Course content © Copyright 2014 - 2018 by William W. Deardorff, Ph.D, ABPP. All rights reserved.

IMPORTANT NOTE:  This course is approved for California Qualified Medical Evaluator (QME) continuing education credits as well as all other accreditations held by BehavioralHealthCE.com.  If you do not need California QME continuing education credits, you should take the MMPI-2 and Chronic Pain course instead.  The following course contains additional information specific to practicing as a QME in the State of California.  In My Account, be sure to add QME to your Degree list and your QME Provider Number to your Licenses.  All Degrees and Licenses will be printed on the CE certificate.     

COURSE OUTLINE

Introduction

Learning Objectives

Development of the MMPI-2

    Clinical Scales

    Changes in MMPI Items

    New Normative Data

    New Psychiatric Sample Comparison Group

    The Development of New Norms

    MMPI-2 Content Scales

    MMPI-2 Validity Scales

MMPI-2-RF: MMPI-2 Restructured Form

MMPI-2 and Chronic Pain Assessment

The Controversy

The Utility of the MMPI-2 in Chronic Pain Assessment

    Identification of Psychopathology

    Identification of Personality and Behavioral Characteristics

    Standardized Scores

    Treatment Planning and Prediction of Outcome

    Forensic Assessment

Using the MMPI-2 with Chronic Pain Patients

    Examining the Overall Profile Elevation

    Sub-grouping by Code Type

    Identification of Sub-groups by Cluster Analysis

    Identifying Characteristics through Factor Analysis

    Use of the Content Scales

An Interpretive Strategy in using the MMPI-2 in Chronic Pain

Invalid Use of the MMPI-2

Case Example

Conclusions

References

INTRODUCTION

Current theories of the etiology and maintenance of chronic pain emphasize its multidimensional nature.  These theories take into account affective, cognitive, behavioral, social, and sensory aspects of the pain experience (Turk and Melzack, 1992; Wall and Melzack, 1984; QME Chronic Pain Management I: Concepts).  These models of pain require a detailed assessment of the patient’s full range of functioning and pain experience (Turk and Melzack, 1992; QME  Chronic Pain Management II: Evaluation and Treatment). Adequate assessment of this nature is important for understanding all aspects of the chronic pain patient’s suffering and to develop successful treatment programs (Turk, 1990). This type of assessment goes far beyond the traditional medical examination.

In order to adequately complete this multidimensional assessment, psychological testing is commonly used.  The original Minnesota Multiphasic Personality Inventory (MMPI; Hathaway and McKinley, 1943) was the most commonly used standardized personality test with chronic pain patients (Keller and Butcher, 1991; Love and Peck, 1987).  As will be discussed in more detail later, the original MMPI was revised and released as the MMPI-2 (Butcher, Dahlstrom, Graham, Tellegen & Kaemmer, 1989).  Additional scales were added to the MMPI-2 in 2001 (Butcher et al., 2001, The MMPI-2 Manual Revised).  Similar to its predecessor, it is likely that the MMPI-2 will also be the most commonly used personality test with chronic pain patients. Of course, any self-report personality test such as the MMPI or MMPI-2 should be only one part of a comprehensive evaluation approach that includes medical, behavioral, psychosocial and demographic data.   Keller and Butcher (1991) conclude that two goals of chronic pain assessment should include those found in Table 1.

This latter goal is especially important since individual characteristics of a chronic pain patient relate to designing treatment programs and predicting outcomes including: individualized multidisciplinary pain programs, spine surgery, implanted pain management devices, among other things.  The MMPI (and now the MMPI-2) can help to achieve these assessment goals. 

This course specifically reviews the use of the MMPI-2 in the assessment of personality and other characteristics of chronic pain patients.  This course will limit its focus to the following:

1.   brief overview of the development of the MMPI-2

2.   the controversy over using the MMPI/MMPI-2 with chronic pain patients

3.   the utility of using the MMPI-2 as part of a chronic pain evaluation

4.   the MMPI-2 for assessment of chronic pain problems

5.   interpretive strategies when using the MMPI-2 with chronic pain patients

6.   invalid use of the MMPI-2

DEVELOPMENT OF THE MMPI-2

This course will provide a brief overview of the development of the MMPI-2 with a special focus on how this relates to assessing chronic pain patients.  As discussed in detail elsewhere (Keller and Butcher, 1991), the need for a revision of the MMPI had been discussed for a number of years (Butcher, 1972) and the revision project was finally embarked upon in 1982 (Butcher, Dahlstrom, Graham, Tellegen, & Kaemmer, 1989; Keller and Butcher, 1991).  The development of the MMPI-2 included several significant changes from the original MMPI and these are summarized in the following.

MMPI-2 Clinical Scales

For those familiar with the MMPI-2, a review of the Clinical Scales will not be necessary; however, it will be provided for those who do not typically utilize this instrument.  Even if you do not use the MMPI-2 in a professional capacity, you may certainly come across its use in clinical practice.  Even for those who do not administer the MMPI-2, it can be very useful to recognize when it may have been used inappropriately with a chronic pain patient.

Table 2 lists the ten MMPI-2 clinical scales and these have not changed from the MMPI to the MMPI-2.  The original scale names and abbreviations (e.g. Hy: Hysteria) are also provided.  The original scale names reflect the original goal of the MMPI which was to identify and measure common diagnoses of the time (e.g. l940’s).  Subsequent research demonstrated that this goal of the MMPI was not possible and the scales were more usefully conceptualized as measures of psychological features, symptoms and characteristics (not formal diagnoses).  For instance, an elevation on Schizophrenia (Scale 8) in a great many cases does not indicate schizophrenia or even psychotic symptoms.  As such, the current standard practice is to refer to a clinical scale by its number (e.g. Scale 1) rather than its original name.  Table 2 illustrates very general summaries of what each scale measures. 

Changes in the MMPI Items

One goal of the MMPI revision was to preserve the established original MMPI clinical correlates while expanding the item pool to cover additional problem areas (Keller and Butcher, 1991; Butcher et al., 1989).  First of all, the 16 repeated items in the original MMPI were deleted and problematic items that used sexist wording, had objectionable content or used confusing syntax were either removed or rewritten.  It has been demonstrated that this re-wording did not affect endorsement frequency (Ben-Porath and Butcher, 1988).  Then, new items were added to cover content areas that were previously under-represented in the original MMPI item pool.  This resulted in the final form of the MMPI-2 that is 567 items long (versus 566 on the MMPI).  Item membership of the basic validity and clinical scales is largely equivalent to the original MMPI although certain scales are shorter (F, Hs, D, Mf, Si). As such, there can be slight shifts in raw score patterns from the MMPI to the MMPI-2 and comparisons must be made carefully. 

New Normative Data

One of the major changes in the MMPI-2 is the inclusion of a new national normative reference group upon which all T-scores are based.  The original MMPI was based upon outdated norms from over 50 years ago that did not represent the general population (Pancoast and Archer, 1989).  The contemporary normative sample was developed with an effort to match the 1980 census data (Butcher, et al, 1989).  This was largely accomplished although the normative sample tends to be biased towards higher socioeconomic levels and educational status as compared to the general U.S. population.  Also, Hispanic and Asian sub-populations are under-represented.  These issues should be taken into account when interpreting clinical profiles generated by chronic pain patients with these demographic characteristics. 

New Psychiatric Sample Comparison Group

As part of the revision of the original MMPI, a new sample of psychiatric patients was used as a comparison group.  This new sample was used to provide information about the performance of the MMPI-2 with a clinical sample as well as to help refine the new content scales (Keller and Butcher, 1991; Butcher, Graham, Williams, & Ben-Porath, 1989).  Relative to using the MMPI-2 with chronic pain patients, this new psychiatric comparison group should help with more accurate identification of psychopathology and personality characteristics as discussed above.

The Development of New Norms

As discussed by Keller and Butcher (1991), the largest differences between the MMPI and the MMPI-2 are likely to result from differences in the norming procedures.  The MMPI normative sample is outdated and even originally was quite limited.  This is in contrast with the norms for the MMPI-2 which are based upon a large national sample.  Scores from this sample provide the raw score distributions used to develop T-scores for each scale. On the MMPI, scales were converted to standardized T-scores using a linear transformation with a mean of 50 and a standard deviation of 10 relative to the original normative sample.  Due to a number of problems in using linear T-score transformations (see Keller and Butcher, 1991 for a discussion of these issues), “uniform” T-scores transformations were used on the MMPI-2 (See Butcher et al, 1989 for a review).  The details of these issues are not important for the current discussion except that the new uniform T-scores result in somewhat lower profile elevations as compared to the linear transformations.  This could affect interpretive strategies based upon scale elevations or code type patterns (Butcher and Keller, 1991).  For instance, this could result in substantially fewer pain patients being classified as being psychologically distressed (Butcher et al, 1989; Keller and Butcher, 1991).  This issue has been somewhat addressed by the shift in the level of a critical scale elevations, a T-score of 70 on the MMPI and a T-score of 65 on the MMPI-2 (Keller and Butcher, 1991).  A comprehensive study of using the MMPI-2 with chronic pain patients has substantiated the validity of this lower critical T-score level on the MMPI-2 (Keller and Butcher, 1991). 

MMPI-2 Content Scales

Since the clinical scales are multidimensional in nature, they can be elevated for several reasons.  For instance, Scale 2 (Depression) has items that focus on various features of depression including subjective depression, fatigue, physical problems, memory and concentration problems.  As such, two individuals with the same score on Scale 2 may have elevated it for different reasons.  One individual may be significantly depressed and the other may have endorsed all the physical and cognitive symptoms without endorsing any of the items related to a mood disturbance.  Analysis of the content scales (and other subscales) can help hone the interpretation of the profile.  This is especially important in assessing a patient with chronic pain due to the presence of a number of physical symptoms.

The MMPI-2 includes the development of many new content scales and some of these may prove useful in assessing chronic pain patients (Keller an Butcher, 1991; Butcher et al, 1989; Strassberg et al., 2000).   Among these content scales, those in Table 3 may be particularly useful and will be discussed later.

MMPI-2 Validity Scales

Another strength of the MMPI-2 is that it provides several validity scales beyond the MMPI to assess the test taking attitudes of the patient.  The MMPI-2 contains the same original validity scales as the MMPI (L, F and K).  New validity scales include both measures of consistency in responding and improved measures of validity. 

Consistency measures.  When first approaching the assessment of the validity of the MMPI-2 results, one must look at the consistency of responding.  This is done in a number of ways including the number of items omitted (Cannot Say) and the Consistency Measures (VRIN and TRIN).  If the VRIN and TRIN are two measures of how consistent the patient was in answering the questions. 

Validity measures. After determining that a patient answered the items in a consistent fashion, analysis of the validity of the profile can commence.  Aside from the original L, F, and K scales, new validity scales have been added to the MMPI-2.  All of the currently available scored validity scales can be seen in Table 5.  The MMPI-2 is scored by Pearson as licensed by the University of Minnesota.  The validity pattern on the MMPI-2 is among some of the most valuable information that can be obtained on chronic pain patients (Bianchini et al., 2008; Butcher and Harlow, 1987).

MMPI-2-RF: THE MMPI-2 RESTRUCTURED FORM

A new psychometrically improved version of the MMPI-2, the MMPI-2-RF has very recently been released.  It is a 338-item self-report measure linked conceptually and empirically to modern theories and models of psychopathology and personality.  This test employs rigorous statistical methods that were used to develop the RC Scales in 2003 (Nine RC scales are provided on the extended score MMPI-2 report). The MMPI-2-RF produces scores on a theoretically grounded, hierarchically structured set of scales, including the RC Scales.  The MMPI-2-RF builds on the foundation of the RC Scales, which have been extensively researched since their publication in 2003 (bibliography for the MMPI-2-RF).  The MMPI-2-RF is not without controversy.  In summary, proponents of keeping the MMPI-2 Clinical Scales argue that the new RC scales are measuring pathology which is markedly different than that measured by the original clinical scales.  However, the research suggests that the RC scales are more pure versions of the original clinical scales because: (1) the interscale correlations are greatly reduced, (2) no items are contained in more than one RC scale and, (3) common variance spread across the older clinical scales due to a general factor common to psychopathology is parsed out and contained in a separate scale measuring demoralization (RCdem).

Critics of the RC scales assert they have deviated too far from the original clinical scales and that previous research done on the clinical scales will no longer be relevant to the interpretation of the RC scales.  Proponents of the RC scales assert that research has adequately addressed these issues with results indicating that the RC scales predict pathology in their designated areas better than their corresponding original clinical scales while using significantly fewer items and maintaining equal to higher internal consistency reliability and validity.

Until the MMPI-2 findings with chronic pain patients can be confidently generalized to the MMPI-2-RF, it should be used with caution.  Studies are currently underway to address these issues. This is similar to what had to occur when the MMPI-2 was developed from the MMPI, but possibly on a more complex level.  

MMPI-2 AND CHRONIC PAIN ASSESSMENT: THE CONTROVERSY

The use of the MMPI or the MMPI-2 as part of a chronic pain assessment is not without controversy and the practitioner using this test should be aware of the issues.  For instance, Main and Spanswick (1995a) contend that the use of the MMPI and MMPI-2 in assessing chronic pain is “understandable but no longer justifiable” (p. 90).  Other researchers have taken a similar stand (Fishbain, 1996; Helmes, 1994; Main and Spanswick, 1995a, 1995b; Turk and Fernandez, 1995).  These investigators posit that the MMPI and the MMPI-2 have several problems that preclude their usefulness with chronic pain patients.  Some of the minor problems with the MMPI were remedied with the development of the MMPI-2 including such things as item content and wording, as well as the outdated norms.   Other criticisms are more fundamental and have been leveled at the MMPI and the MMPI-2.  Anyone using the MMPI-2 to assess the personality characteristics of chronic pain patients (either clinically or as part of a research project) must be aware of these issues to understand its strengths and weaknesses and avoid inappropriate use of the instrument. 

First, the point is made that the MMPI and MMPI-2 are based on outdated theories of psychopathology (Helmes and Reddon, 1993).  It has been noted that concepts underlying the diagnostic categories used to form the clinical scales of the MMPI are no longer used in modern theories of psychopathology (Helmes an Reddon, 1993).  Since these original scale names and most items have been retained on the MMPI-2, there remain links with the original diagnostic system that has been described as “relics of an antiquated psychiatry” (Cronbach, 1990, p.539).  Critics say that the scales clearly measure something other than that implied by the scale names (Turk and Fernandez, 1995).  These various researchers argue that this allegedly outdated approach makes the MMPI-2 not useful in helping to identify personality characteristics and psychopathology among chronic pain patients.

The second major criticism is that there is significant overlap in the item content across both clinical and validity scales which can cause spurious profiles.  Critics of the MMPI-2 believe that this item overlap causes scale elevations that do not accurately reflect what the patient is experiencing.  For instance, two common profiles among chronic pain patients are the so-called “V” profile (also referred to as the “Conversion-V” Profile) and the “neurotic triad”.  The first of these profiles is characterized by elevated T-scores on scales 1 (Hs) and 3 (Hy) with a lower T-score on scale 2 (D).  The latter of these profiles is characterized by elevations on scales 1, 2 and 3 (Hs, D, and Hs).  Critics of the MMPI and the MMPI-2 point out that there is significant item overlap across these and other scales.  For instance, scales 1 and 3 have 20 items in common and other items are scored on up to six scales.  The critics believe that this overlap makes it difficult to interpret profiles and to discriminate among different groups.  Their argument can be summarized as, “The overlap also makes it more difficult for the test to discriminate among different groups, because the responses to a small set of common items can influence several scales” (Helmes, 1994, p.5).

Third, it has been argued that the results of the MMPI-2 profiles are invalid when used with chronic pain patients due to the inclusion of items that reflect features of both a psychiatric disturbance and a chronic illness such as long term pain (Pincus, Callahan, Bradley, Vaughn, & Wolfe, 1986).  Scales 1 and 3 contain a number of items reflecting a general medical or physical condition.  Research has shown that elevations on scale 1 and 3 can be reflective of disease rather than psychological status (Pincus et al., 1986). Thus, there is the possibility that a patient simply reporting symptoms of his or her medical disorder could lead to the inappropriate conclusion of emotional and psychiatric problems due to elevations on these scales (Helmes, 1994).

These criticisms must be weighed against other arguments that substantiate the utility of using the MMPI-2 as part of a chronic pain evaluation.  Related to the first criticism, proponents of the MMPI-2 argue that sophisticated users understand that the scale names on the MMPI-2 do not reflect distinct psychopathological disorders in the patient as implied by the scale names (Bradley, 1995; Keefe, Lefebvre, & Beaupre, 1995; Sanders, 1995). In fact, in order to get away from the original scale meanings, many interpretive guidelines suggest using the scale numbers rather than the original names.

Relative to the second and third criticisms discussed previously, advocates argue that the MMPI-2 provides valuable information about persons with chronic pain and that experienced interpreters take the item overlap and medical symptom content into account as part of their interpretation.  For example, it will be taken into account that chronic pain patients will almost automatically elevate certain scales to a certain degree due to their report of medical symptoms (Bradley, Haile, & Jaworski, 1992; Moore, McFall Kivlahan, & Capestany, 1988; Naliboff, Cohen, & Yellin, 1982; Pincus, Callahan, Bradley, Vaughn, & Wolfe, 1986; Prokop, 1986; Watson, 1982).  These MMPI-2 users will also use the sub-scales to help aid in their interpretation of the profiles and determine exactly what subgroup of items caused the scale elevations (Butcher, 1990; Graham, 1990; Greene, 1991; Keefe et al, 1995; Strassberg and Russell, 2000)  

Special issues related to the utility of the MMPI-2 in assessing chronic pain patients will be reviewed more fully later in the course.  Briefly, MMPI/MMPI-2 assessment provides information about the individual characteristics of the chronic pain patient that have been found to be correlated to other important pain variables such as exhibition of pain behaviors, level of emotional distress, difficulty in performing daily activities, patterns of medication use, among other things (Bradley, 1995; Bradley et al, 1992; Gatchel et al., 2006; Keller and Butcher, 1991).  Also, these individual characteristics as identified by the MMPI-2 have been used to make predictions regarding patient behavior such as response to multidisciplinary treatment and return to work (Bradley, 1995; Gatchel et al., 2006; Moore, McCallum, Holman, & O’Brien, 1991).

THE UTILITY OF THE MMPI-2 IN CHRONIC PAIN ASSESSMENT

The specific areas of utility for MMPI-2 assessment of chronic pain patients can be summarized as follows:

Identification of Psychopathology. 

One of the important uses of the MMPI-2 in assessing chronic pain patients is to aid in the identification of psychopathology, including personality disorders (Gatchel, 1997; Gatchel et al., 2006).  Research has demonstrated a high prevalence of psychiatric disorders among chronic pain patients.  For instance, clinical depression has been found to be at least four times greater in people with chronic back pain than in the general population (Magni, Marchutti, & Moreschi, 1993; Sullivan, Reesor, Mikail, & Fisher, 1992).  It has been shown that, among patients seeking treatment at pain clinics, 32 to 82 percent of patients show some type of depressive problem, with an average of 62 percent (Sullivan et al, 1992).  

Research has also clearly revealed the high prevalence of personality disorders in chronic pain patients.  The prevalence range of personality disorders among chronic pain patients has been shown to be between 37 percent and 66 percent across various studies (Fishbain, 1997).  One study found that a wide range of personality disorders was identified in a sample of chronic pain patients, with seven of the 12 possible disorders represented (Reich, Tupin, & Abramowitz, 1983).  Another well designed study indicates that 51 percent of chronic pain patients met the criteria for one personality disorder and 30 percent met the criteria for more than one personality disorder (Polatin, Kinney, Gatchel, Lillo, & Mayer; 1993).  These rates are well above what has been found in the general population (about 6% in a random community sample; Samuels, Nestadt, Romanoski, Folstein, & McHugh, 1994).

Identifying these psychopathological disorders is extremely important in the management of a chronic pain problem.  Where significant anxiety or depression is present, specific treatment of these problems would be a necessary part of a chronic pain program (e.g. Sullivan et al, 1992; Turk and Melzack, 1992).  An accurate personality disorder diagnosis can help clinicians work with chronic pain patients in many ways. First, the identification of a personality disorder can help clinicians understand that certain behavior patterns and psychopathological features are likely to be long-standing and persist into the future (Gatchel and Weisberg, 2000; Weisberg and Keefe, 1997). 

Second, accurate identification of a personality disorder in a chronic pain patient will help improve patient management and treatment planning.  For instance, patients with personality disorders have been shown to exhibit inadequate coping skills (Gatchel, 1997) which, in turn, have been shown to relate to ineffective pain coping strategies in chronic pain patients (Kleinke, 1994; Turk and Rudy, 1986).    It will also help the treatment team establish realistic goals for treatment outcome.  The MMPI-2 can be a very important part of a multi-component assessment for psychiatric diagnoses, including personality disorders (Weisberg and Keefe, 1995a)

As presented by Keller and Butcher (1991), there is “some evidence that the MMPI-2 can identify psychopathology even within this highly homogeneous sample” of chronic pain patients (p. 105).  However, the authors go on to say that MMPI-2 scales generally lack specificity for the correlates of psychopathology that were identified. Therefore, the presence of psychopathology seems to be related to a general elevation on a number of MMPI-2 scales measuring distress and dysphoria.  The MMPI-2 has been shown to be able to identify the specific nature of the psychopathology in a more heterogeneous sample (Keller and Butcher, 1991).  In their sample of chronic pain patients, they found that MMPI scales designed to measure depression, anxiety, and low self-esteem (D, Pt, ANX, DEP, LSE, A) seem to be useful in identifying patients who are having difficulty with depression in general. They report that depressive symptoms seem to be the major dimension along which the sample of chronic pain patients varied.

Keller and Butcher (1991) also stated there was a small group of other correlates that emerged in their data set.  For instance, there was a history of various characterological problems (such as alcohol and drug use, arrest history, and violence) associated with elevations on scales Pd, Ma (men only), Antisocial Practices (ASP), Anger (ANG), and MacAndrew alcoholism (MAC).  The authors suggest that these data support the conclusion that the MMPI-2 can identify these potential problems even within a chronic pain patient population that has been specifically screened to exclude individuals with psychopathic characteristics.  Keller and Butcher (1991) state that it was not possible to determine whether the scale elevations reflected current difficulties or past history information in these patients.  Other researchers have found MMPI-2 scale correlates with alexithymia in chronic pain patients (Lumley, Asselin, & Norman, 1997).

Recent research has suggested that the certain MMPI-2 profiles may, indeed, be able to successfully screen for psychopathology.  Gatchel et al. (2006) investigated the MMPI-2’s ability to predict psychopathology in 1489 chronic occupational spinal disorders (COSD) patients.  Specifically, they were interested in studying what they termed a “Disability Profile” (DP) on the MMPI-2 (See Personality Characteristics of Patients with Chronic Pain).  As discussed by the authors, this profile had previously been termed a “Floating Profile” in the psychiatric literature; however, it had not been investigated in the chronic pain arena.  The DP profile was defined as 4 or more of the clinical scales being significant (T>65).  In the psychiatric research, this profile was associated with personality disorders (particular Borderline).  Gatchel et al. (2006) were also interested in the prevalence of various MMPI-2 profiles in a musculoskeletal chronic pain population.

The investigators administered the MMPI-2 as part of an initial comprehensive evaluation prior to interdisciplinary chronic pain treatment.  The patients were divided into 4 groups based upon MMPI-2 profiles (the definitions of these profiles groups are based on previous research to be discussed subsequently).  The results of the MMPI-2 profile classification can be seen in Table 6.

This yielded 1185 MMPI-2 profile group patients, of which 868 had DSM-IV diagnoses.  Table 7 is a brief summary of the psychiatric diagnostic results.  The Table shows the percentage of psychiatric diagnosis in each MMPI-2 profile group.

