Insurers may be closer to getting the data they need to improve the design and pricing of long-term care insurance (LTCI) products.

A team at the Society of Actuaries (SOA) has released results of an effort to create tables summarizing the experience insurers have had with claim incidence, claim termination and claim utilization for U.S. LTCI policies from 2000 through 2011.

The SOA hired Towers Watson to create the report. Vincent Bodnar and Matt Morton, the actuaries responsible for the results, and other project team members, collected data from 22 LTCI insurers on a total of 14.8 million people who had LTCI coverage.

The insurers provided data on 14.5 million active lives, or people who had coverage without being on claim, and on 172,281 people who collected LTCI benefits.

About 1.2 percent of the people in the database were people who had filed LTCI claims.

The SOA package includes spreadsheets showing how much different factors, such as age, sex, policy type, policyholder underwriting class, and various health conditions, were to affect the likelihood that insureds would need assisted living facility care, home health care or nursing home care.

The actuaries looked, for example, at the widely accepted idea that “LTCI costs are much lower for married people.”

Many people involved with LTCI assume that LTCI costs are lower, on average, for couples, because married people tend to have better access than single people to family care,

The actuaries found that, in their pool of experience data, the difference was much lower at younger ages than at older ages. Single policyholders were about twice as likely as married policyholders to file claims at age 65, but the gap disappeared around age 87. Once policyholders reached their late 80s, married policyholders were about as likely to file claims as single policyholders.

LTCI issuers have reported having a great deal of trouble with low interest rates in recent years, and also a great deal of trouble with accurately predicting how likely policyholders would be to keep or drop policies. They have also reported having some trouble with predicting how policyholders would use their policies, in part because widespread use of the policies is relatively new, and the holders of the many policies sold in the 1990s are just getting to be old enough to file claims in large numbers.