Population Cost Management:

Transcription

1 Population Cost Management: Using Data to Improve Care Katie Morrow, LBSW, MPA VP of Client Accounts, Streamline Healthcare Solutions

2 We all know the facts.. 20% of Medicaid beneficiaries who have a behavioral health diagnosis account for almost half of the total expenditure. 1 Spending for mental health was $221 billion in Behavioral Health spending is growing at a faster rate than health spending growth and health care inflation rate Herman Soper, M., Matulis, R., & Menschner, C. (2017, June). Moving Toward Value-Based Payment for Medicaid Behavioral Health Services. Retrieved May 1, 2018, from 2. Mandros, A. (2016, October 6). Behavioral Health Spending Reached $231.6 Billion in What does that mean? Retrieved May 16, 2018, from 3. Ross Johnson, S., & Meyer, H. (n.d.). Behavioral Health. Retrieved May 18, 2018, from

3 Population Cost in Value Based Care Less than 25% of providers are using health IT systems designed to handle their population health initiatives. 1 In a 2016 survey of State Medicaid Directors, 40% indicated plans to expand VBP arrangements. 2 Population Based Payments are becoming more prevalent. 2 Inconsistency in the market on what defines Behavioral Health outcomes. 1. Bruessner, V. (2017, January 19). Value-Based Care Drives Progress in Population Health Management. Retrieved May 1, 2018, from 2. Herman Soper, M., Matulis, R., & Menschner, C. (2017, June). Moving Toward Value-Based Payment for Medicaid Behavioral Health Services. Retrieved May 1, 2018, from

4 What if you could predict and plan for use of services across your population? Cost Anticipation Environmental Changes Put your data to work Think forwards, not backwards Planning for Budgetary constraints

5 Michigan Department of Health and Human Services (6/2017). Population Estimates by Local Health Department Michigan Retrieved May 29, 2018 from Jackman, M., & Woodhead, A. (2018). 18 historic (an not so historic) maps of Michigan. Retrieved May 29, 2018 from List of Counties in Michigan. (n.d.. In Wikipedia. Retrieved May 29, from

6 Michigan Environment circa 2009 Capitated funding model with regional entities (PIHP s) overseeing the allocation of Behavioral Health Dollars Shrinking budgets forcing reduction in services Potential political changes on the horizon Lack of trust between the region and County Community Mental Health Boards

7 Case Study MCO Organization in Michigan responsible for Medicaid funding for SUD and Behavioral Health Services for 5 counties. Autonomy vs. Cost Management Maintain autonomy of all 5 counties Manage expenditures up front Benefit Consistency Did not see consistency of benefits across the service area. Current Process Staff time was being wasted approving all authorizations in the system or they were not approving any auths prior to this project

8 Project Timeline Tool Selection Identify the level of care tools Data Analysis Population Creation Population Management Roll out to EHR 2009 Early 10 Mid Late Today Implement Tools Roll out into EHR Build LOC Model Determine Levels of Care and Service Model Outcomes Was it worth it?

9 Tool Selection DLA 20 MTM Services CAFAS FAS Outcomes RAP Davis Deshaies - Modified from the original version Sites for the tools: DLA CAFAS RAP -

10 Implementation of the Tools Integration into staff workflows: Assessment Care Plan Review Discharge Total scores were collected from the system: Initial Score Ongoing Scores Discharge Scores

11 Data Analysis to Create the Foundations of the Program: Historical service utilization used to determine cost of service Analysis of the standard tool scoring to determine Levels of Care by population How many clients were falling into each of the Levels of Care? What was the average cost per client within the Levels of Care? How did counties vary across the region?

12 Building the Level of Care Model Determine units of service by Levels of Care Determining Appropriate Services within Levels of Care Establishing Level of Care Ranges

13

14 Caps are a guide for average levels of care If more is requested, it needs manual review If the request is under the Cap, it is automatically approved The system cannot automatically deny any services

15 Building the Model Into the EHR Assessment Drives LOC Validations on Care Plan to require Assessment Care Plan Creates Auths Under both CAPs = Automatic approval Levels of Review are applied LCM CCM

16

17 Building the Model Into SmartCare Other factors in building out the logic: Flexibility for caps to be by coverage plan Flexibility to consolidate the authorizations into a UM Category Flexibility to update and modify the CAPs as funding changes Flexibility for local counties to decide on if their CAPs for auto approval should be modified for a particular service Build it to work for the counties, despite who their MCO is Flexibility in the tools that would be used to determine Level of Care Ability to guide clinicians in decision making when doing a Care Plan

18

19 Roll Out into the EHR System roll outs started early 2012 Initial Reactions by organizations Skepticism Direction to stay below the review caps New UM Responsibilities for some counties seen as additional work

20 Subjective Outcomes Consistency across the region using a data driven process Clinical Awareness of what was within norms for services Increased consistency of not requesting services outside of the norms Decreased time for access to care Accrediting Bodies satisfaction with the tools and process

21 Sometimes the questions are complicated and the answers are simple. - Dr. Seuss

22 Measuring Success Did the volume of services change? Did the automation improve staff time? Did it impact overall costs? Did the process inadvertently impact the Level of Care tools?

23 NO. OF AUTHS NO. OF AUTHS NO. OF AUTHS Number of Unique Authorizations Organization A MONTHS Organization B Organization C MONTH MONTH

24 Auto-approval of Authorizations Organization A Organization B Organization C Combined 23% 19% c 33% 25% Average of 110 authorizations per month for an average 153 clients per month Average of 103 authorizations per month for an average 162 clients per month Average of 179 authorizations per month for an average 152 clients per month. Average of 3932 authorizations for an average of 467 clients per month.

25 Was the cost of service impacted? Average Cost per Authorization Organization C Post Organizatino C Pre Organization B Post Organization B Pre Organizatin A Post Organization A Pre Average Post Average Pre $0.00 $50.00 $ $ $ $ $ $ $ $450.00

26 Was the level of care impacted? DLA 49% 10% 41% Increase in Functioning Decrease in Functioning Exact Same Score CAFAS RAP 17% 44% Decrease in Functioning Increaase in Functioning 22% 48% Decrease in Functioning Increaase in Functioning 39% Exact Same Score 30% Exact Same Score

27 What Happened Next? Some counties have expanded to use it for other funders besides their capitated dollars The State of Michigan did change the evaluation tools DLA LOCUS RAP SIS Currently revamping based on these changes to create new levels of care for these tools plus the ASAM Flexibility of the E.H.R has made this adaptable

28 Rome wasn t built in a day. - John Heywood Blue Print for Success 1. Look backward so you can look forward 2. Analyze your data to determine your populations, level of care, and utilization norms (caps) 3. Build the process into existing workflows using your EHR 4. Roll it out; continue to monitor and adapt

29 Keys to Success Vision and organizational buy in Staff capable of Data Analytics An EHR that can provide the data collection, flexibility for adaptability, and integrate to your business process Staff to drive the change and make the tough decisions

30 Website Follow us on Social Media: