How to anticipate the assessment of the public health benefit of new medicines?

The Public Health Benefit (PHB) of new medicines is a recent and French-specific criterion (October 1999 decree) which is often only partially documented in the transparency files due to a lack of timely information. At the time of the first reimbursement application for a new medicine to the "Transparency Committee", the file is exclusively based on data from randomised clinical trials. These data are generated from a global clinical development plan which was designed a long time before the new medicine's submission for reimbursement. And this plan does not systematically provide the data needed to assess the PHB. Thus, one easily understands the difficulty to anticipate and document this recent French criterion. In France, the PHB is both one of the necessary criteria for the reimbursement submission and an indicator for the national health policy management. Its assessment also helps to identify the needs and objectives of the post-registration studies (nowadays in the scope of responsibilities of the "Drug Economics Committee"). The assessment of the PHB criterion is carried through after the marketing authorization process and is an addition to it. To understand how to anticipate the assessment of the new medicines' PHB, one needs to consider how it differs from the preliminary step of the marketing authorization process. Whereas the evaluation for marketing authorization seeks to determine if the new medicine could be useful in a specific indication, the PHB assessment aims at quantifying the therapeutic benefit in a population, taking into account the reference treatments in this population. A new medicine receives a marketing authorization based on the data of the registration file which provides information on the clinical benefit of the new medicine in the populations of the trials and in the context of the trials. On the other side, the PHB looks at the effects of the new medicine at the scale of the general population, in real practice. The PHB components of a new medicine at first submission are the expected response of this new medicine to a public health need, the expected benefit on the health status of the population and ultimately the expected impact on the health care system. The benefit of a new medicine on the health status of a population is based on public health criteria which can be morbi-mortality or quality of life criteria. However, few registration files contain these public health criteria from the beginning and the predictive value of the surrogate criteria used in the trials is not always precisely assessed. It is, thus, difficult to quantify the expected benefit on these public health criteria. Moreover, the data that enable to quantify the new medicine's effects according to the various characteristics of the target population, are rarely available. Similarly, the French population epidemiological data related to the indication of the new medicine are often not available at the time of the assessment. Therefore it is difficult to evaluate the expected number of events that could be avoided if the new medicine reached the market. The authors suggest to adapt the clinical development plan for a better documentation of the PHB. They specifically recommend to integrate to the judgment criteria (endpoints) of the trials, criteria that are relevant in terms of public health, and to check for a good heterogeneity of the trial populations. They also suggest to start early enough collecting reliable national epidemiological data and the necessary elements for the assessment of the transposability of the trial results to the French population (ability to target the patients to be treated, adaptation of the healthcare system...). About the epidemiological data, the authors consider that the needs are covered in various ways depending on the diseases. To meet the needs of evaluation of the new medicines' target populations in specific indications, they recommend to use ad hoc studies as much as needed. In addition, epidemiological studies designed for market purpose with an acceptable methodology should not be systematically rejected but deserve to be presented. To be able to assess the importance of the expected theoretical benefit of a new medicine in a population, the authors underline the necessity to have access to study results with criteria related to this objective. They suggest to first define and list the criteria by disease. Regarding the representativity of the populations, it comes out that it would be advisable, but unrealistic to include in trials a population 100% representative of the population to be treated. Therefore the effect of the new medicine must be modelised (the "effect model") to be evaluated in the general population. Yet to obtain a reliable effect model, the study population must be sufficiently heterogeneous, which legitimates the demand to ensure a good population heterogeneity at the time of decision-making about trials methodology. When the criteria assessed during the development plan does not correspond to the PHB criteria, the only way to evaluate the number of events related to the PHB criterion is, again, to use modelisation. However, modelisation is only possible when the scientific literature has established a reliable correlation between the two types of criteria. In this case, the new model should be applied to a French target population to assess the expected benefit. As a conclusion, the possibilities to estimate the expected benefit of a new medicine on the health status of a specific population are currently limited. These limitations are regrettable because such an estimate is feasible without disrupting the development plans. The authors' general recommendations to update the development plans seem especially appropriate as the additions should not only be beneficial to France but to all the health authorities who would wish to assess the expected benefit of a new medicine on their territories. Anticipating the lack of clinical and epidemiological data and the lack of data that enable to evaluate the transposability of the trials results to real clinical practice is a sine qua none condition to improve the PHB assessment. The anticipation of these needs should be planned early enough by the pharmaceutical companies which could in this purpose meet the health authorities and the heads of the French public health policy in a consultation.Finally, because of the PHB's universal dimension, it is suggested that the necessary actions and publications be initiated so that the PHB can be acknowledged at the European level.