Rich SE (VAMC (Bedford) & Boston University, Schools of Dental Medicine and Public Health)
Wehler CJ (VAMC (Bedford) & Boston University, School of Dental Medicine)
McCoy LC (VAMC (Bedford))
Miller DR (VAMC (Bedford) & Boston University, School of Public Health)
Rothendler JA (VAMC (Bedford) & Boston University, School of Public Health)
Christiansen CL (VAMC (Bedford) & Boston University, School of Public Health)
Jones JA (VAMC (Bedford) & Boston University, School of Dental Medicine)
Attempts to minimize attrition in randomized clinical trials would be helped by better information on predictors of attrition. We assessed correlates of withdrawal in our trial of the efficacy of periodontal care on glycemic control in VA patients.
We examined withdrawals in 166 study participants with poorly controlled diabetes (HbA1c>8.5), 8+ teeth, and periodontal treatment need (CPITN > 2 in at least two sextants). Bivariate analyses (Chi-square and t-tests, a=0.05) compared subjects who withdrew from the study to those who did not by study group, self-reported functional status, comorbid medical conditions, medication, study site, military service (era, branch, combat experience), sex, age-group, marital status, race, education, tobacco/substance use, mental health problems, weight change, exercise change, baseline HbA1c, and type of diabetes medication.
Withdrawal was more likely in participants who took insulin only (vs. oral medication only or oral + insulin), reported poor oral health or poor general health, brushed teeth less than twice/day, served during the Vietnam or post-Vietnam era, have been treated for substance abuse, alcohol abuse, psychological problems, PTSD or bipolar disorder, or had decreased their activity level. There was no difference in withdrawal rates by study group.
Certain health behaviors, comorbid conditions, and medical treatments were predictors of subject attrition.
This information may be useful in efforts to retain participants in future trials. It also serves make the VA aware of factors which may reduce a patient’s adherence to a treatment regimen.