Overall, the authors found that the traditional MMPI/MMPI-2 profiles often associated with chronic pain (NP, NT, CV) failed to classify 75% of the patients.  The DP profile has a prevalence of 53.2% in this chronic pain patient population and was associated with significant psychopathology.  This would certainly be important information in terms of treatment planning. 

Identification of Personality and Behavioral Characteristics

The identification of personality and behavioral characteristics of chronic pain patients differs from the identification of psychopathology as discussed above (Weisberg and Keefe, 1995a, 1995b).  This is an important distinction to keep in mind since it has been argued that clinicians will often “label” chronic pain patients as having a personality disorder when this is clearly inaccurate and inappropriate (Weisberg and Keefe, 1995a).  The MMPI has a long history of use in identifying personality characteristics in chronic pain patients (See Weisberg and Keefe, 1995a for a review).  The MMPI-2 is now being used for a similar purpose.  Examples of pain patient personality characteristics that might be identified by the MMPI-2 include such things as dependent personality trends, passive aggressiveness, and obsessive- compulsive behavior.  Identification of these characteristics by the MMPI-2 can help in treatment planning with chronic pain patients by doing such things as developing a time-limited treatment contract  (to address the dependent traits), adding assertiveness training to the treatment program (for the passive-aggressive behaviors), and addressing obsessive-compulsive behaviors through cognitive behavioral approaches (Weisberg and Keefe, 1995a).

Standardized Scores

Another strength of the MMPI-2 is that the scores are standardized.  This allows the clinicians to compare an individual chronic pain patient profile with other groups (Keefe et al, 1995).  The MMPI-2 includes a new national normative reference group upon which all T-scores are based.  This updated normative reference group allows for more accurate comparisons to a population that is more representative of the U.S. population in general.  Of course, the experienced MMPI-2 user will compare a chronic pain patient’s profile not only with the normative reference group but also with other chronic pain patient profiles and subgroups developed through statistical means (Bradley et al, 1992; Keller and Butcher, 1991).

Treatment Planning and Prediction of Outcome

As discussed previously, many users of the MMPI and MMPI-2 use these instruments to make predictions regarding patient behavior including response to treatment and treatment outcome (Bradley et al, 1992; Guck, Meilman, Skultety, Poloni, 1988; Keller and Butcher, 1991).  Many studies have shown that MMPI profiles of chronic pain patients can be reliably classified into three or four major subgroups (Bradley et al, 1992; Love and Peck, 1997; Sanders, 1995).  The subgroups have been found to differ reliably from one another on such variables as pain intensity, medication use, functional disability, and employment status (See Bradley et al, 1992 for a review).  Similar results are now being found with the MMPI-2 (Riley, Robinson, Geisser, & Wittmer, 1993; Riley, Robinson, Geisser, Wittmer, & Smith, 1995).  Other areas also investigated using the MMPI-2 include its ability to predict a poorer return to work rate (Moore, McCallum, Holman, & O’Brien, 1991) and surgery outcome (Riley et al, 1995; Block et al., 2003).

In another study, Maruta, Goldman, Chan, Ilstrup, Kunselman, and Colligan, 1997), examined the relationship between MMPI-2 scale elevations and nonorganic signs (measured by a “Waddell” score) of low back pain on physical examination. These Waddell signs are patient responses to standardized physical examination procedures that suggest a non-organic component to the pain behavior     (Waddell, McCulloch, Kummel, & Venner, 1980)   The Maruta et al (1997) study is important because it expands previous research on how well MMPI-2 scales correlate with other pain-related variables.  As discussed by Maruta et al. (1997), previous research using the MMPI had established a correlation between scales 1, 2, and 3 with nonorganic signs of low back pain.   The nonorganic signs of low back pain have been found to be associated with general affective disturbances, ineffective coping, and abnormal illness behavior in previous research (Waddell, Pilowsky, and Bond, 1989).  Maruta et al. (1997) investigated this issue by dividing male and female patients into high and low groups based on their scores for showing nonorganic physical signs of low back pain.  These were termed High Waddell’s (HW) and Low Waddell’s (LW) after the scoring system that was used.   It was found that among the male patients, statistically significant differences were found between the HW and LW groups on the MMPI-2 scales 1, 3, and 8.  Among females, differences between the HW and LW were statistically significant only on scale 8.  The study provides support for the ability of the MMPI-2 to discriminate between patients showing differing levels of nonorganic physical signs of low back pain.

As discussed previously, Gatchel et al. (2006) identified a Disability Profile on the MMPI-2 amongst a group of chronic pain patients.  Participants with a Normal Profile (NP) were twice as likely as the other three code types to retain work one year after treatment.  The DP code type group was 14 times more likely than the NP group to have an Axis I diagnosis.  In addition, in the DP group there was a comorbidity of Axis II personality disorders almost five times that of the NP group.  The study clearly established that there was a high prevalence of psychopathology associated with the DP code type.

A recent study by Haggard et al. (2008) completed a further investigation of the Disability Profile as identified by Gatchel et al. (2006) and discussed previously.  The participants were 755 patients who, as part of an initial evaluation for pain treatment, produced a valid MMPI-2 profile that could be classified into one of four profiles (NP, CV, NT, and DP).  Unlike the Gatchel et al. (2006) study, the patients were heterogeneous with regard to their pain problems.  Relative to patients in the other groups, the DP group demonstrated more severe behavioral pathology, psychosocial distress and physical impairment across a number of measures.   

The MMPI has also been used extensively as part of a comprehensive screening evaluation when a chronic pain patient is being considered for spine surgery (See the QME Psychological Screening in Spine Surgery course) or placement of a spinal cord stimulator (See Burchiel, Anderson, Wilson, Denison, Olson, & Shatin, 1995; North, Kidd, Wimberly, & Edwin, 1996 for reviews as well as the QME Chronic Pain Management III: Special Issues course).  A spinal cord stimulator is an electronic device with leads placed in the spine to help with pain relief.  In general, MMPI scales 1, 2, and 3 have been found to correlate with outcome from this procedure.  The MMPI-2 was tested as part of a comprehensive evaluation designed to identify a patient population in whom reasonably long-term success could be expected after placement of the spinal stimulators (Burchiel et al., 1995).  In this study of 40 patients (85 percent of whom were diagnosed with a failed back surgery syndrome), the MMPI-2 depression scale was found to be a prognostic indicator.  This seems to substantiate being able to generalize previous findings from the original MMPI to the MMPI-2 for use in this area.

Forensic Assessment

As concluded by Pope, Butcher and Seelan’s review of the literature, the MMPI-2 (and the MMPI previously) is the most widely used personality assessment for evaluating individuals in forensic settings (2006, p. 7).  Although I know of no data on the subject, this conclusion very likely applies to forensic chronic pain evaluations as well.  Many chronic pain patients have onset of their symptoms due to some type of injury for which they allege that someone else is accountable (e.g. workers compensation, personal injury).  Therefore, anyone involved in evaluating or treating chronic pain patients will invariably interact with the legal system and forensic evaluation issues may arise.  For a very detailed review of this area, see the course QME Assessment of Malingering in Chronic Pain.  

As discussed in QME Chronic Pain Management I: Concepts, chronic pain is always subjective in nature and largely based on self-report by the patient as well as the display of other pain behaviors (e.g. limping, grimacing, using assistive devices, etc).  Given the largely subjective nature of the symptoms, and the incentives associated with litigation settlement, the possibility of malingering is always an issue. In the DSM-IV, malingering is defined as, “…the intentional production of false or grossly exaggerated physical or psychological symptoms, motivated by external incentives such a avoiding military duty, avoiding work, obtaining financial compensation, evading criminal prosecution, or obtaining drugs” (p. 683). The DSM-IV differentiates malingering from other conditions in which the production of symptoms is not intentional such as somatoform or conversion disorders or where the incentive is not external (factitious disorder).  This strict differentiation of intentional versus unintentional and internal versus external sets up a “false dichotomy” that is likely to show up in clinical studies and forensic evaluations (Arbisi and Butcher, 2004).  Arbisi and Butcher (2004) and others (Fishbain et al., 1999) make the point that DSM-IV sets up an impossible task for the clinician, that of determining the patient’s incentive for and awareness of a behavior (e.g. being a “mind reader”). 

A more reasonable approach to conceptualizing and evaluating malingering is presented by Rogers (2008).  The definition of malingering presented by Rogers (2008) includes the observation that the level of intention may vary, and the incentives for deception may change over time. This definition allows for gradations of malingering or deception as well as response styles such as impression management and defensiveness (see Arbisi and Butcher, 2004; Rogers, 2008 for a complete discussion of these issues). This continuum conceptualization of malingering is especially important in the area of chronic pain since the symptoms become more and more influenced by non-physical factors over time including cognitive (e.g. catastrophic thinking), affective (e.g. depression) and operant (e.g. nurturance and support from the family). These factors are non-physical but are also unintentional.  Then there may be the very subtle intentional forces supporting the display of pain behaviors and disability such as the case example.  The challenge is to separate out the intentionally deceptive behavior and then determine its level of intensity.  As discussed by Arbisi and Butcher (2004), “Approaching the challenge in an overly simplistic manner by reducing the clinical decision to a simply dichotomy – malingered versus non-malingered – fails to acknowledge that both the pain experience and the motivation to alter self-report is a fluid process” (p. 388). They go on to state that, “The MMPI-2 can provide useful information in specifying the degree to which the individual is altering his or her presentation as well as the psychologic context within which the pain complaint occurs” (p. 389).

Just recently there are more studies being done using the MMPI-2 in assessing malingering of chronic pain and associated symptoms (e.g. disability).  As discussed by Rogers (2008), the preferred method for developing malingering detection techniques is referred to as the “known groups” design.  This research design uses patients in actual clinical settings that have been independently identified as engaging in dissimulation or malingering.  This group is then compared to criterion groups of known “honest responders” on measure of interest in the study.  In the first known group study of its kind, Biachini et al. (2008) assessed the classification accuracy of MMPI-2 validity scales in the detection of pain-related malingering.  The sample consisted of patients without financial incentives, nonmalingering patients with financial incentives, patients definitely determined to be malingering based on published criteria, and college students asked to simulate pain-related disability.  The study found that the MMPI-2 validity scales had the capacity to differentiate malingerers from nonmalingerers with a high degree of accuracy.  As discussed previously, and summarized by Biachini et al. (2008), it is also important to remember that not all exaggeration of pain symptoms is intentional.  Therefore, malingering indicators must differentiate between symptom exaggeration due to psychosocial factors and intentional deception.  Their study suggested that this could be accomplished based upon the elevation of some of the scales investigated.  They found that the malingering patients and college simulators scored on average 10 points higher on the Hs (Scale 1) and Hy (scale 3) scales than those patients more likely to have symptom exaggeration due to psychosocial factors.  The malingerers were almost 4 times more likely to score T=90 or higher compared to the nonmalingerer group on these scales.   

In another series of studies the Meyers validity composite index (MI) for the MMPI-2 has been investigated to detect malingering in chronic pain.  The MI is a composite of many validity scale results in which each scale score is weighted 0, 1, or 2 depending on how strongly the score indicates exaggeration (see Aguerrevere et al., 2008, for a review).  The Aguerrevere et al. (2008) used a known groups design to test the ability of an abbreviated MI to accurately differentiate malingerers from nonmalingerers.  The abbreviated MI consisted of scales that are all available on the current MMPI-2 extended score report from Pearson (not all of the scales from the original MI are scored by Pearson).  The abbreviated MI consisted of F, Fp, FBS, ES, and F-K.  The accuracy of the abbreviated MI was high and comparable to the original MI. 

Lastly, the MMPI-2-RF is beginning to be investigated relative to its use in forensic settings (Wygant et al., 2009).  The research on the ability of the MMPI-2 to identify gradients of malingering in the chronic pain population will continue to refine this valuable tool.  Any practitioner who works with chronic pain patients should have some knowledge of this research since you will no doubt be faced with a case in which it is relevant.

USING THE MMPI-2 WITH CHRONIC PAIN PATIENTS

There are really two important questions that must be answered relative to using the MMPI-2 with chronic pain patients.  First, it must be determined what research and clinical findings are generalizable from the MMPI to the MMPI-2 in this patient population.  Literally hundreds of studies have been done using the original MMPI in assessing chronic pain patients.  It is imperative that the generalization from the MMPI to the MMPI-2 can be made with confidence.  Second, it is important to determine how the new features of the MMPI-2 might make it more useful in the assessment of chronic pain patients. 

Examining the Overall Profile Elevation

Mean profiles for chronic pain patients have been compared between the MMPI-2 and the original MMPI (Keller and Butcher, 1991).  It was found that the new norms and T-score conversions on the MMPI-2 resulted in a generally lower profile even though the overall shape of the profiles were maintained between the two instruments.  As discussed previously, this problem is addressed by using an interpretive range that is five T score points less on the MMPI-2 relative to the MMPI and this appears valid with chronic pain patients (Keller and Butcher, 1991).

Sub-grouping by Code Types

Research in this area has focused on how MMPI-2 code types compare to those found with the MMPI in this patient population.  It also looks at how unique MMPI-2 code types might be useful in providing information about chronic pain patients.

Keller and Butcher (1991) have examined these issues in a large sample of chronic pain patients.  They first compared patients with the highest clinical scale in each profile, regardless of elevation (1-point code type).  Correspondence was measured by the percentage of profiles with a given high-point elevation using MMPI-2 norms versus the original MMPI norms.  It was determined that 64 percent of the men and 78 percent of the women would have obtained the same highest scale elevation on the original MMPI as they had on the MMPI-2.  Of course, this means that 36 percent of men and 22 percent of women received a different high point elevation on the MMPI-2 relative to the original MMPI.

In the same large study, there was also an investigation of the consistency of two-point codes. Using appropriate critical T-score criteria, it was found that 56 percent of the men and 67 percent of the women would have obtained exactly the same two-point code type across the MMPI and the MMPI-2. These results support only a moderate level of consistency from the MMPI to the MMPI-2 relative to chronic pain patients.  An important conclusion from these data is that the “configuration of scales Hs, D, and Hy should not be automatically assumed to correspond to previous research on the scales” (Keller and Butcher, 1991, p. 109).  These investigators warn that, when using the MMPI-2, a larger difference between scales 1, 2, and 3 should be required before strong interpretations of the “Conversion-V” profile versus the neurotic-triad or reactive depression profiles is made. They also conclude that chronic pain patients will look somewhat less depressed and less characterological (Scale 4, Pd elevation relative to other scales) on the MMPI-2 than on the MMPI.  Conversely, patients who obtain high scores on these MMPI-2 scales would have obtained even higher scores on the MMPI.

The Keller and Butcher (1991) study also looked at correlates of various scale elevations among their chronic pain patient sample.  It was determined that several traditional MMPI clinical scales (D and Pt) and the new MMPI-2 content scales (ANX, DEP, LSE) were associated with depressive symptoms.  Also, the Ma and Pd clinical scales, the MAC-R scale and the new ASP (Antisocial Practices) and ANG (Anger) content scales were found to be associated with impulse control, characterological, and “acting out” problems such as alcohol/drug abuse or legal histories. Lastly, Keller and Butcher (1991) conclude that the general consistency in profile classification between the MMPI and MMPI-2 allows for generalization of previously determined behavioral correlates of these code-types as determined using the MMPI with chronic pain patients.  Generalizing these MMPI findings should be done keeping in mind the previous discussion.

Identification of Sub-groups by Cluster Analysis

Initial interest in the identification of MMPI profile sub-types in chronic pain patient populations probably started with Sternbach (1974).  Sternbach (1974) initially described homogeneous subgroups within the chronic pain patient population that were based on his clinical observation of his patients.  He then identified four distinct MMPI profile sub-types based on these observations.  He concluded that these sub-types would require different treatment approaches for maximum effectiveness.

Subsequently, Sternbach's ideas were tested using multivariate cluster analytic techniques to analyze MMPI profiles. As discussed previously, relatively simple code typing rules will use one- or two-point codes to assign chronic pain patients to different sub-groups.  In contrast, cluster analysis uses information from all of the scales in the profile to classify 100% of the patients in the sample.    Many studies have identified either 3 or 4 cluster profiles (Bradley et al, 1992; Costello, Hulsey, Schoenfeld, & Ramamurthy, 1987; Keller and Butcher, 1991). Further, these subgroups have been validated by identifying cluster differences across patient variables such as pain duration and intensity, pain related disruption of activities, number of surgeries, number of hospitalizations, history of mental health treatment, and effectiveness of pain coping strategies (Bradley et al, 1992; Bradley, 1995; Keller and Butcher, 1991). 

This research was furthered by Costello et al (1987) who used a meta-clustering technique to combine the results of 10 previous cluster analytic studies using the MMPI.  These researchers concluded that a 4-cluster typology was the most appropriate one and this can be seen in Table 9.

Costello et al (1987) asserted that these four clusters were common to both genders.  This assertion, as well as the existence of the four clusters, was supported by McGill, Lawlis, Selby, Mooney, and McCoy (1983), who had previously found similar results.

Given the extensive and important cluster analytic findings with the MMPI, it is important to determine that similar subgroups of profiles can be found on the MMPI-2 when assessing chronic pain patients.  This issue has been addressed by many researchers since the development of the MMPI-2.  Using an experimental form of the MMPI-2, Cohen (1987) found the same general patterning of the four clusters that was found by Costello et al. (1987). 

Keller and Butcher (1991) also tested the previous cluster findings using the MMPI-2.  Although they could not determine a clearly defined number of clusters, they tested solutions from two to six.  They found that only a three cluster solution could be replicated across two cohorts (General-Elevation, Neurotic-Triad, and Within-Normal-Limits). Although the authors conclude that their results support the comparability between the MMPI and MMPI-2, the reasons for their unique finding of three clusters is not clear.  Other authors have suggested that it was either a methodological issue (e.g. use of the experimental MMPI-AX form, screening criteria which “screened-out” a large number of the chronic pain patient referrals) or it is the correct cluster solution when using the MMPI-2 (Riley et al, 1993).

Riley et al (1993) investigated whether MMPI-2 cluster solutions would replicate those found by previous researchers with the MMPI specifically in chronic low back pain patients.  Another goal of the study was to help determine which cluster solution was most appropriate with the MMPI-2:  the four-cluster solution identified by Costello et al (1987), Cohen (1987), and many previous researchers using the MMPI or the three-cluster solution found by Keller and Butcher (1991) using the MMPI-AX form.  Riley et al. (1993) found that the four cluster solution was the most appropriate one.  These researchers found the same four clusters that have been previously identified consistently on chronic pain patient MMPI profiles:

•     General-Elevation or Depressed-Pathological

•     Neurotic-Triad

•     Conversion-V

•     Within-Normal-Limits (WNL)

Significant group differences across these sub-types were found for variables of time from injury to evaluation, the number of previous back surgeries, and pre-injury psychiatric treatment.  Also, the WNL and Triad sub-groups had a significantly shorter time from injury to evaluation than the Depressed-Pathological subgroup and a significantly fewer number of surgeries than the V-type subgroup.  Also, the Depressed-Pathological subgroup was significantly more likely to have a reported psychiatric history than the WNL or V-type subgroups.   These authors conclude that, “…clinical application of MMPI-2 with CLBP [chronic low back pain] patients affords little loss in interpretation when compared with the MMPI.  Furthermore, although additional studies are certainly wanted, they show that much of the MMPI empirical database with CLBP patients is also applicable to the MMPI-2” (Riley et al., 1993, p 252). 

In a follow-up study, Riley et al. (1995) investigated how MMPI-2 cluster profiles might predict surgical outcome in a subgroup of their sample that ultimately underwent spine surgery.  Similar to findings with the original MMPI (See Riley et al., 1995 for a review) these researchers found that cluster subgroupings of MMPI-2 profiles were predictive of a number of different surgical outcome measures.  For instance, the study found that patients in the WNL or Triad subgroups reported significantly more satisfaction with post-surgical improvement than did patients in the depressed-pathological or V-type subgroups.  In addition, the Triad subgroup also gave a more favorable subjective rating of surgical outcome than did the depressed-pathological or V-type subgroups.

Slesinger et al. (2002) investigated the use of MMPI-2 with chronic pain patients including prevalence of two-point code types and assignment to one of four clusters as defined by the P-A-I-N algorithm discussed previously. Patients included 209 individuals presenting for treatment in an inpatient pain rehabilitation program.  In this chronic pain patient population, more than 67% produced basic MMPI-2 code types classified as 1-3/3-1, 2-3/3-2, 1-2/2-1.  Furthermore, the only 3-point code types that occurred with a frequency greater than 10 patients were the neurotic triad code type (Type I: 1-2-3/3-2-1) or the Conversion-V configural pattern (with Scale 2 at least 10 points lower than Scale 1 and 3).  The Conversion-V pattern had a frequency of 10% and the neurotic triad 28%.  The P-A-I-N classification is as follows:

As discussed by the authors and previous researchers (Keller and Butcher, 1991), there is support for the 4-cluster solution in a chronic pain patient population and these profiles are associated with unique characteristics; however, many chronic pain patients will produce profiles that do not correspond to any of the cluster prototypes.  In clinical practice, the profiles are still useful to keep in mind when conceptualizing characteristics of a chronic pain patient and associated features.  Strict decision rules that attempt to concretely assign a patient to one of the four clusters have not proven useful in the vast majority of cases.    

Identifying Characteristics through Factor Analysis

One method of improving the utility of the MMPI and MMPI-2 is to identify cluster subgroups as discussed previously.  Another method is through the use of factor analysis.  The objective of this strategy is to identify distinct, relatively independent characteristics of the chronic pain syndrome as assessed by the MMPI.  One study completed by Schmidt and Wallace (1982) of the original MMPI factor analyzed the clinical scales and subscales in a small group of chronic pain patients (N=25 with a cross-validation sample of 25).  Three factors were identified and labeled “Severity of Symptomatology”, “Anger and Aggression”, and “Psychogenic Components”.  This study must be viewed with caution due to the small sample size relative to the number of variables being factor analyzed.

Deardorff, Chino and Scott (1993) sought to replicate the factor analytic findings of Schmidt and Wallace (1982) using the MMPI-2.  In the Deardorff et al. (1993) study, MMPI-2 profiles of 114 chronic pain patients from two different treatment centers were factor analyzed.  The analysis included the clinical scales and the Harris and Lingoes sub-scales for Scales 1, 2, and 3.  Four interpretable factors were identified and labeled: “Psychological Dysfunction”, Interpersonal Isolation”, “Psychomotor Retardation”, and “Physical Dysfunction”.  The different findings of Deardorff et al. (1993) and Schmidt and Wallace (1982) might be due to the differences in patient populations sampled (low back versus a heterogeneous pain group), factor loading criteria for variable inclusion (.5 versus .3), and the use of different sub-scales for analysis.

Use of the MMPI-2 Content Scales

As discussed previously, the new MMPI-2 content scales can provide additional information over what was available on the original MMPI and this applies to the evaluation of a patient with chronic pain (Butcher et al., 1989; Slesinger et al., 2002; Strassberg and Russell, 2000).  As pointed out by Keller and Butcher (1991), the content scales are interpreted according to the attitudes and themes reflected in the items.  For instance, Keller and Butcher (1991) point out that the HEA scale might function as a marker of somatic preoccupation and concerns. They also feel that the other content scales might be useful for determining different treatment needs among chronic pain patients.  As example, the WRK scale designed to measure attitudes and behaviors that are likely to contribute to poor work performance, a lack of ambition, and negative attitudes toward co-workers (Graham, 1990).  Assessment of work attitudes is especially important for at least two reasons.  First, many chronic pain patients have had their problem start with an occupational injury and are disabled from work. Second, job dissatisfaction has been shown to be an important variable in chronic pain problems stemming from work injuries (Bigos, Battie’, and Spengler, 1991; Fordyce, Bigos, Battie’, and Fisher, 1992).

The TRT scale is designed to measure such things as a negative attitude towards doctors, giving up easily when problems are encountered, and showing poor problem solving abilities and judgment (Graham, 1990).  The scale might be useful in predicting problems that might occur during treatment for the chronic pain problem and addressing this early in the intervention program.  Lastly, the MAC-R scale was shown to be a useful indicator of substance abuse history among chronic pain patients in the Keller and Butcher (1991) study.

Recent research has started to empirically test the value of the content scales in assessing the chronic pain patient.  Strassberg and Russell (2000) administered the MMPI-2 to 309 patients presenting to an outpatient, hospital-based pain management program.  The investigators were interested in determining whether certain content scales (ANX, DEP, LSE, and ANG) might demonstrate the ability to add valuable information beyond that provided by the traditional MMPI-2 clinical and validity indices.  Concurrent, criterion-related and discriminant validity of these four content scales was investigated.  The authors concluded that, “The results of the evaluation of the concurrent, criterion-related validity of four of the MMPI-2 content scales suggest that, within this large chronic pain patient sample, the new scales provided significant and substantial information relevant to the constructs of depression, anxiety, self-concept, and anger” (Strassberg and Russell, 2000, p. 57).  The content scales were found to add 6% to 20% to the variance when added to the regression equation that contained a standard scale being correlated with a well-established criterion.  

AN INTERPRETIVE STRATEGY IN USING THE MMPI-2

WITH CHRONIC PAIN PATIENTS

An overview of an interpretative strategy when using the MMPI-2 with chronic pain patients can be seen in Table 11.  This strategy is based upon the MMPI-2 Manual and other sources. 

Current research suggests that MMPI-2 validity, clinical, and supplementary scales will provide the same type of symptom and personality information about the chronic pain patient as did the original MMPI scales (Keller and Butcher, 1991, Riley, et al, 1993, 1995).  Further, preliminary cluster analytic and factor analysis studies also suggest compatibility across the two instruments. As discussed previously, some caution must be used when generalizing one and two-point code type research and clinical data from the MMPI to the MMPI-2. 

Keller and Butcher (1991) and Senior and Douglas (2001) summarize an interpretive strategy based on their research in addition to previous strategies developed on the original MMPI especially by Sternbach (1974), Fordyce (1979), Costello et al. (1987) and Love and Peck (1987).  These guidelines will be summarized, augmented by information from more recent research.

As with any MMPI-2 interpretive strategy, careful evaluation of the validity scales is the initial step.  This includes analysis of the consistency and validity scales.  Keller and Butcher (1991) feel that this is particularly important with chronic pain patients because of the multitude of variables that might be impacting that presentation of pain behaviors and disability (such as compensation/litigation issues, medication use, and family reinforcement patterns).  They point out that the validity configuration can indicate such things as possible drug or alcohol toxicity, cognitive impairment, as well as assessing how "comfortable" the patient is in the current sick-role. It is also discussed that the validity pattern can give the interpreter an idea of the patient’s typical defense mechanisms (such as denial) and an idea of how the patient views his or her own resources to manage the problem.

A study by Dush, Simons, Platt, Nation, and Ayres (1994) underscores Keller and Butcher’s (1991) recommendation regarding the importance of assessing validity scales in the chronic pain population.  Dush et al. (1994) examined chronic pain patients who were in the midst of litigation over settlement for their injuries.  They compared MMPI-2 profiles from two similar groups of chronic pain patients; one of which was in litigation and the other of which was not.  Significant differences between the groups on various MMPI-2 subscales were found.  In addition, a Conversion-V profile was more salient for those in litigation versus those that were not. Similar results have been found in subsequent research (Bianchini et al., 2008)

Several lines of research on the K-scale of the original MMPI have suggested that it is particularly useful in identifying psychological contributions to physical conditions (see McGrath, Sweeney, O’Malley & Carlton, 1998 for a review).  It is assumed that if most chronic pain patients do not show psychopathological characteristics prior to developing pain (although this assumption is not without critics; see Weisberg and Keefe, 1995 for a review of this issue), then an elevated K-scale could be indicative of an individual who demonstrates superior psychological adjustment to the physical injury.  In pursuing this line of research with the MMPI-2, McGrath et al (1998) concluded that "…the K-scale emerges as a useful predictor of the individual’s emotional response to the development of chronic pain.  Patients with relatively elevated scores on the K-scale seem to have demonstrated better psychological adjustment to their pain.  Individuals with low or normal K-scale scores were more likely to be seen as psychologically avoidant, dependent, and self-concerned by the program staff suggesting that the intensity of their pain complaints are more likely to be exacerbated by psychological factors” (p. 457).

The new validity scales on the MMPI-2 (Fb, VRIN, TRIN, Fp) can help in determining the validity of the test.  For instance, an elevated Fb can indicate that the pain patient stopped paying attention to the test items that occurred later in the booklet and shifted to some type of random responding. The VRIN can help indicate if the patient is tending to respond inconsistently to items throughout the test. And, the TRIN can help identify a patient who is showing a response bias towards True or False answers.  These four scales are particularly important in assessing a chronic pain patient.  As example, patients who are completing the MMPI-2 under duress (e.g. involved in medical-legal issues) might elevate these validity scales if they are trying to manipulate the results.  These scales might also be elevated in the chronic pain patient who is fatigued or lacks endurance in completing the task and simply starts responding randomly half-way through the test.  Although beyond the scope of this course, there is an extensive body of research investigating the ability of the MMPI-2 validity scales to identify malingering in chronic pain patients (See Bianchini et al. 2008 for a review).   

Once the validity scales are examined, one should look at the overall elevation of the clinical profile.  This provides important information about the chronic pain patient’s degree of distress, disability, and the "cost" of being sick.  The cluster analytic research reviewed previously has shown that increasing profile elevations are associated with increasing disability in the patient’s life as well as poorer outcome from traditional unilateral treatment interventions. Even though exact assignment to one of the four cluster groups may not be possible, the clinician can generate interpretive hypotheses based upon approximation of assignment.  The importance of the overall elevation of the profile has been discussed previously when we review the Disability Profile.

Next, the profile can be examined for specific code type and additional interpretive hypotheses developed.  General personality characteristics and symptom descriptors can be found for the specific code type in one of the interpretative texts available that summarize empirical findings for specific populations (Butcher, 1990; Butcher et al, 1989; Graham, 1990, Greene, 1991).  Beyond that, there are some common code types that occur frequently in chronic pain samples.  For instance, various combinations of scales 1, 2, and 3 are probably the most common (greater than two-thirds as found by Slesinger et al., 2002).  According to Keller and Butcher (1991), the interpretive guidelines of Fordyce (1979) developed on the MMPI apply to the MMPI-2 as well.  It is beyond the scope of this course to present the entire interpretive strategy and the reader is referred to the original reference for more information.  As a brief example, Fordyce (1979) suggests that scale 1 indicates a general readiness to emit pain behaviors.  Further, the relative elevation of scale 2 compared to scales 1 and 3 can suggest the “cost” to the individual for remaining in the sick role.  Elevations on scales 1 and 3 with a low 2 indicate a low “cost” or distress associated with the condition. Thus, the patient might find aspects of the sick role reinforcing. Alternatively, elevations on all three scales indicate a readiness to emit pain behaviors and a great deal of distress over the situation.  Elevations on scales 1-2-3-4 typically describe a passive-dependent person who has a considerable readiness to seek nurturance and support from others.  A person with this profile might use pain behaviors to attain support from others and shelter from responsibility demands.   

The previous discussion should make it clear that traditional interpretive statements may not apply when using the MMPI-2 with a chronic pain patient.  Ethical principles of psychological testing mandate that the practitioner be aware of these issues (See the course Ethical Issues in Behavioral Health Practice).  Table 12 illustrates the difference between traditional interpretive statements and ones that might be more appropriate for a chronic pain patient.

There is an important caveat related to using one-, two-, or three-point code type interpretive strategies with the MMPI-2 and chronic pain patients:  In using the MMPI-2 with any patient population that is discrepant from the standardization sample (e.g. chronic pain), care must be taken to substantiate interpretive results (either traditional, chronic pain related, or some other).  As discussed by Senior and Douglas (2001), it is useful to employ a method of systematically either accepting or rejecting potential interpretations of scale elevations based upon additional information.  As Senior and Douglas (2001) suggest, “Using this hypothesis testing approach to MMPI-2 interpretation means that the interpretative statements that we as clinicians apply to a particular case are based upon confirmed patterns of endorsement that consistently reflect particular content” (p. 209).  The use of the supplementary and content scale can be useful in this regard.  For instance, an elevation of Scale 2 (D) is interpreted as reflecting depression-related symptoms only when the content scale, DEP, is also elevated.  If DEP is not elevated, then subscale analysis of Scale 2 is completed (sub-scales of Scale 2 include Subjective Depression, Psychomotor Retardation, Physical Malfunctioning, Mental Dullness, and Brooding).  In a chronic pain patient population, scale 2 may be elevated due to reasons more related to the pain than depression including such comorbid symptoms as sleep disruption, physical symptoms, and low energy.  Determination of whether an elevation on scale 2 reflects distress (depression) or something else will also impact the validity of the chronic pain code type interpretations presented in Table 12. 

INVALID USE OF THE MMPI-2

During the course of completing medical-legal evaluations, and doing consultations to attorneys, I have reviewed thousands of MMPI-2 reports used in treatment and forensic arenas (e.g. QME, AME, IME).  Based on this experience, I can summarize some of the common problems I see in using the MMPI-2 in assessing injured workers with chronic pain.  When these problems occur, the validity of the results cannot be substantiated. 

Use of the Short Form MMPI-2

Administration of the first 370 items of the MMPI-2 constitutes the “short form”.  If this form is used only three of the Validity Scales (L, F, K) and the Ten Clinical Scales can be scored.  Limiting the MMPI-2 to only the first 370 items excludes other important Validity scales, the 11 Supplemental Scales, 15 Content Scales, and some of the critical items. In a forensic assessment (e.g. QME, AME, IME), there can be little justification for using the short form of the MMPI2 since many important scales cannot be scored and much information is lost (e.g. not all of the Validity scales are available, no sub-scales, etc).  As discussed by Pope et al. (2006), “However, in the past few years there has been a reemergence of MMPI-2 short forms that make similar promises of ‘saving time’.  And although they may have value in limited application for some situations….until they have established adequate validity, reliability, sensitivity, and specificity for specific assessment, they lack a sufficient research base and track record for use in clinical or forensic assessment”. (p. 28)    

Limited Hand Scoring of the MMPI-2 

There are different methods for scoring the MMPI-2.  By far the most common method currently is some type of automated computer scoring (A computer scoring service must be licensed by the University of Minnesota to provide the service.  The cost is less than 8 dollars per scoring which includes extended scales, but not interpretation).  If you see computer scoring that does not contain copyright information, it may be from an individually developed software program.  If so, it may be a violation of copyright laws and the scoring rules are unknown. Hand scoring is available on an unlimited basis once the templates are purchased from the test publisher.  This is rarely done currently since scoring the vast number of important scales by hand is very tedious, time consuming, and vulnerable to error.   However, I do come across a number of MMPI-2 profiles in QME’s that have been hand scored.  When the profile is hand scored, a number of important validity and clinical scales are often left out. These omitted scales are those that might suggest the results are questionable or provide important information.  The same recommendations for computer scoring also apply to hand scoring – It is important to score all of the scales required according to the most recent MMPI-2 Manual (all the Validity, Clinical, Content, and Sub-Scales).  Anything less is ostensibly using the test in an unethical manner and certainly not appropriate in a forensic setting.

Not Scoring the Entire MMPI-2 

I see many instances in which only the 3 validity scales (L, F, K) and the 10 clinical scales are scored even though the entire MMPI-2 has been administered (scoring is usually by hand in these instances).  In 2001, the University of Minnesota Press released its revision of the MMPI-2 manual.  The 2001 manual identifies and reviews new scales developed since the original MMPI-2 manual was published in 1989.  The MMPI-2 Manual (2001) includes two new Validity Scales (Fp and S) in addition to the other, Content Component Scales for 12 of the 15 Content Scales, and five PSY-5 Scales.  

The Fp scale has been briefly reviewed previously.  The (Fp) Infrequency Psychopathology Scale identifies symptom exaggeration and embellishment that is not necessarily indicative of psychopathology. The Fp Scale is an infrequency Scale that indicates unusual response to the MMPI-2 items through claiming excessive, unlikely symptoms that are not typically endorsed by psychiatric patients In cases of severe psychopathology, the F Scale can be elevated resulting in false-positive classifications (mistakenly concluding that a test taker is exaggerating or embellishing symptoms, when in fact he or she is not).  The Fp Scale was developed to decrease the frequency of false-positive classifications of malingering or symptom exaggeration.  The Fp Scale serves its purpose quite well and should always be part of a forensic assessment. If a patient scores high on the Fp Scale, he or she is endorsing a large number of rare symptoms that even actual psychiatric patient infrequently endorse.  In the high range, (T-scores from 80-89) this indicates an extremely exaggerated response set which reflects an attempt to claim extreme or unusual psychiatric symptoms.  At a higher range, (T-scores from 90-109) the profile is possibly invalid due to the individual claiming an extreme number of rare psychiatric symptoms.  At extremely high levels, (T-score greater than 110) there is an indication of likely malingering of psychiatric symptoms due to extreme item endorsement.  Interpreted hypotheses for extremely elevated Fp scores (according to Pope et al, 2000) include likely symptom exaggeration, faking psychological problems, or malingering of mental health symptoms.  Often, psychologists overlook the Fp Scale in matters that might involve malingering.  Given the availability of peer reviewed research and data supporting the Fp Scale, it can no longer be considered a research scale.  Therefore, the Fp Scale can be invaluable in determining whether or not a patient was attempting to engage in conscious symptom exaggeration or malingering.  The S (Superlative Self Presentation) Scale has been developed to identify defensive test takers seeking to describe themselves in superlative terms.  The 2001 manual includes content component scales for 12 of the 15 content scales. The content component scales detail various characteristics associated with elevated content scales. The Low Self-Esteem scale, for example, can be elevated in response to general low self-esteem, and/or Self-Doubt, and/or Submissiveness. The content component scales should not be interpreted unless the parent scale falls at, or above, a T-score of 65.

Inappropriate MMPI-2 Interpretation 

There are many methods to MMPI-2 interpretation.  The first area of inquiry is to establish that that the profile is actually valid for interpretation. This is done by going over the consistency and validity measures as discussed previously (using of them and not just L, F, and K).  Once that is established, interpretation of the clinical scales can proceed.  If the profile is not valid, the profile should not have been interpreted and reasons for the invalidity explored.

By far the most common interpretative strategy is the 2-point code-type or some type of “configural” analysis.  This approach looks at the two highest scales on the MMPI-2 clinical profile and a lot of other information to be discussed.  In cases where the evaluator may be attempting to portray the patient as psychopathological as possible, interpretation statements for all of the MMPI-2 scales that are elevated above T=65 (the cut off for “abnormality”) will be listed. This is an inappropriate interpretive strategy for the MMPI-2 and not recommended by any text with which I am familiar (including the MMPI-2 Manual).  As referenced previously, it is not recommended by the MMPI-2 Manual. 

Researchers like Hathaway and McKinley (the originators of the MMPI), Dr. Graham (published extensively and a member of the MMPI-2 restandardization committee) area and Dr. Butcher (published extensively and a member of the restandardization committee) are all experts in the MMPI-2.  As reported in Graham (1990 and other sources), “From the MMPI’s inception, Hathaway and McKinley made it clear that the configural interpretation of an examinee’s scores was diagnostically richer and thus more useful than interpretation that examined single scales without regard for relationships among the scales”. Relative to this issue, the MMPI2 manual includes a “step-by-step” approach to interpreting the basic profile and this has been presented in Table 11.  

In summary, the MMPI-2 is almost always interpreted using a 2-point codetype method (the two highest scales) and occasionally and 3-point pattern analysis.  This approach was made clear by the MMPI authors, Drs Hathaway and McKinley and is reiterated in any text on interpreting the MMPI and the MMPI2 Manual (“code pattern analysis”).  If commonly scored sub-scales that can help determine why the “parent” scale was elevated are not scored (as recommended in the Manual), the evaluator has no detailed information about why any particular scale was elevated. An example might be the chronic pain patient who endorses a number of physical symptoms but is not clinically depressed or anxious.  Scales 2 (depression) and 7 (psychasthenia or anxiety) will both be elevated due to the physical symptoms even though significant depression or anxiety is not present. 

CASE EXAMPLE

Case Overview

This is a 42-year-old female, who alleges multiple industrial injuries related both to a specific incident and cumulative trauma during the course of her employment.  The applicant underwent evaluation (including MMPI-2 testing) on two occasions separated by approximately 5 years.  She is alleging a psychological injury derivative to the orthopedic injuries.  The first evaluation yielded a conclusion of no psychological injury on an industrial basis and probable malingering. The follow-up evaluation yielded a conclusion of an industrial psychological injury (primarily related to the development of a chronic pain and disability syndrome).  There were significant apportionment and other issues which are not reviewed here.  The following excerpts focus on the use of the MMPI-2 in the evaluation of the injured worker with chronic pain. 

Current Symptoms and Level of Function.  The following are the patient’s current symptom complaints and her reported level of function:

Psychological.  In an effort to understand the applicant’s symptoms of “stress and anxiety” she was questioned carefully about the type of psychological problems and symptoms she was experiencing.  She summarized her situation as being “physically and emotionally exhausted.” However, she consistently related this to her daughter’s current situation and reporting “I can’t help her.”  As will be discussed in greater detail subsequently, her daughter’s significant other recently died unexpectedly leaving her as a single mother with two young children.  The patient attempts to help take care of her grandchildren and was quite tearful when discussing this specific situation.  When asked about her primary areas of stress, she stated “the primary thing is that I can’t help my daughter or grandkids.”  Upon further questioning, she did report some distress related to her chronic pain “mostly in my lower back.” 

Physical.  The applicant currently complains of “a lot of pain.”  Initially, she was referring to almost total body pain.  Again, on more detailed questioning, she stated “it is mostly in my lower back.” She also reported numbness in the right leg.  She reported neck pain, bilateral shoulder pain (right greater than left), and upper extremity pain. 

Reported Level of Function. Similar to what she reported previously, she does engage in ADLs each day including “house cleaning.”  She stated she is currently “renting a couch” from a friend.  She reported that she is “up frequently during the night” and “I don’t sleep well.”  She will often sleep until 11:00 a.m.  She will then shower, dress, and engage in household ADLs.  She also currently spends a great deal of time caring for her grandchildren due to the situation described previously. 

Brief History

The applicant reported injuries to her lower back, shoulders, and upper extremities occurring as the result of a specific incident in 2002.  The applicant was working in a warehouse at the time of the injury.  She started working at the company in 2000.  Her last day of work was the date of the injury.    She stated that after the injury she was experiencing low back pain, mid back pain, neck pain, shoulder pain, as well as pain and “swelling” in both hands. 

As of 2005, she carried physical diagnoses of bilateral carpal tunnel syndrome and bilateral rotator cuff tears.  She began medical pain management in 2005.  In 2007, the applicant underwent left shoulder arthroscopy with arthroscopic subacromial decompression, mini open repair, and left rotator cuff.  During the course of the clinical interview, the applicant reported that this surgery provided no benefit and “did not work.” 

In early 2008, the treatment began focusing on the lower back.  This included diagnostic studies (e.g. imaging and electrodiagnostic).  Medication management was also being provided including a regimen of Vicodin ES, Compazine, Soma, Lyrica, Prevacid, Zoloft and Xanax.  The applicant underwent an L4-5 epidural. 

The applicant underwent the orthopedic AME in 2009.  The applicant was found to be permanent and stationary with no industrial injury to the lumbar or cervical spine.  Industrial injury was found relative to her shoulder.  The applicant followed-up with medical pain management.  She was continuing to complain of a variety of physical symptoms.  A drug screening was completed and found to be positive for methamphetamines.  At that point, the patient was directed to return to the clinic on a prn basis and instructed to find “an addiction medicine specialist.” 

Psychiatric Treatment History

The applicant reported she has been on a number of antidepressants over the years including Prozac “for a long time,” Zoloft (“for years”), and, most recently, Cymbalta.  She has also been on Xanax for over 20 years.  She has been in some type of psychotherapy for at least the past 25 years. 

Substance Use

The applicant reported that she rarely consumes alcohol and smokes .5 pack of cigarettes per day.  She currently denies use of any illicit or recreational drugs.  When asked specifically about a substance abuse or chemical dependency history, she denied any problems.  She basically states that “I grew up in the sixties.”  She states that she occasionally uses substances, but for each report of an instance of use, it is generally described as “one time,” and there is some justification. 

Comment:    The medical records clearly substantiate a long history of chemical dependency and substance abuse issues.  This includes poly-substance abuse of hallucinogens (mushrooms), methamphetamine, and marijuana.  On each occasion when the applicant has been discovered, she has had some rationalization or justification for the use.  For instance, when her toxicology screen came back positive while in medical pain management, she stated the methamphetamine use was just on one occasion with her daughter in an effort to provide energy to help care for her grandchildren. 

Medications

The applicant reported that her current medications include Xanax (1 mg, 1-3 per day, over the past 30 years), Norco (q.i.d.), Prevacid, Donnatal, and Dramamine.  She stated that she is also on something for hypertension.  She was on Cymbalta as provided by her family physician.  She was getting samples, but “ran out” a few weeks ago.  Currently, her family physician manages all of her medications. 

Psychological and Pain Assessments

The applicant was administered the assessment instruments as listed previously.  The following results were obtained:

Pain and Functional Ratings:

On the Pain Patient Questionnaire, the applicant currently rates her pain (0-100) at a 99, at its highest intensity, a 75 at its usual intensity, and a 70 at its lowest intensity.  Her pain rating during the clinical interview was at an 80 out 100. 

These pain ratings can be compared to what was previously obtained on 2/16/2005.  At that time, she rated her pain at a 100 at its highest intensity, a 97 at its usual intensity, and a 60 at its lowest intensity.  She stated her pain at the time of the clinical interview at that time was a 100 out of 100. 

Currently, the applicant states that the pain interferes with work, family chores, and recreational activities 100% of the time, while it interferes with sexual functioning 80% of the time.  She states she is in pain 99% of the time.  Her pain is exacerbated by standing, lifting, and moving.  It is reduced by lying down. 

Pain Drawing:

On the Pain Drawing, the applicant showed pain in multiple body regions including the posterior cervical region, the shoulders and upper extremities, bilaterally, as well as the low back with radiation to the lateral aspect of the right buttock and down the right lower extremity to the foot. 

MMPI-2:

In this case, we have two MMPI-2 tests available for comparison.  The first is from 2005 and the second is current (2010).  The difference will be discussed subsequently. Only the current MMPI-2 will be fully interpreted. 

MMPI-2 Administrating #1 (2005)

MMPI-2 Administration #2 (2010)

Analysis of the consistency of item endorsements was evaluated using the accepted standards (Greene, 1991; Pope, Butcher, and Seelan, 2000, 2006; Graham, 1990; and others).  There were no item omissions, which is within acceptable limits.  In addition, the VRIN (T=46) and the TRIN (T=58F) were within normal limits.  This suggests that the patient answered the questions in a consistent manner. 

Validity of the profile was assessed using accepted standards (see Greene, 1991; Deardorff, 2000; Keller & Butcher, 1991; and Pope et al., 2000, 2006; MMPI Manual, 2001).  As can be seen, F, Fb, Fp, K, and S were all within normal limits.  The applicant did show an elevation on L (T=71).  Elevations in this range on the L Scale can be due to a number of factors including a normal person who is very self-controlled and lacks insight into her own behavior, persons with religious and moralistic training, occupations that deny even the most common human faults, unsophisticated persons who are trying to create an unusually favorable impression of themselves, clients whose dynamics revolve around denial, and/or naïve self views.  Overall, in this case, the L Scale can be construed as a measure of psychological sophistication with high scores indicating a lack of sophistication (see Greene, 1991, page 107).  This will be taken into account in terms of interpretation of the Clinical Profile. 

The clinical Profile can be interpreted from two different perspectives: 

(1) Personality and psychopathology features

(2) Relative to the chronic pain. 

As can be seen, the patient produced a very clear 1-3/3-1 Profile.  Scale 2 must also be taken into account in terms of its own elevation and the relative differences between Scales 1-3, and 2. 

Personality-Psychopathology Interpretation.  The most commonly accepted interpretive method for the MMPI is to utilize 2-point code types.  This profile is clearly a 1-3/3-1 code type.  In patients with similar profiles, some tension may be reported, but severe anxiety and depression are usually absent, as are clearly psychotic symptoms.  This is certainly consistent with the subclinical elevation on Scale 2.  Rather than being grossly incapacitated in functioning, these patients generally function, but at a reduced level of efficiency.  The somatic complaints represented by the elevations on 1-3 include such things as headaches, chest pain, back pain, and numbness or tremors of the extremities.  Eating problems are common.  Other physical complaints often include weakness, fatigue, dizziness, and sleep disturbance.  The physical symptoms increase in times of stress and there is a clear secondary gain associated with the symptoms.  Patients with this profile present themselves as normal, responsible, and without fault.  They like excessive use of denial, projection, and rationalization, and they blame others for their difficulties.  They prefer medical explanations for their symptoms and often lack insight into psychological factors underlying their symptoms.  These patients tend to be rather immature, egocentric, and selfish.  They are generally insecure and have a strong need for attention, affection, and sympathy.  They are very dependent, but they are uncomfortable with the dependency and experience conflict because of it.  Although they are outgoing and socially extroverted (notice the very low elevation on Scale 0), their social relationships tend to be shallow and superficial.  They will often lack genuine emotional involvement with other people.  Patients with this profile harbor resentment and hostility towards other people, particularly those who are perceived as not fulfilling their needs for attention.  These patients need to convince other people that they are logical and reasonable, and they are conventional and conforming in their attitudes and values.  Because of the unwillingness to acknowledge psychological factors underlying their physical symptoms, these persons are difficult to motivate in traditional psychotherapy.  They are reluctant to discuss psychological factors that might relate to somatic symptoms.  It is occasionally possible for these patients to discuss problems, as long as no direct link to somatic symptoms is suggested. 

Chronic Pain Interpretation. This profile might also be interpreted relative to chronic pain issues.  According to Fordyce, patients with similar profiles present with a wide variety of vague and diffuse somatic complaints.  In these cases, there is often a very low correlation between subjective and objective findings.  These patients show pain behaviors and somatic complaints far beyond what would be expected due to nociceptive input and objective findings.  These patients show a high readiness to admit pain behaviors, but very little emotional distress associated with their reports of pain and other symptoms.  These patients will often use complaints of pain to extricate themselves from stressful situations.  The extreme elevations on Scales 1 and 3, in conjunction with the non-clinical elevation on Scale 2 (depression) suggest that this applicant is not uncomfortable in the sick role and may find aspects of it reinforcing.  As such, the applicant is showing a high readiness to admit pain behaviors, along with multiple somatic complaints, in conjunction with minimal distress regarding these symptoms.  Given the findings on the Validity Scale, it is not believed that this is due to conscious malingering (however, partial malingering was determined at the time of my evaluation five years ago).  Rather, this profile represents someone who unconsciously uses somatic symptoms to achieve specific psychological goals. 

Lastly, this profile suggests a very high level of symptom amplification.  These findings must be taken into account relative to any type of physical or psychiatric/psychological evaluation.  In addition, the MMPI-2 is one of the most objective measures of the various aspects of depression.  As can be seen, Scale 2 (Depression; T=62) is subclinical, and suggests that the applicant is not suffering from any type of depressive disorder.  Analysis of other MMPI-2 Scales that assess depression (DEP, T=55; as well as four of the five Harris-Lingoes Subscales for Depression) were all subclinical.  The only elevation was found on Physical Malfunctioning (D3; T=85), and this was due to endorsement of physical symptoms.  Therefore, although the applicant may report various symptoms of depression (subjective data) validated objective data such as the MMPI-2 suggests a lack of depressive symptoms.  This is consistent with other findings in the evaluation.

These MMPI-2 findings are also very important relative to the lack of elevation on Scale 7 (Pt, a measure of anxiety type symptoms).  As can be seen, on Scale 7, she achieved a T-score of 59, which is subclinical.  In addition, Content Scales measuring Anxiety were not particularly elevated including ANX (T=64),  A (T=44), and the Fears Subscales (both T less than 50).  Although the applicant provides a self report of symptoms consistent with generalized anxiety and panic episodes, this is not substantiated on objective testing.  Again, one must keep in mind, that these Clinical Scale results are part of a valid profile. 

As can be seen, the profile configuration is similar to what was found previously  aside from a significant increase in Scale 1 and 3.  As will be discussed, I believe this represents develop of the chronic pain syndrome over time.

Multi-Dimensional Pain Inventory:

The MPI scores for the two test periods are as follows:

The applicant was re-administered the Multi-Dimensional Pain Inventory.  Previously, Cluster Classification was not possible and was coded as “anomalous.”  At that time, her pattern of scores did not make sense relative to established theories of chronic pain.  On this occasion, Cluster Classification was possible and she was placed in the “Dysfunctional” Cluster Classification Group.  As can be seen, she scored above average for Pain Severity and Affective Distress.  Interestingly, she scored above average for almost all of the Activity Categories including Household Chores, Activities Away From Home, Social Activities, and General Activity Level.  This does suggest that the applicant is trying to stay busy and active, regardless of her reports of chronic pain and disability. 

We can also make comparisons with her previous findings, from five years ago.  As can be seen, her Pain Severity rating has actually improved even though her reported Interference With Functioning has increased.  Her reported sense of Life Control has decreased.  Her reported Affective Distress on the MPI has increased.  In the face of these inconsistent findings, her Activity Scales have generally increased since her previous assessment, especially for Activities Away From Home, Social Activities, and General Activity Level.  Some of these inconsistencies may be due to the fact that the initial MPI results were determined to be “anomalous” and the patterns did not really make sense.  The current assessment is more consistent with the other data that is available for review.  It should be noted that the MMPI Scale II measure is much more sophisticated relative to assessing all dimensions of depression versus the Affective Distress Scale on the MPI. 

Beck Depression Inventory-2.  On the BDI, the applicant scored a 16, which is in the mild depression range of self-rated depression.  This compares with her previous BDI-2 findings as obtained by earlier this year by another evaluator (score of 19).  This is also in the range of mild depression. 

Analysis of individual items on the BDI-2 include a 3-I feel like I am being punished; and a 3 on changes in sleep pattern-I wake up 1-2 hours early and can’t get back to sleep.  These high scoring items must be taken into account relative to the applicant’s complaint of chronic pain and the adversarial nature of the workers’ compensation system.  Chronic pain patients will often artificially elevate BDI-2 scores due to endorsement of physical symptoms unrelated to actual depression as well as an endorsement of items related to being in the workers’ compensation system (e.g. I feel like I am being punished, “by the insurance carrier”).  Taking into account the results of the MMPI-2, even the BDI-2 score of 16 is likely higher than what the applicant is actually experiencing in terms of any depression. 

Epworth Sleepiness Scale.  The Epworth Sleepiness Scale is a very face valid instrument (similar to the BDI) and should be interpreted taking into account establishing credibility of the patient’s self-report (e.g. MMPI-2 results).  In this case, the applicant was considered to be credible based upon the MMPI-2 results and other objective measures.  On the Epworth Sleepiness Scale, the applicant scored a 4, which is in the normal range.  This suggests that she is not experiencing any type of sleepiness or anergia during the day.  It is also suggestive of restorative sleep.  These results are exactly the same as obtained in April, 2009, as part of the orthopedic AME.  Scores on the Epworth Sleepiness Scale between 0 and 10 are considered in the normal range. 

Again, these objective measures of subjective symptoms are somewhat discrepant.  During the course of the clinical interview, the applicant reports sleep disruption and lack of energy during the day.  She also reported this to other evaluators.  Even so, standardized objective measures such as the Epworth indicates she is likely obtaining restorative sleep and not experiencing any significant sleepiness or anergia during the day.  This amplification of symptoms on subjective self-report is also consistent with the MMPI-2 Profile.  Therefore, one must be careful to develop opinions based only on objective data.  In addition, any subjective self-report data provided by the applicant must take into account the high degree of likely symptom amplification.  Again, based on the Validity Scales of the MMPI-2, I do not believe that this is occurring due to malingering.  Rather, it is more likely due to a personality style and various factors that are reinforcing pain behaviors and disability. 

CONCLUSIONS

The MMPI/MMPI-2 is the most commonly used objective psychometric instrument used in the comprehensive evaluation of chronic pain patients.  It is used for a variety of purposes including identification of comorbid psychopathology, elucidation of personality and behavioral characteristics, planning treatment interventions, screening for invasive procedures directed at pain relief, and to evaluate malingering in a forensic setting.  It is important to be aware of special issues when interpreting the MMPI-2 profiles of chronic pain patients.

RESOURCES

Link for MMPI-2 current research:  MMPI-2 Updates

REFERENCES

Aguerrevere et al. (2008).  Detecting malingering in traumatic brain injury and chronic pain with an abbreviated version of the Meyers Index for the MMPI-2.  Archives of Neuropsychology, 23, 831-838.

Arbisi, P.A. and Butcher, J.N. (2004).  Psychometric perspectives on detection of malingering of pain: Use of the MMPI-2.  Clinical Journal of Pain, 20, 383-391.

Ben-Porath, Y.S., & Butcher, J.N.  (1988).  Exploratory analyses of rewritten MMPI items.  Paper presented at the 23rd Annual Symposium on Recent Developments Use of the MMPI, St. Petersburg, Florida.

Bianchini et al. (2008).  Classification accuracy of MMPI-2 validity scales in the detection of pain-related malingering: A known-groups study.  Assessment, 15, 435-449.

Bigos, S.J., Battie’, M.C., & Spengler, D.M. (1991). A prospective study of work perceptions and psychological factors affecting the report of back injury. Spine, 16, 1-6.

Bradley, L.A.  (1995).  Biopsychosocial model and the Minnesota Multiphasic Personality Inventory: Even psychologists make cognitive errors.  Pain Forum, 4, 97-100.

Bradley, L.A., Haile, J.M., & Jaworski, T.M.  (1992).  Assessment of psychological status using interviews and self-report instruments.  In D.C. Turk and R. Melzack (Eds.),  Handbook of Pain Assessment.  New York: The Guilford Press.

Burchiel, K.J., Anderson, V.C., Wilson, B.J., Denison, D.B., Olson, K.A., & Shatin, D.  (1995).  Prognostic factors of spinal cord stimulation for chronic back and leg pain.  Neurosurgery, 36, 1101-1111.

Butcher, J.N.  (1990).  MMPI-2 in Psychological Treatment.  New York: Oxford University Press.

Butcher, J.N.  (1972).  Objective Personality Assessment: Changing Perspectives.  New York:  Academic Press.

Butcher, J.N. , Dahlstrom, W.G., Graham, J.R. Tellegen, A., & Kaemmer, B. (1989).  Manual for administration and scoring.  MMPI-2.  Minneapolis, MN: University of Minnesota Press.

Butcher, J.N., Graham, J.R., Williams, C.L. & Ben-Porath, Y.S.  (1989).  Development and Use of the MMPI-2 Content Scales. Minneapolis, MN: University of Minnesota Press.

Butcher, J.N. & Harlow, T. (1987). Personality assessment in personal injury cases.  In A. Hess & I. Weiner (Eds.), Handbook of Forensic Psychology (pp. 128-154). New York: Wiley.

Cohen, N.  (1987).  The response of chronic pain patients to the original and revised versions of the Minnesota Multiphasic Personality Inventory.  Unpublished doctoral dissertation, University of Minnesota.

Costello, R.M., Hulsey, T.L., Schoenfeld, L.S., & Ramamurthy, S.  (1987).  P-A-I-N: A four-cluster MMPI typology for chronic pain.  Pain, 30, 199-209.

Cronbach, L. (1990).  Essentials of psychological testing (5th ed.).New York: Harper & Row.

Deardorff, W.W., Chino, A.F. & Scott, D.W.  (1993).  Characteristics of chronic pain patients: factor analysis of the MMPI-2.  Pain, 54, 153-158.

Dush, D.M., Simons, L.E., Platt, M., Nation, P.C. & Ayres, S.Y.  (1994).  Psychological profiles distinguishing litigating and nonlitigating pain patients: subtle, and not so subtle.  Journal of Personality Assessment, 62, 299-313.

Fishbain, D.A..  (1996).  Some Difficulties with the Predictive validity of Minnesota Multiphasic Personality Inventory. Pain Forum, 5, 81-82.

Fishbain, D.A.  (1997).  Can personality disorders in chronic pain patients be accurately measured?  Pain Forum, 6, 16-19.

Fishbain et al. (1999).  Chronic pain disability exaggeration-malingering and submaximal effort research.  Clinical Journal of Pain, 15, 244-274.

Fordyce, W.E.  (1979).  Use of the MMPI in the assessment of chronic pain.  Clinical Notes on the MMPI, #3.  Nutley, NJ:Hoffman-LaRoche.

Fordyce, W.E., Bigos, S.J., Battie’, M.C., & Fisher, L.D. (1992).  MMPI Scale 3 as a predictor of back injury report:  What does it tell us?  Clinical Journal of Pain, 8, 222-226.

Gatchel, R.J.  (1997).  The significance of personality disorders in the chronic pain population.  Pain Forum, 6, 12-15.

Gatchel, R.J. and Weisberg, J.N. (2000).  Personality Characteristics of Patients with Pain.  Washington, D.C.: APA Books.

Gatchel et al. (2006).  MMPI Disability Profile: The least known, most useful screen for psychopathology in chronic occupational spinal disorders.  Spine, 31, 2973-2978.

Graham, J.R.  (1990).  MMPI-2:  Assessing Personality and Psychopathology. New York: Oxford University Press.

Greene, R.L.  (1991).  The MMPI-2/MMPI:  An Interpretive Manual.  Boston: Allyn & Bacon.

Guck, T.P., Meilman, P.W., Skultety, M., & Poloni, L.D.  (1988).  Pain-patient Minnesota Multiphasic Personality Inventory (MMPI) subgroups: Evaluation of long-term treatment outcome.  Journal of Behavioral Medicine, 11, 159-169.

Haggard et al. (2008).  Relationship between the MMPI-2 and psychosocial measures in a heterogeneous pain population.  Rehabilitation Psychology, 53, 471-478.

Hathaway, S.R. & McKinley, J.C.  (1943). The Minnesota Multiphasic Personality Schedule (revised).  Minneapolis, MN:  University of Minnesota Press.

Helmes, E. (1994). What types of useful information do the MMPI and MMPI-2 provide on patients with chronic pain?  American Pain Society Bulletin, 4 (1), 1-5.

Helmes, E. & Reddon, J.R. (1993). A perspectve on developments in assessing psychopathology: A critical review of the MMPI and MMPI-2. Psychological Bulletin, 113, 453-471.

Keefe, F.J., Lefebvre, J.C., & Beaupre, P.M. (1995). The Minnesota Multiphasic Personality Inventory in chronic pain:  Security blanket or sound investment?  Pain Forum, 4, 101- 103.

Keller, L.S. & Butcher, J.N.  (1991).  Assessment of Chronic Pain Patients with the MMPI-2.  Minneapolis, MN: University of Minnesota Press. 

Kleinke, C.L.  (1994).  MMPI scales as predictors of pain-coping strategies preferred by patients with chronic pain.  Rehabilitation Psychology, 39, 123-128.

Love, A.W., & Peck, C.L. (1987).  The MMPI and psychological factors in chronic low back pain: A review. Pain, 28, 1-28.

Lumley, M.A.,  Asselin, L.A., & Norman, S.  (1997).  Alexithymia in chronic pain patients.  Comprehensive Psychiatry, 38, 160-165. 

Magni, G., Marchutti, M., & Moreschi, C.  (1993).  Chronic musculoskeletal pain and depressive symptoms in the National Health and Nutrition Examination: I. Epidemiological follow-up study.  Pain, 53, 163-168.

Main, C.J., & Spanswick, C.C. (1995a).  Personality assessment and the Minnesota Muliphasic Personality Inventory, 50 years on: do we still need our security blanket?  Pain Forum, 4, 90-96.

Main, C.J., & Spanswick, C.C. (1995b).  Response to our commentators. Pain Forum, 4, 110-113.

Maruta, T., Goldman, S., Chan, C.W., Ilstrup, D.M., Junselman, A.R., & Colligan, R.C.  (1997).  Waddell’s nonorganic signs and Minnesota Multiphasic Personality Inventory profiles in patients with chronic low back pain.  Spine, 22, 72-75.

McGill, J.C., Lawlis, G.F., Selby, D., Mooney, V., & McCoy, C.E.  (1983).  The relationship of Minnesota Multiphasic Personality Inventory profile clusters to pain behaviors.  Journal of Behavioral Medicine, 6, 77-92.

McGrath, R.E., Sweeney, M., O’Malley, W.B. & Carlton, T.K.  (1998).  Identifying psychological contributions to chronic pain complaints with the MMPI-2: The role of the K Scale.  Journal of Personality Assessment, 70, 448-459.

Moore, J.E., McCallum, S., Holman, C., & O’Brien, S.  (1991).  Prediction of return to work after pain clinic treatment by MMPI-2 clusters.  Paper presented at the meeting of the American Pain Society, New Orleans, LA.

Moore, J.E., McFall, M.E., Kivlahan, D.R., & Capestany, R.  (1988).  Risk of misinterpretation of MMPI Schizophrenia scale elevations in chronic pain patients.  Pain, 32, 207-213.

Naliboff, B.D., Cohen, M.J., & Yellin, A.N.  (1982).  Does the MMPI differentiate chronic illness from chronic pain?  Pain, 13, 333-341.

Nordin et al. (2005).  MMPI-2 subgroups in a sample of chronic pain patients.  Scandinavian Journal of Psychology, 46, 209-216.

North, R.B., Kidd, D.H., Wimberly, R.L., & Edwin, D.  (1996).

Prognostic value of psychological testing in patients undergoing spinal cord stimulation: a prospective study.  Neurosurgery, 39, 301-311.

Pancoast, D.L. & Archer, R.P. (1989). Original adult MMPI norms in normal samples: A review with implications for future developments.  Journal of Personality Assessment, 53, 376-395. 

Pincus, T., Callahan, L.F., Bradley, L.A., Vaughn, W.K., & Wolfe, R.  (1986).  Elevated MMPI scores for hypochondriasis, depression, and hysteria in patients with rheumatoid arthritis reflect disease rather than psychological status.  Arthritis and Rheumatism, 29, 1456-1466.

Polatin, P.B., Kinney, R.K., Gatchel, R.J., Lillo, E., & Mayer, T.G.  (1993). Psychiatric illness and chronic low back pain.  Spine, 18, 66-71.

Pope et al. (2006).  The MMPI, MMPI-2, and MMPI-A in Court, Third Edition.  Washington, DC: APA Books.

Prokop, C.K.  (1986).  Hysteria scale elevation in low back pain patients:  A risk factor for misdiagnosis?  Journal of Consulting and Clinical Psychology, 54, 558-562.

Reich, J., Tupin, J.P., & Abramowitz, S.I.  (1983).  Psychiatric diagnosis of chronic pain patients.  American Journal of Psychiatry, 140, 1495-1498.

Riley, J.L., Robinson, M.E., Geisser, M.E., & Wittmer, V.T.  (1993).  Multivariate cluster analysis of the MMPI-2 in chronic low-back pain patients. The Clinical Journal of Pain, 9, 248-252.

Riley, J.L., Robinson, M.E., Geisser, M.E., Wittmer, V.T., & Smith, A.G.  (1995).  Relationship between MMPI-2 cluster profiles and surgical outcome in low-back patients.  Journal of Spinal Disorders, 8, 213-219.

Rogers, R.  (2008).  Clinical Assessment of Malingering and Deception, Third Edition.  New York: Guilford.

Samuels, J.F., Nestadt, M.D., Romanoski, A.J., Folstein, M.F., & McHugh, P.R. (1994).  DSM-III personality disorders in the community.  American Journal of Psychiatry, 151, 1055-1062.

Sanders, S.H.  (1995).  Minnesota Multiphasic Personality Inventory and clinical pain:  The baby or the bathwater?  Pain Forum, 4, 108-109.

Schmidt, J.P. & Wallace, R.W.  (1982).  Factor analysis of the MMPI profiles of low back pain patients.  Journal of Personality Assessment, 46, 366-369.

Senior, G. and Douglas, L. (2001).  Misconceptions and misuse of the MMPI-2 in assessing personal injury claimants.  Neurorehabilitation, 16. 203-213.

Slesinger et al. (2002).  MMPI-2 characteristics in a chronic pain population.  Assessment, 9, 406-414.

Sternbach, R.A. (1974). Pain Patients: Traits and Treatments.  New York: Academic Press.

Strassberg, D.S. and Russell, S.W. (2000). MMPI-2 content scale validity within a sample of chronic pain patients.  Journal of Psychopathology and Behavioral Assessment, 22, 47-60.

Sullivan, M.J.L., Reesor, S., Mikail, R., & Fisher, R.  (1992).  The treatment of depression in chronic low back pain: review and recommendations, Pain, 50, 5-13. 

Turk, D.C.  (1990).  Customizing treatments for chronic pain patients:  Who, what, and why.  Clinical Journal of Pain, 6, 255-270.

Turk, D.C. & Fernandez, E. (1995). Peronality Assessment and the Minnesota Multiphasic Personality Inventory in chronic pain: underdeveloped and overexposed. Pain Forum, 4, 104-107.

Turk, D.C. & Melzack, R.  (1992).  Handbook of Pain Assessment.  New York: The Guilford Press.

Turk, D.C. & Rudy, T.E.  (1986).  Assessment of cognitive factors in chronic pain: a worthwhile enterprise?  Journal of Consulting and Clinical Psychology, 54, 760-768.

Waddell, G., McCulloch, J., Kummel, E.,  & Venner, R.M. (1980).  Non-organic physical signs in low-back pain.  Spine, 5, 117-125.

Waddell, G., Pilowsky, I.,  & Bond, M.R. (1989).  Clinical assessment and interpretation of abnormal illness in low back pain.  Pain, 39, 41-53.

Wall, P.D. & Melzack, R.  (Eds.). (1984).  Textbook of Pain.  New York: Churchill-Livingstone.

Watson, D.  (1982).  Neurotic tendencies among chronic pain patients:  An MMPI item analysis.  Pain, 14, 365-385.

Weisberg, J.N. & Keefe, F.J.  (1997).  Personality disorders in the chronic pain population:  Basic concepts, empirical findings and clinical implications.  Pain Forum, 6, 1-9.

Weisberg, J.N. & Keefe, F.J.  (1997b).  Methodological considerations for diagnosis of personality disorders in chronic pain patients.  Pain Forum, 6, 20-21.

Wygant et al. (2009).  Examination of the MMPI-2 Restructured Form (MMPI-2-RF) validity scales in civil forensic settings: Findings from simulation and known group samples. Archives of Clinical Neuropsychology, 24, 671-680.