Ch 4 Preferred Risk Underwriting

Question 1

preferred risk

Answer 1

individuals demonstrating lower mortality risk features

Question 2

CAD risk factors (8)

Answer 2

1. age
2. sex
3. blood pressure
4. cholesterol
5. HDL
6. diabetes
7. LVH, left ventricular hypertrophy
8. smoking status

Question 3

Framingham study - 1948

Answer 3

study to help understand modifiable risk factors of coronary artery disease and help reduce risk of dying
used exam, med hx, physical measurements, h/w, bp, resting EKG, chest x-ray, blood chemistries

Question 4

year preferred products enter industry

Answer 4

late 1980s, early 1990s

Question 5

tobacco distinct pricing in early 1980s

Answer 5

divides standard into 2 classes: smokers and non-smokers. preferred risk divides standard risk class

Question 6

residual standard

Answer 6

remaining standard risk that do not qualify for any of the preferred classes. mortality of residual higher than original standard class

Question 7

Conservation of Deaths

Answer 7

there is a range of mortality contained inside the standard class, defined as 100%, if subset of those risks w/ lower mortality expectation are identified and offered a lower-priced rate class, then those remaining in standard class will exhibit higher mortality.
-fewer that qualify for preferred, greater the mortality discount
-after discount is determined for preferred, simple algebra calculates mortality for remaining insureds

Question 8

normal distribution

Answer 8

models assume mortality in standard class is bell shaped, most insureds exhibit mortality close to average, few are very low or very high w/in class
-mortality does not follow normal distribution but works as approximation

Question 9

to create preferred

Answer 9

standard subdivided into unique subsets reflecting individualized mortality assumptions associated w/ UW rules for each unique subset.
-if subdivided into 3 risk classes, mortality by class calculated as weighted average of all insureds w/in each of 3 risk classes
-pricing assumption is average mortality associated w/ specific risk class

Question 10

reasons for overlap between risk classes

Answer 10

1. unmeasured risk factors above and beyond preferred rules that impact mortality
2. companies employ knock-out system that does not definitively stratify risk

Question 11

Multivariate Cox Proportional Hazard Model

Answer 11

• risk is multifactorial, multiple variables (risk factors)
• goal to parse out influence of each risk factor according to its independent contribution to risk
• AX + BY + C*Z = D
• allows researchers to investigate mortality for medical condition or procedure
• A/B/C variables solved for
• X/Y/Z variables have results for (known values)

Question 12

predictor variables (independent variables)

Answer 12

use to predict who will develop heart disease
-build/bp/cholesterol etc

Question 13

target variable (dependent variable)

Answer 13

result depends on predictor variable
ex. heart disease or mortality

Question 14

hazard ratios

Answer 14

-baseline is 1.0 similar to 100% mortality or 0 debits
-1 unit increase = 1.25 HR or 125%, mortality risk increases by 25%
-ratio is multiplied instead of added like debits

Question 15

point system

Answer 15

suggest individual’s risk can be defined by sum of debits and credits associated w/ risk factors
-positive numbers represent increased risk, negative numbers represent decreased risk
-lower points is more favorable
-takes favorable/unfavorable into account w/o allowing 1 factor to overwhelm decision by itself

Question 16

knock-out system

Answer 16

criteria established as series of rules associated w/ each factor, PI either qualifies for risk class or is knocked out based on rule
-favorable factors cannot offset 1 unfavorable, unlike point system
-some create stretch criteria, secondary set of guidelines, to allow some insureds to move to better risk based on secondary considerations

Question 17

stricter criteria equals?

Answer 17

lower mortality assumptions

Question 18

predictors of mortality/baseline mortality features

Answer 18

age, gender, smoking status

Question 19

Mortality Markers

Answer 19

1. age, gender, smoking status
2. bp - below 140/90, lowest risk 115/75, risk doubles for every 20/10 increase
3. cholesterol - less than 200
4. HDL - chol/hdl ratio, lower ratio, lower risk
5. diabetes & LVH on resting EKG
6. build - overwhelming impact on mortality. as weight increases, mortality increases along J curve

Question 20

Addl Mortality Markets

Answer 20

1. Accidental Death Risk
   - Driving: DUI/speeding
   - Hazardous avocations: flying, scuba diving, racing, mountain climbing, sky diving, parasailing, extreme sports
   - Occupations
   - Criminal Hx
2. Drugs/Alcohol: exclude hx of abuse w/in 5-10 yrs
3. Personal Med hx: could cause extra mortality to warrant exclusion
4. Family Med hx: CAD/diabetes/stroke/kidney disease/cancer
5. Treatment: ex. tx of HBP

Question 21

indirect correlation

Answer 21

lower education levels associated w/ higher smoking. correlates education w/ smoking. smokers have higher mortality. correlation of education and smoking is indirect

Question 22

for data to be actionable?

Answer 22

needs to adhere to Fair Credit (FCRA). data must be disclosable, disputable and correctable

Question 23

continuum of risk

Answer 23

core component of UW
continuous sequence in which adjacent elements are not perceptibly different from each other, although extremes are quite distinct
ex. risk associated w/ bp: heart disease can be present w/ low bp levels
-preferred pricing predicated on fact no difference in mortality, a continuum of risk, found inside standard class

Question 24

predictive analytics

Answer 24

practice of extracting info from existing data sets in order to determine patterns and predict future outcomes and trends

Question 25

advantages to preferred

Answer 25

-reward healthy individuals w/ lower premiums
-compete effectively by applying lower-priced products to healthier risks
-driven by competition