MLTC Risk Adjustment Workgroup Meets

Last week, the Department of Health (DOH) and the State's actuary, Deloitte, convened the first of three Managed Long Term Care (MLTC) Risk Adjustment Workgroup meetings. The State’s goal is to gather input to inform their work in making appropriate updates to the risk model that assigns scores to various assessment items and that is then used to calculate individual plan risk scores.

Deloitte and DOH are taking a similar approach to developing and applying the risk model as has been used previously. As in previous years, a single, statewide risk model will be applied separately to each plan type in a regionally-specific way. The first step is to update the risk model by reviewing the set of predictors used previously while seeking recommendations on other variables and interaction of variables, as well as potential scoring changes, that should be taken into account. DOH requests that any additional suggestions be submitted by email to MLTCRS@health.ny.gov and is especially interested in receiving input from plan clinical staff. Meeting participants submitted a series of recommendations.

Highlights of other discussion points and participant input is provided below. 

• Deloitte/DOH intend to use assessment, cost report, and encounter data for CY 2016 for regression modeling in updating the risk model. (2017 encounter data is not usable at this time.)
• HCHN survey data will be used to help inform shadow pricing and be incorporated into the risk model itself, if warranted.
• For April 2019 rates, DOH intends to use a January 2019 enrollment snapshot along with UAS assessments through June 2018.
• Deloitte is reviewing mood and behavior variables as well as assessment data of HCHN members and is considering mobility issues and polypharmacy as potential predictors to incorporate into the risk model.
• Participants recommended that the model development analyses rely on utilization/intensity rather than cost so as to avoid skewing the model based on variation in negotiated rates.
• Participants identified differences in service utilization/intensity patterns in different regions as a potential data concern that DOH/Deloitte should consider when developing a statewide model.
• Participants expressed concern that high cost members who are enrolled for a short time may not be reflected in the data used to develop the model nor in calculating plan risk scores. 
• Participants deliberated the idea of implementing a reconciliation mechanism but agreed that predictability may be more important than retrospective accuracy.
• Participants reminded DOH/Deloitte that UAS data used for analysis should be from 2015 and on, given that initial UAS data was not fully reliable.

LeadingAge Contact: Darius Kirstein, dkirstein@leadingageny.org, 518-867-8841