Double Duty: Can the meningitis vaccine also protect against gonorrhea?

episode 37 GoGoVax
[Regina] (0:00 - 0:15)
The typical one is trending towards statistical significance. But my favorite, can I just go ahead and say it? Oh, please, yes.

Flirting with statistical significance. Flirting. Our results were flirting.

[Kristin]
Oh my gosh, that is great.

[Regina]
And we're like batting their eyelashes and sending little kisses.

[Kristin] (0:15 - 0:34)
Did somebody really say that in a paper? Flirting with statistical significance.

Welcome to Normal Curves.

This is a podcast for anyone who wants to learn about scientific studies and the statistics behind them. I'm Kristin Sainani. I'm a professor at Stanford University.

[Regina] (0:35 - 0:40)
And I'm Regina Nuzzo. I'm a professor at Gallaudet University and part-time lecturer at Stanford.

[Kristin] (0:41 - 0:46)
We are not medical doctors. We are PhDs. So nothing in this podcast should be construed as medical advice.

[Regina] (0:46 - 0:51)
Also, this podcast is separate from our day jobs at Stanford and Gallaudet University.

[Kristin] (0:51 - 1:09)
Regina, we haven't had an episode about sex in a while. So believe it or not, I picked a study that's related to sex.

[Regina]
Wow, you did?

Yes. I'm so proud of you.

[Kristin]
Aren't you though?

Okay, to be fair, it's actually about a sexually transmitted disease. So maybe not as fun as you were hoping for.

[Regina] (1:09 - 1:19)
Okay, I'm going to give you half credit for that one, Kristin. And how about this? I'll try to make jokes about how sex is still indeed fun, even though it can give you germs.

Yes.

[Kristin] (1:20 - 1:33)
So there was a really good study out in the New England Journal of Medicine recently that involved the sexually transmitted disease gonorrhea. And I'm not just picking it for the sex, but it illustrates some great statistical points.

[Regina] (1:33 - 1:42)
Gonorrhea. I feel like gonorrhea is one of those old-fashioned things that, you know, the sailors pick up when they come home from war. Am I thinking about the right disease here?

[Kristin] (1:43 - 1:44)
That sounds right to me.

[Regina] (1:44 - 1:48)
Okay, gonorrhea. It's such a fun word. Okay, what is the study about?

[Kristin] (1:48 - 2:19)
So this is where it actually gets pretty interesting. There is no vaccine for gonorrhea, but there is a vaccine for a different disease that researchers believe may incidentally help reduce the risk of gonorrhea. And it's a specific vaccine against the meningococcal bacterium, which causes meningitis.

And that bacterium happens to be a close cousin of the gonorrhea bacterium, so scientists think that the vaccine may have some cross-protective effects. And there is actually a lot of observational data to back this up, and there are now a few randomized trials.

[Regina] (2:20 - 2:26)
Oh, this is fascinating. Okay, and this is an intervention you can actually randomize. You can give vaccines.

[Kristin] (2:27 - 2:47)
Exactly. The specific vaccine that's being studied is not one that everyone routinely receives. So there are enough adults who have never been vaccinated that you can actually go out and conduct a randomized trial.

And today we're mostly going to focus on a trial called the GOGOVAX study, and results of that trial were published in July of 2026 in the New England Journal.

[Regina] (2:48 - 2:58)
GOGOVAX? Is that really the name of it? That is the name of the study, yes.

I'm picturing like go-go boots. Is it like white leather boots and the little Nancy Sinatra kind of thing?

[Kristin] (2:58 - 3:05)
Every randomized trial needs a really good acronym, and I feel like they were trying to make it really upbeat, so we didn't focus on the fact that it's about gonorrhea.

[Regina] (3:06 - 3:13)
GOGOVAX against gonorrhea! GOGO gonorrhea. Going gone.

Gone.

[Kristin] (3:13 - 3:38)
Gonorrhea gone, yes. Maybe it's the gonorrhea gone vax trial. Maybe that's what it's short for.

Maybe that's what it's short for. Okay, GOGOVAX. So GOGOVAX is the first randomized double-blind placebo-controlled trial that we have results from on this topic.

There are two other trials that have also released results, but they were smaller and they did not use a placebo, and so that's why we're going to focus on GOGOVAX today.

[Regina] (3:38 - 3:39)
Yeah, we need a placebo.

[Kristin] (3:39 - 4:02)
Yeah, placebo is important. And the claim we're going to evaluate is the hypothesis of the trial, which is that a vaccine against meningococcal disease also protects against gonorrhea. And we'll tackle some interesting statistics along the way, the limitations of observational studies, why statisticians hate the term statistical trend, and the difference between per-protocol and intention-to-treat analyses.

[Regina] (4:03 - 4:04)
GOGOVAX. And the stats.

[Kristin] (4:06 - 4:07)
GOGO stats!

[Regina]
GOGO stats!

[Kristin] (4:10 - 4:24)
Okay, so let me start with some biology. There's a bacterium called meningococcus. It can cause meningitis, which is an infection of the lining around the brain and spinal cord.

Have you ever heard of it, Regina?

[Regina] (4:24 - 4:34)
I have, actually, because I think I'm at heightened risk for this with my cochlear implant. Oh, interesting! Or maybe if I get it, I'm more likely to die.

Something bad.

[Kristin] (4:34 - 4:40)
That makes sense, because you've got an implant in that area, and so if you get bacteria in it, you might be more susceptible.

[Regina] (4:40 - 4:44)
That's interesting. Does that mean I'm more susceptible to gonorrhea?

[Kristin] (4:45 - 5:17)
I don't think so. Good. Interestingly enough, different area of the body.

