Advertisement

Case-finding for depression in primary care: a randomized trial∗ ∗

      Abstract

      PURPOSE: Depression is a highly prevalent, morbid, and costly illness that is often unrecognized and inadequately treated. Because depression questionnaires have the potential to improve recognition, we evaluated the accuracy and effects on primary care of two case-finding instruments compared to usual care.
      SUBJECTS AND METHODS: The study was conducted at three university-affiliated and one community-based medical clinics. Consecutive patients were randomly assigned to be asked a single question about mood, to fill out the 20-item Center for Epidemiologic Studies Depression Screen, or to usual care. Within 72 hours, patients were assessed for Diagnostic and Statistical Manual of Mental Disorders Third Revised Edition (DSM-III-R) disorders by an assessor blinded to the screening results. Process of care was assessed using chart audit and administrative databases; patient and physician satisfaction was assessed using Likert scales. At 3 months, depressed patients and a random sample of nondepressed patients were re-assessed for DSM-III-R disorders and symptom counts.
      RESULTS: We approached 1,083 patients, of whom 969 consented to screening and were assigned to the single question (n = 330), 20-item questionnaire (n = 323), or usual care (n = 316). The interview for DSM-III-R diagnosis was completed in 863 (89%) patients; major depression, dysthymia, or minor depression was present in 13%. Both instruments were sensitive, but the 20-item questionnaire was more specific than the single question (75% vs 66%, P = 0.03). The 20-item questionnaire was less likely to be self-administered (54% vs 90%) and took significantly more time to complete (15 vs 248 seconds). Case-finding with the 20-item questionnaire or single question modestly increased depression recognition, 30/77 (39%) compared with 11/38 (29%) in usual care (P = 0.31) but did not affect treatment (45% vs 43%, P = 0.88). Effects on DSM-III-R symptoms were mixed. Recovery from depression was more likely in the case-finding than usual care groups, 32/67 (48%) versus 8/30 (27%, P = 0.03), but the mean improvement in depression symptoms did not differ significantly (1.6 vs 1.5 symptoms, P = 0.21).
      CONCLUSIONS: A simple question about depression has similar performance characteristics as a longer 20-item questionnaire and is more feasible because of its brevity. Case-finding leads to a modest increase in recognition rates, but does not have consistently positive effects on patient outcomes.
      Depression is a common, costly, and sometimes fatal illness (

      Depression Guideline Panel. Depression in Primary Care: Volume 1. Detection and Diagnosis. Clinical Practice Guideline, Number 5. Rockville, MD: US Department of Health and Human Services, Public Health Service, Agency for Health Care Policy and Research, AHCPR Publication No. 93-0550. 1993.

      ,
      • Wells K.B
      • Stewart A.L
      • Hays R.D
      • et al.
      The functioning and well-being of depressed patients results from the Medical Outcomes study.
      ,
      • Spitzer R.L
      • Kroenke K
      • Linzer M
      • et al.
      Health-related quality of life in primary care patients with mental disorders. Results from the PRIME-MD 1000 Study.
      ). In the United States major depression is associated with more than 20,000 suicides and $44 billion in health-care costs annually (
      • Greenberg P.E
      • Stiglin L.E
      • Finkelstein S.N
      • Berndt E.R
      The economic burden of depression in 1990.
      ). Primary-care physicians are the initial health-care contact for most patients with mental illness and can provide integrated, comprehensive health care for individuals with depression and coexisting medical illness (
      • Norquist G.S
      • Regier D.A
      The epidemiology of psychiatric disorders and the de factor mental health care system.
      ). However, primary-care physicians fail to recognize and treat 30% to 50% of depressed patients (
      • Simon G.E
      • Vonkorff M
      Recognition, management, and outcomes of depression in primary care.
      ,
      • Gerber P.D
      • Barrett J
      • Manheimer E
      • Whiting R
      • Smith R
      Recognition of depression by internists in primary care a comparison of internist and “gold standard” psychiatric assessments.
      ). Some experts have recommended using a depression questionnaire as a case-finding tool to enhance recognition (
      • Zung W.W
      The role of rating scales in the identification and management of the depressed patient in the primary care setting.
      ,
      • Zich J.M
      • Attkisson C.C
      • Greenfield T.K
      ).
      Case-finding for depression is intuitively appealing. Patients who present for medical care would complete a simple questionnaire, and those scoring above a specified threshold would be asked more detailed questions for diagnosis. In practice, questionnaire length (most are about 20 items) and time for scoring are pragmatic barriers to implementation; fewer than 25% of primary-care physicians use them (
      • Banazak D.A
      Late-life depression in primary care. How well are we doing?.
      ). Brief, easy to score instruments, such as the PRIME-MD and Symptom Driven Diagnostic Scale, have been developed in response to this need (
      • Spitzer R.L
      • Williams J.B
      • Kroenke K
      • et al.
      Utility of a new procedure for diagnosing mental disorders in primary care. The PRIME-MD 1000 study.
      ,
      • Broadhead W.E
      • Leon A.C
      • Weissman M.M
      • et al.
      Development and validation of the SDDS-PC screen for multiple mental disorders in primary care.
      ). Although theoretically attractive, it remains uncertain if brief instruments will perform similarly to more comprehensive measures.
      We conducted a randomized trial to determine whether a novel single question instrument is as well accepted and has similar accuracy as the Center for Epidemiological Studies Depression Questionnaire, a standard 20-item instrument. Second, we evaluated whether case-finding, as compared to usual care, improved the process of care, patient satisfaction, and depression outcomes.

