The Diagnostic Inventory for Depression: A self-report scale to diagnose DSM-IV major depressive disorder

The Diagnostic Inventory for Depression: A Self-ReportScale to Diagnose DSM-IV Major Depressive Disorder

�

Mark Zimmerman, Thomas Sheeran, and Diane Young

Rhode Island Hospital and Brown University

In this report from the Rhode Island Methods to Improve Diagnostic Assess-ment and Services (MIDAS) project, we describe the development andvalidation of the Diagnostic Inventory for Depression (DID), a new self-report scale designed to assess the Diagnostic and Statistical Manual ofMental Disorders, fourth edition (DSM-IV; American Psychiatric Associa-tion, 1994) symptom inclusion criteria for a major depressive episode,assess psychosocial impairment due to depression, and evaluate subjec-tive quality of life. A large sample of 626 psychiatric outpatients com-pleted the DID and were interviewed with the Structured Clinical Interviewfor DSM-IV (SCID). The measure’s test-retest reliability, discriminant andconvergent validity, and sensitivity to clinical change were investigated.The DID subscales achieved high levels of internal consistency and test-retest reliability. The DID was more highly correlated with another self-report measure of depression than with measures of anxiety, substanceuse problems, eating disorders, and somatization, thereby supporting theconvergent and discriminant validity of the scale. The DID also was highlycorrelated with interviewer ratings of the severity of depression and psy-chosocial functioning, and DID symptom severity scores were significantlydifferent in depressed patients with mild, moderate, and severe levels ofdepression. The DID was a valid measure of symptom change. Finally,the DID was significantly associated with a diagnosis of major depressivedisorder. © 2003 Wiley Periodicals, Inc. J Clin Psychol 60: 87–110, 2004.

Keywords: assessment; depression; diagnosis; questionnaire; majordepressive disorder

More than 15 years ago, Zimmerman, Coryell, Corenthal, and Wilson (1986) developedthe first self-administered questionnaire that was designed to “diagnose” a psychiatricdisorder defined by the specific inclusion and exclusion diagnostic criteria of the Diagnosticand Statistical Manual of Mental Disorders, third edition (DSM-III; American Psychiatric

Correspondence concerning this article should be addressed to: Mark Zimmerman, Bayside Medical Building,235 Plain Street, Providence, RI 02905; e-mail: [email protected].

JOURNAL OF CLINICAL PSYCHOLOGY, Vol. 60(1), 87–110 (2004) © 2004 Wiley Periodicals, Inc.

Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jclp.10207

Association, 1987). The Inventory to Diagnose Depression (IDD) assessed the DSM-IIIinclusion criteria for major depressive disorder (MDD), and the results of the initial vali-dation study demonstrated that a self-report scale could be successful in making a provi-sional psychiatric diagnosis. Subsequent to the initial study of psychiatric inpatients, otherstudies found that the IDD was a reliable and valid diagnostic instrument when used withpsychiatric outpatients (Krause, Philipp, Maier, & Schlegel, 1989), patients attending a pri-mary care practice (Froom & Hermoni, 1993; Jaffe, Froom, & Galambos, 1994), patientswith specific medical illnesses (Doerfler, Pbert, & DeCosimo, 1997; Frank et al., 1992),and nonpatient samples (Goldston, O’Hara, & Schartz, 1990; Zimmerman & Coryell, 1987).

Three developments during the past 15 years led us to develop a new self-reportdepression scale. First, the DSM-III diagnostic criteria for MDD have been slightly mod-ified in the subsequent DSM editions. In the DSM-III, the item referring to dysphoricmood included depression and sadness as well as hopelessness and irritability. In theDSM-IV, hopelessness and irritability are not included as dysphoric mood equivalents.Second, the DSM-IV explicitly requires the presence of psychosocial impairment to diag-nose MDD. DSM-III criteria for MDD did not explicitly include an impairment criterion.Though one could have inferred the necessity of impairment based on the definition ofmental illness given in the introduction of the DSM-III manual, this was generally notdone. For example, research diagnostic interviews used in epidemiological studies didnot require impairment to make a diagnosis of depression (Robins, Helzer, Croughan, &Ratcliff, 1981). Narrow, Rae, Robins, and Regier (2002) recently demonstrated that epi-demiological prevalence rates of major depression drop substantially when an impair-ment criterion is added to the identification of a case. And third, during the past decadethe outcomes literature increasingly has emphasized the importance of including assess-ments of quality of life and psychosocial functioning in outcome measurement.

In the present article from the Rhode Island Methods to Improve Diagnostic Assess-ment and Services (MIDAS) project, we describe the development and validation of theDiagnostic Inventory for Depression (DID). The DID is a new self-report scale designedto assess the DSM-IV symptom inclusion criteria for a major depressive episode, psy-chosocial impairment due to depression, and subjective quality of life.

Methods

Participants

Six hundred twenty-six psychiatric outpatients completed the 38-item DID before theirintake appointment. The group included 229 (36.6%) men and 397 (63.4%) women rang-ing in age from 18 to 78 years (M � 37.7, SD �11.9). More than two fifths of theparticipants were married (42.5%, n � 266); the remainder were single (29.6%, n � 185),divorced (15.7%, n � 98), separated (5.8%, n � 36), widowed (1.3%, n � 8), or livingwith someone as if in a marital relationship (5.3%, n � 33). Ten percent (n � 63) of theparticipants did not graduate high school, 63.2% (n � 396) graduated high school orachieved equivalency, and 26.7% (n � 167) graduated college. The racial composition ofthe sample was 89.1% (n � 558) White, 3.2% (n � 20) Black, 1.9% (n � 12) Hispanic,0.8% (n � 5) Asian, and 5.0% (n � 31) from another or a combination of the previousracial backgrounds.

Procedures

Individuals presenting for an intake evaluation at the Rhode Island Hospital Departmentof Psychiatry outpatient practice were invited to participate in a clinical research assessment

88 Journal of Clinical Psychology, January 2004

project that included a semistructured diagnostic interview and completion of self-administered questionnaires. Not all patients participated in the project because of thelack of availability of diagnostic interviewers and patients’ preference for a briefer clin-ical evaluation. Patients who did and did not participate in the research were similar indemographic characteristics and scores on a self-report screening questionnaire (Zim-merman & Mattia, 1999c).

All patients were interviewed by a trained diagnostic rater who administered theStructured Clinical Interview for DSM-IV (SCID; First, Spitzer, Gibbon, & Williams,1995) supplemented with questions from the Schedule for Affective Disorders and Schizo-phrenia (SADS; Endicott, Andreasen, & Spitzer, 1978) assessing the severity of symp-toms during the week prior to the evaluation. An extracted HRSD score was derived fromthe SADS ratings following the algorithm developed by Endicott, Cohen, Nee, Fleiss,and Sarantakos (1981). Patients also were rated on the Clinical Global Index (CGI) ofdepression severity and the Global Assessment of Functioning (GAF) scale. Details regard-ing interviewer training and diagnostic reliability are available in other publications fromthe MIDAS project (Zimmerman & Mattia, 1998, 1999a, 1999b). In brief, the diagnosticraters included four research assistants with college degrees in the social or biologicalsciences, 12 Ph.D.-level psychologists, and two psychiatrists. One of the four researchassistants had more than six years of experience administering the SCID and had previ-ously trained other research assistants in its use. All diagnostic raters received threemonths of training during which they observed at least 15 interviews (20 for the researchassistants), and they were observed and supervised in their administration of more than15 evaluations (20 for the research assistants). At the end of the training period, the raterswere required to demonstrate exact, or near-exact, agreement with a senior diagnosticianon five consecutive evaluations. During the course of the study, joint-interview diagnos-tic reliability information was collected on 47 patients. For MDD, the Kappa coefficientof agreement was 0.91.

Participants completed the DID as part of their initial paperwork. When scheduling theirappointments, the patients were told to arrive early to complete some standard forms. TheDID takes approximately 15 to 20 min to complete, and its administration did not disruptusual clinical practice. Because we were planning to test the DID’s validity by examiningits relationship with psychiatric diagnoses, the diagnosticians were kept blind to the pa-tients’ responses on the measure. The Rhode Island Hospital institutional review commit-tee approved the research protocol, and all patients provided informed, written consent.

The test-retest reliability of the DID was examined in a consecutive sample of 101 ofthe 626 participants. These patients were given the scale at the conclusion of the intakeevaluation and asked to mail it back in a preaddressed, postage-paid envelope. They weretold that the purpose of the second administration was to test the performance of thescale, not to question the truthfulness or accuracy of their responses. Twelve patients didnot indicate the date they completed the second administration of the scale. The majorityof the remaining patients completed the second administration later the same day (76.4%,n � 68). All patients completed the second administration within two weeks of the first.A problem with examining the test-retest reliability of a state measure in psychiatricpatients who present for treatment is that the patient’s state often changes quickly. Studiesof the placebo response in psychopharmacology studies has shown that it is often early(Quitkin, Rabkin, Ross, & Stewart, 1984). Clinical experience indicates that patientsoften feel much less distressed (and depressed) after the initial evaluation. Consequently,to study test-retest reliability over a longer interval in patients presenting for treatmentwould be inappropriate because symptom improvement is to be expected in many patientsover a short interval. An alternative would be to study test-retest reliability in patients

Diagnostic Inventory for Depression 89

who are already in treatment; however, this would not be appropriate for a test of thediagnostic performance of the scale.

