Phil B.FontanarosaMD, Deputy EditorIndividualAuthorStephen J.LurieMD, PhD, Fishbein FellowIndividualAuthor
In Reply: We agree with Dr Bausell that a patient's
perception of retrospective change will influence the patient's level of satisfaction
and that this could have an acquiescent component, but what the only alternative
is not to ask the patient at all. Also, the patients' personal physician did
not ask these questions; patients were surveyed by mailed questionnaire sent
from the study center.
We understand that change scores may be more or less variable than baseline
or follow-up scores individually. However, if we cannot use retrospective
judgment because of faulty memory or an acquiescent response, and we cannot
use serial change scores because they are more "unreliable," what do we do?
At this point, the criticism appears to depart from reality. At least the
retrospective judgments have only 1 source of variance, and our results showed
that the retrospective variances were much smaller than the variances for
change scores, thus giving a much better sensitivity.
Holman H, Bloch DA, Lorig K, Laurent D, Fischer D, Stewart AL. Assessing Patients' Views of Clinical Changes—Reply. JAMA. 2000;283(14):1824-1825. doi:10.1001/jama.283.14.1821