      Methods

      Subjects

      The study was conducted at a community-based family medicine clinic, a Veterans-Affairs general internal medicine clinic, a university-affiliated general internal medicine clinic located in San Antonio, Texas, and a university-affiliated general internal medicine clinic in Washington, DC. Participating physicians were given a copy of the “Quick Reference Guide for Clinicians” (

      Depression Guideline Panel. Depression in Primary Care: Quick Reference Guide for Clinicians. Clinical Practice Guideline, Number 5. Rockville, MD: US Department of Health and Human Services, Public Health Service, Agency for Health Care Policy and Research, AHCPR Publication No. 93-0552, 1993.

      ) on managing depression in primary care and a continuing medical education session on interpreting case-finding questionnaires and diagnosing and treating depression. Participating physicians were community family physicians (n = 3), faculty general internists (n = 16), and internal medicine housestaff (n = 61). The community and faculty physicians cared for almost 50% of the study participants.
      On designated days, consecutive adult patients were approached for participation. Patients without a telephone or stable address were excluded. Patients giving informed consent were randomly assigned to usual care, case-finding with a single question about mood, or case-finding with the 20-item instrument. Random assignment was stratified by site and was determined by a computer-generated, blocked randomization log. Patients assigned to case-finding completed the depression questionnaire, and the results were reported to their physician on a bright orange report form. Questionnaire results were reported as negative or positive for depressive symptoms that, if positive, warranted additional questioning to determine if clinical depression was present. After the physician visit, patients were contacted by phone to complete the criterion standard diagnostic interview schedule for Diagnostic and Statistical Manual of Mental Disorders Third Revised Edition (DSM-III-R) diagnosis, the CAGE instrument (a screen for alcohol abuse), and a measure of patient satisfaction. The study was approved by the local human subjects committees.

      Measurements

      Case finding instruments

      The single question is “Have you felt depressed or sad much of the time in the past year?” and is answered yes or no (
      • Rost K
      • Burnam M.A
      • Smith G.R
      Development of screeners for depressive disorders and substance disorder history.
      ). The Center for Epidemiological Studies Depression Questionnaire is a 20-item validated instrument that focuses on depressive symptoms within the last week; scores ≥16 identify patients with probable depression (
      • Radloff L.S
      The CES-D scale a self-report depression scale for research in the general population.
      ). Case-finding instruments were self-administered unless the patient could not read or requested that the questionnaire be read to them.

      Psychiatric and medical assessments

      Psychiatric diagnoses were assessed with the mood and anxiety sections of the diagnostic interview schedule, version III-R, by assessors who were blinded to the patient’s random assignment and case-finding results (
      • Robins L.N
      • Helzer J.E
      • Croughan J
      • Ratcliff K.S
      National Institute of Mental Health and Diagnostic Interview Schedule its history, characteristics, and validity.
      ,
      • Helzer J.E
      • Spitznagel E.L
      • McEvoy L
      The predictive validity of lay Diagnostic Interview Schedule diagnoses in the general population. A comparison with physician examiners.
      ,
      • Karno M
      • Burnam M.A
      • Escobar J.I
      • Hough R.L
      • Eaton W.W
      Development of the Spanish-language version of the National Institute of Mental Health Diagnostic Interview Schedule.
      ,
      • Burnam M.A
      • Karno M
      • Hough R.L
      • Escobar J.I
      • Forsythe A.B
      The Spanish Diagnostic Interview Schedule. Reliability and comparison with clinical diagnoses.
      ,
      • Wells K.B
      • Burnam M.A
      • Leake B
      • Robins L.N
      Agreement between face-to-face and telephone-administered versions of the depression section of the NIMH diagnostic interview schedule.
      ). Psychiatric diagnoses were made using DSM-III-R criteria; minor depression was defined as depressed mood or anhedonia plus one to three additional DSM-III-R symptoms of major depression (
      • Broadhead W.E
      • Blazer D.G
      • George L.K
      • Tse C.K
      Depression, disability days, and days lost from work in a prospective epidemiologic survey.
      ). Current depression was defined as meeting the criteria for major depression, dysthymia, or minor depression within the month preceding the clinic visit. In addition, this interview yielded a DSM-III-R symptom count ranging from 0 to 9. The CAGE questionnaire was used to identify patients with a history of probable alcohol abuse based on a score ≥2 (
      • Mayfield D.G
      • Mcleod G
      • Hall P
      The CAGE questionnaire validation of a new alcoholism screening instrument.
      ,
      • Ewing J.A
      Detecting alcoholism the CAGE questionnaire.
      ). Functional status was measured using the Medical Outcomes Study SF-36 instrument (
      • Ware Jr, J.E
      • Kosinski M.A
      • Keller S.D
      ). Interviews were completed in the subject’s preferred language; all instruments were adapted for use in Spanish using a reverse translation process (
      • Guillemin F
      • Bombardier C
      • Beaton D
      Cross-cultural adaptation of health-related quality of life measures literature review and proposed guidelines.
      ,
      • Flaherty J.A
      • Gaviria F.M
      • Pathak D
      • et al.
      Developing instruments for cross-cultural psychiatric research.
      ). Chart audit, using the Duke Severity of Illness Scale (0 to 100, higher scores indicate greater comorbidity), was used to assess patients’ severity of illness (
      • Parkerson G.R.J
      • Michener J.L
      • Wu L.R
      • et al.
      Associations among family support, family stress, and personal functional health status.
      ).

      Patient and physician satisfaction

      Patient judgments about the technical and interpersonal aspects of care were assessed using the Visit Rating Questionnaire from the Medical Outcomes Study (
      • Rubin H.R
      • Gandek B
      • Rogers W.H
      • Kosinski M.A
      • McHorney C.A
      • Ware J.E
      Patients’ ratings of outpatient visits in different practice settings results from the medical outcomes study.
      ). For patients assigned to a case-finding strategy, physicians were asked to rate the impact of the depression screening score on the patient’s management at that visit. The physician chose from five responses ranging from “very helpful” to “very unhelpful.”

