Initiation And Engagement Performance Improvement Project

MHO: Verity Integrated Behavioral Health Care System

Project Leader: Charmaine Kinney

Telephone: 503-988-5464 x24424

Study Period: July 2006-March 2008

I. ACTIVITY 1: ASSESS THE STUDY METHODOLOGY

Step 1. REVIEW THE SELECTED STUDY TOPICS(s)

1.1 Relevant Information

  • Verity members reported that they did not have timely access to services after an initial intake appointment.
  • Verity contracts did not contain contractual requirements or incentives for timely access after intake.
  • The Oregon Administrative Rules do not contain timeframes that support timely access after intake
  • Initiation and engagement performance measures originated with alcohol and other drug services. The measure has been incorporated by the Health Plan Employer Data and Information Set (HEDIS) and is a required A&D performance measures for the National Council on Quality Assurance (NCQA) accreditation process.
  • Second National Forum on Performance Measures Carter Center April 2004, Washington Circle, Outcomes Roundtable for Children and Families, Forum Adult Workgroup-mental health, advocate use of initiation and engagement performance measure to improve access to services and measure impact motivational engagement interventions. The Forum Adult Workgroup completed modifications to the measure and a pilot study was conducted. National benchmarks for using this measure in the mental health field are limited at this time. The study looked at two populations of mental illness, SMI and selected mental disorders. Three sites were chosen as test sites. Rates for sites serving the general public were between 44% and 71% for initiation and 30% for engagement.
  • Verity hired a consultant in 2005 whom conducted two baseline reports for initiation and engagement (I&E) for January 16, 2004 through January 15, 2005 and January 1, 2005 through September 30, 2005. This data included all new treatment episodes and with all outpatient procedure codes and did not factor in continuous eligibility. The exclusion for continuous eligibility was identified as an important factor that could skew the data since we have a mobile population who move from different counties, and care could have already been established before an index appointment paid for by Verity. The measure was not implemented internally until late 2006 when an internal continuous eligibility algorithm was written.
  • An administrative problem came to light after the move to fee-for service in 2006 and early 2007 that may have affected productivity and the ability for providers to give the level of care and follow up required for our sickest clients. Encounters were not being submitted or they are being submitted incorrectly so are being kicked out of the billing system. A coding training session was provided in April and May to increase provider’s expertise in fee for service, and more complete billing has been occurring since.
  • In April 2006, Verity implemented a Fee-For-Service payment system under which agencies would paid for services rendered, rather than on the number of enrolled cases. Under the old capitation system there was no provider incentive to provide access to timely initial services after the intake. However, a significant drop in claims data was discovered. The drop in claims data was investigated, and it was increasingly clear that some of the providers and/or agencies did not have the expertise to fully utilize a fee-for-service system.
  • Initiation and engagement measures were included in some provider contracts, and data was broken out by agency to share the results with each agency. A plan was developed to have QM team is visit all major agencies to share data and discuss how customer service effects the outcome of this type of measure.
  • The 2007/2008 PIP cycle was further refined to eliminate authorizations for crisis and intensive services from the data, and limit the query of specific procedure codes to general outpatient. With these authorization types in the mix, the true level of successful initiation and engagement in general outpatient was potentially skewed. First, crisis services are not always followed up with general outpatient. Second, Verity expects intensive service authorizations to have a higher level of initiation and engagement since these clients are often seen at least twice a week. In order to have a true comparison data the project will limit data collection to July 2006 forward.
  • Age groups were collapsed in 2007/2008 data to simply adult and children’s categories in order to get sufficient numbers for the study.

