Building Resiliency and Increasing Community Hope (BRICH)

B-RICH proposes to pilot and then conduct a randomized, single blind trial of a psychosocial intervention called a "Resiliency Class" (RC), to provide depression education and health promotion to individuals with depressive symptoms, by training non-professionals to offer this class to clients receiving services within diverse community settings (e.g. health care clinic, church, community advocacy organization, social services agency). This class is not designed to be therapy, but rather is designed as a class that is informed on cognitive behavioral therapy (CBT) principles used to address depressive symptoms, on how to improve mood, and to enhance resiliency in the face of stress.

The Resiliency Class is a manualized, CBT informed, seven session, psycho-educational intervention designed to be delivered by non-licensed teachers in both English and Spanish to address depressive symptoms. Each session is designed to be between 90-120 minutes. Some of the goals of the RC are: 1) to provide psycho-education on mood and depressive symptoms; 2) to provide this education through teachers who are not health or mental healthcare professionals in order to improve access to depression education and to offer approaches to improve depressive symptoms; 3) to deliver this class in non-health care or mental health care settings like churches or in trusted community locations (e.g. parks and recreation, primary care, social services agencies, job retraining centers, senior centers) instead of as "therapy" or in mental health settings so that individuals would not feel stigmatized; 4) to design the manual so that it will be usable by class participants with low literacy levels in both English and Spanish.

The specific aims for this study are to:

1. Pilot a revised resiliency class intervention led by non-professionals in 20 individuals with depressive symptoms for acceptability;
2. Revise the resiliency class intervention based on therapist's experience from the pilot based on feedback from both the instructors and the pilot class participants;
3. Conduct a randomized, single blind trial of the resiliency class (study group) versus case management social services assessment and referrals (control group) to assess their differential impact on depressive symptoms in 400 study participants (200 in each arm of the study).

There will be several steps to the implementation of the B-RICH Study. Phase I

1. We will conduct a pilot of the Resiliency Class Manual with 2 instructors and 20 individuals divided in two classes in the fall of 2012. We will screen 100 clients at Healthy African American Families II and the 1st African Presbyterian Church located in South Los Angeles to detect approximately 70 potential participants with depressive symptoms using the Patient Health Questionnaire 2 (PHQ-2). Participants who endorse one item on the PHQ-2, are 18 years or older, can attend six 90-120 minute weekly Resiliency Classes, and have a PHQ-8 score less than 15 (on the baseline interview) will be eligible to enroll in the Resiliency Class. These classes will be delivered by a non-licensed, individual who would be the equivalent of a community health worker. The classes will be supervised by Bowen Chung and Jeanne Miranda. In this phase of the study, we will pilot the participant screener, baseline interview, and follow-up interview for feasibility and to assess the amount of time each instrument will require.
2. Based on the class instructor's feedback and focus groups with the pilot class study participants, the PI, Co-PI, and project manager will then re-revise the manual for a final version which will be utilized in the randomized trial of the resiliency class (study group) versus 2 case management calls (control group) that will be started in January 2013.

Phase II 4) We will conduct a randomized trial of the Resiliency Class (study group) compared to telephone case management social services assessment (control group) and referrals,to ascertain their differential impact on depressive symptoms between January 2013 and January 2014.

4a) We will screen 1000 clients at the partner agencies for potential study participation. Entry criteria will include those individuals age 18 or older who endorse at least one item on the Patient Health Questionnaire- 2 (PHQ-2), have a reliable way to contact them by phone, and are able to attend our six, weekly sessions of the Resiliency Class. We estimate that screening will take place between January 2013 and June 2013.

4b) We plan on enrolling 400 clients into the study. 4c) Enrolled participants will then be offered a 45 minute baseline interview, that will collect data on income, housing, functional status (short form health survey - 12 item or SF-12); medical conditions; mental health conditions; alcohol and substance use; use of health and mental health services; knowledge, attitudes, and beliefs about depression; housing; food insecurity; coping strategies; medication use; satisfaction with services; and health insurance status. We will also collect contact information as well as information about 3 people who will always know how to reach the participant in case their direct contact information becomes invalid due to a disconnected phone or a change in address. The baseline interview will be conducted at the time of recruitment in the community at study agencies after determining study eligibility or on the phone either on a laptop computer or using pencil and paper by a study research assistant. We will exclude enrolled participants with a PHQ-8 score of 20 or greater (severe depression) and refer them to local mental health agencies in the area.

4d) After being enrolled in the study and completing the baseline interview, participants will either be randomized to the control group (two case management phone calls) or to the study group (a seven session resiliency class).

4e) Participants randomized to the control group will receive two calls from a case manager of the participating partner agencies within the first two months of their enrollment. These phone calls will ask the study participants if they should need referrals in the local community for healthcare, mental health care, substance use, job retraining, and social services. If the participants are in need of referrals, they will be offered a phone number and address for an agency that will meet their needs.

4f) Participants randomized to the study group will then be offered enrollment in a Resiliency Class that will take place at the agency they were screened at.

4g) The study group will have staggered starting times that will depend on success of participant enrollment. Each weekly, 90-120 minute class will have 10-12 study participants. Each participant in the Resiliency Class will receive a copy of the study manual at the initial class. In addition, the class instructor will provide light snacks and refreshments to participants at each class. Class instructors will be a combination of study staff as well as staff from partner agencies.

4h) Study Participants in both arms of the study will receive a 3-month, telephone follow-up interview assessing measures on the baseline interview that should take about 45 minutes.

4i) Research staff interviewing participants in baseline and follow-up interviews will be blinded as to the intervention status of the study participants.

5) We will conduct data analysis on the primary outcome measures, depressive symptoms as assessed by the PHQ-8. If the Resiliency Class shows a greater impact on diminishing depressive symptoms, the primary study outcome measure, than the case management phone calls, we will contact all participants in the case management arm of the study and offer the resiliency classes to those participants free of charge, as well as offer training to all staff from agencies interested in implementing the resiliency class intervention.

Phase I: Resiliency Class Pilot Intervention The goal of this step of the study is to pilot the Resiliency Class in order to assess for feasibility and to determine if the intervention might need additional changes to improve delivery of the intervention from the perspective of the class instructors.

Screening and Recruitment of Pilot Participants Study research assistants will use the screening script in to recruit potential participants at the pilot sites: Healthy African American Families II and the 1st African Presbyterian Church. A table with a sign promoting the study will be set up at the recruitment site and a study brochure will be distributed to individuals interested in the study. Research staff will then obtain oral consent and screen potential participants one-on-one by telephone or in person to determine eligibility. The screening instrument contains questions on race/ethnicity, age, gender, language, and depressive symptoms. As part of the screener, we will use the PHQ-2 to assess probable depression. It will take no more than 10 minutes to complete the entire screener. Screeners that are administered in person will be done in a private room at the recruitment site. All measures will be administered in a private room on a one-to-one basis at the recruitment site. Potential study participants will be offered a monetary incentive for being screened for the study as well as a resource guide that lists agencies that provide health care, mental health, substance use, social services, child welfare services, and prison re-entry programs in South Los Angeles.

B-RICH Phase I, Pilot Inclusion Criteria

1. Age 18 or older
2. Endorse one item on the PHQ-2
3. Able to be contacted by phone
4. Able to participate in the six, weekly 90-120 minute sessions of the Resiliency Classes
5. English or Spanish speaker

B-RICH, Phase I Pilot Exclusion Criteria

1. Under age 18 years
2. Endorse no items on PHQ-2 screener
3. PHQ-8 score of 15 or greater on the baseline interview

3) Unable to be contacted by phone 4) Unable to attend the weekly resiliency classes

Pilot Consenting Process If participants are eligible for the study, the UCLA research assistant will offer the participant the IRB approved Pilot Non-Randomized consent form at the time of screening and will explain the full study procedures to the potential study participant. UCLA Research assistants will explain the IRB-approved consent and assure that the participant will have enough time to read and review the consents before signing. The potential participant will be informed verbally in the consent form, that the study involves baseline and follow-up interviews, as well as a manualized class that will teach the participant how to improve their mood, and that this class will be audio tape recorded for review by Bowen Chung and / or Jeanne Miranda to assess the instructors' skill and to provide feedback on how to improve teaching the class.

