Assessing the Responsiveness of a Quality-of-Life Instrument and the Measurement of Symptom Severity in Essential Hypertension

Publisher Website
Google Scholar
Abstract
A pilot study was conducted to compare symptoms elicited with an open-ended question versus a checklist and to measure the responsiveness of quality-of-life measures to symptom severity. The pilot study was part of a multicentre, randomised, double-blind, placebo-controlled study of clentiazem, a calcium channel blocker, in the treatment of essential hypertension. Symptom and quality-of-life data were obtained from 88 patients at baseline and after 10 weeks of therapy by a trained telephone interviewer. Comparison of the symptom checklist and open-ended question method suggests that both methods are necessary to capture severe symptomatology. The 24-item checklist failed to elicit approximately 50% of the severe symptoms reported on the open question list. On the other hand, only 18% of the most severe symptoms subsequently reported on the checklist were first reported by the open question method. The responsiveness of quality-of-life measures to symptom severity was tested using a 20% change in symptom severity obtained from the checklist as the minimal clinically significant difference. Using Guyatt's formula, a minimum sample size of approximately 428 (alpha = 0.05, beta = 0.10) patients per treatment group is required to detect differences in measures of general health perception, anxiety, depression and limitations in social activities. A larger sample is required to show differences in leisure activities. Differences in limitations of the capability to perform house or yard work might be demonstrable with as few as 17 patients per group. This pilot study demonstrated that the severity of symptoms associated with hypertension, and the side effects of its treatment with drugs, are adequately captured by a symptom checklist preceded by an open-ended method of questioning. Responsiveness testing estimated the sample size required to show a statistically significant difference, assuming a 20% change in symptom severity.