STAR*D trial

CLINICAL PRACTICE VERSUS CLINICAL TRIALS THE STAR D TRIAL... 

The Oklahoma study showed the same pattern of results as the STAR D trial. Some treatments seemed to be effective for some patients and other treatments seemed effective for others. More than half of the subjects responded successfully to the first treatment 17 per cent did not respond to the first treatment, but did respond to the second. The third treatment was successful for another 20 per cent who had not responded to prior treatments, and by the time the sixth treatment was tried, 100 per cent of the subjects had successfully responded to at least one of them. 

Like the STAR D trial, this study seemed to show that different people respond to different medications and that the key might be finding the right treatment for the right person -but there was a catch. None of the medications were real treatments for nausea or vomiting. Instead, they were all placebos. 

Most clinical trials, including the ones my colleagues and I analysed, are conducted on volunteers, many of whom are recruited for the trial by advertisements. Perhaps these depressed people are not as responsive to antidepressants as the patients seen in clinical practice. The STAR D trial that I described earlier was designed specifically to evaluate the effect of antidepressants on the kinds of patients who are typically seen in clinical practice. None of the patients in this trial were recruited by advertising. Instead, they were all patients who sought treatment for depression in family practice or psychiatric out-patient treatment facilities. Also, the usual exclusion criteria were relaxed, so that a broader range of patients was evaluated. The trial did exclude patients who had already tried antidepressants but had not responded to them, although this exclusion should result in better response rates, not worse ones. 

The STAR D trial is not alone in finding that all antidepressants are created equal. In meta-analyses of head-to-head comparisons of different antidepressants, statistically significant differences are occasionally found, but these tend to be very small... 

Rush, A. J., Fava, M., Wisniewski, S. R. et al. (2004). Sequenced treatment alternatives to relieve depression (STAR D) rationale and design. Controlled Clinical Trials, 25(1), 119-42. 

Using this very strict criterion of remission, the STAR D researchers reported that 37 per cent of the patients in the trial recovered from depression on the first medication they were given. Another 19 per cent of the full group of patients recovered on the second medication, 6 per cent on the third and 5 per cent on the fourth. Altogether, 67 per cent of the patients recovered. However, the remission of symptoms turned out to be only temporary for many - approximately half of the patients who recovered relapsed within a year. 

Nowhere are these transitions more apparent than in the design of three important studies sponsored by the National Institute of Mental Health. One—the STAR D study—is discussed in detail in Chapter 3. The other two, the CATIE study and the STEP-BD study, are reviewed briefly here as an illustration of current trends in psychiatric research. These three studies break new ground in the sense that they go beyond the concept of efficacy that is established in the industry-sponsored studies The drug is shown to be efficacious for a certain disorder, but in the real world, will physicians prescribe it and will patients take it and stay on it The answer to these questions centers on the concept of effectiveness. An effective drug is one that not only has efficacy against a certain malady but also is accepted by patients because its safety, tolerability, mode and frequency of administration, price, taste, appearance, and so on. These issues are commonly left unexamined in the classic drug trials. 

---