Other State Agency Study

Frequently Asked Questions
Early Intervention


Q1: Our state's Early Childhood Mental Health Program provides mental health services to children as well as training(s) for parents. Program recipients may also exhibit developmental delays and/or mental retardation, other than a mental health issue. Should we report this program? How should we report the number of clients served?

A: Yes, this program should be captured and reported since the intent for which it was established (given the program name) is within the scope of this project. The ideal reporting procedure is to separate out expenditures on mental health services to children with known mental health disorders (including those with developmental delays and/or mental retardation), including trainings to parents of these children. If this is not feasible, you may report the total program expenditures.  With regard to reporting the number of clients, use the program definition and counting method for clients. Please provide adequate (data) notes to ensure appropriate interpretation of data. Include program definition and counting method for clients served. For example, if the count is under-reported because parents who received training are not tracked, please state this in your notes.


File Matching Procedure


Q1: Do you feel it is necessary to pull 'Date of Birth' as an individual identifier? Currently, we have the individual's name and SS# to use to determine OSA overlap but feel that may not be enough to identify the correct person.

A: To use Date of Birth as a single file (matching) key may not allow you to maximize your results from the matching procedure undertaken. According to a presentation given by Dennis Nalty, the use of multiple criteria (identifiers) will give better outcome of the match. Please refer to the NRI website, Technical Assistance, for presentations given on an April 2006 meeting that are related to this topic. Dan Whalen's and Dennis Nalty's presentations provide useful hints about file matching.  A sample of matching algorithm is provided in Dan Whalen's presentation as follows:


Washington Mental Health Division Matching Algorithm

1. First Match - Name, DOB, Gender must have valid date for DOB, and surname is not "unknown", "Doe" or blank.  A match exists when last name, first name, DOB and gender are the same - or last name, DOB and gender are the same

2. Second Match - Last name, SSN Must have valid SSN, and last name is not "unknown", "Doe" or blank.  A match exists when last name and SSN are the same3. Third Match - SSN, DOB.  Must have a valid DOB, SSNA match occurs when DOB and SSN are the same.


General - Reporting Period


Q1: What if only one agency cannot give real up-to-date data? For example, if we can get SFY 06 information from all agencies except for one?

A: It is okay to report that one agency with different reporting period. A bit of caution should be observed in terms of determining the overlap. You have to use the same reporting period roster of clients to be able to calculate the overlap.




Q1: We were told that domestic violence shelters are excluded from having to report to HMIS (even though they may receive HUD homeless funding). We intend to exclude them from our purview even though we might be able to get some client level information through another source (a different state agency). Is that correct?

A: Yes, you may exclude domestic violence shelters. Return to top of page


Q2: Do we count local match (e.g. Shelter plus Care requires a dollar per dollar service match from the local provider)?

A: Yes, report total expenditures - local plus federal. In the Fund Source table, please report the federal dollars under "Federal" and the local counterpart under "Local". In the expenditures table, report them according to the type of service they were expended on. It is not necessary to distinguish how much of the local match went to which type of service. Even though the match comes from the local provider, hence not strictly state dollars, this is however an enabler to drawing down of the federal funds. Return to top of page


Q3: What was the rationale for separating out HUD's Supportive Housing Program from other housing programs (when there are other housing programs that receive HUD McKinney homeless funding and that are required to use one specific data system - the Housing Management Information System (HMIS))?

A: The "Other housing" was meant to be a catch-all category for other state housing programs not funded by HUD. The only rationale for this category is to avoid exclusion of housing programs not funded or receiving funds from HUD. Report the total expenditures of state housing programs that receive as a proportion (1%-100%) of its total funding from the HUD Supportive Housing Program. As a rule of thumb, use the other category (non-HUD) only if the housing program gets 100% of its fund from non-HUD sources. Return to top of page


Q4: Should I pursue contacts with other state agencies that may be responsible for specific housing programs listed in the protocol even if they are not agencies in the original list of OSAs? For example, CDBG projects are funded out of the Governor's Office for Local Development.

