Mark ZimmermanProfessor of Psychiatry and Human Behavior
Mark Zimmerman, MD, is the Director of Outpatient Psychiatry at Rhode Island Hospital and director of the Rhode Island Methods to Improve Diagnostic Assessment and Services (MIDAS) project. The goal of the MIDAS project has been to integrate research methodology into routine clinical practice in order to examine a number of issues related to diagnostic comorbidity and treatment outcome. Dr. Zimmerman has conducted research in psychiatry for more than a two decades resulting in more than 300 publications in peer-reviewed journals. Much of Dr. Zimmerman's research has been in the area of assessment and diagnosis. While a research assistant at the University of Iowa, Dr. Zimmerman developed a self-report questionnaire to diagnose DSM-III major depressive disorder, and was one of the authors of the first semi-structured interview to assess the DSM-III personality disorders. Through his academic career he has continued his interest in scale development.
Mark Zimmerman, M.D., is the director of outpatient psychiatry at Rhode Island Hospital and director of the Rhode Island Methods to Improve Diagnostic Assessment and Services (MIDAS) project. The goal of the MIDAS project has been to integrate research methodology into routine clinical practice in order to examine a number of issues related to diagnostic comorbidity and treatment outcome. Thus far more than 200 papers have been published in peer-reviewed journals from the MIDAS project database.
The Rhode Island Methods to Improve Diagnostic Assessment and Services (MIDAS) project is an ongoing clinical research project that began more than a decade ago. To date, approximately 2,500 patients presenting for treatment at the Rhode Island Hospital Department of Psychiatry outpatient practice have been evaluated with semi-structured diagnostic interviews making this the largest clinical epidemiological study ever conducted. In 2003 we published a book chapter describing the background and methods the MIDAS project and summarizing the results published in several journal articles. To date more than 100 articles have been published based on the MIDAS project dataset. Below is the book chapter that summarizes much of this work.
DSM-III was published in 1980 (American Psychiatric Association 1980). The principal goal of changing the way psychiatric disorders were defined from DSM-II's prototypic descriptions of disorders to the Washington University diagnostic approach of specifying inclusion and exclusion criteria (Feighner, Robins, Guze et al. 1972) was to improve diagnostic reliability. The early reviews of DSM-III suggested that it succeeded in "solving" the reliability problem (Klerman, Vailant, Spitzer et al. 1984), and up until recently few questions had been raised about the adequacy of diagnostic practice in clinical settings in the post-DSM-III era. However, during the past few years, several studies have raised concerns about the thoroughness and accuracy of diagnostic evaluations conducted by mental health professionals in routine clinical practice (Basco, Bostic, Davies et al. 2000; Shear, Greeno, Kang et al. 2000; Zimmerman and Mattia 1999d).
Back in 1980, when DSM-III was published, I was a graduate student in clinical psychology at the University of Iowa. During this time I began working in the Department of Psychiatry, which was then chaired by George Winokur (one of the co-authors of the Washington University criteria), and worked as a researcher in the Department for six years. Working with Bill Coryell on one of the two inpatient units of Psychopathic Hospital I was trained in the administration of the Schedule for Affective Disorders and Schizophrenia (SADS, Endicott and Spitzer 1978) and the Hamilton Rating Scale for Depression (HRSD, Hamilton 1960). Working with Bruce Pfohl, I helped develop the first semi-structured interview to assess the DSM-III personality disordersthe Structured Interview for DSM-III Personality (SIDP, Stangl, Pfohl, Zimmerman et al. 1985). The three of us, along with Dalene Stangl, conducted a large psychobiological study of depression which included research interviews for depressive symptoms, personality disorders, family history of psychiatric disorders, life events, social support, and follow-up status (Pfohl, Stangl, and Zimmerman 1984; Zimmerman, Coryell, Pfohl et al. 1986; Zimmerman, Pfohl, Coryell et al. 1991). In another study conducted during my tenure at Iowa I administered the full SADS to a series of nonmanic psychotic patients, conducted six and twelve-month follow-up interviews of these patients with a standardized instrument, and coordinated a family study of psychiatric disorders in which first-degree family members of patients and controls were administered the Diagnostic Interview Schedule (DIS) and the SIDP (Coryell and Zimmerman 1987; Zimmerman and Coryell 1990). Thus, my initial experience in evaluating psychiatric patients, which lasted for six years, consisted almost entirely of the administration of comprehensive, standardized research interviews.
