Self-reported information and pharmacy claims were comparable for lipid-lowering medication exposure

Centers for Disease Control and Prevention, Atlanta, Georgia 30341, USA.
Journal of Clinical Epidemiology (Impact Factor: 3.42). 06/2007; 60(5):525-9. DOI: 10.1016/j.jclinepi.2006.08.007
Source: PubMed


To examine agreement between self-reported exposure to lipid-lowering medications and objective evidence of filling prescribed lipid-lowering medications.
Using data from 7,918 adults from the Adverse Childhood Experiences (ACE) Study, we calculated the sensitivity, specificity, and positive (PV+) and negative (PV-) predictive values, and likelihood ratios for self-reported exposure to lipid-lowering medications compared to exposure obtained from pharmacy claims (gold standard) both overall and by age, sex, race/ethnicity, education, and ACE Score.
Eight percent (n=655) of adults self-reported lipid-lowering medication exposure, and 379 adults filled at least one lipid-lowering prescription within 60 days of the baseline exam during 1997. The sensitivity of self-reported exposure was nearly 94%; the specificity was 96%; the PV+ was 54%; and the PV- was nearly 100%. Values for sensitivity, specificity, PV+, and PV- were similar across participant characteristics.
A self-reported measure of lipid-lowering medication exposure was accurate with high sensitivity and specificity while the PV+ of self-reported lipid-lowering medication exposure was relatively low. These findings suggest that self-reported exposure to lipid-lowering medications may be useful in surveys that examine the prevalence of hyperlipidemia, but may overestimate actual exposure in studies monitoring trends in use of lipid-lowering medications.

Download full-text


Available from: David W Brown, Dec 24, 2014
  • Source
    • "The correspondence between pharmacy submission of claims and patients’ receipt and consumption of the medication was assumed and not directly measured. However, prior work suggests that medication exposure can be accurately derived from pharmacy claims [21, 22]. The study also assumed that all information needed for cohort stratification was present and similar across the cohorts of interest. "
    [Show abstract] [Hide abstract]
    ABSTRACT: While the liraglutide effect and action in diabetes (LEAD-6) clinical trial compared the efficacy and safety of liraglutide once daily (LIRA) to exenatide twice daily (EXEN) in adult patients with type 2 diabetes, few studies have explored the associated per-patient costs of glycemic goal achievement of their use in a real-world clinical setting. This retrospective cohort study used integrated medical and pharmacy claims linked with glycated hemoglobin A1C (A1C) results from the IMS Patient-Centric Integrated Data Warehouse. Patients' ≥18 years and naïve to incretin therapies during a 6-month pre-index period, with ≥1 prescription for LIRA or EXEN between January 2010 and December 2010, were included. Patients with evidence of insulin use (pre- or post-index) were excluded. Only patients who were persistent on their index treatment during a 180-day post-index period were included. Follow-up A1C assessments were based on available laboratory data within 45 days before or after the 6-month post-index point in time. Diabetes-related pharmacy costs over the 6-month post-index period were captured and included costs for both the index drugs and concomitant diabetes medications. 234 LIRA and 182 EXEN patients were identified for the analysis. The adjusted predicted diabetes-related pharmacy costs per patient over the 6-month post-index period were higher for LIRA compared to EXEN ($2,002 [95% confidence interval (CI): $1,981, $2,023] vs. $1,799 [95% CI: $1,778, $1,820]; P < 0.001). However, a higher adjusted predicted percentage of patients on LIRA reached A1C < 7% goal (64.4% [95% CI: 63.5, 65.3] vs. 53.6% [95% CI: 52.6, 54.6]; P < 0.05), translating into lower average diabetes-related pharmacy costs per successfully treated patient for LIRA as compared to EXEN ($3,108 vs. $3,354; P < 0.0001). Although predicted diabetes-related pharmacy costs were greater with LIRA vs. EXEN, a higher proportion of patients on LIRA achieved A1C < 7%, resulting in a lower per-patient cost of A1C goal achievement with LIRA compared to EXEN.
    Preview · Article · Jan 2014 · Advances in Therapy
  • [Show abstract] [Hide abstract]
    ABSTRACT: To compare self-reported (SR) medication use and pharmacy data for major psychoactive medications and three classes of medications used for different indications, and to determine the socio-economic factors associated with the congruence. Postal questionnaire data collected in 1997 were compared with the register of the Social Insurance Institution of Finland on the reimbursed prescriptions purchased during 1997. Altogether 7625 subjects were included in this study. Drugs were categorized according to the Anatomical Therapeutic Chemical (ATC) system. Kappa values were 0.77, 0.68, 0.84, 0.92 and 0.55 for antipsychotics, antidepressants, antiepileptics, antidiabetics and beta-blocking agents, respectively. Prevalence-adjusted and bias-adjusted kappa values were almost perfect (0.98-1.00). Reliability of antipsychotics use was better for married subjects than for those who were not married; and of antidepressants use for highly educated and married subjects than for those who were less educated and were not married. Altogether 414 (5.4%) responders and 285 (7.1%) non-responders had used at least one of the selected medications. Agreement between the SR and pharmacy data was moderate for psychoactive medication use. Even though data collected by postal questionnaire may underestimate the prevalence of medication use due to non-participation it can be assumed accurate enough for study purposes.
    No preview · Article · Jun 2010
  • [Show abstract] [Hide abstract]
    ABSTRACT: Despite the importance of comorbidity in multiple sclerosis (MS), methods for comorbidity assessment in MS are poorly developed. We validated and applied administrative case definitions for diabetes, hypertension, and hyperlipidemia in MS. Using provincial administrative data we identified persons with MS and a matched general population cohort. Case definitions for diabetes, hypertension, and hyperlipidemia were derived using hospital, physician, and prescription claims, and validated in 430 persons with MS. We examined temporal trends in the age-adjusted prevalence of these conditions from 1984-2006. Agreement between various case definitions and medical records ranged from kappa (κ) =0.51-0.69 for diabetes, κ =0.21-0.71 for hyperlipidemia, and κ =0.52-0.75 for hypertension. The 2005 age-adjusted prevalence of diabetes was similar in the MS (7.62%) and general populations (8.31%; prevalence ratio [PR] 0.91; 0.81-1.03). The age-adjusted prevalence did not differ for hypertension (MS: 20.8% versus general: 22.5% [PR 0.91; 0.78-1.06]), or hyperlipidemia (MS: 13.8% versus general: 15.2% [PR 0.90; 0.67-1.22]). The prevalence of all conditions rose in both populations over the study period. Administrative data are a valid means of tracking diabetes, hypertension, and hyperlipidemia in MS. The prevalence of these comorbidities is similar in the MS and general populations.
    No preview · Article · Feb 2012 · Multiple Sclerosis
Show more