Mental Illness, Substance Related Disorders
Conditions
Keywords
Implementation Science, Supportive Housing, Co-occurring disorders, Cluster randomized controlled trial, Hybrid implementation-effectiveness trial
Brief summary
A major goal for the Department of Veterans Affairs is to end Veteran homelessness by 2015. The VA's largest homelessness initiative is the joint Departments of Housing and Urban Development (HUD) and Veterans Affairs (VA) Supportive Housing program (HUD-VASH), which has been expanded greatly over recent years via the allocation of 30,000 Housing First vouchers between 2008 and 2010 and increased funding to hire 1,000 new program case managers. However, recent expansion has resulted in a number of implementation challenges including delays in the distribution of housing vouchers and dropout among program participants (25% of those housed in HUD-VASH drop out within a year). Much of this dropout can be attributed to untreated issues facing many Veterans enrolled in HUD-VASH. The most common among these untreated issues are mental health and substance use disorders. The presence of these disorders is due in large part to the fact that much of HUD-VASH case management focuses on housing placement and maintenance, with limited attention to mental health, substance abuse, and other related psychosocial issues, which when left untreated, negatively impacts voucher distribution and housing stability. This project will test an implementation model-Getting To Outcomes (GTO)-designed to assist in the delivery of an intervention for Veterans with co-occurring mental health and substance use disorders (MISSION-Vet) in the HUD-VASH program. The proposed study will compare implementation of MISSION-Vet currently being planned through VA Office of Patient Care Services to an enhanced approach using the GTO model. Thus, this project can contribute to ending all Veteran homelessness by 2015, a pledge made by President Obama.
Detailed description
This project tests an implementation platform-Getting To Outcomes (GTO)-designed to assist in the delivery of an evidence-based intervention for Veterans with co-occurring mental health and substance use disorders (MISSION-Vet) in the HUD-VASH program. This project will be a cluster randomized controlled trial that compares implementation of MISSION-Vet augmented by GTO to MISSION-Vet Implemented as Usual (IU) at three of the largest HUD-VASH programs in the country: VA Central Western Massachusetts Healthcare System (Northampton, MA), VA Capitol Health Care Network (Washington, D.C.), and VA Eastern Colorado Health Care System (Denver, CO). This project will randomly assign 150 HUD-VASH case managers and 1106 Veterans on their caseloads who have received HUD-VASH vouchers and case management services into these two groups on a 1-year rolling admission basis determined by when the Veteran receives a housing voucher. The control group will receive MISSION-Vet in addition to HUD-VASH case management services and the intervention group will receive the same as the control, however the HUD-VASH case manager will have access to GTO implementation support. To compare case managers implementing MISSION-Vet augmented with GTO to HUD-VASH case managers using IU strategies on the following variables: (1) fidelity to the MISSION-Vet intervention; (2) proportion of time the Veteran is housed; (3) mental health, substance use, and functional outcomes among Veterans; and (4) factors key to the successful deployment of a new treatment as specified by the Reach, Effectiveness, Adoption, Implementation, & Maintenance (RE-AIM) model. Data will be collected on fidelity to MISSION-Vet in both IU and GTO groups. Data will also be collected on all subjects' substance use, overall mental health functioning, engagement in substance abuse treatment services, the length of time housed, and community participation. The investigators will collect data from individuals at baseline and three time points for up to one year. The investigators' data analysis strategy will be to use a repeated-measures model to test for the significance of the treatment-by-time interaction while accounting for the clustered design of case manager within site. This study intends to serve a dual function of comparing implementation of MISSION-Vet currently being planned through VA Office of Patient Care Services to an enhanced implementation approach using the GTO model. The proposed research will help to guide policy and practice actions to implement MISSION-Vet with fidelity and effectiveness to achieve maximum outcomes among homeless Veterans.