The research assistants will determine each potential participant's level of comprehension of the material presented by asking the participants to summarize in their own words how they understand the study procedures, what will be required of them, potential risks and benefits of study participation, the voluntary nature of the study, and their right to withdraw at any time. They will be encouraged to ask questions about the study, study procedures, as well as interviews. Potential participants will be allowed to consent at the initial visit or they will offered up to one week to consider their participation and to either mail or fax their consent form back to the UCLA research assistant so that they will have an opportunity to think about study participation and / or discuss potential study participation with family members or friends. All study participants will understand that their acceptance or refusal to participate in the study will in no way affect their ability to receive services or care at the agency they were screened at or any other B-RICH partner agency. In addition, they will be informed that they will be able to withdraw their consent and discontinue study participation at any time without any adverse consequences. If the potential participant has any questions, they can call and request to meet with the study investigators and the project manager at any time.

Pilot Baseline Assessments If they agree to participate in the study, participants will be asked to take part in the 45 minute baseline interview either in person when the consent has been signed or over the phone at a later time and date. Baseline interview will be conducted by a UCLA Research Assistant.

Pilot Resiliency Class After completion of the interview, the participant will be offered and assigned a time and date to attend the classes at either Healthy African American Families II or at the 1st African Presbyterian Church. In order to facilitate their participation in the class, they will be offered bus tokens for transportation to attend the first three classes. Each class will have up to 10 participants.

The instructors for the class will be a UCLA research study staff member and a staff member from Healthy African American Families II or the 1st African Presbyterian Church.

Classes will be 90-120 minutes in duration and will be held weekly for six weeks. At the classes, participants will be offered a copy of the Resiliency Class Manual. And at each class, light snacks and refreshments will be offered.

Each class will be audio recorded on a digital audio recorder for review by Bowen Chung and / or Jeanne Miranda to provide feedback on the class instructor's technique of delivering this educational intervention. The audio tapes will be uploaded to a HIPAA compliant, password protected website housed on a secure server at the UCLA Center for Health Services and Society. After review of the audio taped sessions, the data for these sessions will be destroyed. These audio recordings will not be used for research purposes, but merely to assess the instructor's skill and to provide feedback on how to improve teaching the class.

The Pilot Resiliency Class manual covers the following topics:

Session 1 - "What Affects Your Mood and Resilience" Session 2 - "Pleasant Activities Can Help Improve Your Mood and Make You Resilient" Session 3 - "What Gets In The Way of Pleasant Activities: Harmful Thoughts and How to Change Them" Session 4 - "How to Increase Your Resilience Through Support from Others" Session 5 - "My Personal Resiliency Plan: Goal Setting" Session 6 - "Celebrate Your Resiliency: Graduation"

Pilot Post Class Assessment Focus Groups with pilot study participants We will invite all pilot study participants to participate in one of two post pilot study focus groups of 90 minutes. Focus groups will be conducted by Bowen Chung. All focus groups will be held at either 1st African Presbyterian Church or Healthy African American Families II and audio-taped for later transcription and analysis. The goal of these focus groups will be to see from the pilot participant's perspective: what the experience was like to participate in the Resiliency Class; what about the class worked well; what aspects of the class needed improvement; and how they would suggest the class be improved.

Pilot Two month follow-up interview: Two months after the completion of the class, the pilot participants will be asked to take the follow-up interview mostly to assess the feasibility and length of time it takes to administer the study instrument. The post-class assessment will be administered over the phone or in person.

Revisions to Pilot Resiliency Class manual and to study instruments: During class implementation and after the completion of the pilot, revisions will be made to the manual and to the study instruments based on feedback from the instructors and the UCLA study research staff.

Recruiting Additional Agencies for the randomized trial: During phase I, we will also recruit 3 additional agencies for the randomized trial phase of this study.