The structure and content of these interviews were imprinted, and when I began my clinical career I found that I was essentially administering the instruments. Consequently, my initial intake appointments typically lasted at least two hours whereas my colleagues completed theirs in less than an hour. When I examined my colleagues' charts I noticed that they were unlikely to diagnose patients with more than one psychiatric disorder. This was clearly in contrast to the high comorbidity rates found in studies of patients based on research instruments, and lower than the rates I found in my patients. It was this observation, in the context of my research background, that was the impetus for initiating the Rhode Island Methods to Improve Diagnostic Assessment and Services (MIDAS) project in which research assessment methods have been incorporated into routine clinical practice.
The MIDAS project consists of two major componentsa structured initial diagnostic evaluation and standardized follow-up outcome ratings. An expanded version of the Structured Clinical Interview for DSM-IV (SCID, described below) has thus far been administered to more than 1,800 psychiatric outpatients presenting for treatment. The study is ongoing; consequently, the sample size in our more recent publications is larger than initial papers. After the study was underway and running smoothly, the full SIDP for DSM-IV was introduced and more than 1,000 patients have been evaluated on DSM-IV Axis II. From the outset we assumed that comprehensive structured interviews were unlikely to be incorporated into many other clinical practices; thus, we developed a self-administered questionnaire to screen for the most common DSM-IV Axis I disorders diagnosed in outpatient settings (Zimmerman and Mattia 1999f; Zimmerman and Mattia 2001a; Zimmerman and Mattia 2001b). The goal was to develop a measure with good psychometric properties that could be incorporated into routine clinical practice.
Adequate outcome evaluation requires attention to comorbidity; a wealth of data indicates that comorbidity is the rule rather than the exception. An outcome instrument should also monitor psychosocial functioning and quality of life. Structured research interviews, either on a regular basis or simply as pretreatment and post-treatment endpoint assessments, are too time consuming and thus expensive for use in routine outpatient mental health settings. A major obstacle towards implementing a comprehensive outcome program in routine clinical practice is the lack of a single, user-friendly instrument that assesses each of these important domains of treatment outcome. Consequently, as part of the MIDAS project we developed a DSM-IV based self-report questionnaire and clinician-rating system that can provide information to clinicians and clinics in a cost-effective manner.
To date the focus of the publications from the MIDAS project has been on the initial assessment. Consequently, in this chapter I focus on this aspect of our research-clinical practice integration. Readers interested in how we integrated the outcome ratings into our practice, and how this enabled us to examine such clinically important questions as the relative efficacy of switching or augmenting antidepressants in treatment refractory depressed patients, are referred to these publications (Posternak and Zimmerman 2001; Posternak, Zimmerman, and Solomon 2002).
Methods of the MIDAS Project
Patients who call our practice are offered the option of receiving a standard clinical evaluation or a more comprehensive diagnostic interview. Patients are told that the comprehensive diagnostic interview lasts half a day. Patients are asked to arrive at 8:00 a.m., are given two questionnaires to complete, and then are interviewed with the SCID, the SIDP-IV, and the FH-RDC. A trained diagnostic rater, currently a Ph.D. level clinical psychologist, conducts the interviews. The patient is usually scheduled to be seen by the treating psychiatrist in the afternoon of the same day. After completing the standardized interviews, while waiting to see the psychiatrist, patients are asked to complete additional self-report scales. Sometimes, the interview with the psychiatrist is scheduled for another day, and patients are asked to bring the completed questionnaires at this time. For all evaluations, the diagnostic rater meets with the psychiatrist to summarize the findings from the structured interviews before the psychiatrist meets with the patient.
When this methodology was first proposed my colleagues did not believe that this protocol could be successfully implemented in a clinical practice because it would interfere with the development of a therapeutic relationship with patients. Because the patient spent more time with the diagnostic rater than the treating psychiatrist it was suggested that it would be more difficult to develop a therapeutic alliance. To avoid jeopardizing the clinical practices of my colleagues, the first patients enrolled in the MIDAS project were limited to those who were scheduled to see me. When seeing patients for the intake evaluation I began my clinical interview by summarizing the results of the research interviews, and then clarified areas of uncertainty and elaborated on topics that had been insufficiently covered. Even though I was provided with the information from the SCID and SIDP-IV, my evaluation typically lasted for an hour. It was my impression this method posed no threat to the development of a therapeutic alliance and in fact might have enhanced it because it enabled me to collect developmental and psychosocial history information that I otherwise would not have had time to ascertain because of the time needed to evaluate current psychiatric symptoms. A comparison of dropout rates between my patients and patients seen in the rest of the practice revealed a higher retention rate in my patients thereby further suggesting that the SCID/SIDP interview did not compromise the therapeutic alliance. As a result, the comprehensive diagnostic interviews were offered to patients being treated by all psychiatrists in the practice.
