Randomization Process in Question: Efficacy Trials Evaluating Psychotherapy vs Medications May Not Be Valid

Irving Kuo, MD


November 01, 2007

Acceptability of Second-step Treatments to Depressed Outpatients: A STAR*D Report

To determine factors that affect patients' willingness to accept different treatment options in a population of treatment-resistant individuals, 4041 individuals with a diagnosis of depression from the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) trial were entered into the level 1 portion where they received the antidepressant medication citalopram. Patients who had not achieved remission or who had intolerable side effects to the citalopram were encouraged to enter the level 2 study phase, which included the following options:

  1. Switch to bupropion sustained-release

  2. Switch to sertraline

  3. Switch to venlafaxine extended-release

  4. Switch to cognitive therapy

  5. Add bupropion sustained-release to citalopram

  6. Add buspirone to citalopram

  7. Add cognitive therapy to citalopram

The equipoise randomization strategy allowed patients to indicate a preference to which of the level 2 options they would accept or refuse.

Only 1% of the cohort would accept all 7 treatment options. Only 26% of patients were willing to accept cognitive therapy as a switch or augmentation strategy. This group tended to have a higher education level and family history of depression or bipolar disorder. Most subjects were only willing to accept either a medication switch or augmentation strategy. Those who desired a switch had a higher side effect burden and less symptom improvement with the citalopram, while those accepting augmentation had fewer side effects and a higher level of symptom improvement.

This study puts into question the validity of the absolute randomization strategy used in many trials because patients may be randomized to treatments that they do not find acceptable and, thus, the likelihood of being effective is much lower (either through diminished adherence or altered placebo effect). This may be especially true in studies comparing efficacies of psychotherapy vs medications because this study indicated that a certain demographic is much more willing to accept psychotherapy as a treatment. It also raises the possibility that the clinician's bias toward various treatment options may influence a patient's perspective about these treatments -- in other words, how strongly clinicians "sell" their personal treatment preferences to the patient. In addition, the initial treatment experience of the patient appears to be the largest influence on his or her choice of either a treatment switch or augmentation strategy, which makes intuitive sense.



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: