Results Summary
What was the research about?
In mental health care, doctors often make the decisions about treatment. To make care more focused on people’s needs, people with a mental illness and their doctors can make decisions together. This process is called shared decision making.
In this study, the research team compared two programs to help people with mental illness make decisions about treatment:
- Measurement-Based Care, or MBC, used a web-based portal to ask people about their health, like their symptoms and medicines. These answers helped doctors work with patients to choose medicines and treatment that would work best for each person.
- Person-Centered Care, or PCC, used peer support staff to help people fill out an online survey. Peer support staff are people recovering from mental illness trained to give peer support. The survey asked questions about people’s health, like how they are feeling, what they do to feel better, and what concerns they have about their medicine. The peer support staff also helped people figure out their treatment goals and prepare to talk with their doctors. Doctors got the patients’ answers before the visit and talked with them about what treatment would work best.
The team looked at whether people with mental illness felt involved in their care and medicine choices.
What were the results?
At the end of the study, people in the two programs didn’t differ in their
- Experience managing their medicines
- Shared decision making with their doctors
- Involvement in their health and mental health care
- Management of their medicines’ side effects
- Quality of life
- Seriousness of their mental health symptoms
- Ability to do daily activities
Who was in the study?
The study included 2,363 adults who had Medicaid health insurance and received mental health care at one of 14 community mental health centers in Pennsylvania. Of these adults, 86 percent were white, 9 percent were black, and 5 percent were another race; 2 percent were Hispanic. The average age was 42, and 63 percent were women. In the study
- 24 percent of people had bipolar disorder
- 23 percent had major depression
- 20 percent had a substance use problem
- 17 percent had another type of depression
- 16 percent had schizophrenia or schizoaffective disorder
- 10 percent had anxiety or posttraumatic stress disorder
- 10 percent had another diagnosis
What did the research team do?
The research team assigned 14 community mental health centers by chance to either MBC or PCC programs. People took part in the programs at the centers where they received mental health care. They filled out four surveys between the start of the program and two years later. The team looked at the surveys and at data from Medicaid insurance claims.
People with mental illness, mental health organizations, peer support staff, policy makers, mental health providers, and researchers gave input on the study.
What were the limits of the study?
Four centers didn’t stay in the study. The centers that replaced them may have differed in ways that changed the results. The questions the research team used to look at people’s experiences with managing their medicine had not been used before and may not have captured patients’ experiences fully. Not as many people filled out surveys as expected at the start of the study. Having information from more people might have led to different results.
Future research could develop ways to make the programs easier for clinic staff and doctors to use.
How can people use the results?
Clinics can use the results when considering ways to improve mental health care.
Professional Abstract
Objective
To compare the effectiveness of two technology-supported approaches to improving people’s experience with medication management and shared decision making in mental health care
Study Design
| Design Element | Description |
|---|---|
| Design | Randomized controlled trial |
| Population | 2,363 adults enrolled in Medicaid and receiving mental health medication treatment at 1 of 14 community mental health centers |
| Interventions/ Comparators |
|
| Outcomes |
Primary: patient experience of care with medication management, collaborative decision making Secondary: patient engagement, engagement in outpatient mental health services, management of medication side effects, quality of life, mental health symptom severity, functional status |
| Timeframe | 2-year follow-up for primary outcomes |
This cluster randomized controlled trial compared two approaches to improve medication treatment outcomes focusing on collaborative decision making between people with mental illness and their providers. Researchers randomly assigned community mental health centers to provide either
- Measurement-Based Care (MBC): A web-based portal tracked participant data on symptoms, medication management, and side effects to help providers tailor medications and make treatment recommendations.
- Person-Centered Care (PCC): Peer specialists, people who have experienced mental illness trained to provide peer support, helped participants complete a web-based health self-assessment, which providers received before each visit. With support from peer specialists, participants identified treatment and recovery goals and prepared to collaborate with their providers during visits.
The study included 2,363 Medicaid-enrolled adults who received mental health treatment at one of 14 community mental health centers in Pennsylvania. Of these participants, 86% were white, 9% were black, and 5% were another race; 2% were Hispanic. The average age was 42, and 63% were female. Among these patients, 24% had bipolar disorder, 23% had major depression, 20% had a substance use disorder, 17% had another type of depression, 16% had schizophrenia or schizoaffective disorder, 10% had anxiety or posttraumatic stress disorder, and 10% had another diagnosis.