To examine the convergent and discriminant validity of the DID, participants wereasked to complete a booklet of questionnaires at home that included measures ofsymptoms related to depression and nondepressive disorders (Measures are describedlater.) Five hundred patients completed the questionnaires. There was no difference indemographic characteristics between patients who did and did not return the booklet ofquestionnaires.

Measures

The DID contains 38 items. Nineteen questions assess symptom severity, and three ques-tions assess symptom frequency. The three symptom-frequency questions assess the fre-quency of the necessary symptoms to diagnose DSM-IV MDD (depressed mood, loss ofinterest in usual activities, or loss of pleasure in usual activities). The 19 symptom-severity questions assess the DSM-IV inclusion criteria for MDD. Compound DSM-IVcriteria referring to more than one construct (e.g., problems concentrating or makingdecisions; insomnia or hypersomnia) were subdivided into their respective components,and a DID item was written for each component. The individual symptoms assessed bythe DID are depressed mood, loss of interest in usual activities, loss of pleasure in usualactivities, low energy, psychomotor agitation, psychomotor retardation, guilt, worthless-ness, thoughts of death, suicidal ideation, impaired concentration, indecisiveness, decreasedappetite, weight loss, increased appetite, weight gain, insomnia, hypersomnia, and hope-lessness. The senior author (M.Z.) wrote the items and circulated a draft of the measureto colleagues for comment. Items were revised based on the feedback, the revision wasrecirculated, and the final version of scale drafted.

The six-item psychosocial functioning subscale assesses the amount of difficultysymptoms of depression have caused in usual daily responsibilities, relationships withsignificant others such as spouse, relationships with close family members, relationshipswith friends, participation in leisure activities, and overall level of function. Items arerated on a Likert scale of 0 (no difficulty) to 4 (extreme difficulty). An additional itemassessed the number of days in the past week the respondent was completely unable toperform their usual daily responsibilities. The quality of life subscale assesses satisfac-tion with the same areas covered by the psychosocial functioning scale as well as globalsatisfaction with mental health and physical health. Items are rated on a Likert scale of 0(very satisfied ) to 4 (very dissatisfied ).

The instructions indicate that the respondent is to pick the item that best describeshow he/she has been feeling during the past week. The DSM-IV criteria require symp-toms to last at least two weeks; nonetheless, we chose a one-week time frame to enableuse of the scale as a repeated measure over a one-week time interval. The time frame forthe symptom-frequency items is the past two weeks, thereby corresponding to the DSM-IVtime frame to diagnose MDD. The DID is reprinted in Appendix A.

The DID differs from other self-administered depression inventories in four ways.First, the DID assesses each of DSM-IV symptom criteria with 19 groups of five state-ments. In each group, one symptom of depression is assessed, and the items are arrangedin order of increasing severity. Second, the items were written so that a cutoff value canbe used to evaluate the presence or absence of the symptom. Similar to other depressionmeasures, the DID can be used to quantify the severity of the symptoms of depression byadding up the item scores. However, unlike other scales, the items were designed for a


dichotomous decision regarding symptom presence or absence to be made. For all itemsexcept loss of interest or pleasure in usual activities, an item score of 0 represents nodisturbance, a score of 1 represents subclinical severity, and a score of 2 or more indicatesthat the symptom is present. For the loss of interest and pleasure in usual activities items,a score of 3 or more indicates that the symptom is present. Third, the DID assesses thefrequency of depressed mood, loss of interest in usual activities, and loss of pleasure inusual activities during the two weeks before the evaluation. Symptom-frequency assess-ments were limited to these items because previous research has suggested that addingthe symptom-frequency questions to the symptom-severity assessments did not improvethe overall diagnostic performance of self-report depression scales (Kuhner & Veiel,1993; Zimmerman et al., 1986). We retained the frequency questions, albeit for only twoDSM-IV criteria, to reexamine the impact of this assessment on the diagnostic perfor-mance of the measure. Finally, the fourth novel feature of the DID is that it assessespsychosocial impairment related to depression and subjectively perceived life satisfaction.

We diagnosed MDD on the DID in four ways: symptom presence alone, symptompresence and symptom frequency, symptom presence and impairment, and symptom pres-ence, frequency, and impairment. We followed the DSM-IV’s diagnostic algorithm forcriterion A of a major depressive episode. That is, the patient must score above the apriori-defined thresholds on the DID depressed mood, decreased interest, or decreasedpleasure items, and above the threshold on at least one item in five of the nine DSM-IVsymptom groupings. Following DSM-IV’s algorithm, a score above threshold on twoitems which are components of the same DSM-IV criterion (e.g., impaired concentrationand indecisiveness) only counts as one symptom. To meet the symptom-frequency require-ment, the patient needed to indicate that the low mood, loss of interest, or loss of pleasurewas present every day or nearly every day of the past two weeks. The impairment crite-rion was met if the patient reported at least a moderate level of impairment on the globalrating of functioning. Directions for scoring the DID according to the DSM-IV diagnosticalgorithm for MDD are provided in Appendix B.

The booklet of questionnaires completed at home to establish the discriminant andconvergent validity of the DID included measures commonly used to evaluate symptomsof depression and nondepressive disorders. The nondepressive anxiety, substance use,eating, and somatoform disorders assessed are the most frequently diagnosed in clinicaland epidemiological studies (Kessler et al., 1994; Koenigsberg, Kaplan, Gilmore, &Cooper, 1985; Mezzich, Fabrega, Coffman, & Haley, 1989; Robins, Locke, & Regier,1991; Zimmerman & Mattia, 1999b). Moreover, these disorders are often comorbid withdepression; thus, it is important to demonstrate that a measure of depression can bediscriminated from measures of these constructs.

Eating Disorder Inventory Anorexia and Bulimia Subscales (EDI; Garner, Olmstead,& Polivy, 1983). The anorexia and bulimia subscales of the EDI consist of 23 self-report,6-point Likert-type items. The mean item-total correlation was .63. Criterion validity wasexamined with correlations between eating-disorder patients and clinician’s ratings(anorexia subscale � .53; bulimia subscale � .57). In a sample of eating-disorder patientsand noneating-disorder patients, the subscales correctly classified 89% of eating-disorderpatients.

Beck Depression Inventory (BDI; Beck, Ward, Mendelson, Mock, & Erbaugh,1961). The BDI is a 21-item self-report inventory that measures depressed mood andvegetative symptoms of depression. It is the most widely used self-administered measureof depression with well-established reliability and validity.


Brief Fear of Negative Evaluation (Brief-FNE; Leary, 1983). The Brief-FNE is a12-item, 5-point Likert-type, self-report measure that is widely used to measure anxietyat the prospect of being evaluated negatively—a core feature of social anxiety. The Brief-FNE correlated highly with the original 30-item scale (r � .96, p � .0001). Internalconsistency correlation (Cronbach’s �) was .90, and four-week test-retest reliability was.75. The Brief-FNE showed convergent correlations with other measures of social anxiety.

Fear Questionnaire (FQ; Marks & Mathews, 1979), Social Phobia and AgoraphobiaSubscales. The FQ is a 24-item, 8-point Likert-type, self-report scale that yields foursubscale scores used to measure severity and outcome of phobic symptoms. Inter-itemand item-subscale correlations were .50 or greater. One-week test-retest reliability cor-relations for the social phobia and agoraphobia subscales were .82 and .89, respectively.During development, the scale was used in 171 phobic patients, and the ratings corre-sponded well with the patients’ clinical state, relatives’ reports, and other ratings of adjust-ment. The scale also was found to be sensitive to clinical improvement after treatment ina sample of 26 phobic patients treated with exposure.

Social Phobia and Anxiety Inventory, Agoraphobia Subscale (SPAI; Turner, Beidel,Dancu, & Stanley, 1989). The agoraphobia subscale of the SPAI is a 13-item self-reportsubscale for agoraphobia, based on DSM-III criteria, that was added to the SPAI to deter-mine whether reported social distress might be due to fears of having panic attacks orbeing trapped rather than to fear of negative evaluation. Cronbach’s alpha for the internalconsistency of the agoraphobia subscale was .85. The two-week test-retest reliabilitycoefficient was .86. The scale significantly differentiated socially anxious college stu-dents from nonsocially anxious college students as well as social phobic patients frompatients with panic disorder and obsessive-compulsive disorder.

The Agoraphobic Cognitions Questionnaire (ACQ; Chambless, Caputo, Bright, &Gallagher, 1984). The ACQ is a 14-item, 5-point Likert-type, self-report scale com-prised of thoughts concerning negative consequences of experiencing anxiety. Internalconsistency (Cronbach’s �) was .80; item-total correlations were .26 or greater, and test-retest reliability over a median interval of 31 days ranged from .49 to .87 with a medianof .74. The ACQ demonstrated statistically significant convergent correlations with otherscales as predicted: Body Sensations Questionnaire, BDI, and State–Trait Anxiety Inventory.

