A routine audit revealed that the analytical method used to measure digoxin concentrations by our statewide pathology provider in 2009 was underestimating digoxin concentrations by 10%. The assay was recalibrated by the manufacturer in 2010, but clinical outcomes of the underestimation were never measured. This is a pilot study to describe the prescribing behavior around out-of-range digoxin concentrations and to assess whether miscalibrated digoxin immunoassays contribute to clinically relevant effects, as measured by inappropriate alterations in digoxin doses.
About 30,000 digoxin concentrations across the State Hospital system were obtained in 2 periods before and after recalibration of the digoxin assay. Digoxin concentration means were calculated and compared and were statistically significantly different. Subsequently, a single-centered retrospective review of 50 randomly chosen charts was undertaken to study the clinical implications of the underestimated concentrations.
Mean digoxin concentrations for 2009 and 2011 were significantly different by 8.8% (confidence interval, 7.0%–10.6%). After recalculating the 2009 concentrations to their “corrected” values, there was a 16% increase in the number of concentrations within the range when compared with the 2011 concentrations (41.48% versus 48.04%). However, overall, this did not cause unnecessary dose changes in patients who were “borderline” or outside the therapeutic range when compared with controls (P = 0.10). The majority of decisions were based on the clinical impression rather than concentration alone (85.1% versus 14.9%), even when the concentration was outside the “therapeutic range.”
Although recalculating digoxin concentrations measured during 2009 to their corrected values produced a significant change in concentration and values inside and outside the range, this does not seem to have had an influence on patient treatment. Rather, clinicians tended to use the clinical impression to dose digoxin.