A Clinical Decision Support Tool for Electronic Health Records
| Status: | Completed |
|---|---|
| Conditions: | Psychiatric |
| Therapuetic Areas: | Psychiatry / Psychology |
| Healthy: | No |
| Age Range: | 18 - Any |
| Updated: | 8/15/2018 |
| Start Date: | March 17, 2016 |
| End Date: | March 31, 2018 |
For behavioral health clinicians who are interested in getting tailored treatment and level
of care recommendations, "BH-CDS" is a desktop/tablet web-based application that provides
clinicians with data and a rationale for better decision-making to improve patient care.
Few Clinical Decision Support (CDS) systems are available for Behavioral Health, and unlike
existing CDS this product will compile relevant patient data and organize these data into
general treatment recommendations linked to the patient's presenting circumstances, symptoms
and substance use issues.
The BH-CDS solution shall factor patient characteristics into a Latent Class Analysis (LCA)
that will group patients according to their responses with other patients with similar
responses (i.e., a subgroup or "class"). Once patients have been assigned to a class, the
solution shall present recommendations to counselors that use the software.
of care recommendations, "BH-CDS" is a desktop/tablet web-based application that provides
clinicians with data and a rationale for better decision-making to improve patient care.
Few Clinical Decision Support (CDS) systems are available for Behavioral Health, and unlike
existing CDS this product will compile relevant patient data and organize these data into
general treatment recommendations linked to the patient's presenting circumstances, symptoms
and substance use issues.
The BH-CDS solution shall factor patient characteristics into a Latent Class Analysis (LCA)
that will group patients according to their responses with other patients with similar
responses (i.e., a subgroup or "class"). Once patients have been assigned to a class, the
solution shall present recommendations to counselors that use the software.
Summary of the specific aims and impact on public health of the Phase II. Substance abuse
treatment is often complicated by a client's family, employment, psychiatric, or legal
problems. When these co-existing issues are addressed with evidence-based practices (EBPs),
outcomes improve. The inclusion of behavioral health evidence-based practices to enhance
Medication-Assisted Treatment (MAT) is the subject of a number of federal and state treatment
initiatives. However, the integration of such evidence-based practices into clinical settings
continues to lag, despite extensive efforts to educate clinicians through training. Since it
is often difficult to integrate EBPs into the clinical workflow, clinicians rely on
established (and often ineffective) patterns of care. This grant proposed to (1) use
electronic health record data on patients with a diagnosis of opioid use disorder to create
profiles of patient groups using latent class analysis (LCA) analysis and determine, for each
class, which combination of services are empirically associated with positive outcomes; (2)
develop clinical decision support (CDS) software to help counselors classify patients and
match them to appropriate services, and (3) conduct a field trial (randomized controlled
trial or RCT) to test the impact of the CDS software on clinical practice.
Provide a succinct account of published and unpublished results, indicating progress toward
achievement of the originally stated aims.
Latent Class Analysis: The first aim (using electronic health record data on patients with a
diagnosis of opioid use disorder to create profiles of patient groups using LCA and
determining which combinations of services are empirically associated with positive outcomes
for each class of opioid users) was successfully achieved, as discussed in previous progress
reports.
Four classes were identified: Class 1: Individuals in this class tend to have relatively high
medical and mental health problems, be taking psychiatric medications and tend to experience
control problems with their temper. Class 2: Individuals in this class tend to have mental
health problems, but are not taking psychiatric medications. They do not generally snort or
inject opiates and tend not to have serious medical problems. Class 3: Individuals in this
class tend to have medical and mental health problems and are taking psychiatric medications.
They have a tendency to snort or inject opiates and may have some problems controlling their
temper. Class 4: Individuals in this class tend to have a high tendency to snort or inject
opiates. They have medium medical problems and low mental health issues.
Software Development: Based on the LCA results, CDS software was developed to help counselors
classify patients and match them to appropriate services.
Field Trial: The purpose of this field trial was to evaluate the effectiveness of this new
CDS software when compared to clinical care as usual or treatment-as-usual (TAU), and to
gather information about feasibility and perceived usefulness of the CDS software from the
counselor's perspective. It was anticipated that when compared to TAU, clients in the
experimental condition would (1) have significantly greater matched evidenced-based and
wraparound services, (2) have greater engagement in treatment, (3) have less frequent use of
substances, (4) have greater biopsychosocial functioning, and (5) have greater cost
effectiveness (i.e., less cost to achieve successful outcomes).
treatment is often complicated by a client's family, employment, psychiatric, or legal
problems. When these co-existing issues are addressed with evidence-based practices (EBPs),
outcomes improve. The inclusion of behavioral health evidence-based practices to enhance
Medication-Assisted Treatment (MAT) is the subject of a number of federal and state treatment
initiatives. However, the integration of such evidence-based practices into clinical settings
continues to lag, despite extensive efforts to educate clinicians through training. Since it
is often difficult to integrate EBPs into the clinical workflow, clinicians rely on
established (and often ineffective) patterns of care. This grant proposed to (1) use
electronic health record data on patients with a diagnosis of opioid use disorder to create
profiles of patient groups using latent class analysis (LCA) analysis and determine, for each
class, which combination of services are empirically associated with positive outcomes; (2)
develop clinical decision support (CDS) software to help counselors classify patients and
match them to appropriate services, and (3) conduct a field trial (randomized controlled
trial or RCT) to test the impact of the CDS software on clinical practice.
Provide a succinct account of published and unpublished results, indicating progress toward
achievement of the originally stated aims.
Latent Class Analysis: The first aim (using electronic health record data on patients with a
diagnosis of opioid use disorder to create profiles of patient groups using LCA and
determining which combinations of services are empirically associated with positive outcomes
for each class of opioid users) was successfully achieved, as discussed in previous progress
reports.
Four classes were identified: Class 1: Individuals in this class tend to have relatively high
medical and mental health problems, be taking psychiatric medications and tend to experience
control problems with their temper. Class 2: Individuals in this class tend to have mental
health problems, but are not taking psychiatric medications. They do not generally snort or
inject opiates and tend not to have serious medical problems. Class 3: Individuals in this
class tend to have medical and mental health problems and are taking psychiatric medications.
They have a tendency to snort or inject opiates and may have some problems controlling their
temper. Class 4: Individuals in this class tend to have a high tendency to snort or inject
opiates. They have medium medical problems and low mental health issues.
Software Development: Based on the LCA results, CDS software was developed to help counselors
classify patients and match them to appropriate services.
Field Trial: The purpose of this field trial was to evaluate the effectiveness of this new
CDS software when compared to clinical care as usual or treatment-as-usual (TAU), and to
gather information about feasibility and perceived usefulness of the CDS software from the
counselor's perspective. It was anticipated that when compared to TAU, clients in the
experimental condition would (1) have significantly greater matched evidenced-based and
wraparound services, (2) have greater engagement in treatment, (3) have less frequent use of
substances, (4) have greater biopsychosocial functioning, and (5) have greater cost
effectiveness (i.e., less cost to achieve successful outcomes).
Counselor Inclusion Criteria:
- full or part-time counselors
- English speaking
- Treat clients with opioid use problems
- Have an active e-mail account
Client Inclusion Criteria:
- Currently meet with a counselor in the study at least once a month
- able to read and speak English
- in treatment for an opioid use problem
- completed detox, if it was necessary
Exclusion Criteria:
We found this trial at
1
site
Click here to add this to my saved trials
