Impact of a collaborative pharmacy practice model on the treatment of ...

5 downloads 0 Views 566KB Size Report
Treatment of depression prescribing privileges by KPMCSR's pharmacy and therapeutics commit- tee. This permitted the pharmacists to modify the dosages of ...
REPORTS

Treatment of depression

REPORTS

Impact of a collaborative pharmacy practice model on the treatment of depression in primary care PATRICK R. FINLEY, HEIDI R. RENS, JOAN T. PONT, SUSAN L. GESS, CLIFTON LOUIE, SCOTT A. BULL, AND LISA A. BERO

P

atients with depressive symptoms are more likely to seek help in the primary care setting than any other health care environment.1,2 Observational studies continue to demonstrate, however, that the outcomes of patients with depression managed in primary care are less than optimal.3-5 Disease detection rates are low, the intensity of treatment is usually inadequate, and patients are often left with significant clinical and functional impairments.3,6-9 Deficiencies in the management of depression have become a major health policy concern.10 Depression is a remarkably common illness associated with substantial disability. The one-year prevalence of major depression (acute episodes) in the United States is 6.4%.11 Recent projections from the World Health Organization suggest that depression will be the second leading cause of disability worldwide, exceeding hypertension, arthritis, diabetes, and chronic back

Abstract: The effects of a collaborative pharmacy practice model, in which clinical pharmacy specialists provided medication maintenance and follow-up patient care services at a clinic, on patients’ adherence to treatment and satisfaction and costs were studied. A cohort of 13 primary care providers (PCPs) was designated to refer patients diagnosed with depression to the practice model at a staff-model health maintenance organization (HMO) immediately after the initiation of antidepressant medications. Clinical pharmacy specialists proceeded to coordinate follow-up with the patients for six months through a combination of scheduled office visits and telephone calls. Working closely with psychiatric liaisons, pharmacists were granted limited prescribing privileges to provide medication comanagement. These patients’ adherence to treatment and satisfaction and costs to the HMO were compared with a control group of patients being treated for depression by the remaining 17 PCPs at the facility. A total of 91 patients were referred to the intervention group and received care

PATRICK R. FINLEY, PHARM.D., BCPP, is Associate Clinical Professor, Department of Clinical Pharmacy, School of Pharmacy, University of California at San Francisco (UCSF), San Francisco. HEIDI R. RENS, PHARM.D., is Clinical Pharmacist; JOAN T. PONT, M.D., is Assistant Chief of Medicine; and SUSAN L. GESS, PHARM.D., is Drug Education Coordinator, Kaiser Permanente Medical Center, San Rafael, CA. CLIFTON LOUIE, DPA, is Vice Chair, Department of Clinical Pharmacy, School of Pharmacy, UCSF. SCOTT A. BULL, PHARM.D., is Outcomes Researcher, Kaiser Permanente Division of Research, Oakland, CA. LISA A. BERO, PH.D., is Professor of Clinical Pharmacy and Health Policy, School of Pharmacy, UCSF. Address correspondence to Dr. Finley at the University of California at San Francisco, 521 Parnassus Avenue, Room C-152, San Francisco, CA 94143-0622. The following individuals are acknowledged for their support,

1518

Am J Health-Syst Pharm—Vol 59 Aug 15, 2002

from the pharmacists during the 10-month enrollment phase; 129 patients were included in the control group. There were no significant differences between groups regarding age, sex and chronic disease scores. An intent-to-treat analysis of medication adherence revealed that adherence was significantly higher in the intervention group (medication possession ratio, 0.81 versus 0.66) (p = 0.0005). Medication switch rates were higher among intervention patients as well (24% versus 5%) (p = 0.0001). There was a greater decline in the number of visits to PCPs for patients in the intervention group (39% versus 12%) (p = 0.029). A collaborative practice model in which clinical pharmacy specialists managed the medication therapy of patients with mild to moderate depression increased patients’ adherence to treatment and their satisfaction and reduced the patients’ subsequent visits to PCPs. Index terms: Antidepressants; Clinical pharmacy; Depression; Models; Primary care Am J Health-Syst Pharm. 2002; 59:1518-26

encouragement, and guidance: Enid Hunkeler, M.S.; Bruce Fireman, Ph.D.; William Pigeon; Joseph Guglielmo, Pharm.D.; Elior Vas, M.D.; David Richman, M.D.; and Barry Pierce, M.D. Francois Collin and Janelle Y. Lee are acknowledged for conducting the statistical analysis for this study. Funding assistance was provided by a grant from the Kaiser Permanente Sidney Garfield Memorial Fund (as part of the Interregional Depression Initiative) and by an unrestricted educational grant from Pfizer Inc. Presented at the ASHP Midyear Clinical Meeting, Las Vegas, NV, December 7, 2000. Copyright © 2002, American Society of Health-System Pharmacists, Inc. All rights reserved. 1079-2082/02/0802-1518$06.00.

REPORTS

pain.12 The total direct medical costs for patients with depression are much greater than for the general population, owing, in part, to the increased comorbidity of those afflicted.13,14 Indirect costs incurred by patients with depression are substantial as well, with work-absenteeism rates two to three times greater than the rest of society.15 From the employer’s perspective, the annual indirect cost of the illness is estimated as $4200 per worker with depression, or $250 per employee.16 While there is a general consensus that systematic efforts to improve the outcomes of patients with depression are worthwhile, there is no agreement as to the best approach.10,17,18 Clinical practice guidelines have been developed by many organizations, intensive physician education has been emphasized, and numerous depression-screening strategies have been exercised, but evidence supporting the clinical success and costeffectiveness of these efforts has not been forthcoming.19-25 Researchers have begun to focus on integrating various mental health specialists into the primary care arena as a means to improve outcomes, and results have generally been favorable.26-28 Katon et al.29 conducted a randomized controlled trial that examined the effectiveness of a collaborative approach, combining primary care providers (PCPs) and psychiatrists for follow-up care (i.e., study subjects seen by PCPs at weeks 1 and 3 and by psychiatrists at weeks 2 and 4). Patient outcomes were significantly better in the intervention group than the cohort receiving usual care, but the benefits of this approach were only apparent in patients with depression that was moderate to severe in psychopathology. As mild or subsyndromal depression may be seen more in the primary care setting than more severe varieties, the cost-effectiveness and feasibility of implementing this model on a large-scale basis remain to be determined.30

Several efforts have also been undertaken to improve the welfare of patients with depression, primarily through telephone follow-up. 31-33 Two investigations focused on the impact of specially trained nurses or therapists with a master’s degree to contact patients weekly during the acute phase of the disease.32,33 In the larger study, Hunkeler et al.32 found that patients receiving frequent telephone calls from these professionals were very satisfied with the additional attention, and there was evidence of higher recovery rates than those seen in randomized controls. Neither study, however, was able to demonstrate a favorable impact on medication adherence rates. In an effort to improve patients’ adherence to antidepressant drug therapy and outcomes, the pharmacists at Kaiser Permanente Medical Center, San Rafael (KPMCSR), CA, developed a collaborative care approach to managing depression in primary care that emphasized the role of clinical pharmacy specialists in the recovery process.34 Previous investigations have found that pharmacist interventions have a positive and substantial impact on patient care in both inpatient and ambulatory care settings.35-37 This approach has been particularly successful in managing chronic disease states that rely heavily on skilled medication management and follow-up (e.g., diabetes, asthma, hyperlipidemia, anticoagulation).38,39 For the treatment of mental illness, the intervention of pharmacists has also been associated with an improvement in the safety, efficacy, and cost-effectiveness of pharmacotherapy.40-44 For the purposes of our investigation, a multidisciplinary practice model was created in which clinical pharmacy specialists worked, in effect, at the juncture of primary care and psychiatry. Patients in this model were to receive a combination of frequent scheduled visits and telephone follow-up with the pharma-

Treatment of depression

cists in the hopes of forging a strong therapeutic alliance. Our hypothesis was that the emotional support, improved access, and skilled medication management provided by pharmacists would be associated with greater medication adherence and patient satisfaction in comparison to patients with depression who received usual care. In addition, this model of care might alter resource utilization patterns and decrease the number of patient visits to PCPs. Methods This investigation was conducted at KPMCSR, a medium-sized primary care facility within a large nonprofit staff-model health maintenance organization (HMO) (Kaiser Permanente of Northern California). The clinic is staffed by approximately 30 PCPs (28 physicians and 2 nurse practitioners) and provides medical care to over 52,000 members. Patient selection. For the purposes of this pilot project, a total of 13 PCPs were designated to refer eligible patients to the intervention group. The designation of these PCPs was at the discretion of the assistant chief of medicine (i.e., nonrandomized allocation) and occurred before commencing study activities. This allocation was necessitated by the limited clinic capacity of the clinical pharmacy service and the physicians’ willingness to participate. Both of the pharmacists participating in this practice model had received doctor of pharmacy degrees and participated in direct patient care activities for several years before the model was implemented. In addition, the senior member had completed two years of postdoctoral work and was board-certified in psychiatric pharmacy. He was responsible for training the other pharmacist during a two-month lead-in phase immediately preceding this investigation. In order to provide skilled medication management activities, the pharmacists were granted limited

Am J Health-Syst Pharm—Vol 59 Aug 15, 2002

1519

REPORTS

Treatment of depression

prescribing privileges by KPMCSR’s pharmacy and therapeutics committee. This permitted the pharmacists to modify the dosages of antidepressants prescribed by the patient’s PCP and add ancillary medications (e.g., trazodone), if necessary. If the pharmacists felt that an alternative antidepressant medication was indicated or that the original medication should be discontinued, this recommendation had to be approved by the PCP. The patient population consisted of HMO members who were identified by their PCP as suffering from depression and subsequently received prescriptions for antidepressant medications for this purpose. The first antidepressant prescription written by one of the PCPs was marked as the index prescription. Patients were excluded from analysis if they had received a prescription for antidepressant medication during the six-month period preceding the index prescription, if they were referred to the collaborative practice model without an antidepressant prescription (i.e., PCP referred the patient for the purpose of medication recommendation), or if they were already under the care of a psychiatrist within the HMO. Additional exclusion criteria included imminent suicidality, psychotic symptoms, active substance abuse or dependence, and a history of manic episodes. Patients were enrolled in the study from March 1998 to December 1998, followed by a six-month period of treatment and follow-up after which they were discharged from the pharmacists’ care. Resource utilization analysis included data from the 12 months before the index prescription was written and data from the 12 months after the index date for each patient. Patients in the control group consisted of HMO members who were treated for depression with antidepressants by the remaining 17 PCPs at this facility in accordance with the

1520

usual standard of care. The control group was identified retrospectively through a comprehensive review of automated pharmacy refill records for patients starting on antidepressant medications during the same nine-month time frame. Medical charts were reviewed to verify that depression was the expressed purpose of the prescription and to ensure that the same exclusionary criteria were applied to both study groups. The health care utilization data of these patients were collected for the period of July 1999 to June 2000. A comparison of various demographic variables and treatment patterns was completed to investigate potential referral bias or other baseline differences. In addition to age and sex, the medical comorbidity of the patients in both groups was compared. For this purpose, the investigators calculated the chronic disease score, a quantitative measure of a patient’s total disease burden based on International Classification of Diseases—9th Revision codes and automated pharmacy refill records.45 Diagnoses were obtained from the outpatient summary of clinical records, and medication profiles were accessed through the clinical information profiling system when the index prescription was received. Process of care. The 13 providers who participated in the collaborative practice model were encouraged to refer all patients diagnosed with depression and started on antidepressant medications to the pharmacist-managed clinic for medication management and follow-up. The pharmacists were contacted by the PCP immediately after the conclusion of an index visit. The pharmacists subsequently conducted an intake interview (20–30 minutes) with the patient, usually on the same day that the first antidepressant was prescribed. This interview was modeled after the standard psychiatric intake interview, emphasizing active

