Comorbidity-adjusted life expectancy optimizes screening recommendations

Calculating comorbidity-adjusted life expectancy could help physicians determine whether to continue or cease cancer screening in elderly patients, according to a study published Nov. 18 in Annals of Internal Medicine.

As age increases, the benefits of early cancer detection decline because older patients are more likely to die of comorbid conditions or other causes. However, many guidelines differ when recommending elderly patients to stop undergoing cancer screening. For many, no survival benefit would result from the screening. Potential complications from screening or follow-up tests must be considered, as well as the discovery of an asymptomatic disease in the latter portion of a patient’s life.

“For all of these reasons, it is crucial to weigh potential benefits and harms of cancer screening in elderly persons in relation to life expectancy,” wrote lead author Hyunsoon Cho, PhD, of the National Cancer Institute, and colleagues.

In order to estimate the life expectancy for elderly persons without a history of cancer and to take into account comorbid conditions, Cho and colleagues designed a population-based cohort study. A five percent sample of Medicare beneficiaries who were at least 66 years old without a history of cancer from selected geographic areas participated between 1992 and 2005.

To identify comorbid conditions in the Charlson index, the researchers utilized the patricipants’ medicare claims. Survival probabilities were estimated by comorbidity for the three most prevalent conditions: diabetes, chronic obstructive pulmonary disease, and congestive heart failure by using the Cox proportional hazards model.

Comparisons of survival models with U.S. life tables helped calculate comorbidity-adjusted life expectancy. Survival probabilities from the U.S. life tables offered the most similar survival experience to the cohort of interest.

The study’s results revealed that those with higher levels of comorbidity had shorter life expectancies. Those without comorbid conditions, however, including the very elderly, had favorable life expectancies comparable to an average person of the same chronological age.

Estimated life expectancy at 75 years old was about three years longer for participants with no comorbid conditions and approximately three years shorter for those with high comorbidity relative to the average U.S. population.

“The comorbidity-adjusted life expectancy developed here may facilitate clinical decision-making and recommendations tailored to individual patients,” wrote Cho and colleagues.