Researchers long have known many people end cognitive behavioral therapy before the recommended course of treatment has ended, but why that happens has remained something of a mystery.

"We don't know why," said Partha Krishnamurthy, director of the Institute for Health Care Marketing at the University of Houston's Bauer College of Business. "The person just doesn't show up."

New research by Krishnamurthy and his colleagues involving patients being treated for anxiety has suggested some answers. The work, "Survival Modeling of Discontinuation from Psychotherapy: A Consumer Decision-Making Perspective," was published in the Journal of Clinical Psychology.

The researchers found patients who improve quickly are more likely to drop out before completing treatment, losing the potential for future benefit.

"We believe the heartbeat of the finding is the speed of improvement, rather than the level of improvement," said Krishnamurthy, lead author of the paper and professor of marketing at UH. "The faster they get better compared to where they started, the more likely they are to abandon treatment."

The stigma associated with mental health problems may be one reason. Although there is no direct evidence, he said the patient might be thinking of the stigma and reasoning, "I've gotten better, so why should I continue being seen seeking treatment?"

The study involved 139 people enrolled in a 12-week course of cognitive behavioral therapy at an anxiety disorder research clinic run by the University. During each session, the patients' anxiety levels were assessed. In addition to Krishnamurthy, the researchers included Adwait Khare, associate professor of marketing at the University of Texas-Arlington, Suzanne C. Klenck, then a UH graduate student, and Peter J. Norton, former associate professor of psychology at UH.

Because cognitive behavioral therapy and other forms of psychotherapy are intended to take a number of sessions, patients regularly must decide if the benefits are worth the financial and emotional costs, the researchers wrote.

Traditional decision-making research suggests patients should continue treatment as long as they are improving. But the researchers found a correlation between reduced symptoms and stopping treatment, meaning those traditional decision-making factors may not apply to mental health care.

"As the patient starts experiencing improvement, the desire to get better becomes less pronounced compared to the social, emotional, financial and time costs of continuing therapy," they wrote. "Clients who approached their symptom reduction goal were more likely to discontinue treatment activities. More importantly, we found ... that faster improvement leads to greater discontinuation."

People who began with the highest anxiety levels also were more likely to drop out. That could be because the illness affected their ability to decide they need treatment, Krishnamurthy said, as well as their ability to get to appointments and other issues.

Previous research had shown that people often end therapy prematurely - more than half of patients, according to some studies - making it harder to address mental health needs, Krishnamurthy said.

The researchers suggested patients with the highest anxiety levels could benefit "from immediate symptom relief strategies, such as relaxation exercises or medication."

And Krishnamurthy said that, while the work isn't intended to offer specific solutions, strategies such as focusing on future gains, rather than simply recognizing progress already made, could be helpful.

Financial incentives - reducing copays for patients after they reach a certain point in the treatment, for example - also might help, he said.

"How do you make people do those things that are not easy or fun?" he asked. "We need to reframe the issue. At the end of the day, good health outcomes are the result of patients' choice process, as much as they are based on drugs and devices. Understanding how patients make decisions is a critical component of improving health care."