Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Randomized Trial of a Depression Management Program in High Utilizers of Medical Care

Randomized Trial of a Depression Management Program in High Utilizers of Medical Care BackgroundHigh utilizers of nonpsychiatric health care services have disproportionally high rates of undiagnosed or undertreated depression.ObjectiveTo determine the impact of offering a systematic primary care–based depression treatment program to depressed "high utilizers" not in active treatment.DesignRandomized clinical trial.SettingOne hundred sixty-three primary care practices in 3 health maintenance organizations located in different geographic regions of the United States.PatientsA group of 1465 health maintenance organization members were identified as depressed high utilizers using a 2-stage telephone screening process. Eligibility criteria were met by 410 patients and 407 agreed to enroll: 218 in the depression management program (DMP) practices and 189 in the usual care (UC) group.InterventionThe DMP included patient education materials, physician education programs, telephone-based treatment coordination, and antidepressant pharmacotherapy initiated and managed by patients' primary care physicians.Main Outcome MeasuresDepression severity was measured using the Hamilton Depression Rating Scale (Ham-D) and functional status using the Medical Outcomes Study 20-item short form (SF-20) subscales. Outpatient visit and hospitalization rates were measured using the health plan's encounter data.ResultsBased on an intent-to-treat analysis, at least 3 antidepressant prescriptions were filled in the first 6 months by 151 (69.3%) of 218 of DMP patients vs 35 (18.5%) of 189 in UC (P<.001). Improvements in Ham-D scores were significantly greater in the intervention group at 6 weeks (P=.04), 3 months (P=.02), 6 months (P<.001), and 12 months (P<.001). At 12 months, DMP intervention patients were more improved than UC patients on the mental health, social functioning, and general health perceptions scales of the SF-20 (P<.05 for all).ConclusionIn depressed high utilizers not already in active treatment, a systematic primary care–based treatment program can substantially increase adequate antidepressant treatment, decrease depression severity, and improve general health status compared with usual care.DEPRESSION among primary care patients is associated with significant functional impairment, lost productivity, and use of health services. Epidemiologic studies consistently demonstrate that approximately 10% of primary care patients suffer from significant depressive disorders.Depressed primary care patients report poorer quality of life,greater impairment of daily functioning,and more disability and/or lost productivity because of illness.Use of general medical services by depressed primary care patients is 50% to 100% higher than utilization by similar patients without depressive illness.The economic burden of depression in primary care (in terms of increased health care utilization and lost productivity) vastly exceeds the resources currently devoted to treatment.Among primary care patients with clinically significant depression, as many as half go unrecognized.Fewer than half of patients with recognized depression receive treatment of proven efficacyand of those who do obtain treatment, half or fewer receive levels of treatment consistent with expert guidelines.These high rates of underrecognition and undertreatment are especially concerning given the strong evidence for the efficacy of both pharmacotherapy and depression-specific psychotherapy.A series of recent randomized trials have demonstrated that systematic efforts to provide high-quality depression treatment can significantly improve clinical outcomes.These studies show that depression treatment effectiveness can be increased by using careful monitoring of patients and active involvement of specially trained clinicians.The need for improved management of depression is especially clear among "high utilizers" of general medical services. High utilizers of health care services have a high prevalence of depressive disorders,with considerable associated disabilityand excessive use of general medical services.High utilizers who are depressed are also more likely to continue to be high utilizersand often go either untreated or undertreated.Katon et alhave reported a randomized trial of a program to increase physician recognition of psychiatric disorders among high utilizers of outpatient care. This program (including feedback of psychiatric diagnosis, treatment recommendations to primary care physicians, and physician education) had no significant effects on treatment received or patient outcomes.Mindful of that negative result, we have developed a program that follows-up depression screening and physician education with a systematic treatment program to ensure depression treatment of sufficient intensity. This depression management program (DMP) includes many of the features used in the successful consultant/liaisondepression treatment model but is based with the patient's own primary care physician. A small pilot study using this program resulted in improved clinical outcomes and a reduction in total health care utilization.This article describes a randomized trial of a program to identify and treat depression among high utilizers of general medical care. Our intention was to develop and test a population-based screening program and a primary care–based treatment program that could be implemented in most managed health care systems. Results of the screening program (prevalence of depression, association of depression with functional impairment, and increased utilization) are described in a separate publication.This article uses data from the randomized trial to examine the impact of the DMP on treatment received, clinical outcomes, functional status, and health care utilization.SUBJECTS AND METHODSThe study was conducted within selected primary care clinics of 3 large prepaid health plans representing 3 geographic regions (Midwest, Northwest, and New England) in the United States: DeanHealth Plan (DHP), Wisconsin; Group Health Cooperative of Puget Sound (GHC), Washington; and Harvard Pilgrim Health Care (HPHC), Massachusetts. Both HPHC and GHC are not-for-profit health maintenance organizations (HMOs); DHP is a for-profit HMO. DeanHealth Plan has 175 000 members, GHC has 450 000 members, and the staff model division of HPHC has 300 000 members. The study protocol was approved by the institutional review boards of all 3 HMOs. Computerized databases from participating physician practices in all 3 HMOs were used to identify health plan members aged 25 to 63 years with continuous health plan enrollment for the previous 2 years. Members in this sample with ambulatory visit counts above the 85th percentile for both of the 2 previous years were considered high utilizers. For screening purposes, ambulatory office visits were defined as primary care visits, medical specialty visits, and walk-in clinic visits. Ambulatory office visits did not include emergency department visits, all mental health provider visits, routine obstetric visits, physical therapy/occupational therapy visits, optometry visits, x-ray/diagnostic visits, home health visits, telephone contacts, chiropractic visits, and allergy injections. Computerized record data were then used to assess the following exclusion criteria: recent treatment for alcohol or other substance abuse; past treatment for schizophrenia or bipolar disorder; life-threatening medical disorders (eg, metastatic malignant neoplasm); or active treatment for depression. Active treatment was defined as current specialty mental health treatment or minimal adequate trial of antidepressant medication. Minimal adequate trial was defined as an antidepressant prescription(s) filled in the past 90 days with at least 30 days of an adequate dose based on American Psychiatric Association guidelines.We used a 2-stage telephone-screening process to identify those high utilizers likely to benefit from depression treatment.The first screening call included a modified version of the Structured Clinical Interview for DSM-IV(SCID)to identify 1465 enrollees who screened positive for either current major depression or major depression in partial remission. Of these, 1295 enrollees agreed to complete a second telephone interview to assess depression severity using the 17-item Hamilton Depression Rating Scale (Ham-D).Potential study subjects were excluded if they had a Ham-D score less than 15 or met any of the following conditions: contraindications to taking an antidepressant, receiving treatment by a psychiatrist within the past 4 months, pregnancy, planned pregnancy within the next year, breastfeeding, positive screen for alcohol abuse,and intent to disenroll from the HMO.The 163 participating physician practices were randomly assigned to either the DMP (n=82) or usual care (UC) (n=81) groups. While randomization of physicians occurred prior to recruitment, study personnel at each site were not informed of an individual patient's assignment until that patient had completed the enrollment process. Eligible and consenting patients from the practices of UC physicians were informed that telephone screening suggested depression and were advised that care was available with their primary care physician. Patients from the practices of DMP physicians were invited to participate in the protocol described below. All analyses classified patients according to the original assignment of their primary care physicians, regardless of the treatment actually received.The principal elements of the DMP were physician education, patient