english.prescrire.org > Positions > Therapeutic progress > Combined end points

Theme: Therapeutic progress

New products or techniques are only worthwhile if they translate into better care for patients. Clinical data may mean little in terms of improving patients’ health, and lives. How can we know genuine progress when we see it?

Themes

Combined end points

The pharmaceutical industry and regulatory agencies make do increasingly with methodological shortcuts. These shortcuts may gain time and money, but they also make it difficult to obtain clearcut answers.

The real benefits of "new" treatments can only be determined in comparative clinical trials with end points relevant to patients’ own concerns.

Yet industry, which sponsors most clinical trials, and regulatory agencies too, make do increasingly with methodological shortcuts. These shortcuts may gain time and money, but they also make it difficult to obtain clearcut answers. Combining end points is one such shortcut.

Combined end points lump together distinct or ill assorted events so as to increase the chances of showing a statistically significant difference between two treatments, particularly when the trial is small or treatments are likely to be similarly effective. This may be useful in early stages of assessment, but what about defining benefits for patients?

Take, for example, eptifibatid and tirofiban. The combined end points used in studies comprised death, non fatal myocardial infarction, refractory ischaemia and, in some cases, hospital re-admission for unstable angina, which is totally different from patients’ perspective.

Although statistically significant differences have emerged, clinical relevance is not settled. Does the drug lower mortality? Or does it only cut the risk of non fatal myocardial infarction?

How can we balance the clear risks of treatment against a combination of ill assorted benefits? Patients deserve more than sham innovations.

Prescrire Int 2000; 9 (49): 130.