Elsevier

Epilepsy & Behavior

Volume 34, May 2014, Pages 92-98
Epilepsy & Behavior

Elevation discrepancies between MMPI-2 clinical and MMPI-2-RF restructured clinical (RC) scales in people with seizure disorders

https://doi.org/10.1016/j.yebeh.2014.03.016Get rights and content

Highlights

  • The popular MMPI-2 has recently been revised in the form of the MMPI-2-RF.

  • There is little research to date on the MMPI-2-RF in seizure disorders.

  • Psychopathology scale elevations are compared across the MMPI-2 and the MMPI-2-RF.

  • Variance explained in MMPI-2 scores by relevant and irrelevant sources is also examined.

  • It is shown that the MMPI-2-RF provides better assessment of psychopathology.

Abstract

Purpose

People with seizure disorders experience elevated rates of psychopathology, often undiagnosed and untreated. Accurate diagnosis of psychopathology remains an important goal of quality health care for people with seizure disorders. One of the most widely used dimensional measures of psychopathology is the Minnesota Multiphasic Personality Inventory—Second Edition (MMPI-2). Research in heterogeneous mental health samples suggests that the 2008 revision of this measure, the Minnesota Multiphasic Personality Inventory—Second Edition—Restructured Form (MMPI-2-RF), offers better construct fidelity and more cost-effective administration. This study seeks to extend research on MMPI-2-RF scale elevations to a sample of people with seizure disorders.

Methods

In a consecutive, heterogeneous sample of people with seizure disorders, MMPI-2 and MMPI-2-RF scores were compared in terms of categorical classification agreement (clinically elevated versus not clinically elevated). Scores were also compared in terms of variance attributable to diagnosis-specific items, general demoralization, subtle items, social desirability, and demographic factors.

Key findings

Scores on MMPI-2 and MMPI-2-RF provided a statistically significant level of agreement between corresponding clinical diagnostic scales ranging from 68% to 84%. Most classification disagreement was attributable to MMPI-2 clinical scale elevations when MMPI-2-RF scales were not elevated. Regression analysis supported the interpretation that general demoralization, subtle items, social desirability, and demographic factors led to MMPI-2 clinical scale elevations.

Significance

The results provide evidence that in the context of strong psychopathology classification agreement, the MMPI-2-RF restructured clinical scales provide better construct fidelity compared with the more trait heterogeneous MMPI-2 clinical scales. These results should encourage clinicians to use the MMPI-2 Restructured Form (MMPI-2-RF) for improved psychopathology assessment compared with the MMPI-2 in patients with seizure disorders.

Introduction

People who have epilepsy are at an increased risk of behavioral, emotional, and cognitive disorders compared with the general population [1]. Estimated rates of psychopathology in patients with epilepsy vary across studies, but many report psychopathology diagnoses in approximately 19–48% of adults [2]. Comorbid psychopathology in people with epilepsy has important therapeutic and diagnostic implications [3] including decreased quality of life, increased disability, increased medical expenses, and increased utilization of medical services [1], [4].

However, comorbid psychopathology in people with epilepsy often remains undiagnosed [1]. For example, a study by Hermann et al. [4] found that only 55% of patients with epilepsy and a major depressive episode had been accurately diagnosed, and only 11% were receiving adequate treatment for their depression. Without accurate diagnosis, patients may not be offered appropriate treatment [4]. There are many reasons why psychopathology may remain underrecognized, a lack of screening measures with adequate sensitivity and specificity being one [5].

The Minnesota Multiphasic Personality Inventory–Second Edition (MMPI-2) is commonly used for psychological assessment in diverse patient populations including patients with seizure disorders [6], [7]. In 2008, the MMPI-2 underwent a significant revision, culminating in the release of the MMPI-2 Restructured Form (MMPI-2-RF). The MMPI-2-RF was designed to represent the ‘clinically significant substance of the MMPI-2 with a comprehensive set of psychometrically adequate measures’ with the ultimate aim of developing a measure with improved efficiency and enhanced construct validity [8], [9]. The MMPI-2-RF is considerably shorter than the MMPI-2, making administration less taxing for patients [10]. Such restructuring has the potential to improve the usefulness of the MMPI-2-RF, however, the utility of the MMPI-2 Restructured Form remains to be established in the wide array of populations in which the MMPI-2 has been used. As with transitions between versions of other widely used tests, clinicians may be disinclined to make the transition to the MMPI-2-RF without published evidence of criterion-related validity in relevant populations [11].

Recent research suggests that the revision of the MMPI-2 succeeded in creating a measure with enhanced construct validity and improved efficiency, and differences have been noted in the elevation of key diagnostic scores, namely, the restructured clinical (RC) scales of the MMPI-2-RF when compared with the clinical scales of the MMPI-2. Both Van Der Heijden et al. [12] and Osberg et al. [13] have observed a lowered elevation pattern in the RC scales compared with the clinical scales. In view of the recommendation that the RC scales initially be used to aid in the interpretation of equivalent MMPI-2 clinical scales due to their enhanced convergent and discriminant validity, an interpretive dilemma arises when there are noticeable elevation discrepancies between equivalent scales [14]. Understanding the origin of such discrepancies is important in the clinical application of these scales.

