Modified visual analog scale symptom-intensity and overall-bother measures for the assessment of symptoms in studies of pharmacologic stress agents

Abstract

Background: Drugs used to induce stress for cardiac imaging studies often cause discomfort. Patient-reported intensity of discomfort should be considered when comparing such agents.

Objective: The purpose of this study was to assess the psychometric properties of a modified visual analog scale (VAS) symptom-intensity measure and an overall-bother measure adapted to assess patient-reported intensity of the adverse events (AEs) associated with pharmacologic stress testing with adenosine, a pharmacologic stress myocardial perfusion imaging (PS-MPI) agent.

Methods: Data were based on 2 complementary, open-label, multicenter, naturalistic, observational studies among patients completing a PS-MPI procedure. Content, concurrent, and discriminant validity was examined by correlating modified VAS symptom-intensity scores obtained from patients with investigator-rated intensity at time of event, overall bother, and patient-reported measures obtained during a structured interview. Test-retest reliability of the overall-bother measure was examined using 1- and 2-hour assessments, and concurrent validity was assessed by correlating counts of symptoms and other patient-reported measures. Responsiveness was examined by calculating change scores of the VAS symptom-intensity measures from baseline to 1 hour among patients who reported symptoms during the 1-hour monitoring period after PS infusion. Low to moderate correlation was defined as a coefficient between 0.3 and 0.5.

Results: A total of 324 patients enrolled in the 2 studies. Content validity of the VAS symptom-intensity and overall-bother measures was established, with nearly all patients reporting that they were a useful way to rate symptom intensity and overall bother. VAS ratings were moderately to highly correlated with physician-rated AE intensity, and patient-reported assessments of symptom intensity, discomfort, and concern (r(s) = 0.21 r(s) = 0.84). Findings to support discriminant validity were inconclusive because of small sample size. Responsiveness was demonstrated with VAS symptom-intensity change scores ranging from 2.8 for headache to 4.9 for chest pain; effect sizes for these differences were large, ranging from 1.6 to 7.3, respectively.

Conclusion: Findings support the validity, reliability, and responsiveness of the modified VAS symptom-intensity measure, and the reliability and validity of the overall-bother measure for use in patients completing PS procedures.