1. Early On® Redesign Update Michigan SICC Meeting November 16, 2006

2. SICC Presentation November 2006 Topics To Be Addressed Completed Components Completed Components Estimated Prevalence Report Estimated Prevalence Report Next Steps Next Steps

3. SICC Presentation November 2006 Completed Components Causes and Forces – April 2005 Causes and Forces – April 2005 Define Eligible Population* November 2005- Fall 2006 Define Eligible Population* November 2005- Fall 2006 Define Results – ECO Center Child and Family Outcomes Define Results – ECO Center Child and Family Outcomes Convene Results Teams – Summer 2005 Convene Results Teams – Summer 2005

4. Estimated Prevalence Report

5. SICC Presentation November 2006 Concept of prevalence What is prevalence? What is prevalence? prev.a.lence, n. The degree to which something occurs or exists. For example, the percentage of a population that meets a certain definition, or has a certain diagnosis.

6. SICC Presentation November 2006 Purpose of Early On Prevalence Study Estimate the number of children that could have a developmental delay and to understand the size of the potentially eligible population. Estimate the number of children that could have a developmental delay and to understand the size of the potentially eligible population. Identify the gap between who is currently being served and who could potentially be served. Identify the gap between who is currently being served and who could potentially be served. Leverage existing funds and advocate for new funding to help fill service gaps and meet compliance requirements. Leverage existing funds and advocate for new funding to help fill service gaps and meet compliance requirements.

7. SICC Presentation November 2006 Purpose of Early On Prevalence Study Inform the Early On system redesign efforts and strengthen connections with the broader early childhood system. Inform the Early On system redesign efforts and strengthen connections with the broader early childhood system. Support the recommendation that the ECIC use the prevalence study in their work of supporting the development of comprehensive early childhood systems. Support the recommendation that the ECIC use the prevalence study in their work of supporting the development of comprehensive early childhood systems. Recognize the potential numbers of children who will not meet the Early On eligibility criteria and will need to be served through other community resources. Recognize the potential numbers of children who will not meet the Early On eligibility criteria and will need to be served through other community resources.

8. What percentage of infants and toddlers participate in Early On?

9. SICC Presentation November 2006 Gogebic-Ontonagon ISD * 5.48% Traverse Bay Area ISD 3.41% Kalamazoo Valley ISD 2.16% Delta-Schoolcraft ISD * 5.41% Marquette-Alger ISD 3.27% Newaygo ISD 2.16% Hillsdale County ISD * 5.32% Lapeer County ISD 3.15% St. Joseph County ISD 2.16% Midland County ESA 4.74% Berrien County ISD 3.00% Genesee ISD 2.12% Van Buren ISD 4.36% Jackson County ISD 2.84% Eastern Upper Peninsula ISD 2.04% Saginaw ISD/ERC * 4.33% Shiawassee RESA 2.78% Livingston ESA 2.03% Monroe County ISD 4.27% Muskegon Area ISD 2.77% Macomb ISD 1.91% Ionia County ISD * 4.25% Clinton County RESA 2.75% Lenawee ISD 1.83% Ingham ISD 4.13% Calhoun ISD 2.74% Mecosta-Osceola ISD 1.74% Dickinson-Iron ISD 4.10% Clare-Gladwin ISD 2.72% Tuscola ISD 1.74% Ottawa Area ISD 4.09% COOR ISD 2.70% Barry ISD 1.69% Gratiot-Isabella RESD 3.90% Sanilac ISD 2.55% Washtenaw ISD 1.64% Lewis Cass ISD 3.77% Kent County ISD 2.49% Alpena-Montmorency-Alcona ESD 1.60% Iosco County ISD 3.77% Mason-Lake ISD 2.47% Huron ISD 1.45% Allegan County ISD 3.70% Charlevoix-Emmet ISD 2.39% St. Clair County ISD 1.41% Montcalm Area ISD 3.65% Eaton ISD 2.32% Oceana ISD 1.35% Wexford-Missaukee 3.56% Copper Country ISD 2.31% Wayne County RESA 0.87% Branch ISD 3.50% Bay-Arenac ISD 2.22% Oakland ISD 0.85% Manistee ISD 3.47% COP ISD 2.21% State average State average2.12% Menominee County ISD 3.47% Green plus 1 SD from Mean - Yellow minus 1 SD from mean December December 1, 2004 Child Count, Percentage of Children Served