education, antidepressant treatment, and treatment coordination. Prior to patient enrollment, all physicians in the DMP group participated in a standardized 2-hour training program focused on the initial assessment of depression and the initiation of pharmacotherapy. At each health plan, 1 or 2 psychiatrists were identified as consultants for the DMP. At the time of enrollment, all study patients in practices of the DMP group physicians were asked to schedule an evaluation visit with their primary care physician. Prior to this visit, patients received a booklet created for the study "Depression Isn't Just a Mental Problem"and videotaped educational materials from the treatment coordinator designed to increase acceptance of depression treatment. At the initial visit, primary care physicians confirmed the diagnosis of depression, assessed contraindications to pharmacotherapy, and (if indicated) recommended antidepressant treatment. Patients who had psychotic symptoms, mania, or acute suicidality were immediately referred to psychiatrists.Primary care physicians were advised to follow a specific pharmacotherapy algorithm, but were allowed to adjust treatment according to individual clinical need. The treatment algorithm recommended that patients who had previously been successfully treated and tolerated an antidepressant be given the previous antidepressant. All others were to be given sertraline hydrochloride at a recommended starting dose of 50 mg/d. It was also suggested that patients with panic attacks or who weighed less than 45 kg be given 25 mg/d. The dose was to be increased at 4 weeks if the patient had not improved, with a maximum dose of 200 mg/d. Patients who did not respond to or tolerate sertraline were to be treated with nortriptyline, if clinically appropriate. Nortriptyline was to be started at 25 mg/d, titrating up to 100 mg/d if needed. It was strongly recommended that psychiatric consultation be initiated if the preceding steps were unsuccessful. Follow-up visits with the primary care physician were prescheduled at approximately 1, 3, 6, and 10 weeks. Subsequent visits occurred approximately every 10 weeks. The DMP patients were also offered the RHYTHMS®depression education program at the first physician visit if appropriate. The program provided information on the nature of depression, the use of antidepressant medications, and other related topics aimed at increasing patient antidepressant adherence.Primary care treatment was supported by an ongoing program of monitoring, feedback, and as-needed specialty consultation. Coordinators reviewed DMP patient prescription refills and office visits to identify unplanned treatment discontinuation. In addition, treatment coordinators contacted DMP patients for telephone monitoring of treatment adherence; treatment response (based on administration of the 17-item Ham-D); and medication adverse effects after 2 and 10 weeks (and, if needed, after 18, 30, and 42 weeks). Treatment coordinators did not have access to the Ham-D scores generated by the blinded raters. The coordinators all had bachelor's or master's degrees and at least some clinical mental health experience. Data from this monitoring program, as well as recommendations for adjustments in treatment, were provided to treating primary care physicians after each monitoring contact. A written response was generated if the subject was doing well and a call to the physician was placed if the subject was not doing well. The study physicians did not have to acknowledge receipt of written messages. Study psychiatrists had ongoing contact with all intervention group primary care physicians via periodic case reviews and as-needed telephone consultation. A psychiatric consultation visit was strongly encouraged for all DMP patients not responding to treatment by 10 weeks and for patients with more complicated depression (eg, major psychiatric comorbidity, past treatment failures). While UC patients could self-refer to any specialty services normally available to health plan members, no additional monitoring, case management, or psychiatric liaison services were provided.All patients in both the DMP and UC groups were contacted for blinded telephone assessments 6 weeks, 3 months, 6 months, and 12 months after randomization. These assessments included a second administration of the Ham-D, a patient-rated 7-point scale of overall improvement since enrollment (patient global improvement score),and the Medical Outcomes Study 20-item short form (SF-20)functional status questionnaire. Health care utilization was collected from the HMOs' automated claims databases for 1 year before and 1 year after study randomization. Outpatient visits were defined as unique patient, provider, and date of services for all clinic or outpatient hospital encounters, excluding visits to departments of radiology, pathology, and anesthesiology. Specialty mental health outpatient visits are defined as those visits provided by a mental health specialist and include visits to the psychiatrists involved in the study. Hospital admissions are defined as a stay in the hospital requiring at least 1 night's stay, with stays not separated by at least 1 outpatient day considered a single admission. Only people who had a full year of health plan coverage were included in the analysis of outpatient visits and inpatient stays.Participants provided informed consent to participate prior to the initial telephone screening, prior to the second-stage screening, and prior to enrollment in the randomized trial. The DMP patients provided additional informed consent to participate in the treatment program. An independent data safety committee monitored unblinded patient outcomes and adverse events throughout the study.Primary comparisons of the DMP and UC groups were based on change scores (follow-up minus baseline scores) using an intent-to-treat approach. Estimation was done by analysis of variance with robust or sandwich estimates for the variance/covariance matrix corrected for intracluster correlation at the patient level.We also obtained very similar estimates controlling for site, age, sex, and baseline mental and physical health status, using a random-effects generalized least-squares estimator, generalized estimating equations for the health status measures, and negative binomial regression for outpatient visits and inpatient stays (not shown). For simplicity, we report the analysis of variance results on change scores, with robust estimates of the inference statistics and corrected for intracluster correlation among patients of the same physician.RESULTSOf the 1295 members who completed the second screening call, 410 (32%) had Ham-D scores of 15 or more and were eligible to participate. Almost all (n=407) agreed to participate. Based on randomization of treating physicians, 218 were assigned to DMP and 189 to UC (with the imbalance reflecting unequal assignment of full-time and part-time physicians to the 2 groups). Demographic and health data were similar for the 2 groups (Table 1), with the exceptions of ethnicity and recent inadequate antidepressant use, which are commented on below. The overall F test for differences in the means for all of the variables in Table 1for DMP vs UC demographics was not significant (P=.13). Differences between DMP and UC in ethnic distribution were not significant after controlling for study site. There were no significant or appreciable differences in mental health status between the DMP and UC (P=.80). The mean baseline Ham-D score was 19.2 for the DMP group and 19.1 for UC. Medical comorbidity was common: 46.7% of randomized patients had at least 1 chronic medical disorder and 14.5% had 2 or more.The number of study patients who completed the 12-month blinded assessment call was 380: 203 (93.1%) of 218 in DMP vs 177 (93.7%) of 189 in UC (P=.83) (Figure 1). Sample loss was not significantly related to DMP vs UC over time, to baseline Ham-D scores, or to other health, site, or demographic characteristics.Demographic and Patient Characteristics*DMP (n=218)UC (n=189)P†Age, y, mean (95% CI)45.6 (44.0-47.1)45.4 (43.9-46.9).89Female, % (95% CI)76.6 (70.9-82.4)78.3 (71.7-84.9).69White, % (95% CI)87.6 (82.4-92.8)77.3 (70.5-84.0).02Married, % (95% CI)71.1 (64.6-77.6)65.6 (58.8-72.5).25Years of education, mean (95% CI)13.6 (13.4-13.9)13.8 (13.4-14.1).52Depression status, % (95% CI)Current64.7 (59.4-70.0)63.5 (56.4-70.6).79Partial remission35.3 (30.0-40.6)36.5 (29.4-43.6).79SF-20 subscales, mean (95% CI)Ham-D19.1 (18.7-19.6)19.2 (18.7-19.7).75MHI-547.4 (44,7-50.0)48.5 (45.7-51.4).56GHP37.4 (34.2-40.6)38.6 (34.8-42.4).64PF54.6 (50.2-58.9)50.4 (45.6-55.1).20Patient seeing mental health provider (last 2 y), % (95% CI)33.9 (27.8-40.1)34.4 (27.2-41.6).93Inadequate trial of antidepressant (last 90 d), % (95% CI)24.3 (18.5-30.1)13.8 (7.7-19.8).01*DMP indicates depression management program; UC, usual care; SF-20, Medical Outcomes Study 20-item short form; Ham-D, Hamilton Depression Rating Scale; MHI-5, mental health inventory 5; GHP, general health perception; CI, confidence interval; and PF, physical functioning.