In-Home Sleep Monitoring to Detect Suicide Risk in Veterans

NCT ID: NCT04744740

Last Updated: 2021-02-09

Basic Information

• Recruitment Status: UNKNOWN
• Total Enrollment: 100 participants
• Study Classification: OBSERVATIONAL
• Study Start Date: 2020-10-01
• Study Completion Date: 2022-06-30

Brief Summary

This research will assess the ability of nightly objective sleep scheduling and sleep heart rate data to predict variation in suicidal ideation in U.S. military veterans identified as high-risk according to actuarial and/or clinical indicators. These data will be recorded using one of two low-cost, commercial, zero-burden, mattress actigraphic systems that enable remote, centralized, data access. Participants will be veterans accrued both from the REACH VET and local high-risk lists, a process involving collaboration with the Suicide Prevention Coordinator at the VA Palo Alto HCS. In order to minimize burden, participants will be screened by telephone, consented by mail and telephone, administered a brief admission psychometric interview by telephone, and mailed a sleep sensor system and instructions for its installation in their home. Once sleep data collection has commenced, a trained interviewer will contact the participant by telephone weekly and administer the Columbia-Suicide Severity Rating Scale. Data collection will continue for three months. The study targets a sample of 40 veterans. Longitudinal mixed effects modeling will be employed to determine whether changes in sleep scheduling and sleep heart rate predict changes in suicidal ideation.

Detailed Description

Suicide is a devastating event that occurs at a very low base rate, rendering detection of acute suicide risk one of the signal challenges in behavioral science. Advancement in the detection of imminent risk requires new modalities of intensive, longitudinal data collection that are both acceptable and scalable. Recent developments in commercial, actigraphic sleep recording systems have opened a new avenue of approach to this challenge, enabling the automated collection and daily review of sleep scheduling and sleep heart rate (HR), both of which have demonstrated associations with suicidality in prior studies. Our team, comprising Drs. Steven Woodward (Director of the Sleep and Psychophysiology Laboratory of the National Center for PTSD), Rebecca Bernert (Director of the Suicide Prevention Center, Stanford University School of Medicine), Christina Khan (Director of the Thrive Clinic, Stanford University School of Medicine), and Wilfred Pigeon (Director of the Center of Excellence in Suicide Prevention, Department of Veterans Affairs) will pursue the following Aims.

AIM 1: To assess, in U.S. military veterans diagnosed with posttraumatic stress disorder (PTSD), whether sleep schedule variability and sleep heart rate, recorded in the home for one month using a low-cost, zero-burden, actigraphic sleep measurement system, can distinguish those who are at elevated actuarial risk for suicide from those who are not.

AIM 2: To assess, in U.S. military veterans diagnosed with PTSD, whether elevated sleep schedule variability and elevated sleep heart rates can predict increased suicidal ideation over a clinically-relevant surveillance period of two months.

The proposal is founded on observational and epidemiological studies supporting associations between sleep disturbance, basal heart rate and suicidality, including important contributions from Co-Is Bernert and Pigeon (1-12). Also supportive of the proposal are new observations from a sample of 71 U.S. Military veterans with deployment-related PTSD studied by the PI during inpatient PTSD treatment. In this sample, after adjusting for depressed mood and body mass index (BMI), sleep heart rate was 3.8 BPM higher in those endorsing current ideation and a history of attempts than in those denying both. The investigators also describe in detail below a supportive case observed in Dr. Woodward's laboratory which, to our knowledge, represents the first extended acquisition of sleep behavior prior to a serious suicide attempt.

The proposal leverages the PI's expertise in zero-burden, mattress-actigraphic sleep recording originally motivated by the observation that PTSD patients rarely manifest sleep disturbances in the laboratory. The aforementioned data were obtained employing a high-accuracy, non-commercial, mattress actigraphic system developed by the PI (13-15), which was found acceptable to > 95% of the 900+ veteran PTSD inpatients with whom it has been used at the Trauma Recovery Program (TRP) at the VA Palo Alto Health Care System (~40,000 nights of recording). The proposed study will employ a low-cost, commercially-available alternative, the Sleepace RestOn, that can be mailed and user-installed. This system is also zero-burden as there is nothing to put on, take off, charge, or upload. In our pilot work, it has proven acceptable to veteran outpatients at high risk for suicide. Moreover, though its connection to a tablet computer, this system provides daily, web-based access to the raw actigraphic data enabling the direct verification of data quality and the application of custom algorithms to extract movement, heart rate, and respiratory amplitude.

Aim 1 will attempt to replicate and extend the aforementioned observations made in PTSD inpatients with and without suicidal ideation and histories of attempts in PTSD outpatients studied in the home. Aim 2 will attempt to replicate and extend Co-I Bernert's study demonstrating, in high-risk young adults, that weekly changes in sleep schedule variability predict weekly changes in suicidal ideation (1). Bernert's study period will be extended from three weeks to two months, and sleep heart rate will be included as a predictor. Suicide symptoms will be assessed weekly by telephone using the Depression Symptom Index - Suicide Severity (DSI-SS; 16), a brief, psychometrically-sound measure of suicide ideation intensity yielding a score from 0 to 12 (17). Mixed effects modeling (18) will be employed, and with respect to Aim 2, the temporal precedence of changes in sleep to changes in suicidal ideation intensity will be tested in order confirm that sleep changes meet criteria for true risk-factors (19). An established partnership with VA Suicide Prevention Coordinators (SPCs) will facilitate recruitment from the actuarial REACH VET list (20) and local hospital high-risk list while maximizing participant safety. An exploratory aim will estimate the contribution of certain participant factors likely to impact sleep scheduling or its measurement, in-bed media consumption, hypnotic medication, CPAP use, and alcohol/substance use. In the case of a positive study outcome, a multi-site R01 will be proposed including centralized data collection and review at the VA Center of Excellence for Suicide Prevention in Canandaigua, NY, under the direction of Co-I Pigeon, implementation science and health economic aims developed in collaboration with Co-I Marshall, and exploratory integration into the REACH VET algorithm.

Conditions

Suicidal Ideation; Sleep Disturbance

Study Design

• Observational Model Type: CASE_CONTROL
• Study Time Perspective: PROSPECTIVE

Study Groups

PTSD+Suicidal Ideation

US Military Veterans diagnosed with PTSD and identified via the REACH VET or local high-risk list as requiring intensified surveillance by their VA's Suicide Prevention Coordinators in collaboration with their Primary Care Providers

in-home sleep recording

Intervention Type: DEVICE

in-home, nightly, zero-burden objective quantification of sleep scheduling and sleep heart rate via a mattress-top sensor (Sleepace RestOn) + ad lib quantification of blood SaO2 via a ring-based sensor (Bodimetrics Circul) + ad lib quantification of sleep and waking activity and heart rate via a wrist-worn sensor (Garmin Vivosmart 4)

PTSD-Suicidal Ideation

US Military Veterans diagnosed with PTSD not identified as requiring intensified surveillance by their VA's Suicide Prevention Coordinators in collaboration with their Primary Care Providers

in-home sleep recording

Intervention Type: DEVICE

in-home, nightly, zero-burden objective quantification of sleep scheduling and sleep heart rate via a mattress-top sensor (Sleepace RestOn) + ad lib quantification of blood SaO2 via a ring-based sensor (Bodimetrics Circul) + ad lib quantification of sleep and waking activity and heart rate via a wrist-worn sensor (Garmin Vivosmart 4)

Interventions

in-home sleep recording

in-home, nightly, zero-burden objective quantification of sleep scheduling and sleep heart rate via a mattress-top sensor (Sleepace RestOn) + ad lib quantification of blood SaO2 via a ring-based sensor (Bodimetrics Circul) + ad lib quantification of sleep and waking activity and heart rate via a wrist-worn sensor (Garmin Vivosmart 4)

Intervention Type: DEVICE

Eligibility Criteria

Inclusion Criteria

1. PTSD or partial PTSD

2. Status on clinical high-risk list and subject to enhanced surveillance by the suicide prevention activity and staff at VAPAHCS (or not)

3. Adequate cell phone service and/or wireless wife connectivity in sleeping quarters.

Exclusion Criteria

1. Schizophrenia or psychotic disorder, autism-spectrum diagnoses medically unstable

2. unable to speak and read English or otherwise incapable of providing informed consent

• Minimum Eligible Age: 18 Years
• Maximum Eligible Age: 75 Years
• Eligible Sex: ALL
• Accepts Healthy Volunteers: No

Sponsors

Stanford University

OTHER

Sponsor Role: collaborator

Palo Alto Veterans Institute for Research

OTHER

Sponsor Role: lead

Responsible Party

Responsibility Role: SPONSOR

Locations

VA Palo Alto Health Care System

Palo Alto, California, United States

Site Status: RECRUITING

Countries

United States

Central Contacts

Steven H Woodwawrd, PhD

Role: CONTACT

steve.woodward@va.gov

6502450450

Diana Villasenor, BA

Role: CONTACT

diana.villasenor@va.gov

650-3915983

Facility Contacts

Esther Eng, BA

Role: primary

eng@pavir.org

650-239-2818

Elaine Staats, BA

Role: backup

estaats@pavir.org

650-239-2803

Other Identifiers

WOS0023AGG

Identifier Type: -

Identifier Source: org_study_id