Family Care Residential Rate Setting Methodology Project July

Family Care Residential Rate Setting Methodology Project July 12, 2010 1

Introduction l Early 2010, DHS initiated residential rate setting project for the Family Care program. – l 2 Purpose of project: establish a uniform methodology for determining all-inclusive rates for services provided by community-based residential programs that reflect the reasonable costs of services in relation to the acuity of members served. In mid-April, DHS held listening sessions to gather initial input from stakeholders on issues to be considered in the development of the methodology.

Introduction l Since April the Department has been reviewing data and possible methodological approaches, including approaches used in other states. – l 3 On July 1, briefing was provided by John Villegas-Grubbs, a consultant who developed residential rate setting models in several states and localities Purpose of today’s meeting: provide an update on the status of the Department’s residential rate setting project and discuss possible approaches to the development of a rate setting methodology.

Residential Rate-Setting Team Members l l l l 4 Fredi Bove, DLTC Deputy Administrator Tom Lawless, Director, DLTC Bureau of Financial Management Rebecca Hotynski, DLTC Bureau of Financial Management Monica Deignan, Director, Managed Care Section Charlie Jones, Managed Care Section Deb Rathermel, Managed Care Section Dana Raue, Project Manager Jim Robinson, Director, UW Center for Health Systems and Research Analysis (CHSRA)

Key Elements in Residential Rate Setting Model § § § 5 Client Acuity Provider cost Regional cost factors Provider characteristics, including type of facility (AFH, RCAC, CBRF), size, etc. Different approaches utilize these key factors in different ways

Client Acuity l After reviewing possible client acuity measures, at this point DHS views that the long-term care functional screen (LTCFS) is the most appropriate tool to use as the base for client acuity. Strengths of the LTCFS are: – – – 6 Includes a broad and comprehensive number of characteristics. We recognize that it is not perfect; and are reviewing whether it should be supplemented in certain areas for purposes of the residential rate setting project; Strong inter-rater reliability, an important quality highlighted by Mr. Villegas-Grubbs; Subject to an established, strong quality oversight system which identifies attempts at manipulation for financial gain (i. e. , “upcoding”), an important quality highlighted by Mr. Villegas-Grubbs; Measures the needs of an individual in the absence of any current supports, the third important quality identified by Mr. Villegas-Grubbs; Already operative, avoiding the imposition of new administrative activities and costs on MCOs or providers;

Provider Cost l The most direct source of provider cost data is audited cost data from providers. However, DHS does not have a centralized, automated data set that includes audited cost data from every residential provider in Family Care. – l 7 DHS has a limited set of provider budget data collected by two MCOs (Community Care and Western Wisconsin Cares) and data transmitted voluntarily by providers in the past four weeks. DHS has detailed, comprehensive payment data by provider through the encounter system. We are seeking to define a universe of encounter data that represents payments by mature MCOs in a stable environment, which would be expected to be closely aligned with provider cost and therefore can serve as a good proxy for provider costs.

Provider Cost l DHS intends to use both sources of data, encounter data and provider data, in the residential rate setting model. – – 8 Because encounter data is comprehensive and can be linked to individual members and their acuity, appears most feasible and sound to use the encounter data as the base data to develop a preliminary model, which would be subject to further review and refinements The more limited provider cost data set will be used as a subsequent step to test the validity of the preliminary model and identify and develop refinements.

Provider Cost l Final model will need to include a mechanism to handle “exceptions”, as noted in the listening sessions and by Mr. Villegas-Grubbs – 9 Some providers will have higher than predicted costs due to unique programmatic or other factors

Provider Characteristics l l 10 Division of Quality Assurance regulatory database includes data on provider characteristics, such as size, target group served, and location for residential providers regulated by the state (CBRFs, RCACs, and 3 -4 bed AFHs). DHS plans to explore use of this data base in the residential rate setting model.

Regional Variation l Two sources of measures for regional cost differences: – – l 11 External, federally-established regional labor wage rates published by the Bureau of Labor Statistics, as noted by Mr. Villegas-Grubbs; and identification of regional cost differentials as determined through a regression model. DHS plans to explore use of both approaches to determine which is stronger

Possible Methodologies l l Allowable cost reimbursement model (used under the legacy waiver programs) l JVGA model l 12 MCO acuity-based model developed with WIPFLI Inclusive model

MCO Acuity-based model developed with WIPFLI l l Assigns an acuity score to each client, based on the LTCFS; and an additional behavioral score For each provider type, total funding allocated is divided by the total number of acuity points of the current caseload to determine a rate per acuity point For each individual, the individual’s number of acuity points is multiplied by the rate/point to determine the residential service rate for the individual Key characteristics of current model – – – 13 Incorporates client acuity Does not explicitly incorporate provider costs or provider characteristics (these were being considered as future refinements) Does not reflect policy objectives regarding best practice service or business models

Allowable Cost Reimbursement l Providers are reimbursed for costs reported, provided the costs are allowable services under federal Medicaid waivers l Key characteristics – – 14 Incorporates provider costs Does not explicitly incorporate client acuity Does not explicitly incorporate provider characteristics Does not reflect policy objectives regarding best practice service or business models

JVGA Model l l l 15 Financial (general ledger) data from providers is collected and compiled into four cost categories: Direct Support Professional (DSP) wages, employment-related expenses, program supports, and general and administrative costs Policy makers review and refine level of staffing based on service standards/best practice Wage adjusted through use of federal BLS wage data Results in a standard rate Can be related to acuity: a higher tier rate can be created for higher acuity individuals Impact analysis completed before implementation Mechanism established for handling exceptions (outliers)

JVGA Model l Key characteristics – – 16 Incorporates provider cost data Reflects policy objectives regarding best practice service and business models Can reflect client acuity to some degree, but does not appear to link client acuity with provider cost data at the individual level Requires extensive new data collection system and process for provider cost data

Inclusive Model l Seeks to incorporate the strong points from all of the above models l Key characteristics – – – 17 Incorporate client acuity Incorporate provider cost data and/or proxy data Incorporate provider facility characteristics Link provider cost/proxy data to client acuity data on an individualized basis and to provider characteristic data Reflect policy objectives regarding best practice service and business models

Inclusive Model l Development of Model – – – – 18 Regression analysis of selected encounter data as cost proxy on acuity and provider characteristics to identify likely acuity and provider characteristics that are cost drivers Clinical review of cost drivers Policy review of cost drivers Use of provider cost data to test validity of cost drivers Stakeholder comment on cost drivers/preliminary model Refinement of preliminary regression model based on clinical, policy, and provider cost validation reviews and stakeholder input Impact analysis by MCO, county, and providers Identify outliers and mechanism for addressing them

Inclusive Model (continued) l l 19 Develop implementation strategy, including transition mechanism(s) Determine timing and process for periodic updates

Questions/Comments and Next Steps l l 20 We welcome questions and comments DHS will continue to provide updates and opportunities for stakeholder comment as we proceed with the residential rate project