Psychotherapeutic case formulation: Plan analysis for narcissistic personality disorder.

BACKGROUND: One of the relevant case formulation methods for personality difficulties is plan analysis. The present study aimed at delivering a prototypical plan analysis for clients presenting with a diagnosis of narcissistic personality disorder (NPD). The sample consisted of 14 participants diagnosed with an NPD. Based on audio clinical material, we developed 14 individual plan analyses that we then merged into a single prototypical plan analysis. For explorative purposes, we ran an ordinary least squares regression model to predict the narcissistic symptoms severity (NAR) measured on a scale of 1-7 of the 14 clients by the presence (respectively absence) of certain plans in their individual plan analysis. The synthesis revealed that clients with pathological narcissism share common basic motives. Results of the regression model reveal that the presence of the plan 'be strong' reduces the NAR scale by 1.52 points (p = 0.011). DISCUSSION: In the treatment of psychological disorders, precise case formulations allow therapists for making clinically appropriate decision, personalizing the intervention and gaining insight into the client's subjective experience. In the prototypical plan structure we developed for NPD, clients strive to strengthen their self-esteem and avoid loss of control, criticism and confrontation as well as to get support, understanding and solidarity. When beginning psychotherapy with a client presenting with NPD, the therapist can use these plans as valuable information to help writing tailored, and therefore more efficient, case formulations for their patients presenting with an NPD.

INTRODUCTION

Over the past years, research in the field of personality disorders reached the consensus that the clinical presentation of clients with pathological narcissism is mainly characterized by its heterogeneity (Bender, 2012; Caligor et al., 2015; Ronningstam, 2020). In an attempt to understand and structure this clinical variability, experts relied on a categorical approach—as embodied by the Diagnostic and Statistical Manual of Mental Disorders paradigm and its diagnosis of narcissistic personality disorder (NPD). Yet, the sole reliance on standardized diagnostic criteria focusing on the ‘overt type’ (characterized by the grandiose manifestation of the pathology) failed to cover the core psychological features of the disorder including vulnerable self‐esteem, feelings of inferiority, emptiness and boredom as well as affective reactivity and distress (Levy, 2012; Ogrodniczuk, 2013; Ronningstam, 2009; Ronningstam, 2011). As a result, this categorical approach did not encompass the heterogeneity of pathological narcissism (Kernberg, 2009; Pincus et al., 2009; Roberts & Huprich, 2012; Skodol et al., 2014).

An alternative route to understanding and structuring the clinical variability associated with pathological narcissism is to focus on its dimensionality. Clinicians and scholars acknowledge that narcissistic phenomena are not strictly pathological but that they are an essential part of general personality functioning. Narcissism has its roots in normal development during which it can be disturbed to varying degrees by environmental stress and failures of nurturing (Bender, 2012) and ranges from ‘healthy and exaggerated to pathological, including high and low functioning NPD, as well as severe forms with malignant or psychopathic functioning’ (Ronningstam, 2020, p.2).

Recent research suggests that pathological narcissism is associated with significant functional impairment and psychosocial disability as well as decreased life satisfaction and lower quality of life (for a brief review of relevant investigations on the subject and putative explanations, see Ellison et al., 2020), making the accurate diagnosis, effective case formulation and the development of tailored interventions a priority. Pathological narcissism is associated with the prognosis of difficulties in building a good therapeutic relationship and in the success of a therapy (Caligor et al., 2015; Levy & Clarkin, 2006; Ronningstam, 2017). In order to understand and explain the heterogeneity in personality disorders, case formulations may be crucial (Eells, 2011; Kramer, 2019). They link ‘the clinical theory with the unique case, and the general with the particular’ (Kramer, 2019), thereby providing clinicians with a tool to integrate clinical observations with the explanatory model, with the aim of personalizing psychotherapy. Case formulation appears thus a necessary step to understand qualitatively the heterogeneity observed in clients with NPD.

Examples of case conceptualizations methods may focus on emotional experiences (Strating & Pascual‐Leone, 2019) rely on cognitive and behavioural theory (Sturmey & McMurran, 2019), psychodynamic theory (Levy et al., 2019) or clarification‐oriented theory (Eells, 2011; Sachse, 2019). Regardless of the method, they all enable clinicians to elaborate fundamental therapeutic hypotheses to guide their interventions.

