NEJM Interview: Arnold Monto on the value and feasibility of various methods for evaluating the protection offered by vaccines.

Both randomized control trials and observational studies have been used to assess the effectiveness of vaccines. But observational studies are often dismissed as lacking the methodologic rigor necessary to inform policy making and regulatory decisions. I'm Stephen Morrissey, Managing Editor of the New England Journal of Medicine, and I'm talking with Arnold Monto, a professor emeritus of epidemiology and of public health and co-director of the Center for Respiratory Virus Research and Response at the University of Michigan School of Public Health. Dr. Monto has co-authored a perspective article about the value of various methods for studying vaccine effectiveness. Dr. Monto, how have regulators in the United States traditionally regarded evidence from various types of studies when they're evaluating vaccines?

When regulators such as the FDA examine data so that they can license a vaccine, the sponsors typically have carried out a randomized control trial or maybe even two randomized control trials, depending on the product and the regulations, which are used to make sure that the product, let's say a vaccine, is working as expected and at the level that is required in terms of efficacy to be licensed. But that is only the start, because we never know how well a product will work in the real world when we go to people who do not necessarily volunteer to participate in these trials.

You say in your perspective article that respiratory viruses, especially influenza, present particular challenges for vaccine research. What factors complicate research and vaccine development in that area?

Well, a couple of factors. First of all, in the United States, we now have a universal recommendation, and typically, the most rigorous clinical trials use a comparator, which is placebo. And we couldn't do that in the United States now. Because of the recommendation, it would not be ethical. The other reason that flu vaccines present an issue is that we update the vaccine each year, and the recommendations come in typically from WHO in February, and the FDA meets in March, and we couldn't conduct randomized trials to meet the necessity of having the vaccine available the following September. So there are logistic challenges.

So getting back to the ethical issue you raised, what strategies have investigators developed for maintaining rigor in trial design while they don't violate ethical constraints, given the preexisting recommendations for vaccination?

One of the ways that this is handled is the setting up trials against an active comparator. In other words, if you are trying to figure out how a new preparation works, you will compare it in a randomized design with another product which is already licensed. This happened recently with the new mRNA influenza vaccines, where the comparator was the regular vaccine. Now, there was a problem with this because the comparator, which was regular vaccine, is not the comparator which needs to be used in sixty-five plus individuals. So they're going to be doing another study with a less rigorous but still a randomized design with a different comparator. And that's the kind of issue that you get into. There's no black and