†Pvalues for individual variables are unadjusted for multiple comparisons. An overall F test of DMP vs UC for all variables was not significant (P=.13).Figure 1.A flowchart mapping randomization of patients participating in the trial to either usual care (UC) or the depression management program (DMP).Of the 218 patients in the DMP, 198 scheduled and 189 completed the first physician visit and 166 started the RHYTHMS program. A treatment coordinator telephone contact was made at 2 weeks (177 patients), 10 weeks (194 patients), 18 weeks (131 patients), and 30 weeks (90 patients), for an average of 2.7 contacts per DMP patient. Computerized administrative databases were examined to determine the treatment actually received by DMP and UC patients. The DMP patients were much more likely to receive adequate antidepressant treatment. Pharmacy refill data for the 6-month period after enrollment indicated that 179 (82.1%) of 218 DMP patients and 61 (32.3%) of 189 UC patients filled at least 1 antidepressant prescription (P<.001); 151 (69.3%) of 218 DMP patients and 35 (18.5%) of 189 UC patients had filled at least 3 antidepressant prescriptions (P<.001) (Figure 2). The mental health treatment provided in both groups was given largely in primary care. At least 1 specialty mental health visit was made by only 59 (27.1%) of 218 DMP patients and 32 (16.9%) of 189 UC patients (P=.03). Only 29 (13.3%) of 218 DMP patients vs 18 (9.5%) of 189 UC patients made 3 or more specialty mental health visits in the first 6 months (P=.29).Figure 2.Patterns of antidepressent treatment for patients who received either the depression management program (DMP) or usual care (UC).Patients assigned to the intervention group experienced significantly better outcomes at every follow-up assessment. Improvements in Ham-D scores were significantly greater in the DMP than in the UC group: −3.3 vs −2.0 at 6 weeks (P=.04); −5.6 vs −3.9 at 3 months (P=.02); −7.3 vs −4.0 at 6 months (P<.001); and −9.2 vs −5.6 at 12 months (P<.001). The difference between DMP and UC groups increased significantly over time (P=.005) (Figure 3). At 12 months, 108 (53.2%) of 203 DMP patients had responded (defined as at least 50% reduction in Ham-D scores from baseline) compared with 58 (32.8%) of 177 UC patients (P<.001), and 92 (45.3%) of 203 DMP patients were in remission (defined as a Ham-D score <7) compared with 49 (27.7%) of 177 UC patients (P<.001). Forty-one (23.2%) of 177 UC patients vs 26 (12.8%) of 203 DMP patients reported an increase in Ham-D scores (an increase indicates worse health) from baseline to 12-month assessment (P=.01). One hundred seventeen (57.6%) of 203 DMP patients rated themselves as "much" or "very much" improved at 12 months compared with 60 (33.7%) of 178 UC patients (P<.001), concordant with the Ham-D results. At 12 months, DMP patients reported significantly better social functioning, mental health, and general health perceptions than UC patients on the SF-20 (P<.05 for all), but not on physical functioning, role functioning, and pain perception (Figure 4). Given the baseline imbalance in the proportion of patients receiving recent inadequate antidepressant treatment, the above analyses were repeated, excluding all patients receiving inadequate antidepressant treatment during the 90 days prior to enrollment. Results were nearly identical.Figure 3.Patients' scores on the Hamilton Depression Rating Scale (Ham-D) during the 12-month study. The Pvalues were based on the change in scores since the baseline assessment.Figure 4.Change in patients' scores on the Medical Outcomes Study 20-item short form (SF-20) subscales after receiving either the depression management program (DMP) or usual care (UC) for 12 months. The Pvalue was based on the change in scores since the baseline assessment.Comparing the year before randomization with the year after randomization, mean visit counts in DMP increased by 1.6 visits (18.4 to 19.9 visits), 0.5 of which was attributable to a single case. Mean visit counts in UC decreased by 2.0 visits (19.4 to 17.4 visits). The change in visits did differ significantly between groups (P=.02). Outpatient visits included nonphysician encounters, such as physical therapy and speech therapy. Comparing the year before randomization with the year after randomization, mean inpatient admissions in the DMP group increased by 0.04 (0.23 to 0.27 admissions) and mean inpatient admissions in the UC group decreased by 0.08 (0.26 to 0.18 admissions). The change in inpatient admissions between groups was not significantly different (P=.09).COMMENTDepressed high-utilizer patients not already in active treament who received the DMP had significantly better clinical outcomes and improved general health status compared with patients who received UC. The effectiveness of the DMP in this real world clinical setting is notable, given that some DMP patients did not pursue treatment and some UC patients received adequate treatment through the usual channels. The results show that the DMP did not decrease outpatient visits compared with UC and our estimate is actually a relative increase in outpatient visits. While the study protocol recommended 9 or more outpatient visits for initiating and monitoring depression treatment, the DMP group increased by only 1.6 visits, while the UC group decreased by 2.0 visits, suggesting that most visits for depression could be combined with other visits.The following limitations should be considered when generalizing the results of this study. First, the UC patients may have been more likely to seek depression treatment than usual because the telephone screeners recommended that they contact their physicians. At the exit interview, 74 (41.6%) of 178 UC patients reported they took action in some way based on this recommendation. The low rate of specialty mental health visits and prescription of antidepressants in the UC group suggests that this had a limited effect. Because of this limitation, the study results probably represent a lower bound of the impact of the intervention. Second, the initial imbalance of more DMP patients than UC patients having an inadequate antidepressant trial prior to entering the study could have affected the results. This seems unlikely, as the results did not change when randomized patients on inadequate antidepressant trial at baseline were excluded from the analyses. Third, because the study was not powered to detect a change in inpatient admissions, we cannot determine whether the intervention affected inpatient utilization. Fourth, the 1-year study period may have been too short to capture the full benefit of the DMP intervention. The difference in Ham-D scores between the DMP and UC groups was greatest at the 12-month assessment point. Fifth, care must be taken in generalizing these results to depressed populations other than high utilizers.How substantial is the effect of the DMP intervention? One way to assess the impact of the DMP is to consider the number of patients with depression who would need treatment in the DMP group to achieve 1 more remission than would occur in the UC group. Based on our data, this number needed to treat was 5:1. This ratio compares favorably with the 9:1 number of patients with diastolic blood pressure of 115 to 129 mm Hg needed to treat to achieve 1 less stroke, myocardial infarction, or death.Another way to assess the DMP is to compare it with efficacy studies of antidepressants vs placebo. The differences of 17.6% in remission rates and 3.6 in Ham-D scores between the DMP and UC groups in this study is similar to the difference seen between efficacious antidepressant and placebo in clinical trials.While the difference between the DMP and UC patients in mean Ham-D scores is similar to that between active treatment and placebo groups in other primary careand specialtystudies, rates of response and remission in both groups seem low. This study used a 2-stage screening process that required a diagnosis of depression at the first call and a Ham-D severity greater than or equal to 15 at the second call 2 weeks later. Requiring participants to have depressive symptoms that persisted over time may have decreased the number of patients likely to spontaneously recover.The DMP patients on average took longer to respond to treatment than has been seen in clinical trials.One reason is that time zero was defined as the date of study enrollment, not the first physician visit, which was at least 2 weeks later. Another reason may be that high utilizer patients have more medical and psychiatric comorbidity, which has been shown to be associated with delayed onset of treatment response and lower response rates.Most DMP patients (69.3%) received at least adequate antidepressant treatment (defined here as 3 or more antidepressant prescriptions) during the first 6 months of the study, even though the large majority of DMP patients (72.9%) received no specialty mental health treatment. This is consistent with the intervention's goal of providing primary care–based treatment for the majority of depressed patients and increasing the intensity of specialty care for nonresponding patients.Our randomized trial of a multicomponent intervention does not allow specific conclusions regarding the benefits of individual intervention components. However, considering our findings along with those of other primary care intervention studies suggests some general conclusions. Several programs limited to physician training or depression screening have not improved patient outcomes.Common elements of effective programsinclude guideline-based treatment, organized follow-up care, and active monitoring of patient outcomes.Our findings suggest that real-world primary care physicians can deliver effective depression treatment, if those physicians are supported by organized systems for patient education, proactive psychiatric consultation for nonresponders, physician education, and patient monitoring. Previous studies of programs that focused solely on increasing recognition of depression have not demonstrated differences in patient outcomes.The DMP tested in this study was not as tightly controlled as previous successful primary care interventions, but it was more intensive than several earlier unsuccessful programs. The successful programs described by Katon et aland Schulberg et alall involved direct treatment by specially trained study personnel. In the study by Katon et al,mental health specialists participated in the care of all patients. Schulberg et alused 2 different systematic depression treatment programs for patients in primary care that were more effective than UC. The major difference between the DMP used in this study and the successful Katon and Schulberg models is that patient's depression treatment in the DMP was based with their own primary care physician.The DMP used in this study could be a model for use with high utilization patients in many health care organizations. The DMP treatment model has the advantage of being initiated by the patient's own primary care physician. In addition, it encourages integration of medical and psychiatric treatment. Another substantial advantage is that patient visit and pharmacy tracking programs already present in many existing managed care organizations will facilitate the implementation of the DMP. Further research will be needed to determine the effectiveness of this systematic treatment model in other depressed populations.RLSpitzerJBWilliamsKKroenkeUtility of a new procedure for diagnosing mental disorders in primary care: the PRIME-MD 1000 study.JAMA.1994;272:1749-1756.TUstunNSartoriusMental Illness in General Health Care: An International Study.London, England: John Wiley & Sons Inc; 1995.RLSpitzerKKroenkeMLinzerHealth-related quality of life in primary care patients with mental disorders: results from the PRIME-MD 1000 Study.JAMA.1995;274:1511-1517.JLCoulehanHCSchulbergMRBlockMJMadoniaERodriguezTreating depressed primary care patients improves their physical, mental, and social functioning.Arch Intern Med.1997;157:1113-1120.JOrmelMVonKorffTBUstunSPiniAKortenTOldehinkelCommon mental disorders and disability across cultures: results from the WHO Collaborative Study on Psychological Problems in General Health Care.JAMA.1994;272:1741-1748.GSimonJOrmelMVonKorffWBarlowHealth care costs associated with depressive and anxiety disorders in primary care.Am J Psychiatry.1995;152:352-357.JUnutzerDLPatrickGSimonDepressive symptoms and the cost of health services in HMO patients aged 65 years and older: a 4-year prospective study.JAMA.1997;277:1618-1623.GESimonMVon KorffWBarlowHealth care costs of primary care patients with recognized depression.Arch Gen Psychiatry.1995;52:850-856.HCSchulbergMRBlockMJMadoniaThe "usual care" of major depression in primary care practice.Arch Fam Med.1997;6:334-339.GESimonMVonKorffRecognition, management, and outcomes of depression in primary care.Arch Fam Med.1995;4:99-105.KBWellsWKatonBRogersPCampUse of minor tranquilizers and antidepressant medication by depressed outpatients: results from the medical outcomes study.Am J Psychiatry.1994;151:694-700.GESimonMVonKorffEHWagnerWBarlowPatterns of antidepressant use in community practice.Gen Hosp Psychiatry.1993;15:399-408.DJKatzelnickKAKobakJHGreistJWJeffersonHJHenkEffect of primary care treatment of depression on service use by patients with high medical expenditues.Psychiatr Serv.1997;48:59-64.HCSchulbergMRBlockMLMadoniaTreating major depression in primary care practice: eight-month clinical outcomes.Arch Gen Psychiatry.1996;53:913-919.ATBeckSDHollonJEYoungRCBedrosianDBudenzTreatment of depression with cognitive therapy and amitriptyline.Arch Gen Psychiatry.1985;42:142-148.GKlermanMWeissmanInterpersonal psychotherapy (IPT) and drugs in the treatment of depression.Pharmacopsychiatry.1987;20:3-7.US Dept of Health and Human ServicesDepression in Primary Care: Treatment of Major Depression (Clinical Practice Guideline).Rockville, MD: US Dept of Health and Human Services; 1993.WKatonMVon KorffELinCollaborative management to achieve treatment guidelines: impact on depression in primary care.JAMA.1995;273:1026-1031.WKatonPRobinsonMVon KorffA multifaceted intervention to improve treatment of depression in primary care.Arch Gen Psychiatry.1996;53:924-932.LMMynors-WallisDHGathARLloyd-ThomasDTomlinsonRandomised controlled trial comparing problem solving treatment with amitriptyline and placebo for major depression in primary care.BMJ.1995;319:441-445.WKatonMVon KorffELinDistressed high utilizers of medical care: DSM-III-Rdiagnoses and treatment needs.Gen Hosp Psychiatry.1990;12:355-362.MVon KorffJOrmelWKatonEHLinDisability and depression among high utilizers of health care: a longitudinal analysis.Arch Gen Psychiatry.1992;49:91-100.HJHenkDJKatzelnickKAKobakJHGreistJWJeffersonMedical costs attributed to depression among patients with a history of high medical expenses in a health maintenance organization.Arch Gen Psychiatry.1996;53:899-904.WKatonMVon KorffELinTBushJOrmelAdequacy and duration of antidepressant treatment in primary care.Med Care.1992;30:67-76.WKatonMVon KorffELinA randomized trial of psychiatric consultation with distressed high utilizers.Gen Hosp Psychiatry.1992;14:86-98.SDPearsonDJKatzelnickGESimonWGManningCPHelstadHJHenkDepression among high utilizers of medical care.J Gen Intern Med.1999;14:461-468.American Psychiatric AssociationPractice guideline for major depressive disorder in adults.Am J Psychiatry.1993;150(suppl 4):S1-S26.MFirstRSpitzerMGibbonJBWilliamsStructured Clinical Interview for DSM-IV Axis I Disorders-Patient Edition (SCID-I/P, Version 2.0).New York: New York State Psychiatric Institute; 1995.MHamiltonA rating scale for depression.J Neurol Neurosurg Psychiatry.1960;23:56-62.RLBrownLARoundsConjoint screening questionnaires for alcohol and other drug abuse: criterion validity in a primary care practice.WMJ.1995;94:135-148.DJKatzelnickGESimonSDPearsonDepression Isn't Just a Mental Problem[available on request from authors&rsqb;.1995.Not AvailableThe RHYTHMS Patient Compliance Program.New York, NY: Carlson HealthCare Publications Inc;1995.WGuyECDEU: Assessment for Psychopharmacology, Revised.Rockville, Md: National Institute of Mental Health; 1976. US Dept of Health, Education, and Welfare publication ADM 76-338.ALStewartRDHaysJEWareThe MOS Short-Form General Health Survey: reliability and validity in a patient population.Med Care.1988;26:724-735.PHuberThe Behavior of Maximum Livelihood Estimates Under Non-Standard Conditions: Fifth Berkeley Symposium in Mathematical Statistics and Probability.Berkeley: University of California Press; 1967.HWhiteA heteroscedasticity: consistant covariance matrix estimator and a direct test for heteroscedasticity.Econometrica.1980;48:817-830.ALaupacisDLSackettRSRobertsAn assessment of clinically useful measures of the consequences of treatment.N Engl J Med.1988;318:1728-1733.LFFabreFSAbuzzahabMAminSertraline safety and efficacy in major depression: a double-blind fixed-dose comparison with placebo.Biol Psychiatry.1995;38:592-602.JFeighnerA double-blind comparison of paroxetine, imipramine and placebo in depressed outpatients.Int Clin Psychopharmacology.1992;6:31-35.JEWernickeSRDunlopBEDornseifRLZerbeFixed-dose fluoxetine therapy for depression.Psychopharmacol Bull.1987;23:164-168.SKasperClinical efficacy of mirtazapine: a review of meta-analyses of pooled data.Int Clin Psychopharmacol.1995;10:25-35.JLClaghornCQCraigDDWalczakFluvoxamine maleate in the treatment of depression: a single-center, double-blind, placebo-controlled comparison with imipramine in outpatients.J Clin Psychopharmacol.1996;16:113-120.IElkinMTSheaJTWatkinsNational Intsitute of Mental Health Treatment of Depression Collaborative Research Program: general effectiveness of treatments.Arch Gen Psychiatry.1989;46:971-982.HMullerHJMollerMethodological problems in the estimation of the onset of the antidepressant effect.J Affect Disord.1998;48:15-23.CBrownHCSchulbergMJMadoniaMKShearPRHouckTreatment outcomes for primary care patients with major depression and lifetime anxiety disorders.Am J Psychiatry.1996;153:1293-1300.GIKeitnerCERyanIWMillerRKohnNBEpstein12-month outcome of patients with major depression and comorbid psychiatric or medical illness (compound depression).Am J Psychiatry.1991;148:345-350.CMCallahanHCHendrieRSDittusDCBraterSLHuiWMTierneyImproving treatment of late life depression in primary care: a randomized clinical trial.J Am Geriatr Soc.1994;42:839-846.CDowrickIBuchananTwelve month outcome of depression in general practice: does detection or disclosure make a difference?BMJ.1995;311:1274-1276.HIGoldbergEHWagnerSDFihnA randomized controlled trial of CQI teams and academic detailing: can they alter compliance with guidelines?Jt Comm J Qual Improv.1998;24:130-142.EHLinWJKatonGESimonAchieving guidelines for treatment of depression in primary care: is physician education enough?Med Care.1997;35:831-842.HCSchulbergMRBlockMJMadoniaERodriguezCPScottJLaveApplicability of clinical pharmacotherapy guidelines for major depression in primary care settings.Arch Fam Med.1995;4:106-112.Accepted for publication October 27, 1999.Funded by a grant from Pfizer Pharmaceuticals Inc, New York, NY (Drs Katzelnick, Simon, and Pearson).We thank the members of the study's data safety and monitoring board: David DeMets, PhD (chair), James Jefferson, MD, William McKinney, MD, Colleen McHorney, PhD, and Jeff Douglas, PhD, and advisory committee: Jonathan Borus, MD (chair), Donald Klein, MD, and Herbert C. Schulberg, PhD. We also thank Wilma Harrison, MD, and Robert Miceli, PhD, for their assistance in designing the study.Corresponding author: David J. Katzelnick, MD, Dean Foundation for Health, Research and Education, 2711 Allen Blvd, Middleton, WI 53562 (e-mail: katzelnick_david_j@ssmhc.com). http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Archives of Family Medicine American Medical Association

Randomized Trial of a Depression Management Program in High Utilizers of Medical Care

Download PDF
7 pages

Loading next page...