      Process and patient outcomes

      Physician recognition of depression was assessed from medical records. Patients were classified as “recognized” if depression was listed as a diagnosis in the assessment and plan or if there was a listing of depressive symptoms in the physician’s note. Patients were considered to have been treated for depression if the medical record documented counseling for depression, a mental health referral, or an antidepressant prescribed for depression. Medical records were supplemented by administrative and pharmacy databases to identify visits to a mental health professional (within 3 months) or a dispensed antidepressant medication for depression within 1 month of the initial visit. Record reviews were completed by a physician blinded to patient assignment.
      At the three San Antonio clinics, patients with a criterion diagnosis of current depression and a random sample of nondepressed patients were selected for follow-up. The sample of nondepressed patients was designed to oversample those who had depressive symptoms but did not meet the DSM-III-R criteria for depression. Patients were contacted by telephone at 3 months, and the diagnostic interview schedule was re-administered to determine current affective status. Symptom reduction was defined as ≤1 DSM-III-R symptom of depression. For patients unavailable at the planned 3-month assessment, continued efforts were made for up to 1 year to complete the diagnostic interview schedule. In patients contacted after the planned 3-month follow-up, a retrospective assessment of their affective status at 3 months was determined by asking them to date the onset of any change in symptom status.

      Analysis

      Sensitivity and specificity of the 20-item instrument and the single question for detecting depression were calculated. Analysis of the time required to complete the single question or 20-item instrument was restricted to individuals who were able to complete the questionnaire without assistance. The influence of language, ethnic background, medical comorbidity, and depression severity on use of the questionnaire was examined using logistic regression analysis (
      • Harrell F.E
      The LOGIST procedure.
      ). Between group comparisons were made using chi-square analysis for categorical data and the Wilcoxon statistic for continuous data. The associations between depression recognition or treatment and depression severity or physician experience were examined using logistic regression analysis to control for case-finding assignment.
      The process of care outcomes (recognition and treatment rates) were restricted to patients who met the criteria for a diagnosis of current depression. The effects of case-finding on 3-month affective outcomes were examined in three ways. First, the prevalence of depression at that time was determined in patients assigned to case-finding (single question or 20-item instrument) compared to that in the usual-care group. The other analyses were limited to patients with current depression at baseline. Recovery rates (defined as ≤1 DSM-III-R symptom) were compared between case-finding and usual-care groups. Finally, the changes from baseline to 3 months in DSM-III-R symptom counts were compared between the case-finding and usual-care groups. Recognition, treatment, and recovery rates were adjusted for baseline depression severity (DSM-III-R symptom count) using hierarchical logistic regression analysis. Statistical significance was set at P <0.05; 95% confidence intervals (CI) were estimated (29,30).

      Results

      From November 1993 through August 1995, 1,083 patients were approached for study entry (Figure 1). Of these, 969 patients (89%) were eligible and randomly assigned to usual care (n = 316), the single question (n = 330), or the 20-item instrument (n = 323). The 114 patients who were ineligible or refused participation were more likely to be men but did not differ from participants by age or ethnic background. Of the participants, 863 (89%) completed the criterion standard interview; patients were predominately female (71%), of Hispanic ethnic background (60%), and had low income (76% with personal income <$16,800; Table 1). Eleven percent were making their first visit to the study physician. Medical comorbidity was moderate, with a median Duke Severity of Illness Score of 34, and a mean of three chronic medical conditions.
      Figure thumbnail GR1
      Figure 1Patient recruitment, outcomes assessment, and follow-up.
      Table 1Patient Characteristics by Group Assignment among Those Who Completed the Diagnostic Interview
      Values are percentages or medians with interquartile range in brackets. For all variables, there were no significant between-group differences.
      CharacteristicUsual Care (n = 276)Single Question (n = 291)20-Item Instrument (n = 296)
      Age, years56 [47,66]58 [49,65]59 [49,65]
      Female gender71%68%74%
      Ethnic background
      Hispanic58%59%61%
      White30%28%29%
      Black10%12%9%
      Other2%1%1%
      Spanish-speaking25%26%27%
      Education, years11 [6,14]11 [6,12]10 [6,12]
      Annual income
      <$7,20036%40%42%
      $7,200–$16,79940%35%34%
      ≥$16,80024%25%24%
      Depression diagnosis13.8%13.8%12.4%
      Major depression
      Includes major depression with concurrent dysthymia.
      9.8%6.9%7.4%
      Dysthymia1.5%2.1%0.3%
      Minor depression2.5%4.8%4.7%
      History of alcohol abuse15%15%17%
      Anxiety disorder2.5%1.7%2.7%
      Medical diagnoses3 [2,5]3 [2,4]3 [2,4]
      Duke Severity of Illness35 [24,47]34 [24,49]34 [24,48]
      Prescription medications3 [1,5]3 [1,4]3 [1,4]
      Values are percentages or medians with interquartile range in brackets. For all variables, there were no significant between-group differences.
      Includes major depression with concurrent dysthymia.
      A diagnosis of current depression that met DSM-III-R criteria was made in 115 patients (13%). Of these, 37 patients had major depression, 35 had minor depression, 32 had major depression and dysthymia (double depression), and 11 had dysthymia. Panic or generalized anxiety disorder was present in 35 patients; in 22 patients the anxiety disorder coexisted with depression. Probable alcohol abuse was present in 131 patients (15%); in 28 patients the alcohol abuse coexisted with depression.

      Performance characteristics of case-finding instruments

      More patients completed the single question without assistance than the 20-item instrument (90% vs 54%, P <0.0001). The single question took significantly less time to complete (15 vs 248 seconds, P <0.0001). Low educational level, increasing age, black ethnic background, and male gender were associated with interviewer administration of the questionnaire.
      A substantial proportion of patients answered the single question “yes” (41%) or scored ≥16 on the 20-item instrument (33%). The sensitivity of both the single question (85%) and 20-item instrument (88%) was highly similar, though the 20-item instrument was somewhat more specific (75%) than the single question (66%; Table 2). Case-finding instruments were more sensitive (94% vs 67%) and less specific (65% vs 77%) among Hispanic patients (P <0.01). There was a trend for the instruments to be more sensitive at detecting major depression and dysthymia than minor depression.
      Table 2Performance of the Case-finding Instruments
      Sensitivity and specificity were estimated based on results in 291 patients in a single-question group and 296 patients in 20-item-questionnaire group who had a diagnostic interview for depression.
      CI = confidence interval.
      Performance MeasureSingle Question (n = 330)20-Item Instrument (n = 323)Difference (95% CI)
      Self-administered90%54%36% (26%,46%)
      Administration time (seconds)15248−233 (−204,−262)
      Positive screen41%33%0.08 (−0.05,0.21)
      Sensitivity85%88%−0.03 (−0.20,0.14)
      Specificity66%75%−0.09 (−0.19,0.01)
      Sensitivity and specificity were estimated based on results in 291 patients in a single-question group and 296 patients in 20-item-questionnaire group who had a diagnostic interview for depression.
      legend CI = confidence interval.

      Effects on process of care

      Since case-finding with the single question and 20-item instrument were similar, these two groups of patients were combined and compared with the usual-care group. Among patients assigned to case-finding, 30 (39%) of the 77 patients with current depression had that diagnosis recognized by their physician, compared with 11 (29%) of the 38 patients in the usual-care group (difference = 10%, 95% CI for difference = −23% to 43%; Figure 2). A new diagnosis of depression was made in 10 (13%) of 77 case-finding patients and 1 (3%) of 38 usual-care patients (difference = 10%, 95% CI for difference = 1 to 19%).
      Figure thumbnail GR2
      Figure 2Rates of recognition of depression and assessment for suicide risk (n = 77 for case-finding and 38 for usual care).
      Physician counseling, drug treatment, and referral treatment were similar in the two groups (Figure 3). Patients with major depression were more likely to be recognized (45% vs 24%, P <0.02) and treated (55% vs 28%, P = 0.005) than those with minor depression or dysthymia. Similarly, patients with more functional impairment were more likely to be recognized and treated. These differences persisted after controlling for case-finding assignment. Compared with resident physicians, practicing physicians were no more likely to recognize (36% vs 36%) or treat (51% vs 38%, P = 0.17) patients with depression.
      Figure thumbnail GR3
      Figure 3Treatment and referral rates for patients with current depression in the case-finding (n = 77) and usual-care (n = 38) groups.

      Effects on outcomes

      Longitudinal follow-up was planned for 230 patients: 101 patients with current depression and a random sample of 129 patients without current depression. (There was no planned follow-up for patients enrolled at the Washington site.) Follow-up was completed for 216 (94%) of these patients. Among the 101 depressed patients with planned follow-up, 84 were assessed at 3 months, 13 were assessed 6 to 12 months later, and four could not be contacted (three usual care, one case-finding).
      At 3 months, the prevalence of depression was similar in the two groups: 37% (56/153) in the case-finding groups compared with 46% (30/65) in usual care (P = 0.19). However, patients in the case-finding group were more likely to recover [48% (32/67) had ≤1 DSM-IIIR symptom] than those assigned to usual care [27% (8/30); difference = 21%, 95% CI for the difference = 1% to 41%]. Results were similar, albeit not quite significant, among only those patients who had follow-up at 3 months (case-finding 48% (29/61) had ≤1 DSM-III-R symptom; usual care 30% (7/23; difference = 18%, 95% CI for the difference = −4% to 40%). After controlling for baseline severity of depression, the mean reduction in DSM-III-R symptom counts was similar for the case-finding (1.6 symptoms) and usual-care (1.5 symptoms) groups (P = 0.21).

      Physician and patient acceptance

      Case-finding did not affect patient satisfaction. The proportion of patients rating their care as “excellent” varied from 56% for physicians’ technical skill to 17% for waiting time, with no differences among the three groups.
      Physicians returned 433 (67%) of 649 rating forms describing the impact of case-finding on the visit. Physicians rated the impact of the information on their management: very helpful (38%), somewhat helpful (29%), no impact (29%), somewhat unhelpful (2%), and very unhelpful (2%). General internal medicine faculty and community physicians were more likely than resident physicians to rate the information as very or somewhat helpful (78% vs 51%, P = 0.001). There were no differences in patient satisfaction or physician acceptance between the single question and 20-item instrument groups.