10. What percentage of infants and toddlers participate in Part C nationally?

11. SICC Presentation November 2006

12. Methodology The premise of an estimated prevalence model is rooted in the notion that all communities within a state Part C system should serve the same percentage of children; except for accounting (indexing) for community differences in population characteristics that are likely predictors of participation in early intervention.

13. SICC Presentation November 2006 Steps in the prevalence study: Define the community for which you are estimating prevalence (ISDs). Calculate the percentage of children currently in service in each community. Select a projection model (i.e. population variables that are predictors of participation in early intervention). Compute an index. Estimated prevalence projection. Decide how results will play a role in your system. Continue to review, refine and update.

14. SICC Presentation November 2006 Step 1. Step 1. Define the community for which you are estimating prevalence For consistency with the current system, we are defining ‘community’ as ISD service areas. For consistency with the current system, we are defining ‘community’ as ISD service areas. Some data was only available on a county basis, in which case we had to combine county data to approximate ISD service areas. Some data was only available on a county basis, in which case we had to combine county data to approximate ISD service areas.

15. SICC Presentation November 2006 Step 2. Step 2. Calculate the percentage of children currently in service in each community Determine the number of children in service at a given point in time (‘shapshot’ count). Determine the number of children in service at a given point in time (‘shapshot’ count). Divide by the total number of children in the birth-3 age group (‘birth cohort’). Divide by the total number of children in the birth-3 age group (‘birth cohort’). Result is your percentage of children currently in service. Result is your percentage of children currently in service.

16. SICC Presentation November 2006 Step 3. Se Step 3. Select a projection model The projection model is made up of population variables that are predictors of participation in early intervention, typically social risk factors or health risk factors. There should be evidence of a link between the population variable and early intervention (e.g. through epidemiological studies, National Early Intervention Longitudinal Study, identified in IDEA as a target population, etc.)

17. SICC Presentation November 2006 Step 3. Step 3. REQUIRED Data Characteristics! 1. 1.Data must be readily available; 2. 2.Must be population-based data rather than participatory counts (e.g. birth certificate data, Census data); 3. 3.Must have a long history of consistent data collection; 4. 4.The data must be statistically reliable; 5. 5.The data must be available as both numbers and rates; and 6. 6.The data must be available for the state and county/ISD.

18. SICC Presentation November 2006 Step 3. Step 3. Population variables considered Michigan Natality statistics % of births to teen parents % of births to women with < high school education % of births to women >35 % of births to women who smoked % of births to unmarried women % of births that were low birthweight (< 2500g) % of births that were very low birthweight (< 1500g) % of births that were preterm (<37 weeks) Rate/1000 for infant mortality

19. SICC Presentation November 2006 Step 3. Step 3. Population variables considered CPS data% population aged 0-4 # of substantiated abuse/neglect CPS cases (meeting CAPTA criteria) Rate/1000 of substantiated CPS cases Rank for rate of CPS cases

20. SICC Presentation November 2006 Step 3. Step 3. Population variables considered Other descriptive dataRate/1000 of live births with reported birth defects % of 0-18 year olds insured by Medicaid % of 0-18 year olds insured by MIChild (SCHIP) % of children aged 0-17 living in poverty

21. SICC Presentation November 2006 Step 3. Step 3. Population variables considered Ethnicity% White % Black or African American % Native American/Alaskan % Asian % Hawaiian/Pacific Islander % of Hispanic/Latino origin

22. SICC Presentation November 2006 Step 3. Step 3. Population variables chosen Poverty Source: Bureau of the Census, Small Area Income Estimates; Poverty rate children 0-17 (2002) Low Birthweight (< 2500 grams) Natality statistics from Michigan 2003 resident birth files Mother < High School education at time of child’s birth Natality statistics from Michigan 2003 resident birth files Birth Defects Birth Defects Cases among Resident Live Births - Cases Diagnosed within 1 year of Birth by Residence County and Birth Year; Michigan Resident Birth Cohorts - 1999 through 2002 Preterm birth (< 37 weeks) Natality statistics from Michigan 2003 resident birth files Race (non-white population) Ethnicity - US Census 2000 (http://factfinder.census.gov/servlet/GCTTable?_b m=n&_lang=en&mt_name=DEC_2000_PL_U_GCTPL _ST2&format=ST-2&_box_head_nbr=GCT- PL&ds_name=DEC_2000_PL_U&geo_id=04000US26)

23. SICC Presentation November 2006 Step 3. Step 3. Population variables chosen, cont. Poverty Index 2002 Low Birth Weight 2003 <12 Yrs Education 2003 Birth Defects 99-02 Preterm Birth 2003 Race/ Ethnicity 2000 State of Michigan14.2%8.2%16.90%6.3%11.20%19.85%

24. SICC Presentation November 2006 Step 3. SMALL GROUP WORK Review Step 3. SMALL GROUP WORK Review Population data for your community Use Worksheet 1, and Table 1 from your LICC’s Data folder. Take about 15 minutes to review the questions on Worksheet 1 with your LICC members.

25. SICC Presentation November 2006 Step 4. Step 4. Compute an index The ‘Index’ is the number that accounts for community differences. Divide the percent of children in service by the sum of the population variables chosen for the model. The highest resulting number from across the communities will be the index that will be applied universally. It is important to verify that the index community has a reliable and valid percent of children in service, is using the state definition of eligibility, is using an eligibility determination process that is in compliance with rules and regulations, and their data represents an established pattern of service.

26. SICC Presentation November 2006 Step 5. Step 5. Estimated prevalence projection Sum 12-1-2004 Child Count % in serviceIndex Universal (State) Index Estimated Prevalence Percentage State of Michigan76.7%2.12% 0.02760.1014967.8%

27. SICC Presentation November 2006 Step 6. How will Michigan use the results of the Early On prevalence study? Estimate the number of children that could have a developmental delay and to understand the size of the potentially eligible population. Estimate the number of children that could have a developmental delay and to understand the size of the potentially eligible population. Identify the gap between who is currently being served and who could potentially be served. Identify the gap between who is currently being served and who could potentially be served. Leverage existing funds and advocate for new funding to help fill service gaps and meet compliance requirements. Leverage existing funds and advocate for new funding to help fill service gaps and meet compliance requirements.

28. SICC Presentation November 2006 Step 6. How will Michigan use the results of the Early On prevalence study? Inform the Early On system redesign efforts and strengthen connections with the broader early childhood system. Inform the Early On system redesign efforts and strengthen connections with the broader early childhood system. Support the recommendation that the ECIC use the prevalence study in their work of supporting the development of comprehensive early childhood systems. Support the recommendation that the ECIC use the prevalence study in their work of supporting the development of comprehensive early childhood systems. Recognize the potential numbers of children who will not meet the Early On eligibility criteria and will need to be served through other early childhood community resources. Recognize the potential numbers of children who will not meet the Early On eligibility criteria and will need to be served through other early childhood community resources.

29. Next Steps

30. SICC Presentation November 2006 Next Steps Completion of Cost Study Completion of Cost Study Development of Purchasing Plan Development of Purchasing Plan Triangulate: Triangulate: Causes and Forces Causes and Forces Results Teams’ Recommendations Results Teams’ Recommendations Revenue Information Revenue Information Allocate Resources Allocate Resources Development of Interagency Agreement Development of Interagency Agreement Final Report in February Final Report in February