A treatment must above all be evaluated according to pertinent clinical criteria that are obvious to any patient, such as mortality, pain, disability and effect on quality of life.
In clinical trials, pharmaceutical companies often use other so-called "intermediate" criteria such as lower blood pressure or lower cholesterol levels – effects which are faster and easier to obtain. For patients however, these criteria are not necessarily either useful or sufficient. For example, a treatment which increases bone density in menopausal women may have no impact on the prevention of symptomatic bone fractures.
In some cases, when clinically pertinent evaluation criteria take a long time to gather, or are particularly serious in nature, it can be helpful to use "surrogate" criteria that are systematically correlated with the development of the disease, such as, for example, CD4 lymphocyte count and viral load in the evaluation of AIDS drugs.
However, a treatment which appears effective according to surrogate criteria can be ineffective, even harmful, according to criteria that matter to patients. For instance, a drug that lowers cholesterol can increase cancer mortality.
The evaluation of a drug's benefits must not overshadow the evaluation of its adverse effects.
©Prescrire 1 January 2008
Source: "Evaluer les bénéfices d'un traitement : d'abord les critères cliniques utiles aux patients" Rev Prescrire 2008; 28 (291): 69-70.
- More articles in Prescrire's "Spotlight"...