1.2 Systematic and Prioritization process

  • Rationale: Verity QM reviewed the problem areas and prioritized access due to the frequency of complaints, feedback and discussions from CMHP advisory groups, member focus groups, Verity QM Committee discussions and during direct client interviews during agency certification reviews, and an EQR finding. In 2006, Access problems consisted of 38% of all complaints and had the highest number of complaints in any category. Access continues to be a concern for consumers and providers are beginning to seriously implement interventions at the agency level to improve overall access. After reviewing several methods for intervention improvement strategies, such as access reports, or having initial appointments set up through the call center, the Verity QM Committee and Verity Management decided on initiation and engagement as the measure, and to include the requirement to achieve the baseline goal and to work on developing a goal in all future contracts as the intervention. Initiation and Engagement was selected and approved by the QM Committee in April 2005. Consumers wanted more timely access, and this topic was chosen because the measure quantified timely access in a way that was validated in the addictions medicine field as a Hedis measure. Mental Health and Addictions medicine have a similar populations and consumers often overlap treatment for these problems.
  • However, consumers and agencies agreed that the initial appointment was not as much of a concern as getting follow-up appointments after the intake. Access problems to ongoing appointments after the index appointment was identified through the following means:
  • Verity received complaints that revealed an increased trend of lack of access problems after the initial index appointment.
  • Call center was receiving member calls to intervene because members could not get in follow-up services due to unavailable appointments.
  • Call Center needed to call agencies to arrange for priority appointments or assisting with alternative agency referrals or arranging for walk-in appointments for urgent care more frequently than considered appropriate.
  • National HEDIS measures for mental health is follow up within 7 days after an inpatient hospitalization. However, this measure excludes clients who access outpatient mental health without a serious condition that leads to hospitalization. Verity wanted a measure that would include the whole population when measuring access.
  • Verity strategy to have timely initiation of care and engagement in service should decrease need for higher levels of care, because members decompensate while waiting for services. Evidence reviewed by national groups above demonstrates that quicker access and consistent appointments during the initial episode of care will decrease over all utilization by improving client outcome.
  • Verity then tried to validate what members and call center staff reported through a data analysis. An external consultant was hired in 2005 to conduct baseline measurements for I&E and to develop the measurement specifications. Although baseline data was within the baseline range for initiation and above the engagement findings in the Pilot Study results, (Teague & Trabin, 2006), the percentage was still unacceptable to consumer representatives and Verity staff. In addition since we did not use diagnostic criteria and used only index appointments as the denominator, the comparison is not exact. Baseline data for I&E for the 2004 and 2005 time periods are:

Children and Adults / 1/16/2004-1/15/2005 / 1/1/2005-9/30/2005 / Comparative Data from Teague & Trabin Pilot Project
Met Initiation / 53% / 50% / 44.7%-71.9% depending on site and MH severity
Met Engagement / 32% / 32% / 12.7%-31.5% depending on site and MH severity
  • A long-term goal of 70% for initiation and 50% for engagement was discussed in 2005, but the providers felt this goal was unattainable under present systems. It was agreed to set a baseline and determine a goal in the future. The long-term goals above were approved in 2007.

Step 2. REVIEW THE STUDY QUESTIONS(s)

2.1 Clearly define the questions the study is designed to answer.

  • Identification: The number of Verity clients who have not received services 120 prior to an index appointment and are receiving routine outpatient services for a new episode of care?
  • Initiation: What percent of clients receive a second visit within 14 day after the index appointment for a new episode of care?
  • Engagement: What percent of clients who successfully met initiation also receive a third and fourth visit within 30 days of the second visit? What percent of clients who did not successfully meet initiation have 4 visits within 45 days by providers? Generally seen as 4 visits within 45 days by providers.

Step 3. REVIEW SELECTED STUDY INDICATOR(s)

3.1 Indicators must be objective, measurable, clearly defined, unambiguous statements of an aspect of quality to be measured.

Initiation Numerator:

Total children and adult Verity members who have an additional mental health service within 14 days of an index appointment.

Initiation Denominator:

Total Verity members who have above the line diagnosis and who have not received mental health services 120 days prior to an index general mental health outpatient appointment.

Engagement Numerator:

Total children and adult Verity members who have two additional mental health services within 30 days after initiation.

Engagement Denominator:

Total Verity members who have above the line diagnosis and who have not received mental health services 120 days prior to an index general mental health outpatient appointment.

3.2 Indicators must be capable of measuring enrollee outcomes, enrollee satisfaction, or process of care strongly associated with improved enrollee outcomes.

There are no direct data on the correlation between successful initiation and engagement and successful outcomes given the newness of the measure in the mental health field. There is evidence of successful continuity of care and good healthcare outcomes in several disciplines including mental health. National HEDIS measures for mental health is follow up within 7 days after an inpatient hospitalization. However, this measure excludes clients who access outpatient mental health without a serious condition that leads to hospitalization. Verity wanted a measure that would include the whole population when measuring access. Verity consumers and staff feel that access is a critical component of providing effective mental health treatment. Initiation and Engagement has proven to be an effective intervention in successful treatment strategies for addictions medicine. These two fields have similarities that would suggest that this would also be an effective intervention in Mental Health consumers. Also, Verity has a high level of dual diagnosis clients and the consumers and staff believe that this proven method will work to have more clients successfully engage in and complete treatment.

