“Preferred” Prognostications

We operate two life underwriting study groups for the benefit of the industry. Taken together, they include over 40 chief underwriters plus nearly all reinsurers and major service firms.

Based on what was discussed at these open forums, its time to go out on the proverbial limb with some predictions and opinions about preferred underwriting practices.

Debits and Credits

Preferred criteria will no longer be quite as “cast in bronze” as they have been. There will be more widespread use of favorable findings to offset adverse risk factors. This is a highly desirable development because it will—that is, it should—result in more accurate determination of preferred status. One example is crediting normal NT-proBNP tests against suboptimal lipids, blood pressure, and cardiac family history. Fact is, screening older applicants with this remarkable test will likely lead to more preferred credits than adverse actions.

Drunken Driving

The most significant thing about DWI/DUI convicts is their high probability of having an undiagnosed alcohol use disorder. We know over 50% have alcohol use disorders. For this reason alone, insurers need to rethink how they address drunken (and drugged; but it’s mostly alcohol) driving risks on three counts:

■ Table ratings should replace flat extras

■ Preferred criteria should be tightened up

■ The CDT test should be obligate for any DWI/DUI conviction within 10 years

“Preferred” Diabetic

This confabulation is oxymoronic. There are no genuinely preferred diabetics. Indeed, a solid case could be framed for denying preferred to middle age pre-diabetics. Of course, any company can blatantly jack up the premium rate and call the concoction “preferred.”

Glycosylated Hemoglobin

Screening with HbA1-c appears poised to becoming darned near universal. Screening is replacing reflexive testing because elevated HbA1-c is a protean cardiovascular risk factor in non-diabetics. There should be a maximum HbA1-c beyond which preferred is denied.

Uric Acid

Years ago, uric acid was dropped from most insurance screening blood profiles. Burgeoning evidence from clinical and epidemiological studies makes an airtight case for reinstating this test. Now that carriers can instruct insurance labs to customize their profile content, the hope is that more of them will add this underappreciated CV risk marker and use it as a preferred criterion. If they want to balance the equation by jettisoning a current component, how about parting company with LDL-C? Total cholesterol is superior because LDL-C is a calculation, not a quantified measurement.

Gamma-glutamyltransferase(GGT)

The preponderance of evidence shows that GGT is even better as a CV risk marker than as an indicator of alcohol excess.

It has every bit as much legitimacy for preferred risk assessment as cholesterol or blood pressure. A solid argument can be fashioned for denying best preferred to applicants with otherwise unexplained GGT elevations…and also for setting a conservative maximum GGT elevation allowable for preferred on any basis.

Rest assured that we can blow away, in a manner of speaking, any attending physician who challenges us in this regard.

Family History

The only aspect of family history that should be retained for preferred is premature circulatory events in first-degree relatives. And these guidelines should reflect the key difference in age threshold between genders (e.g., age 60 for men vs. age 70 for women). Events matter; deaths are incidental.

Whether a proposed insured’s father who had an MI at age 55 succumbed to or survived is immaterial to us. What matters is that he had the event. Does a family history of cancer tell us anything?

It depends entirely on whether that history is considered on a site-specific basis. In a Swedish Cancer Registry study of over 1.2 million cancer patients, the risk of cancer at the same site as one’s first-degree relatives was increased 1.8-fold. But when all cancers were considered together, the excess family history risk evaporated.

If a female has a mother and a sister who developed premenopausal breast carcinoma, she is at increased risk for breast malignancy. On the other hand, if her mother had a lymphoma and her sister was diagnosed with pancreatic cancer, this does not influence her insurability.

Therefore, unless guidelines are applied on a site-specific basis, there appears to be little justification for using a general cancer family history criterion in preferred underwriting.

All of the current rhetoric about the “diabetes epidemic” is centered on its profound link to obesity. While the association between family history and the risk of type 2 diabetes is indisputable, we already address “build” in our preferred criteria.

Does diabetic family history have enough stand-alone risk impact, independent of obesity, or would we be better served to select a more durable preferred benchmark?

“Build”

Many companies have moved on from “build” to body mass index (BMI). This is timely considering BMI is how obesity and underweight are defined clinically. Waist circumference (WC) and waist-to-hip ratio are highly touted anthropometric markers for further assessing BMI. They single out individuals with abdominal adiposity, the aspect of obesity that has insurability implications. We have no illusions about getting both of these measurements done on paramedicals!

On the other hand, when I questioned representatives of four major paramedical providers at the study group sessions, no convincing arguments were made as to why they could not routinely measure WC.

WC would be a major upgrade for our “build” criteria, favoring health conscious applicants. If enough insurers speak up, this will surely come to pass.

Closing Thought

Some of you will take issue with and dispute one or more of the foregoing observations. Good. Because when disagreement begets dialogue/debate – not obstinate insistence on the illogical/ absurd—we are all well served.

By Hank George for November/December 2012 Issue of NAILBA’s Perspectives Magazine.

Advertisements

Comments are closed.

%d bloggers like this: