To the Editor: The interesting study by Dr
Sawicki1 provides further evidence of the potential value
of patient self-management in oral anticoagulation. However, several
peculiarities of the study design make the results difficult to
extrapolate to patients currently receiving optimal care.
The intervention group, in addition to self-assessment of their
prothrombin time and international normalized ratio (INR), had
extensive weekly teaching sessions that included topics and issues that
could clearly affect future coagulation control (effect of diet,
medications). More important, they were instructed in a model of care
regarding how to increase or decrease the anticoagulant dosages to
achieve target INR values. The algorithm they were taught is not
available to the reader. At our ambulatory center, we have been able to
see considerable improvement in achieving target INR values by
standardizing all physicians' adjustments in dosages with a standard
of care algorithm. Was the same model of change in dosage used by the
family physicians who were adjusting the anticoagulants for the control
group? If not, this confounder makes the results nearly
uninterpretable. If used, the algorithm should be provided for review
of practicing physicians. Could differing clinical care explain the
significant effect seen between centers?