Interventions
BEHAVIORALMaintaining Independence and Sobriety Through Systems Integration, Outreach, and Networking (Veterans Edition)
MISSION-Vet has been developed to target mental health, substance abuse and related issues facing homeless Veterans through assertive outreach, psychoeducation, and linkages to community-based resources
OTHERGetting To Outcomes
GTO strengthens the knowledge, attitudes, and skills practitioners need to carry out evidence based programs. GTO consists of a series of steps practitioners should follow in order to obtain positive results and then provides them with the guidance necessary to complete those steps with quality. According to GTO, carrying out an evidence based program includes a series of steps corresponding to three general areas: (1) planning - e.g., developing goals and performance targets, ensuring staff are trained in the evidence based program; (2) implementation - e.g., monitoring program activities, maintaining adherence to an evidence based program model, supervision; and (3) self-evaluation - e.g., tracking patient outcomes, using data to improve program operations.
Sponsors
VA Office of Research and Development
Study design
Allocation
RANDOMIZED
Intervention model
PARALLEL
Primary purpose
OTHER
Masking
NONE
Eligibility
Sex/Gender
ALL
Healthy volunteers
Yes
Inclusion criteria
VA HUD-VASH case manager at the Northampton, Denver, and Washington D.C. HUD-VASH programs
Exclusion criteria
N/A
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| MISSION Fidelity Index | 12-months | The fidelity index assesses the presence or absences of activities within MISSION-Vet - DRT, peer led sessions, self-guided exercises, referrals made, and/or delivery of the workbook, to each participating Veteran. We analyzed fidelity as the percent of adoption of MISSION-Vet. The threshold for fidelity to adopt MISSION-Vet is 1 contact between the case manager and each participating Veteran. There is no composite score, fidelity to MISSION-Vet is if the case manager conducted at least 1 session with a Veteran.This measure was embedded into the VA Electronic Medical Record System. The investigators will assess the impact GTO has in facilitating adoption and use with fidelity to the MISSION-Vet 12-month service delivery platform, in comparison to implementation as usual. |
Countries
United States
Participant flow
Pre-assignment details
All case managers were invited to participate and verbally consented before randomization into the IU or the GTO group. 10 refused to participate.
Participants by arm
| Arm | Count |
|---|---|
| MISSION-Vet - IU Case Management Maintaining Independence and Sobriety Through Systems Integration, Outreach, and Networking (Veterans Edition):MISSION-Vet has been developed to target mental health, substance abuse and related issues faced by homeless Veterans through assertive outreach, psychoeducation, and linkages to community-based resources.
Implementation as Usual (IU) - standard training on the MISSION model via a 1.5 hour webinar | 22 |
| MISSION-Vet - IU Veterans Maintaining Independence and Sobriety Through Systems Integration, Outreach, and Networking (Veterans Edition):MISSION-Vet has been developed to target mental health, substance abuse and related issues faced by homeless Veterans through assertive outreach, psychoeducation, and linkages to community-based resources.
Implementation as Usual (IU) - staff receive standard training on the MISSION model via a 1.5 hour webinar and then deliver MISSION services to Veterans. | 87 |
| Enhanced Implementation Approach GTO Case Management Getting To Outcomes (GTO) is used to strengthens the knowledge, attitudes, and skills practitioners need to carry out evidence based programs. In GTO, staff receive ongoing technical assistance using the GTO implementation platform. | 37 |
| GTO Veterans Getting To Outcomes (GTO) is used to strengthens the knowledge, attitudes, and skills practitioners need to carry out evidence based programs. In GTO, staff receive ongoing technical assistance using the GTO implementation platform, while delivering MISSION services to Veterans | 81 |
| Total | 227 |
Baseline characteristics
| Characteristic | MISSION-Vet - IU Veterans | GTO Veterans | Total | MISSION-Vet - IU Case Management | Enhanced Implementation Approach GTO Case Management |
|---|---|---|---|---|---|
| Age, Continuous | 49.7 years STANDARD_DEVIATION 10.8 | 55.9 years STANDARD_DEVIATION 9.5 | 52.8 years STANDARD_DEVIATION 10.6 | — | — |
| Sex/Gender, Customized Unknown | 87 participants | 81 participants | 227 participants | 22 participants | 37 participants |
Adverse events
| Event type | EG000 affected / at risk | EG001 affected / at risk | EG002 affected / at risk | EG003 affected / at risk |
|---|---|---|---|---|
| deaths Total, all-cause mortality | — / — | — / — | — / — | — / — |
| other Total, other adverse events | 0 / 22 | 0 / 87 | 0 / 37 | 0 / 81 |
| serious Total, serious adverse events | 0 / 22 | 0 / 87 | 0 / 37 | 0 / 81 |
Outcome results
Primary
MISSION Fidelity Index
The fidelity index assesses the presence or absences of activities within MISSION-Vet - DRT, peer led sessions, self-guided exercises, referrals made, and/or delivery of the workbook, to each participating Veteran. We analyzed fidelity as the percent of adoption of MISSION-Vet. The threshold for fidelity to adopt MISSION-Vet is 1 contact between the case manager and each participating Veteran. There is no composite score, fidelity to MISSION-Vet is if the case manager conducted at least 1 session with a Veteran.This measure was embedded into the VA Electronic Medical Record System. The investigators will assess the impact GTO has in facilitating adoption and use with fidelity to the MISSION-Vet 12-month service delivery platform, in comparison to implementation as usual.
