changing say every 3-5 years, actuaries will likely find it very difficult to generate hypothetical quotes for old policies with obsolete fixed populations, given that insurers’ policies and enrollees have changed. Even more problematic is the issue of how actuarial quotes would take into account the premium effects of variations in the health status of those selecting into (and leaving) a particular policy’s enrollee population. Such data are typically not observable.

The panel appreciates the difficulties faced by the BLS in obtaining third-party price quotes of insurers’ reimbursements for physician and hospital services and prescription drugs and in creating diagnostic-specific episode treatment price indexes. Replacing indirect pricing with the direct pricing of health insurance policies offers very attractive possibilities, and we believe that this direct pricing alternative merits close scrutiny. But we also strongly recommend that no change from indirect to direct pricing be made without extensive experimentation and reliability assessment that includes consultation with leading health economists, actuaries, clinicians, and health insurance specialists.

RECOMMENDATIONS

Pricing Diagnostic Treatments

The changes that the BLS implemented in 1998 regarding the MCPI reduce the bias caused by the aggregation of goods and services to levels that do not allow for substitution among alternative medical treatments. BLS should continue its efforts to eliminate this bias. The major change BLS implemented in 1998 concerned aggregating inpatient and outpatient hospital services into a single stratum, to allow for substitution between them, but that is only one area of potential bias from input substitution.

The panel favors the use of diagnosis-based rather than input-based measures wherever this is feasible. The advantages of diagnosis-based methods have been highlighted in a number of research papers, of which the treatment for depression is only one example. Thus the panel recommends using a disease- or diagnosis-based unit for pricing rather than the current “industry” or medical care strata. This recommendation does not imply (or require) outcome-based measures. It implies only that inputs be priced and aggregated by the changing treatment mix for a particular diagnosis, rather than by the traditional BLS medical care strata.

Recommendation 6-1: BLS should select about 15 to 40 diagnoses from the ICD (International Classification of Diseases), chosen randomly in proportion to their direct medical treatment expenditures and use information from retrospective claims databases to identify and quantify the inputs used in their treatment and to estimate their cost. On a monthly basis, the BLS could reprice the current set of



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement