The Impact of Disclosing Personalized Depression Risk Information on High-risk Individuals' Outcomes

Rationale: The proposed study stems from a CIHR project and is driven by the results of knowledge translation (KT) activities. The investigators developed the first sex-specific risk prediction models for major depressive episode (MDE) using data from over 10,000 Canadians. The models include age, personal and family history of MDE, childhood trauma, and predictors that may change over time. Using the models, individuals can answer the question: what is my risk of having a MDE in the next 4 years? Knowing this personalized risk information may assist health professionals in communicating with patients and planning preventive actions; disclosing the information may also empower consumers to actively engage in self-help. In the past 5 years, the KT activities with over 500 stakeholders informed that the investigators need to clearly delineate the benefits (enhanced risk perceptions and self-help) and potential risk (increased psychological distress and impaired function) of disclosing the information before implementing the tools.

The goal of this study is to produce evidence about the benefits and potential risks associated with risk disclosure (i.e., the expected outputs). The primary objectives are to compare those who do and do not receive personalized depression risk information in (1) accuracy of risk perceptions, the use of self-help strategies, and (2) changes in psychological distress and functional measures over one year among those who are at high risk of MDE. The secondary objective is to examine the differential impacts of the disclosure on the outcomes by demographics and levels of baseline risk. To achieve these objectives, the investigators propose a 3-year randomized controlled trial (RCT) with an embedded qualitative component.

The proposed study will be a RCT with an embedded qualitative component. The RCT will have one interventional arm (receiving personalized depression risk information) and one control arm. The personalized depression risk will be generated using the sex-specific prediction algorithms for MDE that we developed in Canadians aged 18+ years old. Because the prediction algorithms are sex-specific, the investigators propose to recruit 350 men and 350 women at baseline. After baseline assessment for eligibility, participants will be randomized into intervention and control groups, in men and women separately. To obtain in-depth information about how the personalized depression risk information is processed by participants and how the information affects them emotionally, the research team will conduct qualitative interviews 1 month after the personalized risk information is disclosed. To understand how the personalized risk information affect participants' health behaviors, the investigators will conduct another round of qualitative interviews at 12-month.

Recruitment: The target population of future preventive studies are high risk individuals in the general population, who reside in different regions across the country. For the proposed study, a feasible approach for recruiting eligible participants is the random digit dialing method (RDD). In the past 10 years, we have used the RDD for recruitment in CIHR-funded longitudinal studies and an ongoing national RCT funded by the Movember Foundation. Recruitment, screening and baseline assessment will be done by a telephone interview firm identified through a competitive bidding process.

A random sample of land line and cell phone numbers will be selected. When a household is reached, the person who is 18+ years will be assessed for eligibility. If a household has 2+ persons aged 18+ years, one will be randomly selected. The interviewers will explain the study objectives and procedures and answer questions. Potential participants will be ensured about confidentiality, that participation is voluntary and that they may withdraw at any time. Oral consent will be obtained before assessment of eligibility. Outcome measures will be assessed at baseline, 6 and 12-month.

Baseline Assessment and Randomization Screening: Once a potentially eligible participant is identified, the interviewer will confirm the participant's age and administer the World Health Organization's Composite International Diagnostic Interview - Short Form for Major Depression (CIDI-SFMD) and the sex-specific prediction algorithms. Interviewees who are in a MDE or are below the risk thresholds based on the risk calculators, will be excluded. Individuals with MDE will be encouraged to contact family doctors and information about local mental health resources will be provided. For those who are at low risk, the web site of the sex-specific risk prediction algorithms (www.predictingdepression.com) will be provided so they may monitor their risk in the future.

Baseline assessment: In eligible participants, the interviewer will administer the Non-specific Psychological Distress Scale (K10), Self-help Strategy Use Scale (SSUS), and ask questions about absenteeism and perceived risk of MDE. Our pilot study showed that the screening and baseline assessment take 20 to 25 minutes.

Randomization will be carried out in men and in women separately. Eligible participants who complete the baseline telephone interview will be randomized into intervention and control groups. The randomization will be conducted and managed by the project coordinator affiliated with the project. 350 random numbers (between 0 and 1) will first be generated for men and women separately using Excel ("RAND()") for 350 study identification numbers (ID), ranging from 1 to 350. The 350 random numbers will then be sorted at the descending order and categorized into 2 equal groups:

• the ID numbers in the first group (n = 175) will be allocated to the control group
• the ID numbers in the second group (n = 175) will be allocated to the intervention group

Intervention and Control: For the participants in the intervention group, the personalized risk will be disclosed and the interviewer will inform them that they will be contacted again at 6 months and 12 months. The interest in receiving such personalized depression information has been confirmed by our recent pilot study using the same sampling method. The team's pilot data (n = 200) showed that 100% of high-risk individuals were interested in knowing their risks. Participants in the intervention group will also be informed that some may be contacted in one month for a 30-minute qualitative interview. A package including the following materials will be mailed to intervention participants: (1) thank-you letter, (2) a website containing general information about MDE, and self-help strategies. (3) $20 incentive as appreciation of their participation. For participants in the control group, the project coordinator will inform them that they will be contacted again at 6 and 12 months. Their personal risks will be provided at the 12-month interview. The control group participants will receive the same package as those in the intervention group.

Blinding and follow-up assessments: The telephone interviewers will securely transfer encrypted baseline data to the PI on a bi-weekly basis. The group assignment data will be transferred in a separate file. The follow-up assessments will be conducted at the telephone interview laboratory at the Mathison Mental Health Research \& Education Center. One month before the scheduled follow-up interviews, letters will be sent to participants to remind them of the upcoming interview. After the 12-month interview, participants' group status will be linked with interview data by study ID numbers.

Over the study period, investigators will be blinded to participants' group status. The interviewers who conduct randomization, will not be involved in follow-up interviews. The interviewers who conduct the follow-up interviews in Calgary will not have access to participants' group status. Given the description of study objectives, participants may know their group status. Therefore, it is possible that some participants in the control group may try to find more information about personalized depression risk. At the follow-up assessments, the research staff will ask if they have used any risk prediction tools over the study period. At the follow-up assessments, if participants develop a MDE, they will be encouraged to contact family doctors and information about local mental health resources will be provided.

Qualitative Interviews: To obtain in-depth information about how disclosing personalized depression risk affects participants' decision processes, mental health and health behaviors, the investigators will conduct two rounds of qualitative interviews via telephone, 1 month after these participants receive the personalized depression risk and at 12 months. Each will include a initial random sub-sample of 20 men and 20 women from the intervention group. The qualitative interviews strengthen this study as the researchers will use the findings to "triangulate" our quantitative results and to guide interpretation of the quantitative results. The interviews will be audio recorded. Qualitative interviews will be transcribed verbatim then analyzed inductively for themes. As per Berg themes will be chosen to "reflect all relevant aspects of the messages." The analysis will follow the interpretive practices of constant comparison and attempt to uncover patterns both within and between interviews. Nvivo 10 software will be used to support thematic analysis. It is expected to achieve theoretical saturation with the initial sample. However, if new themes continue to emerge in final interviews, more interviews will be conducted until no new themes emerge.

Risk and mitigation: The investigators acknowledge concerns about the changes in response rates in telephone surveys due to cell phone use and telemarketing. Including eligible participants across the country will enhance the generalizability of the study. Given the vast geographic area of Canada, RDD is the only feasible method. The goal of this study is to recruit participants for a RCT, rather than selecting a representative sample. In a RCT, selection bias is not a serious concern as long as the bias is the same across the intervention and control groups. To mitigate the risk, the interviewers will also access the validated cellphone database. However the use of cellphone numbers is associated with increased costs. Another potential risk of the proposed study is attrition which may incur selection bias. The population-based cohort studies on mental disorders in the workplace, conducted in our lab, showed that we could achieve 77% response rate at one year follow-up without any financial incentives. The strategies for reducing attrition will include appropriately designed introductory scripts, a minimum of nine call back attempts spaced over weekdays and times of day and provision of $20 incentive for each completed interview. Finally, those deemed low risk but who go on to develop MDE will be excluded from the RCT at the screening stage, which is a limitation. the investigators have planned to provide the risk prediction algorithms so that they can monitor their risks in the future.