Phase II - Randomized trial of the Resiliency Class In the randomized trial phase, we expect to screen 1000 adults for depressive symptoms using the PHQ-2 to obtain about 400 potential participants for study entry. Our study plans on having between 200 and 300 eligible adults to consent to participate in a randomized trial of our seven session resiliency class (study group) versus two case management phone calls (control group). We will assess the resiliency class's differential impact (versus the case management calls) on study participants' depressive symptoms (primary outcome measure) as well as other study measures (e.g. SF-12, substance use, use of health and mental health services, physical activity).

Screening and Recruitment of Randomized Trial Participants UCLA research assistants will recruit potential participants from community-based agencies. A table with a sign (promoting the study will be set up at the recruitment site and a study brochure will be distributed to individuals interested in the study. Research staff will screen potential participants one-on-one by telephone or in person to determine eligibility. The screening instrument contains questions on race/ethnicity, age, gender, language, and depressive symptoms. As part of the screener, we will use the Patient Health Questionnaire 8 (PHQ-8) to assess probable depression. When the PHQ-8 is administered by phone, it will be done on a one-on-one basis. It will take no more than 15 minutes to complete the entire screener. Screeners that are administered in person will be done in a private room at the recruitment site. All measures will be administered in a private room on a one-to-one basis at the recruitment site. Potential study participants will be offered a monetary incentive for being screened for the study as well as a resource guide that lists agencies that provide health care, mental health, substance use, social services, child welfare services, senior center, and prison re-entry programs in South Los Angeles.

B-RICH Phase II, Randomized Trial Inclusion Criteria

1. Age 18 or older
2. Endorse one of the first two items on the PHQ-8: "In the past two weeks, how often have you been bothered by 1) little interest or pleasure in doing things, 2) feeling down depressed or hopeless."
3. Able to be contacted by phone (voice or text message), e-mail, or Facebook.
4. Able to participate in the seven, weekly 90-120 minute sessions of the Resiliency Classes
5. English or Spanish speaker

If score is between 15-19, we will allow them to enroll in the study if they are in existing treatment. If they are not, we will make a referral and measure their mood throughout the intervention. If at the end of the class, anyone is still in the range of 15-19 we will highly recommend that they get treatment and we will make a referral to a local mental health agency.

Patient Engagement Approach: The class instructor will do some follow-up with participants who enter the study with a PHQ-8 score of 15-19 to see if they went to the referral agency, ask if they saw a provider, are they receiving treatment, if not what were their barriers.

The PHQ-8 will be administered during class 1, 4, and 7 and 3-month follow-up of the intervention to track whether depressive symptoms are improving or worsening.

B-RICH Phase II, Randomized Trial Exclusion Criteria

1. Under age 18 years
2. PHQ-8 score of 20 or greater on the baseline interview
3. PHQ-8 score of 15 or greater and not currently in care and unwilling to get a referral for care
4. Currently homeless
5. Has a prior diagnosis of Bipolar Disorder
6. Has a prior diagnosis of Schizophrenia
7. Current alcohol and substance abuse
8. Unable to attend the weekly resiliency classes
9. Does not currently have a phone, an email address, or a Facebook profile

Consenting Process: The same consenting process will be used as described in the pilot above.

Baseline Assessments If they agree to participate in the study at the time of screening, participants will be asked to take part in the 45 minute baseline interview either in person when the consent has been signed or over the phone at a later time and date. For those who do not consent at the time of screening but within a week of being screened, the baseline interview will be administered over the phone. Baseline interview) will be conducted by a UCLA Research Assistant. Each participant will be offered a monetary incentive for completion of the baseline interview. Payment will offered in person to those who elect to take the baseline interview in person. Participants who elect to take the interview over the phone will have the choice of picking up their incentive payment at the UCLA Center for Health Services and Society or mailed to them at an address of their choice.

Randomization As part of the Study Recruitment packet, there will be a sealed envelope that will indicate whether the study participant has been randomized to either the resiliency class or the case management phone calls. If the enrollment and baseline survey are completed in person, the study staff that consents and completes the baseline interview with the study participant will open the envelope to determine which condition the study participant has been assigned: control or study group. If the baseline interview is conducted over the phone, the envelope assigning the study participant will be opened by research staff and inform them of their assignment over the phone.

In order to assure that that research staff are unaware of the intervention condition at follow-up, we will re-assign study participants to different research staff person to conduct the 3 month follow-up interview.

Case management social services assessment and referrals (control group) Study participants randomized to the case management phone calls will be told that they will receive two calls over two months by study staff to offer them referrals based on their perceived need for services to local health care, mental health, substance use, social services, child welfare, housing, and food. In addition, study participants in this arm of the study will be told that if the study determines that the resiliency classes are better at improving mood than the case management calls, they will receive a call to invite them to participate in resiliency classes for free at the B-RICH partner agencies.

Resiliency Classes (study group) Study participant randomized to the Resiliency classes will be offered and assigned a time and date to attend the classes at Healthy African American Families II, 1st African Presbyterian Church, Los Angeles Biomedical Research Institute, and at the additional partner agencies that will be recruited. In order to facilitate their participation in the class, they will be offered bus tokens for transportation to attend the first three classes. Each class will have up to 10-12 participants.

The instructors for the class will be a UCLA research study staff member and a staff member from Healthy African American Families II or the 1st African Presbyterian Church.

Classes will be 90-120 minutes in duration and will be held weekly for seven weeks. At the classes, participants will be offered a copy of the Resiliency Class Manual. And at each class, light snacks and refreshments will be offered.

Each class will be audio recorded on a digital audio recorder for review by Bowen Chung and / or Jeanne Miranda to provide feedback on the instructor's class technique of delivering this educational intervention. The audio tapes will be uploaded to a HIPAA compliant, password protected website housed on a secure server at the UCLA Center for Health Services and Society. After review of the audio taped sessions, the data for these sessions will be destroyed. These audio recordings will not be used for research purposes, but merely to assess the instructors' skill and to provide feedback on how to improve teaching the class.

The Resiliency Class manual covers same topics described above.

3 month follow-up interview Participants will be asked to participate by phone in a 45 minute follow-up interview 3 months after being enrolled in the study.

Data Analysis

Phase I Quantitative analysis: Descriptive analysis will be conducted using our screener measures. Univariate distributions will be examined to identify outliers and assess their influence on the analyses. Issues to be considered include errors in data input, scoring or other sources of invalidity in the measurements. The data will be carefully inspected to determine the need for scale transformations and to identify missing data, outliers, or other unusual features. Assumptions of normality will also be evaluated. Change scores in depressive symptoms from baseline to post-class interviews will be assessed by 95% confidence intervals. Due to small sample size (n=20) in this pilot phase, no formal testing hypotheses will be conducted.

Qualitative / Focus Group data: Qualitative materials will be transcribed by a confidential transcriber and de-identified by a confidential research assistant. The research assistant will transform field code identifiers into analytic code identifiers and maintain the code list in a secure file. Each dataset - discussion notes, recorded feedback, and interview responses - will be entered into a file in the project Dedoose file (www.dedoose.com), a password-protected online site, using the UCLA Dedoose program (www.dedoose.com), software designed specifically to facilitate mixed-methods analysis. We will then use constructive grounded theory analysis to develop coding categories based on the themes embedded in the datasets that fall within domains of interest. (12) In particular, we will be examining the qualitative data for themes describing: 1) what their experiences were like with the resiliency class; 2) what aspects of the resiliency class worked to improve their mood; 3) what aspects of the resiliency class did not improve their mood; 4) what they felt needed to be improved about the resiliency class and the manual; 5) how they suggest improving the resiliency class and manual.

Phase II Analysis of Baseline and 3 month follow-up interview data The main analyses will be intent-to-treat comparing main effects of CBT intervention versus case management calls control. We hypothesizes that CBT class will have a greater impact on reducing depressive symptoms (PHQ-8) than control. We will also examine the intervention effects on other study measures (e.g. MCS-12, PCS12, substance use, use of health and mental health services). We will use an ANCOVA model that specifies the outcome measure observed at the 3 months as the dependent variable (e.g. PHQ-8), treatment group as the independent variable, controlled for the baseline measure for the same outcome as the covariate. For binary outcomes (e.g. substance use or prior use of antidepressant medication) we will fit a logistic regression model.

General Analytic Strategies Univariate distributions will be examined to identify outliers and assess their influence on the analyses. Issues to be considered include errors in data input, scoring or other sources of invalidity in the measurements. The data will be carefully inspected to determine the need for scale transformations and to identify missing data, outliers, or other unusual features. To improve the precision of the estimated intervention effect, we will conduct a series of bivariate analyses to identify the potential covariates to be considered for a multiple regression model. Confounding will be assessed by comparing the unadjusted coefficient for treatment condition with the adjusted coefficient. Assumptions of normality will also be evaluated. For example, we will explore various transformations (e.g., log) for number of visits due to the often skewed distribution of event counts. Smearing estimates will be used, if necessary, for retransformation, applying separate factors for each intervention group to ensure consistent estimates (13) In general, we leave open the possibility of transforming variables with non-normal distributions (14) In certain cases, the outcome of interest will have a skewed distribution, either because there will be many zeros (e.g. days of work missed due to illness) and/or because the distribution of observed values has a very long tail (e.g. services utilization). In such cases, we will draw upon statistical models developed for these type of data, in particular a two-part model to separately handle zero values and a skewed distribution among non-zeros (13). Tobit-type sample selection models (15); split-sample techniques to distinguish between different functional forms and to avoid overfitting (13); and dynamic models, such as episodes of care analytic models, duration analysis, and count models (16).

To improve the precision of the estimated intervention effect, we will conduct a series of bivariate analyses to identify the potential covariates to be considered for a multiple regression model. Confounding will be assessed by comparing the unadjusted coefficient for treatment condition with the adjusted coefficient. We will consider multiple imputation to account for unit level non-response (17) We will inspect patterns of missing data. If non-response appears to be non-ignorable we will employ statistical techniques such as pattern mixture models and related methods ; and otherwise imputation strategies will assume that non-response is ignorable (i.e. missing at random). (13, 17-19)

Description of measures B-RICH PARTICIPANT STUDY SCREENER Age PHQ-2 Whether they know provider or staff member they see at agency

BASELINE INTERVIEW Health status Gender Marital status Children under age 18 living at home Health Insurance coverage Education Employment Born in the U.S. Ethnicity Housing PHQ-8 Function: SF-12 Fordyce Emotions Questionnaire Brief Resiliency Scale Use of Services (health, mental health, substance use, social services, religious, parks and recreation) Current Medication Use Satisfaction with Services Beliefs about people with depression Beliefs about treating depression Chronic Conditions (medical, smoking, weight, height, physical activity) Current and Past Use of Alcohol / Drugs Life Difficulties in Past 6 months Coping Strategies Social Supports Employment, work missed for those employed Income Income, income from governmental sources Food security

3 MONTH FOLLOW-UP INTERVIEW Demographics Date of birth Marital status Housing PHQ-8 Function: SF-12 Fordyce Emotions Questionnaire Brief Resiliency Scale Use of Services (health, mental health, substance use, social services, religious, parks and recreation) Current Use Medications Beliefs about people with depression Beliefs about treating depression Chronic Conditions (medical, smoking, weight, height, physical activity) Current and Past Use of Alcohol / Drugs Life Difficulties in Past 6 months Coping Strategies Social Supports Employment, work missed for those employed Income, income from governmental sources Food security

Phase II RCT:

All data collection was completed in 2014. Study participants were screened (n=310) and enrolled (n=168) from community-based organizations located in South Los Angeles including a Women, Infants, and Children (WIC) program; a senior center; a church; a faith-based, low-income senior housing facility; a health advocacy organization, a YMCA, and a federally qualified health center. 84 participants were randomized into the case management assessment and referral arm and 84 into the resiliency class intervention.

Data analyses are currently under way and we are preparing a manuscript with the findings for publication in the next several months.