Background: Nonadherence to treatment recommendations is associated with poorer outcomes in inflammatory bowel disease and may increase the cost of care. We examined the longitudinal relationship between nonadherence and health care costs and hypothesized that at least 3 distinct trajectories of nonadherence would be observed and that increasing nonadherence would account for significantly greater health care costs after controlling for disease activity.
Methods: Ninety-nine patients aged 2 to 21 years with inflammatory bowel disease were recruited into this 2-year longitudinal study. Medication possession ratios were calculated from pharmacy refill data, disease activity ratings were obtained from medical charts, and hospital and physician charges associated with an International Classification of Diseases, Ninth Revision code for ulcerative colitis or Crohn's disease were obtained from the hospital's accounting database.
Results: An average total cost effect size of d = 0.68 was observed between the increasing severity and stable low severity groups, but the confidence intervals overlap. Conversely, patients with increasing nonadherence demonstrated significantly higher health care costs than patients with stable ≤10%, stable 11% to 20%, or decreasing nonadherence.
Conclusions: Medication nonadherence is related to increased health care costs after controlling for disease severity. Patients with increasing nonadherence over time demonstrate more than a 3-fold increase in costs compared with adherent patients. In addition, patients whose adherence improves over time incur approximately the same costs as those who are consistently adherent. This suggests that, in addition to leveraging prevention efforts to keep patients from becoming more nonadherent as treatment continues, efforts aimed at modifying adherence behavior may result in significant cost savings over time.
Article first published online 14 June 2017.
*Division of Behavioral Medicine and Clinical Psychology, Cincinnati Children's Hospital Medical Center, Cincinnati, Ohio;
†Department of Pediatrics, College of Medicine, University of Cincinnati, Cincinnati, Ohio; and
‡Department of Pediatrics, University of Michigan School of Medicine, Ann Arbor, Michigan.
Address correspondence to: Kevin A. Hommel, PhD, Division of Behavioral Medicine and Clinical Psychology, Cincinnati Children's Hospital Medical Center, 3333 Burnet Avenue MLC-7039, Cincinnati, OH 45229 (e-mail: firstname.lastname@example.org).
Funding for this study was provided by NIH T32 DK007727 and NIH T32 HD068223.
The authors have no conflict of interest to disclose.
Received December 27, 2016
Accepted April 25, 2017