa 0.75), as reported in a series of studies (Angus et al. , 2004. The Narrative Process Coding System was applied with different therapeutic modalities. However, years later, Angus and Greenberg (2011) realized that in good Emotion-Focused Therapy sessions, narrative processes and emotional processes do not act in isolation but co-occur in the context of enriched storytelling and emotional meaning-making sequences.
To enhance narrative-emotion integration in Emotion-Focused Therapy sessions, Angus and Greenberg (2011) created eight clinically derived narrative-emotion integration markers. These markers were initially identified for the implementation of process-guiding therapeutic responses. The authors subdivide these Narrative-Emotion Process markers into two categories, to discern the client's problematic vs productive narrative indicators. The Problem Markers include: Same Old Storytelling, Empty Storytelling, Unstoried Emotion, Superficial Storytelling. The Change Markers include: Competing Plotlines Storytelling, Inchoate Storytelling, Unexpected Outcome Storytelling, Discovery Storytelling. The identification of these critical Narrative-Emotion Process markers makes it possible to improve clinical practice, adjust therapist interventions and support future studies.
Boritz and collaborators developed a first version of Narrative-Emotion Process Coding System (version 1.0) (Boritz et al. 2012) to include these Narrative-Emotion Process markers. The Narrative-Emotion Process Coding System 1.0 enables the identification of verbal and non-verbal indicators of narrative and emotion integration by researchers in video-based therapy sessions. This video-based coding system allows access to paralinguistic and non-verbal indicators of narrative-emotion processes, not presented in therapy transcriptions. Several process-outcome studies applied the Narrative-Emotion Process Coding System 1.0.
Three years later, Angus Narrative-Emotion Marker Lab (2015) refined the coding system and standardized a videobased manual to bring different levels of client reflection and meaning-making processes in videotaped therapy sessions. These second version (2.0) includes one new marker subgroup-Transition Markers, with three individual markers. The present article is the first review of all empirical studies that used one of the three coding systems: either the Narrative Process Coding System (Angus et al. 1992(Angus et al. , 1996, or the Narrative-Emotion Process Coding System 1.0 (Boritz et al. 2012) or Narrative-Emotion Process Coding System 2.0 (Angus Narrative-Emotion Marker Lab. 2015. The review will address the research findings emerging from the application of these two measures to a range of clinical disorders and theoretical orientations, followed by a critical discussion of future research directions and implications for practice.

Method
Systematic electronic searches were independently performed by two researchers without a time limit, for the following formula: narrative-emotion processes or NEPCS or Narrative-Emotion Process Coding System or narrativeemotion markers or narrative processes or NPCS or Narrative Process Coding System (full text) and Angus, Lynne (author) in the electronic databases Book Collection, Psy-cINFO, PsycARTICLES, PsycBOOKS, PEP Archive, Psychology and Behavioral Sciences Collection, Academic Search Complete and Web of Knowledge. We examined the list of references from the review articles, original articles and book chapters to find any more potential study. As inclusion criteria, studies had to apply either the Narrative Process Coding System (Angus et al.1992(Angus et al. , 1996, or Narrative-Emotion Process Coding System 1.0 (Boritz et al. 2012) or Narrative-Emotion Process Coding System 2.0 (Angus Narrative-Emotion Marker Lab. 2015. After careful screening of the records, 27 articles were eligible for inclusion. 1 We integrated the findings into a narrative review, which involved describing critically, appraising, and comparing the studies into a coherent theoretical framework.

Results
We identified 27 empirical studies, 14 applying the Narrative Process Coding System (Angus et al. 1992(Angus et al. , 1996, 3 using the Narrative-Emotion Process Coding System 1.0 (Boritz et al. 2012) and 10 the Narrative-Emotion Process Coding System 2.0 (Angus Narrative-Emotion Marker Lab. 2015;Angus et al. 2017). The empirical findings from the 27 studies are thematically clustered and reviewed in the context of three major subsections: (a) Studies with the Narrative Process Coding System; (b) Studies with the Narrative-Emotion Process Coding System and (c) New Directions in Narrative Process Coding System and Narrative-Emotion Process Coding System Research Studies. The Narrative Process Coding System focuses on narrative modes and the Narrative-Emotion Process Coding System on Narrative-Emotion Process markers. For this reason, we chose to analyze separately the studies carried out with both instruments. In each section we decided to group the studies according the following classification: process-outcome studies, studies using client, therapist, or treatment variables.

