[Identification and prediction of responders to a therapy. A model and its preliminary application to hypertension]

The effect of a given treatment for a given disease may be estimated from randomized controlled clinical trials, expressed as a single treatment effect averaged over the trial population. However, in recent years there has been an increasing willingness to individualize therapeutic decisions. The method we report here identifies responders by assessing the individual probability of an event, according to the treatment group. We used a treatment-stratified Cox regression model including interaction between treatment and patient's covariates, with common regression coefficients for treated and untreated, except for the special case of a prognostic variable which has an interaction with treatment. Further, we used a discriminate function based on the final model, representing the absolute individual therapeutic effect, to identify the patients to be treated according to a given threshold of clinical efficacy. The model was explored on the INDANA database (which pools individual patient data from clinical trials of anti-hypertensive drug intervention). Data on 36,444 patients, from five randomized controlled trials were included. The results show the relationship between the proportion of avoided events among the avoidable ones, and the proportion of patients treated who were responders, as a function of the threshold of absolute benefit defining responders. The confidence intervals of the absolute therapeutic benefit for each individual were calculated, by using the Monte Carlo simulation method. A comparison of the survival of treated and controlled individuals, in both subgroups of responders and non responders, illustrated the relevance of the model. We conclude that the tools for predicting individual therapeutic benefit do exist. It will be important to assess the reproducibility of these results in other models or in other populations before widespread application. It will be necessary to have a properly computerized environment and to train doctors to use these tools.