News & Views: Why Some Vaccine Trials Don’t Use a Saline Placebo

Editor’s Note: This feature article includes a video option. In the 12-minute video, Dr. Lori Handy discusses the information in this article. Check out the video.

A recent mantra has been that “gold-standard” science is informed by randomized controlled trials (RCTs) and that the only appropriate control in such trials is a simple saltwater (saline) placebo. These ideas are misleading in two ways. 

1. While RCTs are important scientific tools, we can learn from other types of studies, and we should not disregard information uncovered using alternative methods simply because they aren’t an RCT. 
2. Saline-based placebos are not a requirement of a well-done RCT — nor should they be. The bigger picture is more nuanced because clinical trial design must be informed by what researchers are trying to learn and what can be done safely and ethically. 

Because families are hearing the discussions around RCTs and the types of placebos used, we wanted to do a deeper dive this month and offer resources, so you are prepared if families and patients come to you with questions. 

Clear and unbiased answers

RCTs have characteristics that decrease the risk for bias while also controlling for external variables that may suggest a relationship when one doesn’t actually exist, known as confounding variables. In short, RCTs are designed to give clear, unbiased answers. The characteristics of an RCT are:

• Randomized, meaning participants are assigned to different study groups by chance. This ensures that groups are similar, so differences seen in the study results are due to the intervention (e.g., receipt of a candidate vaccine) and not the participants.
• Double blind, meaning neither the participants nor the researchers know who received which intervention. This decreases opportunities for bias in the results. For example, we know there is a true “placebo effect,” meaning simply being in the study may cause some individuals to change their behavior or perceive changes in how they feel. Since participants in a double-blinded study won’t know if they have received the intervention or placebo, any placebo effect should be similar across groups, thereby canceling out the effect. Similarly, if researchers know who received an intervention, they may act differently or interpret data differently — not out of malice, simply out of unconscious bias.
• Placebo controlled, meaning the group not receiving the intervention still receives something that standardizes the interaction with the research team. This, too, helps decrease the risk for bias among participants and researchers. The presence of a placebo group also removes another risk for bias, and that is the risk associated with random events that can occur simply by living, like accidents, disease diagnoses, or other occurrences that could be questioned as whether they were caused by the intervention if no comparison group existed. 

Defining a placebo

A placebo is an inert or innocuous substance used in controlled experiments. Many people interpret this to mean a saline injection or sugar solution — and sometimes that is exactly what is used. However, in clinical research studies of vaccines, a control group can also receive another vaccine, an earlier version of a vaccine, or some comparator that allows researchers to isolate the effect of the specific vaccine that stimulates the immune system. This means that a placebo can be all the ingredients of the experimental vaccine except the active ingredient that generates immunity (called the antigen.)

Let’s take a look at some landmark vaccine trials to highlight the value of elegantly designed control groups. 

Using saline as a placebo: COVID-19 vaccine trials

The mRNA COVID-19 vaccine trials were textbook double-blinded, placebo-controlled RCTs. Individuals and investigators did not know who received vaccine or saline. They were followed over time for side effects, and because the disease was prevalent at the time of the trial, the end point of the trial — meaning the thing being measured to determine if the vaccine worked — could reasonably be infection and severe disease from infection. Two key facts made this type of trial the ideal choice. 

1. There was no existing vaccine, so providing saline placebo was not limiting someone from receiving a known benefit. 
2. The disease was prevalent, so investigators could study the disease outcome reliably and quickly. 

While this was the ideal design during the pandemic, many other vaccine trials differ from this classic approach for good reason. 

Using a licensed vaccine as a placebo: Acellular pertussis vaccine trials

The first pertussis vaccines, composed of whole, dead pertussis bacteria, became available in the early 1900s. However, these versions contained about 3,000 immunologic components, and recipients often experienced side effects, including some that were severe. In 1996, the acellular pertussis vaccine became available, containing fewer immunologic components (two to five) and causing fewer side effects. (Acellular meant that the vaccine didn’t contain the whole bacterial cell.) Because pertussis can be deadly, particularly for young infants, scientists compared the new acellular vaccine with the whole-cell pertussis vaccine.

Using the whole-cell pertussis vaccine as the control group defies the notion of a saline-based placebo. However, knowing that the whole-cell pertussis vaccine prevented suffering and death, it would have been unethical to leave participants in the control group susceptible to pertussis by using a saline placebo. By comparing a new vaccine to an existing one, researchers can ensure that no one in the study is denied effective protection. This approach also makes sense from the perspective that researchers — and frankly providers and the public — would want to know whether the new vaccine is equal to, or more effective than, the existing vaccine. Likewise, it is critical to understand the relative rates of side effects, particularly in a situation like that of the pertussis vaccine where the side effects were one of the main reasons for developing a new vaccine. 

Using an adjuvant as a placebo: Hepatitis B vaccine trials

Current hepatitis B vaccines contain aluminum as an adjuvant. Adjuvants allow vaccines to induce a better immune response and be given in fewer doses and in lesser quantities. Aluminum has been used in vaccines as an adjuvant for about 100 years.

Because aluminum has been used for so long, we know what side effects to expect. For example, when a vaccine contains aluminum, injection-site reactions, like pain and redness, tend to be more common. As such, when hepatitis B vaccines were studied, they were compared to a placebo that contained the adjuvant. With this design, researchers were able to isolate side effects due to the active ingredient (antigen) in the vaccine. Additionally, in this case, the placebo helped maintain study blinding because researchers and participants were not accidentally unblinded based on the anticipated side effects of injection-site reactions. In this example, using a saline placebo would have weakened the study design.

Using another vaccine as a placebo: Conjugate pneumococcal vaccine trials

The study of pneumococcal conjugate vaccine (called PCV7 because it contained seven different pneumococcal types) in the late 1990s was done at a time when there was not an effective pneumococcal vaccine for infants. However, instead of saline, the placebo in this study was a conjugate meningococcal vaccine. The study was designed to measure whether the pneumococcal vaccine was safe and worked. The choice to use the meningococcal vaccine was driven by three factors.

1. Since the researchers would be giving infants in the control group an injection and taking blood samples, by using another vaccine, they could learn information about that vaccine during the study. 
2. The infants in the control group would get the benefit of gaining immunity to another potentially severe disease. 
3. As the meningococcal vaccine would not impact rates of pneumococcal disease, investigators could still accomplish the study goal of measuring the vaccine’s effectiveness. 

In this manner, the study design conserved resources, gave the babies an additional benefit for participation, and answered the important questions about PCV7 safety and effectiveness. 

Using an earlier version of the vaccine: Later pneumococcal conjugate vaccine trials

The past two decades have seen advances in the development of conjugate pneumococcal vaccines, and research has built on the initial PCV7 trials. After PCV7, which protected against seven types of pneumococcus, vaccines were changed to offer protection against more types of pneumococcus: 13 (PCV13), 15 (PCV15), 20 (PCV20), and 21 (PCV21) types. With each new version, clinical studies have been based on measuring the immune response to determine whether the vaccine could be licensed. This approach was the most appropriate choice for two reasons:

1. Similar to pertussis vaccines, since there was already a pneumococcal vaccine available, researchers had an ethical obligation to protect study participants from pneumococcus. 
2. Because a vaccine was available, rates of infection were low, so comparing rates of infection would have required large study groups and a long time to complete the trials. While this may not seem problematic, it’s also important to consider that since the newer vaccines protected against more types, they would be expected to offer better protection. As such, a long study would potentially mean children getting illnesses that could have been prevented. To address these concerns, study designs in this type of situation often compare immune responses between participant groups, allowing for smaller, more affordable and shorter studies. These are often referred to as “immunobridging studies.” Importantly, for immunobridging studies to be an option, researchers must be able to reliably measure part of the immune response and compare the measurements across study groups.

The bottom line for families

Vaccine trials are not one-size-fits-all. Each study is carefully designed to answer specific questions while prioritizing participant safety and ensuring participants are treated ethically. Importantly, all approved vaccines are supported by extensive evidence on both safety and effectiveness — gathered using the most appropriate and sound scientific methods available.

Beyond considering clinical trial studies designs, families can also take comfort in three additional points:

1. The manufacturers do not design or carry out these studies in a vacuum. FDA staff scientists, institutional review boards, and data monitoring committees are all reviewing and monitoring the studies.
2. Clinical trials are not the only source of information used to inform policies, such as vaccine recommendations. Other types of studies, related to both the pathogen and the vaccine, are also analyzed.
3. If a vaccine is being used in multiple countries, studies and experiences in those countries provide additional information — and additional oversight bodies. 

Understanding how and why clinical studies are designed in the way they are, as well as considering the larger context of information that’s available, can support confident and informed decision making.

Related content

• Technically Speaking: 75 Years of Placebo-Controlled Vaccine Testing in the U.S.
• Vaccine Science: Process of Vaccine Development
• The Persistent Misleading Claim That Vaccines Aren’t Properly Tested for Safety

Contributed by: Lori Handy, MD, MSCE , Charlotte A. Moser, MS, Paul A. Offit, MD