Among these methodologies, plan analysis is a case conceptualization instrument in psychotherapy developed by Grawe and Dziewas (1978) and Caspar (2007). Historically, its origin traces back to the 1970s when Grawe observed that so‐called difficult clients—many would nowadays likely receive a personality disorder diagnosis—would not engage in therapy or struggled with their therapist despite irreproachable technique from their part. Based on the concept of Plan as coined by Miller et al. (1960) and on the assumption that understanding and psychotherapeutic care of clients could only succeed if their motivational structure was understood, Grawe and Dziewas (1978) developed the vertical behaviour analysis in complement to the horizontal analysis of behaviour that explains the sequential unfolding of stimuli and responses on the time axis. Vertical behaviour analysis emphasized the importance of identifying and understanding clients' important motives and how they related with instrumentally relevant behaviours. The assumption was that doing so should lead to a simplified representation of the complexity and uniqueness of clients' experience and behaviour(s). Vertical behaviour analysis later developed into plan analysis.

Plan analysis incorporates the conceptualization of thoughts, beliefs and emotions by taking verbal and para/non‐verbal aspects into account. It is compatible with most therapeutic approaches. It is used to develop an individualized case conceptualization, which may serve therapy planning and aims at guiding the therapeutic process and improving the relationship between therapists and client. Central to plan analysis is the assumption that behaviours are repeated and consolidated into implicit structures of action organized to serve a specific purpose. Even instrumental behaviours are not necessarily conscious (Caspar, 2019), as exemplified by the case of Charles, a 30‐year‐old psychotherapy client diagnosed with NPD, who failed his math studies and who presented himself to others in a grandiose fashion by insisting that he may eventually solve a major mathematical problem. For this clinical case, it appears that the self‐presentation ‘show that you are capable of solving a still unsolved math problem’ serves the higher Plans of ‘present as a genius’ and ‘avoid admitting your failures’, which may serve to strengthen his self‐esteem in interpersonal situations. Another case is the one of Barbara, a 45‐year‐old psychotherapy client diagnosed with NPD, who works as a nurse. Facing her current psychotherapist, she describes herself as a ‘therapist too’, not without expressing contempt. For this clinical case, it appears that her self‐presentation ‘show that you are competent’ and ‘explain that you have high therapeutic skills’ may serve the higher Plans of ‘present as competent’ and ‘avoid that the therapist asks intrusive questions’, which may serve to strengthen both her integrity and self‐esteem in interpersonal situations. For each of these individual case conceptualisations, the client's individual plans as hypothetically inferred by the therapist are taken into account. Such plans are units consisting of a motivational component (motive, purpose, goal) and one or more means to achieve these goals.

Plan analysis is a useful tool to facilitate the development of meaningful and coherent explanations or conceptualizations of the client's symptoms, disorders and problems. A prototypical plan structure aims at assisting psychotherapy trainees in the elaboration of a case conceptualization/formulation and refers to a framework outlining what is frequently observed among clients presenting with a particular diagnosis or clinical problem and can therefore serve as default hypotheses. Such a prototypical plan structure has already been developed for the neighbouring disorder of borderline personality disorder (Berthoud et al., 2013) and several other problems. In their study, the authors could highlight two main prototypical tendencies (“dependent” and “autonomous”) along with the plans aiming at emotion regulation in both subtypes. They also found that all clients in their sample intended to seek support (‘make sure you get support’). They also discussed the prototypical plan structure's subsequent clinical implications, namely, for the building of an individualized, or motive oriented, therapeutic relationship (Caspar, 2019). For instance, ‘facing a client with the “make sure you get support” Plan activated, the therapist, after deciding if this Plan serves the basic motives “get healed”, “avoid being alone” or “stay in control” (and/or any other motive involved), proactively focuses on this motive. If the motive is “avoid being alone”, the therapist will have a soothing non‐verbal attitude and will assure the patient that the therapist does not intend to abandon the patient and conveys acceptance to the patient as a person’ (Berthoud et al., 2013). More generally, motive‐oriented therapeutic relationship (MOTR) has proven to be a promising intervention, as based on plan analysis, in treatments for borderline personality disorder. In two randomized controlled trials (Kramer et al., 2011; Kramer, Kolly, et al., 2014), small but consistent outcome advantages in a brief treatment have been found favouring MOTR. Several process advantages, for example, a stronger session‐by‐session evolution of the therapeutic alliance, were also observed. Apart from a case study, no evidence exists for the neighbouring disorder of NPD and pathological narcissism more generally (Kramer, Berthoud, et al., 2014).