It is a pretty serious disease, meningitis, though. It can progress very quickly, and it can be life-threatening. And it tends to spread through close contact, like living in the same household.

And this is why outbreaks can occur in places like college dorms. So it turns out that a lot of teenagers get the meningococcal vaccines right before they go to college. And in fact, my son just got two meningococcal vaccines because he just started college.

[Regina] (5:17 - 5:19)
Oh, that's right. He's a banana slug now, isn't he?

[Kristin] (5:20 - 5:22)
He is. He joins us as banana slugs, Regina.

[Regina] (5:22 - 5:28)
Yes, UC Santa Cruz. Just to explain what banana slug means for people who are not in the know.

[Kristin] (5:28 - 5:50)
We both did the science writing program there, and it was wonderful. Amazing campus. It was.

So meningococcus bacterium comes in several different flavors called serogroups, and they are identified by letters. We don't make these very exciting. Serogroup A, B, C, W, and Y.

And the vaccine that we're talking about today specifically protects against serogroup B, which is also known as meningococcal B or MenB.

[Regina] (5:50 - 6:01)
Do they really call it MenB? MenB is the shorthand for it, yeah. MenB cool.

It's relevant. MenB hot. MenB sexy.

Okay. This is MenB vaccine.

[Kristin] (6:02 - 6:14)
There are actually two different MenB vaccines used in the United States, and the study we're going to talk about today looked at only one of those called Bexsero. That's the brand name. The technical name is the 4C MenB vaccine.

[Regina] (6:15 - 6:19)
Okay. 4C MenB vaccine. Why that one?

[Kristin] (6:19 - 6:55)
So out of these two MenB vaccines available, Bexsero is the one that they think might protect against gonorrhea because that vaccine contains several proteins, several antigens that are found on the surface of the meningococcus bacterium. And that makes your body make antibodies to those proteins. And it turns out that some of those same proteins are found on the surface of gonorrhea.

Again, because the bugs are close cousins. So it's totally plausible to think that the antibodies that you generate against MenB might also recognize gonorrhea. And in fact, there is laboratory and animal data to show that the antibodies are indeed cross-reactive.

[Regina] (6:56 - 7:17)
Oh, that's interesting. So it's that we're training our immune system to recognize a particular type of bacterium, and then it gets smarter. And so it's like, oh, okay, you're not exactly the same, but you look like this other thing.

So I'm going to protect against you too. That's the idea. So that is the vaccine.

But let's get back to the sex. Tell me about gonorrhea.

I want to hear some good things.

[Kristin] (7:17 - 7:45)
All right, well, a lot of good things here. But gonorrhea is a sexually transmitted bacterial infection.

You transmit it through contact between mucous membranes. And it doesn't always cause symptoms. But if it's left untreated, it can cause pelvic inflammatory disease, infertility, and pelvic pain.

It can also increase the risk of transmitting HIV. And it can be passed from a mother to her baby during childbirth, which is bad for the baby.

[Regina] (7:45 - 8:02)
Okay, transmitted through contact between mucous membranes. So you need a condom. Yes, exactly.

So condoms are protective. Condoms are protective. And this is not like, what was the one that we decided could be passed through kissing?

Oh, HPV.

[Kristin] (8:02 - 8:07)
Yes, right. We had an episode on the HPV vaccine. I don't think this one is passed through kissing.

[Regina] (8:07 - 8:15)
Okay, wear condoms, people. Yes. Wear condoms.

Okay, and I feel like that's part of the routine testing that we all get if we're responsible.

[Kristin] (8:16 - 8:26)
Yes, it is part of the routine panel that's normally tested for. And that is a way to help prevent spread is to get regularly tested because you can treat it with antibiotics.

[Regina] (8:27 - 8:33)
Can we have a brief moment here? STD versus STI, what is the difference?

[Kristin] (8:33 - 8:52)
That is a great question.

I kind of thought, well, infection implies an infection, like you're infected with something, disease, D implies a disease. So maybe you can get an infection without having any disease. That is, you might be asymptomatic.

So like we mentioned HPV, HPV infections a lot of times have no symptoms or anything, right? But it can lead to cancer.

[Regina] (8:52 - 8:54)
It can lead to cancer. Which is a disease.

[Kristin] (8:54 - 8:56)
So it's an infection when it starts and a disease later.

[Regina] (8:57 - 9:04)
Does this make sense? I feel like when I was coming of age, they talked about STDs, everything was that. And then recently I've been hearing STIs.

[Kristin] (9:04 - 9:12)
I think it might be like they're trying to de-stigmatize. I'm just guessing that that's why that term is used sometimes is to try to make it sound less bad.

[Regina] (9:13 - 9:15)
Okay. But it can be treated with antibiotics.

[Kristin] (9:15 - 9:41)
It can be treated, which is why they test regularly so that you can get it treated. And then you don't have these bad effects. You don't pass it on to other people.

Exactly. One thing that's really important to understand about gonorrhea today is that you can get it more than once. So unlike some other bugs that you get where you can only get them once and then your body mounts immunity after that, like chickenpox you get once, right?

With gonorrhea, it doesn't work that way. So you can get reinfected many times and your body doesn't build an immunity to it.

[Regina] (9:42 - 9:44)
Oh, that makes it a little bit scarier.

[Kristin] (9:44 - 9:50)
Yeah. What we're going to see in this study that there's a lot of reinfection going on. So I just wanted to point that out.

[Regina] (9:50 - 9:55)
Do we know anything about how prevalent this is in the population? I should have looked that up, Regina.

[Kristin] (9:55 - 10:13)
Okay. I'm going to tell you though, it's surprisingly prevalent in the population they studied. So we're going to talk about that.

I don't know how prevalent it is in general, and it probably depends a lot on who you're asking about. So maybe certain populations have very high rates, but then other populations have very low rates is my guess.

[Regina] (10:14 - 10:21)
So maybe during the break, I will quickly go and Google men of dateable age for Regina. How prevalent is it?

[Kristin] (10:21 - 10:25)
I'm guessing that at our age, it's not quite as prevalent, but you never know.

[Regina] (10:25 - 10:25)
You never know.

[Kristin] (10:25 - 10:30)
We might be dating. Yes. Okay.

But yeah, good question, Regina. We should find this out.

[Regina] (10:30 - 10:36)
All right. So how did people even make this link between this MenB vaccine and gonorrhea?

[Kristin] (10:37 - 11:33)
So I think biologically, people were aware that these were related, and then they started to get some clues from what we call ecologic studies. So these are studies where you're just looking at trends in the population, not at individuals. And one of the earliest ones came from Cuba.

So after Cuba introduced a nationwide MenB vaccination campaign, researchers noticed that gonorrhea rates fell, but the rates of other sexually transmitted infections did not. Fascinating. So it's like you can't conclude causality from that, right?

I mean, two things can happen at the same time or close in time and not be related at all. It could just be two trends happening at the same time coincidentally. So it started with some ecological studies like this, but then people followed up with stronger types of observational studies where you actually look at individual people, not at just trends over the population.

There were several of these that were published. One of the most interesting ones comes from Australia. They used a case control study.

[Regina] (11:33 - 11:58)
I love case control studies. So this is where I almost think about it as going and reversing time, right? So you find people who have the disease and people who don't, and then you go back and look at what kind of exposures they had or what kind of risk factors.

So here I'm guessing they looked at people who had gonorrhea and then a group of similar people who did not, and then they went back and asked about whether they had the MenB vaccine. Perfect.

[Kristin] (11:58 - 12:13)
Exactly. That's exactly it, Regina. And they actually used kind of a clever design.

The cases were people diagnosed with gonorrhea, but the controls, this is interesting. They were people who didn't have gonorrhea, but instead had chlamydia, which is a different sexually transmitted disease.

[Regina] (12:13 - 12:27)
Right. But this way we're not just saying, okay, the type of person who would get gonorrhea and then the churchgoing people over here that, you know, are still virgins or whatever. No, we're getting similar types of people.

The only difference being gonorrhea or not. Yes, exactly.

[Kristin] (12:28 - 12:45)
That's the idea. If you pick a control group that's nothing like the case group, well, maybe your control group, yeah, if they're virgins, they have no way of having gotten gonorrhea, but also they may be more health conscious and have actively gone out and sought a MenB vaccine, right, because they're health conscious. So you don't want these systematic differences between cases.

[Regina] (12:45 - 12:45)
Right, right.

[Kristin] (12:45 - 13:23)
And they did find that the cases here were significantly less likely to have gotten the Bexsero vaccine in the past than the chlamydia controls. Oh, interesting. Okay.

So there was an association and a bunch of other studies from other countries have also found similar results. And that leads us up to 2024 when researchers did a meta-analysis to pool all of these observational studies. And the pooled estimate came out to a one-third, a 33% reduction in the risk of gonorrhea in vaccinated versus unvaccinated individuals.

So observational data still, but, you know, it seems pretty consistent and pretty robust.

[Regina] (13:23 - 13:24)
33% is not nothing.

[Kristin] (13:24 - 13:28)
No, especially for like something you're getting as a side effect of another vaccine.

[Regina] (13:28 - 13:31)
Right. Okay. But this was just observational.

[Kristin] (13:31 - 14:16)
Yeah. So we always have to worry about these are all observational studies and observational studies, as we've talked about on this podcast, can be affected by confounding because people who bother to go out and get this vaccine may be exactly the kind of people who are also more careful about having protected sex or having less sex or getting tested regularly, right? Or maybe they're people who went to college, right?

And college goers might be different than non-college goers. So we can try to, as we've talked about on this podcast, statistically adjust for confounders, but statistical adjustment is not perfect. We're always worried about unmeasured or residual confounding, and that's hard, as we've talked about on this podcast, to get rid of.

So that's why scientists have now followed up on these observational studies with randomized trials.

[Regina] (14:16 - 14:43)
Randomized trials. I love how this is the typical progression, right? Starting with these, like, hmm, these ecological, I noticed this, and then let's move on to increasingly more sophisticated observational studies, and then we end up with randomized trials, which are expensive to do, which is why we don't start with them right away.

We don't start with those until there is a reason to fund that. Okay. So you mentioned, I think, that there were a few trials that came out before this GoGoVax one.

[Kristin] (14:44 - 14:49)
Yeah. There were two studies, randomized trials, that were published earlier than GoGoVax.

[Regina] (14:49 - 14:49)
Okay.

[Kristin] (14:49 - 15:06)
One was a very small, 130-person randomized trial called MENGO. You are making all of these up. I am not making up the acronyms.

Every randomized trial has a cool acronym. MENGOGO. Yeah, MEN is just coincidental with MEN.

It's not MEN. It's meningococcal.

[Regina] (15:06 - 15:12)
There are no coincidences in life, Kristin. MENGO, MENGOGO, MEN...

[Kristin] (15:12 - 15:29)
Well, I guess GO maybe is for gonorrhea. MEN... meningococcal gonorrhea.

MENGO. Okay, anyway. So that trial, very small, results were out in April of 2026.

But it was an open-label trial. So they did randomize people to get the vaccine or not, but there was no placebo. So people knew what they were getting.

[Regina] (15:29 - 15:39)
Okay, so that can be a little bit of a problem. Because if you know that you're getting the vaccine, maybe that's changing your behavior a little bit. Maybe it's just going to be a little more risky.

Is that the idea?

[Kristin] (15:39 - 15:48)
Yeah, you might think I'm protected, and then you might be like, okay, I don't have to use a condom. And so then maybe the vaccine doesn't show any protective effects because you've engaged in riskier behavior than the control group, right?

[Regina] (15:49 - 15:53)
Right, right. So that's a weakness. It's a weakness, yes.

What'd they find?

[Kristin] (15:53 - 16:05)
They found a 22% reduction in the rate of gonorrhea in vaccinated versus unvaccinated people. But it was not statistically significant or, as you like to say, statistically discernible.

[Regina] (16:06 - 16:16)
Thank you for including that one in there. Okay, so before we had 33% reduction in that observation. Now we're doing at least the beginning of a randomized trial, and we're getting 22%.

[Kristin] (16:16 - 16:17)
Yeah, shrunk.

[Regina] (16:18 - 16:21)
Not statistically discernible. But it was a small trial.

[Kristin] (16:22 - 16:23)
130 people, very small.

[Regina] (16:23 - 16:28)
So maybe we just didn't have enough people to see the difference? Could be, right. Okay, what about the second one?

[Kristin] (16:29 - 16:58)
So the second one was a trial in France called the DOXYVAC study. You are making all of these up. It was like a vacuum.

No, that's just what they named it. Some easy to say acronym. That one had 545 people, so bigger.

Results were released in 2024. Now it used what we call a factorial design. So the researchers were testing two different interventions at the same time, antibiotics and the vaccine.

It was also open label. So again, the control group got nothing and everybody knew what they got.

[Regina] (16:58 - 17:00)
Okay, and what did they find here?

[Kristin] (17:00 - 17:30)
This study also found a null result. Interestingly, the vaccine group had again a 22% lower rate of gonorrhea. That's coincidental.

It was not statistically significant, although the p-value was 0.06. And Regina, I just want to make a note here that the authors were good and they get a pat on the back because they didn't try to say something like there was a trend toward significance, which is something that authors sometimes do when they get a p-value that's just above that 0.05 threshold.

[Regina] (17:30 - 17:52)
I hate this. So this is an object of fascination of mine, but also like a soapboxy rant. Statisticians hate this.

It's something about that weird gray zone between 0.06 and 0.10 that's just so tempting for us to go in and say, oh, but it was so close to 0.05.

[Kristin] (17:52 - 17:55)
The researchers feel like, oh, I just missed it. If I just reached a little farther, I would have grabbed it, you know?

[Regina] (17:55 - 18:00)
They can't resist. And then they end up putting these crazy things in their publications, right?

[Kristin] (18:01 - 18:02)
Oh, yes, yes.

[Regina] (18:02 - 18:05)
Because they cannot actually say statistically significant.

[Kristin] (18:05 - 18:05)
Right.

[Regina] (18:05 - 18:09)
So they start adding in all these adjectives and adverbs to kind of wriggle around.

[Kristin] (18:09 - 18:34)
The most common one, they say, is there's a trend towards significance. The problem with that is trends suggest direction, like the p-value is speeding down the road, heading somewhere. Yeah.

It makes it sound like if we just collected a few more subjects, the p-value probably would have made it to below 0.05. And the p-value is not moving anywhere. It's not moving towards anything. It is just, you miss significance, done.

[Regina] (18:35 - 18:43)
Either you have 0.05 or you do not have it. It's kind of like pregnancy or gonorrhea. Either you have it or you don't.

[Kristin] (18:43 - 18:44)
You can't be trending towards pregnancy.

[Regina] (18:44 - 18:48)
No, you cannot. OK, so they did this or they did not do this?

[Kristin] (18:48 - 18:53)
They did not do this. It was p equals 0.06, and they resisted the temptation.

[Regina] (18:53 - 19:11)
So my favorite thing is many years ago, a guy named Matt Hankins, he published a blog post, and he had gone through the literature to find out all the crazy things that people were saying when they were publishing something that was greater than 0.05, but they still wanted to count it somehow.

[Kristin] (19:11 - 19:19)
I love this website. We're going to put the link in the show notes. But can you also read a few of these for us, Regina, just because they're a riot.

[Regina] (19:19 - 19:30)
Well, I mean, the typical one is trending towards statistical significance. But my favorite, can I just go ahead and say it? Oh, please, yes.

Flirting with statistical significance. Flirting. Our results were flirting.

[Kristin] (19:30 - 19:32)
Oh, my gosh, that is great.

[Regina] (19:32 - 19:35)
They were like batting their eyelashes and sending little kisses.

[Kristin] (19:35 - 19:44)
Did somebody really say that in a paper? And flirting with statistical... Wow.

I mean, that brings the sex back into the episode, though. I feel like that's much more positive than the sexually transmitted disease.

[Regina] (19:44 - 19:56)
I know, I know. I feel like there are more on here that are sexually suggestive. Oh, can you give us a few?

Let me find it. I have it open right now. Suggestive of statistical significance.

Suggestive.

[Kristin] (19:56 - 19:57)
Suggestive.

[Regina] (19:57 - 20:02)
I'm picturing a little striptease. I'm picturing a striptease, like, you know, like, ba-da-bum.

[Kristin] (20:02 - 20:04)
I'm going to read a few of my favorites. Oh, a slight slide towards significance. P less than 0.20.