PTSD Symptom Scale-Self-Report (PSS-SR; Foa, Riggs, Dancu, & Rothbaum,1993). The PSS-SR is a 17-item, 4-point Likert-type scale used to diagnose PTSD accord-ing to DSM-III criteria and assess the severity of PTSD symptoms. Internal consistency(Cronbach’s �) was .91, and the average item-total correlation was .60. The one-monthtest-retest correlation for the total score was .74. The PSS-SR was positively and signif-icantly correlated with the Rape Aftermath Symptom Test (Kilpatrick, 1988), the Impactof Event Scale (Horowitz, 1979), the BDI, and the State–Trait Anxiety Inventory. Whencompared to diagnoses based on SCID interviews, the sensitivity of the PSS-SR was62%, and the specificity was 100%.

Maudsley Obsession-Compulsion Questionnaire (MOC; Hodgson & Rachman,1977). The MOC is a 30-item, dichotomous (true/false), self-report scale used to assessthe existence and extent of four obsessive-compulsive complaints. The four subscales(checking, cleaning, slowness, and doubting) had internal consistency correlations (Cron-bach’s �) of .7 to .8. Test-retest reliability correlation was .8. The MOC was validated


against retrospective clinician ratings, and significantly correlated with the Leyton Obses-sional Inventory (Cooper, 1970).

Penn State Worry Questionnaire (PSWQ; Meyer, Miller, Metzger, & Borkovec,1990). The PSWQ is a 16-item, 5-point Likert-type, self-report scale used to measure thetrait of worry. Internal consistency (Cronbach’s �) was .91, and the one-month test-retestcorrelation was .93. It significantly discriminated college samples who met all, some, ornone of the DSM-III-R criteria for generalized anxiety disorder (GAD), and who metcriteria for GAD versus posttraumatic stress disorder. The PSWQ significantly and pos-itively correlated with measures related to pervasive worry such as the Self-ConsciousnessScale (Fenigstein, Scheier, & Buss, 1975), Reactions to Tests (Sarason, 1984), Perfec-tionism Scale (Burns, 1980), and Time Urgency Scales (Landy, Thayer, & Colvin, 1988).

Beck Anxiety Inventory (BAI; Beck, Brown, Epstein, & Steer, 1988). The BAI is a21-item, 4-point Likert-type, self-report scale used to measure anxiety severity. It dem-onstrated high internal consistency (Cronbach’s � � .92) and item-total correlations rangedfrom .30 to.71 (median � .60). Test-retest reliability over one week was .75. The BAIdiscriminated anxious diagnostic groups from nonanxious diagnostic groups. The corre-lations of the BAI with the Hamilton Anxiety Rating Scale-Revised (Hamilton, 1959) andthe Hamilton Rating Scale for Depression-Revised (HAM-D; Hamilton, 1960) were .51and .25, respectively.

Anxiety Sensitivity Index (ASI; Peterson & Heilbronner, 1987). The ASI is a 16-item,5-point Likert-type, self-report scale developed to assess beliefs about the social andsomatic consequences of anxiety symptoms. Internal consistency was .88, and spit-halfreliability was .85. Test-retest reliability coefficients ranged from .71 to .75. The scalesignificantly differentiated self-reported anxious college students from a normal sample.Factor analysis indicated that the ASI measured sensitivity to fear, or reaction to anxiety,relatively independently of the amount of anxiety expressed at a given time.

Michigan Alcoholism Screening Test (MAST; Selzer, 1971). The MAST is a 25-item,dichotomous (yes/no), self-report measure that is widely used to detect problem drink-ing. The MAST was validated by obtaining independent evidence of alcohol problemsfrom medical, social, driver, and criminal records on 526 participants in five groups,including a control group.

Drug Abuse Screening Test (DAST; Skinner, 1982). The DAST is a 28-item, dichot-omous (yes/no), self-report scale that measures the extent and severity of problems relatedto drug misuse. Internal consistency was .92. Concurrent validity was examined by cor-relating the DAST with background variables, frequency of drug use, and indices ofpsychopathology.

Whitely Index (WI; Pilowsky, 1967). The WI is a 14-item, dichotomous (yes/no),self-report measure of hypochondriasis. The WI significantly discriminated between hypo-chondriacal patients, nonhypochondriacal patients, and normal controls. Validity wasexamined by factor analysis and by correlating spouses’ responses about the patients topatients’ responses on the WI (r � .59, p � .001).

Physical Symptom Scale (PSS; Othmer & DeSouza, 1985; Swartz et al., 1986). ThePSS is a 17-item, 5-point Likert-type, self-report screening measure for somatization


disorder. It combines the Seven-Symptom Screening Test for Somatization Disorder (SSST;Othmer & DeSouza, 1985) and a somatization screening index (Swartz et al., 1986) usedin the Diagnostic Interview Schedule (DIS; Robins et al., 1981). The presence of three ofthe seven SSST symptoms correctly identified 91% of patients with somatization disorder(sensitivity � 87%, specificity � 95%). With the somatization index for the DIS, thepresence of 5 of 11 symptoms correctly identified 98% of cases with somatization dis-order and 99% without the disorder.

Symptom Rating Test, Paranoia and Psychosis Subscales (SRT; Kellner, 1973). TheSRT is a state measure of psychological distress. The paranoia and psychosis subscalesare each 7-item, 4-point Likert-type, self-report subscales. The SRT was validated as ascreening measure and as a preliminary assessment instrument for symptom severity. Thescales were sensitive to clinical changes during treatment.

Self-Report Mania Inventory (SRMI; Shugar, Schertzer, Toner, & DiGasbarro,1992). The SRMI is a 49-item, dichotomous (true/false), self-report scale for measuringmania. Internal consistency was high (Cronbach’s � � .94), as was the scale’s test-retestreliability (.93). A cutoff score of 14 provided the greatest sensitivity (80%) and thegreatest negative predictive value (88%).

Quality of Life Enjoyment and Satisfaction Questionnaire (Q-LES-Q; Endicott, Nee,Harrison, & Blumenthal, 1993). The Q-LES-Q is a 93-item, 5-point Likert-type, self-report measure of the degree of enjoyment and life satisfaction experienced in a varietyof domains. Internal consistency coefficients of subscales ranged from .90 to .96. Validitywas examined by correlating the Q-LES-Q with the CGI, HRSD, and BDI (range ��.33—.73).

Twenty-six patients completed the DID a second time during the course of treatment.At the follow-up evaluation, the treating clinicians rated the participants on the GAF, theCGI, and on a 6-point scale of depression severity (1 � full remission with no symptomsof MDD; 6 � meets full DSM-IV criteria for MDD, and either prominent psychoticsymptoms or extreme impairment in functioning) used in other studies of the outcome ofthe treatment for depression (Posternak & Zimmerman, 2001).

Statistical Analysis

For each of the three DID subscales (symptom severity, psychosocial functioning, andquality of life), we examined two types of reliability—test-retest reliability and internalconsistency. Internal consistency was evaluated for each subscale with Cronbach’s �.

We examined the convergent and discriminant validity (Campbell & Fiske, 1959) ofthe symptom-severity scale by comparing the correlation between the DID and the BDIwith the correlation between the DID and measures of anxiety, substance use, eatingdisorders, and somatization. We predicted that the DID would be significantly correlatedwith measures of nondepressive symptom domains, but the correlation with another mea-sure of depression would be higher.

Finally, we examined the diagnostic properties of the DID—sensitivity, specificity,positive and negative predictive value, and kappa (Kessel & Zimmerman, 1993). Depend-ing on the DID’s purpose, algorithms might be selected to optimize the sensitivity orspecificity of the scale (Hsiao, Bartko, & Potter, 1989; Mossman & Somoza, 1989). Inthe present report, we describe the diagnostic performance of four DID algorithms basedon symptom presence only, symptom presence plus symptom frequency, symptom pres-ence plus psychosocial impairment, and symptom presence, frequency, and impairment.


Results

The data in Table 1 show the diagnostic characteristics of the sample. The most frequentDSM-IV diagnoses were MDD (52.1%), social phobia (26.5%), GAD (17.3%), and panicdisorder (17.0%).

Internal Consistency and Item–Scale Correlations

The three DID subscales demonstrated excellent levels of internal consistency: symp-toms (Cronbach’s � � .91), psychosocial functioning (� � .89), quality of life (� � .90).

We computed the correlation between each item and the total DID subscale score,done separately for the symptom, psychosocial functioning, and quality of life items.When computing the item–scale correlations, the total score did not include the valuefrom the item. All item–scale correlations were significant (Tables 2 and 3). The mean ofthe item–scale correlations was .55 for the DID symptom severity items, .71 for thepsychosocial functioning items, and .67 for the quality-of-life items.