The aim of this article is to contribute to the existing literature by developing a prototypical plan structure for NPD using the plan analysis approach (Caspar, 2019). Ultimately, the goal is to provide a basis that will help elaborating NPD case formulations more easily in order to optimize treatment planning and eventually enhancing treatments.

METHODS

Sample

A total of 14 clients at a German outpatient clinic were included in this study. The client's ages ranged between 25 and 58 years old with a mean of 40.36 (SD = 10.49). Six of them were women (43%) and eight men (57%). All of them fulfilled the SCID‐II (First et al., 1995) criteria for an NPD. In addition, we assessed dimensionally the narcissistic symptom severity (NAR) of each client on an ordinal scale ranging from 1 (mild symptoms) to 7 (extremely severe and pervasive symptoms). In line with the SCID‐II, values of 2 (symptoms are present) and above indicate a clinically relevant narcissistic symptomatology and the presence of the disorder. In this sample, the values varied between 2 and 5 (M = 3.43, SD = 1.09). Aside from the NPD diagnoses, six clients had a co‐morbid diagnosis of major depression (43%), three clients had a diagnosis of substance abuse disorder (21%), and two has somatoform disorder (14%). On Axis 2, two clients were also diagnosed with a histrionic personality disorder (14%). In this sample, we also evaluated the participant's depressive symptomatology using the ‘Beck Depression Inventory II’ (BDI‐II; Beck et al., 1996). The German translation has satisfactory validity (r = .68 to.89) and reliability (internal consistence: .89 ≤ α ≤ .94) coefficients. The BDI‐II's values in this sample ranged from 1 to 39 (M = 19.38, SD = 13.07).

Plan analysis (Caspar 2019)

To elaborate a plan analysis, we need to follow three steps: First, watch the video recording from the psychotherapy session or listen to its audio recording (given the importance of para/non‐verbal aspects, video‐recordings should always be preferred if they are available). Because in this study only audio recordings and no video recordings of the clients were available as source material, it is important to note that the present analysis focuses on para‐verbal and verbal aspects of behaviour and experiences. In the second step, potentially relevant synthetic information is noted, so‐called extensions. These are based both on verbal (and possibly para‐ and/or non‐verbal), as well as information that appears necessary for a better understanding of the client's behaviour. These intermediate steps are important to make the link from the observable presumably instrumental behaviours to the development of a plan structure as transparent and comprehensible as possible. Then, based on the information extracted, the plans are developed and the client's individualized plan structure is created. The plans are described in the imperative (e.g. ‘present as particularly competent’), whereas behaviours are formulated in the indicative (e.g. ‘expresses contempt of the person of the therapist’, for the case of Barbara in the introduction). This results in a hierarchical structure where lower plans are intermediate motives serving to achieve the ones (basic motives) at the top.

In the present study, two raters applied the three aforementioned steps for all 14 clients and each came up with 14 individualized plan structures. The inter‐rater reliability was determined using Benkert's method (Benkert, 1997) on a randomly 14% of the data (two cases). In these two cases, the 10 most important plans of a client from the first plan structure (selected by the rater) were compared with all plans of the second plan structure. In order to determine a value for each of these 10 selected plans, the following matching criteria were applied, and points were distributed accordingly: 1 point was awarded if the plan itself occurs in both raters' structures, 2 points if the higher level plans in the hierarchy matched and 2 points if the lower level ones matched. Finally, the maximum score of 5 points for a plan, with the raters' complete agreement. The average agreement should be at least 60% in order to be considered sufficient (Benkert, 1997).

Procedure

Once the 14 individual plan Analyses' were formulated and drawn on paper, we elaborated a synthesized plan structure (Berthoud et al., 2013; Kramer et al., 2009):

Plans of different clients whose meaning content overlapped to a sufficient extent were combined into one formulation item.

