Uncover Hidden Population using Predictive Modeling Tool

1
Uncover Hidden Opportunity to Prevent Unnecessary
Admissions
2

• Predictive Modeling Tool for Sustainable Care
  Management
• Using Predictive Modeling Tool to Identify at
  Risk Patients who has a chance of becoming users
  of High-Cost Healthcare service and subsequently
  Reducing PMPM (Per Member Per Month) Costs While
  Increasing Member Satisfaction.

3
Challenge

• Over the years, Health Maintenance Organizations
  (HMOs) have been using traditional care
  management programs to reduce healthcare cost and
  improve quality of care for their plan members.
• GlobalHealth, an Oklahoma based HMO has been
  doing well at managing care after a disease or
  condition is identified.
• But they didnt have necessary data to identify
  clinically high risk members who may not be high
  utilizers of Care Management programs.
• This resulted increase in ED encounters and
  readmissions of members thus increase PMPM.

4
Scenario at a Glance

• GlobalHealth identified group of members who are
  low utilizer of services.
• Some of these members experience a sudden acute
  event that requires a hospitalization and,
  consequently, become high users.
• GlobalHealths clinical and administrative
  leaders found clear indicators of these acute
  events after reviewing statistics on daily
  inpatient admissions (e.g. reasons for
  admissions, diagnosis, health and claims history)

5

• GlobalHealth realized the need of Predictive
  Modeling Tool.
• Predictive modeling tool could prevent large
  percentage hospitalizations by evaluating the
  member data and identifying at-risk patients.
• Overall the challenge was to prevent acute
  incidents by identifying at-risk patients.
• Indentifying at-risk patients was important from
  a cost as well as quality of care and service
  standpoint.

6
Solution

• GlobalHealth used VitreosHealths predictive
  modeling tool to retrospective review of health
  plan member data.
• Care Managers found approximately 4,000 health
  plan members who were low-cost healthcare user at
  the end of 2011, had a high risk of facing an
  acute event.
• These members became high utilizers of services
  between 2012 and 2013 compared to 2011.
• Healthcare costs associated with them tripled in
  2012 and again in 2013.

7

• Surprisingly every year about 12 to 15 percent
  members from hidden category moved to the
  high-utilizer of the service category.
• Predictive risk modeling tool can help the HMOs
  to identify hidden populationwho are not high
  users of healthcare services but have high risk
  for an acute event in future.
• At present this hidden cohort members do not
  take routine care (in the form of primary care
  visits, health screenings, and diagnostic tests).
• Routine care could prevent more acute and
  expensive emergency admissions.

8

• VitreosHealths predictive modeling tool creates
  a profile for each plan members after analyzing
  clinical and nonclinical data.
• Predictive modeling tool use various data like
  HMOs Electronic Health Record (EHR), Utilization
  data from claims, Medication data from pharmacy
  systems, Scheduling data from practice management
  systems and Demographic data to create members
  profile
• Then the tool calculates a members risk (State
  of Health score) for clinical chronic conditions,
  including congestive heart disease, diabetes,
  asthma, and hypertension.

9

• Importantly, it also assesses five nonclinical
  factors that can affect a members risk of
  experiencing an acute event
• Utilization Utilization score is derived from
  the claims data and calculated based on members
  resource utilization patterns. E.g. Number of
  hospitalizations, ER visits and medication.
• Compliance Compliance score is calculated by
  measuring members adherence to evidence-based
  care protocols such as appointments, lab test and
  medications.
• Access to care Ease of access to appropriate
  care services have an effect on availing care
  management services.

10

• Socioeconomic Demographic data (e.g. education
  levels, household incomes, family size, and
  native languages) may also affect health status.
• Perceived well-being This measures how patient
  feels about his/ her health condition. Studies
  have shown that patients perceptions about their
  health may affect their actual health condition.

11

• Vitreos uses a patent pending transformational
  approach for predicting risk of hospitalization /
  ER visits / High Cost Intervention that takes
  into account 6 dimensions.

12

• Based on clinical and non-clinical factors,
  plan-members are categorized into four cohorts
• High risk, high cost
• Low risk, high cost
• Low risk, low cost
• High risk, low cost
• Opportunity for cost savings lies in the last
  cohort Hidden. Healthcare leaders need to
  identify hidden members and put them into
  effective care management to maximize PMPM
  savings.

13
Outcomes

• GlobalHealth care managers receive a daily,
  weekly and monthly report from the Vitreos
  predictive modeling tool.
• The report indicates of high-risk members,
  including the hidden cohort.
• It also identifies problem areas based on the
  members clinical and nonclinical scores,
  prioritizes the members who should receive
  outreach first, and recommends care management
  actions.

14

• Sample report are sent to the care managers
  daily, weekly or monthly

15
Key observations

• The preliminary result shows GlobalHealth has
  been successful in reducing ED encounters and
  readmissions by 20 in among all members between
  January 2014 and May 2014.
• Hospital admissions have declined by 5
• 20 reduction of ED visits resulted cost savings
  of 2 to 3 in PMPM ( one ED visit 10 of PMPM
  costs )
• Saving of 3 to 4 PMPM due to 5 reduction in
  inpatient admission ( one inpatient admission
  represent about one-third of PMPM costs )
• Member satisfaction has increased by 3.5 among
  all health plan members

16

• Based on the preliminary results of this program
  the improvement in both member satisfaction and
  cost savings exceeded the cost of adding staff.
• As a result, GlobalHealth is planning to double
  its care management staff to accommodate the
  additional member outreach in one year.
• The results of the program will become more
  apparent in August or September 2014, but the
  overall trend is promising. Right now it is
  indicating very positively that were putting our
  resources in the right spot, says J. David
  Thompson, GlobalHealths vice president of health
  plan operations.