Sn 209-02 Economic Incentives (Antiselection) In Health Ins.

Front 1

Reasons Why HI Is Susceptible to Antiselection

Back 1

Note: A similar list is found in IHI Ch. 6. See the list “HI / DI Underwriting is Harder Than LI Underwriting, because...” at the top of that document.
 Insureds know their own health
 Elective nature of health care
 High cost
 antiselective lapses
 choice is socially desirable / demanded
 Ees often have opt-out option & spousal option

 Regulations prevent premiums from matching risk:
 Mandated benefits
 guaranteed issue and renewability
 community rating
 HIPAA Regulations
 pregnancy coverage
 Guaranteed Issue to small groups
 limits on Rating Factor usage
 can’t uprate or reject individuals in a group
 portability.

Front 2

ANTISELECTION SOLUTIONS for HEALTH INSURERS AND MCO’S

Back 2

 Quote the entire “Antiselection Controls” list from McKay Ch. 2.

In addition, this SN suggests the following:

 Use models to measure and predict antiselection:
 Choice model, Memb Utilz Cost model, CAST model
 study durational claims pattern
 vary premium rates as much as possible (when legal)
 have a Government Affairs Staff
 beware of self-insured groups suddenly buying insurance
 rerate a group if it acquires another group

 Base premium rates on ultimate claims
 Build up reserves

 use Field Underwriting
 Commissions are key
 proper training of Underwriters and Agents.

 redesign and improve policy provisions
 for competitivity; to prevent antiselective lapses
 Use balanced provider panels
 Manage pharmacy benefits

Front 3

Antiselection Controls for an HMO in a Dual-Choice Environment:
(That is, where the Employer is offering an Indemnity Plan as well as the HMO plan)

Back 3

HMO should have similar:
 rate tiers
 benefit plan
 premium rates s/change at same time as indemnity’s
 Be aware if Er changes his contributions

Front 4

The Choice Model of measuring and predicting adverse selection costs

Back 4

1. Build a model office based on claims distribution
2. Identify the plan options
3. Temporarily assume each ee knows his optimal plan
4. Compute (cost of his optimal plan) minus (cost of plan-he-would-have-had-without-choice). The difference is called the Maximum Adverse Selection Cost.

5. Reduce the maximum adverse selection cost to reflect:
 imperfect knowledge
 risk avoidance
 family members having different needs

 use computer simulations

Front 5

Plan Characteristics that Affect the Adverse Selection Cost
Antiselection is worse if:

Back 5

 wider deductible range
 (healthy people choose high deductible plan and sick people choose low deductible plan)
 Fixed-dollar ee contribution
 Less Er contribution (fixed $)
 Opt-out ability
 fewer family members

Front 6

The Member Utilization Cost Model of measuring and predicting adverse selection costs

Back 6

1. Split costs by type of service
2. To each type of service, apply a Utilization Adjustment Factor (antiselection factor)
 primary services elective  adjustment factor high
 hospital inpatient nonelective  adjustment factor low.

3. Lower cost-sharing option  adjustment factor higher
4. The Utilz Adjustment is added to the utilization of the “base plan”.

See example in chapter notes.

Front 7

Other models

Back 7

Durational Model (moved to SN 206)

CAST Model (see TSA 34)
Durational Model (moved to SN 206)

CAST Model (see TSA 34)

Done