All clients' plans and motives (excluding observed behaviours) were grouped into a single list with occurrence of each plan (ranging between 1 and 14; see Supporting Information). Based on the standard of five (Berthoud et al., 2013) and in an attempt to find an acceptable trade‐off between sensitivity and specificity, we only included plans present in at least four distinct clients in the prototypical structure.

A thematic analysis of these “prototypical plans” revealed groupings and instrumental connections between them so that a single prototype plan structure could be created.

RESULTS

Inter‐rater plan analysis reliability

In our study, the reliability of the plan structure achieved 60.5%.

Prototypical plan structure

Once we regrouped all the 14 clients' initial plans into semantically identical units, we elaborated a list of 98 plans (see Appendix S1). Out of those, 29 prototypical plans were more frequent than the abovementioned criterion of 4, whereas the other 69 plans' frequency did not make the cut. Figure 1 shows the resulting NPD prototypical plan structure (behaviours are left out). Drawn lines represent a direct instrumental relationship between plans and motives. The structure is a vertical hierarchy in which lower level plans serve higher order plans, goals and motives. The numbers in brackets indicate the frequency of plans' occurrence in the sample. Certain plans are highly prevalent in the prototypical plan structure of NPD: Namely, on the higher motive level, ‘strengthen self‐esteem’ appears in every individual plan structure, ‘avoid loss of control’ in 12 (86%), ‘get support, understanding and solidarity’ in 10 (71%), ‘establish bond/relationship’ in 9 (64%), ‘get recognition and appreciation’ in 7 (50%) and ‘maintain integrity’ in 6 (43%). On the lower plan levels, ‘avoid criticism and confrontation’ appears in 10 (71%) individual plan analyses, whereas ‘show yourself especially reflected and accessible’, ‘show your skills’ and ‘show yourself independent’ appear in 8 (57%) of them.

FIGURE 1

Prototypical plan structure for narcissistic personality disorder (n = 14). In brackets is the number representing the plan's presence within the structure of a client

A plan structure has implications for treatment planning and relationship building by the therapist, in particular by using the MOTR. If Figure 1 would represent the case formulation of an individual client, then the therapist could use it to understand the client's individualized inter‐ and intrapersonal functioning in order to create a tailored idiosyncratically safe therapeutic relationship (Kramer, Berthoud, et al., 2014). To achieve and foster this, the therapist should choose the lowest plan in the structure that is also acceptable in terms of how it relates to relationship and cooperation within psychotherapy.

For explorative purposes, we ran an ordinary least squares regression model to predict NAR of the 14 clients by the presence of plans that appeared in 7 clients (n = 5). Results reveal that the presence of the plan ‘be strong’ reduces the NAR scale by 1.52 points (p = 0.011) (Table 1).

TABLE 1

Summary of linear regressions predicting NAR

Predictors	NAR
	Estimates	CI	p
(intercept)	4.59	3.08–6.10	<0.001
be_strong	−1.52	−2.59 to −0.45	0.011
show_feelings	−0.52	−1.78–0.75	0.376
be_likeable	−0.84	−2.09–0.41	0.161
get_recognition	0.13	−1.04–1.30	0.806
avoid_attack	0.42	−0.78–1.62	0.442
Observations	14
R2/R2 adjusted	0.640/0.415
F(5, 8)	2.842

Note: Bold emphasis indicates significant results.

DISCUSSION

In the field of research of personality disorders, there exist two competing approaches: the categorical and the dimensional one. The problem in this debate is that both perspectives do not provide sufficient support for the essential personalization of psychotherapy. A third perspective integrating both approaches within an evidence‐based case formulation approach is needed in order to provide clear guidance to the practicing clinician working in the field of personality disorders. A case formulation approach has the additional advantage to take into account clinically essential idiosyncratic information from each client and manages to accommodate and integrate both perspectives.

Pathological narcissism is a particular case in point. The present study has as objective to develop a prototypical case formulation template using the qualitative methodology of plan analysis. It appears that several aspects of the present NPD prototypical plan structure are in line with the existing literature. The structure encompasses basic motives and plans consistent with dimensions of pathological narcissism ranging from the conceptual grandiose type (‘get recognition and appreciation’, ‘strengthen self‐esteem’, ‘avoid inferiority’, ‘show your skills’) to the vulnerable one (‘get support, understanding and solidarity’, ‘establish bond/relationship’, ‘show how bad you feel’) and the presumed core features (‘make yourself important’, ‘be something special’).

Prototypical plan structures are designed for education and research purposes. When used in the analysis of individual patients, they can be used as default assumptions that can speed up an individual analysis but have to be verified with the individual patient while self‐critically controlling a possible confirmation bias. For whole groups of patients, such as NPD patients, they have clinical implications mainly considering how crucial the alliance building is in the treatment of NPD (Adler, 2000; Bender, 2005; Ronningstam, 2012). They provide a good basis for training psychotherapists to write a case conceptualization and implement a corresponding individualized complementary, or MOTR (Caspar, 2007; Grawe, 1992). As reported by Ronningstam (2017) in pursuance of admiration and heightened self‐esteem, clients presenting with an NPD may use emotion‐regulating strategies (‘avoid negative feelings’, ‘do not offer any surface to be attacked’, ‘avoid getting hurt’). To achieve their various goals of grandiosity and/or bonding, they may also use interpersonal control strategies (Caligor et al., 2015) such as ‘get therapists on your side’, ‘show that you have been treated unjustly’, ‘make sure that you are taken seriously’ or ‘show yourself especially reflected and accessible’. Using the MOTR concept, the therapist can look for the plans that do not threaten or limit the therapeutic alliance. Upper plans do not, by definition, threaten nor limit the therapeutic procedure, but the therapist should look for the lowest acceptable motive in the structure and adjust therapy accordingly.

In light of our explorative analyses, the therapist could focus on the plan ‘be strong’ and develop complementary techniques to foster it. Indeed, because it appears to be a predictor of a less severe narcissistic symptomatology, working on the fulfilment of this motive on a relationship level could prove useful to reduce NPD severity. The therapist could have a reinforcing attitude, highlighting the strength and competences of the client when faced with adversity in order to let him/her know that he/she is strong. Clinical implications of the use of plan analysis facing clients with pathological narcissism are numerous. An illustration of alliance‐building moment‐by‐moment processes has been provided in a case study by Kramer, Berthoud, et al. (2014). In this case, the client named Mark presented a set of plans (‘present yourself as responsible’, ‘present as a flawless employee’ and ‘present as someone who has success’) serving the basic needs of maintenance of control and of a positive self‐image. Using MOTR principles, a therapist may productively underline that Mark is a good father and a good employee (both serving the need of positive self‐image) and/or convey this on a non‐verbal level.

For Charles, the failed math student mentioned in the introduction, an MOTR consistent intervention may consist in highlighting some elements of the extraordinary competencies this math student may have (i.e. complementary to strengthen a good self‐esteem, by showing your skills and present as strong), despite the difficulty he encountered and to express clear acceptance of his value unconditioned to his performances. The therapist may monitor his affective reaction to such an intervention and, as soon as he shows signs of readiness, offer genuine astonishment about his failure given these extraordinary skills. The latter may then shift the therapeutic discussion towards an effective problem solution, which should enhance the therapeutic collaboration between the client and the therapist. For Barbara, the client expressing contempt in the therapeutic relationship and considering herself a ‘therapist too’, the therapist could behave in a complementary fashion to plans like ‘present as competent’ while at the same time avoiding to label her problem. He/she may for example offer a discussion ‘among therapists’ by saying: ‘As you know, as a therapist, it is important to continually improve oneself, so would you be interested in using this therapy to becoming an even more effective therapist?’, depending on the readiness of the client. For both Charles and Barbara, we would assume that these offers of collaboration—all consistent with the client's acceptable plans (yet still specific enough to each individual)—may increase their collaboration and strengthen the therapeutic alliance. Empirical research should examine this hypothesis for clients with pathological narcissism.

The study presented here has several limitations. Both the small sample size and absence of video material hinder reporting the complexity and heterogeneity of the NPD's clinical presentations. However, despite these limitations, the present prototypical plan structure still has the potential to inform clinicians when dealing with clients presenting with NPD and helping them come up with individualized case formulations to tailor the treatment to the (motives) need of their client. Future research should try to replicate this methodology on a bigger sample to investigate the validity of this NPD prototypical structure and investigate ‘be strong’ as a predictor of symptom severity.

ETHICS STATEMENT

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

CONFLICT OF INTEREST

They were no conflicts of interest.

Data S1. Supporting information

Click here for additional data file.