Given the inherent difficulties encountered in assessing national patterns of outpatient practice, it is not surprising that Drs Kakaiya and Warnhoff identify several limitations of our analysis, as well as areas of possible misunderstanding. While reinforcing the need for caution in interpreting our findings, their comments do not invalidate our conclusion that the use of β-blockers in coronary artery disease appears to be suboptimal.
Kakaiya and Warnhoff may have misunderstood several aspects of our methods. For instance, our analyses did not rely on patient recollection because survey forms were filled out by physicians, not patients. Furthermore, the coded medication lists include all medications regardless of how they were obtained (office sample or pharmacy bought). Second, our sample sizes were large enough to make annual estimates of β-blocker usage. The year with the fewest visits was 1996, where 710 visits were available for analysis. Annual rates of β-blocker use did rise, from 17% in 1993 to 25% in 1996, a point we highlight in the discussion and Figure 1 of our article.1 Even so, we found it surprising that in 1996 three fourths of patients with a diagnosis of coronary artery disease were not on a regimen that included β-blockers.