Process-Outcome Studies
Most of the studies in this section looked for the relationship between the proportion of narrative modes at different stages of therapy and outcomes. Angus and Hardtke (1994) illustrate a pioneer study exploring the relationship between narrative process modes and therapeutic outcomes in Brief Dynamic Therapy. The authors applied the Narrative Process Coding System (Angus et al. 1992) to the three best and the three worst therapeutic outcomes in different phases of therapy. The results show that successful clients had a higher frequency of Reflexive Narrative modes than unsuccessful clients (42% vs 28%) and a lower frequency of External Narrative modes (47% vs 57%) and Internal Narrative modes (11% vs 15%). Also, the researchers noted that the percentage of Internal and External Narrative modes increased over time in the group with the worst results, compared to the group with the best results, while the frequency of Reflexive Narrative modes increased throughout the therapy sessions (33%, 45%, 48%).
Clients often express themselves with metaphors, but there is a lack of studies focusing on metaphoric expression in psychotherapy. To bridge this gap,  sought to understand the relationship between 'burden' metaphors and different therapeutic outcomes comparing one recovered vs one unrecovered client in process-experiential therapy. The results demonstrate a significant statistical difference between the two outcomes [ X 2 = 13; df = 2; p < 0.001]. While in the best outcome, as the therapy progresses, the burden metaphors are transformed into metaphors "in which the burden is unloaded", there is no evident transformation in poor therapeutic outcome. Besides, the percentage of metaphors in Reflexive Narrative modes is higher in the case with a good outcome, than in the case with a poor therapeutic outcome (67% vs 65%). The relationship is similar concerning Internal Narrative modes (32% vs 15%). The main conclusion of this study is that successful clients express metaphors to represent personal and internal experiences during the established therapeutic relationship through the expression of internal narratives. Banham and Schweitzer (2015) examined the relationship between the results of therapist-client dyads and the narrative processes that occurred in session in initial, mid-and late phases of therapy. They compared the six best therapeutic results with the six worst therapeutic results of a sample of patients diagnosed with depression, followed up in three different therapeutic modalities. The researchers concluded that dyads with better therapeutic outcomes show significant statistical increases [F(2,15) = 19.09; p < 0.001] in the total number of Reflexive Narrative modes throughout the therapy and a higher percentage of Internal Narrative modes in the middle phase [F(1,10) = 14.76; p < 0.01], compared with the group with the worst outcomes. The results also highlight that while at the beginning of therapy, both groups tend to express External Narrative modes, clients with the best results evidenced more Reflexive Narrative modes at the end of the treatment. In contrast, the clients with the worst results continue to engage External Narrative modes, regardless of the stage of therapy.
Lewin's study corresponds to a methodological progress because it addressed narrative mode shifts instead of the proportion of different narratives. Lewin (2010) sought to understand the prediction effect of Internal-Reflexive Narrative mode shifts to therapeutic outcome. So, she compared early, middle and late phases of Emotion-Focused Therapy and Client-centered Therapy of recovered vs unrecovered depressed clients. The results revealed no significant statistical association between narrative mode shifts and treatment outcome in Emotion-Focused Therapy sample, using Beck Depression Inventory (p = 0.80) and Global Symptom Index as outcomes (p = 0.76). Client-centered Therapy sample revealed similar results (p = 0.89).
Most of the External Narrative modes identified in psychotherapeutic sessions are reports of autobiographic memories. Some studies have used the Narrative Process Coding System to identify autobiographic memories in External narrative modes and investigate the relationship with treatment outcomes for clinical depression (Boritz et al. 2008(Boritz et al. , 2011. Boritz et al. (2008) focused the relationship between autobiographic memories' specificity and depression. They conducted a study to investigate the relationship between the change in the level of client's depression and autobiographic memories' specificity, comparing initial, middle and final phases of Client-centered Therapy and Emotion-Focused Therapy. The authors report the existence of a pattern, regardless of the treatment type and level of clients' symptomatology at the end of therapy: autobiographic memories' specificity increases significantly throughout therapy [early to late: t(147) = 2.002; p = 0.047 and middle to late: t(147) = 2.373; p = 0.019]. Boritz et al. (2011) carried out a similar study to investigate the role of emotional intensity in session. Results suggested a significant statistically positive relationship between autobiographic memories' specificity and emotional intensity peaks [t(744) = 2.396; p = 0.016] for recovered clients. Thus, the specificity of autobiographic memories if accompanied by high levels of emotional expression seems to be related to better therapeutic outcomes. Literature support this relationship. Researchers highlight that autobiographic memories' narration, the symbolization of emotions in narrative form, the reflection that takes place in session and the creation of new meanings, allows clients to identify, differentiate and organize their subjective internal experiences (Angus 2012;Angus et al. 2004).

Treatment and Client Variables
Shortly after being created, Narrative Process Coding System started to be used for comparing narrative modes in different treatment models.  compared Process-Perceptual Therapy, Process-Experiential Therapy and Client-Centered Psychodynamic Therapy. The therapists selected three clients with good therapeutic outcomes for the study. The results demonstrated significant differences in terms of the number of narrative sequences [ X 2 (2) = 11.61; p = 0.003] and the type of narrative modes [ X 2 (4) = 25.40; p = 0.0001]. There was a higher frequency of Internal Narrative mode in Emotion-Focused Therapy (29%), in comparison to the other two therapies. The therapy with most Reflexive Narrative mode was Process-Perceptual Therapy (54%); with the same frequency of External Narrative mode in Client-Centered Psychodynamic Therapy. Emotion-Focused Therapy demonstrated more narrative mode sequences, emphasizing more narrative mode shifting in therapy. This study stresses the importance of the focus on the internal experiences and self-questioning for good results in therapy and the importance of experiential and information processing as promoters and facilitators of narratives that contribute to good therapeutic outcomes.
Lewin (2010) pursued a second objective: understand the relationship between narrative process mode shifts of recovered clients and their Experiencing levels (using the Experience Scale, Klein et al. 1970). So, regarding recovered Emotion-Focused Therapy and Client-centered Therapy clients, she found a significant statistically relationship (p = 0.00001; p = 0.004), respectively, between Internal -Reflexive Narrative mode shifts and higher levels of clients' Experiencing, from early stages to late sessions of therapy. Daniel (2011) looked for differences in narrative processes in patients with different attachment types. She used the Adult Attachment Interview (AAI) to classify different kinds of attachment. Four dismissing clients and four preoccupied clients, with a diagnosis of bulimia nervosa (DSM-IV-TR), were randomly selected for Cognitive-Behavioral Therapy (CBT) or Psychoanalytic Therapy. The author used six of the first 20 sessions from each of the eight therapy dyads (the most difficult experienced three sessions vs the least difficult experienced three sessions). The authors rated only External modes and Reflexive modes. The results show that the preoccupied clients initiate more narrative mode shifts. However, the study was inconclusive as to the differences between the predominance of External mode and Reflexive mode shifts, for dismissing and preoccupied clients.
In 2013, Armstrong conducted two studies with novice trainee therapists. In the first study, he explored the relationship between narrative process modes, the therapeutic alliance and the evaluation of the session by the client (depth, smoothness, positivity and intensity). The results indicated that Internal to External Narrative process mode shifts and External to Reflexive Narrative process mode shifts predicted client-rated smoothness [(β = − 2.41; p = 0.044); (β = 0.30; p = 0.01)] and client positivity [(β = 0.23; p = 0.030); (β = − 2.82; p = 0.021)], respectively. The narrative process shifts not predicted the client-rated alliance. This study emphasizes the importance of emotion-reflective shifts to a therapy session. As Internal to External mode shifts increased in frequency, client-rated smoothness and positivity decreased. In contrast, as the shifts between External Narrative modes and Reflexive Narrative modes increased, client smoothness and client positivity increased.