 
/lp/american-medical-association/randomized-trial-of-a-depression-management-program-in-high-utilizers-cFshNoH28y
Publisher
American Medical Association
Copyright
Copyright 2000 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.
ISSN
1063-3987
eISSN
1063-3987
DOI
10-1001/pubs.Arch Fam Med.-ISSN-1063-3987-9-4-foc9088
Publisher site

Abstract

BackgroundHigh utilizers of nonpsychiatric health care services have disproportionally high rates of undiagnosed or undertreated depression.ObjectiveTo determine the impact of offering a systematic primary care–based depression treatment program to depressed "high utilizers" not in active treatment.DesignRandomized clinical trial.SettingOne hundred sixty-three primary care practices in 3 health maintenance organizations located in different geographic regions of the United States.PatientsA group of 1465 health maintenance organization members were identified as depressed high utilizers using a 2-stage telephone screening process. Eligibility criteria were met by 410 patients and 407 agreed to enroll: 218 in the depression management program (DMP) practices and 189 in the usual care (UC) group.InterventionThe DMP included patient education materials, physician education programs, telephone-based treatment coordination, and antidepressant pharmacotherapy initiated and managed by patients' primary care physicians.Main Outcome MeasuresDepression severity was measured using the Hamilton Depression Rating Scale (Ham-D) and functional status using the Medical Outcomes Study 20-item short form (SF-20) subscales. Outpatient visit and hospitalization rates were measured using the health plan's encounter data.ResultsBased on an intent-to-treat analysis, at least 3 antidepressant prescriptions were filled in the first 6 months by 151 (69.3%) of 218 of DMP patients vs 35 (18.5%) of 189 in UC (P<.001). Improvements in Ham-D scores were significantly greater in the intervention group at 6 weeks (P=.04), 3 months (P=.02), 6 months (P<.001), and 12 months (P<.001). At 12 months, DMP intervention patients were more improved than UC patients on the mental health, social functioning, and general health perceptions scales of the SF-20 (P<.05 for all).ConclusionIn depressed high utilizers not already in active treatment, a systematic primary care–based treatment program can substantially increase adequate antidepressant treatment, decrease depression severity, and improve general health status compared with usual care.DEPRESSION among primary care patients is associated with significant functional impairment, lost productivity, and use of health services. Epidemiologic studies consistently demonstrate that approximately 10% of primary care patients suffer from significant depressive disorders.Depressed primary care patients report poorer quality of life,greater impairment of daily functioning,and more disability and/or lost productivity because of illness.Use of general medical services by depressed primary care patients is 50% to 100% higher than utilization by similar patients without depressive illness.The economic burden of depression in primary care (in terms of increased health care utilization and lost productivity) vastly exceeds the resources currently devoted to treatment.Among primary care patients with clinically significant depression, as many as half go unrecognized.Fewer than half of patients with recognized depression receive treatment of proven efficacyand of those who do obtain treatment, half or fewer receive levels of treatment consistent with expert guidelines.These high rates of underrecognition and undertreatment are especially concerning given the strong evidence for the efficacy of both pharmacotherapy and depression-specific psychotherapy.A series of recent randomized trials have demonstrated that systematic efforts to provide high-quality depression treatment can significantly improve clinical outcomes.These studies show that depression treatment effectiveness can be increased by using careful monitoring of patients and active involvement of specially trained clinicians.The need for improved management of depression is especially clear among "high utilizers" of general medical services. High utilizers of health care services have a high prevalence of depressive disorders,with considerable associated disabilityand excessive use of general medical services.High utilizers who are depressed are also more likely to continue to be high utilizersand often go either untreated or undertreated.Katon et alhave reported a randomized trial of a program to increase physician recognition of psychiatric disorders among high utilizers of outpatient care. This program (including feedback of psychiatric diagnosis, treatment recommendations to primary care physicians, and physician education) had no significant effects on treatment received or patient outcomes.Mindful of that negative result, we have developed a program that follows-up depression screening and physician education with a systematic treatment program to ensure depression treatment of sufficient intensity. This depression management program (DMP) includes many of the features used in the successful consultant/liaisondepression treatment model but is based with the patient's own primary care physician. A small pilot study using this program resulted in improved clinical outcomes and a reduction in total health care utilization.This article describes a randomized trial of a program to identify and treat depression among high utilizers of general medical care. Our intention was to develop and test a population-based screening program and a primary care–based treatment program that could be implemented in most managed health care systems. Results of the screening program (prevalence of depression, association of depression with functional impairment, and increased utilization) are described in a separate publication.This article uses data from the randomized trial to examine the impact of the DMP on treatment received, clinical outcomes, functional status, and health care utilization.SUBJECTS AND METHODSThe study was conducted within selected primary care clinics of 3 large prepaid health plans representing 3 geographic regions (Midwest, Northwest, and New England) in the United States: DeanHealth Plan (DHP), Wisconsin; Group Health Cooperative of Puget Sound (GHC), Washington; and Harvard Pilgrim Health Care (HPHC), Massachusetts. Both HPHC and GHC are not-for-profit health maintenance organizations (HMOs); DHP is a for-profit HMO. DeanHealth Plan has 175 000 members, GHC has 450 000 members, and the staff model division of HPHC has 300 000 members. The study protocol was approved by the institutional review boards of all 3 HMOs. Computerized databases from participating physician practices in all 3 HMOs were used to identify health plan members aged 25 to 63 years with continuous health plan enrollment for the previous 2 years. Members in this sample with ambulatory visit counts above the 85th percentile for both of the 2 previous years were considered high utilizers. For screening purposes, ambulatory office visits were defined as primary care visits, medical specialty visits, and walk-in clinic visits. Ambulatory office visits did not include emergency department visits, all mental health provider visits, routine obstetric visits, physical therapy/occupational therapy visits, optometry visits, x-ray/diagnostic visits, home health visits, telephone contacts, chiropractic visits, and allergy injections. Computerized record data were then used to assess the following exclusion criteria: recent treatment for alcohol or other substance abuse; past treatment for schizophrenia or bipolar disorder; life-threatening medical disorders (eg, metastatic malignant neoplasm); or active treatment for depression. Active treatment was defined as current specialty mental health treatment or minimal adequate trial of antidepressant medication. Minimal adequate trial was defined as an antidepressant prescription(s) filled in the past 90 days with at least 30 days of an adequate dose based on American Psychiatric Association guidelines.We used a 2-stage telephone-screening process to identify those high utilizers likely to benefit from depression treatment.The first screening call included a modified version of the Structured Clinical Interview for DSM-IV(SCID)to identify 1465 enrollees who screened positive for either current major depression or major depression in partial remission. Of these, 1295 enrollees agreed to complete a second telephone interview to assess depression severity using the 17-item Hamilton Depression Rating Scale (Ham-D).Potential study subjects were excluded if they had a Ham-D score less than 15 or met any of the following conditions: contraindications to taking an antidepressant, receiving treatment by a psychiatrist within the past 4 months, pregnancy, planned pregnancy within the next year, breastfeeding, positive screen for alcohol abuse,and intent to disenroll from the HMO.The 163 participating physician practices were randomly assigned to either the DMP (n=82) or usual care (UC) (n=81) groups. While randomization of physicians occurred prior to recruitment, study personnel at each site were not informed of an individual patient's assignment until that patient had completed the enrollment process. Eligible and consenting patients from the practices of UC physicians were informed that telephone screening suggested depression and were advised that care was available with their primary care physician. Patients from the practices of DMP physicians were invited to participate in the protocol described below. All analyses classified patients according to the original assignment of their primary