New drug application

In the United States, the New Drug Application (NDA) is the process by which the Food and Drug Administration authorizes a drug to be sold, after verification of the drug's safety and efficacy. The NDA authorizes the drug to be advertised for specific indications, or diseases for which the FDA has accepted evidence that the drug is an effective part of treatment for those specific conditions. Prior to the NDA, the Investigational New Drug Application (IND) effectively approves the use of the drug in specific randomized controlled trials that will produce the evidence of safety and efficacy.

Prior to any trials, either for pure research or as part of an IND, an Institutional Review Board (IRB) must consent to experimental use of drugs in human subjects. The IRB will review in vitro and in vivo results before consenting to the trial, weighing risk against potential benefit.

The IND and NDA process involve four stages, three before approval and one afterwards:

1. Phase I: Intended primarily as a means of determining safety and bioavailability, the classic Phase I trial involves giving a single dose of the drug to healthy volunteers, observing for adverse effects, and sampling blood or otherwise verifying that a pharmacologically active level of the drug is available in the tissues where it will work. In relatively recent years, Phase I trials have been permitted to use patients with active disease, for which there is no effective treatment or all other treatments have failed. To proceed to Phase II, the drug must be bioavailable and have acceptable side effects, although it is understood that all side effects may not be seen in healthy individuals who received a single, or several, doses.
2. Phase II: Using a larger group of patients than the Phase I trial, this stage looks for evidence that the drug has efficacy is an effective treatment for the indication for which the IND was submitted. Phase II trials typically involve a single institution, but the trial is usually double-blind and controlled. Under current ethical guidelines, participants in a trial must give informed consent. The use of placebo controls is acceptable only if there is no recognized therapy for the indication; if there is such a treatment, the control group patients must receive the best currently accepted treatment.
3. Phase III: Assuming Phase II showed efficacy, the study population is extended to a larger number of independent clinical trial centers. Success in this phase demonstrates efficacy in a larger number of patients, treated by different clinicians. As in Phase II, if adverse events outweighing the potential benefit are observed, the trial must be stopped.
4. Review and approval of NDA
5. Phase IV: Postmarketing surveillance receives reports of adverse events from the much larger population when the drug is in general use. Significant numbers of such events will trigger FDA action from requiring additional warnings to prescribers, to removal of the drug from the market.

Phases I-III are under the supervision of independent safety monitoring IRBs, who have access to patient records and which are assigned to the control or treatment group. The IRB can and will stop the trial if the drug's risks exceed any potential benefit. If the experimental drug is significantly more effective than the best standard treatment, the IRB will also stop the trial on ethical grounds, and all patients in the control arm will be offered therapy with the new drug.

The FDA has an equivalent approval process for medical devices (e.g., pacemakers, defibrillators, artificial skeletal joints, etc.), but not for surgical procedures or methods of treatment using approved drugs. A licensed prescriber has the right to order the use of a drug for an indication for which the FDA has not approved it, if, in the professional judgment of that prescriber, it would be appropriate in a specific situation. Some insurers, in the U.S., will not pay for drugs prescribed for other than approved indications.

"Compassionate use" applications can be made, by clinicians, to obtain and use an unapproved drug in a patient for which there are no treatment alternatives, and the manufacturer allows such use. The FDA evaluates these on a case-by-case basis.

A related regulatory consent is administered by the Centers for Disease Control, which may buy, or have manufactured under contract, drugs for diseases for which no manufacturer has or will submit an NDA. The most common class of such diseases are parasitic diseases endemic to tropical countries, but not acquired naturally in the United States, so there is no regular U.S. market that would justify a manufacturer spending the millions of dollars that the IND/NDA process costs.