Therapist Variables
Armstrong' (2013) second study explored the relationship between narrative mode shifts and therapists' facilitative interpersonal skills measured by Facilitative Interpersonal Skills-In Session (Uhlin et al. 2010). Researchers applied this measure to the middle-session or working-phase of each dyad, to explore the influence of other in-session interaction variables and compare the facilitative interpersonal skills' ratings with client-rated measures linked to outcome. The results point out that shifts between External and Reflexive Narrative mode and between External and Internal narrative mode predicted therapist' persuasiveness [(β = − 0.03; p = 0.04); (β = 0.33; p = 0.04)] and empathy [(β = − 0.03; p = 0.03) (β = 0.35; p = 0.01)], respectively. Besides, External to Reflexive Narrative mode shifts predicted alliance bond capacity (β = − 0.03; p = 0.05). This study highlights the importance of the therapists' contribution to Internal -Reflexive mode shifts for facilitating the movement between emotion and meaning-making processes in the session.
The Narrative Process Coding System can also be used to analyze the therapists' narrative mode. Goates-Jones et al. (2009) examined the relationship between the timing and the effectiveness of the therapist response modes (using the Helping Skills System, Hill 2004) and client narrative modes in the exploration stage of Hill's helping skills model. The results demonstrated a significant association between therapist response modes and client narrative modes in at least one dyad [ X 2 (8) = 93.46; p < 0.001]. Besides, 31% of the cases used more Internal Narrative modes than External or Reflexive Narrative modes when therapists used open questions about feelings and reflections of feelings. Therefore, the authors suggest that, perhaps, some clients are more receptive than others, to respond with feelings or need to highlight their emotional arousal to focus their attention on emotions.

Process-Outcome Studies
Most of studies used Narrative-Emotion Process Coding System to look for the relationship between the proportion of narrative markers in different phases of therapy and outcomes. Boritz et al. (2014) were the first to apply Narrative-Emotion Process Coding System 1.0 at three phases of the therapeutic process in Client-Centered Therapy, Emotion-Focused Therapy and Cognitive Therapy. Comparing two recovered patients with two patients not recovered from depression, the researchers explored the role of Narrative-Emotion Process markers in psychotherapeutic change. The results emphasize significant higher proportions of Problem Markers in the unrecovered group, namely Abstract Story, in contrast to the recovered clients, in the middle phase of the therapy [t(28) = 2.02; p = 0.05]. Besides, results show significantly higher proportions of Change Markers in the recovered group [F(2,12) = 4.79; p = 0.03], compared to the unrecovered group, independently of the therapy and in all its phases. Among Transition Markers, Inchoate Storytelling [F(1,6) = 7.041; p = 0.037] and Discovery Story [F(1,6) = 25.113; p = 0.002] are significantly associated with client recovery. Besides, the group recovered in Emotion-Focused Therapy demonstrated a significantly higher proportion of Competing Plotlines [F(1,12) = 5.97; p = 0.031] in the middle phase of therapy than the unrecovered group. In turn, the group recovered in Client-centered Therapy showed a significantly higher proportion of this marker, in the initial [F(1,12) = 26.88; p = 0.0002] and middle phases of therapy [F(1,12) = 10.67; p = 0.0067], compared to the non-recovered group. Researchers suggests that Inchoate Storytelling can pave the way to clients' symbolization and elaboration of their inner experiences and turn Same Old Stories into new views of the self. Thus, an Inchoate Storytelling can be considered a promoter of a Discovery Story, an indicator of the clients' agency capacity to reflect on new emotional experiences and visions of self.
A new research area is addressing the role of narrative and emotion integration when treating clients for Complex Trauma. Initially, Carpenter et al. (2016) applied Narrative-Emotion Process Coding System 1.0 to two initial, middle, and final sessions of Emotion-Focused Therapy for Trauma for two recovered clients vs two unrecovered clients. Researchers randomly selected their sample for Imaging Confrontation and Empathic Exploration. The researchers found higher proportions of Transition Markers, namely, Competing Plotlines in the initial (M = 0.08) and mid (M = 0.11) Emotion-Focused Therapy for Trauma phases, in the recovered group, compared to the unrecovered group (initial: M = 0.05; mid: M = 0.03). In turn, the proportion of Competing Plotlines in the last phase of therapy was higher (M = 0.11) in the unrecovered group, compared to the group that recovered (M = 0.00) from the trauma. Regarding Change Markers, Unexpected Outcome Story and Discovery Story, results show a significant stage by outcome interaction [ X 2 (2) = 7.74; p = 0.021; X 2 (2) = 6.31; p = 0.04], respectively. Thus, in the last phase of therapy, the proportion of Unexpected Outcome Story (M = 0.15) and Discovery Story (M = 0.07) were higher in recovered clients compared to unrecovered clients (M = 0.01, in both markers). Regarding Problem Markers, the unrecovered group showed higher proportions (M = 0.10) than the recovered group (M = 0.02) of Unstoried Emotion. The main conclusion of Carpenter et al. (2016) is that while unchanged trauma clients have difficulty in elaborating their emotional experiences and solving their problems, the changed clients show greater ease in constructing more positive and cohesive narratives, with new associated meanings.
To expand the research of narrative processes to another diagnosis, implementing the Narrative-Emotion Process Coding System 2.0, Macaulay et al. (under review) carried out a study to identify which Narrative-Emotion Process markers occurred in the sessions associated with good vs poor therapeutic outcomes. Their sample suffered from Generalized Anxiety Disorder (GAD) and went through a  [β = 8.06; t(32) = 3.82; p < 0.001], when compared to unchanged clients. The main conclusion of this study is that Competing Plotlines and Inchoate Storytelling seem to be processes of potential change when combined with Reflective Storytelling markers. In this sense, it seems essential for some clients, the access to conflicts and the expression of contrary emotions, as indicated by Angus et al. (2017). Competing Plotlines seems functioning as a sort of catalyst that unlocks the Same Old Storytelling and provides new opportunities for the emergence of more adaptive self, other visions of their emotions and different visions of others. Khattra et al. (2018), using a similar methodology to Macaulay et al. (under review), extended the application of Narrative-Emotion Process Coding System 2.0 to CBT for GAD. The results indicate that, throughout therapy, there were no significant differences in the proportions of Problem Markers between recovered vs unrecovered clients. In turn, Transition Markers [t(32) = − 0.09; p = 0.0028] are more frequent in recovered CBT clients compared to unchanged clients, namely Reflective Storytelling [t(32) = − 5.91; p < 0.001]. Researchers also recorded a double proportion of Change Markers in recovered clients [t(32) = − 0.04; p = 0.0206], compared to unrecovered clients, specifically, Unexpected Outcome Storytelling [t(32) = − 0.03; p = 0.0232]. These results agree with Macaulay et al. (under review), supporting the evidence that successful CBT informed therapies for GAD allow the client to reflect on the events that are highlighting a joint experiential work between therapist and client to counter the ambivalence and typical concern of patients with GAD. On the other hand, less successful therapies do not allow the elaboration of new meanings and the adoption of more adaptive actions and emotions.
Using a similar methodology, Bryntwick (2016) extended her research to a more significant number of participants. The author applied the Narrative-Emotion Process Coding System 2.0 to two early, two middle and two late Emotion-Focused Therapy for Trauma sessions selected from 12 clients (six recovered from the trauma vs six unchanged). The results are consistent with Carpenter et al. (2016) and Boritz et al. (2014). Recovered clients have significant statistically higher proportions of Transition Marker (Inchoate Storytelling) in early and mid-therapy phase [Wald X 2 (1) = 13.05; p = 0.003; Wald X 2 (1) = 12.87; p = 0.003, respectively] and Change Marker (Discovery Storytelling) in mid and latetherapy phases [Wald X 2 (1) = 6.67; p = 0.049; Wald X 2 (1) = 8.84; p = 0.021, respectively]. In contrast, clients who remained unchanged by treatment termination evidenced significantly more Problem Marker at all stages of therapy [early: Wald X 2 (1) = 9.96; p = 0.013; mid: Wald X 2 (1) = 16.02; p = 0.0008; late: Wald X 2 (1) = 13.66; p = 0.002]. The author suggests that the study findings indicate that unchanged clients evidence heightened emotional avoidance throughout treatment phases, while clients who can access, symbolize, and story their inner experiences are more likely to recover (Bryntwick 2016).
