Pharmacy chain and prescription benefits manager CVS and the Brigham and Women’s Hospital took a look at medication-adherence data and are proposing a new way to consider patient patterns. Instead of focusing on what’s called “proportion of days covered,” which looks at the number of pills available during a given time span, researchers said a better way to understand a patient’s medication-taking habits is to adopt a longer, more holistic perspective, also known as a group trajectory model. Medical Care published the study in its September issue.

The researchers explain that moving toward a trajectory model, which accounts for multiple behaviors, as opposed to a prescription-fill count, would mean employing the same type of models used in disease-progression and health-behavior research. The compiled data reveals the sort of audience data that appears in the audience part of a marketer’s creative brief, merging data points like income, education and geography.

The researchers explain that adding these variables gives a better perspective than the traditional proportion-of-days covered accounting because an identical PDC score could make a patient with good initial habits that degrade over time look the same as a patient with intermittent use.

The basis of the trajectory study were statin users who used local pharmacies. Researchers created supply diaries that tracked 15 months worth of medication supply. This diary was the basis of 15 variables that determined patients were adherent for around 24 days out of a 30-day calendar month. For contrast, they also calculated proportion of days covered for this same period.

In addition to fill numbers, researchers also pulled information including age, sex, geography, co-payment and total co-payment burden for the statin as well as other medications.

Using the trajectory approach, researchers were able to identify six adherence patterns, which broke out as follows:

  • 23.4% were always adherent
  • 11.4% had a gap in adherence, but improved at the end of the examined time period
  • 11.3% had declining adherence throughout the 15 months
  • 15% used statins sporadically over 15 months
  • 19.3% had a rapid decline in use after starting the medication
  • 23.4% “had virtually no fills” after the first

Adding texture to the numbers is the additional information researchers were able to mine. Among the findings:

  • The highest adherent group—dubbed Trajectory 1—were generally older, likely had a higher income, likely to have graduated high school, more likely to be on Medicare, live in New England. This group was also less likely to be African-American.
  • The least adherent group—Trajectory 6—tended to be younger, generally male and less likely to have 30 days’ worth of pills than those who fell into patterns one through five.

This perspective has some overlap with previous adherence discussions, by addressing factors such as total prescription cost burden. This added information is a nice overlay to previous studies which have addressed adherence trends and costs. The researchers note that trajectory models may be a way for investigators to hone in on particular adherence patterns and test out distinct interventions to see what strategies work with each patient track.