Another attitudinal problem is the widespread belief that the CMI cannot be helped. Based on the work that is going on in Dallas County it appears that someone forgot to tell folks there that the CMI could not perform well in regular job situations. In May of this year I attended an awards banquet in Dallas, Texas, where dozens of these fine citizens were honored for their work records, and seven of them received recognition for working continuously for five or more years. Truly these citizens can be helped, and they can become productive, if given a chance. JTPA FUNDING I just want to underscore the point made earlier in the hearing by Congressman Don Payne...Job Training Partnership Training Act funding is available to those organizations serving disabled persons. However, our state Independent Living Council has noted that such organizations do not take advantage of this source of funding. So I would urge these organizations to contact the JTPA Private Industry Council in their community and apply for a grant/contract. DISCRIMINATION AGAINST DISABLED PERSONS Title IV of the Americans with Disabilities Act will provide redress for the type of discrimination that took place in Dallas with a program called Anchor Services. This was a supported employment program for CMI, and it occupied a suite of offices on the fifth floor of an office building. After being in operation less than six months, the building manager wrote letters and put on a lot of pressure, trying to force Anchor Services to move out...not because of any breach in the lease, but only because the clients "look strange, and they act strange, and the other tenants are afraid of them". Thank you for giving me this opportunity to testify before your committee. I strongly urge passage of the ADA. The Functioning and Well-being of Results From the Medical Outcomes Study Kenneth B. Wells, MD, MPH; Anita Stewart, PhD; Ron D. Hays, PhD; M. Audrey Bumam, PhD; William Rogers, PhD; We describe the functioning and well-being of patients with depression, relative to patients with chronic medical conditions or no chronic conditions. Data are from 11 242 outpatients in three health care provision systems in three US sites. Patients with either current depressive disorder or depressive symptoms in the absence of disorder tended to have worse physical, social, and role functioning, worse perceived current health, and greater bodily pain than did patients with no chronic conditions. The poor functioning uniquely associated with depressive symptoms, with or without depressive disorder, was comparable with or worse than that uniquely associated with eight major chronic medical conditions. For example, the unique association of days in bed with depressive symptoms was significantly greater than the comparable association with hypertension, diabetes, and arthritis. Depression and chronic medical conditions had unique and additive effects on patient functioning. THE POLICY and clinical impact of both medical and psychiatric conditions is increasingly being assessed not just by their prevalence and associated mortality, but by their impact on subjective patient well-being and functional sta See also pp 907, 925, and 943. tus." Estimates of that impact are used to justify funds for training, research, and service provision. Well-being, or the subjective assessment of quality of life and health, and functional status, or the capacity to perform tasks and activi From The RAND Corp. Santa Monica. Call (Drs Wells, Hays. Burnam, Rogers, Daniele Greentield, and Ware and Ms Berry): Department of Psychiatry and Biobehavioral Sciences, UCLA Neuropsychainc insh lute and Hospital. UCLA School of Medicine, Los Angeles. Call (Dr Wells). Institute for Heath and Aging. University of California, San Francisco (Dr Stewart), and Institute for the improvement of Medical Care and Heath, New England Medical Center Hospitals, Boston, Masa (Drs Greenfield and Ware) The conclusions are those of the authors and do not necessanty reflect the opinion of the sponsors or the authors instutions Reprint requests to The RAND Corp. 1700 Main St, Santa Monica, CA 90406-2138 (Dr Wats) WAMA. 1989-202-914-919) ties, are primary concerns of patients, The awareness that chronic health While the literature indicates that depression is associated with limitations in social and role functioning, few studies have comprehensively examined the well-being and functioning of depressed patients. Herein, we describe unique associations of depression with multiple domains of well-being and functioning. We believe there is a tendency for many clinicians to view the limitations associated with depression as more subjective or of less clinical significance than those associated with major chronic medical conditions. While previous studiesTM have estimated associations of specific chronic medical conditions, such as hypertension, diabetes, and heart disease, with multiple domains of patient func tioning, such studies have not directly compared patients with depression with those with specific chronic medical conditions. We do so herein, to provide a clinical "yardstick" of the morbidity associated with depression. In the United States, at least half of patients who receive any mental health care receive that care only from general medical clinicians, and the remainder have at least some visits to mental health specialista. Previous authors have reached conflicting opinions about the clinical comparability of patients treated by the two provider sectors. Thus, to understand the functioning of patients with depression in any health care provision system, it is necessary to contrast these two patient groups, as we do herein. In current clinical practice and research, there are two major paradigms for defining depression: the general phenomenon of depressive symptoms and specific depressive disorders (eg. discrete psychiatric disorders defined by the Diagnostic and Statistical Manual of Mental Disorders, Third Edition (DSM-IID. While the general medical sector tends to conceptualize depression according to the former definition, much of the mental health specialty sector conceptualizes depression Functioning of Depressed Patents-Walls et al according to the latter. One of the specific goals of this article is to inform this debate by contrasting the morbidity associated with depression, as defined by these two paradigms. METHODS The MOS is an observational study of adults who receive care either in a group practice-style health maintenance organization, a large multispecialty group practice, or a single-specialty solo or amall group practice. The MOS was designed to asaris processes and outcomes of care for patients with hypertension, diabetes, coronary heart disease, or depression. This article uses baseline MOS data collected in 1986 from patients at the time of an office visit in these three systems of care in three US sites. For fuller details on the design of the MOS, see the article by Tarlov et al.' The three study sites are Los Angeles, Calif; Boston, Mass; and Chicago, Ill. At each site, one large health maintenance organization and several large multispecialty group practices were selected. A clinician sample was selected to represent the specialty groups providing the majority of care to patients with the four target conditions. The groups were internists, family practitioners, cardiologists, endocrinologists, diabetologists, psychiatrists, and psychologists. Clinicians eligible for participation in the MOS (1) were between the ages of 31 and 65 years; (2) were board eligible/certified or licensed for independent practice; and (3) had direct patient care as their primary professional activity. The MOS attempted to enroll all eligible clinicians in the participating health maintenance organizations and large multispecialty group practices. The health maintenance organizations and large multispecialty group prac tices included 266 eligible clinicians, of whom 225 (85%) agreed to participate in the MOS. In the single-specialty solo or small group practice sector, a multistage selection process was used. In the first stage, 2216 clinicians were initially selected by stratified random sampling from lists provided by national professional associations. Of these, 1525 (69%) were contacted. Through a series of two telephone interviews, 511 physicians were identified as eligible and agreed to a final selection interview. Of these, 298 (58%) agreed to participate in the main study. Among single-specialty solo or small group practice clinicians, participants and nonparticipants were similar in demographics and clinical training, but participants spent more time in direct patient care. Participants and nonJAMA, August 18, 1989-Vol 282, No. 7 participants at the final selection stage did not differ in experience with or attitudes toward depression, as assessed by a self-administered questionnaire. The total number of participating clinicians in all settings was 523. Each participating clinician was asked to invite all adult English-speaking patients visiting the practice over a specified period (9 days on average) to complete a self-administered questionnaire (the Patient Screener). Screening occurred between March 13 and October 31, 1986. Complete questionnaires were obtained for 74% of eligible patients in multispecialty group practices and 65% of patients in single-specialty solo or small group practices (N-22 462). To reduce respondent burden, two versions of the Patient Screener were fielded, each administered to a random 50% of patients. This article wes data from the random half of patients (n=11242) who completed a version that included a battery on wellbeing, functioning, and chronic medical conditions. Depressive Disorder and Symptoms To facilitate assessment of depressive disorders in such a large patient sample, a two-stage case-finding procedure was used. First, patients completed an eight-item depression symptom scale included in the Patient Screener. This scale, developed specifically for the MOS, elicits information on intensity of symptoms of depression (eg, feeling sad or crying spells) over the past week and on periods of depressed mood over the past year. Burnam et al," using data from two independent secondary data sources, developed a scoring algorithm for the eight items (each item was weighted by its coefficient in a logistic regression predicting probability of having depressive disorder) and identified a cutoff score (.06) that has excellent sensitivity and acceptable positive predictive value for identifying persons who have DSM-III depressive disor der. By depressive disorder we mean major depression alone, dysthymia alone, or the combination of both conditions, ie, "double depression." We use exceeding the cutoff as our indicator of current depressive symptoms. Second, patients who exceeded the cutoff score on the depression symptom scale and who were eligible for a longitudinal phase of the MOS were contacted for a follow-up telephone interview that included the depression section of the National Institute of Mental Health's Diagnostic Interview Schedule (DIS)." The DIS is a highly structured diagnostic interview that determines the presence of psychiatric disorders according to DSM-III criteria." Wells et al" demonstrated that telephone and face-toface administration of the DIS are equivalent in terms of identifying lifetime depressive disorders, and that the test-retest reliability of the DIS is acceptable for identifying a sample of depressed patients in a two-stage procedure. The longitudinal phase of the MOS was limited to patients with hypertension, diabetes, coronary heart disease, or depression who had an ongoing relationship with the MOS clinician, who could complete self-administered questionnaires, and who did not have an acute major physical condition that could severely and temporarily limit their functioning. Among the 11 242 patients in the sample, 2467 exceeded the cutoff on the depression symptom scale. Of the 1876 who were also eligible for the longitudinal study, 1137 (61%) completed the telephone DIS interview. Those who did not complete the telephone interview had 0.6 more years of education, on average, than noncompleters (t=2.7; df-1874; P<.01; two tailed), but there were no significant differences in mean age, sex, recentness of last health care visit, or mean depression screener score (f's range from 0.3 to 1.3; each P>.10). We control for education in the analyses reported herein. Thus, we do not think that nonresponse on the telephone interview poses a serious bias to our results. Those scoring above the cutoff score on the depression symptom scale were defined as having depressive symptoms. Current depressive disorder was defined as meeting all of the following criteria: (1) lifetime major depression or dysthymia, by DSM-III criteria (the only exclusion criterion was lifetime mania); (2) an episode of major depression or period of dysthymia during the last 12 months; and (3) no remission (ie, >2 months with two or fewer depressive symptoms) since the onset of the recent episode. Helzer et al found that the lay-administered DIS tended to underdetect true cases of major depression in persons who just missed meeting full DSM-III criteria, ie, those with clustering of symptoms in three rather than four Criteria B symptom groups of major depression. To enhance the sensitivity of the DIS, we modified our criteria for a recent episode of major depression to include persons with clustering in only three symptom groups. Only 43 patients entered the sample owing to this modification. Chronic Medical Conditions The definitions of eight chronic medical conditions are given in Table 1. Each Functioning of Depressed Patients-Welle et al 915 Table 1.-Definitions of Patient-Reported Chronic Conditions History of hypertension History of diabetes Current advanced coronary artery disease Current engine only Current arthritis Current back problems Curent lung problems Current gastrointestinal disorder Patient was told by a physician, nurse, or other health care professional that he or she has high blood pressure or hypertension. Patient was told by a physician, nurse, or other health care professional that he or she has a high blood sugar level or diabetes. Patient was told by a physician, nurse, or other health care professional that he or she has had a heart attack, myocardial infarction, or coronary and that it occurred within the last 12 months. Or the patient reports now having heart failure or an enlarged heart (with or without angina). Patient reports now having angina, but neither recent myocardial infarction nor heart Pesent reports now having arthritis. Patient reports now having back problems, including disk or spine problema. Patient reports now having ulcer (duodenal, stomach, or peptic) or chronic inflamed condition was assessed by one or two items in the Patient Screener. The items required a simple yes-or-no response. We also identified patients who had no chronic conditions. These patients had neither depressive symptoms nor any chronic medical condition defined in Table 1, nor any other chronic problem assessed by the Patient Screener: cancer in the last 3 years; major neurological problems; cardiac pacemaker, limb amputation; kidney disease; legal blindness; or other problem. Sociodemographic Factors and Clinical Specialty The Patient Screener elicited data on age, sex, education, and income. The self-designated medical specialty of the treating clinician was determined from data from professional societies and telephone interviews of the MOS clinicians. surance Experiment." All measures except days in bed are scored from 0 to 100, with 100 representing perfect health on that construct. The "bed days" measure is the number of days spent in bed in the past 30 days. Statistical Analysis We used the general linear model to conduct analyses of covariance to compare the functioning of patients with current depressive disorder (the "disorder" group), those with depressive symptoms in the absence of depressive disorder (the "symptoms unly" group), and those with no chronic conditions. The covariates were age, sex, education, specialty (.nental health specialist vs medical clinician), presence or absence of each chronic condition, interac tions between specialty and each chronic medical condition, and interactions between specialty and depression group status (current disorder, symptoms only, or no chronic conditions) For these analyses, we excluded the 1330 persons with depressive symptoms who were ineligible or refused the telephone interview, because we did not know if they had depressive disorder. Least-squares multiple linear regression was used to estimate the unique associations of each functional status and well-being variable with depressive symptoms (with or without depressive disorder) and with each chronic medical condition. These models used the full sample. The independent variables were sociodemographic characteristics (age, sex, education, and income), specialty, each chronic medical condition, and depressive symptoms (exceeding the screener cutoff). The SEs and inference statistics presented herein are unadjusted for the clustered sampling design. The intraclass (within-provider) correlations in patient functioning were quite small, ie, .03 to 12, and thus would have minimal effects on the results. In a sensitivity analysis, we determined that the cor rection was, in fact, trivial. Because we report many significant findings in a consistent direction, a Bonferroni correction" for multiple comparisons is too conservative. We comment on results significant at the P<.05 level or better, but also present the actual level of significance (P<.05, .01, .006, 001 0.00011 We determined that our main conclusions would not be affected by a strict Bonferroni correction, which would impose a significance level between .001 and .005 for most comparisons. RESULTS Sample Characteristica On average, patients of general medical providers are 48 years old, while patients of mental health specialists average 40 years of age (P<.05). As shown in Table 3, compared with patients of general medical providers, patients of mental health specialists are more likely to be white, more likely to be highly educated, less likely to be married, and more likely to be female. Patients of general medical providers have a relatively higher prevalence of five out of eight chronic medical conditions and a lower prevalence of back problems. Well-being and Functioning of Table 4 presents the adjusted mean well-being and functioning of patients in each specialty group who have depressive disorder, depressive symptoms only, and no chronic conditions, based on the analyses of covariance. Out of the 24 possible pairwise comparisons (types 1 through 4 in Table 4) between the two depressed groups (one at a time) and the group with no chronic conditions, the depressed group has worse functioning in 17 comparisons and significantly better functioning in 1 comparison. We compared patients with depressive disorder vs those with depressive Functioning of Depressed Paberts-Walls at al symptoms only, for each type of provider sector (comparison types 5 through 6 in Table 4). Among patients of medical clinicians, those with depressive disorder have worse current health than those with symptoms only. Among patients of mental health specialists, those with depressive disorder have worse physical and social functioning and perceived current health and more bed days and bodily pain than those with depressive symptoms but no disorder. We also contrasted depressed patients treated by the two provider sectors (comparison types 7 through 8 in Table 4). Among patients with depressive disorder, those treated by mental health specialists have poorer social functioning than those treated by gen Table 3.-Patient Characteristics by Provider Group* eral medical clinicians; but among patients with depressive symptoms but no disorder, patients of general medical providers have worse physical functioning and more bodily pain and days in bed. Depressive Symptoms Compared We estimated the unique associations of each functioning indicator with any depressive symptoms (with and without depressive disorder), relative to comparable associations with specific chronic medical conditions. Table 5 presents the mean score on each functioning indicator, estimated from the multiple linear regression models and adjusted for all covariates in the models. As shown in Table' 5, the physical functioning of patients with depressive symptoms, other factors being equal, is significantly worse than that of patients with four of the chronic medical conditions; it is significantly better than that of patients with current advanced coronary artery disease and current angina only. Patients with depressive symptoms have significantly worse social functioning than patients with each of the eight chronic medical conditions. Role functioning is significantly worse for patients with depressive symptoms than for patients with six of the chronic medical conditions; but it is significantly worse for patients with current advanced coronary artery disease than for those with depressive symptoms. Patients with depressive symptoms have significantly more recent days in bed, other factors being equal, than do patients with six of the chronic conditions; but they have fewer days in bed than patients with current advanced coronary artery disease. Current health is significantly, worse for the depressive symptom group than for those with each of the conditions except the two heart conditions. Patients with depressive symptoms have significantly more bodi-. ly pain than do patients with five of the medical conditions, including angina only. They have significantly less pain, other factors being equal, than do patients with current arthritis. We found no significant interactions between depressive symptoms and any of the chronic medical conditions, across all dimensions of well-being and funetioning. Thus, depressive symptoms Table 4.-Aqusted Mean Well-being and Functioning of Patients With No Chronic Conditions, Depressive Disorder, and Depressive Symptoms Alone, by Specialty *Adjusted for sociodemographics, presence of chronic medical conditions, and interactions of speciality with each medical condition and with an indicator of depression group statue (no chronic conasons, depressive disorder or depreserve symptoms alone) For (+) higher score indicates better functioning and well-being; for (-) higher score indicases poorer functioning and well-being NCC indicates no chronic conditions, DIS, depressive disorder. SYMP. depreserve symptoms alone; and NS, not significant. The companson types are as follows 1, medical sector compenson of DIS and NCC groups: 2, mental health sector compenson of DIS and NCC groups: 3, medical sector companion of SYMP and NCC groups, 4, mental health sector comparison of SYMP and NCC groups: 5. medical sector, compenson of DIS and SYMP groups: 6, mental neann sector, companson of DIS and SYMP groups; 7, among DIS patents, comparison of medical and mental health sectors; and 8, among SYMP patients, comparison of medical and mental heath sectors. JAMA, August 18, 1989-Vol 262, No. 7 Functioning of Depressed Patients-Wells et al 917 |