Recently, some authors tried to understand how narrative flexibility (shifting between the three categories of Narrative-Emotion process markers or movement between individual Narrative-Emotion Process markers) contributes to therapeutic outcome. Boritz et al. (2016) used the same methodology of the previous study with depressed clients. The results indicate that recovered clients (50.6%) are more likely to have narrative changes than clients who remain depressed (37%) at the end of therapy regardless of the therapeutic modality. Also, the study points out that this probability of change remains constant throughout therapy in recovered clients (β = − 0.0002; t = − 0.53; p = 0.59) and decreases in unrecovered clients (β = − 0.002; t = − 6.63; p < 0.001). On the other hand, unrecovered clients have more difficulty switching from Problem Markers to Change Markers (18.8%) than clients recovered at the end of therapy (33.4%) Wald X 2 = 57.77; p ≤ 0.0001). Results provided preliminary support for the contribution of narrative flexibility to treatment outcome.
The main conclusion to be drawn from the first studies described is that, on the one hand, there appear to be unique patterns of Transition and Change markers that facilitate client change processes in different therapy approaches . Lastly, Boritz et al. (2016) also highlight that greater narrative flexibility between Narrative-Emotion Process markers and individual markers is associated with recovery. Furthermore, recovered clients have a higher likelihood of productive narrative shifting. Examples of productive shifting are: from a Problem marker to either a Transition or Change marker, from Transition marker to another Transition marker, a Transition marker to a Change marker, Change marker to another Change marker, or from No Client Marker to a Transition marker or a Change marker ). On the other hand, unrecovered clients appear to tend to unproductive narrative changes (Bryntwick 2016). There is a significant duration of time in any Narrative-Emotion Process Coding System marker and Narrative-Emotion Process Coding System shifting through session (Wald X 2 = 4.01; p = 0.045) (Boritz et al. 2016). Narrative shifting appears to be particularly tricky in clients who remain unchanged throughout therapy ).

Therapist Variables
Recently, Duarte (2019) sought to identify specific interventions used by therapists that promote productive narrative shifting. She used 20 video recorded sessions of the American Psychological Association (APA) from 12 different therapeutic approaches. All videos were rated with the Narrative Emotion Process Coding System 2.0, in minute units, by two trained raters and productive narrative shifting were identified. Two other trained raters rated the therapist's interventions that preceded each productive shift using the items of the Multitheoretical List of Therapeutic Interventions items (MULTI-60, McCarthy and Barber 2009). The results demonstrated that a wide range of interventions facilitated productive narrative shifts. The most frequently used interventions were: (1) focusing on the here-and-now and emotions, exploring the patient's present feelings and encouraging the patient to talk about avoided emotions; (2) focusing on interpersonal functioning, specifically, identifying the patient's problematic relational patterns; (3) supporting the patient to examine their interpersonal relationships. These results confirm and expand the results of Friendlander et al. (2019) looking for a better understanding of what a therapist can do to promote clients' productive shifting.

New Directions in Narrative Process Coding System and Narrative-Emotion Process Coding System Research Studies
Several authors used Narrative Process Coding System or Narrative-Emotion Process Coding System in intensive case-studies with an exploratory purpose: to explore and analyze narrative processes in different therapeutic modalities (Angus and Bouffard-Bowes 2003;Angus and Hardtke 2006;Laitila et al. 2001;Macaulay and Angus 2019;Paivio and Angus 2017); to understand the change' mechanisms or the narrative sequences of change Friendlander et al. 2018;Levenson et al. 2020), or identifying specific interventions used by therapists that promote productive narrative shifting (Friendlander et al. 2019).

Conclusions and Future Directions
Compared to previous studies, the present empirical review has the benefit of being the first to date that brings together the results of studies applying the Narrative Process Coding System and Narrative-Emotion Process Coding System (1.0 and 2.0). Most studies have been process-outcome studies that looked for the contribution of narrative, emotion, meaning-making patterns, or narrative-emotion markers, to treatment outcomes. In some studies, early, middle, and late phases of therapy were compared, while other studies conducted intensive analyses of Narrative Process Coding System and Narrative-Emotion Process Coding System patterns comparing recovered vs unchanged clients. The research of change patterns in narrative modes and Narrative-Emotion Process markers, using Narrative Process Coding System and Narrative-Emotion Process Coding System, have shown remarkable results, irrespective of therapeutic approaches.