Through the years the percentage of patients presenting for treatment in our practice who were administered the SCID/SIDP varied according to the availability of diagnostic interviewers and patients' interest in receiving a more comprehensive interview. More recently, we have changed the policy in our practice and now require almost all patients to receive the comprehensive evaluation. Since implementing this change fewer than 10% of patients calling for an appointment decide to seek treatment elsewhere because they are unwilling to undergo such a lengthy evaluation. An important contributor to the success of the MIDAS project has been the support staff who schedule the initial evaluations and who have been educated about the benefits of comprehensive diagnostic evaluations so this can be explained to patients calling to make an initial appointment.
An Overview of Results from the MIDAS Project
Community-based epidemiological studies of psychiatric disorders provide important information about the public health burden of these problems. While the frequency of treatment seeking may be increasing (Olfson, Marcus, Druss et al. 2002), epidemiological studies indicate that most patients in the community do not get treatment for psychiatric disorders. Seeking treatment is related to a number of clinical and demographic factors. Consequently, studies of the frequency and correlates of psychiatric disorders in the general population should be replicated in clinical populations to provide the practicing clinician with information that might have greater clinical utility. The gap between general population and clinical epidemiological research might be greatest when examining disorder prevalence and diagnostic comorbidity. Obviously, one cannot extrapolate from community-based prevalence rates to clinical settings where the disorder rates are higher. Comorbidity rates are also expected to be higher in clinical settings because help-seeking may be related to comorbidity (Berkson 1946).
Most clinical epidemiological surveys have been based on unstructured clinical evaluations (Koenigsberg, Kaplan, Gilmore et al. 1985; Mezzich, Fabrega, Coffman et al. 1989; Oldham and Skodol 1991). However, several recent studies have questioned the accuracy and thoroughness of clinical diagnostic interviews. Shear and colleagues (Shear et al. 2000) studied diagnostic accuracy in two community mental health centers, one in urban Pittsburgh and the other in rural western Pennsylvania. They questioned whether clinicians apply DSM-IV diagnostic criteria in a rigorous manner, and suggested that clinical diagnoses may not be very accurate. Shear et al. interviewed 164 psychiatric outpatients with the SCID after they were evaluated clinically. More diagnoses were made on the SCID. Shear and colleagues found that more than one-third of patients were diagnosed with adjustment disorder by the clinicians versus only 7% by the SCID interviewers. Only 13% of the patients diagnosed by clinicians were given an anxiety disorder diagnosis whereas more than half (53%) of the patients interviewed with the SCID were diagnosed with a current anxiety disorder. The authors also found that half of the patients with a current primary diagnosis of major depressive disorder (MDD) on the SCID were diagnosed with adjustment disorder by clinicians. Shear and colleagues concluded that clinicians' diagnoses are often inaccurate, and that this poses a barrier to the implementation of treatments that have proven effective for specific disorders.
In another study of community mental health patients, this one conducted in Texas, (Basco et al. 2000) psychiatric nurses administered the SCID to patients as a test of the utility of research diagnostic procedures in clinical practice. They found that supplementing information from the patients' charts with the information from the SCID resulted in more than five times as many comorbid conditions being diagnosed. A gold standard diagnosis, consisting of the SCID diagnosis supplemented by chart information and then confirmed by a research psychiatrist or psychologist after interviewing the patient, was made for all patients, and the level of agreement with this standard was higher for the nurse-administered SCID than the clinical diagnoses.
Miller and colleagues (Miller, Dasher, Collins et al. 2001) compared diagnoses of 56 psychiatric inpatients evaluated with the traditional diagnostic assessment, SCID, and a computer assisted diagnostic evaluation. Consistent with the other studies they found that diagnoses were missed by the unstructured clinical diagnostic evaluation compared to the computer-assisted interview. In addition, diagnoses based on the SCID and computer assisted interview were significantly more highly associated with an all sources of information consensus diagnosis than was the unstructured clinical interview.
These studies, together with the findings from the MIDAS project (described below) suggest that clinical epidemiological studies should be based on structured research evaluations. To obtain accurate disorder prevalence rates in clinical settings it may also be important to assess a broad range of pathology in contrast to a single disorder. Melartin and colleagues (Melartin, Rytsala, Leskela et al. 2002) suggested that studies that focus on a single disorder find higher rates of that disorder compared to studies that assess several different disorders. It is possible that researchers who have expertise in the study of a particular disorder may be inclined to more frequently diagnose that disorder.
The MIDAS project is one of the first clinical epidemiological studies using structured interviews assessing a wide range of psychiatric disorders conducted in general clinical practice. Among the strengths of the study are that diagnoses are based on the reliable and valid procedures used in research studies, and the patients are presenting to a community-based psychiatric outpatient practice rather than a research clinic specializing in the treatment of one or a few disorders. A limitation of the study is that it is based in a single site.
The characteristics and correlates of several DSM-IV disorders hypothesized to be underdetected by clinicians have been described in case series. One of the first reports from the MIDAS project focused on one of these disordersbody dysmorphic disorder (BDD), Zimmerman and Mattia 1998). BDD is a distressing and impairing preoccupation with an imagined or slight defect in appearance. In a large case series of patients with BDD, Phillips and colleagues (Phillips, McElroy, Keck et al. 1994) reported that the disorder was associated with significant impairment in academic, occupational, and social functioning. BDD was also associated with a risk of suicidal behavior (29% of patients had attempted suicide). Despite its associated suicidal risk and psychosocial impairment, many individuals are so humiliated or ashamed of their BDD symptoms that they keep their concerns secret even from clinicians who have been treating them for years. The underdiagnosis of BDD has been consistently described in case series and research reports (Phillips 1991; Phillips, McElroy, Keck et al. 1993; Phillips et al. 1994). There are some studies of the prevalence of BDD in psychiatric patients; however, these studies were limited to patients with selected axis I disorders. The MIDAS project was the first to assess the presence of BDD in an unselected sample of patients presenting for treatment in an outpatient psychiatric setting.
In a sample of 500 patients interviewed with the SCID, 16 (3.2%) patients were diagnosed with BDD. BDD was the principal diagnosis for three (0.6%) patients and an additional diagnosis for 13 (2.6%) patients. In a separate sample of 500 patients The 500 patients presented to the same practice but not concurrently. The non-SCID sample was ascertained prior to the SCID sample. They were evaluated with a standard, unstructured clinical interview the prevalence of BDD was 0%. Patients with BDD received more axis I diagnoses than the patients without BDD. Looking at the specific axis I diagnoses, BDD patients were significantly more likely to have current diagnoses of social phobia and obsessive-compulsive disorder. The most frequent diagnosis in the BDD patients was MDD; however, BDD patients were no more likely to have MDD than were patients without BDD. The BDD patients were rated significantly lower on the GAF, indicating that their overall level of functioning was poorer. Across all patients the BDD patients were more severely depressed, despite a lack of difference in prevalence rate of MDD. The patients with BDD were not significantly more likely to have a lifetime history of suicide attempts or psychiatric hospitalization.
This study illustrated some of the power of the MIDAS project. We were able to examine the prevalence of a disorder that is rarely diagnosed in clinical practice and empirically demonstrate that BDD is, in fact, underdiagnosed by clinicians. We were also able to examine the strength of association between BDD and other psychiatric disorders, and the frequency of BDD in patients with these other disorders. This information is useful to clinicians who, because they have limited time to conduct diagnostic evaluations, can target higher risk individuals for inquiry about BDD. Finally, we were able to establish that patients with BDD, compared to patients without BDD, are more severely ill and functionally impaired, but not at greater risk for suicidal behavior.
We examined the clinical epidemiology of all Axis I disorders in a separate publication. Disorder frequency was examined in the first 400 patients interviewed with the SCID (Zimmerman and Mattia 2000). For patients with more than one disorder, the diagnoses were assigned as principal or additional according to the DSM-IV convention of whether it was the patient's stated primary reason for presenting for treatment. The most frequent diagnosis was MDD, which was present in nearly half of the patients. MDD was also the most common principal diagnosis, with more than three-quarters of the depressed patients having this as their principal diagnosis. The second most common diagnosis was social phobia. In contrast to MDD which, when present, was usually the principal diagnosis, very few patients with social phobia had it as their principal diagnosis. The other diagnoses that were present in at least 10% of the sample were posttraumatic stress disorder (PTSD), panic disorder with agoraphobia, specific phobia, and anxiety disorder not otherwise specified (NOS). Most disorders were more frequently diagnosed as additional disorders rather than the principal disorder. Only the mood and adjustment disorders were more frequently diagnosed as the principal disorder rather than as an additional disorder.
