When evaluating targeted therapies in chronic lymphocytic leukemia, the absolute measure of treatment effect should be preferred

In randomized clinical trials (RCT), treatment effect is often measured using the hazard ratio (HR), with HR < 0.8 being of clinical interest. We hypothesized that absolute measures of effect are more meaningful for both practitioners and patients. We selected all articles in the English literature reporting results of RCTs evaluating targeted therapies in chronic lymphocytic leukemia. We digitized the survival curves and computed HR, differences in survival rates and in mean survival times, and net survival benefit by 6 and 12 months. A total of 20 articles were analyzed. All measures were roughly consistent, except the mean survival benefit. For HR < 0.8, 78% of comparisons showed a mean benefit of at least 2.5 months while 53% showed a mean benefit of 6 months. These figures dropped to 43% and 7%, respectively when the benefit was measured by OS. We suggest that absolute measures of treatment effect should be reported more systematically.