care physicians, regardless of the treatment actually received.The principal elements of the DMP were physician education, patient education, antidepressant treatment, and treatment coordination. Prior to patient enrollment, all physicians in the DMP group participated in a standardized 2-hour training program focused on the initial assessment of depression and the initiation of pharmacotherapy. At each health plan, 1 or 2 psychiatrists were identified as consultants for the DMP. At the time of enrollment, all study patients in practices of the DMP group physicians were asked to schedule an evaluation visit with their primary care physician. Prior to this visit, patients received a booklet created for the study "Depression Isn't Just a Mental Problem"and videotaped educational materials from the treatment coordinator designed to increase acceptance of depression treatment. At the initial visit, primary care physicians confirmed the diagnosis of depression, assessed contraindications to pharmacotherapy, and (if indicated) recommended antidepressant treatment. Patients who had psychotic symptoms, mania, or acute suicidality were immediately referred to psychiatrists.Primary care physicians were advised to follow a specific pharmacotherapy algorithm, but were allowed to adjust treatment according to individual clinical need. The treatment algorithm recommended that patients who had previously been successfully treated and tolerated an antidepressant be given the previous antidepressant. All others were to be given sertraline hydrochloride at a recommended starting dose of 50 mg/d. It was also suggested that patients with panic attacks or who weighed less than 45 kg be given 25 mg/d. The dose was to be increased at 4 weeks if the patient had not improved, with a maximum dose of 200 mg/d. Patients who did not respond to or tolerate sertraline were to be treated with nortriptyline, if clinically appropriate. Nortriptyline was to be started at 25 mg/d, titrating up to 100 mg/d if needed. It was strongly recommended that psychiatric consultation be initiated if the preceding steps were unsuccessful. Follow-up visits with the primary care physician were prescheduled at approximately 1, 3, 6, and 10 weeks. Subsequent visits occurred approximately every 10 weeks. The DMP patients were also offered the RHYTHMS®depression education program at the first physician visit if appropriate. The program provided information on the nature of depression, the use of antidepressant medications, and other related topics aimed at increasing patient antidepressant adherence.Primary care treatment was supported by an ongoing program of monitoring, feedback, and as-needed specialty consultation. Coordinators reviewed DMP patient prescription refills and office visits to identify unplanned treatment discontinuation. In addition, treatment coordinators contacted DMP patients for telephone monitoring of treatment adherence; treatment response (based on administration of the 17-item Ham-D); and medication adverse effects after 2 and 10 weeks (and, if needed, after 18, 30, and 42 weeks). Treatment coordinators did not have access to the Ham-D scores generated by the blinded raters. The coordinators all had bachelor's or master's degrees and at least some clinical mental health experience. Data from this monitoring program, as well as recommendations for adjustments in treatment, were provided to treating primary care physicians after each monitoring contact. A written response was generated if the subject was doing well and a call to the physician was placed if the subject was not doing well. The study physicians did not have to acknowledge receipt of written messages. Study psychiatrists had ongoing contact with all intervention group primary care physicians via periodic case reviews and as-needed telephone consultation. A psychiatric consultation visit was strongly encouraged for all DMP patients not responding to treatment by 10 weeks and for patients with more complicated depression (eg, major psychiatric comorbidity, past treatment failures). While UC patients could self-refer to any specialty services normally available to health plan members, no additional monitoring, case management, or psychiatric liaison services were provided.All patients in both the DMP and UC groups were contacted for blinded telephone assessments 6 weeks, 3 months, 6 months, and 12 months after randomization. These assessments included a second administration of the Ham-D, a patient-rated 7-point scale of overall improvement since enrollment (patient global improvement score),and the Medical Outcomes Study 20-item short form (SF-20)functional status questionnaire. Health care utilization was collected from the HMOs' automated claims databases for 1 year before and 1 year after study randomization. Outpatient visits were defined as unique patient, provider, and date of services for all clinic or outpatient hospital encounters, excluding visits to departments of radiology, pathology, and anesthesiology. Specialty mental health outpatient visits are defined as those visits provided by a mental health specialist and include visits to the psychiatrists involved in the study. Hospital admissions are defined as a stay in the hospital requiring at least 1 night's stay, with stays not separated by at least 1 outpatient day considered a single admission. Only people who had a full year of health plan coverage were included in the analysis of outpatient visits and inpatient stays.Participants provided informed consent to participate prior to the initial telephone screening, prior to the second-stage screening, and prior to enrollment in the randomized trial. The DMP patients provided additional informed consent to participate in the treatment program. An independent data safety committee monitored unblinded patient outcomes and adverse events throughout the study.Primary comparisons of the DMP and UC groups were based on change scores (follow-up minus baseline scores) using an intent-to-treat approach. Estimation was done by analysis of variance with robust or sandwich estimates for the variance/covariance matrix corrected for intracluster correlation at the patient level.We also obtained very similar estimates controlling for site, age, sex, and baseline mental and physical health status, using a random-effects generalized least-squares estimator, generalized estimating equations for the health status measures, and negative binomial regression for outpatient visits and inpatient stays (not shown). For simplicity, we report the analysis of variance results on change scores, with robust estimates of the inference statistics and corrected for intracluster correlation among patients of the same physician.RESULTSOf the 1295 members who completed the second screening call, 410 (32%) had Ham-D scores of 15 or more and were eligible to participate. Almost all (n=407) agreed to participate. Based on randomization of treating physicians, 218 were assigned to DMP and 189 to UC (with the imbalance reflecting unequal assignment of full-time and part-time physicians to the 2 groups). Demographic and health data were similar for the 2 groups (Table 1), with the exceptions of ethnicity and recent inadequate antidepressant use, which are commented on below. The overall F test for differences in the means for all of the variables in Table 1for DMP vs UC demographics was not significant (P=.13). Differences between DMP and UC in ethnic distribution were not significant after controlling for study site. There were no significant or appreciable differences in mental health status between the DMP and UC (P=.80). The mean baseline Ham-D score was 19.2 for the DMP group and 19.1 for UC. Medical comorbidity was common: 46.7% of randomized patients had at least 1 chronic medical disorder and 14.5% had 2 or more.The number of study patients who completed the 12-month blinded assessment call was 380: 203 (93.1%) of 218 in DMP vs 177 (93.7%) of 189 in UC (P=.83) (Figure 1). Sample loss was not significantly related to DMP vs UC over time, to baseline Ham-D scores, or to other health, site, or demographic characteristics.Demographic and Patient Characteristics*DMP (n=218)UC (n=189)P†Age, y, mean (95% CI)45.6 (44.0-47.1)45.4 (43.9-46.9).89Female, % (95% CI)76.6 (70.9-82.4)78.3 (71.7-84.9).69White, % (95% CI)87.6 (82.4-92.8)77.3 (70.5-84.0).02Married, % (95% CI)71.1 (64.6-77.6)65.6 (58.8-72.5).25Years of education, mean (95% CI)13.6 (13.4-13.9)13.8 (13.4-14.1).52Depression status, % (95% CI)Current64.7 (59.4-70.0)63.5 (56.4-70.6).79Partial remission35.3 (30.0-40.6)36.5 (29.4-43.6).79SF-20 subscales, mean (95% CI)Ham-D19.1 (18.7-19.6)19.2 (18.7-19.7).75MHI-547.4 (44,7-50.0)48.5 (45.7-51.4).56GHP37.4 (34.2-40.6)38.6 (34.8-42.4).64PF54.6 (50.2-58.9)50.4 (45.6-55.1).20Patient seeing mental health provider (last 2 y), % (95% CI)33.9 (27.8-40.1)34.4 (27.2-41.6).93Inadequate trial of antidepressant (last 90 d), % (95% CI)24.3 (18.5-30.1)13.8 (7.7-19.8).01*DMP indicates depression management program; UC, usual care; SF-20, Medical Outcomes Study 20-item short form; Ham-D, Hamilton Depression Rating Scale; MHI-5, mental health inventory 5; GHP, general health perception; CI, confidence interval; and PF, physical functioning.†Pvalues for individual variables are unadjusted for multiple comparisons. An overall F test of DMP vs UC for all variables was not significant (P=.13).Figure 1.A flowchart mapping randomization of patients participating in the trial to either usual care (UC) or the depression management program (DMP).Of the 218 patients in the DMP, 198 scheduled and 189 completed the first physician visit and 166 started the RHYTHMS program. A treatment coordinator telephone contact was made at 2 weeks (177 patients), 10 weeks (194 patients), 18 weeks (131 patients), and 30 weeks (90 patients), for an average of 2.7 contacts per DMP patient. Computerized administrative databases were examined to determine the treatment actually received by DMP and UC patients. The DMP patients were much more likely to receive adequate antidepressant treatment. Pharmacy refill data for the 6-month period after enrollment indicated that 179 (82.1%) of 218 DMP patients and 61 (32.3%) of 189 UC patients filled at least 1 antidepressant prescription (P<.001); 151 (69.3%) of 218 DMP patients and 35 (18.5%) of 189 UC patients had filled at least 3 antidepressant prescriptions (P<.001) (Figure 2). The mental health treatment provided in both groups was given largely in primary care. At least 1 specialty mental health visit was made by only 59 (27.1%) of 218 DMP patients and 32 (16.9%) of 189 UC patients (P=.03). Only 29 (13.3%) of 218 DMP patients vs 18 (9.5%) of 189 UC patients made 3 or more specialty mental health visits in the first 6 months (P=.29).Figure 2.Patterns of antidepressent treatment for patients who received either the depression management program (DMP) or usual care (UC).Patients assigned to the intervention group experienced significantly better outcomes at every follow-up assessment. Improvements in Ham-D scores were significantly greater in the DMP than in the UC group: −3.3 vs −2.0 at 6 weeks (P=.04); −5.6 vs −3.9 at 3 months (P=.02); −7.3 vs −4.0 at 6 months (P<.001); and −9.2 vs −5.6 at 12 months (P<.001). The difference between DMP and UC groups increased significantly over time (P=.005) (Figure 3). At 12 months, 108 (53.2%) of 203 DMP patients had responded (defined as at least 50% reduction in Ham-D scores from baseline) compared with 58 (32.8%) of 177 UC patients (P<.001), and 92 (45.3%) of 203 DMP patients were in remission (defined as a Ham-D score <7) compared with 49 (27.7%) of 177 UC patients (P<.001). Forty-one (23.2%) of 177 UC patients vs 26 (12.8%) of 203 DMP patients reported an increase in Ham-D scores (an increase indicates worse health) from baseline to 12-month assessment (P=.01). One hundred seventeen (57.6%) of 203 DMP patients rated themselves as "much" or "very much" improved at 12 months compared with 60 (33.7%) of 178 UC patients (P<.001), concordant with the Ham-D results. At 12 months, DMP patients reported significantly better social functioning, mental health, and general health perceptions than UC patients on the SF-20 (P<.05 for all), but not on physical functioning, role functioning, and pain perception (Figure 4). Given the baseline imbalance in the proportion of patients receiving recent inadequate antidepressant treatment, the above analyses were repeated, excluding all patients receiving inadequate antidepressant treatment during the 90 days prior to enrollment. Results were nearly identical.Figure 3.Patients' scores on the Hamilton Depression Rating Scale (Ham-D) during the 12-month study. The Pvalues were based on the change in scores since the baseline assessment.Figure 4.Change in patients' scores on the Medical Outcomes Study 20-item short form (SF-20) subscales after receiving either the depression management program (DMP) or usual care (UC) for 12 months. The Pvalue was based on the change in scores since the baseline assessment.Comparing the year before randomization with the year after randomization, mean visit counts in DMP increased by 1.6 visits (18.4 to 19.9 visits), 0.5 of which was attributable to a single case. Mean visit counts in UC decreased by 2.0 visits (19.4 to 17.4 visits). The change in visits did differ significantly between groups (P=.02). Outpatient visits included nonphysician encounters, such as physical therapy and speech therapy. Comparing the year before randomization with the year after randomization, mean inpatient admissions in the DMP group increased by 0.04 (0.23 to 0.27 admissions) and mean inpatient admissions in the UC group decreased by 0.08 (0.26 to 0.18 admissions). The change in inpatient admissions between groups was not significantly different (P=.09).COMMENTDepressed high-utilizer patients not already in active treament who received the DMP had significantly better clinical outcomes and improved general health status compared with patients who received UC. The effectiveness of the DMP in this real world clinical setting is notable, given that some DMP patients did not pursue treatment and some UC patients received adequate treatment through the usual channels. The results show that the DMP did not decrease outpatient visits compared with UC and our estimate is actually a relative increase in outpatient visits. While the study protocol recommended 9 or more outpatient visits for initiating and monitoring depression treatment, the DMP group increased by only 1.6 visits, while the UC group decreased by 2.0 visits, suggesting that most visits for depression could be combined with other visits.The following limitations should be considered when generalizing the results of this study. First, the UC patients may have been more likely to seek depression treatment than usual because the telephone screeners recommended that they contact their physicians. At the exit interview, 74 (41.6%) of 178 UC patients reported they took action in some way based on this recommendation. The low rate of specialty mental health visits and prescription of antidepressants in the UC group suggests that this had a limited effect. Because of this limitation, the study results probably represent a lower bound of the impact of the intervention. Second, the initial imbalance of more DMP patients than UC patients having an inadequate antidepressant trial prior to entering the study could have affected the results. This seems unlikely, as the results did not change when randomized patients on inadequate antidepressant trial at baseline were excluded from the analyses. Third, because the study was not powered to detect a change in inpatient admissions, we cannot determine whether the intervention affected inpatient utilization. Fourth, the 1-year study period may have been too short to capture the full benefit of the DMP intervention. The difference in Ham-D scores between the DMP and UC groups was greatest at the 12-month assessment point. Fifth, care must be taken in generalizing these results to depressed populations other than high utilizers.How substantial is the effect of the DMP intervention? One way to assess the impact of the DMP is to consider the number of patients with depression who would need treatment in the DMP group to achieve 1 more remission than would occur in the UC group. Based on our data, this number needed to treat was 5:1. This ratio compares favorably with the 9:1 number of patients with diastolic blood pressure of 115 to 129 mm Hg needed to treat to achieve 1 less stroke, myocardial infarction, or death.Another way to assess the DMP is to compare it with efficacy studies of antidepressants vs placebo. The differences of 17.6% in remission rates and 3.6 in Ham-D scores between the DMP and UC groups in this study is similar to the difference seen between efficacious antidepressant and placebo in clinical trials.While the difference between the DMP and UC patients in mean Ham-D scores is similar to that between active treatment and placebo groups in other primary careand specialtystudies, rates of response and remission in both groups seem low. This study used a 2-stage screening process that required a diagnosis of depression at the first call and a Ham-D severity greater than or equal to 15 at the second call 2 weeks later. Requiring participants to have depressive symptoms that persisted over time may have decreased the number of patients likely to spontaneously recover.The DMP patients on average took longer to respond to treatment than has been seen in clinical trials.One reason is that time zero was defined as the date of study enrollment, not the first physician visit, which was at least 2 weeks later. Another reason may be that high utilizer patients have more medical and psychiatric comorbidity, which has been shown to be associated with delayed onset of treatment response and lower response rates.Most DMP patients (69.3%) received at least adequate antidepressant treatment (defined here as 3 or more antidepressant prescriptions) during the first 6 months of the study, even though the large majority of DMP patients (72.9%) received no specialty mental health treatment. This is consistent with the intervention's goal of providing primary care–based treatment for the majority of depressed patients and increasing the intensity of specialty care for nonresponding patients.Our randomized trial of a multicomponent intervention does not allow specific conclusions regarding the benefits of individual intervention components. However, considering our findings along with those of other primary care intervention studies suggests some general conclusions. Several programs limited to physician training or depression screening have not improved patient outcomes.Common elements of effective programsinclude guideline-based treatment, organized follow-up care, and active monitoring of patient outcomes.Our findings suggest that real-world primary care physicians can deliver effective depression treatment, if those physicians are supported by organized systems for patient education, proactive psychiatric consultation for nonresponders, physician education, and patient monitoring. Previous studies of programs that focused solely on increasing recognition of depression have not demonstrated differences in patient outcomes.The DMP tested in this study was not as tightly controlled as previous successful primary care