[Regina] (20:05 - 20:33)
Oh, that one's fun. Approached but did not reach orgasm.

No, wait. Approached but did not reach significance. Sorry, I couldn't help myself.

You can edit that out.

[Kristin]
No, I might keep it.

[Regina]
A little significant.

Tantalizingly close to significance. So close I could just reach out and taste it and touch it.

[Kristin] (20:33 - 20:35)
I think that's my favorite.

[Regina] (20:35 - 20:38)
Oh, here's one. Just tottering on the brink of significance.

[Kristin] (20:39 - 20:51)
That's actually from a published paper. I love this. Barely escapes.

Sorry, I'm trying to get through that one. Barely escapes being statistically significant at the 5% risk level.

[Regina] (20:52 - 20:54)
Oh, no. Arguably significant. You can argue anything.

[Kristin] (20:55 - 21:10)
Just very slightly missed the significance level. P equals 0.086. Leaning towards significance. P equals 0.15. Hovered around significance. P equals 0.061. A hint of significance.

[Regina] (21:10 - 21:13)
A hint. Oh, baby, it's a taste. Vaguely significant.

[Kristin] (21:14 - 21:20)
On the cusp of significance. Yours are the best. Narrowly evaded statistical significance.

[Regina] (21:20 - 21:24)
Very closely brushed the limit of statistical significance.

[Kristin] (21:24 - 21:25)
You're getting all the poetic ones.

[Regina] (21:25 - 21:35)
Well, I'm finding all the ones that seem like they belong in a romance novel. I think so. Has anyone done like a 50 shades of gray for statisticians?

[Kristin] (21:35 - 21:43)
On the very fringes of significance. P equals 0.099.

[Regina] (21:44 - 21:51)
OK, I'm going to make a novel. 50 shades of P values. Tickling someone's bare abdomen with a feather of significance!

[Kristin] (21:52 - 22:11)
P equals 0.057. OK, so that was a fun statistical detour on why statisticians hate this idea of a statistical trend and you should never do it. Back to gonorrhea now. I feel like we didn't lose the sex thread too much in that little statistical detour though.

[Regina] (22:11 - 22:11)
We did not.

[Kristin] (22:12 - 22:12)
Yeah.

[Regina] (22:12 - 22:14)
I made sure that we were keeping that one going.

[Kristin] (22:14 - 22:16)
Back to gonorrhea.

[Regina] (22:16 - 22:17)
Gonorrhea.

[Kristin] (22:17 - 22:40)
Let's summarize the background so far. So we have lots of observational data showing an effect of this vaccine on gonorrhea. We have two smaller open label trials that have null results.

So we cannot conclude from those trials that the vaccine does not work. We just can conclude that we don't yet have evidence from randomized trials that there's an effect here. And that brings us up to now the GOGOVAX trial.

GOGOVAX.

[Regina] (22:40 - 23:34)
All right, let's take a short break first.

Welcome back to Normal Curves. Today we're looking at a randomized clinical trial that asked whether a specific vaccine against MenB can also reduce the risk of gonorrhea.

And Kristin, you were about to tell us about the GOGOVAX trial. But I did promise over the break to go check out incidence and prevalence because I'm curious like how big this is. OK.

It looks like in the U.S. in recent years, it's between one and two per 1,000 people every year gets reported a diagnosis with gonorrhea.

[Kristin] (23:34 - 23:48)
OK, so that's not like terribly common. That's good.

[Regina]
OK, but you were going to tell us about GOGOVAX.

[Kristin]
Right. So this was a randomized, double-blind, placebo-controlled trial. And I thought it might be fun actually in talking about the study design here to review PICOT.

[Regina] (23:48 - 24:05)
PICOT. I love PICOT. It's a mnemonic for remembering the elements of study design.

And we use this in our holiday guide episodes. And PICOT, P-I-C-O-T. So P is for population.

Who was in the population study group here?

[Kristin] (24:05 - 24:30)
So the study enrolled 654 men who have sex with men ages 18 to 50, so relatively young. This was in Australia. And everyone had to have recently had either gonorrhea or syphilis.

So this was definitely a high-risk population. And of course, they also could not have previously received the Bexsero vaccine. So Regina, let's stop and talk for a minute about why they would focus on such a high-risk population.

Do you want to explain that?

[Regina] (24:30 - 25:02)
Yeah, I think it gets back to what I just said about the incidence, right? If you're only getting one or two per 1,000 in the general population, and these clinical trials are expensive to do, then, OK, you got 1,000 people in each group. Maybe you don't have anyone who gets gonorrhea at the end.

And then you're just kind of wasted all that time. So if you focus on a high-risk group, then you are more likely to actually get the outcome where you can see it. And you had said that you can get gonorrhea, you can get reinfected with gonorrhea over and over again.

[Kristin] (25:02 - 25:12)
That's why they chose people who had already had it, because you can get reinfected. Yeah, you're more likely to see enough outcomes. And you don't want to have to have a trial with 50,000 people in each arm where that's going to be much more expensive.

[Regina] (25:12 - 25:15)
OK, so that was P, PICOT, I, intervention.

[Kristin] (25:16 - 25:28)
Yeah, so participants were randomly assigned to the intervention or placebo, 327 per group. The intervention group got two doses of the Bexsero vaccine given three months apart, because this is a vaccine that requires two doses.

[Regina] (25:29 - 25:31)
OK, and then C, control, was placebo.

[Kristin] (25:31 - 25:36)
Yeah, the control group, the placebo group, got a saline shot, salt water.

[Regina] (25:36 - 25:43)
So this was not open-label. They were blinded to knowing whether they had it or not. And you said it was double-blinded.

