Reanalysis of data from ten years of koala chlamydia vaccine trials reveals inherent bias and a lack of protective efficacy

Koalas, the iconic marsupials of Australia, are facing a perfect storm of threats. Their forests are shrinking, the climate is shifting, and a pervasive bacterial infection known as chlamydia is spreading through their populations. This disease is not merely a nuisance; it causes blindness, kidney failure, and reproductive damage that leads to infertility and death. For years, conservationists have hoped that a vaccine could be the silver bullet to save these animals, offering a way to halt the decline without needing to solve the complex political and environmental issues of habitat loss. The idea is straightforward: if scientists can protect koalas from this specific infection, they can give the species a fighting chance to recover.

In late 2025, a vaccine named Klavax received a permit for limited use in Australia, hailed as a major breakthrough. The decision to approve it was based on a highly cited study that claimed the vaccine reduced deaths from chlamydia by 64 percent. However, a new, independent analysis of the data behind that study suggests the approval was granted on a foundation of errors. Two researchers from the University of Queensland, Keith Chappell and Michaela Blyton, re-examined the original numbers and found that the way the data was organized and analyzed had created a false picture of success. Their work indicates that the vaccine, as currently tested, does not actually protect koalas from the disease or from dying from it.

The story begins with how the original study was put together. The researchers who approved the vaccine had combined data from ten years of different trials, involving hundreds of koalas. They compared the health of vaccinated animals against those that were not. On the surface, the vaccinated group looked much healthier. But when Chappell and Blyton looked closer at the raw records, they found a fundamental mistake in how the information was coded. In the original analysis, the time a koala spent alive and healthy before it ever received a shot was incorrectly counted as time spent after vaccination.

Imagine a student who is given a test, fails, and then is given a tutor. If you later looked at their grades and counted the time they spent studying before the tutor arrived as time they spent with the tutor, you might wrongly conclude the tutor was responsible for their earlier good grades. Similarly, the original study counted the years koalas survived naturally before they were vaccinated as if the vaccine had saved them. Since the animals had to survive to get the shot, this automatically made the vaccinated group look like they had a higher survival rate, simply because the clock started ticking for them at the wrong moment. When the researchers corrected this coding error, the apparent survival advantage vanished.

There was a second, more subtle problem with the study design. The original analysis included koalas that were already very sick when they were first found by researchers. These sick animals were almost never vaccinated, likely because they were too ill to be treated or were humanely euthanized. Instead, they were placed in the control group—the group meant to represent the "unvaccinated" population. This created a skewed comparison. The control group was filled with animals that were already on the path to death or severe illness, while the vaccinated group was made up of animals that were healthy enough to receive the shot.

This is like comparing the health of a group of people who just walked into a hospital with a broken leg against a group of people who were healthy enough to go for a jog. If you then give the joggers a vitamin and find they stay healthy while the hospital patients get worse, you cannot say the vitamin worked; you are just comparing two very different starting points. In the koala study, the control group was nearly six times more likely to have the disease at the very start of the record than the vaccinated group. This massive difference in starting health made the vaccine look effective when it was actually just a victim of bad grouping.

To get a true answer, the researchers stripped the data down to its core. They removed every koala that was sick at the first moment they were seen, ensuring that both the vaccinated and unvaccinated groups started the comparison in the same state of health. They also removed the time periods before vaccination to ensure no one was being credited for survival they hadn't earned yet. When they ran the numbers again on this clean, fair dataset, the results were starkly different. There was no significant difference in how many vaccinated koalas got sick compared to the unvaccinated ones. About 12.6 percent of the vaccinated group developed the disease, while 14.4 percent of the unvaccinated group did. Similarly, the death rates were nearly identical, with 2.1 percent of vaccinated animals dying from the disease compared to 2.5 percent of the unvaccinated.

Perhaps most surprisingly, the data suggested that vaccinated koalas might actually be less likely to recover from the disease once they caught it. In the corrected group, only a small fraction of vaccinated animals cleared the infection by their last check-up, whereas a larger portion of the unvaccinated animals did. While the researchers caution that this could be due to other factors, such as age or when veterinary care was given, it certainly does not support the idea that the vaccine is helping the animals fight off the infection.

The findings of this reanalysis stand in direct contrast to the earlier, optimistic study that led to the vaccine's approval. They align instead with a different, more rigorous study that used a randomized, blinded design—where neither the researchers nor the animals knew who got the real vaccine and who got a fake one. That earlier, stricter study had found no protective effect at all. The new analysis confirms that the positive results from the approved vaccine were likely an illusion created by statistical errors and biased grouping, rather than a real biological benefit.

The implications for koala conservation are significant. The authors argue that before any vaccine is rolled out to protect an endangered species, it must be tested with the highest standards of scientific rigor. The current data does not support the widespread use of this vaccine to prevent chlamydia in wild koala populations. Instead, the researchers suggest that efforts should focus on developing a better vaccine formulation and testing it in a way that eliminates these sources of error. The goal remains the same: to save the koala from a devastating disease. But the path to that goal requires clear eyes and honest numbers, ensuring that the tools used to help these animals are truly effective.