Is Comparative Effectiveness Research the Best Way to Choose Your Next Treatment?

Portrait of Dr.Jeffrey mcCombsWith a $1.1 billion commitment for comparative effectiveness research in health care stipulated by the American Recovery and Reinvestment Act of 2009, it’s no surprise this type of research has the health care world buzzing.

Comparative effectiveness research, also called evidence-based health care, evaluates how various drugs, treatments and services stack up against one another in real-word effectiveness and cost. There is a current emphasis on physicians considering this research when selecting a patient’s treatment course, as it is thought to be the best way to compare the many available treatment options and to rationalize care across the country.

So on your next doctor’s visit, should your treatment be based upon this type of research to ensure the best (and most cost-effective) option?

A new study from a team at the USC School of Pharmacy details potential sources of bias that clinicians, pharmacy and therapeutic committees, health insurance companies, HMOs and government programs must be aware before using comparative effectiveness research based on retrospective data.

The study, “Impact of Drug Treatment History on Comparative Effectiveness Research in Schizophrenia,” which appears in the June 13th issue of Value in Health, looks at comparative effectiveness research relating to patients prescribed treatments for schizophrenia. Study authors are associate professor Jeffrey McCombs, director of graduate studies at the School of Pharmacy, and graduate research assistants Sara Zolfaghari and Vaidynathan Ganapathy.

“These studies typically consider very restricted sub-samples of all observable patient treatment episodes, focusing instead on patients with extended periods in which no relevant drugs are used,” explains McCombs who is lead author of the journal entry.

Because comparative effectiveness research inevitably samples from patient populations that are not representative of a wide range of patients, the results of these studies do not apply to the majority of patients seeking care. In McCombs’ research with a wide range of patients taking medications for schizophrenia, he discovered that patient outcomes varied significantly with patient treatment history, especially episode type. Unfortunately, treatment history is not always taken into account when conducting research analysis which includes a wide array of patients.

Although McCombs’ study looked specifically at schizophrenia treatment, he argues it can be generalized to any type of comparative effectiveness research.

“Concern should extend beyond severe mental disorders to include any disease state in which long-term drug therapy is indicated but difficult to achieve,” he explains in his article. “Hopefully, the standard of practice for observational research will move toward including the full range of treated patients and accounting for each patient’s treatment history in their analyses.”

The article abstract is available at