Am J Health-Syst Pharm—Vol 59 Aug 15, 2002

listening and thorough data gathering. The pharmacists assessed target symptoms, current environmental stressors, previous mental health history, medical history, social history, and family history. Patients were also assessed for the presence of certain conditions that mandated immediate referral to a psychiatrist, in which case the patient was excluded from the study. The latter half of the intake interview was devoted to patient education regarding the nature of depression as an illness (predisposing factors, natural course, prognosis, and anticipated recovery process), pharmacologic treatment for depression (potential adverse effects, drug interactions, therapeutic response pattern, and duration of treatment), and the importance of adhering to treatment. Patients were also given a single-page handout of written information reinforcing these messages and were advised of other treatment options available to HMO members (e.g., selfreferral to psychiatry, group therapy for depression). At the end of the intake interview, pharmacists documented pertinent findings in the patients’ medical record and provided future treatment recommendations. After the intake interview, pharmacists scheduled additional clinic visits with the patients at week 6 to assess the success of a therapeutic trial and week 24 to discuss the necessity of maintenance treatment. Frequent telephone contacts were also scheduled with the patient at weeks 1, 2, 4, 10, and 16. Adverse drug effects, therapeutic efficacy, adherence, and potential stressors were monitored during all telephone contacts, and the results were documented in the patients’ medical record. The pharmacists also helped patients identify pleasurable activities that they might have been neglecting during their acute illness and encouraged them to participate in these activities during the recovery process. Telephone calls were generally 5–10

REPORTS

minutes in length; patients were advised to call the clinic if they needed to talk to the pharmacist between scheduled contacts. If the pharmacist felt that a medication change was indicated or if other significant developments had occurred in the patient’s condition, he or she notified the patient’s PCP via e-mail with a recommendation for changing the treatment and awaited the PCP’s approval. A hard copy of this communication was then placed in the patient’s medical chart for documentation. The clinical pharmacists also worked under the auspices of a designated psychiatric liaison at the HMO. This psychiatrist met with the pharmacists for one hour weekly to discuss the condition and progress of referred patients. In addition, the liaison psychiatrist was available offsite for consultation during the hours of clinic operation. Outcomes of care. The primary outcome measures for this controlled cohort study were medication adherence rates, patient satisfaction, and resource utilization patterns (from an institutional perspective). Other outcome variables included an analysis of prescribing patterns and assessment of clinical and functional improvements (intervention group only). Medication adherence was assessed for both groups of patients at months 3 and 6. This was accomplished through a systematic review of automated prescription refill records. The medication possession ratio (MPR) was calculated by dividing the number of days supply of antidepressant medication a patient received during the six-month period by 180 days (six months).46 This calculation was inclusive of all antidepressant prescriptions the patients had picked up from the pharmacy during this period and included medication changes and switches. The days supply of medication was determined from the strength and

quantity of antidepressant dispensed, factoring in the dose and frequency indicated on the prescription. As all patients included in this study were identified by virtue of receiving a prescription for antidepressant medication and the minimum (and most common) supply of dispensed medication was 30 days, the effective range of MPR values was 0.17–1.0. An alternative method of comparing medication adherence rates was to calculate the percentage of study patients obtaining refills of an antidepressant between three and six months after the index prescription was dispensed. This particular value represents the actual percentage of patients receiving medication beyond the acute phase of treatment and has been found to correlate highly with the percentage of patients completing a full six months of therapy. To evaluate patient satisfaction with the treatment received, an identical two-page survey was mailed to both study groups at least six months after the index prescription date.33 This was accompanied by a stamped, self-addressed envelope with a discrete code identifying the subject’s study group (intervention versus control). As the surveys were completed anonymously, the investigators were effectively blinded to the identity of the individual respondents. In the survey, subjects were asked to indicate their level of satisfaction with the care they received on 11 specific items by rating them on a 5-point Likert scale, where 1 = very dissatisfied and 5 = very satisfied. The remaining three items on the survey were in a dichotomous format (yes or no), with two questions related to psychiatric history and the other to self-reported medication adherence. The influence of this practice model on resource utilization and the workload of PCPs was analyzed by a comparison of office visits with the referring prescriber. The number

