changed outcome is better eyesight, not extended mortality. But how does one place an economic value on additional years of better eyesight?

In other cases, medical advances have no significant effect on medical outcomes for the medical problem being treated but are superior from the patient’s viewpoint because they involve fewer side effects. The introduction of non-sedating antihistamines has been significant, for example, not because the new generation of antihistamines is demonstrably more efficacious in reducing symptoms from allergies but because patients receiving the new antihistamines no longer have to face side effects from sedation—side effects that are important in an individual’s ability to function and perform normal activities of daily living. In these cases, narrowly defined medical outcomes might not be changed at all, but the quality of life for patients who receive the new treatment is much improved. How is BLS to evaluate such quality-of-life improvements from new medications?

Health services researchers are currently attempting to measure the value of quality-of-life advances. They have constructed metrics such as QALYs (quality-adjusted life years) and DALYs (disability-adjusted life years). Although the conceptual foundations underlying such measures have been criticized, they are increasingly used within the health policy community. While the panel does not believe that BLS should use outcomes-adjusted price indexes based on QALY or DALY measures in its official price indexes in the near future, high priority should be given to research examining the feasibility and sensitivity of alternative ways of incorporating quality-of-life aspects of medical treatments, both at the BLS and elsewhere.


The CPI is designed to price goods and services for which consumers make direct out-of pocket-expenditures. Because this definition does not include the employee health insurance premiums paid by employers or government payments for medical services (Medicare Part A and Medicaid), the weight given medical care items in the CPI (slightly less than 6%) is much smaller than the share of medical care expenditures in total personal consumption expenditures (about 17.6%). Under certain circumstances, these weighting procedures can result in biases. One possible bias will occur if the health insurance payment part of the MCPI moves differently from other components of the index (e.g., if prices of insurance-covered services, such as treatments for cancer, moved differently from uncovered services, such as cosmetic surgeries or nonreimbursed purchases of nonprescription drugs). The likely direction of such a bias is a priori unclear.

Much more importantly, the panel concluded, is that including only patients’ out-of-pocket outlays inappropriately understates the medical care portion in the total CPI. Arguably, this biases the CPI as a whole. In the last 75 years, the MCPI

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