Depression
Conditions
Keywords
major Depressive Disorder, Diabetes, Heart Disease
Brief summary
This study will evaluate the effectiveness of a nurse-led case management intervention in improving disease control and depression symptoms in adults with diabetes and/or heart disease who are also depressed.
Detailed description
Depression is a serious medical illness that has been associated with increased risk for heart disease and diabetes. Depression may negatively impact aspects of self-care that are required to effectively manage such long-term diseases. In depressed people who have heart disease and/or diabetes, treatment for depression appears to result in only limited improvements in depression symptoms and no improvements in heart disease and diabetes symptoms. An integrated treatment approach may be more effective in improving all three conditions. This study will evaluate the effectiveness of a nurse-led case management intervention in improving disease control and depression symptoms in adults with diabetes and/or heart disease who are also depressed. Participants in this single-blind study will be randomly assigned to take part in the case management intervention or receive usual care. All participants will attend 5 in-person study evaluation visits and receive 4 follow-up phone calls over 24 months. At each of the study visits, measurements of height, weight, waist size, and blood pressure will be taken. At study evaluations, blood and urine samples will also be taken. Participants will be asked not to eat for 8 hours before providing blood samples at 3 of the visits. During follow-up phone calls participants will answer various questions. The case management intervention will entail approximately 10 visits with a trained nurse at the clinic or by telephone. Participants in this group will receive educational materials about how to manage diabetes and/or heart disease and stress or depression. Nurses will also provide guidance and support in managing medications, phone calls to check participants' progress, and assistance in setting personal goals and in managing physical health problems and symptoms of depression or stress. Outcomes will be measured at Months 6, 12, 18, and 24 months.
Interventions
BEHAVIORALNurse-led case management
The case management intervention will entail approximately 10 visits with a trained nurse at the clinic or by telephone. Participants in this group will receive educational materials about how to manage diabetes and/or heart disease and stress or depression. Nurses will also provide guidance and support in managing medications, phone calls to check participants' progress, and assistance in setting personal goals and in managing physical health problems and symptoms of depression or stress.
BEHAVIORALTreatment as usual
Participants will attend 10 study visits and receive 4 follow-up phone calls over 24 months. During this time, participants will receive usual care.
Sponsors
National Institute of Mental Health (NIMH)
University of Washington
Study design
Allocation
RANDOMIZED
Intervention model
PARALLEL
Primary purpose
TREATMENT
Masking
SINGLE (Outcomes Assessor)
Eligibility
Sex/Gender
ALL
Age
18 Years to 80 Years
Healthy volunteers
No
Inclusion criteria
* Diagnosis of diabetes and/or heart disease * Poor disease control (defined as an HbA1c level of at least 8.5%, blood pressure greater than 140/90 mm Hg, LDL cholesterol greater than 130 mg/dL) * Diagnosis of major depressive disorder
Exclusion criteria
* History of psychosis * At high risk for suicide * Cognitive impairment * Current alcohol or substance abuse disorder * Does not own a telephone * Currently seeking psychiatric care * Pregnant or breastfeeding * Currently enrolled in a Group Health Cooperative disease management program * Terminal illness * Plans to leave Group Health Cooperative in less than a year * Does not speak English
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Symptom Checklist-20 Score at Baseline, 6 Months and 12 Months | Measured at Baseline, 6 Months, 12 months | SCL-20 is a 20 question checklist in which items are averaged to yield a potential score of 0 to 4 with higher scores indicating more severe depression symptoms. For the Primary Outcome (Outcome Measure #1 above), a scaled marginal model approach was used to jointly describe the four 12 month outcomes (SCL-20, HbA1c, systolic BP, LDL) and allowed use to test for a primary effect of the intervention among outcomes, scaling each outcome by its standard error, so the intervention effects could be interpreted as effect sizes. |
| LDL Cholesterol at Baseline and 12 Months | Measured at Baseline and 12 months | LDL Cholesterol was measured at Baseline and 12 months For the Primary Outcome (Outcome Measure #1 above), a scaled marginal model approach was used to jointly describe the four 12 month outcomes (SCL-20, HbA1c, systolic BP, LDL) and allowed use to test for a primary effect of the intervention among outcomes, scaling each outcome by its standard error, so the intervention effects could be interpreted as effect sizes. |
| Glycated Hemoglobin (HbA1c) at Baseline, 6 Months and 12 Months | Measured at Baseline, 6 months and 12 months | Glycated hemoglobin (HbA1c) was measured at Baseline, 6 months and 12 months For the Primary Outcome (Outcome Measure #1 above), a scaled marginal model approach was used to jointly describe the four 12 month outcomes (SCL-20, HbA1c, systolic BP, LDL) and allowed use to test for a primary effect of the intervention among outcomes, scaling each outcome by its standard error, so the intervention effects could be interpreted as effect sizes. |
| Combined Effect of Intervention on SCL-20, Systolic Blood Pressure, LDL and HbA1c | Baseline to 12 months | A scaled marginal model approach was used to jointly describe the four 12 month outcomes (SCL-20, HbA1c, systolic BP, LDL: all data submitted as Outcome Measures #2-5 below) and allowed use to test for a primary effect of the intervention among outcomes, scaling each outcome by its standard error, so the intervention effects could be interpreted as effect sizes.The model was estimated by iterating between estimation of the covariance associated with the outcomes and generalized-estimating equation estimation of scaled outcomes. Effect size is estimated as Cohen d effect size that was use for the depression outcome is the difference in change from baseline to 12 months in the intervention and usual care groups divided by the pooled base line standard deviation. Thus, a d of 0.25 indicates that one-quarter of a standard deviation separates the two means. Cohen has suggested that an effect size of 0.20 would be considered small, 0.50 medium and 0.80 large. |
| Systolic Blood Pressure at Baseline, 6 Months and 12 Months | Measured at Baseline, 6 Months, 12 months | Systolic Blood Pressure was measured at Baseline, 6 months and 12 months For the Primary Outcome (Outcome Measure #1 above), a scaled marginal model approach was used to jointly describe the four 12 month outcomes (SCL-20, HbA1c, systolic BP, LDL) and allowed use to test for a primary effect of the intervention among outcomes, scaling each outcome by its standard error, so the intervention effects could be interpreted as effect sizes. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Health Care Costs | Cumulative outpatient costs over 24 months | Mean total outpatient costs for 2 years post baseline adjusted for age, gender and previous 12 months of outpatient costs |
| Functional Impairment | Measured at Months 6, 12 months | Disability was measured by the Sheehan Disability scale which measures the extent to which health interferes with social, vocational and familial functioning each on a 0 to 10 Likert scale where 0 is not at all and 10 is extremely. This scale consists of 3 items which are averaged together to create the average disability score, which ranges from 0 to 10. |
Countries
United States
Participant flow
Recruitment details
Participants were recruited from may 2007 to october 2009 from 14 primary care clinics of Group Health Cooperative in Washington State
Pre-assignment details
Exclusion criteria: terminal illness, residence in long-term care facility, bipolar/schizophrenia, mental confusion suggesting dementia, pregnancy/breast-feeding, use of antipsychotic/mood stabilizer med, severe hearing loss, planned bariatric surgery within 3 months, ongoing psychiatric care
Participants by arm
| Arm | Count |
|---|---|
| Usual Care Patients in usual care arm were advised to consult with their primary care physician to receive care for depression and for diabetes, coronary artery disease or both. With the patient's permission their primary care physician was notified about their PHQ-9 score of 10 or greater and poor medical disease control. In addition primary care doctors received laboratory results at baseline, 6 and 12 months regarding HbA1c and at baseline and 12 months on fasting LDL | 108 |
| Care Management Intervention Case management intervention: patients assigned to care management received intervention visits with a medical nurse supervised weekly by both a psychiatrist and primary care doctor. Nurses provided psychoeducation, motivational interviewing, behavioral activation and problem solving and carefully tracked medications, side effects, PHQ-9 scores and blood pressure and lab results. Using a registry the medical supervisors recommended changes in medications that the nurse communicated to the patient's individual primary care doctor (who wrote all prescriptions). Nurses also worked with the patient to set self care goals regarding health behaviors such as increasing exercise or improving diet | 106 |
| Total | 214 |
Withdrawals & dropouts
| Period | Reason | FG000 | FG001 |
|---|---|---|---|
| Overall Study | Death | 2 | 1 |
| Overall Study | Lost to Follow-up | 14 | 11 |
Baseline characteristics
| Characteristic | Care Management Intervention | Usual Care | Total |
|---|---|---|---|
| Age, Categorical <=18 years | 0 Participants | 1 Participants | 1 Participants |
| Age, Categorical >=65 years | 25 Participants | 25 Participants | 50 Participants |
| Age, Categorical Between 18 and 65 years | 81 Participants | 82 Participants | 163 Participants |
| Age, Continuous | 57.4 years STANDARD_DEVIATION 10.5 | 56.3 years STANDARD_DEVIATION 12.1 | 56.8 years STANDARD_DEVIATION 11.3 |
| Region of Enrollment United States | 106 participants | 108 participants | 214 participants |
| Sex: Female, Male Female | 51 Participants | 60 Participants | 111 Participants |
| Sex: Female, Male Male | 55 Participants | 48 Participants | 103 Participants |
Adverse events
| Event type | EG000 affected / at risk | EG001 affected / at risk |
|---|---|---|
| deaths Total, all-cause mortality | — / — | — / — |
| other Total, other adverse events | 23 / 106 | 27 / 105 |
| serious Total, serious adverse events | 2 / 108 | 1 / 106 |
Outcome results
Primary
Combined Effect of Intervention on SCL-20, Systolic Blood Pressure, LDL and HbA1c
A scaled marginal model approach was used to jointly describe the four 12 month outcomes (SCL-20, HbA1c, systolic BP, LDL: all data submitted as Outcome Measures #2-5 below) and allowed use to test for a primary effect of the intervention among outcomes, scaling each outcome by its standard error, so the intervention effects could be interpreted as effect sizes.The model was estimated by iterating between estimation of the covariance associated with the outcomes and generalized-estimating equation estimation of scaled outcomes. Effect size is estimated as Cohen d effect size that was use for the depression outcome is the difference in change from baseline to 12 months in the intervention and usual care groups divided by the pooled base line standard deviation. Thus, a d of 0.25 indicates that one-quarter of a standard deviation separates the two means. Cohen has suggested that an effect size of 0.20 would be considered small, 0.50 medium and 0.80 large.
Time frame: Baseline to 12 months
Population: This was an intent to treat analysis of the 12 month SCL-20, HbA1c, LDL and systolic blood pressure outcomes
| Arm | Measure | Value (NUMBER) | Dispersion |
|---|---|---|---|
| Effect Size of Invervention Group to Standard Care | Combined Effect of Intervention on SCL-20, Systolic Blood Pressure, LDL and HbA1c | -1.57 unitless | 95% Confidence Interval 0.66 |
p-value: 0.00195% CI: [-2.12, -1.02]Scaled Marginal Model
Primary
Glycated Hemoglobin (HbA1c) at Baseline, 6 Months and 12 Months
Glycated hemoglobin (HbA1c) was measured at Baseline, 6 months and 12 months For the Primary Outcome (Outcome Measure #1 above), a scaled marginal model approach was used to jointly describe the four 12 month outcomes (SCL-20, HbA1c, systolic BP, LDL) and allowed use to test for a primary effect of the intervention among outcomes, scaling each outcome by its standard error, so the intervention effects could be interpreted as effect sizes.
Time frame: Measured at Baseline, 6 months and 12 months
Population: This was an intent to treat analysis of the 12 month SCL-20, HbA1c, LDL and systolic blood pressure outcomes
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| Effect Size of Invervention Group to Standard Care | Glycated Hemoglobin (HbA1c) at Baseline, 6 Months and 12 Months | Glycated Hemoglobin (HbA1c) at Baseline | 8.04 percent glycated hemoglobin | Standard Deviation 1.87 |
| Effect Size of Invervention Group to Standard Care | Glycated Hemoglobin (HbA1c) at Baseline, 6 Months and 12 Months | Glycated Hemoglobin (HbA1c) at 6 months | 7.87 percent glycated hemoglobin | Standard Deviation 1.93 |
| Effect Size of Invervention Group to Standard Care | Glycated Hemoglobin (HbA1c) at Baseline, 6 Months and 12 Months | Glycated Hemoglobin (HbA1c) at 12 months | 7.81 percent glycated hemoglobin | Standard Deviation 1.9 |
| Care Management Intervention | Glycated Hemoglobin (HbA1c) at Baseline, 6 Months and 12 Months | Glycated Hemoglobin (HbA1c) at Baseline | 8.14 percent glycated hemoglobin | Standard Deviation 2.03 |
| Care Management Intervention | Glycated Hemoglobin (HbA1c) at Baseline, 6 Months and 12 Months | Glycated Hemoglobin (HbA1c) at 6 months | 7.42 percent glycated hemoglobin | Standard Deviation 1.32 |
| Care Management Intervention | Glycated Hemoglobin (HbA1c) at Baseline, 6 Months and 12 Months | Glycated Hemoglobin (HbA1c) at 12 months | 7.33 percent glycated hemoglobin | Standard Deviation 1.21 |
95% CI: [-0.85, -0.27]Generalized-Estimating-Equation Model
Primary
LDL Cholesterol at Baseline and 12 Months
LDL Cholesterol was measured at Baseline and 12 months For the Primary Outcome (Outcome Measure #1 above), a scaled marginal model approach was used to jointly describe the four 12 month outcomes (SCL-20, HbA1c, systolic BP, LDL) and allowed use to test for a primary effect of the intervention among outcomes, scaling each outcome by its standard error, so the intervention effects could be interpreted as effect sizes.
Time frame: Measured at Baseline and 12 months
Population: This was an intent to treat analysis of the 12 month SCL-20, HbA1c, LDL and systolic blood pressure outcomes
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| Effect Size of Invervention Group to Standard Care | LDL Cholesterol at Baseline and 12 Months | LDL cholesterol at Baseline | 109.4 mg/dL | Standard Deviation 36.7 |
| Effect Size of Invervention Group to Standard Care | LDL Cholesterol at Baseline and 12 Months | LDL cholesterol at 12 months | 101.4 mg/dL | Standard Deviation 36.6 |
| Care Management Intervention | LDL Cholesterol at Baseline and 12 Months | LDL cholesterol at 12 months | 91.9 mg/dL | Standard Deviation 36.7 |
| Care Management Intervention | LDL Cholesterol at Baseline and 12 Months | LDL cholesterol at Baseline | 106.8 mg/dL | Standard Deviation 35.4 |
95% CI: [-17.5, -0.8]Regression, Linear
Primary
Symptom Checklist-20 Score at Baseline, 6 Months and 12 Months
SCL-20 is a 20 question checklist in which items are averaged to yield a potential score of 0 to 4 with higher scores indicating more severe depression symptoms. For the Primary Outcome (Outcome Measure #1 above), a scaled marginal model approach was used to jointly describe the four 12 month outcomes (SCL-20, HbA1c, systolic BP, LDL) and allowed use to test for a primary effect of the intervention among outcomes, scaling each outcome by its standard error, so the intervention effects could be interpreted as effect sizes.
Time frame: Measured at Baseline, 6 Months, 12 months
Population: This was an intent to treat analysis of the 12 month SCL-20, HbA1c, LDL and systolic blood pressure outcomes
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| Effect Size of Invervention Group to Standard Care | Symptom Checklist-20 Score at Baseline, 6 Months and 12 Months | SCL-20 at Baseline | 1.65 scores on a scale | Standard Deviation 0.6 |
| Effect Size of Invervention Group to Standard Care | Symptom Checklist-20 Score at Baseline, 6 Months and 12 Months | SCL-20 at 6 mos | 1.26 scores on a scale | Standard Deviation 0.72 |
| Effect Size of Invervention Group to Standard Care | Symptom Checklist-20 Score at Baseline, 6 Months and 12 Months | SCL-20 at 12 mos | 1.14 scores on a scale | Standard Deviation 0.66 |
| Care Management Intervention | Symptom Checklist-20 Score at Baseline, 6 Months and 12 Months | SCL-20 at Baseline | 1.74 scores on a scale | Standard Deviation 0.59 |
| Care Management Intervention | Symptom Checklist-20 Score at Baseline, 6 Months and 12 Months | SCL-20 at 6 mos | 0.84 scores on a scale | Standard Deviation 0.68 |
| Care Management Intervention | Symptom Checklist-20 Score at Baseline, 6 Months and 12 Months | SCL-20 at 12 mos | 0.83 scores on a scale | Standard Deviation 0.68 |
p-value: <0.00195% CI: [-0.56, -0.26]Generalized-Estimating-Equation Model
Primary
Systolic Blood Pressure at Baseline, 6 Months and 12 Months
Systolic Blood Pressure was measured at Baseline, 6 months and 12 months For the Primary Outcome (Outcome Measure #1 above), a scaled marginal model approach was used to jointly describe the four 12 month outcomes (SCL-20, HbA1c, systolic BP, LDL) and allowed use to test for a primary effect of the intervention among outcomes, scaling each outcome by its standard error, so the intervention effects could be interpreted as effect sizes.
Time frame: Measured at Baseline, 6 Months, 12 months
Population: This was an intent to treat analysis of the 12 month SCL-20, HbA1c, LDL and systolic blood pressure outcomes
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| Effect Size of Invervention Group to Standard Care | Systolic Blood Pressure at Baseline, 6 Months and 12 Months | Systolic Blood Pressure at Baseline | 131.9 mmHg | Standard Deviation 17 |
| Effect Size of Invervention Group to Standard Care | Systolic Blood Pressure at Baseline, 6 Months and 12 Months | Systolic Blood Pressure at 6 months | 133.5 mmHg | Standard Deviation 20.4 |
| Effect Size of Invervention Group to Standard Care | Systolic Blood Pressure at Baseline, 6 Months and 12 Months | Systolic Blood Pressure at 12 months | 132.3 mmHg | Standard Deviation 17.4 |
| Care Management Intervention | Systolic Blood Pressure at Baseline, 6 Months and 12 Months | Systolic Blood Pressure at Baseline | 135.7 mmHg | Standard Deviation 18.4 |
| Care Management Intervention | Systolic Blood Pressure at Baseline, 6 Months and 12 Months | Systolic Blood Pressure at 6 months | 131.9 mmHg | Standard Deviation 15.2 |
| Care Management Intervention | Systolic Blood Pressure at Baseline, 6 Months and 12 Months | Systolic Blood Pressure at 12 months | 131.0 mmHg | Standard Deviation 18.2 |
95% CI: [-6.9, 0.1]Generalized-Estimating-Equation Model
Secondary
Functional Impairment
Disability was measured by the Sheehan Disability scale which measures the extent to which health interferes with social, vocational and familial functioning each on a 0 to 10 Likert scale where 0 is not at all and 10 is extremely. This scale consists of 3 items which are averaged together to create the average disability score, which ranges from 0 to 10.
Time frame: Measured at Months 6, 12 months
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| Effect Size of Invervention Group to Standard Care | Functional Impairment | 6-month follow-up | 4.2 units on a scale | Standard Deviation 2.6 |
| Effect Size of Invervention Group to Standard Care | Functional Impairment | 12-month follow-up | 4.5 units on a scale | Standard Deviation 2.9 |
| Care Management Intervention | Functional Impairment | 6-month follow-up | 3.7 units on a scale | Standard Deviation 3.2 |
| Care Management Intervention | Functional Impairment | 12-month follow-up | 3.8 units on a scale | Standard Deviation 3 |
Comparison: Disability outcomes were measured at six and 12 months after randomization using linear regression models adjusted for baseline values. The disability models combined information across time points and were estimated using general estimating equations to account for correlation. All analyses were based on intent to treat principlesp-value: <0.01general estimating equations
Secondary
Health Care Costs
Mean total outpatient costs for 2 years post baseline adjusted for age, gender and previous 12 months of outpatient costs
Time frame: Cumulative outpatient costs over 24 months
| Arm | Measure | Value (LEAST_SQUARES_MEAN) | Dispersion |
|---|---|---|---|
| Effect Size of Invervention Group to Standard Care | Health Care Costs | 21,513 US dollars | Standard Deviation 9201 |
| Care Management Intervention | Health Care Costs | 20,918 US dollars | — |