Sellbom et al. [15] have suggested three variables which may contribute to these discrepancies in elevation between equivalent clinical and RC scales. Firstly, the presence of a ‘first (or common) factor’ of nonspecific psychological distress pervading the MMPI-2 clinical scales may contribute to the elevation of the clinical scales over and above elevations in the RC scales. This first factor, a result of the empirical keying approach used to create the clinical scales of the MMPI-2, is thought to reflect a dimension of generalized emotional distress and unhappiness not related to any specific psychopathological diagnoses. During the construction of the RC scales, items considered to reflect this first factor were removed and grouped into a new scale called “demoralization” (RCd: for a full list of scale abbreviations used in the text, see Table 1). As a consequence, Sellbom et al. [15] have suggested that when a clinical scale is elevated and the corresponding RC scale is not, elevation on the RCd scale may be contributing to these differences.

A second set of variables, in addition to demoralization, involves the inclusion of so-called subtle items in the clinical scales which may also contribute to elevation differences [15]. A series of subtle scales were developed by Wiener [16] for several of the original MMPI clinical scales. The subtle items that formed the basis of these scales were originally thought to be less susceptible to falsified responding; however, they have since been shown to have no direct relationship to the constructs targeted by the clinical scales and were, thus, removed from the RC scales [17].

Finally, Sellbom et al. [15] have suggested that the routine application of K-correction to several of the clinical scales may also contribute to elevation differences. The K-correction process was intended to correct for the effects of defensiveness by adding a percentage of the K-raw score to the clinical scale raw score. However, the majority of research studies indicate that K-correction does not increase the validity of the clinical scales and may even have the opposite effect [18].

The investigation conducted by Sellbom et al. [15] in two large and heterogeneous mental health clinic samples demonstrated that demoralization, subtle items, and K-correction substantially contributed to the observed elevation differences between corresponding clinical and RC scales. It remains to be seen, however, if these sources of elevation discrepancy between the MMPI-2 and the MMPI-2-RF generalize to other clinical populations. To date, there has been no study of equivalence of the primary diagnostic scales from these two versions of the MMPI in a sample of patients with seizure disorders. In addition, the study by Sellbom and colleagues [15] only examined categorical diagnostic classification discrepancies. In contrast, a multivariate approach may provide more sensitive measurement of the relationship between scale elevation and the intervening variables of demoralization, subtle items, and K-correction.

Locke et al. [10] contrasted MMPI-2 clinical and MMPI-2-RF RC scales in samples of patients with epileptic seizures versus those with nonepileptic seizures (NESs). These authors found significant elevations in the sample with NESs on validity scales (Fs and FBS-r), RC scales (e.g., RC1), and several supplementary somatic scales from the MMPI-2-RF. In contrast, the group with NESs scored significantly lower on RC3. In a diagnostic classification model, RC1 added significantly to demographic and medical history predictors of NESs. Locke and colleagues concluded that the MMPI-2-RF should be regarded as a valid alternative to the MMPI-2 in the assessment of psychopathology in patients with NESs and that clinicians can use the newer version for the assessment of NESs without loss of diagnostic utility [10].

The primary purpose of the present research was to extend the findings of the Sellbom et al. [15] study to a population of patients with seizure disorders using a multivariate approach. In particular, we wanted to examine the prevalence of discrepant elevation patterns on the clinical versus RC scales. Based on the findings of Sellbom et al. [15], we hypothesized that when differences in elevation occurred, the clinical scales would be more frequently elevated compared with the corresponding RC scale. In addition, we also wanted to examine the contribution of the three variables proposed by Sellbom et al. [15] to the explanation of discrepant elevations. It was hypothesized that these variables would significantly contribute to the explanation of variance in the MMPI-2 clinical scales over and above the variance explained by the corresponding RC clinical scale.

Section snippets

Participants

Medical records and assessment data from a consecutive sample of 140 patients with seizure disorders, for whom full records were available, were utilized in this study. The sample included 88 females and 62 males with a mean age of 35.0 years and a mean education level of 11.6 years of schooling. The patients were a consecutive, representative sample of the individuals referred to the Neuropsychology Service in the Department of Clinical Neurosciences at St Vincent's Hospital, Melbourne for

Clinical and validity scale scores

Application of the MMPI-2 criteria described above resulted in 19 cases being identified as invalid. In contrast, the MMPI-2-RF criteria resulted in 23 cases being excluded as invalid, and, as noted for consistency with previous studies [15], these latter criteria were used to derive the study sample of 117 (see Table 3). Means and standard deviations (SDs) for MMPI-2 and MMPI-2-RF validity and clinical scales are presented in Table 3. Within the MMPI-2 validity scales, F and FB had the highest

Discussion

The results of the present study extend the findings of Sellbom et al. [15] in heterogeneous mental health settings to a heterogeneous sample of patients with seizure disorders. The high prevalence of mental health needs in people with seizure disorders underscores the importance of accurate, dimensional assessment of psychopathology in this high-needs population. The recently released DSM-V [22] also highlights the value of dimensional assessment of psychopathology, and, historically, the

Acknowledgments

We thank Simon Scalzo for assistance with manuscript preparation.

Conflict of interest statement

Yossef Ben-Porath is a paid consultant to the MMPI Publisher, the University of Minnesota, and the distributor, Pearson, Inc. As coauthor of the MMPI-2-RF, he receives royalties on sales of the test. The remaining authors have no conflicts of interest.

References (22)

  • D. Loring et al.

    Testing the limits: cautions and concerns regarding the new Wechsler IQ and memory scales

    Neurology

    (2010)
  • Cited by (12)

    • Predictive utility of the Minnesota Multiphasic Personality Inventory-2-RF (MMPI-2-RF) in differentiating psychogenic nonepileptic seizures and epileptic seizures in male veterans

      2021, Epilepsy and Behavior
      Citation Excerpt :

      No additional substantive subscales were included given that previous investigations have not demonstrated elevations in these scales to be predictive, and there are no clear conceptual links with PNES [15]. Data were considered invalid and not included when validity scales were elevated based on manual recommendations and consistent with prior research [13,24]. However, a conservative approach was used with regard to inclusion in order to maximize construct validity of clinical and substantive scales.

    • Clinical identification of psychogenic nonepileptic events using combinations of psychological tests in a veteran sample

      2021, Epilepsy and Behavior
      Citation Excerpt :

      Manualized T-scores for the following scales were used based on previous evidence that they discriminate PNEE from ES or otherwise appear conceptually linked with PNEE based on the MMPI-2-RF manual description: infrequent somatic responses (Fs), symptom validity (FBS-r; noncredible somatic complaints), infrequent responses (F-r; general over-reporting of symptoms), response bias scale (RBS; noncredible cognitive complaints), revised clinical demoralization scale (RCd; general emotional distress), revised clinical scale 1 (RC1; somatic complaints), revised clinical scale 2 (RC2; low positive emotions), revised clinical scale 3 (RC3; cynicism), revised clinical scale 7 (RC7; dysfunctional negative emotions), malaise (MLS), head pain complaints (HPC), gastrointestinal complaints (GIC), neurological complaints (NUC), cognitive complaints (COG), and suicidality (SUI) [6,7]. Consistent with previous research and the MMPI-2-RF manual [7,31], profiles with the following scores were considered invalid/uninterpretable and excluded from analyses: variable response inconsistency (VRIN-r; representing an uncooperative test taking approach) T ≥ 80, true response inconsistency (TRIN-r; representing a true or false fixed-response style) T ≥ 80, infrequent psychopathology responses (Fp-r) T ≥ 100, or infrequent responses (F-r) T ≥ 120. The remaining MMPI-2-RF validity indices (i.e., CNS, Fs, FBS-r, L-r, K-r, RBS) were not used as criteria for invalid MMPI-2-RF profiles.

    • Diagnostic utility of the Minnesota Multiphasic Personality Inventory-2 Restructured Form in the epilepsy monitoring unit: Considering sex differences

      2018, Epilepsy and Behavior
      Citation Excerpt :

      As most patients with PNES are women [15–18], relatively few studies have evaluated sex differences in a large sample. In one prior study of patients with seizure disorders (95% epilepsy, 5% PNES), sex contributed significantly to the variance explained in regression models where MMPI-2-RF RC1 and RC7 scales were used to predict the corresponding MMPI-2 scales (Hypochondriasis and Psychasthenia, respectively), with women showing decreased odds of scale elevations compared with men; sex was not a significant contributor for other scales [19]. These results support the notion that sex can be a meaningful consideration when using the MMPI-2-RF as a diagnostic tool in seizure patient samples, particularly for clinical scales that have demonstrated diagnostic reliability in distinguishing PNES from epilepsy [5,9,10].

    • Increased odds and predictive rates of MMPI-2-RF scale elevations in patients with psychogenic non-epileptic seizures and observed sex differences

      2017, Epilepsy and Behavior
      Citation Excerpt :

      For the clinical scales, invalid profiles were removed from the analysis. Out of 169 total patients, a small number of patients (n = 5) exhibited invalid MMPI-2-RF profiles as defined by previously established guidelines [30]. Per these guidelines, for a profile to be considered invalid, elevations on any one of the following scales were required.

    View all citing articles on Scopus
    View full text