interventions, but it was more intensive than several earlier unsuccessful programs. The successful programs described by Katon et aland Schulberg et alall involved direct treatment by specially trained study personnel. In the study by Katon et al,mental health specialists participated in the care of all patients. Schulberg et alused 2 different systematic depression treatment programs for patients in primary care that were more effective than UC. The major difference between the DMP used in this study and the successful Katon and Schulberg models is that patient's depression treatment in the DMP was based with their own primary care physician.The DMP used in this study could be a model for use with high utilization patients in many health care organizations. The DMP treatment model has the advantage of being initiated by the patient's own primary care physician. In addition, it encourages integration of medical and psychiatric treatment. Another substantial advantage is that patient visit and pharmacy tracking programs already present in many existing managed care organizations will facilitate the implementation of the DMP. Further research will be needed to determine the effectiveness of this systematic treatment model in other depressed populations.RLSpitzerJBWilliamsKKroenkeUtility of a new procedure for diagnosing mental disorders in primary care: the PRIME-MD 1000 study.JAMA.1994;272:1749-1756.TUstunNSartoriusMental Illness in General Health Care: An International Study.London, England: John Wiley & Sons Inc; 1995.RLSpitzerKKroenkeMLinzerHealth-related quality of life in primary care patients with mental disorders: results from the PRIME-MD 1000 Study.JAMA.1995;274:1511-1517.JLCoulehanHCSchulbergMRBlockMJMadoniaERodriguezTreating depressed primary care patients improves their physical, mental, and social functioning.Arch Intern Med.1997;157:1113-1120.JOrmelMVonKorffTBUstunSPiniAKortenTOldehinkelCommon mental disorders and disability across cultures: results from the WHO Collaborative Study on Psychological Problems in General Health Care.JAMA.1994;272:1741-1748.GSimonJOrmelMVonKorffWBarlowHealth care costs associated with depressive and anxiety disorders in primary care.Am J Psychiatry.1995;152:352-357.JUnutzerDLPatrickGSimonDepressive symptoms and the cost of health services in HMO patients aged 65 years and older: a 4-year prospective study.JAMA.1997;277:1618-1623.GESimonMVon KorffWBarlowHealth care costs of primary care patients with recognized depression.Arch Gen Psychiatry.1995;52:850-856.HCSchulbergMRBlockMJMadoniaThe "usual care" of major depression in primary care practice.Arch Fam Med.1997;6:334-339.GESimonMVonKorffRecognition, management, and outcomes of depression in primary care.Arch Fam Med.1995;4:99-105.KBWellsWKatonBRogersPCampUse of minor tranquilizers and antidepressant medication by depressed outpatients: results from the medical outcomes study.Am J Psychiatry.1994;151:694-700.GESimonMVonKorffEHWagnerWBarlowPatterns of antidepressant use in community practice.Gen Hosp Psychiatry.1993;15:399-408.DJKatzelnickKAKobakJHGreistJWJeffersonHJHenkEffect of primary care treatment of depression on service use by patients with high medical expenditues.Psychiatr Serv.1997;48:59-64.HCSchulbergMRBlockMLMadoniaTreating major depression in primary care practice: eight-month clinical outcomes.Arch Gen Psychiatry.1996;53:913-919.ATBeckSDHollonJEYoungRCBedrosianDBudenzTreatment of depression with cognitive therapy and amitriptyline.Arch Gen Psychiatry.1985;42:142-148.GKlermanMWeissmanInterpersonal psychotherapy (IPT) and drugs in the treatment of depression.Pharmacopsychiatry.1987;20:3-7.US Dept of Health and Human ServicesDepression in Primary Care: Treatment of Major Depression (Clinical Practice Guideline).Rockville, MD: US Dept of Health and Human Services; 1993.WKatonMVon KorffELinCollaborative management to achieve treatment guidelines: impact on depression in primary care.JAMA.1995;273:1026-1031.WKatonPRobinsonMVon KorffA multifaceted intervention to improve treatment of depression in primary care.Arch Gen Psychiatry.1996;53:924-932.LMMynors-WallisDHGathARLloyd-ThomasDTomlinsonRandomised controlled trial comparing problem solving treatment with amitriptyline and placebo for major depression in primary care.BMJ.1995;319:441-445.WKatonMVon KorffELinDistressed high utilizers of medical care: DSM-III-Rdiagnoses and treatment needs.Gen Hosp Psychiatry.1990;12:355-362.MVon KorffJOrmelWKatonEHLinDisability and depression among high utilizers of health care: a longitudinal analysis.Arch Gen Psychiatry.1992;49:91-100.HJHenkDJKatzelnickKAKobakJHGreistJWJeffersonMedical costs attributed to depression among patients with a history of high medical expenses in a health maintenance organization.Arch Gen Psychiatry.1996;53:899-904.WKatonMVon KorffELinTBushJOrmelAdequacy and duration of antidepressant treatment in primary care.Med Care.1992;30:67-76.WKatonMVon KorffELinA randomized trial of psychiatric consultation with distressed high utilizers.Gen Hosp Psychiatry.1992;14:86-98.SDPearsonDJKatzelnickGESimonWGManningCPHelstadHJHenkDepression among high utilizers of medical care.J Gen Intern Med.1999;14:461-468.American Psychiatric AssociationPractice guideline for major depressive disorder in adults.Am J Psychiatry.1993;150(suppl 4):S1-S26.MFirstRSpitzerMGibbonJBWilliamsStructured Clinical Interview for DSM-IV Axis I Disorders-Patient Edition (SCID-I/P, Version 2.0).New York: New York State Psychiatric Institute; 1995.MHamiltonA rating scale for depression.J Neurol Neurosurg Psychiatry.1960;23:56-62.RLBrownLARoundsConjoint screening questionnaires for alcohol and other drug abuse: criterion validity in a primary care practice.WMJ.1995;94:135-148.DJKatzelnickGESimonSDPearsonDepression Isn't Just a Mental Problem[available on request from authors&rsqb;.1995.Not AvailableThe RHYTHMS Patient Compliance Program.New York, NY: Carlson HealthCare Publications Inc;1995.WGuyECDEU: Assessment for Psychopharmacology, Revised.Rockville, Md: National Institute of Mental Health; 1976. US Dept of Health, Education, and Welfare publication ADM 76-338.ALStewartRDHaysJEWareThe MOS Short-Form General Health Survey: reliability and validity in a patient population.Med Care.1988;26:724-735.PHuberThe Behavior of Maximum Livelihood Estimates Under Non-Standard Conditions: Fifth Berkeley Symposium in Mathematical Statistics and Probability.Berkeley: University of California Press; 1967.HWhiteA heteroscedasticity: consistant covariance matrix estimator and a direct test for heteroscedasticity.Econometrica.1980;48:817-830.ALaupacisDLSackettRSRobertsAn assessment of clinically useful measures of the consequences of treatment.N Engl J Med.1988;318:1728-1733.LFFabreFSAbuzzahabMAminSertraline safety and efficacy in major depression: a double-blind fixed-dose comparison with placebo.Biol Psychiatry.1995;38:592-602.JFeighnerA double-blind comparison of paroxetine, imipramine and placebo in depressed outpatients.Int Clin Psychopharmacology.1992;6:31-35.JEWernickeSRDunlopBEDornseifRLZerbeFixed-dose fluoxetine therapy for depression.Psychopharmacol Bull.1987;23:164-168.SKasperClinical efficacy of mirtazapine: a review of meta-analyses of pooled data.Int Clin Psychopharmacol.1995;10:25-35.JLClaghornCQCraigDDWalczakFluvoxamine maleate in the treatment of depression: a single-center, double-blind, placebo-controlled comparison with imipramine in outpatients.J Clin Psychopharmacol.1996;16:113-120.IElkinMTSheaJTWatkinsNational Intsitute of Mental Health Treatment of Depression Collaborative Research Program: general effectiveness of treatments.Arch Gen Psychiatry.1989;46:971-982.HMullerHJMollerMethodological problems in the estimation of the onset of the antidepressant effect.J Affect Disord.1998;48:15-23.CBrownHCSchulbergMJMadoniaMKShearPRHouckTreatment outcomes for primary care patients with major depression and lifetime anxiety disorders.Am J Psychiatry.1996;153:1293-1300.GIKeitnerCERyanIWMillerRKohnNBEpstein12-month outcome of patients with major depression and comorbid psychiatric or medical illness (compound depression).Am J Psychiatry.1991;148:345-350.CMCallahanHCHendrieRSDittusDCBraterSLHuiWMTierneyImproving treatment of late life depression in primary care: a randomized clinical trial.J Am Geriatr Soc.1994;42:839-846.CDowrickIBuchananTwelve month outcome of depression in general practice: does detection or disclosure make a difference?BMJ.1995;311:1274-1276.HIGoldbergEHWagnerSDFihnA randomized controlled trial of CQI teams and academic detailing: can they alter compliance with guidelines?Jt Comm J Qual Improv.1998;24:130-142.EHLinWJKatonGESimonAchieving guidelines for treatment of depression in primary care: is physician education enough?Med Care.1997;35:831-842.HCSchulbergMRBlockMJMadoniaERodriguezCPScottJLaveApplicability of clinical pharmacotherapy guidelines for major depression in primary care settings.Arch Fam Med.1995;4:106-112.Accepted for publication October 27, 1999.Funded by a grant from Pfizer Pharmaceuticals Inc, New York, NY (Drs Katzelnick, Simon, and Pearson).We thank the members of the study's data safety and monitoring board: David DeMets, PhD (chair), James Jefferson, MD, William McKinney, MD, Colleen McHorney, PhD, and Jeff Douglas, PhD, and advisory committee: Jonathan Borus, MD (chair), Donald Klein, MD, and Herbert C. Schulberg, PhD. We also thank Wilma Harrison, MD, and Robert Miceli, PhD, for their assistance in designing the study.Corresponding author: David J. Katzelnick, MD, Dean Foundation for Health, Research and Education, 2711 Allen Blvd, Middleton, WI 53562 (e-mail: katzelnick_david_j@ssmhc.com).

Journal

Archives of Family Medicine – American Medical Association

Published: Apr 1, 2000

References