The Noncompliance Epidemic

Can We Get Patients to Be More Compliant?

Neil Chesanow

Disclosures

January 16, 2014

In This Article

Identifying Patients Before They Become Noncompliant

Insurers, retail pharmacies, and integrated health systems have begun to use a technique called "predictive modeling" to comb through their massive databases on patients' fill and refill patterns at the pharmacy.[21,22,23] Algorithms are used to identify those who fit the profile of a patient who is likely to be noncompliant.

At-risk patients are then monitored. If it appears that they might go off-regimen, based on when or whether they pick up their scripts, automated reminder systems or live pharmacists then intervene.

According to Walgreens' Kristi Rudkin, predictive modeling has been instrumental in identifying "new-to-therapy patients, who are at significantly higher risk than existing or ongoing patients of deciding not to take their medication," she says.

Last year, researchers at CVS Caremark and Brigham and Women's Hospital in Boston developed a new approach to classifying patients by their long-term medication compliance behavior, which they found is more accurate than simpler traditional approaches.[23] The new method, called "group-based trajectory modeling," is based on observed patterns of medication filling over time.

Using the new method, researchers tracked more than 264,000 statin users over a 15-month period and created measures to account for different compliance behaviors.[23] They then identified key groups into which various patients would fall, based on these measures, to predict their long-term compliance patterns.

By using the new method, "we could help identify patients with distinct patterns of adherence so that healthcare professionals could appropriately target interventions," lead investigator Niteesh Choudhry, MD, PhD, an internist, said in a statement when the results were published.[23]

"In addition," he said, "data related to the quantity and timing of medication availability could help improve our understanding of the effects of nonadherence on clinical outcomes."[23]

Comments

3090D553-9492-4563-8681-AD288FA52ACE
Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as:

processing....