Observations on Mortality Trends
June  2016

At a recent industry meeting, SCOR presented findings from two mortality experience studies. One study considered trends in causes of death (COD) and the other compared mortality experience between fully underwritten and "traditional" simplified issue business. This article recaps the findings.

Cause of Death Reliability

No conversation on COD is complete without a disclaimer - COD must be taken with a grain of salt (see "The Death Certificate Challenge" sidebar). The complexity of humans makes it difficult to capture the true cause of death. When should we capture the most direct problem and when is the underlying cause more meaningful? A person may have a pulmonary embolism listed as the immediate COD on their death certificate, but perhaps the fact that the individual was admitted to the hospital with advanced lung cancer is more valuable in understanding our risks.

COD reliability varies with age and cause. Younger people and non-natural causes of death may receive further investigation or be more likely to have an autopsy – both good ways to capture an accurate COD.

Cause of Death Trends

The study covers claims between 2006 and 2012 with actual-to-expected ratios (A/E) based on the 2015 VBT. We excluded claims occurring within the first two years of issue to avoid skew from the contestability period.

The results over the observed period indicate a decrease in all-cause mortality of 1.3% annually for both males and females (Figures 1 and 2). For males, circulatory and external CODs decreased, while cancer and respiratory impairment-related deaths remained more or less flat.

MortalityTrendsMalesFemales.jpg 

CODs not related to natural causes among males varied over the timeframe. Motor vehicle accidents dropped significantly over the observed period. However, death from drug overdose or poisoning increased slightly. Suicide deaths increased from 2007 to 2009 but have since returned to their lower 2006 levels (Figure 3).

For females, non-natural CODs decreased most markedly in motor vehicle accidents, though this improvement deteriorated in the most recent years (Figure 4). Similarly, deaths caused by drugs or poisoning and "other" increased towards the end of the observed period.

NonNaturalCOD.jpg 

Fully Underwritten vs. "Traditional" Simplified Issue

As carriers consider options in changing underwriting requirements, it is worth reviewing how mortality experience varies under fully underwritten and "traditional" simplified issue (questionnaire only, no e-data) business.

We considered claims experience on males who underwent full underwriting under four-class non-tobacco products (Figure 5). As expected, mortality worsens in line with the insured’s risk class. In this example, the A/E for the worst class ranges from about 135%-145% of 2015 VBT.

To compare fully underwritten to simplified issue, we isolated males age 18-49 from both blocks. (Figure 6). All groups of business benefit from the contestable period. However, beginning in duration 3, the A/E for simplified issue policies increases significantly compared to fully underwritten groups.

FullyUnderwrittenSimplifiedIssue.jpg 

Post-Contestable Causes of Death

Immediately following the contestable period, external factors (accidents, suicide, etc.) drive up mortality, especially in simplified issue business (Figure 7).

In later durations (6-10), simplified underwriting wears off more quickly and individual medical CODs begin to overtake external causes. COD patterns in fully underwritten business remain relatively stable. Full underwriting provided much more protective value across all non-external causes (Figure 8).

AEbyCOD.jpg 

For more information on mortality reviews or other mortality-related R&D initiatives, please contact me at kwarner@scor.com.