Object Relations Theory Model of Personality Disorders

One of Kernberg’s most influential contributions has been the introduction of an ORT-based approach to assessment, diagnosis, and classification of personality pathology. This diagnostic framework, often referred to as a “structural diagnosis” or “level of personality organization” framework (6), is derived from extensive clinical experience treating patients who have a wide range of personality disorders and builds on a psychoanalytic model of the mind and mental functioning. Diagnosis focuses on characterizing personality pathology, first in terms of severity of impairment and then in terms of personality style or personality disorder type. Severity of personality pathology is defined structurally in terms of the degree of integration and hierarchical organization of “psychological structures,” where structures are conceptualized as the latent variables or processes that organize descriptive aspects of personality functioning. For example, the psychological structure called “identity” organizes the individual’s sense of self, sense of others, and capacity for goal directedness. Identity formation can be identified by evaluating its expression in these functions. Similarly, the structure called “object relations” refers to the working models of relatedness that organize interpersonal functioning. One can assess object relations by evaluating the quality of interpersonal relationships and internal expectations of relationships.

The ORT model focuses on six core psychological structures as the organizers of self- and interpersonal functioning: identity, defenses, object relations, moral values, quality and management of aggression, and reality testing. The model was formally introduced in the 1960s (6) and has been further developed during the ensuing decades through an accumulation of clinical research on the evaluation and treatment of personality pathology; see the article by Clarkin et al. (7) in this special issue for more details. Level of organization and severity of pathology have proven to be robust predictors of clinical outcome (8), and a structural diagnosis can be used to guide treatment planning and to anticipate developments in treatment (9).

The ORT model identifies three overall levels of severity of pathological personality organization. Each level of severity is characterized according to the nature of the identity formation (consolidated vs. failure of consolidation, capacity for investment in relationships, work, internal values, and personal interests and goal directedness vs. impairment), defenses (higher-level repression-based and mature defenses vs. splitting-based defenses), and reality testing (intact and stable, intact but vulnerable and nonpsychotic, and frank failure of reality testing and psychotic), and is further refined on the basis of the quality of object relations (object relations in depth vs. superficial, need-fulfilling, or exploitative), the nature of moral functioning (internalized and stable vs. inconsistent or lacking), and the quality of aggression (integrated and well modulated vs. poorly integrated and poorly modulated). At the mildest level of severity is the neurotic level of personality organization, differentiating pathological personality functioning from subclinical and healthy functioning and characterized by consolidated identity, predominance of repression-based and mature defenses, and intact and stable reality testing. At a greater level of severity is the borderline level of personality organization, corresponding to all clinical personality disorders and characterized by failure of identity consolidation, predominance of splitting-based defenses, and intact but vulnerable reality testing. The borderline level of personality organization is further divided into severity levels of high (mild personality disorder), middle (moderate personality disorder), and low (extreme personality disorder). Level of organization is determined by the six domains that are central to differentiating healthy from pathological personality functioning, with specific attention given to quality of object relations, moral values, and aggression to differentiate among the levels (see Table 1). More severe than the borderline personality level of organization is the psychotic level of personality organization, corresponding to a subset of psychotic disorders that may superficially present as personality pathology, which is characterized by failure of identity consolidation, predominance of splitting-based and psychotic-level defenses, and frank failure of reality testing. By focusing on six clinically and theoretically interdependent domains of functioning, the ORT model provides a highly nuanced approach to assessment and classification of personality pathology that is used to guide treatment planning and predict clinical course (9) and that has been used to develop an evidence-based treatment for severe personality disorders (10).

Although the ORT model (5, 11) predated the AMPD by several decades, considerable overlap exists between the two models. Both embrace a dimensional, in contrast to a categorical, approach to personality disorder diagnosis; both identify self- and interpersonal functioning as core features of personality functioning; both emphasize severity of impairment in the classification of pathology across the range of personality disorder presentations; and both identify three general levels of severity of personality disorders, a framework also adopted in the ICD-11 (12). The convergence of these models suggests the construct validity of recent developments in the understanding and classification of personality pathology that focus on dimensional assessment of severity of impairment of self- and interpersonal functioning.

Whereas significant overlap exists between the ORT model and the AMPD, and it is possible to move between AMPD and ORT classifications with relative ease (see Tables 1 and 2), important distinctions between the two models are highly relevant to clinical work. Notably, reflecting its position within an integrated psychodynamic model of personality functioning and treatment, the ORT model is unique in identifying defensive operations, moral functioning, management of aggression, and reality testing as central domains of personality functioning that are integral to self- and interpersonal functioning and to assessing severity of personality pathology. In contrast, the AMPD characterizes moral functioning, aggression, and psychoticism as traits, rather than central domains of personality functioning, and entirely omits consideration of defenses. Furthermore, reflecting its foundation in psychodynamic theory, the ORT model differs in that it adopts a developmental perspective, in which level of personality organization has its roots in early attachment relationships.

The Structured Interview of Personality Organization (STIPO) (13) was developed to address the need for a standardized approach to the assessment of personality disorders in research and clinical settings. Built on the basis of experience with the Inventory of Personality Organization (14)—a self-report instrument widely used in clinical assessment and screening—the STIPO was created as a semistructured interview to evaluate personality organization through a series of standard questions, follow-up probes, and scoring guidelines in order to ensure reliability. Structured questions, coupled with structured follow-up probes, allow the interviewer to evaluate vague or imprecise patient answers, and the structured scoring system allows the possibility of investigations of the reliability of administration and scoring. The STIPO consists of 100 questions across seven domains: identity consolidation versus identity pathology, use of primitive defenses, quality of object relations, coping strategies, use of self-directed and other-directed aggression, moral values, and reality testing. The aim of the STIPO is to obtain clinical information relevant to a diagnosis of personality structure according to the ORT model.

In support of the ORT model’s construct validity, research has highlighted the specificity of the STIPO dimensions in determining type and severity of personality pathology. Stern et al.’s study (15) revealed unique contributions of the STIPO identity domain to measures of negative affect as well as DSM cluster A personality disorders, whereas the STIPO primitive defenses domain has been associated with multiple measures of aggression as well as with DSM cluster B personality disorders. Both sets of findings are consistent with Kernberg’s underlying psychodynamic model of personality functioning (5). Di Pierro and colleagues (16) compared the STIPO profiles of three groups: inpatients with co-occurring disorders, psychiatric outpatients, and healthy control participants. In findings consistent with the ORT model, those authors found that a poorly integrated identity; difficulties in the capacity to invest in relationships, work, internal values, and personal interests; poorly integrated moral values; and high levels of self-directed and other-directed aggression were characteristics of patients with severe personality pathology.

More recent studies have examined the empirical relationship between structural diagnosis obtained through the ORT model with the STIPO and diagnosis obtained through the AMPD. Preti et al. (17) showed that clinical information gathered through the STIPO could successfully be used by untrained and clinically inexperienced students to rate severity of pathology among patients with personality disorders on the Level of Personality Functioning Scale (LPFS) (2) of the AMPD. Di Pierro and colleagues (18) showed that clinicians could use the STIPO to make ratings on the LPFS that differentiated patients with personality disorders, psychiatric outpatients, and healthy control participants. Kampe et al. (19) investigated the convergence between the Structured Clinical Interview for the DSM-5 AMPD (SCID-5-AMPD) Module 1 and the STIPO. They found significant correlations between overall scores and domain scales of the two interviews, supporting convergent validity. Ratings from both interviews showed high correlations among criteria of clinical severity, including number of suicide attempts, frequency of psychiatric hospitalization, and severity of ICD-10 diagnoses. Notably, all the STIPO domains correlated with clinical criteria for pathology, although the domains of identity and primitive defenses correlated most strongly with the frequency of psychiatric hospitalization, considered a metric of personality disorder burden.

These findings were consistent with those of Stern et al. (15), as well as those of Hörz et al. (20), who found that assessment of primitive defenses is central to the evaluation of cluster B disorders. In examining partial correlations, Kampe et al. (19) reported that the STIPO domains of identity, primitive defenses, aggression, and reality testing showed unique significant correlations with clinical criteria, and the authors concluded that these domains contain specific content, reflecting features of personality pathology linked to clinical severity, that is not operationalized in the LPFS. Conversely, only the SCID-5-AMPD score for the empathy domain was significantly correlated with severity of diagnosis when the analyses were controlled for STIPO total score.

Diagnosis of level of personality organization has direct and reliable implications for case formulation and treatment planning (5, 9). For individuals with subsyndromal personality pathology (i.e., those organized at the neurotic level), psychodynamic treatment is organized around the goal of reducing rigidity in personality functioning (21). These individuals have a very favorable prognosis and can benefit from relatively unstructured treatments. They typically do not have difficulty establishing and maintaining a therapeutic alliance, and transference distortions tend to be slow in developing, consistent, and subtle. For individuals with a full, syndromal personality disorder (i.e., those organized at a borderline level), the treatment goal is to ameliorate identity pathology and promote normal identity consolidation (9, 10). Progress toward these goals will manifest in improved relationships with friends, romantic partners, and work associates and in the relationship between the patient and therapist. More progress is expected among those with high and middle levels of personality organization than among those with a low level of organization. In contrast to those with a neurotic level of organization, individuals with a borderline level of organization, particularly those in the low borderline spectrum, require a highly structured treatment setting. They have great difficulty establishing and maintaining a therapeutic alliance; transference distortions develop rapidly and are highly affectively charged and extreme, often leading to disruption of treatment. Across the range of severity, the overall therapeutic approach is to explore the patient’s representations of self in relation to others as they are activated in interpersonal relationships with significant others and with the clinician (9).