Merck web widget points to future adherence fix
The three-question survey, which groups patients into low, medium and high risk for non-adherence to a newly-prescribed medicine, aims to keep patients on treatment by providing them with information specific to their concerns.
“Merck's research over the past several years shows that health beliefs can be strong predictors of who may or may not take their prescribed medical treatment,” said Merck VP, US medical affairs Sethu Reddy, MD, in a statement. “Our studies point specifically to three health beliefs that are the best predictors: understanding why the medicine is needed and being committed to its use, concern about taking the medicine, and perceived cost issues.”
It's an early effort at building an algorithm to predict non-adherence and allow for preemptive intervention by drug companies, healthcare professionals and health insurers—something all parties in the healthcare industry are looking at as healthcare reform law incentives keeping patients on-treatment and the advent of electronic medical records promises a continuous feedback loop between patients and professionals.
“It's the best practical tool we've seen because it's so simple,” said Jay Bolling, president and CEO of Roska Healthcare Advertising, which is working with clients on similar tools based on the same questions. “If you ask people to fill out a 30-question survey, forget it.”
The Adherence Estimator asks patients to give yes or no answers to the following questions: "I worry that my prescription medication will do more harm than good to me;" "I am convinced of the importance of my prescription medication;" "I feel financially burdened by my out-of-pocket expenses for my prescription medication."
A study of the survey, designed by Merck's Colleen McHorny and Abhijit Gadkari, found it highly accurate in predicting non-adherence but less so in forecasting adherence. The study correctly pegged non-adherers 88% of the time, based on their answers, but was right only 59% of the time in identifying adherers from a group of 100 patients.
The other pitfall of this approach is that the EMR infrastructure isn't quite there to support automated reporting, but it suggests how the system could work to keep patients on treatment and out of the ER once it is.
“If it's not almost automatic it can be hard for the office to act on it, but if I'm a nurse looking at a patient's EMR and it shows that patient is likely to be non-adherent, I can spend extra time on that patient,” says Bolling.