So the investigators were blinded.

[Kristin] (25:43 - 25:48)
That's right. Participants didn't know what they were getting. Investigators didn't know who was getting the real vaccine.

[Regina] (25:48 - 25:52)
OK, PICOT, 0 outcomes, what did they look at? Whether you got gonorrhea or not?

[Kristin] (25:52 - 26:04)
Yeah, the primary outcome actually was the first laboratory-confirmed gonorrhea infection. But it had to occur at least four weeks after the second vaccine dose, because they wanted to give enough buffer for immunity to develop.

[Regina] (26:04 - 26:08)
Right, that's pretty typical for vaccine studies to give you a little time. You said first.

[Kristin] (26:09 - 26:17)
Yeah, this was a time-to-event outcome, time to the first gonorrheal infection, because some people had multiple infections, but they only wanted to look at the first one.

[Regina] (26:17 - 26:25)
OK, so that means they were coming in for testing, I guess, which gets us to the T, timing. Did they test them frequently?

[Kristin] (26:25 - 26:34)
Yes, they had to test them frequently. So the participants were followed for 24 months, and they got testing for STIs every three months for those two years.

[Regina] (26:35 - 26:41)
OK, so that is pretty frequent. So that was the backing up to the outcome. That was the primary outcome.

Did they have secondary outcomes?

[Kristin] (26:41 - 26:52)
They did have secondary outcomes. The secondary outcomes looked at symptomatic versus asymptomatic infections, because not everybody has symptoms, infections at different body sites, recurrent infections if people had multiple infections, and safety.

[Regina] (26:52 - 26:58)
All right, that sounds pretty well thought out. So you said time-to-event. They're doing survival analysis.

[Kristin] (26:58 - 27:17)
Survival analysis, yes. So let's talk about how they analyzed the results. They used survival analysis.

So that means they were comparing the rate of infection in the two groups. But before we get to that, Regina, I want to talk about who they analyzed from the dataset.

[Regina]
Wasn't it everyone?

[Kristin]
So it should be.

[Regina] (27:17 - 27:17)
It should be.

[Kristin] (27:17 - 27:28)
Usually should be. But can we do a little statistical detour on the difference between intention-to-treat versus modified intention-to-treat versus per-protocol analyses in randomized trials?

[Regina] (27:28 - 27:33)
Yeah, these are so confusing, right, because they all kind of sound alike. But they're very important. They are important.

[Kristin] (27:33 - 27:47)
So intention-to-treat is the standard analysis that we use in randomized trials, because in intention-to-treat, you analyze everyone. Everyone is analyzed according to the group they were randomized to, even if they didn't do anything in the study, if they never followed the protocol.

[Regina] (27:47 - 27:53)
Once randomized, always analyzed. So even if they didn't get the vaccine.

[Kristin] (27:53 - 28:15)
Yeah, or if they were later found to be ineligible for the study. And that might seem a little counterintuitive, right? Why would you include them if they weren't eligible and they didn't get the vaccine, right?

But the whole point of this is to preserve the benefits of randomization. Once you've randomized people, that gives us all of the balance. And we don't want to lose that by allowing people to then be taken out of either of the groups.

[Regina] (28:15 - 28:20)
By balance, you're just talking about this is the whole point of why we do randomized trials, because before we had this idea of confounding.

[Kristin] (28:20 - 28:21)
Yes, yes.

[Regina] (28:21 - 28:42)
So it's like we're following the people who chose to get a vaccine versus those who didn't. And we want to eliminate that part. And also, often they use this to test what it would look like in the real world, right?

So it's like, OK, we're telling people they should get the vaccine, whether they actually do or not. That's important. We want to see what it looks like in practicality.

[Kristin] (28:43 - 29:10)
Exactly, yes. The real world effectiveness as opposed to the efficacy in perfect conditions. Now, that's intention to treat, but sometimes people bend the rules a little.

So maybe they realize after randomization that someone didn't meet those eligibility criteria, or someone literally never comes back after they sign the consent form, so they have no follow-up data. So researchers sometimes justify doing a modified intention to treat, which means they're dropping a handful of people for reasons that they think are legitimate.

[Regina] (29:11 - 29:12)
That is a little squishy.

[Kristin] (29:12 - 29:13)
It is so squishy.

[Regina] (29:14 - 29:21)
You need to pre-specify that. It's not like you can just look at the data afterwards and say, you know what, we're just going to get rid of these few over here.

[Kristin] (29:21 - 29:44)
Yeah, this can be a little dicey, because you absolutely need to pre-specify how you're going to modify your intention to treat if you're going to do it. People don't always do that. I think this is sometimes done after the fact for practical reasons.

And the bad thing is there is no standard definition of modified intention to treat. You can basically mean anything by that. So that can muck up a result if you're not careful.

[Regina] (29:44 - 29:53)
It's a little like that trending towards significance, right? Oh, it's not quite 0.05, but we're still going to count it anyway. Yeah, it wasn't quite intention to treat, but good enough.

[Kristin] (29:53 - 30:00)
Close enough to, it's trending towards intention to treat. Tantalizingly close to intention to treat.

[Regina] (30:01 - 30:03)
Okay, and then the last one, per protocol.

[Kristin] (30:03 - 31:00)
So her protocol analysis is getting out of intention to treat and saying, it's just what it sounds like. We're only going to analyze people who actually followed the protocol. So we're going to exclude anybody who didn't actually get the intervention that they were supposed to get, didn't come back for their follow-up visit.

So it's now asking, instead of what happened after randomization, it's asking what happened among people who actually followed the protocol as intended. So let me bring this back now to the GOGOVAX trial. So remember, we started with 654.

So an intention to treat analysis there would analyze all 654, 327 per group as they were randomized. They, in fact, did that as one of their analyses in GOGOVAX. They also did a modified intention to treat where they included only 620 people.

They got rid of some people who were found to not be eligible, who never came back for any of the shots.

[Regina] (31:00 - 31:03)
Hopefully they were somewhat balanced between the two groups.

[Kristin] (31:03 - 31:33)
Yeah, the 34 people we lost were pretty balanced between the two groups. And they also did a per protocol analysis that included only 587 participants. So we lost even more.

And in that analysis, you had to have gotten both of your shots and come back for at least two follow-up visits. Now, here's what's interesting. Normally, the intention to treat analysis would be primary.

They pre-specified that their primary analysis was the per protocol analysis, not the intention to treat.

[Regina] (31:34 - 31:39)
That's unusual. That is interesting because usually that's considered like a weaker analysis.

[Kristin] (31:39 - 31:47)
It is, right, because of the fact that this confounding can creep back in once you start following what people actually did instead of what they were told to do.

[Regina] (31:47 - 31:50)
Did they explain why they did that? Did they give reasons?

[Kristin] (31:50 - 32:14)
Yeah, they did. So they pre-planned this. So this was not after the fact.

But they pre-planned this as an efficacy trial, meaning the researchers wanted to know, does the vaccine work biologically when administered as intended rather than simply when we told somebody to get the vaccine? So efficacy trials are fine. So this is not unreasonable because they really wanted to prove that biological connection here as opposed to real-world effectiveness.

[Regina] (32:15 - 32:28)
That actually makes some sense once you put it that way, because they're kind of repurposing this vaccine that they're using for one thing and saying, does it even work biologically in the body before we take it out and offer it?

[Kristin] (32:28 - 32:48)
Right. So you can make an argument from this. It's still not as strong, though.

It's still, you know, when a trial, when I'm reviewing trials, I still expect to see intention to treat. And if I don't see that, I'm going to have some skepticism because excluding participants can introduce bias. Now, it turns out here that it didn't matter.

All three analyses gave pretty much the same result. So that's great.

[Regina] (32:49 - 32:49)
That is reassuring.

[Kristin] (32:49 - 32:51)
Yeah, but it is something to point out.

[Regina] (32:51 - 33:06)
Yeah. I think it's really good that we're pointing this out here, because if you're not paying attention to these details or reading very carefully, you know, they just throw these words in here. We analyze it according to this and this, and it all kind of feels like jargon.

But underneath, it's really important.

[Kristin] (33:06 - 33:07)
It's really important, yeah.

[Regina] (33:07 - 33:11)
And so it's like they should put this in bold and highlight it.

[Kristin] (33:11 - 33:11)
Yeah.

[Regina] (33:11 - 33:17)
So, you know, I mean, good thing here that it's consistent. It ended up consistent, yeah. But it's good that we're highlighting it.

[Kristin] (33:17 - 33:25)
Yeah, yeah. It would be very important if they got different results in those three analyses. They treated the intention to treat and modified intention to treat as secondary.

[Regina] (33:26 - 33:31)
As secondary. Okay. And there.

Okay. So what did they find? You said it didn't matter.

Yeah. What they actually found.

[Kristin] (33:31 - 33:48)
Yeah. Let's start with the main result there. Per protocol analysis, their primary analysis, it found nothing.

Nothing. The incidence of gonorrhea was 48.1 infections per 100-person years in the vaccine group. And in the placebo group, it was 47.8 per 100-person years. Pretty much the same.

[Regina] (33:48 - 34:11)
Ah, that is. Okay. So let's unpack because there are like numbers and units in there.

Okay. So 100-person years means if you were to follow 100 people for one year, that's essentially what it is. You said they got 48.1 infections. Almost half the people essentially in this got a gonorrhea diagnosis within a year.

[Kristin] (34:11 - 34:13)
Yes, that is a lot.

[Regina] (34:13 - 34:19)
Even with the vaccine. And it didn't matter whether they got the vaccine. 48 versus kind of 48 right away.

[Kristin] (34:19 - 34:36)
That is much higher than that one in a thousand, two in a thousand we were talking about. This is a really high-risk population. They were recruiting people from sexual health clinics.

So I'm thinking these are kind of like repeat customers at sexual health clinics, right? They might have even joined the study because they've had multiple infections and they wanted to be screened and treated.

[Regina] (34:37 - 34:47)
So this is showing kind of the benefit then of using this high-risk population because they made sure that they had enough people. Yeah, there may be too many.

[Kristin] (34:47 - 35:03)
So if you divide the rate of infection in the two groups, intervention group divided by placebo group, that gives us an incidence rate ratio. And that came out here to be 1.01 with a p-value of 0.97. So that p-value is not flirting with anything.

[Regina] (35:03 - 35:18)
That is flirting with nothingness, with the void. 1.01, just as a reminder, so an incidence rate ratio of 1 would mean that there's no difference between the vaccine and the control group. So 1.01 means it's basically nothing.

[Kristin] (35:18 - 35:20)
Basically identical, yeah, exactly.

[Regina] (35:21 - 35:31)
Wow, that was a big null result. Yeah. That's like shockingly null.

Very null result. Yes, very null. But the other ones basically found the same thing.

[Kristin] (35:31 - 35:35)
Yeah, the intention to treat and modified intention to treat also showed the same thing. Very little difference between the groups.

[Regina] (35:35 - 35:38)
Okay, all right. So we like the consistency at least.