Test-Retest Reliability

A total of 101 patients completed the DID two times. Reliability was examined for totalscale scores and individual items as well as DID diagnoses. The test-retest reliabilitycoefficients were above .75 for the total scores of all three subscales (symptoms r � .91,psychosocial functioning r � .82, and quality of life r � .78). The reliability coefficientsof dimensional scores and dichotomous distinctions for the individual items are presented

Table 1DSM–IV Axis I Diagnoses of 626 Psychiatric Outpatientsa

DSM–IV Diagnosis n %

Major Depression 326 52.1Bipolar I Depression 12 1.9Bipolar II Depression 17 2.7Dysthymic Disorder 46 7.3Generalized Anxiety Disorder 108 17.3Panic Disorder without Agoraphobia 21 3.4Panic Disorder with Agoraphobia 85 13.6Social Phobia 166 26.5Specific Phobia 78 12.5Obsessive-Compulsive Disorder 53 8.5Posttraumatic Stress Disorder 65 10.4Adjustment Disorder 37 5.9Schizophrenia 4 0.6Schizoaffective Disorder 4 0.6Bulimia Nervosa 8 1.3Binge Eating Disorder 16 2.6Alcohol Abuse/Dependence 49 7.8Drug Abuse/Dependence 23 3.7Somatization Disorder 4 0.6Undifferentiated Somatoform Disorder 19 3.0Hypochondriasis 11 1.8

Note. aIndividuals could be given more than one diagnosis.


in Tables 2 and 3. The mean of the dimensional and categorical item reliabilities were .79and .69, respectively, for the DID symptom severity items, .74 and .60, respectively, forthe psychosocial functioning items, and .71 and .58, respectively, for the quality of lifeitems. Reliability was good for the four approaches used to make DID diagnoses (symp-tom presence � � .60, symptom presence plus symptom duration � � .71, symptompresence plus impairment � � .62, and symptom presence plus symptom duration plusimpairment � � .77).

Association Between the DID and Clinical Ratings of Depressive Symptoms

The severity of depressive symptoms was rated in all patients, regardless of whether theywere diagnosed with a mood disorder. The SADS was used to make the clinical ratings ofsymptom severity, and the SCID was used to rate symptom presence or absence. Thecorrelations between the DID and corresponding SADS and SCID ratings were all sig-nificant (Table 4). The mean of the correlations with the SADS ratings was .64, and themean kappa coefficient for the symptoms was .50. To confirm the validity of the a prioriitem thresholds, we examined the agreement between the DID and the SCID item ratingsfor each of the DID scores. The mean kappa was .34 when a DID item threshold of 1 ormore was used, .50 for a cutoff score of 2 or more, .41 for a cutoff of 3, and .23 for acutoff of 4.

Table 5 compares the frequency of depressive symptoms in depressed and nonde-pressed patients according to both the SCID and the DID. According to both measures,

Table 2DID Symptom Severity Items: Means, Standard Deviations, Item-Scale Correlations,and Test-Retest Reliability

Test-Retest Reliabilitya

DID Item Mean (SD) ICC �Item-Total

Correlationb

Low Mood 1.95 (1.16) .88 .68 .78Decreased Interest in Usual Activities 1.84 (1.29) .72 .58 .75Decreased Pleasure in Usual Activities 1.91 (1.26) .69 .67 .77Decreased Energy 2.26 (1.37) .87 .76 .72Psychomotor Agitation 1.32 (1.14) .77 .68 .53Psychomotor Retardation 1.75 (1.42) .66 .67 .73Guilt 1.70 (1.29) .79 .70 .63Worthlessness 1.77 (1.22) .80 .79 .74Thoughts About Death 1.00 (1.23) .84 .83 .60Suicidal Ideation 0.50 (0.91) .78 .69 .50Decreased Concentration 2.03 (1.32) .78 .61 .76Indecisiveness 1.66 (1.34) .83 .70 .73Decreased Appetite 1.08 (1.36) .79 .70 .40Weight Loss 0.49 (0.87) .92 .79 .33Increased Appetite 0.62 (1.16) .73 .61 .15Weight Gain 0.35 (0.73) .83 .77 .11Insomnia 1.28 (1.23) .77 .64 .30Hypersomnia 0.69 (1.26) .80 .73 .23Hopelessness 1.79 (1.26) .77 .67 .72

aICC � intraclass correlation coefficient.bAll correlations are significant at p � .001, except weight gain which was significant at p � .005.


depressed patients significantly more often reported every symptom of depression, andthe means of the odds ratios based on the SCID (3.6) and DID (2.9) were comparable.

Discriminant and Convergent Validity of the DID Symptom Subscale

Five hundred patients completed a package of questionnaires at home an average of1.2 days (SD � 16.9; range � 0–35 days) after the intake evaluation. The data in Table 6show that the DID symptom subscale was more highly correlated with the BDI than withmeasures of the other symptom domains. Moreover, the DID was nearly as highly cor-related with the extracted HRSD and CGI, clinician ratings of the severity of depressivesymptoms, as with the self-rated BDI.

The total DID QOL score significantly correlated with the Q-LES-Q total score (r ��.72, p � .001). The total QOL score also was significantly correlated with the DIDsymptom severity score (r � .78, p � .001) as well as the HRSD (r � .60, p � .001) andCGI (r � .62, p � .001).

Ability of the DID to Discriminate Between Levels of Depression Severity

The ability of the DID to discriminate between different levels of depression severity wasexamined with an analysis of variance on the CGI severity rating. Because only seven ofthe 626 patients were rated at the highest level of severity (extreme depression), the twohighest rating levels were combined. The overall ANOVA was significant, F � 179.4,df (4,620), p � .001. The DID symptom scores increased with increasing global severity

Table 3DID Psychosocial Impairment and Quality of Life Items: Means, Standard Deviations,Item-Scale Correlations, and Test-Retest Reliability

Test-Retest Reliabilitya

DID Item Mean (SD) ICC �Item-Total

Correlationb

Psychosocial Impairment SubscaleWork Performance 1.99 (1.35) .88 .63 .72Marital Relationship 2.13 (1.43) .64 .66 .61Family Relationships 1.67 (1.36) .67 .47 .69Friendships 1.36 (1.32) .78 .68 .70Leisure 2.14 (1.42) .73 .68 .78Global Rating of Impairment 2.39 (1.23) .77 .58 .75

Quality of Life SubscaleWork 2.37 (1.23) .70 .59 .65Marital Relationship 2.29 (1.32) .67 .64 .54Family Relationship 1.90 (1.26) .68 .44 .65Friendships 1.66 (1.26) .71 .65 .66Leisure 2.37 (1.28) .75 .55 .73Mental Health 2.81 (1.20) .72 .50 .74Physical Health 2.33 (1.28) .50 .32 .56Global Rating of Life Satisfaction 2.60 (1.07) .82 .71 .79Global Rating of Quality of Life 2.35 (0.94) .82 .80 .72

aICC � intraclass correlation coefficient.bAll correlations are significant at p � .001.


Table 4Association Between DID Ratings and Symptom Severity Ratingson the SADS and Symptom Rating on the SCID

DID ItemPearson CorrelationWith SADS Rating

� WithSCID Ratinga

Low Mood .71 .52Decreased Interest in Usual Activities .60 .48Decreased Pleasure In Usual Activities .67 .51Decreased Energy .67 .53Psychomotor Agitation .51 .35Psychomotor Retardation .39 .21Guilt .60 .49Worthlessness .71 .53Thoughts About Death .66 .58Suicidal Ideation .67 .56Decreased Concentration .67 .53Indecisiveness .55 .44Decreased Appetite .77 .63Weight Loss .61 .54Increased Appetite .69 .60Weight Gain .64 .57Insomnia .64 .54Hypersomnia .64 .47Hopelessness .67 .44

Note. All correlations and kappa coefficients are significant at p � .001.aWe used a cutoff of 2 to dichotomize the DID items as present or absent, except for decreased interestor pleasure in usual activities in which a cutoff of 3 was used.

Table 5Symptom Frequencies According to the DID and SCID in Depressed and Nondepressed Patients

Symptom FrequencyBased on DID (%)

Symptom FrequencyBased on SCID (%)

SymptomDepressed

PatientsNondepressed

PatientsOdds Ratio

(CI)Depressed

PatientsNondepressed

PatientsOdds Ratio

(CI )

Depressed Mood 83.4 27.3 4.1 (3.2–5.1) 81.0 14.4 6.9 (5.0–9.4)Anhedonia 54.1 8.3 6.1 (4.0–9.2) 76.4 7.4 11.6 (7.5–18.0)Fatigue 87.2 40.9 3.1 (2.6–3.7) 84.3 35.0 3.3 (2.7– 4.0)Agitation 51.4 21.2 2.4 (1.8–3.1) 38.2 21.4 1.7 (1.3–2.1)Retardation 72.8 23.3 3.6 (2.9– 4.7) 23.3 5.4 3.3 (2.0–5.4)Guilt 66.6 30.4 2.5 (2.0–3.0) 53.1 19.1 2.7 (2.1–3.6)Worthlessness 72.3 25.4 3.3 (2.6– 4.2) 50.4 10.5 4.4 (3.1– 63)Thoughts of Death 40.9 11.1 3.2 (2.3– 4.6) 49.6 9.3 4.8 (3.3–7.1)Suicidal Ideation 24.5 5.1 3.7 (2.2– 6.2) 23.6 6.6 2.8 (1.8– 4.4)Decreased Concentration 84.7 40.4 2.9 (2.4–3.5) 77.2 33.5 2.9 (2.4–3.6)Indecisiveness 71.3 27.7 3.0 (2.4–3.7) 52.6 10.5 4.6 (3.2– 6.7)Decreased Appetite 43.8 15.2 2.6 (2.0–3.5) 42.3 11.7 3.2 (2.3– 4.5)Weight Loss 21.0 6.7 2.5 (1.6–3.8) 22.8 8.6 2.2 (1.5–3.2)Increased Appetite 18.2 11.3 1.4 (1.0–2.0) 17.6 9.7 1.6 (1.1–2.2)Weight Gain 11.0 7.2 1.3 (0.9–2.0) 14.6 7.8 1.6 (1.1–2.3)Insomnia 53.2 28.5 1.9 (1.5–2.4) 68.3 30.0 2.6 (2.1–3.2)Hypersomnia 22.4 11.7 1.7 (1.2–2.3) 15.7 7.4 1.8 (1.2–2.6)