Rationale for initiation and engagement: Verity believes that the best practice for initiating and engaging consumers in care suggests that an individual receive two visits within the first 14 days of care and an additional 2 visits within the next 30 days (a total of 4 visits within the first 45 days of service). These timelines provide the best opportunity for an individual to become fully engaged in services that can promote recovery and stability. Mental health best practices promote recovery and stability for the client and initiation and engagement is believed to support continuity of services and positive outcomes.

The following documentation shows the history of successful initiation and engagement measures and the relation between continuity of care and improved outcomes are attached:

Process Of Care Measures For Mental Health: Report Of A Pilot Study, Gregory B. Teague, Ph.D. and Tom Trabin, Ph.D.,Draft, April 5, 2006 See Attachment 3

This report describes the measures development in the alcohol and drug services (AOD) and the specific criteria for the original measure. The authors then explain the differences between the AOD populations and mental health populations and where the measure could be modified for the mental health field. Diagnostic inclusion criteria was developed to address the differences in treatment intensity in mental health, selection of the denominator, the service-free period, and required data elements. The report then gives detailed measure methods and reports the results for three sites. The chart in section 1.1 has the total data for SEL and SMI initiation and engagement by setting and gender, (The chart only includes all study subjects not broken out by gender.) The authors concluded that the measure was feasible to apply the measure to the mental health field with the modifications.

A Practical Approach to Performance Measurement and Quality Improvement for Children’s Services, Bridging the Gap Between Research and Practice

2003 Tools That Work Conference, Miami, Florida, November 2003 Astrid Beigel, Ph.D., Doreen A. Cavanaugh, Ph.D. ,Ann Doucette, Ph.D See Attachment 4

This PowerPoint presentation was from a training and discussion of relevant performance improvement tools that could be implemented in the mental health field. The mission of the group was to develop common performance measures across systems, public and private in order to improve the quality of services and to influence practice and policy. Measurement tools and methods were covered, the core principles of federal initiative requiring performance measurement and how to select measures that were relevant, scientifically sound, and feasible. I&E measures from the AOD field were presented as meeting all these criteria in both adult and children’s populations.

Continuity of Care and Health Care Costs Among Persons With Severe Mental Illness, Mitton, Craig R., et al Psychiatric Services, September 2005 Vol. 56 No. 9 See Attachment 5

The journal article reviewed the association between continuity of care and health care costs for a severally mentally ill population. Improved continuity of care was associated with lower hospital costs but higher community costs. Total costs were not significant for overall total costs. However, total costs for patients who had higher self-rated quality of life indicators and those with higher functioning were significantly lower.

Step 4. REVIEW THE IDENTIFIED STUDY POPULATION and sampling methods

4.1 Study population is clearly defined so that all the Verity’s Medicaid enrollees are eligible for the study is included.

Study population includes all members in service who had no mental health services for 120 days before an index appointment, who are continuously eligible for Verity coverage for the full study period (120 before and 45 days after an index appointment), who were initiating a new outpatient episode of care, and all age-groups.

Study does not involving sampling

4.2 Study includes Verity’s entire eligible population, the data collection approach captures all eligible enrollees

Study captures all eligible enrollees, baseline data and follow-up data collected from claims and encounter data system and follow-up data is collected from same system. Age groups were collapsed in 2007/2008 data to adult and children’s categories and not stratified by age in order to get sufficient numbers for the study. The data is grouped into two age groups, children 0-17 and adults 18+.

4.3 Verity has described the method for determining the sample size.

N/A

4.4 Sampling methodology is valid and protects against bias

N/A

4.5 Sample size should be large enough to allow calculation of statistically meaningful measures

N/A

Step 5. REVIEW the MHO’s Data Collection Procedures

5.1 Study design clearly specifies the data to be collected

Encounter/Claims submissions and authorization information from PHTech.

5.2 Data sources are clearly identified

Outpatient encounter data used in the analysis comes from the Verity Encounter Master table, the PHTech Referral Table, and the PHTech Claims Extract Table. Eligibility data used in this analysis comes from the Verity enrollee table. All of these tables are contained within SQL server databases. An Access 2000 database is utilized to connect to these tables via ODBC links. Queries and tables are set up in Access to extract and analyze the data. For specific codes, please see Performance Specifications at the end of this document.