      Discussion

      Our study confirms that the recognition of depression in the primary-care setting is low in usual care. Encouragingly, patients with more severe depression and more functional impairment were more likely to be diagnosed or treated. These data confirm that primary-care physicians incorporate the level of disability into diagnostic and treatment decisions and are appropriately focusing their efforts on those with greater impairment (
      • Moore J.T
      • Silimperi D.R
      • Bobula J.A
      Recognition of depression by family medicine residents the impact of screening.
      ,
      • Shapiro S
      • German P.S
      • Skinner E.A
      • et al.
      An experiment to change detection and management of mental morbidity in primary care.
      ,
      • Magruder-Habib K
      • Zung W.W
      • Feussner J.R
      Improving physicians’ recognition and treatment of depression in general medical care. Results from a randomized clinical trial.
      ,
      • Linn L.S
      • Yager J
      The effect of screening, sensitization, and feedback on notation of depression.
      ). Nevertheless, many patients with major depression and most with milder forms of depression are unrecognized.
      Case-finding for depression might improve recognition, but it was not recommended by the US Preventive Services Task Force or the clinical practice guidelines supported by the Agency for Health Care Policy and Research (

      Depression Guideline Panel. Depression in Primary Care: Volume 1. Detection and Diagnosis. Clinical Practice Guideline, Number 5. Rockville, MD: US Department of Health and Human Services, Public Health Service, Agency for Health Care Policy and Research, AHCPR Publication No. 93-0550. 1993.

      ,
      • Zich J.M
      • Attkisson C.C
      • Greenfield T.K
      ,
      US Preventive Services Task Force
      ). Prior trials of case-finding have shown generally favorable effects on intermediate outcomes, such as depression recognition and treatment (
      • Moore J.T
      • Silimperi D.R
      • Bobula J.A
      Recognition of depression by family medicine residents the impact of screening.
      ,
      • Shapiro S
      • German P.S
      • Skinner E.A
      • et al.
      An experiment to change detection and management of mental morbidity in primary care.
      ,
      • Magruder-Habib K
      • Zung W.W
      • Feussner J.R
      Improving physicians’ recognition and treatment of depression in general medical care. Results from a randomized clinical trial.
      ,
      • Linn L.S
      • Yager J
      The effect of screening, sensitization, and feedback on notation of depression.
      ,
      • Callahan C.M
      • Hendrie H.C
      • Dittus R.S
      • Brater D.C
      • Hui S.L
      • Tierney W.M
      Improving treatment of late life depression in primary care a randomized clinical trial.
      ,
      • Rand E.H
      • Badger L.W
      • Coggins D.R
      Toward a resolution of contradictions. Utility of feedback from the GHQ.
      ,
      • Rucker L
      • Frye E.B
      • Cygan R.W
      Feasibility and usefulness of depression screening in medical outpatients.
      ,
      • Linn L.S
      • Yager J
      Recognition of depression and anxiety by primary physicians.
      ), but the only study to examine recovery from depression showed no benefit (
      • Callahan C.M
      • Hendrie H.C
      • Dittus R.S
      • Brater D.C
      • Hui S.L
      • Tierney W.M
      Improving treatment of late life depression in primary care a randomized clinical trial.
      ). That study focused on elderly, low-income patients with substantial barriers to effective treatment and medical and social conditions that may have limited the effectiveness of treatment.
      For case-finding to be incorporated into evidence-based guidelines, it must meet a series of criteria that establish beneficial effects on patient outcomes (
      • Sackett D.L
      • Haynes R.B
      • Guyatt G.H
      • Tugwell P
      ,
      • Schwenk T.L
      Screening for depression in primary care a disease in search of a test.
      ). First, the illness in question must be common and have a substantial effect on the quality or quantity of life. Major depression meets these criteria: it affects 6% to 10% of primary-care patients, is associated with 20,000 suicides annually, and is estimated to cost $44 billion in direct and indirect costs (
      • Greenberg P.E
      • Stiglin L.E
      • Finkelstein S.N
      • Berndt E.R
      The economic burden of depression in 1990.
      ). Minor depression and dysthymia are more prevalent than major depression and are associated with important functional impairment, excess disability days, and suicide (
      • Broadhead W.E
      • Blazer D.G
      • George L.K
      • Tse C.K
      Depression, disability days, and days lost from work in a prospective epidemiologic survey.
      ,
      • Williams Jr, J.W
      • Kerber C.A
      • Mulrow C.D
      • Medina A
      • Aguilar C
      Depressive disorders in primary care prevalence, functional disability, and identification.
      ). Second, the proposed case-finding test must be sufficiently accurate, safe, and low in cost to justify its use. At least nine case-finding instruments, ranging in length from two to 28 questions, have been used to detect depression in primary-care settings (
      • Mulrow C.D
      • Williams Jr, J.W
      • Gerety M.B
      • Ramirez G
      • Montiel O.M
      • Kerber C
      Case-finding instruments for depression in primary care settings.
      ). Our data show that a single question about depression is a sensitive measure for depressive disorders and is only slightly less specific than the longer 20-item instrument. A study comparing a two-item instrument with longer questionnaires showed similar results (
      • Whooley M.A
      • Avins A.L
      • Miranda J
      • Browner W.S
      Case-finding instruments for depression two questions are as good as many.
      ). The single question was much more likely to be self-administered and took less time to complete, important advantages for a busy primary-care practice. Both instruments were well accepted by patients and physicians.
      The third criterion is that case-finding must be able to detect the illness before it becomes clinically obvious. Several studies in primary-care settings have shown relatively low recognition rates for depression (

      Depression Guideline Panel. Depression in Primary Care: Volume 1. Detection and Diagnosis. Clinical Practice Guideline, Number 5. Rockville, MD: US Department of Health and Human Services, Public Health Service, Agency for Health Care Policy and Research, AHCPR Publication No. 93-0550. 1993.

      ,
      • Gerber P.D
      • Barrett J
      • Manheimer E
      • Whiting R
      • Smith R
      Recognition of depression by internists in primary care a comparison of internist and “gold standard” psychiatric assessments.
      ,
      • Zich J.M
      • Attkisson C.C
      • Greenfield T.K
      ,
      • Magruder-Habib K
      • Zung W.W
      • Feussner J.R
      Improving physicians’ recognition and treatment of depression in general medical care. Results from a randomized clinical trial.
      ). Our data show a modest, but statistically nonsignificant, increase in the recognition of depression among patients who undergo case-finding. Prior studies have also shown increased recognition, particularly when the intervention focused on depression rather than a broad array of mental illness (
      • Moore J.T
      • Silimperi D.R
      • Bobula J.A
      Recognition of depression by family medicine residents the impact of screening.
      ,
      • Shapiro S
      • German P.S
      • Skinner E.A
      • et al.
      An experiment to change detection and management of mental morbidity in primary care.
      ,
      • Magruder-Habib K
      • Zung W.W
      • Feussner J.R
      Improving physicians’ recognition and treatment of depression in general medical care. Results from a randomized clinical trial.
      ,
      • Linn L.S
      • Yager J
      The effect of screening, sensitization, and feedback on notation of depression.
      ,
      • Callahan C.M
      • Hendrie H.C
      • Dittus R.S
      • Brater D.C
      • Hui S.L
      • Tierney W.M
      Improving treatment of late life depression in primary care a randomized clinical trial.
      ,
      • Rand E.H
      • Badger L.W
      • Coggins D.R
      Toward a resolution of contradictions. Utility of feedback from the GHQ.
      ,
      • Linn L.S
      • Yager J
      Screening of depression in relationship to subsequent patient and physician behavior.
      ,
      • Hoeper E.W
      • Kessler L.G
      • Nycz G.R
      • Burke Jr, J.D
      • Pierce W.E
      The usefulness of screening for mental illness.
      ). Taken together, these data suggest that case-finding leads to moderate increases in the recognition of depression. Given the many other issues that the primary-care physicians in our study needed to address, a larger impact was unlikely from this simple intervention.
      The final criterion is that case-finding should improve patient outcome. Implicit in this criterion is that effective treatments must be available and that treatment given early must be more effective than treatment given when the diagnosis becomes clinically obvious. For minor depression, antidepressant medication may improve outcomes (
      • Parnetti L
      • Sommacal S
      • Morselli L.A.M
      • Senin U
      Multicentre controlled randomised double-blind placebo study of minaprine in elderly patients suffering from prolonged depressive reaction.
      ), but data are too limited to determine if treatment is efficacious. Antidepressant drug therapy and psychotherapy are effective treatments for major depression and dysthymia, increasing response rates from about 30% for placebo to 70% for active treatment (

      Depression Guideline Panel. Depression in Primary Care: Treatment of Major Depression. Clinical Practice Guideline, Number 5. Rockville, MD: US Department of Health and Human Services, Public Health Service, Agency for Health Care Policy and Research, AHCPR Publication No. 93-0051. 1993.

      ). The efficacy of these therapies has been demonstrated in the primary-care setting (
      • Schulberg H.C
      • Bock M.R
      • Madonia M.J
      • et al.
      Treating major depression in primary care practice eight month clinical outcomes.
      ). Further, patients treated earlier in their illness are more likely to recover (
      • Light R.W
      • Merrill E.J
      • Despars J
      • Gordon G.H
      • Mutalipassi L.R
      Doxepin treatment of depressed patients with chronic obstructive pulmonary disease.
      ,
      • Keller M.B
      • Shapiro R.W
      • Lavori P.W
      • Wolfe N
      Recovery in major depressive disorder analysis with the life table and regression models.
      ). Thus, there is a strong scientific basis for efficacious therapies and a rationale for treatment early in the illness.
      Our data showed no discernible effect of case-finding for the intermediate outcomes of counseling, filled antidepressant prescriptions, or mental health referrals. There are several possible explanations. First, our methods for detecting recognition and treatment may have missed some treatments. Second, the barriers to treatment, particularly in low-income populations, may be greater than the barriers for recognition. Third, to provide effective therapy, the primary-care physician must be knowledgeable and have effective communication skills. We reported previously that our study physicians were moderately knowledgeable about drug therapy but, like their peers, had little confidence in their ability to treat depression with counseling (
      • Shao W
      • Williams Jr, J.W
      • Lee S
      • Badgett R.G
      • Aaronson B
      • Cornell J.E
      Knowledge and attitudes about depression among non-generalists and generalists.
      ). Finally, low-income patients may not have been able to afford antidepressant drug therapy or mental health specialty care.
      There was a trend for case-finding to benefit patient outcomes, but these effects were not usually statistically significant. The lack of a consistent benefit is related in part to the modest improvement in recognition rates and the lack of effect on treatment. It also may be related to characteristics of the study population. The chronic social stresses seen in an urban, low-income, and ethnically diverse group of patients may make depression more refractory to treatment. Further, cross-cultural studies have shown that Hispanic patients may interpret symptoms of depression more benignly, are more likely to use prayer and other nonmedical therapies, and are less likely to receive treatment in the mental health specialty setting than non-Hispanic white patients (
      • Edgerton R.B
      • Karno M
      Mexican-American bilingualism and the perception of mental illness.
      ,
      • Sartorius N
      Cross-cultural research on depression.
      ). Although antidepressant medications appear equally efficacious for different ethnic groups, their effectiveness in clinical use may be attenuated by cultural beliefs and patient preferences. For similar populations, a more potent and culturally sensitive intervention may be needed to produce consistent, clinically significant changes.
      The validity of our findings is strengthened by the randomized design, the varied clinical settings, the careful categorization of subjects using DSM-III-R criteria, and the assessment of multiple outcomes, such as patient and physician satisfaction and affective status. By using independent samples to assess performance of the case-finding instruments, we were able to avoid order effects and improve our ability to detect true differences between instruments. An important study limitation is the reliance on medical record review for measuring recognition of depression, which was addressed by the careful assessment of drug therapy and mental health visits, de facto measures of recognition. Nevertheless, our estimates of depression recognition are a lower bound for the true rates.
      In summary, case-finding is well accepted by patients and physicians. Because of its brevity and comparable performance, we recommend a single question for clinicians wishing to use depression case-finding in their practice. Its use can be expected to increase recognition rates modestly and may improve patient affective outcomes at 3 months.
      • Cohen J
      Differences between proportions.

      References

      1. Depression Guideline Panel. Depression in Primary Care: Volume 1. Detection and Diagnosis. Clinical Practice Guideline, Number 5. Rockville, MD: US Department of Health and Human Services, Public Health Service, Agency for Health Care Policy and Research, AHCPR Publication No. 93-0550. 1993.

        • Wells K.B
        • Stewart A.L
        • Hays R.D
        • et al.
        The functioning and well-being of depressed patients.
        JAMA. 1989; 262: 914-919
        • Spitzer R.L
        • Kroenke K
        • Linzer M
        • et al.
        Health-related quality of life in primary care patients with mental disorders. Results from the PRIME-MD 1000 Study.
        JAMA. 1995; 274: 1511-1517
        • Greenberg P.E
        • Stiglin L.E
        • Finkelstein S.N
        • Berndt E.R
        The economic burden of depression in 1990.
        J Clin Psychiatr. 1993; 54: 405-418
        • Norquist G.S
        • Regier D.A
        The epidemiology of psychiatric disorders and the de factor mental health care system.
        Ann Rev Med. 1996; 47: 473-479
        • Simon G.E
        • Vonkorff M
        Recognition, management, and outcomes of depression in primary care.
        Arch Fam Med. 1995; 4: 99-105
        • Gerber P.D
        • Barrett J
        • Manheimer E
        • Whiting R
        • Smith R
        Recognition of depression by internists in primary care.
        J Gen Intern Med. 1989; 4: 7-13
        • Zung W.W
        The role of rating scales in the identification and management of the depressed patient in the primary care setting.
        J Clin Psychiatr. 1990; 51: 72-76
        • Zich J.M
        • Attkisson C.C
        • Greenfield T.K
        Depression in Primary Care. Routledge, New York & London1990
        • Banazak D.A
        Late-life depression in primary care. How well are we doing?.
        J Gen Intern Med. 1996; 11: 163-167
        • Spitzer R.L
        • Williams J.B
        • Kroenke K
        • et al.
        Utility of a new procedure for diagnosing mental disorders in primary care. The PRIME-MD 1000 study.
        JAMA. 1994; 272: 1749-1756
        • Broadhead W.E
        • Leon A.C
        • Weissman M.M
        • et al.
        Development and validation of the SDDS-PC screen for multiple mental disorders in primary care.
        Arch Fam Med. 1995; 4: 211-219
      2. Depression Guideline Panel. Depression in Primary Care: Quick Reference Guide for Clinicians. Clinical Practice Guideline, Number 5. Rockville, MD: US Department of Health and Human Services, Public Health Service, Agency for Health Care Policy and Research, AHCPR Publication No. 93-0552, 1993.