Treatment of depression

of total office visits in the 12 months before the index prescription date, the number in the 12 months after, and the differences were compared between the two groups. As this investigation was conducted within a closed staff-model system, information on primary care visits was readily accessible through claims data, available in the institution’s medical information system. For the purposes of this particular comparison, the date on which the index prescription was written was included in the pretreatment phase. A comparison of the prescribing patterns for the two groups of patients was also conducted, including the percentage of patients prescribed selective serotonin-reuptake inhibitors (SSRIs) at baseline, the average initial dose of SSRI, the average maintenance dose of medication, and the percentage of patients who were switched to different agents. These comparisons were also based on prescription information included in the automated pharmacy records. Maintenance doses were defined as the same antidepressant and the same dosage taken for more than 90 days (i.e., the patient has been effectively stabilized on medication). The pharmacists providing care in this practice model assessed all patients enrolled in the treatment protocol for their clinical and functional outcomes at baseline, six weeks, and six months after they were initially seen. The pharmacists used the 28item clinician-rated Inventory for Depressive Symptoms (IDS) to assess psychopathology, which rates specific symptoms on a four-point scale (range, 0–90).47 They also used the one-item Work and Social Disability Scale (WSDS) to evaluate the patients’ functional outcomes.48 The WSDS uses a 5-point scale to measure the degree of a patient’s disability from absent to severe. The Clinical Global Impression (CGI) and Change in Clinical Global Impression scales were also completed at

Am J Health-Syst Pharm—Vol 59 Aug 15, 2002

1521

REPORTS

Treatment of depression

baseline, six weeks, and six months.49 These assessment tools use 7-point scales to rate a patient’s overall psychopathology from not ill to extremely ill and very much improved to very much worse, respectively. Statistical analysis. Baseline differences between the treatment and control groups were compared using Student’s t test for continuous data and the chi-square test for categorical data. The a priori level of significance was 0.05. Potential differences in medication adherence were examined by an intent-to-treat analysis, comparing the mean three-month and six-month MPRs for all patients in the intervention group versus those in the control group. After the data were analyzed for normal distribution, MPR values and the percentage of patients receiving antidepressants between months 3 and 6 were subjected to a Student’s t test and chi-square test, respectively, to investigate statistical differences. For the patient satisfaction survey, dichotomous variables were subjected to a chi-square analysis of paired data. Answers to items ranked on a 5point scale were analyzed as continuous variables, with 5 representing the greatest degree of satisfaction. Differences in clinical and functional outcomes were only available for the intervention group. Population means for these outcomes were compared between values obtained at baseline and 6 weeks and baseline and 24 weeks, respectively, using a Student’s t test. Results Patient characteristics. A total of 91 patients were referred to the intervention group from March 1998 to December 1998, and 129 patients were identified in the control group during the same time frame. In the intervention group, 61 patients (67%) completed the six months of care under the collaborative model. The primary reasons for dropout were referral to psychiatry (n = 11)

1522

and lack of efficacy from the antidepressants or refusal to consider antidepressants (8). Other reasons included serious medical complications (5), adverse effects (3), moved out of area (2), and death (1). On average, the patients in the intervention group talked to the pharmacists 6.0 times on the telephone (median duration, approximately 7 minutes per phone call) and saw the pharmacists at 1.6 office visits (20 minutes per visit) for a total investment of 71 minutes of pharmacist time in six months. Pairwise comparison of demographic variables revealed no statistically significant differences between the two study groups (Table 1). Analysis of the antidepressant initially prescribed by PCPs at baseline revealed that the vast majority of patients in both groups received SSRIs (paroxetine or fluoxetine)—85.1% in the intervention group versus 92.2% in the control group (p = 0.01), with starting doses slightly lower in the intervention group (p < 0.05 for both medications). Before receiving the index prescription, the average clinician-rated IDS score for the intervention patients was 19.8 (n = 60), reflecting mild-tomoderate disease severity. The corre-