A: Yes, we encourage you to seek participation of other state agencies (including Governor's Office) even if they are not in the original list of OSAs if they have control over the funds for providing housing for persons with mental illness. Return to top of page


Q5: While the Housing Management Information System holds client files and will have diagnostic information, there is great concern about data accuracy. It is understood that at the program level the Annual Performance Report (APR) is more accurate as to total numbers of persons served in programs but does not have client level data. Could we use APR numbers to report on the "number of persons served" chart while using HMIS client level data to conduct data matching with our public mental health system information?

A: Yes, but when reporting please qualify your data. It is also recommended that you conduct a comparative analysis of the number of clients reported by both databases and determine the percentage of error/variance.


Q6: Housing structure expense might be broken down into three categories:

·         Development cost

·         Operating cost

·         Rental Assistance cost


Are we looking at housing structure expense just for those projects that came on line (began operating) during the most recent state fiscal year? Or are we looking at all projects in the housing portfolio no matter when they were put into service? Or just those projects that we're sure house individuals with mental illness?

A: Housing structure expense should include expenses that aid individuals and families to obtain and maintain housing such as rental or utility subsidies, housing repairs, relocation services, move-in costs, and other similar supportive housing services provided to assist persons with mental illness gain residential stability.


To reflect the above definition on the data table, the "Housing Structure Expense" category has been changed to "Supportive Housing Wraparound Costs." This is to distinguish it from actual mental health services like skills training, case management, etc. which may also be provided under the generic classification of supportive housing services but have to be reported under "Mental Health and Other Support Services".


Development and operating costs involved in the construction of new housing, renovation of existing housing structures for persons with mental illness which include brick and mortar cost, mortgage amortization, cost of remodeling a housing facility, and building depreciation are excluded in reporting for this OSA Study. States, however, may opt to collect this information and report the data using the prescribed procedure for project extension (see Main Text Protocol, page 5).  Please see the Revised Housing Data Table. (Excel, 46 KB)


Q7: Can NRI get state level data from HUD (like with the VR data)? Basically this would be aggregate counts from the individual Annual Performance (AP) Reports from programs like SHP, ESG and Shelter plus Care. Mental illness is apparently identified in these reports as a % of total individuals served. Or should we simply use the numbers that are available to us under our state's Homeless Management Information System although we are aware the data is understated than what was stated in the programs' AP Reports?

A: If the reported housing agency/program total expenditures for persons with mental illness are determined by applying an assumed percentage of the overall program/agency housing expenditures, then it is appropriate to apply the same method (not necessarily the same percentage) in determining the total number of individuals with mental illness who were served. Please use this method only in cases where the housing program/agency does not have a mental health identifier (see Protocol for various types of identifiers you may use) for people receiving housing support.  If you are to use this method, it is important that you include in the Data Notes the percentage used, how it was determined or source of assumed percentage (e.g. based on past 5 years of state data, assumed rate by HUD, etc.) and the method of application. It is also important to state in your Data Notes the attempt made to use the methods proposed in the Protocol and reason for using an alternate method.  NRI will approach the HUD main office to access a national performance report database of state-level HUD programs relevant to our project, if one exists. However, states are still encouraged to use local contacts to get the information due to possible delay in getting the data at this level. If we do get the data, there is a reasonable chance of at least a one-year reporting lag time.


Q8: Within the Section 8 Housing Voucher program, we plan to either match our CMHC client database with their housing voucher roster or use PPE. If we do the matching using either method, is it appropriate to use the percentage matched (in a backwards way) to apply to the total voucher holders (about 4160) to arrive at the number of persons served?

A: You may use this method as a last resort. States are encouraged, whenever feasible, to identify OSA clients with mental illness using the methods described in the protocol because this will provide more meaningful results with strong policy implications. The calculation of overlap and the count of those who are served solely by the OSA will be more precise. Again, it is important to document why an alternate method is used.


Q9: Given that we can likely get the total number of participants in most of the housing agency's programs, is it ever appropriate to apply some percentage (estimate) from the literature or some other study to the total to arrive at the number of individuals with mental illness in the database for programs where disability is not identified?

A: As explained in a related question, this method is acceptable only in certain cases and it is important for the state to document the following information:

·         Rate applied

·         Source of data for the rate or how the rate was determined

·         Method used in applying the rate

·         Note the attempts made in using the proposed methods in the Protocol and why the proposed methods were not feasible




Q1: Should we be matching with our community mental health center data only or should we be also matching with our state hospital records so that a "match" is anyone who we find either in our community service system or in our hospital system during state fiscal year 2006?