ratings (nondepressed, 9.3 6 8.8; borderline, 16.0 6 8.1; mild, 23.1 6 9.5; moderate32.469.8; severe, 42.1611.3). Duncan follow-up tests found that the difference betweeneach adjacent level of severity (e.g., nondepressed vs. borderline depressed; mild vs.moderate) was significant. We repeated this analysis limiting the sample to the 321 patientsdiagnosed with MDD. This truncated the range of CGI ratings, and all but seven patientswere rated mild, moderate, or severe. Despite the reduced variability in CGI and DIDscores, the overall ANOVA was still significant, F � 46.3, df (2,318), p � .001, as wereall three follow-up comparisons (mild, 28.5 6 7.6; moderate 33.0 6 9.0; severe,42.26 11.3).

Association with Psychiatric Diagnosis

Patients with MDD scored higher than patients without MDD on the symptom severity(34.7 6 10.7 vs. 16.5 6 10.7, df � 623, t � 21.3, p � .001), psychosocial impairment(14.3 6 4.9 vs. 7.4 6 5.5, df � 615, t � 16.3, p � .001), and quality of life subscales(23.96 6.0 vs. 15.36 7.6, df � 616, t � 15.7, p � .001).

The diagnostic performance of the DID was evaluated by comparing it to the re-sults of the SCID interview. As described earlier, a DID case was identified in four

Table 6Discriminant and Convergent Validity of the DID Symptom Subscale

ScaleCorrelation

With DID (r)

Measures of DepressionBeck Depression Inventory .83Extracted Hamilton Rating Scale for Depression .73Clinical Global Index of Depression Severity .73

Measures of Nondepressive Symptom DomainsEating Disorder Inventory Bulimia Subscale .27Eating Disorder Inventory Anorexia Subscale .24Self-Report Mania Inventory .28Brief Fear of Negative Evaluation Scale .37Fear Questionnaire Social Phobia Subscale .42Anxiety Sensitivity Inventory .46Agoraphobia Cognitions Questionnaire .43Fear Questionnaire Agoraphobia Subscale .35Social Phobia and Agoraphobia Inventory–Agoraphobia Subscale .42Symptom Rating Test Paranoia Subscale .44Symptom Rating Test Psychosis Subscale .28Posttraumatic Stress Disorder Scale .40Maudsley Obsession–Compulsive Questionnaire .26Penn State Worry Scale .42Beck Anxiety Inventory .53Michigan Alcohol Screening Test .04Drug Abuse Screening Test .10Whitely Index .27Somatic Symptom Index .47

Note. Because of missing data, the sample sizes ranged from 470 to 500. All correlations are signif-icant at p � .001, except Michigan Alcohol Screening Test (nonsignificant) and Drug Abuse Screen-ing Test ( p � .05).


ways: symptom presence, symptom presence plus symptom duration, symptom presenceplus impairment, and symptom presence plus symptom duration plus impairment. Thedata in Table 7 show that, as expected, as the threshold to identify a case increases, thesensitivity of the DID decreases and specificity increases. The level of diagnostic agree-ment between the DID and SCID, however, was approximately the same with all diag-nostic algorithms.

Sensitivity to Change

Twenty-six depressed patients completed the DID a second time an average of four monthsafter presentation for treatment. At follow-up, scores on all three subscales indicatedsignificant improvement over baseline scores (symptom severity, 34.56 10.0 vs. 18.0615.5, paired t � 4.58, p � .001; psychosocial impairment, 15.76 4.5 vs. 9.26 8.1, pairedt � 3.60, p � .001; quality of life, 25.26 6.3 vs. 17.36 10.3, paired t � 5.58, p � .001).Likewise, scores on the GAF were significantly higher at follow-up (52.5 6 6.5 vs.63.76 12.7, paired t � 4.85, p � .001). All correlations between the change in GAF andDID scores from baseline to follow-up were significant (symptom severity, r � .81, p �.001; psychosocial impairment, r � .78, p � .001; quality of life, r � .72, p � .001). Onthe CGI, 16 patients were rated much or very much improved, 1 patient was a littleimproved, and 9 patients had not improved or were worse. All correlations between thefollow-up CGI and change in DID scores were significant (symptom severity, r � �.81,p � .001; psychosocial impairment, r � �.78, p � .001; quality of life, r � �.71, p �.001). Finally, the follow-up depression outcome rating was significantly associated withthe change from baseline to follow-up in all DID subscale scores (symptom severity, r ��.79, p � .001; psychosocial impairment, r � �.79, p � .001; quality of life, r � �.72,p � .001).

Discussion

The results of this initial validation study of the DID in an outpatient setting suggest thatit is a reliable and valid measure of the severity of depression and the psychosocialimpairment associated with depression. There is no shortage of self-report questionnairesthat assess depression, therefore it is reasonable to question the development of yet anotherscale. The DID is unique in that it is the only case-finding instrument that is directly tiedto the DSM-IV diagnostic criteria for major depression. All major depression inclusioncriteria are assessed and, most importantly, specified cutoffs are used to determine the

Table 7Diagnostic Performance of the DID

DID Case Definition

No. ofDID

CasesSensitivity

(%)Specificity

(%)

PositivePredictive

Value(%)

NegativePredictive

Value(%) �

Symptoms 360 90.1 77.0 80.6 88.0 .67Symptoms and Duration 311 80.1 82.6 83.0 79.7 .63Symptoms and Impairment 337 85.7 79.9 81.9 84.1 .66Symptoms and Duration and Impairment 294 76.7 84.5 84.0 77.4 .61


presence or absence of each criterion. Thus, the determination of whether someone is acase follows the DSM-IV algorithmic approach. This is in contrast to scales in whichtotal scale scores are computed, and caseness is based on whether the individual scoresabove or below a threshold score. A vexing problem with this latter approach is thatdifferent studies find that different thresholds are optimal for distinguishing cases fromnoncases. For example, studies of the BDI as a case-finding instrument have used cutoffpoints of 10 (Craven, Rodin, & Littlefield, 1988; Deardorff & Funabiki, 1985; Oliver &Simmons, 1984; Zich, Attkisson, & Greenfield, 1990), 11 (Gallagher, Breckenridge, Stein-metz, & Thompson, 1983; Harris, Huckle, Thomas, Hohns, & Fung, 1989), 12 (Lasa,Ayuso-Mateos, Vazquez-Barquero, Diez-Manrique, & Dowrick, 2000), 13 (Hesselbrock,Hesselbrock, Tennen, Meyer, & Workman, 1983; Turner & Romano, 1984), 15 (Bishop,Edgley, Fisher, & Sullivan, 1993), 16 (Holcomb, Stone, Lustman, Gavard, & Mostello,1996), 23 (Martinsen, Friis, & Hoffart, 1995), and even as high as 29 (Lykouras et al.,1998) to classify individuals as depressed. Admittedly, the BDI was not intended by itsdevelopers to be a diagnostic or screening instrument. However, even when a scale’sdevelopers recommend a particular threshold to identify cases based on a total scalescore, researchers will use varying cutoff scores. For example, the Center for Epidemio-logical Studies-Depression (CES-D; Radloff, 1977) scale was developed as a brief screen-ing tool with a recommended threshold score to identify depressed individuals. Otherinvestigators, however, have used different cutoff scores to identify depressed cases(Coulehan, Schulberg, Block, Janosky, & Arena, 1990). The DID’s standardized approachto determining whether someone is or is not a case is a clear advantage over existingmeasures. Admittedly, we defined a case on the DID in four ways; thus, there can be somevariability in using the DID to identify cases of depression. However, the purpose ofexamining different methods of identifying cases of depression was to examine how thesymptom-persistence and symptom-impairment factors impacted on the diagnostic per-formance of the scale. The cutoffs used on the DID to suggest clinical significance wereidentified a priori, and validated empirically. This is in contrast to studies that selectcutoff scores to maximize agreement between the instrument being studied and a diag-nostic standard.

The DID also distinguishes itself from other measures of depression because it assessessymptom persistence and duration as well as symptom severity. The diagnosis of majordepression depends on both symptom presence as well as symptom persistence, and noother self-report depression scale assesses both constructs. Some measures such as theZung Depression Scale (Zung, Magill, Moore, & George, 1983) and the CESD ask respon-dents to indicate the frequency of symptoms. Other measures such as the BDI and theInventory of Depressive Symptoms (Rush, Gullion, Basco, Jarrett, & Trivedi, 1996) askrespondents to choose which of a group of four statements representing different levels ofseverity best describes them. The DID assesses both symptom severity and persistence,though the persistence questions did not improve the overall diagnostic performance ofthe scale. This is consistent with prior research on the IDD suggesting that symptomduration questions do not improve the case identification performance of the measure.We nevertheless chose to retain the symptom-persistence questions on the measure becauseuse of these questions improves the specificity of the scale (at the cost of reduced sensi-tivity), and in some instances, it may be preferable to identify a more homogeneous groupof depressed individuals.

Finally, the DID is unique in that it is the only depression scale that assesses bothpsychosocial impairment due to depression and quality of life. The importance of theseconstructs has been increasingly recognized during the past decade; however, no self-report measure of depressive symptoms evaluates all of these domains.


The DID follows the DSM-IV diagnostic algorithm to identify individuals with aprovisional diagnosis of MDD. However, the DID cannot distinguish between MDD andnormal grief reactions, depression due to medication, physical illness, or substance abuse,or depression occurring during the course of a chronic psychotic disorder. The results ofthe DID should be interpreted by appropriately trained personnel, and should be verifiedby a clinical diagnostic evaluation.

The level of agreement between DID classification and the diagnosis of DSM-IVMDD according to the SCID was not perfect; however, perfect agreement should not beexpected. The level of agreement between the two instruments will be limited, in part, bythe respective test-retest reliabilities of each. The kappa coefficients between the DIDand SCID ranged from .61 to .67, within the range of test-retest reliability of DID case-ness and SCID diagnoses (Williams et al., 1992).

The present sample was drawn from a large, general adult-outpatient private-practicesetting in which the most common presenting problems were mood and anxiety dis-orders, and the patients were predominantly White and had private health insurance. Itwill be important to replicate and extend the present findings to samples with differentdemographic and clinical characteristics. Another direction for future research is to exam-ine the performance of the DID in primary care practice as well as in inpatient and partialhospital settings.

Appendix A

INSTRUCTIONS: This questionnaire is about how you have been feeling during the past week. After eachquestion there are 5 statements (numbered 0– 4). Read all 5 statements carefully. Then decide which one bestdescribes how you have been feeling. Choose only one statement per group. If more than one statement in agroup applies to you, choose the one with the higher number.

(1) During the past week, have you been feeling sad or depressed?0 No, not at all.1 Yes, a little bit.2 Yes, I have felt sad or depressed most of the time.3 Yes, I have been very sad or depressed nearly all the time.4 Yes, I have been extremely depressed nearly all the time.

(2) How many days in the past 2 weeks have you been feeling sad or depressed?0 No days1 A few days2 About half the days3 Nearly every day4 Every day

(3) Which of the following best describes your level of interest in your usual activities during the pastweek?0 I have not lost interest in my usual activities.1 I have been less interested in 1 or 2 of my usual activities.2 I have been less interested in several of my usual activities.3 I have lost most of my interest in almost all of my usual activities.4 I have lost all interest in all of my usual activities.

(4) How many days in the past 2 weeks have you been less interested in your usual activities?0 No days1 A few days2 About half the days3 Nearly every day4 Every day


(5) Which of the following best describes the amount of pleasure you have gotten from your usual activitiesduring the past week?0 I have gotten as much pleasure as usual.1 I have gotten a little less pleasure from 1 or 2 of my usual activities.2 I have gotten less pleasure from several of my usual activities.3 I have gotten almost no pleasure from most of the activities that I usually enjoy.4 I have gotten no pleasure from any of the activities that I usually enjoy.

(6) How many days in the past 2 weeks have you gotten less pleasure from your usual activities?0 No days1 A few days2 About half the days3 Nearly every day4 Every day

(7) During the past week, has your energy level been low?0 No, not at all.1 Yes, my energy level has occasionally been a little lower than it normally is.2 Yes, I have clearly had less energy than I normally do.3 Yes, I have had much less energy than I normally have.4 Yes, I have felt exhausted almost all of the time.

(8) Which of the following best describes your level of physical restlessness during the past week?0 I have not been more restless and fidgety than usual.1 I have been a little more restless and fidgety than usual.2 I have been very fidgety, and it has been somewhat difficult to sit still.3 I have been extremely fidgety, and I have been pacing a little bit almost every day.4 I have been pacing more than an hour a day, and I have been unable to sit still.

(9) Which of the following best describes your physical activity level during the past week?0 I have not been moving more slowly than usual.1 I have been moving a little more slowly than usual.2 I have been moving more slowly than usual, and it takes me longer than usual to do most activities.3 Normal activities are difficult because it has been tough to start moving.4 I have been feeling extremely slowed down physically, like I am stuck in mud.

(10) During the past week, have you been bothered by feelings of guilt?0 No, not at all.1 Yes, I have occasionally felt a little guilty.2 Yes, I have often been bothered by feelings of guilt.3 Yes, I have often been bothered by strong feelings of guilt.4 Yes, I have been feeling extremely guilty.

(11) During the past week, what has your self esteem been like?0 My self-esteem has not been low.1 Once in a while, my opinion of myself has been a little low.2 I often think I am a failure.3 I almost always think I am a failure.4 I have been thinking I am a totally useless and worthless person.

(12) During the past week, have you been thinking about death or dying?0 No, not at all.1 Yes, I have occasionally thought that life is not worth living.2 Yes, I have frequently thought about dying in passive ways (such as going to sleep and not waking up).3 Yes, I have frequently thought about death, and that others would be better off if I were dead.4 Yes, I have been wishing I were dead.

(13) During the past week, have you been thinking about killing yourself?0 No, not at all.1 Yes, I had a fleeting thought about killing myself.2 Yes, several times I thought about killing myself, but I would not act on these thoughts.3 Yes, I have been seriously thinking about killing myself.4 Yes, I have thought of a specific plan for killing myself.


(14) Which of the following best describes your ability to concentrate during the past week?0 I have been able to concentrate as well as usual.1 My ability to concentrate has been slightly worse than usual.2 My attention span has not been as good as usual and I have had difficulty collecting my thoughts, but

this hasn’t caused any serious problems.3 I have frequently had trouble concentrating, and it has interfered with my usual activities.4 It has been so hard to concentrate that even simple things are hard to do.

(15) During the past week, have you had trouble making decisions?0 No, not at all.1 Yes, making decisions has been slightly more difficult than usual.2 Yes, it has been harder and has taken longer to make decisions, but I have been making them.3 Yes, I have been unable to make some decisions that I would usually have been able to make.4 Yes, important things are not getting done because I have had trouble making decisions.

(16) During the past week, has your appetite been decreased?0 No, not at all.1 Yes, my appetite has been slightly decreased compared to how it normally is.2 Yes, my appetite has been clearly decreased, but I have been eating about as much as I normally do.3 Yes, my appetite has been clearly decreased, and I have been eating less than I normally do.4 Yes, my appetite has been very bad, and I have had to force myself to eat even a little.

(17) How much weight have you lost during the past week (not due to dieting)?0 None (or the only weight I lost was due to dieting)1 1–2 pounds2 3–5 pounds3 6–10 pounds4 More than 10 pounds

(18) During the past week, has your appetite been increased?0 No, not at all.1 Yes, my appetite has been slightly increased compared to how it normally is.2 Yes, my appetite has clearly been increased compared to how it normally is.3 Yes, my appetite has been greatly increased compared to how it normally is.4 Yes, I have been feeling hungry all the time.

(19) How much weight have you gained during the past week?0 None1 1–2 pounds2 3–5 pounds3 6–10 pounds4 More than 10 pounds

(20) During the past week, have you been sleeping less than you normally do?0 No, not at all.1 Yes, I have occasionally had slight difficulty sleeping.2 Yes, I have clearly been sleeping less than I normally do.3 Yes, I have been sleeping about half my normal amount of time.4 Yes, I have been sleeping less than 2 hours a night.

(21) During the past week, have you been sleeping more than you normally do?0 No, not at all.1 Yes, I have occasionally slept more than I normally do.2 Yes, I have frequently slept at least 1 hour more than I normally do.3 Yes, I have frequently slept at least 2 hours more than I normally do.4 Yes, I have frequently slept at least 3 hours more than I normally do.

(22) During the past week, have you been feeling pessimistic or hopeless about the future?0 No, not at all.1 Yes, I have occasionally felt a little pessimistic about the future.2 Yes, I have often felt pessimistic about the future.3 Yes, I have been feeling very pessimistic about the future most of the time.4 Yes, I have been feeling that there is no hope for the future.


05 no difficulty 15mild difficulty 25moderate difficulty 35marked difficulty 45 extreme difficulty

INSTRUCTIONSIndicate below how much symptoms of depression have interfered with, or caused difficulties in, the followingareas of your life during the past week (Circle DNA [Does Not Apply] if you are not married or have aboyfriend/girlfriend.)

During the PAST WEEK, how much difficulty have symptoms of depression caused in your. . .

