A paper by Proudman et al. (2024) finds that doing so would be extremely expensive.

Including all preclinical, clinical and post-approval R&D activities, total annual costs to maintain the average lifecycle level of 49.4 approved drugs across all disease areas during the 2018–22 period was estimated to be $139.6 billion. Including only pre-launch costs, total annual costs of development for all drugs is estimated to be $110.9 billion. These figures are not capitalized.

When stratified by the 3 innovation designations, the total annual estimated R&D costs for all approved first-in-class drugs were $55.6 billion; for all approved advance-in-class drugs were $25.6 billion; and for all approved addition-to-class drugs were $58.2.1 billion (total for the 3 categories sums to the $139.6bn).

Comparing to NIH funding for fiscal year 2022, the total estimated lifecycle costs of development were equivalent to 302% of the entire budget or 25 times larger than the dedicated budget the NIH is currently spending on clinical research activities as estimated by the GAO.

This also assumes that NIH would be able to ‘pick the winners’ in terms of the best ROI clinical investments as well as the private sector, which is unlikely You can read the full paper and methods description here.