Aetna 2006 Annual Report Download - page 25

Page 23
Each quarter, we re-examine previously established health care costs payable estimates based on actual claim
payments for prior periods and other changes in facts and circumstances. Given the extensive degree of judgment
in this estimate, it is possible that our estimates of health care costs payable could develop either favorably (i.e., our
actual health care costs for the period were less than we estimated) or unfavorably. The changes in our estimate of
health care costs payable may relate to a prior fiscal quarter, prior fiscal year or earlier periods. We also consider
the results of these re-examinations when we determine our current year liabilities. Because of the uncertainty
involved in establishing estimates of health care costs payable each period, changes in prior period health care cost
estimates may be offset by current period health care costs when we establish our estimate of current period health
care costs. Our reserving practice is to consistently recognize the actuarial best estimate of our ultimate liability for
health care costs payable. When significant decreases (increases) in prior periods’ health care cost estimates occur
that we believe significantly impact our current period results of operations, we disclose that amount as favorable
(unfavorable) development of prior period health care cost estimates. In 2006 and 2005, we recorded favorable
development of prior period health care cost estimates of approximately $18 million pretax and $250 million pretax
(reflects the release of approximately $103 million of reserves related to the New York Market Stabilization Pool),
respectively (refer to Health Care beginning on page 5 and Note 6 of Notes to Consolidated Financial Statements
on page 58 for additional information).
During 2006, our claim submission and processing times were consistent with 2005. During 2005, claim
submission and processing times decreased due to increased electronic submissions of claims, accelerated
processing of claims (as a result of increased automatic processing of claims and other factors) and other
efficiencies in our processes as well as those of the medical service providers submitting claims. As a result, in
2005, older dates of service (i.e., those claims with dates of service three months or more before the financial
statement date) were more complete than we originally estimated, which contributed to the 2005 favorable
development of prior period health care cost estimates we reported in results of operations (refer to our discussion
of Health Care results in this MD&A beginning on page 5 for more information on development). Specifically,
after considering the claims paid in 2006 and 2005 with dates of service prior to the fourth quarter of the previous
year, we observed the assumed weighted average completion factors were approximately .2% lower and .6% higher,
respectively, than previously estimated. When establishing our reserves at December 31, 2006, the weighted
average assumed completion factors were reasonably consistent with those previously used in 2005. Based on our
historical claim experience, it is reasonably possible that our assumed completion factors may vary by plus or minus
.75% from actual.
Also during 2006 and 2005, we observed that our health care cost trend rates for claims with dates of service three
months or less before the financial statement date were slightly lower than previously estimated which contributed
to the favorable development of prior period health care cost estimates we reported in results of operations.
Specifically, after considering the claims paid in 2006 and 2005 with dates of service for the fourth quarter of the
previous year, we observed health care cost trend rates that were approximately .5% and 2.6%, respectively, lower
than previously estimated for claims associated with combined Commercial Risk and Medicare IBNR. The lower
than anticipated health care cost trend rates we observed in 2006 for claims incurred in 2005 were due to
moderating inpatient, outpatient and primary care physician service trends. The lower than anticipated health care
cost trend rates we observed in 2005 for claims incurred in 2004 were due to moderating physician, outpatient and
inpatient service trends. Historical health care cost trend rates are not necessarily representative of current trends.
Therefore, we consider historical trend rates together with our knowledge of recent events that may impact current
trends when developing our estimates of current trend rates. When establishing our reserves at December 31, 2006,
we decreased our assumed health care cost trend rates to account for the lower than anticipated health care cost
trend rates recently observed. Based on our historical claim experience, it is reasonably possible that our estimated
health care cost trend rates may vary by plus or minus 4.0 percentage points from actual.
The following table illustrates the sensitivity of our results of operations to changes in our assumption for our
estimate of our health care costs payable at December 31, 2006 (in millions) that would be caused by certain
reasonably possible changes to our estimated weighted average completion factors and health care cost trend rates
and the impact that such changes could have on current period operating results. However, it is possible that the
actual completion factors and health care cost trend rates will develop differently from our historical patterns and
therefore could be outside of the ranges illustrated in the following table.