Time frame: 12-months
Population: Clinician fidelity to MISSION-Vet using IU or GTO. Only assessed Staff population.
| Arm | Measure | Value (NUMBER) |
|---|---|---|
| MISSION-Vet - IU Case Management | MISSION Fidelity Index | 0 percentage of MISSION-Vet staff use |
| GTO Case Management | MISSION Fidelity Index | 68 percentage of MISSION-Vet staff use |
Post Hoc
Drug and Alcohol Dependence
Examined monthly alcohol and drug use by Veterans. Assessed on a monthly basis for 12 months following study enrollment. These measures were obtained from the standardized monthly and quarterly status reports completed by case managers, which were extracted from HOMES.
Time frame: 12 months
Population: All MISSION-Vet IU Veterans and GTO Veterans were analyzed from HUD-VASH data monitoring system HOMES (Homeless Operations, Management and Evaluation).
| Arm | Measure | Group | Value (LEAST_SQUARES_MEAN) | Dispersion |
|---|---|---|---|---|
| MISSION-Vet IU Veterans | Drug and Alcohol Dependence | Drug Use | 1.291 Monthly Substance Use | Standard Error 0.62 |
| MISSION-Vet IU Veterans | Drug and Alcohol Dependence | Alcohol Use | 1.668 Monthly Substance Use | Standard Error 0.79 |
| GTO Veterans | Drug and Alcohol Dependence | Drug Use | 1.786 Monthly Substance Use | Standard Error 0.588 |
| GTO Veterans | Drug and Alcohol Dependence | Alcohol Use | 2.278 Monthly Substance Use | Standard Error 0.752 |
Comparison: Compared changes in alcohol dependence among Veterans served by case managers in the GTO group relative to the comparison group of Veterans, by fitting a series of mixed-effects regression models. Final mixed-effect models for all outcomes included fixed terms for group (GTO group vs. comparison), time, group x time interaction as well as control covariates.p-value: =0.21Mixed Models Analysis
Comparison: Compared changes in drug dependence among Veterans served by case managers in the GTO group relative to the comparison group of Veterans, by fitting a series of mixed-effects regression models. Final mixed-effect models for all outcomes included fixed terms for group (GTO group vs. comparison), time, group x time interaction as well as control covariates.p-value: =0.07Mixed Models Analysis
Post Hoc
Estimated Proportion of Veterans Without Negative Housing Exits
We obtained information on negative housing exits from the HUD-VASH Exit form, which is available in HOMES and tracks the dates and reasons for all exits from the program. We measured whether and when a Veteran experienced a negative housing over the period from study enrollment (i.e. the first date of MISSION-Vet for those served by case managers in the GTO group and the date of the first MISSION-Vet session provided to any Veteran at the appropriate study site for this in the comparison group) until the end of the study observation period; the maximum follow-up time was 2.4 years, but only for HUD-VASH Exits. A negative exit was defined as an exit from HUD-VASH that occurred due to non-compliance with case management, eviction, Veteran dissatisfaction with housing, inability to locate Veteran and incarceration. Kaplan-Meier survival curves to calculate an estimate of the proportion of Veterans in each arm/group not experiencing a negative exit from housing at the study's endpoint.
Time frame: 12 months
Population: All MISSION-Vet IU Veterans and GTO Veterans were analyzed from HUD-VASH data monitoring system HOMES (Homeless Operations, Management and Evaluation).
| Arm | Measure | Value (NUMBER) |
|---|---|---|
| MISSION-Vet IU Veterans | Estimated Proportion of Veterans Without Negative Housing Exits | 0.808 Proportion of Veterans |
| GTO Veterans | Estimated Proportion of Veterans Without Negative Housing Exits | 0.884 Proportion of Veterans |
Comparison: We used Kaplan-Meier survival curves to estimate the extent and timing of negative housing exits over time and Cox proportional hazards regression to assess the relationship between membership in the GTO group and risk of experiencing a negative housing exit, adjusting for a set of relevant covariates.p-value: =0.481Regression, Cox
Post Hoc
Number of Services Provided by Case Managers and Peers to Veterans
Services provided by Case Managers OR Peers was computed by a count of the number of contacts between case managers OR peers and Veterans. This computed a total of all Veteran contacts, which was the sum of face-to-face contacts with Veterans and contacts on behalf of Veterans (described as Veteran contacts) with other stakeholders (e.g. family, medical providers). Assessed on a monthly basis for 12 months following study enrollment. These measures were obtained from the standardized monthly and quarterly status reports completed by case managers, which were extracted from HOMES. We used a linear-mixed effects model to calculate a Least-Squares Mean for each group at the study's end-point holding all variables in the model (Veteran age, Veteran sex, presence of an SMI diagnosis, employment) constant at their mean value.
Time frame: 12 months
Population: All MISSION-Vet IU Veterans and GTO Veterans were analyzed from HUD-VASH data monitoring system HOMES (Homeless Operations, Management and Evaluation).
| Arm | Measure | Value (LEAST_SQUARES_MEAN) | Dispersion |
|---|---|---|---|
| MISSION-Vet IU Veterans | Number of Services Provided by Case Managers and Peers to Veterans | 3.466 Number of Service Contacts | Standard Error 2.824 |
| GTO Veterans | Number of Services Provided by Case Managers and Peers to Veterans | 3.469 Number of Service Contacts | Standard Error 3.051 |
Comparison: Compared changes in service intensity among Veterans served by case managers in the GTO group relative to the comparison group of Veterans, by fitting a series of mixed-effects regression models. Final mixed-effect models for all outcomes included fixed terms for group (GTO group vs. comparison), time, group x time interaction as well as control covariates.p-value: =0.04Mixed Models Analysis
Post Hoc
Predicted Values of Veteran Emergency Department Visits for Medical and Mental Health Concerns
Number of emergency department visits for medical and for mental health. Assessed on a monthly basis for 12 months following study enrollment. These measures were obtained from the standardized monthly and quarterly status reports completed by case managers, which were extracted from HOMES. A dichotomous measure of whether a veteran experienced an emergency department visit for medical or mental health conditions , and not a count of number of emergency department visits. We used a mixed-effects logistic regression model to calculate a predicted values (i.e. a Least-Squares Mean) for each group/arm at the study's end-point holding all variables in the model (Veteran age, Veteran sex, presence of an SMI diagnosis, employment) constant at their mean value. Predicted value reported in log-odds (logit) units.
Time frame: 12 months
Population: All MISSION-Vet IU Veterans and GTO Veterans were analyzed from HUD-VASH data monitoring system HOMES (Homeless Operations, Management and Evaluation).
| Arm | Measure | Group | Value (LEAST_SQUARES_MEAN) | Dispersion |
|---|---|---|---|---|
| MISSION-Vet IU Veterans | Predicted Values of Veteran Emergency Department Visits for Medical and Mental Health Concerns | Mental Health Emergency Dept Visits | -7.96 Log Odd Units of Emergency Dept. Visits | Standard Error 2.312 |
| MISSION-Vet IU Veterans | Predicted Values of Veteran Emergency Department Visits for Medical and Mental Health Concerns | Medical Emergency Dept Visits | -4.663 Log Odd Units of Emergency Dept. Visits | Standard Error 3.063 |
| GTO Veterans | Predicted Values of Veteran Emergency Department Visits for Medical and Mental Health Concerns | Medical Emergency Dept Visits | -3.983 Log Odd Units of Emergency Dept. Visits | Standard Error 2.923 |
| GTO Veterans | Predicted Values of Veteran Emergency Department Visits for Medical and Mental Health Concerns | Mental Health Emergency Dept Visits | -6.792 Log Odd Units of Emergency Dept. Visits | Standard Error 1.779 |
Comparison: Compared changes in mental health emergency department visits among Veterans served by case managers in the GTO group relative to the comparison group of Veterans, by fitting a series of mixed-effects regression models. Final mixed-effect models for all outcomes included fixed terms for group (GTO group vs. comparison), time, group x time interaction as well as control covariates.p-value: =0.24Mixed Models Analysis
Post Hoc
Predicted Values of Veteran Inpatient Hospitalizations for Medical and Mental Health Conditions
Number of inpatient hospitalization for medical and mental health conditions. Assessed on a monthly basis for 12 months following study enrollment. These measures were obtained from the standardized monthly and quarterly status reports completed by case managers, which were extracted from HOMES. A dichotomous measure of whether a veteran experienced an inpatient hospitalization for medical or mental health conditions , and not a count of number of hospitalizations. Specifically, we used a mixed-effects logistic regression model to calculate a predicted values (i.e. a Least-Squares Mean) for each group/arm at the study's end-point holding all variables in the model (Veteran age, Veteran sex, presence of an SMI diagnosis, employment) constant at their mean value. Predicted value reported in log-odds (logit) units.
Time frame: 12 months
Population: All MISSION-Vet IU Veterans and GTO Veterans were analyzed from HUD-VASH data monitoring system HOMES (Homeless Operations, Management and Evaluation).
| Arm | Measure | Group | Value (LEAST_SQUARES_MEAN) | Dispersion |
|---|---|---|---|---|
| MISSION-Vet IU Veterans | Predicted Values of Veteran Inpatient Hospitalizations for Medical and Mental Health Conditions | Mental Health Inpatient Hospitalizations | -5.311 Inpatient Hospitalization Log Odds Units | Standard Error 6.008 |
| MISSION-Vet IU Veterans | Predicted Values of Veteran Inpatient Hospitalizations for Medical and Mental Health Conditions | Medical Inpatient Hospitalizations | -6.597 Inpatient Hospitalization Log Odds Units | Standard Error 1.903 |
| GTO Veterans | Predicted Values of Veteran Inpatient Hospitalizations for Medical and Mental Health Conditions | Mental Health Inpatient Hospitalizations | -3.825 Inpatient Hospitalization Log Odds Units | Standard Error 5.802 |
| GTO Veterans | Predicted Values of Veteran Inpatient Hospitalizations for Medical and Mental Health Conditions | Medical Inpatient Hospitalizations | -6.555 Inpatient Hospitalization Log Odds Units | Standard Error 1.498 |
Comparison: Compared mental health inpatient hospitalizations among Veterans served by case managers in the GTO group relative to the comparison group of Veterans, by fitting a series of mixed-effects regression models. Final mixed-effect models for all outcomes included fixed terms for group (GTO group vs. comparison), time, group x time interaction as well as control covariates.p-value: =0.14Mixed Models Analysis
Comparison: Compared medical inpatient hospitalizations among Veterans served by case managers in the GTO group relative to the comparison group of Veterans, by fitting a series of mixed-effects regression models. Final mixed-effect models for all outcomes included fixed terms for group (GTO group vs. comparison), time, group x time interaction as well as control covariates.p-value: =0.19Mixed Models Analysis