        • Rost K
        • Burnam M.A
        • Smith G.R
        Development of screeners for depressive disorders and substance disorder history.
        Med Care. 1993; 31: 189-200
        • Radloff L.S
        The CES-D scale.
        Appl Psychol Meas. 1977; 1: 385-401
        • Robins L.N
        • Helzer J.E
        • Croughan J
        • Ratcliff K.S
        National Institute of Mental Health and Diagnostic Interview Schedule.
        Arch Gen Psychiatr. 1981; 38: 381-389
        • Helzer J.E
        • Spitznagel E.L
        • McEvoy L
        The predictive validity of lay Diagnostic Interview Schedule diagnoses in the general population. A comparison with physician examiners.
        Arch General Psychiatr. 1987; 44: 1069-1077
        • Karno M
        • Burnam M.A
        • Escobar J.I
        • Hough R.L
        • Eaton W.W
        Development of the Spanish-language version of the National Institute of Mental Health Diagnostic Interview Schedule.
        Arch Gen Psychiatr. 1983; 40: 1183-1188
        • Burnam M.A
        • Karno M
        • Hough R.L
        • Escobar J.I
        • Forsythe A.B
        The Spanish Diagnostic Interview Schedule. Reliability and comparison with clinical diagnoses.
        Arch Gen Psychiatr. 1983; 40: 1189-1196
        • Wells K.B
        • Burnam M.A
        • Leake B
        • Robins L.N
        Agreement between face-to-face and telephone-administered versions of the depression section of the NIMH diagnostic interview schedule.
        J Psychiatr Res. 1988; 22: 207-220
        • Broadhead W.E
        • Blazer D.G
        • George L.K
        • Tse C.K
        Depression, disability days, and days lost from work in a prospective epidemiologic survey.
        JAMA. 1990; 264: 2524-2528
        • Mayfield D.G
        • Mcleod G
        • Hall P
        The CAGE questionnaire.
        Am J Psychiatr. 1974; 131: 1121-1123
        • Ewing J.A
        Detecting alcoholism.
        JAMA. 1984; 252: 1905-1907
        • Ware Jr, J.E
        • Kosinski M.A
        • Keller S.D
        SF-36 Physical and Mental Health Summary Scales: A User’s Manual. The Health Institute, Boston MA1994
        • Guillemin F
        • Bombardier C
        • Beaton D
        Cross-cultural adaptation of health-related quality of life measures.
        J Clin Epidemiol. 1993; 46: 1417-1432
        • Flaherty J.A
        • Gaviria F.M
        • Pathak D
        • et al.
        Developing instruments for cross-cultural psychiatric research.
        J Nerv Ment Dis. 1988; 176: 257-263
        • Parkerson G.R.J
        • Michener J.L
        • Wu L.R
        • et al.
        Associations among family support, family stress, and personal functional health status.
        J Clin Epidemiol. 1989; 42: 217-229
        • Rubin H.R
        • Gandek B
        • Rogers W.H
        • Kosinski M.A
        • McHorney C.A
        • Ware J.E
        Patients’ ratings of outpatient visits in different practice settings.
        JAMA. 1993; 270: 835-840
        • Harrell F.E
        The LOGIST procedure.
        in: SAS/STAT User’s Guide, Release 6.04. SAS Institute, Cary, NC1990: 175-230
        • Cohen J
        Differences between proportions.
        in: Statistical Power Analysis for the Behavioral Sciences. Lawrence Erlbaum Associates, Hillsdale, NJ1988: 179-213
        • Moore J.T
        • Silimperi D.R
        • Bobula J.A
        Recognition of depression by family medicine residents.
        J Fam Pract. 1978; 7: 509-513
        • Shapiro S
        • German P.S
        • Skinner E.A
        • et al.
        An experiment to change detection and management of mental morbidity in primary care.
        Med Care. 1987; 25: 327-339
        • Magruder-Habib K
        • Zung W.W
        • Feussner J.R
        Improving physicians’ recognition and treatment of depression in general medical care. Results from a randomized clinical trial.
        Med Care. 1990; 28: 239-250
        • Linn L.S
        • Yager J
        The effect of screening, sensitization, and feedback on notation of depression.
        J Med Educ. 1980; 55: 942-949
        • US Preventive Services Task Force
        Guide to Clinical Preventive Services. 1996. 2nd ed. International Medical Publishing, Alexandria, VA1996
        • Callahan C.M
        • Hendrie H.C
        • Dittus R.S
        • Brater D.C
        • Hui S.L
        • Tierney W.M
        Improving treatment of late life depression in primary care.
        J Am Geriatr Soc. 1994; 42: 839-846
        • Rand E.H
        • Badger L.W
        • Coggins D.R
        Toward a resolution of contradictions. Utility of feedback from the GHQ.
        Gen Hosp Psychiatr. 1988; 10: 189-196
        • Rucker L
        • Frye E.B
        • Cygan R.W
        Feasibility and usefulness of depression screening in medical outpatients.
        Arch Intern Med. 1986; 146: 729-731
        • Linn L.S
        • Yager J
        Recognition of depression and anxiety by primary physicians.
        Psychosomatics. 1984; 25: 593-600
        • Sackett D.L
        • Haynes R.B
        • Guyatt G.H
        • Tugwell P
        Clinical Epidemiology A Basic Science for Clinical Medicine. Little Brown, Boston/Toronto/London1991: 153-170
        • Schwenk T.L
        Screening for depression in primary care.
        J Gen Intern Med. 1996; 11: 437-439
        • Williams Jr, J.W
        • Kerber C.A
        • Mulrow C.D
        • Medina A
        • Aguilar C
        Depressive disorders in primary care.
        J Gen Intern Med. 1995; 10: 7-12
        • Mulrow C.D
        • Williams Jr, J.W
        • Gerety M.B
        • Ramirez G
        • Montiel O.M
        • Kerber C
        Case-finding instruments for depression in primary care settings.
        Ann Intern Med. 1995; 122: 913-921
        • Whooley M.A
        • Avins A.L
        • Miranda J
        • Browner W.S
        Case-finding instruments for depression.
        J Gen Intern Med. 1997; 12: 439-445
        • Linn L.S
        • Yager J
        Screening of depression in relationship to subsequent patient and physician behavior.
        Med Care. 1982; 20: 1233-1240
        • Hoeper E.W
        • Kessler L.G
        • Nycz G.R
        • Burke Jr, J.D
        • Pierce W.E
        The usefulness of screening for mental illness.
        Lancet. 1984; 7: 33-35
        • Parnetti L
        • Sommacal S
        • Morselli L.A.M
        • Senin U
        Multicentre controlled randomised double-blind placebo study of minaprine in elderly patients suffering from prolonged depressive reaction.
        Drug Invest. 1993; 6: 181-188
      3. Depression Guideline Panel. Depression in Primary Care: Treatment of Major Depression. Clinical Practice Guideline, Number 5. Rockville, MD: US Department of Health and Human Services, Public Health Service, Agency for Health Care Policy and Research, AHCPR Publication No. 93-0051. 1993.

        • Schulberg H.C
        • Bock M.R
        • Madonia M.J
        • et al.
        Treating major depression in primary care practice.
        Arch Gen Psychiatr. 1997; 53: 913-919
        • Light R.W
        • Merrill E.J
        • Despars J
        • Gordon G.H
        • Mutalipassi L.R
        Doxepin treatment of depressed patients with chronic obstructive pulmonary disease.
        Arch Intern Med. 1986; 146: 1377-1380
        • Keller M.B
        • Shapiro R.W
        • Lavori P.W
        • Wolfe N
        Recovery in major depressive disorder.
        Arch General Psychiatr. 1982; 39: 905-910
        • Shao W
        • Williams Jr, J.W
        • Lee S
        • Badgett R.G
        • Aaronson B
        • Cornell J.E
        Knowledge and attitudes about depression among non-generalists and generalists.
        J Fam Pract. 1997; 44: 161-168
        • Edgerton R.B
        • Karno M
        Mexican-American bilingualism and the perception of mental illness.
        Arch Gen Psychiatr. 1971; 24: 286-290
        • Sartorius N
        Cross-cultural research on depression.
        Psychopathology. 1986; 19: 6-11