|
The Application of Population Estimation (PE) to Mental Health Research: An Evaluation of Methods
|
| Presenter(s):
|
| Tanmaykumar Patel,
University of South Florida,
drtany@gmail.com
|
| Marion Becker,
University of South Florida,
becker@fmhi.usf.edu
|
| Ezra Ochshorn,
University of South Florida,
ochshorn@fmhi.usf.edu
|
| Abstract:
Although administrative data sets are available for public use, personal identifying information is often restricted from use because of laws regarding the privacy, and confidentiality of protected health information. This study examined the comparative usefulness and accuracy of two leading methods of population estimation (PE) used to produce valid measures of the population overlap and treatment outcomes for mental health consumers when personal identification information (IDs) such as social security numbers (SSNs) are unavailable. The authors merged anonymous extracts of Medicaid and involuntary civil commitment datasets to produce estimates of population overlap and treatment outcome for Medicaid enrolled individuals diagnosed with major depressive disorder who were also experienced an involuntarily psychiatric examination. Results obtained using PE method developed by Dr. Steven Banks and the method proposed by Dr. Eugene Laska are compared with those obtained using SSNs. Implications for program planning and evaluation are discussed.
|
|
Successes and Challenges in Using a Web-Based Survey for Community Data Collection
|
| Presenter(s):
|
| Shelly Kowalczyk,
MayaTech Corporation,
skowalczyk@mayatech.com
|
| Kristianna Pettibone,
MayaTech Corporation,
kpettibone@mayatech.com
|
| Abstract:
This session will show the successes and challenges associated with using a Web-based survey to collect data from communities participating in CSAP’s Strategic Prevention Framework State Incentive Grant Initiative. As part of the cross-site evaluation, we developed a Web-based survey to evaluate communities’ progress through the framework, including: needs assessment, capacity building, strategic planning, intervention implementation and evaluation. The survey includes closed-ended and open-ended responses for the collection of quantitative and qualitative data.
Challenges encountered and resultant successes include maximizing user-friendliness by developing capabilities such as automated skip patterns; reducing respondent burden while effectively tracking process measures and providing real-time data; and providing resource intensive training and TA, allowing users to efficiently complete the survey.
Using the Web-based survey resulted in high completion rates. Data from 326 communities implementing 599 interventions were collected. The completion rate at the due date was 83%, increasing to 98% after two weeks.
|
|
Mental Health Treatment Study: Preliminary Evidence on the Demand for Employment Assistance and Supports by SSDI Beneficiaries with Mental Disorders
|
| Presenter(s):
|
| David Salkever,
University of Maryland Baltimore County,
salkever@umbc.edu
|
| Mustafa Karakus,
Westat,
mustafakarakus@westat.com
|
| William Frey,
Westat,
williamfrey@westat.com
|
| Dave Marcotte,
University of Maryland Baltimore County,
marcotte@umbc.edu
|
| Abstract:
This paper reports on our preliminary analysis of the factors that influence beneficiaries’ decisions to enroll in the MHTS. Binary logistic regressions of enrollment outcome include demographic and diagnostic information on each beneficiary, level of their monthly SSDI benefit, length of time they have been an SSDI beneficiary, SSI recipient status, distance to the intervention service site, county unemployment rate, per cent African-American and per capita income level in the 5-digit zip code. Results indicate that most of these predictors are highly significant. Findings for age and time-on-benefits are of particular interest in testing the common “culture of disability” thesis that long-time recipients have the least interest in working. Unemployment rate results show a positive impact of labor-market weakness on enrollment. Viewing the distance variable as a proxy for the time plus travel “price” of participation, we find a significant but relatively small negative impact of price on enrollment.
|
|
Strategies to Increase Parental Consent in Active Consent Situations: Lessons Learned from Two Studies
|
| Presenter(s):
|
| Kristen Ogilvie,
Pacific Institute for Research and Evaluation,
kogilvie@pire.org
|
| Matt Courser,
Pacific Institute for Research & Evaluation,
mcourser@pire.org
|
| Jennifer Norland,
Pacific Institute for Research and Evaluation,
jnorland@pire.org
|
| Melodie Fair,
University of Alaska Anchorage,
anmdf@uaa.alaska.edu
|
| Abstract:
This paper explores the strategies used in two NIDA-funded studies in Alaska to obtain parental consent for a drug use and attitude survey administered to middle-school-aged children. The Alaska legislature passed an active consent law in 1999 that requires schools to obtain positive written consent prior to the administration of any survey that requests personal information from students in public schools. Under these active consent conditions, the parental consent rates in the two studies, which involved nearly identical surveys in similar communities, were vastly different. The first study had an overall baseline survey consent form return rate of 69% while the second study increased this rate to 92%. The second study’s strategy to obtain parental consent was informed by the first study’s shortcomings. This paper examines what lessons learned in the first study helped increase consent form return rates in the second study.
|
| | | |