P-414: Predictors of “lost-to-follow-up” status in the convince trial

The Controlled ONset Verapamil INvestigation of Cardiovascular Endpoints (CONVINCE) Trial randomized 16,602 hypertensive volunteers from 15 countries to initial antihypertensive therapy with either Controlled-Onset, Extended Release (COER)-verapamil or “Standard of Care” (SOC), a physician-directed...

Full description

Saved in:
Bibliographic Details
Published in:American journal of hypertension 2003-05, Vol.16 (S1), p.187A-187A
Main Authors: Black, Henry R., Elliott, William J., Grambsch, Patricia M., Li, Bing B., Wu, Xiaoting, Grandits, Gregory, Neaton, James D., Anders, Robert J.
Format: Article
Language:English
Subjects:
Censoring
Dropouts
Protocol adherence
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The Controlled ONset Verapamil INvestigation of Cardiovascular Endpoints (CONVINCE) Trial randomized 16,602 hypertensive volunteers from 15 countries to initial antihypertensive therapy with either Controlled-Onset, Extended Release (COER)-verapamil or “Standard of Care” (SOC), a physician-directed choice of atenolol or hydrochlorothiazide (HCTZ). The primary outcome (PO) was the first occurrence of myocardial infarction, stroke, or cardiovascular disease-related death. Median follow-up was 3 years. Substantial efforts were made at the end of the study to ascertain the PO status of every volunteer, but were unsuccessful for 1113 volunteers. A Cox model was developed to identify characteristics associated with an increased risk of having unknown PO status; all risk factors were entered simultaneously into the model. The results were: Other baseline parameters, including SOC choice (atenolol v. HCTZ), gender, diabetes, left ventricular hypertrophy, dyslipidemia, obesity, and vascular disease, were not significant predictors of having unknown PO status at the end of the study. As in prior international hypertension trials, volunteers from the USA were significantly more likely to become lost. Those who were lost were more commonly minorities, older, tobacco users, and had higher baseline systolic BPs as well as prior cardiovascular events. These data indicate that, in CONVINCE, higher risk volunteers were more likely to become lost-to-follow-up. This should increase the number of unreported events beyond that expected if volunteers became lost at random. (See Table) Risk Factor Comparison Relative Risk (95% CI) Age 10 years 1.11 (1.02–1.21) Systolic BP 10 mm Hg 1.07 (1.03–1.11) Country of Origin Canada v. USA 0.46 (0.39–0.55) Country of Origin Non-North American vs. USA 0.46 (0.39–0.55) Race/ethnicity White v. Other 0.76 (0.66–0.88) Prior CV Event Yes v. None 1.32 (1.12–1.57) Cigarette Use During last 3 years v. Not 1.53 (1.33–1.75)
ISSN:0895-7061
1879-1905
1941-7225
DOI:10.1016/S0895-7061(03)00586-7