Underrecognition of Psychiatric Comorbidity
The recognition of comorbidity has important clinical significance. Comorbidity predicts poorer outcome for patients with depressive and anxiety disorders, and the presence of multiple psychiatric disorders is associated with greater levels of psychosocial impairment (Grunhaus 1988; Keller, Klerman, Lavori et al. 1984; Noyes, Reich, Christiansen et al. 1990). In routine clinical settings, an unstructured interview is typically used to assess patients. Unstructured interviews, however, may result in missed diagnoses, with potential negative clinical consequences.
In an early report from the MIDAS project, the goal was to examine whether diagnostic comorbidity is less frequently identified during a routine clinical evaluation than a semistructured diagnostic interview (Zimmerman and Mattia 1999d). Axis I diagnoses derived from structured and unstructured clinical interviews were compared in two groups of psychiatric outpatients seen in our practice. Five hundred patients underwent a routine unstructured clinical interview. Subsequent to the ascertainment of the first sample a second sample of 500 was collected though the individuals in this sample were interviewed with the SCID. The two groups had similar demographic characteristics and scored similarly on symptom questionnaires.
More current diagnoses were made in the SCID sample than the clinical sample (2.3+1.4 vs. 1.4+0.8, t=11.6, p
Principal Investigator - "Measurement and Validity of DSM Personality Disorders."
National Institute of Mental Health (NIMH) (MH-48732), 9/30/92 - 8/31/97, $349,966.
Principal Investigator - "Development of the Diagnostic Inventory for Depression."
National Computer Systems Assessments, 12/1/95-10/14/96, $128,682.
Co-Principal Investigator - "A Scale to Improve Mental Health Diagnostic Assessment."
NIMH (MH-56404) 9/1/97-2/28/98, $100,000.
Principal Investigator - "Development of the Psychiatric Diagnostic Screening Questionnaire."
Western Psychological Services, 7/1/97-6/30/98, $100,000.
Co-Principal Investigator "A Comprehensive Outcome Scale for Outpatient Psychiatry."
MIMH (MH-57603-01A1) 7/1/98-12/30/98, $100,000.
Co-Principal Investigator "A Scale to Improve Mental Health Diagnostic Assessments 2."
NIMH (MH-56404-01A1) 5/1/99-4/30/00, $781,508.
Principal Investigator "An Open Label Study of the Effectiveness of Citalopram in the Treatment of Pathological Gambling."
Forest Laboratories, 9/1/99-8/31/00, $16,350.
Principal Investigator "A Double-Blind, Placebo-Controlled Study of the Safety and Efficacy of Depakote in Impulsive Aggression across Psychiatric Disorders."
Abbott Laboratories, 11/1/99-12/30/00, $110,772.
Principal Investigator "An 8-Week, Randomized, Multicenter, Flexible-Dose, Double-Blind, Placebo-Controlled Study of Pagaclone in Patients with Panic Disorder."
Parke-Davis, 11/1/99-12/30/00, $110,772.
Principal Investigator "A 6-Week, Randomized, Multicenter, Flexible-Dose, Double-Blind, Placebo-Controlled Study of Pagaclone in Patients with Generalized Anxiety Disorder."
Parke-Davis, 9/1/00-6/1/01, $122,500.
Co-Principal Investigator "A Comprehensive Outcome Scale for Outpatient Psychiatry."
NIMH (MH-57603-02A2) 10/1/02-9/30/04, $999,162.
Co-Principal Investigator - "Rhode Island Gambling Outcomes Rating Scale (RIGORS)."
9 R44 DA20166-02A1, 9/1/05 - 8/31/07: $812,983.
The goal of this project is to develop a theoretically derived, broad-based, self-report questionnaire that evaluates the course and outcome of treatment for pathological gambling (PG). The RIGORS will combine simplicity of use and clinical utility with incremental predictive validity over other existing methods of assessing PG severity.