Participants completed four follow-up surveys between baseline and two years later. Researchers analyzed Medicaid claims data for the two-year period.
People with mental illness, mental health organizations, peer support staff, policy makers, mental health care providers, and researchers provided input on the study design.
Results
Patients receiving MBC and PCC did not differ on either experience with medication management or collaborative decision making. Both MBC and PCC participants reported improved experience with medication management over time (p<0.0001). However, this change was not clinically meaningful. Patients receiving PCC and MBC also did not have clinically meaningful differences for any of the secondary outcomes.
Limitations
Four participating community mental health centers withdrew from the study, requiring researchers to recruit replacement centers after the initial allocation. This change may have resulted in study population selection bias. Researchers created the measure for experience with medication management for this study because no other suitable assessment was available. The measure may not have adequately captured the construct. At one follow-up, participants completed fewer surveys than anticipated. Exposure to the approaches varied because of the nature of visits with psychiatric care providers, which may have impacted results.
Conclusions and Relevance
The approaches compared in this study did not differ meaningfully on any of the outcomes measured. Clinics may use these results when considering approaches for improving people’s experience with mental health medication management.
Future Research Needs
Future research could develop strategies for efficient and effective implementation of technology-supported interventions in under-resourced mental health care settings. Researchers could also examine how provider characteristics affect the adoption of technology-based collaborative decision making.
Final Research Report
View this project's final research report.
Journal Citations
Results of This Project
Related Journal Citations
Peer-Review Summary
Peer review of PCORI-funded research helps make sure the report presents complete, balanced, and useful information about the research. It also assesses how the project addressed PCORI’s Methodology Standards. During peer review, experts read a draft report of the research and provide comments about the report. These experts may include a scientist focused on the research topic, a specialist in research methods, a patient or caregiver, and a healthcare professional. These reviewers cannot have conflicts of interest with the study.
The peer reviewers point out where the draft report may need revision. For example, they may suggest ways to improve descriptions of the conduct of the study or to clarify the connection between results and conclusions. Sometimes, awardees revise their draft reports twice or more to address all of the reviewers’ comments.
Peer reviewers commented, and the researchers made changes or provided responses. The comments and responses included the following:
- Reviewers did not find the comparison of person-centered care (PCC) versus management-based care (MBC) compelling. They asked for more justification for this comparison based on theoretical and empirical evidence. The researchers expanded their description of the two interventions and added a conceptual framework comparing PCC, MBC, and a typical visit.
- Reviewers expressed concern about the problems in the study design created when several clinics dropped out postrandomization, and the researchers replaced the clinics without randomization. The researchers acknowledged these limitations, noting that this was a notable example of the problems in conducting research in the real world. They also added information about why clinic sites dropped out of the study and included analyses demonstrating that the new sites did not differ significantly from the dropped-out sites.
- Reviewers further questioned why the researchers did not use propensity score weighting or similar measures to account for the lack of randomization when completing their outcome analyses. The researchers noted that given their large number of participants and few significant baseline differences between the sites in the two intervention groups, they felt there was no need to seek additional precision through propensity scoring. The researchers went on to say that the small magnitude of the significant differences between PCC and MBC groups, and the results from sensitivity analyses looking at the effects of non-random site assignment both supported this decision.
- Reviewers asked why the researchers had completed subgroup analyses for Aim 3 given the problems with site randomization, substantial missing data, and the lack of differences between intervention groups. The researchers explained that along with these analyses being a part of their contracted work, they conducted the analyses with the assumption that the results would be exploratory or hypothesis generating rather than confirmatory or hypothesis testing. The researchers stated that they were cautious in their interpretation of the results.
- Reviewers commented that the assessment of side effects was simplistic and that unmeasured differences might appear and be the result of selection bias given the lack of true randomization. The researchers agreed that collecting more-detailed information about side effects would have been helpful, but said they wanted to minimize the number of survey questions so as to limit the burden on participants. The researchers felt that a single question on side effects would provide adequate information and prompt discussion between prescribers and patients. They went on to say that the single item provided the information that patient stakeholders identified as most important to them.