sponding WSDS score was 2.88 for these patients (i.e., moderate impairment in work and social functioning). Intervention effects. Based on data obtained from pharmacy refill records, patients followed by pharmacists in the intervention group demonstrated a significantly higher level of medication adherence. As shown in Figure 1, six-month MPRs were significantly greater for the intervention group than the control group (intent-to-treat analysis, 0.811 versus 0.659) (p < 0.005). The percentage of patients continuing antidepressant therapy beyond three months was also higher in the intervention group (intent-to-treat analysis, 76% versus 51%) (p < 0.001). Patient satisfaction surveys were returned by 55 (61%) of the 91 patients in the intervention group and 56 (43%) of the 129 patients in the control group receiving usual care. Responses to the last two questions (regarding treatment history) suggest that the intervention group was more likely to have received an antidepressant previously (25% versus 19%) (p = 0.044) and slightly more likely to have received previous psychological counseling before this study. (33% versus 27%) (p = 0.051).

Table 1.

Patient Characteristics and Demographics at Baseline

Characteristic No. (%) female Mean ± S.D. age (yr) No. (%) age distribution 76 yr Mean ± S.D. chronic disease score No. (%) patients started on drugs Fluoxetine Paroxetine Other antidepressant Mean ± S.D. dose (mg) of SSRIs Start dose fluoxetine Start dose paroxetine Maintenance dose fluoxetine Maintenance dose paroxetine

Patient Group Control Intervention (n = 129) (n = 91) 92 (71) 61.1 ± 16.2

71 (79) 59.9 ± 15.9

31 (24) 33 (26) 30 (23) 35 (27) 3400 ± 2627

27 (30) 20 (22) 25 (27) 19 (21) 3732 ± 4550

55 (42.6) 65 (50.4) 9 (7.0)

28 (31.1) 45 (48.9) 18 (20.0)

11.6 ± 5.3 11.0 ± 4.5 15.2 ± 5.3 15.0 ± 5.8

9.55 ± 2.9 9.55 ± 1.5 18.4 ± 9.4 16.4 ± 9.3

pa 0.207 0.561 0.367

0.534 0.010

0.031 0.015 0.139 0.481

a Student’s t test was performed for continuous variables; chi-square analysis was performed for categorical variables. SSRIs = selective serotonin-reuptake inhibitors.

Am J Health-Syst Pharm—Vol 59 Aug 15, 2002

REPORTS

Figure 1. Medication possession ratios (MPRs) at three and six months. MPRs were determined from antidepressant prescription records and are defined as the number of days supply of medication divided by the reference time period. The MPRs for the intervention and control groups were significantly different at three and six months (p < 0.005). 1

1

0.901 0.9 0.811

MPR

0.805

Intervention (n = 91)

0.8 0.7

0.659

0.6

Control (n = 129)

0.5 0

3

6

Time (Months)

Both of the study groups expressed high levels of satisfaction with their overall care. For several measures there was a greater level of satisfaction among patients in the intervention group. The items that achieved significance in this regard included overall satisfaction with the HMO, satisfaction with provider follow-up, the personal nature of care, access to the provider, and the likelihood of the patient referring a friend to the HMO for the treatment of his or her depression (Table 2). Patients receiving pharmacist followup were also more likely to recall completing six months of antidepressant treatment (76% versus 53%) (p = 0.008), confirming objective evidence that medication adherence was substantially higher in the intervention group. Analysis of resource utilization patterns (12 months before the index prescription and 12 months after) revealed that there was a greater decline in primary care provider contacts among patients assigned to the intervention group. The total number of visits decreased from 335 to 203 (39.4%) in the intervention group and from 411 to 361 (12.2%) in the control group (p = 0.007 for the difference between the changes). Overall, the average maintenance dosages of antidepressants were comparable between study groups. The in-

tervention patients, however, were more likely to switch antidepressant medications during treatment followup (22 of 91, or 24.2%, in the intervention group versus 7 of 129, or 5.4%, in the control group) (p = 0.001). The associated clinical outcome among intervention patients featured an average decline of 12.1 points in IDS scores at six weeks and 14.6 points at six months. Functional impairments decreased from an average of 2.88 at baseline to 1.65 at six weeks and 1.35 at six months. The change in CGI scores was 1.92 at six weeks and 2.43 at six months on patient-rated scales and 2.22 and 2.63, respectively, on clinician-rated forms. All of these values represented significant changes from baseline. Discussion The implementation of a multidisciplinary practice model emphasizing specialized clinical pharmacy services was associated with a significant improvement in medication adherence rates and patient satisfaction in comparison to the outcomes of patients receiving usual care. Survey results indicated that the patients were particularly pleased with the enhanced provider accessibility, reliability of follow-up, and overall treatment at the HMO. The improvement seen in medication adherence rates during this in-

Treatment of depression

vestigation was relatively robust in comparison to other published studies. Currently, the Agency for Healthcare Research and Quality recommends that patients continue antidepressant treatment for four to seven months after the resolution of acute symptoms.19 Six-month medication adherence rates, however, have ranged from 4% to 40% in naturalistic settings.50-52 Recently, the National Committee on Quality Assurance adopted specific measures to evaluate the treatment that patients with depression received in the primary care setting, based on adherence rates (at three and six months) and provider contacts (minimum of three during first 12 weeks of treatment).53 Early benchmark values for six-month adherence rates continue to demonstrate that the majority of patients do not complete a full course of therapy. In this study, 75% of patients referred for pharmacist medication comanagement completed six months of antidepressant treatment, representing a significant improvement over a control group with relatively high adherence rates (compared with medical literature and other published values). Interestingly, these results contrast with those of randomized studies investigating other practice models, all of which have failed to demonstrate an improvement in adherence rates among patients with mild to moderate depressive symptoms.29,32,33,54 The unique benefits of this approach to managing depression in primary care may be due to several factors. For example, in comparison to usual care (or other practice models), this intervention relied heavily on thorough medication management services of pharmacists with specialized training in the appropriate use of psychotropic medications. In a retrospective analysis of patients with depression included in the Medical Outcomes Study, Sturm and Wells9 reported that two of the stron-

Am J Health-Syst Pharm—Vol 59 Aug 15, 2002

1523

REPORTS

Treatment of depression

Table 2.

Patient Satisfaction Survey Results Survey Item The overall treatment of your depression The personal nature of the care you received for this condition The ability of the Kaiser personnel to listen and acknowledge your concerns The explanation about why the antidepressants were prescribed The information you were given about how to take your antidepressant medication The information you were given on what to expect and how to deal with side effects The effectiveness of the medication prescribed Availability of advice for this condition Kaiser’s follow-up with you for this condition Kaiser Permanente, overall Would you recommend Kaiser to a friend

Mean ± S.D. Responsea Control Group Intervention (n = 56) Group (n = 55)

pb

3.87 ± 1.06

3.98 ± 1.08

0.581

3.93 ± 1.08

4.31 ± 0.96

0.053

3.85 ± 1.19

4.29 ± 0.10

0.037

3.94 ± 0.89

3.92 ± 1.14

0.924

4.11 ± 0.64

4.39 ± 0.79

0.050

3.63 ± 1.00

4.00 ± 1.05

0.062

3.89 ± 1.12 3.67 ± 1.12

3.65 ± 1.24 4.29 ± 0.96

0.309 0.003

3.42 ± 1.17 3.86 ± 1.00 3.98 ± 1.00

4.41 ± 1.00 4.20 ± 0.85 4.36 ± 0.79