A: Under the URS, states usually report the state hospital count in the total number of persons served by the state mental health agency even if they are captured under different data systems. To remain consistent with URS, please include the state hospital roster with the community MH roster when matching files with other state agencies.


Medicaid/Elderly Services


Q1: Some of the clients were eligible for more than one category for some of their claims. Do we report it to all eligible categories?

A: Medicaid eligibility criteria vary by states and so do state eligibility code classification. State would need to do a crosswalk of the state eligibility code classification with the Medicaid Eligibility Categories used in the OSA data table (note: these categories were not been changed), as follows:

·         Disabled (SSI)

·         TANF (Temporary Assistance for Needy Families)

·         Dually Eligible for Medicaid and Medicare Foster Care and Adoption Assistance

·         Other (e.g. Medically Needy, Refugees)


We encourage states to enlist the assistance and agreement of the State Medicaid agency in developing the crosswalk and for states to report the crosswalk in the Contextual Report. Technically, an individual may receive eligibility from more than one Medicaid eligibility group. However, under this circumstance, states usually register the person under one eligibility group only. Please consult the state Medicaid agency the specific data field you would need to use to identify the appropriate eligibility criteria. Throughout the reporting year, an individual may successively be enrolled in more than one eligibility group (i.e. after disenrollment from TANF, an individual may be enrolled under Foster Care). Under this circumstance, you may report the expenditures and the count of person in all respective eligibility groups. To put the data in context, however, it is important for the state to document in the Contextual Report the unduplicated count of persons who received a Medicaid service. Individuals who are eligible for SCHIP cannot be dually qualified for Medicaid.


Q2: When Medicare pays for the service do we need to include that in the expenditures for Medicaid? Do we include the premiums for Medicare that Medicaid pays? The premium for Medicare includes physical health and mental health, should they be included in the report?

A: If the service is provided by Medicare, it should not be reported under Medicaid. However, Medicare premiums paid by Medicaid should be reported.


Q3: National studies have found that somewhere between 40 - 70% of all nursing home residents have some type of mental health disorder. However Indiana's NF (Nursing Facilities) level of care does not, with the exception of dementia/Alzheimer's disease, allow admission based solely on a mental health condition. Payment for mental health services and medications is not included in the daily per diem rate so we can identify those costs. Since the individuals are qualifying for nursing home care based on their physical condition my assumption is that the nursing home costs (the daily rate provided) would not be counted in this study. Am I correct?

A: Yes, the nursing home costs (based on the per diem) are excluded in this study. However, please report the cost of mental health services and medications provided to residents who are receiving these services.


Q4: If the NF placement is because of Alzheimer's disease/dementia should the per diem rate be counted? My feeling is no. The percentage of nursing home residents with dementia is so high that I think it would give a picture of the resources being expended for mental health treatment. Even if explained it could easily be misinterpreted.

A: The per diem rate should not be reported.


Q5: Indiana has a housing program called the Residential Care Assistance Program (RCAP). It is the basically the housing of last resort for individuals that are elderly and/or disabled and poor. The qualifying disability can be mental illness and a very large percentage of residents do have an MI diagnosis, although it may not be their primary diagnosis or reason they are there. The daily rate covers room, board, laundry and minimal supervision. Any health or mental health services are paid for separately, primarily by Medicaid, and data for that is available. Should the daily rate be counted?

A: No, do not include the daily rate. However, please report the cost of mental health services and medications provided to residents of this housing program.


Vocational Rehabilitation


Q1: We re-read the protocol (sections 2 and 3) and realized that the expenditures include ALL services provided by DVR and the distinctions to be made require us to break out the costs into the different types (supported employment, VR/Training and Other). If this is not accurate, please let us know. 

A: Yes, please use the service types specified in the data template for VR to sort out expenditures for people with mental illness who are receiving vocational rehabilitation services. For this particular agency, the expenditures are not limited to mental health services only because VR is basically treated as a 'MH support service'. Alternatively stated, the expenditures that should be captured include both the mental health service expenditures and the MH support service, i.e. the employment rehabilitation services provided by the VR agency.


Q2: The population called 'Eligible for VR due to Non-Psychiatric Disability (but receiving MH services)'-- should this be reported under 'all others'? Do we include all clients (with other disabilities) in that larger bucket or just clients with other disabilities, who also have received MH-specific services?

A: The VR dataset has data fields for primary and secondary disability. For example, a person's primary disability is "blindness" and the secondary disability is "Cognitive Impairment". Searching for mental health as a reported disability in both fields is the way to count individuals to be reported. Individuals with mental illness as the primary disability will be reported under 'Eligible for VR due to Psychiatric Disability' and those with mental illness as the secondary disability will be reported under 'Eligible for VR due to Non-Psychiatric Disability'.


Q3: Two reports from the state VR agency were received showing "purchased" services received based upon a mental health diagnosis. One report was for active case status; the other was for closed case status. The number of consumers and total expenditures for each service is provided in the report. When we provide a count of the "number of persons served" for the Vocational Rehabilitation grid I am assuming we will combine active and closed cases?
A: Yes, please combine both the active and closed cases when reporting the total number of persons served. However, it is advisable to report in your data notes the number and percentage of active and closed cases.


Q4: When we attempt to provide an unduplicated number of persons served across services (column 1), should we add in individuals with psychiatric disabilities who may not have received a "purchased service" during the fiscal year? There are evidently a sizable number of clients who are in different status categories such as "referred", "applied" and "closed ineligible" to name a few.
A: No, please report only those individuals enrolled to receive the vocational rehabilitation services. (This is assuming that the referred, applied, and closed ineligible were not provided any services by the VR agency other than the conduct of eligibility evaluation/assessment.)

Please report in your data notes the number of people with psychiatric disabilities who were referred, applied, and closed ineligible. You may include the total cost and service type (e.g. evaluation) incurred by the VR agency for this group, if available.


Q5: There is evidently a separate database at our Vocational Rehabilitation office for individuals who are blind. Some of these individuals may have a mental health diagnosis as a secondary diagnosis. Should we attempt to count these individuals even though the database is a different one and not under the control of the individuals I am working with?
A: Yes, the intent is to capture all individuals with psychiatric disabilities who received/are receiving vocational rehabilitation service where their impairment (primary or secondary) or source of impairment is due to mental illness.


Q6: The "expenditures" and "number of persons served" portions of the VR grid differentiate between "eligible for VR due to psychiatric disability" and "eligible for VR due to non-psychiatric disability (but receiving MH Services)". Is this language referring only to people who have gotten far enough along in the process as to be deemed eligible for VR services? Or, is this simply referring to all individuals who apply for services based on the fact that they have a mental health diagnosis? (Refer to question 2 above).
A: This refers only to individuals who are already deemed eligible by the vocational rehabilitation agency to receive vocational rehabilitation services.


Q7: Revenue for VR is primarily federal but based on a 25% state match. Expenditures are evidently about equally split between "purchased services" and costs associated with VR personnel and other operating costs. Should we attempt to place a value on the non-purchased services cost such as multiplying the total other operating cost by the percent of overall cases that are based on psychiatric disability? Or, should we simply disregard this cost as the services that are provided are provided equally to all VR clients?

A: Counseling and/or evaluation is a standard cost to the agency since this service is provided to all VR clients. It is not necessary to impute a cost for this.

Please report the 25% state match under the state column of the Revenue Table.


Q8: Our state VR agency is on a federal fiscal year (10/1/05 - 9/30/06) as all state VR agencies would be (in sync with RSA reporting). Is it okay to have them pull data for that 12 month period while we would be matching with our SMHA data on state fiscal year basis? (July 1, 2005 - June 30, 2006).
A: Yes, the VR data can be reported using federal fiscal year to remain in sync with RSA reporting. You may use the SFY to match with SMHA data for the overlap, with proper annotation.


Q9: For individuals who are receiving VR services with psychiatric disability, do we include all services VR is paying for under "other" or just mental health related services?
A:The data template provides a breakdown by service type: supported employment, vocational rehabilitation, and others. Sort out expenditures using these service types and for services that do not fall in either the SE or VR categories, please report them under 'Others'. In your data notes, please cite specific examples on what services were included in 'Others'. Non-mental health services may be reported only if they were essential to the delivery of VR services to persons with mental illness. If the non-mental health service is provided to all VR recipients (whether they have psychiatric disability or not), then this becomes a 'standard' service and therefore excluded in reporting.