23. usual daily responsibilities (at a paid job, at home, or at school)........................................ 0 1 2 3 424. relationship with your husband, wife, boyfriend, girlfriend, or lover .....................DNA 0 1 2 3 425. relationships with close family members............................................................................. 0 1 2 3 426. relationships with your friends ............................................................................................ 0 1 2 3 427. participation and enjoyment in leisure and recreation activities ......................................... 0 1 2 3 4

28. Overall, how much have symptoms of depression interfered with or caused difficulties in your life?0) not at all1) a little bit2) a moderate amount3) quite a bit4) extremely

29. How many days during the past week were you completely unable to perform your usual daily responsibilities(at a paid job, at home, or at school) because you were feeling depressed? (circle one)

0 days 1 day 2 days 3 days 4 days 5 days 6 days 7 days

05very satisfied 15mostly satisfied 25 equally satisfied/dissatisfied 35mostly dissatisfied 45very dissatisfied

INSTRUCTIONSIndicate below your level of satisfaction with the following areas of your life (Circle DNA [Does Not Apply] ifyou are not married or have a boyfriend or girlfriend.)

During the PAST WEEK how satisfied have you been with your. . .

30. usual daily responsibilities (at a paid job, at home, or at school)........................................ 0 1 2 3 431. relationship with your husband, wife, boyfriend, girlfriend, or lover .....................DNA 0 1 2 3 432. relationship with close family members .............................................................................. 0 1 2 3 433. relationships with your friends ............................................................................................ 0 1 2 3 434. participation and enjoyment in leisure and recreation activities ......................................... 0 1 2 3 435. mental health........................................................................................................................ 0 1 2 3 436. physical health ..................................................................................................................... 0 1 2 3 4

37. In general, how satisfied have you been with your life during the past week?0) very satisfied1) mostly satisfied2) equally satisfied & dissatisfied3) mostly dissatisfied4) very dissatisfieded

38. In general, how would you rate your overall quality of life during the past week?0) very good, my life could hardly be better1) pretty good, most things are going well2) the good and bad parts are about equal3) pretty bad, most things are going poorly4) very bad, my life could hardly be worse

Copyright © Mark Zimmerman, M.D. All rights reserved. Not to be reproduced without the author’s permission.


Appendix B. Scoring the DID According to the DSM-IV DiagnosticAlgorithm for Major Depression

Criterion A—Symptoms

DSM-IV Criterion DID Item(s) Item Score Criterion Met

1. Low Mood 1 2, 3, or 4 yes/no2. Anhedonia 3 or 5 3 or 4 yes/no3. Appetite/Weight 16, 17, 18, or 19 2, 3, or 4 yes/no4. Sleep 20 or 21 2, 3, or 4 yes/no5. Psychomotor 8 or 9 2, 3, or 4 yes/no6. Fatigue 7 2, 3, or 4 yes/no7. Worthless/Guilt 10 or 11 2, 3, or 4 yes/no8. Concentration/Indecisive 14 or 15 2, 3, or 4 yes/no9. Thoughts of Death/Suicide 12 or 13 2, 3, or 4 yes/no

Criterion C—Impairment

DID Item(s) Item Score Criterion Met

23, 24, 25, 26, or 27 2, 3, or 4 yes/no

DID provisional diagnosis of major depression. All 3 are necessary.

1. Low Mood or Anhedonia is present Yes No2. Five Items from criteria A are present Yes No3. Criterion C is present Yes No

References

American Psychiatric Association. (1987). Diagnostic and statistical manual of mental disorders(3rd ed.). Washington, DC: Author.

American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders(4th ed.). Washington, DC: Author.

Beck, A.T., Brown, G., Epstein, N., & Steer, R. (1988). An inventory for measuring clinical anxi-ety: Psychometric properties. Journal of Consulting and Clinical Psychology, 56, 893–897.

Beck, A.T., Ward, C.H., Mendelson, M., Mock, J., & Erbaugh, J. (1961). An inventory for measur-ing depression. Archives of General Psychiatry, 4, 561–571.

Bishop, S.R., Edgley, K., Fisher, R., & Sullivan, M.J.L. (1993). Screening for depression in chroniclow back pain with the Beck Depression Inventory. Canadian Journal of Rehabilitation, 7,143–148.

Burns, D. (1980). The perfectionist’s script for self-defeat. Psychology Today, 4, 34–52.

Campbell, D.T., & Fiske, D.W. (1959). Convergent and discriminant validation by the multitraitmultimethod matrix. Psychological Bulletin, 56, 81–105.

Chambless, D.L., Caputo, G.C., Bright, P., & Gallagher, R. (1984). Assessment of fear in agora-phobics: The Body Sensations Questionnaire and the Agoraphobic Cognitions Questionnaire.Journal of Consulting and Clinical Psychology, 52, 1090–1097.


Cooper, J. (1970). The Leyton obsessional inventory. Psychological Medicine, 1, 48– 64.

Coulehan, J.L., Schulberg, H.C., Block, M.R., Janosky, J.E., & Arena, V.C. (1990). Depressivesymptomatology and medical comorbidity in a primary care clinic. International Journal ofPsychiatry in Medicine, 20, 335–347.

Craven, J.L., Rodin, G.M., & Littlefield, C. (1988). The Beck Depression Inventory as a screeningdevice for major depression in renal dialysis patients. International Journal of Psychiatry inMedicine, 18, 365–374.

Deardorff, W.W., & Funabiki, D. (1985). A diagnostic caution in screening for depressed collegestudents. Cognitive Therapy and Research, 9, 277–284.

Doerfler, L.A., Pbert, L., & DeCosimo, D. (1997). Self-reported depression in patients with coro-nary heart disease. Journal of Cardiopulmonary Rehabilitation, 17, 163–170.

Endicott, J., Andreasen, N., & Spitzer, R.L. (1978). Family history research diagnostic criteria (3rded.). New York: Biometrics Research, New York State Psychiatric Institute.

Endicott, J., Cohen, J., Nee, J., Fleiss, J., & Sarantakos, S. (1981). Hamilton depression rating scale.Archives of General Psychiatry, 38, 98–103.

Endicott, J., Nee, J., Harrison, W., & Blumenthal, R. (1993). Quality of life enjoyment and satis-faction questionnaire: A new measure. Psychopharmacology Bulletin, 29, 321–326.

Fenigstein, A., Scheier, M., & Buss, A. (1975). Public and private self-consciousness: Assessmentand theory/ Journal of Consulting and Clinical Psychology, 35, 837–844.

First, M.B., Spitzer, R.L., Gibbon, M., & Williams, J.B.W. (1995). Structured Clinical Interview forDSM-IV Axis I Disorders—Patient edition (SCID-I/P, Version 2.0). New York: BiometricsResearch Department, New York State Psychiatric Institute.

Foa, E.B., Riggs, D.S., Dancu, C.V., & Rothbaum, B.O. (1993). Reliability and validity of a briefinstrument for assessing posttraumatic stress disorder. Journal of Traumatic Stress, 6, 459–473.

Frank, R.G., Chaney, J.M., Clay, D.L., Shutty, M.S., Beck, N.C., Kay, D.R., Elliott, T.R., & Gram-bling, S. (1992). Dysphoria: A major symptom in persons with disability of chronic illness.Psychiatry Research, 43, 231–241.

Froom, J., & Hermoni, D. (1993). The Inventory to Diagnose Depression. Family Practice, 10,312–316.

Gallagher, D., Breckenridge, J., Steinmetz, J., & Thompson, L. (1983). The Beck Depression Inven-tory and Research Diagnostic Criteria: Congruence in an older population. Journal of Con-sulting and Clinical Psychology, 51, 945–946.

Garner, D.M., Olmstead, M.P., & Polivy, J. (1983). Development and validation of a multidimen-sional eating disorder inventory for anorexia nervosa and bulimia. International Journal ofEating Disorders, 2, 15–34.

Goldston, D.B., O’Hara, M.W., & Schartz, H.A. (1990). Reliability, validity, and preliminary nor-mative data for the Inventory to Diagnose Depression in a college population. Journal ofConsulting and Clinical Psychology, 2, 212–215.

Hamilton, M. (1959). The assessment of anxiety states by rating. British Journal of Medical Psy-chology, 23, 50–55.

Hamilton, M. (1960). A rating scale for depression. Journal of Neurology, Neurosurgery, & Psy-chiatry, 23, 56– 62.

Harris, B., Huckle, P., Thomas, R., Hohns, S., & Fung, H. (1989). The use of rating scales toidentify postnatal depression. British Journal of Psychiatry, 154, 813–817.

Hesselbrock, M.N., Hesselbrock, V.M., Tennen, H., Meyer, R.E., & Workman, K.L. (1983). Meth-odological considerations in the assessment of depression in alcoholics. Journal of Consultingand Clinical Psychology, 51, 399– 405.

Hogson, R.J., & Rachman, S. (1977). Obsessional compulsive complaints. Behaviour Research andTherapy, 10, 111–117.


Holcomb, W.L., Stone, L.S., Lustman, P.J., Gavard, J.A., & Mostello, D.J. (1996). Screening fordepression in pregnancy: Characteristics of the Beck Depression Inventory. Obstetrics andGynecology, 88, 1021–1025.

Horowitz, M. (1979). Impact of event scale: A measure of subjective distress. Psychomotor Med-icine, 41, 207–218.

Hsiao, J.K., Bartko, J.J., & Potter, W.Z. (1989). Diagnosing diagnoses: Receiver operating charac-teristic methods and psychiatry. Archives of General Psychiatry, 46, 664– 667.

Jaffe, A., Froom, J., & Galambos, N. (1994). Minor depression and functional impairment. Archivesof Family Medicine, 3, 1081–1086.