[Kristin] (35:38 - 35:49)
Yeah, that's reassuring. And they looked at the secondary outcomes, recurrent infections, symptomatic infections at particular sites of the body, nothing. So overall, a pretty robust null result.

Really?

[Regina] (35:49 - 35:58)
Okay, so we started with these ecological studies, the analyses, and then the observational, and then poof. Poof, yeah. It just disappeared.

[Kristin] (35:59 - 36:17)
Yeah. So I mean, this happens sometimes in epidemiology. This happens a lot with vitamin studies.

We talked about in vitamin D, for example. But this was just men, right? Yes.

So there are some caveats. And I think the authors did a good job in their paper of pointing out the caveat. So this was men and not women.

So this doesn't generalize to women.

[Regina] (36:17 - 36:20)
No, they have different mucous membranes. They do.

[Kristin] (36:20 - 37:04)
So this could help women still. We don't know. Yes.

Another thing they pointed out that I think is really important, 90% of the participants had had a gonorrheal infection within 18 months of joining the study. So most people had had a gonorrhea infection before. And we've said that being infected once does not prevent future infections.

But it is possible that having been exposed to the bug, they have developed some amount of natural immunity.

[Regina]
Oh, that makes sense.

[Kristin]
Right.

And so maybe that's obscuring a vaccine effect. Like maybe if we had tested this in a infection-naive population, maybe then there would be some benefit of the vaccine, right? Because nobody's claiming it's offering perfect protection.

So maybe this kind of moderate 33% protection really depends on you being naive. That could be.

[Regina] (37:04 - 37:13)
I think this is why it's important to really look at the population when you start talking about these studies, right? Because it's easy to say, vaccine has no effect, but in whom? Right.

[Kristin] (37:14 - 37:33)
Exactly. That's why we talk about population. Yes.

Now, Regina, I should point out there is one more large trial of this that's still ongoing, and we don't have the results yet. And that trial includes 2,200 people, including women. So the evidence could change.

We could change our minds about this one when new evidence comes in.

[Regina] (37:33 - 37:41)
Go, go, women. Yeah. This is interesting.

I think we're ready to wrap up the chunk of evidence then for this claim. Remind us of the claim.

[Kristin] (37:42 - 37:48)
The claim is that a vaccine against meningococcal disease also protects against gonorrhea.

[Regina] (37:48 - 38:02)
OK. And we are going to rate that with our smooch scale. One to five smooches.

One means little to no evidence in favor of the claim. Five means really strong evidence in favor of the claim. Go ahead, Kristin.

Kiss it or Diss it.

[Kristin] (38:02 - 38:23)
So I'm going to go 1.5 smooches here. You know, we don't really have evidence from randomized trials. So I think the bottom line is, you know, we don't have evidence.

But there was enough hints from the observational data. There's some biological plausibility. And I'm holding out for that bigger trial with women to see if there's really anything there.

So I'm going to go 1.5. What about you?

[Regina] (38:23 - 38:41)
I think I'm going to go with one smooch on this. I'm going to be a hard line on this one. But it sounds like the authors did a good job.

And I am going to hold out hope for the go-go women trial. Yes, right. But just one smooch here.

All right. What about methodologic morals? I feel like we have a lot of good ones to choose from here.

[Kristin] (38:42 - 38:57)
Yeah. You know, I decided to go with this, how we hate this idea of statistical trending, how statisticians really, that gets under our skin. So mine is going to be, don't mistake a p-value of 0.06 for a p-value of 0.04 with bad luck.

[Regina] (38:59 - 39:00)
Oh, I like that. I like that.

[Kristin] (39:00 - 39:01)
What about you, Adina?

[Regina] (39:01 - 39:14)
I think I'm going to go with the STI theme on this one because, you know, clearly I've been missing the sex in our episodes lately. How about this one? Practice safe statistics, randomize whenever possible.

[Kristin] (39:15 - 39:15)
Oh, I love it.

[Regina] (39:16 - 39:16)
Yes.

[Kristin] (39:16 - 39:17)
Safe statistics.

[Regina] (39:17 - 39:20)
And use condoms. I think randomization is the condom here.

[Kristin] (39:21 - 39:22)
Oh, good metaphor.

[Regina] (39:22 - 39:24)
It protects against confounding.

[Kristin] (39:25 - 39:30)
Wow. And confounding is definitely a bad infection. It is a bad infection.

But we want to avoid.

[Regina] (39:31 - 39:47)
Infection. Okay. I'm going to put that in my 50 Shades of P-Values stats porn book.

Maybe we should send like a coffee mug to the person that gives us the best p greater than 0.05 phraseology.

[Kristin] (39:47 - 39:49)
I love it. To put in the papers. Okay.

Yes. That's a challenge for listeners.

[Regina] (39:50 - 39:52)
Okay. Kristin, this has been fun and.

[Kristin] (39:53 - 39:54)
Who knew gonorrhea could be so much fun.

[Regina] (39:54 - 39:56)
I know. And that was.

[Kristin] (39:56 - 39:57)
And statistics.

[Regina] (39:57 - 39:58)
And, well, of course.

[Kristin] (39:58 - 40:02)
What do people dread more, statistics or gonorrhea, do you think?

[Regina] (40:03 - 40:05)
Well, statistics about gonorrhea.

[Kristin] (40:05 - 40:08)
Oh, yeah. That's the worst. Yes.

Now we've made it fun.

[Regina] (40:08 - 40:13)
The interaction effect. Okay. So this has been fun.

Kristin, thank you so much.

[Kristin] (40:14 - 40:15)
Thanks, Regina. Thanks, everyone, for listening.