Unchanged clients show a significantly higher proportion of External Narrative modes (Narrative Process Coding System) or Problem Markers (Narrative-Emotion Process markers) than successful clients. Results also reveal that recovered clients evidence a significantly higher proportion of Transition Markers (Narrative-Emotion Process markers) earlier in treatment than unchanged clients, that expresses Transition Markers in late-stage sessions, or not at all. These findings have been cross validated in a range of studies and therapy approaches, addressing hundreds of hours of videotaped sessions. Generalization of these findings remains limited by the relatively small number of participants.
Unchanged clients also evidence significantly less narrative flexibility when compared to recovered clients. The literature emphasizes that although a high proportion of Reflexive Narrative modes and Internal Narrative modes is associated with good therapeutic outcomes, the narrative flexibility between different narrative processes modes contributes uniquely, promoting the therapeutic success. While successful clients are more likely to engage in productive narrative shifts, less successful therapy clients have greater difficulty in doing so.
These results allow concluding that recovered clients begin to engage in productive narrative-emotion processing soon after the initiation of treatment. We may think that brief therapies are specifically short for unchanged patients who need more time to elaborate on their emotional experiences and solve their problems. These clients perhaps need more time to adopt more adaptive emotions and actions, as well as creating more coherent narratives of their history. We may also hypothesize that therapeutic interventions for these patients must be specially carved.
Some studies highlight the importance of Transition Markers as susceptible to promote therapeutic change, namely Competing Plotlines and Inchoate Storytelling. In Competing Plotlines, different views connote incoherence and create tension that leads clients to question. In Inchoate Story, the client focuses inward, contacting emergent experience, searching for new meaning. Inchoate Storytelling can pave the way for the symbolization and elaboration of inner experiences and promote the transformation of the Same Old Stories into new views of the self.
The study of autobiographic memories helped to realize the role of specific memories, their reconsolidation and the linking of these memories and emotion in clients' change.
Recently, researchers conclude that the emotions' exploration and differentiation throughout specific autobiographic memories supports adaptive memories reconsolidation and self-narrative change . Narrative-Emotion Process Coding System 2.0 is an instrument that describes these change processes in session. It identifies storytelling marks that show the degree that client discloses specific autobiographic memories, express, explore, symbolize emotions and bodily felt experiences, reflect on their own or others' minds, actions and stories and coherently integrate change in a meaning personal story .
Since there is substantial evidence that Transition Markers and narrative flexibility are related to good outcome, and can increase through therapy, clinical implications should be drawn helping therapists to enhance their clinical practice. At the end of a session, therapists can usually tell whether or not it was productive. But they have a harder time saying why, and specially knowing what to do to promote productive sessions. Knowledge of the Narrative-Emotion Process Coding System can help therapists to identify, during the session, if the client is in a non-productive process, repeating his problems, making impersonal, and superficial narratives (Problem Markers), or if he is in a process of change (Transition Markers). Knowledge of the various forms of change that should be happening during the session, such as having conflicting views, believes or action (Competing Plotlines), noticing, exploring and symbolizing bodily felt experiences (Inchoate Storytelling), re-experiencing some events (Experiential Story), or exploring general patterns (Reflective Story) can, in turn, guide the therapist to find interventions to bring about that processes.
Regarding future research, we must emphasize the need to repeat the studies but performed with larger samples because most of the samples studied are small. Several studies do not have more than six clients. On the other hand, the researchers applied Narrative Process Coding System or Narrative-Emotion Process Coding System to short-term therapies. Future research needs to replicate the findings in long-term therapies and consider the severity of symptoms to understand how and when the more difficult patients change. Future studies should also focus on the characteristics and interventions of the therapist for client change in Narrative-Emotion Process markers, namely, what specific interventions and features of the therapist promote productive transitions. Only a few studies addressed these topics. Within the same therapeutic approach there is such a wide variety of ways of intervening therapeutically and a wide range of efficacy among therapists that exploring the variety of ways to bring about productive change, and especially with the most challenging cases, should be explored. We need further studies to amplify the promising results of the narrative change in Emotion-Focuded Therapy for Trauma.

Compliance with Ethical Standards
Conflict of interest The authors declare that they have no conflict of interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.