Kellner, R. (1973). The Symptom Rating Test. In N. Sartorius & T. Ban (Eds.), Assessment ofdepression (pp. 213–220). New York: Springer-Verlag.

Kessel, J.B., & Zimmerman, M. (1993). Reporting errors in studies of the diagnostic performanceof self-administered questionnaires: Extent of the problem, recommendations for standardizedpresentation of results, and implications for the peer review process. Psychological Assess-ment, 5, 395–399.

Kessler, R.C., McGonagle, K.A., Zhao, S., Nelson, C.B., Hughes, M., Eshelman, S., Wittchen,H.U., & Kendler, K.S. (1994). Lifetime and 12-month prevalence of DSM-III-R psychiatricdisorders in the United States. Archives of General Psychiatry, 51, 8–19.

Kilpatrick, D. (1988). Rape Aftermath Symptom Test. Oxford, England: Pergamon Press.

Koenigsberg, H.W., Kaplan, R.D., Gilmore, M.M., & Cooper, A.M. (1985). The relationship betweensyndrome and personality disorder in DSM-III: Experience with 2,462 patients. AmericanJournal of Psychiatry, 142, 207–212.

Krause, J., Philipp, M., Maier, W., & Schlegel, S. (1989). A German validation study of the Inven-tory to Diagnose Depression. Psychopathology, 22, 57– 64.

Kuhner, C., & Veiel, H.O. (1993). Psychometrische und diagnostiche Eigenschaften einer duet-schsprachigen Version des Inventory to Diagnose Depression (IDD). Diagnostica, 39, 229–321.

Landy, F., Thayer, J., & Colvin, C. (1988). The measurement of time urgency. Unpublished manuscript.

Lasa, L., Ayuso-Mateos, J.L., Vazquez-Barquero, J.L., Diez-Manrique, F.J., & Dowrick, C.F. (2000).The use of the Beck Depression Inventory to screen for depression in the general population:A preliminary analysis. Journal of Affective Disorders, 57, 261–265.

Leary, M.R. (1983). A brief version of the Fear of Negative Evaluation Scale. Personality andSocial Psychology Bulletin, 9, 371–375.

Lykouras, L., Oulis, P., Adrachta, D., Daskalopoulou, E., Kalfakis, N., Triantaphyllou, N., Papa-georgiou, K., & Christodoulou, G.N. (1998). Beck Depression Inventory in the detection ofdepression among neurological inpatients. Psychopathology, 31, 213–219.

Marks, I.M., & Mathews, A.M. (1979). Brief standard self-rating for phobic patients. BehaviourResearch and Therapy, 17, 263–267.

Martinsen, E.W., Friis, S., & Hoffart, A. (1995). Assessment of depression: Comparison betweenBeck Depression Inventory and subscales of Comprehensive Psychopathological Rating Scale.Acta Psychiatrica Scandinavica, 92, 460– 463.

Meyer, T.J., Miller, M.L., Metzger, R.L., & Borkovec, T.D. (1990). Development and validation ofthe Penn State Worry Questionnaire. Behaviour Research and Therapy, 28, 487– 495.

Mezzich, J.E., Fabrega, H., Coffman, G.A., & Haley, R. (1989). DSM-III disorders in a largesample of psychiatric patients: Frequency and specificity of diagnoses. American Journal ofPsychiatry, 146, 212–219.

Mossman, D., & Somoza, E. (1989). Maximizing diagnostic information from the dexamethasonesuppression test. Archives of General Psychiatry, 46, 653– 660.

Narrow, W.E., Rae, D.S., Robins, L.N., & Regier, D.A. (2002). Revised prevalence estimates ofmental disorders in the United States. Archives of General Psychiatry, 59, 115–123.


Oliver, J.M., & Simmons, M.E., (1984). Depression as measured by the DSM-III and the BeckDepression Inventory in an unselected adult population. Journal of Consulting and ClinicalPsychology, 52, 892–898.

Othmer, E., & DeSouza, C. (1985). A screening test for somatization disorder (hysteria). AmericanJournal of Psychiatry, 142, 1146–1149.

Peterson, R.A., & Heilbronner, R.L. (1987). The Anxiety Sensitivity Index: Construct validity andfactor analytic structure. Journal of Anxiety Disorders, 1, 117–121.

Pilowsky, I. (1967). Dimensions of hypochondriasis. British Journal of Psychiatry, 113, 89–93.

Posternak, M.A., & Zimmerman, M. (2001). Switching versus augmentation: A prospective, natu-ralistic comparison in depressed, treatment-resistant patients. Journal of Clinical Psychiatry,62, 135–142.

Quitkin, F., Rabkin, J., Ross, D., & Stewart, J. (1984). Identification of true drug response toantidepressants: Use of pattern analysis. Archives of General Psychiatry, 41, 782–86.

Radloff, L.S. (1977). The CES-D scale: A self-report depression scale for research in the generalpopulation. Applied Psychological Measurement, 3, 385– 401.

Robins, L.N., Helzer, J.E., Croughan, J., & Ratcliff, K.S. (1981). National Institute of Mental HealthDiagnostic Interview Schedule: Its history, characteristics, and validity. Archives of GeneralPsychiatry, 38, 381–389.

Robins, L.N., Locke, B.Z., & Regier, D.A. (1991). An overview of psychiatric disorders in Amer-ica. In L.N. Robins & D.A. Regier (Eds.), Psychiatric disorders in America: The epidemio-logic catchment study (pp. 328–386). New York: Free Press.

Rush, A.J., Guillion, C.M., Basco, M.R., Jarrett, R.B., & Trivedi, M.H. (1996). The Inventory ofDepressive Symptomatology (IDS). Psychological Medicine, 26, 477– 486.

Sarason, I. (1984). Stress, anxiety, and cognitive interference: Reactions to tests. Journal of Per-sonality and Social Psychology, 46, 929–938.

Selzer, M.L. (1971). The Michigan Alcoholism Screening Test: The quest for a new diagnosticinstrument. American Journal of Psychiatry, 127, 1653–1658.

Shugar, G., Schertzer, S., Toner, B.B., & DiGasbarro, I. (1992). Development, use, and factoranalysis of a self-report inventory for mania. Comprehensive Psychiatry, 33, 325–331.

Skinner, H.A. (1982). The Drug Abuse Screening Test. Addictive Behaviors, 7, 363–371.

Swartz, M., Hughes, D., George, L., Blazer, D., Landerman, R., & Bucholz, K. (1986). Developinga screening index for community studies of somatization disorder. Journal of PsychiatricResearch, 20, 335–343.

Turner, J.A., & Romano, J.M. (1984). Self-reporting screening measures for depression in chronicpain patients. Journal of Clinical Psychology, 40, 909–913.

Turner, S.M., Beidel, D.C., Dancu, C.V., & Stanley, M.A. (1989). An empirically derived inventoryto measure social fears and anxiety: The Social Phobia and Anxiety Inventory. PsychologicalAssessment: Journal of Consulting and Clinical Psychology, 1, 35– 40.

Williams, J.B.W., Gibbon, M., First, M.B., Spitzer, R.L., Davis, M., Borus, J., Howes, M.J., Kane,J., Pope, H.G., Roundsaville, B., & Wittchen, H.U. (1992). The Structured Clinical Interviewfor DSM-III-R (SCID). Archives of General Psychiatry, 49, 630– 636.

Zich, J.M., Attkisson, C.C., & Greenfield, T.K. (1990). Screening for depression in primarycare clinics: The CES-D and the BDI. International Journal of Psychiatry in Medicine, 20,259–277.

Zimmerman, M., & Coryell, W. (1987). The Inventory to Diagnose Depression (IDD): A self-report scale to diagnose major depressive disorder. Journal of Consulting and Clinical Psy-chology, 55, 55–59.

Zimmerman, M., Coryell, W., Corenthal, C., & Wilson, S. (1986). A self-report scale to diagnosemajor depressive disorder. Archives of General Psychiatry, 43, 1076–1081.


Zimmerman, M., & Mattia, J.I. (1998). Body dysmorphic disorder in psychiatric outpatients: Rec-ognition, prevalence, comorbidity, demographic, and clinical correlates. Comprehensive Psy-chiatry, 39, 265–270.

Zimmerman, M., & Mattia, J.I. (1999a). Is posttraumatic stress disorder underdiagnosed in routineclinical settings? Journal of Nervous and Mental Disease, 187, 420– 428.

Zimmerman, M., & Mattia, J.I. (1999b). Psychiatric diagnosis in clinical practice: Is comorbiditybeing missed? Comprehensive Psychiatry, 40, 182–191.

Zimmerman, M., & Mattia, J.I. (1999c). The reliability and validity of a screening questionnaire for14 DSM-IV Axis I disorder (The Psychiatric Diagnostic Screening Questionnaire). Journal ofClinical Psychiatry, 60, 677– 683.

Zung, W.W.K., Magill, M., Moore, J.T., & George, D.T. (1983). Recognition and treatment ofdepression in a family medicine practice. Journal of Clinical Psychiatry, 44, 3– 6.


Documents

The Diagnostic Inventory for Depression: A self-report scale to diagnose DSM-IV major depressive disorder