A randomized trial of cardiovascular risk factor reduction: patterns of attrition after randomization and during follow-up

Abstract

Background: The Preventive Cardiology Center compared two intensities of behavior modification for cardiovascular disease (CVD) risk factors in persons at risk and their families. Characteristics of drop-outs both before and after intervention were compared with subjects who completed the 6-month trial.

Methods: A total of 333 individuals of all ages were enrolled in the study and randomly assigned by family to a single-session ('skills' group-low-intensity) or a five-session ('practice' group-high intensity) intervention. Baseline and follow-up assessments included a personal and family health questionnaire, nutritional intake survey, and clinic visit to obtain blood pressure, lipids, and height and weight data.

Results: Two hundred and forty adults over 18 years of age were randomly assigned to one of the two intervention groups. Of these, 68 subjects (28.3%) did not participate in the intervention. Multivariate analysis revealed that these 'early drop-outs' were significantly more likely to be non-white and to have had a lower LDL cholesterol. Of the 172 subjects attending the intervention, 70 (40.7%) did not attend the 6-month follow-up ('late drop-outs'). Multivariate analysis revealed that, compared with follow-up attendees, non-attendees were significantly more likely to be white and to be current smokers.

Conclusions: Both early and late drop-outs in a randomized trial of CVD risk reduction were significantly different than continuing participants in several key factors. These differences suggest the use of caution in both interpreting and making generalizations about behavioral trials of risk factor reduction when attrition is high.