Journal Jam 24 Antibiotics for Strep Throat: Evidence, Myths and Misperceptions

Summary

Podcast Summary: Emergency Medicine Cases – Journal Jam 24: Antibiotics for Strep Throat: Evidence, Myths and Misperceptions

Release Date: June 2, 2026
Host: Dr. Anton Helman
Guests: Dr. Justin Morgenstern (First10EM), Dr. Casey Parker (Broom Docs, Australia)

---

Episode Overview

This episode takes a deep dive into the evidence behind prescribing antibiotics for strep throat. Drs. Helman, Morgenstern, and Parker critically appraise the literature, dissect myths and misconceptions, and weigh the risks and benefits of antibiotics in pharyngitis. The discussion goes beyond simple symptom control and addresses deeper questions about evidence interpretation, harm measurement, and context-based practice—especially regarding preventing rare but serious complications (like rheumatic fever), public health versus individual patient care, and empowering patients in decisions.

---

Key Discussion Points, Insights & Timestamps

1. Why Spend an Entire Episode on Sore Throats?

[03:12]

• The seemingly simple question—antibiotics or not—conceals complex evidence and substantial practice implications.
• “The details let you use your clinical judgment. It’s no longer just black and white ... you can take slightly different approaches to different patients.” – Dr. Justin Morgenstern [03:41]

2. Framing the Population

[05:40]

• Focus: routine pharyngitis, not deep space infections, mono, epiglottitis, thyroiditis, etc.
• Studies reviewed exclude severely ill or complicated cases.

3. Do Antibiotics Improve Symptoms?

[08:19]

• The 2021 Cochrane Review suggests antibiotics modestly shorten symptom duration (12–24 hours; NNT 6 for improvement by day 3).
• Dr. Helman: “Depending on how you slice the data, we’re talking about maybe a 12 to 24 hour reduction in symptom duration.” [08:36]
• The panel argues this benefit is likely overstated due to flaws in the evidence base.

a. Blinding Issues

[10:53]

• About half of foundational trials were unblinded (mostly from the 1950s, often enrolling hospitalized military patients).
• “Subjective outcomes like a sore throat… are at very high risk of bias in unblinded trials.” – Dr. Casey Parker [11:31]
• Properly blinded modern trials often show smaller or no benefit.

b. Inclusion Criteria & Swabbing

[13:14]

• Huge heterogeneity: Many trials included only hospitalized patients, with variable timing and criteria.
• Swabbing introduces potential bias—delays mean most patients improve before antibiotics could even be started.
• “Swabbing has never made a lot of sense, because either you’re treating everybody up front and you’re still going to get all the harms, or you’re waiting 48 hours and then everybody’s better anyway.” – Dr. Morgenstern [15:27]

c. Publication Bias

[20:21]

• High likelihood of bias due to era (pre-trial registration days)—many negative studies likely went unpublished.

d. Use of Analgesics/Steroids in Trials

[21:45]

• Some trials banned analgesics/antipyretics—“absurd and cruel” by modern standards.
• Steroids (like dexamethasone) are effective for symptom relief; unclear if antibiotics add benefit over/above this.

e. Reporting and Significance of Harms

[26:32]

• Very few trials reported antibiotic harms.

• “You get the hint from some of the studies that report on patients who had to stop antibiotics … somewhere in that sort of 10 to 15% of patients have some sort of side effect … bad enough to actually stop taking the antibiotics.” – Dr. Parker [27:15]

• Trade-off: For a small, possibly illusory symptom benefit, 10–15% get side effects (nausea, vomiting, diarrhea).

4. Panel’s Practical Approach to Symptom Relief

[29:03], [30:32]

• Bottom line: Antibiotics should not be routinely used for symptom improvement alone. Analgesics and dexamethasone are preferable.

---

5. Can Antibiotics Prevent Suppurative Complications?

[35:17]

• Data is extremely weak—numbers for suppurative complications (like peritonsillar abscess) are derived from decades-old, flawed studies.
• Main Cochrane signal relies on a single, unblinded 1951 trial, with dubious math and implausibly high abscess rates (5% in placebo).

Memorable Moment:
“We need to drop the word prevention from our vocabulary. That’s not the goal. … The worst 1 to 2% of pharyngitis that you see definitely warrants antibiotics.” – Dr. Morgenstern [49:50]

• Observational modern data (e.g., Lidl 2013): complication rates ~1%, no difference between those who received antibiotics (immediately, delayed, or not at all).

• Most peritonsillar abscesses present early (days 1–2), suggesting a primary pathology rather than a downstream complication.

• Implication: It is reasonable to give antibiotics for severe/unilateral pharyngitis that raises suspicion for early abscess, but not to prevent abscess in run-of-the-mill sore throats.

---

6. Preventing Glomerulonephritis

[50:24], [52:34]

• Glomerulonephritis after strep throat is vanishingly rare in most high-income settings.
• Only 1 case in the Cochrane review—no robust data that antibiotics make a difference.
• Outbreaks can occur in specific high-risk populations (as Dr. Parker describes from Northern Australia)—in these settings, public health interventions/antibiotics may be justified.

---

7. Invasive Group A Strep (iGAS) & Sore Throat

[56:48], [57:33]

• iGAS is a feared deadly disease, but evidence does not show a link to antecedent strep throat.
• Only 2/63 iGAS pediatric cases in a recent European wave had preceding pharyngitis.
• “Invasive group A strep is not coming from pharyngitis. That shouldn’t be our focus.” – Dr. Morgenstern [59:48]
• Bottom line: Do not change your sore throat management during iGAS outbreaks.

---

8. Does Treating Strep Throat Prevent Rheumatic Fever?

[60:16], [60:39]

• YES, but only in specific high-risk populations.

• In the 1950s, untreated rheumatic fever risk was ~3%, antibiotics reduced by ~2% (absolute).

• However, current rates in high-income countries are extremely low (e.g., 3 per million in Canada), except for Indigenous/First Nations populations or in parts of Australia/New Zealand/SE Asia.

  • “If you’re working in suburban Toronto or Montreal or Halifax, your risk is essentially zero … But if you’re working with our First Nation populations … your approach needs to be very different.” – Dr. Morgenstern [66:02]

• Prevention is necessary in high-risk settings; for average low-risk patients, antibiotics provide no benefit for rheumatic fever.

---

9. When Should You Prescribe Antibiotics or Swab?

[69:28]

• Give antibiotics only when:

  • Patient is at significant risk for complications (immunocompromised, high-risk for rheumatic fever, e.g., certain Indigenous populations)
  • Severe, early pharyngitis, especially with features suggesting risk for evolving deep space infection/abscess (“maybe the most severe 1 or 2%”)
  • Rarely, when faced with insistent patient expectations (panel generally recommends against this)

• Swabbing:

  • “If you’re a well person [from a] low-risk population… swabbing is a wasted swab.” – Dr. Parker [74:11]
  • Swab only if local public health protocols call for it due to elevated risk for rheumatic fever AND swab results would influence management.
  • Otherwise, routine swabbing offers no benefit—delays management and doesn’t change outcomes.

---

Notable Quotes & Moments

• “Subjective outcomes like a sore throat or throat pain—they’re at very high risk of bias in unblinded trials.” – Dr. Casey Parker [11:31]
• “Swabbing has never made a lot of sense … either you’re treating everybody up front and you’re still going to get all the harms, or you’re waiting 48 hours and then everybody’s better anyway.” – Dr. Justin Morgenstern [15:27]
• “If a complication is really rare, then you have a much bigger number needed to treat. In other words, more patients need to be exposed to antibiotics just to see one get a benefit.” – Dr. Parker [39:46]
• “Invasive group A strep is not coming from the strep from a strep throat. It’s not coming from pharyngitis. That shouldn’t be our focus.” – Dr. Morgenstern [59:48]
• “Antibiotics absolutely reduce rheumatic fever … in the right population.” – Dr. Morgenstern [60:54]
• “If you really want to prevent post-glomerulonephritis, you’re probably better off being either a public health physician or maybe a plumber.” – Dr. Parker [56:48]

---

Bottom-Line Takeaways

Antibiotics in Strep Throat:

• Do not routinely prescribe antibiotics for symptom relief—benefits are small, harms are not.
• Do not rely on antibiotics to prevent peritonsillar abscess, glomerulonephritis, or iGAS in low-risk populations—evidence does not support this.
• Antibiotics prevent rheumatic fever only where this is a real risk (some high-risk Indigenous populations, parts of Australia, New Zealand, SE Asia)—most patients with sore throat in North America & similar contexts do not need antibiotics for this reason.
• Swabbing for strep is rarely helpful—unless in a context where results will truly change practice (and patient population is at meaningful risk).

---

“Cheat Sheet” for Practice

• Symptom control?
  – Use acetaminophen/ibuprofen and consider dexamethasone. Antibiotics add little, with more harm.
• Preventing suppurative complications (abscess)?
  – Only consider antibiotics in the sickest 1–2% with severe, early, worrisome features.
• Preventing rheumatic fever?
  – YES, but only for specific high-risk populations.
• Preventing iGAS or glomerulonephritis?
  – Not supported for routine pharyngitis.
• Swab?
  – Mostly “no,” unless following a public health protocol in high-risk communities.

---

Final Thoughts

• Routine antibiotic use for sore throats is not justified for most patients.
• Public health and context are critical; blanket approaches do not fit all.
• Patient communication is essential: explain the small benefits, the real harms, and focus on better methods for symptom relief.
• “This isn’t just about sore throats. It’s about how we interpret imperfect evidence ... about public health versus the patient in front of us. And it’s about empowering patients to make informed decisions.” – Dr. Anton Helman [00:39]

---

For further details, check the episode's full write-up and resources at emergencymedicinecases.com and First10EM.com.

Transcript

[A]

I'm anton hellman.

[B]

I'm justin morgenstern and this is the journal jam podcast.

[A]

It's been way too long since our last Journal Jam and I've missed it dearly. There's something deeply satisfying about slowing down, putting aside the hot takes and the headlines and really digging into the evidence. Especially now, in the era of AI summaries and confident hallucinations, I think it's more important than ever that we keep real evidence based medicine alive in em. So today we're doing something that on the surface seems almost embarrassingly strep throat. Should we swab strep throat? Should we prescribe antibiotics for it? Do antibiotics meaningfully reduce symptoms? Do they prevent peritonsillar abscess? Glomerulonephritis? Rheumatic fever? Invasive group A, strep? More importantly, does understanding the nuances actually change what we do at the bedside? Because this isn't just about sore throats. It's about how we interpret imperfect evidence. It's about balancing small benefits against poorly measured harms. It's about public health versus the patient in front of us. And it's about empowering patients to make informed decisions that instead of reflexively reaching for a prescription. Now, new to EM Cases is the host of the excellent Broom Docs podcast, Casey Parker. I can't believe after all these years we've never done a podcast together. So I'm really psyched for this. Welcome to EM Cases, Kasey. Kasey, tell us a little bit about your preferential background for those EM casers who haven't heard your voice yet and what got you interested in ebm.

[C]

Yeah, so thank you for having me. Anton. It's amazing. I think we've passed like podcasts in the night in that I always listen to your podcast and I presume that someone in Canada listens to mine and we've never crossed over. So that's an interesting thing that hasn't happened in 10 or 15 years. So thanks for making this happen at last. I'm a what we call a rural generalist in Australia, so I trained initially as a gp, then I did some anesthetics, some obstetrics, neonates, all sorts of things. But about half of my care is emergency department care. And in the last five years I've also taken on a role in a big quaternary hospital as a sonologist, which is someone that practices as a doctor, hybrid radiology, zoologist. So it's a very diverse set of skills and I seem to be average at most of them. So thanks for, for that.

[A]

Fantastic. You know, in a world where everyone's becoming more and more and more subspecialized, I really appreciate that there's some folks out there like you, who are becoming more generalist. That takes a lot of courage, actually. And we all know Justin Morgenstern from EM Cases and FirstTen EM. Justin, great to be with you here, buddy.

[B]

It's an absolute pleasure to be back and doing some journal jam.

[A]

All right, so let's crack this open. So, Justin, first, why on earth are we spending a full journal jam on sore throats?

[B]

Yeah, you would think on the face this would be a really simple question, like maybe a five minute quick hit, right? Like a pretty simple yes or no antibiotics or no antibiotics. But I think it's worth diving into this literature. And I don't just say that because I'm biased, because I've literally spent three and a half years reading all of these papers. It's taken me that long. But I do think that there is some value here. And it's more than just the general purpose of the journal jam series, which is always to gain a better understanding of science and critical appraisal by going through some of these details. But I do think that the nuance in this topic actually does change practice, or at least understanding these details have impacted my practice, and I think they've helped my patients. The details let you use your clinical judgment. It's no longer just black and white. Here's what the textbook tells me to do. And it means that you can take slightly different approaches to different patients. And I think in doing so, provide better care. And I really think it's helped my patients. I see so many patients who for a decade have gotten six or seven or eight bouts of pharyngitis in a year, and they're convinced that they need antibiotics and a swab every single time. And they always swab positive, presumably because they are carriers. And really going through some of these details with them really helps them make better decisions. It helps them make decisions about whether they want to spend their time waiting in an emergency department waiting room, paying for parking, coming to see me. And so I think a minute or two, explaining some of these details to my patients has really drawn dramatically helped their care, their lives, what they. The decisions they make in their future. And to me, that's a very rewarding part of the job, a part of the job that makes me happy to be an emergency physician. And so that's why I chose this as a topic, because I think it's both Clinically important, but also it helps us learn some of those evidence based medicine skills.

[A]

Yeah, I think in a world where it seems like we're doing way more testing for just about everything in emergency medicine, I think what you've just said, Justin, is especially important. But before we jump in further, I think we need to clarify exactly what we're talking about in this podcast. Because sore throat has a wide variety of presentations and a wide variety of diagnoses. You know, the vast majority are just going to be simple pharyngitis. But Justin, can you just clarify for us all these studies that we're going to talk about, like, what are the general sort of inclusion criteria? What are the sore throat patients that we're going to be talking about in an EBM way in this podcast?

[B]

Yeah, it's a really important question because I guess technically strep throat isn't a medical diagnosis. It's, it's a lay diagnosis. We're going to try to focus on pharyngitis. And so that means you're going to look into the throat and you're going to see a red swollen throat. If you don't see that and somebody's complaining of significant pain, obviously we had to go down a different pathway. We're going to start thinking epiglottitis and other more severe pathology. We're also going mostly exclude the patients who clearly have complications. If the uvula is way off midline, if they can't move their neck back and forth, their voice is muffled. We're thinking about deep space infections. We're going to presume that we've already ruled out the very sick patient, the patients with deep space infections, patients with alternative diagnoses, and we're, we're really focusing on the bread and butter. You know, why are you in my emergency department? Why didn't you go to a family doctor or a walk in clinic pharyngitis. Now, from there, there's still a broad, broad differential based on the causative organisms. And we'll talk a little bit about how that impacts our evidence because not everything is caused by strep. There's obviously a huge amount of viral illness and it's hard to piece that out. And we'll talk about how important that might be later in the episode in terms of swabbing. But I think we can loop it all in right now. You see somebody with a bad red sore throat, but you don't think they have significant other pathology. Just basic pharyngitis.

[A]

Got it. Okay. So just a reminder to our Listeners, then we're not really talking about mono in particular. We're not talking about epiglottitis, we're not talking about Ludwig's angina, we're not talking about other deep space infections, we're not talking about thyroiditis. You know, there's a huge differential of sore throat that we really need to think about. Every time we see someone with a sore throat, it's not always going to be just a viral pharyngitis or strep throat. Okay, great. Okay. So when someone walks into the ED with a sore throat, they're not thinking about peritonsillar abscess or glomerulonephritis or rheumatic fever. You know, some of the complications of strep throat. They're thinking, this hurts, I need to feel better, I've got to go to work tomorrow. And most patients want something that will make the pain and the fever just go away faster. And for many, that something is antibiotics. So let's start with the most basic patient centered question. Do antibiotics actually make people feel better faster? You know, if you look in the 2021 Cochrane Review, the answer appears to be yes. When it comes to the duration of fever, it doesn't seem like there's any significant difference with antibiotics. But duration of sore throat and headache seems to be modestly improved with antibiotics, with the number needed to treat at day three of about six. And by one week it's about 18. So depending on how you slice the data, we're talking about maybe a 12 to 24 hour reduction in symptom duration. So on the surface, that sounds pretty reasonable. You know, you treat six people, one feels better by day three. That doesn't kind of seem outrageous to me. But the answer isn't nearly as straightforward as the abstract may make it look. So, Justin, if Cochrane makes this sound relatively simple, why are we even doing this episode? Are antibiotics truly improving symptoms in a clinically meaningful way or is there more going on under the hood?

[B]

So it's a great question, Anton, and unfortunately, I think the best answer I can give you is just a maybe or perhaps, or I don't know. At the end of the day, the Cochrane may be right and maybe there's a small benefit, but I think there's also a big question about whether that benefit is worth being the harm, which you always need to bring up. So the whole point of this is to go through the actual data so that listeners can make up their own minds. And I'll say, traditionally, in journal jam, we've Literally gone through every single paper published on a topic so we can get a really deep understanding of literature when it comes to sore throat. As I already said, it took me three and a half years to do this. I don't think that's a great idea. Today there are dozens of papers, at least half of them were published before 1960. And so instead of reading every paper individually, I think what we should do is just go through some big critical appraisal topics that came up over and over again in the literature so that people can understand why I might not have the exact same conclusion as the Cochrane Review. And if they want more details, there will be a full write up that includes all the papers on, on the first Tenem site. So I think the big topics that I that came up time and time again in these papers were blinding, or the lack thereof, potentially publication bias, the inclusion of viral pharyngitis or the need for testing to try to identify these patients, the use of analgesics or antipyretics, or the exact opposite. Actually, in some trials they weren't allowed to use pain control. And then I think by far the most important topic we'll talk about is harms, or really the lack of reporting of harms. So if we could talk about those general topics, I think you might understand why the point estimate in the Cochrane Review might not be as exact or as scientific as we would like to think.

[A]

Okay, so let's dig into that list that you just outlined and start with blinding. We all know that without proper blinding, you're introducing some serious bias that pretty much destroys credibility. So Casey, was there decent blinding in the strep throat studies that the Cochrane drew from?

[C]

Yeah, so with blinding, it's really tricky. So about half of this data comes from trials that were in the 1950s, and these trials were far from perfect. About half of the trials were blinded, but if you do the math, that means about half of them weren't blinded. And the most common sorts of trials were run in the 1950s. They tended to randomize patients by having one ward of patients receiving antibiotics and another ward full of patients not having antibiotics. And so that's not proper randomization as we, as we see it in the 21st century. And this brings up another issue as well. This is a group of patients in the 1950s, mostly in military hospitals, and these patients were admitted to a hospital with a sore throat, which seems quite crazy. And it makes me think that we're probably talking about a completely different disease process, because I personally haven't admitted one of these simple pharyngeal patients in my whole career. So subjective outcomes like a sore throat or throat pain, they're at very high risk of bias in unblended trials. And so the fact that about half of the Cochrane data is unblinded certainly reduces the trust that we can have in the estimations.

[B]

Now, it would be nice to see an analysis of just the blinded trials. And there are some modern trials that look a lot better. And I think the important thing, you know, if you see really consistent data across the board, you can be confident. But there's a lot of negative trials mixed in here and a lot of them are the ones that are properly blinded. So it's not like every trial shows a benefit and we're just quibbling about whether they were blinded or not. There are a lot of trials that show no benefit in terms of source throat or fever that are mixed in as well. And we'll talk about some of the other reasons that might be as we go through here.

[A]

Great. So that's the blinding next on your list, Justin, was inclusion criteria. And one thing that really stood out in these trials is how variable the inclusion criteria were. For example, did all patients require a positive strep swab before starting antibiotics? You know, this brings up the issue of timing of antibiotics because so many patients will be better by the time the culture results come back. Then there's the so called rapid strep test, which you can get back within an hour or so, but we all know that it's nowhere near as sensitive as a proper culture. Justin, what were the inclusion criteria in these studies and how do we sort out the timing of swab's issue?

[B]

Yeah, so when I teach evidence based medicine, I always tell people to think about the population of a trial for two major reasons. Number one is do you think it's fair? Do you think they included things that are scientifically fair? And number two, do they actually look like the patients you're looking after? So does the data apply to your current patients? And Casey already mentioned one of the big things, almost all of these 1950s trials, everybody was admitted to hospital. And that just blows my mind. And so I wonder, was the strain different or were outcomes just very different in the world prior to antibiotics or something must have been different. So number one, I have a question about whether these, this data really does apply to my current patient, which is always just if you're treating a prescription for amoxicillin and send them home, not consult medicine and get them into hospital. So that's question Number one, I think the bigger question that we're going to have to tackle here, looking through this data, is we're trying to isolate strep throat from all of the other pharyngitis that is a very specific pathogen. In order to do that, you have to do the swab and you have to wait 48 hours. Now, rapid strep has changed that a little bit, but most of the trials, you'd have to wait 48 hours. And in terms of a fairness of a trial standpoint, that can shift things in one of two ways. Some of these trials included everybody, but then 48 Hours decided to only look at the patients who actually had strep. And by ignoring the 75% of patients that swabbed negative, what you're essentially doing is ignoring harms in 75% of your patients and only looking at the benefits. That makes sense. Right? So if I have a thousand patients and 750 didn't have strep throat, those 750 might still had diarrhea and nausea and vomiting and abdominal PA pain. And by just eliminating them from the trial, you're effectively biasing towards benefit from antibiotics because you're ignoring harms in a large number of patients. So that's one thing to consider. The other thing to consider is you could do it the other way, which is we're just going to treat everybody, in which case, you know that your antibiotics can only possibly help a small minority, which again is going to cause more harm than benefit. Or there's the practical issue. You could wait, you could wait 48 hours for the swab results to come back and only try start your trial then. But most pharyngitis is just better naturally by two or three days. So by the time you have the swab results back, the groups are going to look almost identical. So that is sort of a big problem. Looking through all of this literature is who actually has strep or a bacterial pharyngitis. And when do you decide to treat? And I think at the end of this, we'll talk a little bit about when we might decide to dwell. But swabbing has never made a lot of sense, because either you're treating everybody up front and you're still going to get all the harms, or you're waiting 48 hours and then everybody's better anyway. So that 48 hour lag becomes a bit of a problem in these trials.

[A]

All right. You know, it's interesting when you say harms, you didn't even get to mentioning antibiotic resistance. Which, you know, in the big picture that might be the biggest harm to society is antibiotic resistance. Casey, any thoughts about the inclusion criteria on these studies?

[C]

Yeah, it's a tricky one, Anton, and I think there's a little bit of North Americanism going on because in Australia and a lot of parts of the world, swabbing for throats is not a done thing. And so it's very difficult for us because it's just not part of our culture to understand who would be in these trials. And certainly the external validity of these trials that include swabs certainly wouldn't cross over into my practice where we pretty much never swab people. So it's very tricky in that sense. I think that's the main thing I take away from the Cochrane and a lot of these trials is that there's a lot of swab bias, I think, which is hard to apply to my population.

[A]

That's just mind blowing, Casey. So in Australia, as far as you know, no one swabs for strep throat?

[C]

No, it's, it's just not a done thing. And I think maybe back in the 80s or 90s it was, it was more of a thing, but certainly not in the duration of my career. I can't remember swabbing someone's throat before COVID came along and changed our point of care testing dramatically.

[A]

Yeah, interesting. So do you know what spurred that change to not swabbing anyone?

[C]

Yeah, I'm not really sure, but I think that it was around this data and recognizing that maybe the antimicrobial stewardship and movement sort of wanted us to push more towards not treating people, even if they did have a positive swab, because we know that people get better on their own. So it may be that sort of microbial stewardship process that stopped the swabbing through it.

[A]

Sure, yeah. We're going to talk a little bit about this later. But for the patient who is really convinced they need antibiotics and you, let's say, think that they don't need antibiotics, I guess one strategy is to swab the patient and say, okay, we're going to swab. I mean, assuming that they haven't swabbed positive a million times before, you can use the swab as a way of placating patients and say, only if it's positive. Only if it's positive, then we'll consider giving antibiotics. Anyways, that was just a thought, but we're going to come later to how we might counsel patients if we think that they do not need antibiotics.

[B]

So there's One other interesting bit of this data, which is not in the methodology, but actually in the results, which is interesting to me, the outcomes of these trials, whether your pain went down, whether your fever went down, whether your symptoms improved, was exactly the same, whether or not you swabbed positive for strep. Think about that. Antibiotics had the exact same effect if you had viral pharyngitis as if you had strep pharyngitis in a bunch of these trials. That's strange. And I don't know what to make of that. I can think of two possible reasons for that. Number one, these trials are biased, right? Something is going on in these trials that tells us that it's not actually, and we've already talked about it, these are unblinded trials looking at subjective outcomes. So it's possible the fact that antibiotics work for viral infections just tells us that these are not great trials, that we're seeing biased outcomes. The other possibility is that we know some antibiotics have an anti inflammatory effect. So maybe we're using antibiotics not to kill the bacteria, but as analgesics or antipyretics in and of themselves. But if that's the case, why would you choose an antibiotic as compared to, say, something like acetaminophen or ibuprofen? So I think it's an interesting thing that it came up in not just one trial, but multiple trials, that the results are identical whether or not strep is present.

[A]

All right, great. So so far we've covered blinding and inclusion criteria, and the next on the list was publication bias. Justin, you've already mentioned a bit about bias in general. Publication bias in particular is the tendency for studies with positive or statistically significant results to be published more often and more quickly than studies with neutral or negative results. In other words, if 10 studies are done on a treatment and only the three that show benefit get published, the medical literature will make that treatment look far more effective than it actually is. Casey, to what degree did the studies in the Cochrane Review suffer from publication bias?

[C]

Yeah, that's a good point, Anton. We have a long history in medicine of only publishing the positive results, and a lot of this data comes from the 1950s, long before trial registration was a thing. And so you sort of have to assume that there's some degree of publication bias occurring. There was probably a bunch of small trials that showed no benefit, that probably never got published in the relatively small era of medical publishing at the time. To my eye, the funnel plots in the Cochrane Review suggest exactly that, in that we know that Small studies can skew positive, but small studies are exactly the ones that are likely to go unpublished if they have a negative outcome. So it isn't the whole studies that are impacted by publication bias in the era before trial registration. And when you look at the outcomes that were used in a lot of these trials, you can really see how researchers could pick and choose which ones they wanted to publish. So older studies often just list a bunch of outcomes without any single primary outcome. There's no power calculation. It's basically we'll just throw a heap of symptoms at the wall and see which one sticks. And this significantly increases the degrees of researchers freedom. And you can imagine that the outcomes seem to disagree with what they sort of hypothesis was, then the researcher may have just left that out of the manuscript. So, very different era of publication which we're dealing with.

[A]

Yeah, totally makes sense. You know, one totally absurd and cruel thing I noticed about some of these studies is that they banned patients from using analgesics or antipyretics. I mean, one of our primary responsibilities as physicians is to decrease suffering. That would certainly not fly by ethics boards in 2026, that they were banned from giving anything to make the patient actually feel better. And if we're talking about getting patients better, that brings up in my mind dexamethasone, which my understanding is, we've talked about this on EM cases before, that dexamethasone actually has some pretty good evidence for improving symptoms in patients with severe pharyngitis. Justin, what's your take on this banning analgesics and antipyretics issue of some of these studies? And then we can talk a bit about steroids.

[B]

Yeah. So in fairness to people running studies in the 1950s, from a pure science standpoint, if you want to know the impact of antibiotics, you try to control every other variable so that the benefit won't be masked by the other treatments. So I can sort of understand the logic, but from a clinical standpoint, it just means that this data doesn't help us at all.

[C]

Right.

[B]

It makes no sense because we don't want to know if antibiotics provide some pain control. We want to know if antibiotics provide analgesia in addition to, to the Tylenol or acetaminophen or ibuprofen that we're clearly going to be recommending anyway. Like, that's the baseline for the patients. And so if antibiotics don't add anything to what the patient's already going to be taken, they aren't adding anything to my, to my practice. And it just doesn't make any sense. Nobody's ever going to suggest swapping ibuprofen for amoxicillin. Right. That's not a swap you're going to make. So you want to know if it adds to your acetaminophen or your ibuprofen. The steroids question is a really interesting question, and it's not actually in any of this literature. But if your goal is to make. Make patients feel better, as you say, we know that steroids are wildly effective. What I don't know, and no trial that I could find to ever look at it, is are antibiotics any better than a dose of dexamethasone? Do they add anything to the dose of dexamethasone? I think we should be careful. I know everyone loves dexamethasone in this day and age, but steroids do also have a lot of side effects. Both antibiotics and steroids have a lot of side effects. And I'm actually not sure which would have be better for symptom control in and of its own. So I think it's. I would love to see a modern RCT directly comparing dexamethasone to antibiotics, both for which does a better job of pain control, and I'm pretty sure it's going to be dexamethasone, but which has more side effects. I think we underestimate dexamethasone's side effects a lot. And so I think that's some research that really needs to be done going going forward. I don't think we need that research to compare to acetaminophen, because that's just going to be your baseline. But I don't know which is better, amoxicillin or dexamethasone?

[A]

Yeah. What I would really want to know is are there significant side effects of one dose of dexamethasone? Because that's what I'm often giving patients with pharyngitis. And we know that prolonged use of dexamethasone has a whole variety of horrible side effects, but one dose, there's probably not much in the way of side effects. And that's really the study that I'd love to see is one dose of dexamethasone versus a week of penicillin.

[B]

I think when you talk through this data, it's really important to clarify what types of side effects we're talking about, because as emergency doctors, we're often thinking about the worst. The anaphylaxis, the Steven Johnson syndrome. But in the context of this conversation, we're talking about symptom control. So the whole point of giving dexamethasone is to make their throat pain go better. And I think there's pretty good data that 10 to 15% of patients get a pretty uncomfortable mild psychosis or mania. They can't sleep in the context of dexamethasone. So even though you might write it off as being minor adverse events, I do think the dexamethasone actually does have a lot of side effects, even in a single dose, just not in the kind we're not going to cause bone necrosis. But I think there's still, when you're, when you're thinking about controlling symptoms, you also have to think about the symptoms we're causing. And I. And I don't know is my answer.

[A]

Sure, yeah. I mean, I've been hypomanic before and I kind of like being manic. Maybe I should try some dexamethasone. Just one hit. So, so far in our list of topics we've covered blinding, publication bias, inclusion of viral pharyngitis, the need for testing, and use of analgesics and antipyretics. The last thing on the list, Justin, that you mentioned near the top is harms, which we've touched on a little bit. Casey, did the studies report harms, and if so, were they important ones?

[C]

Yeah. So I think this is the biggest issue with this data that we have, especially in the Cochrane Review and the older trials, in that almost none of these studies reported on the harms from antibiotics. And it's very difficult or possibly impossible to make a harm risk analysis if you don't know the harms. And on that side of the equation. So the question is, as Justin's pointed out, does a slight reduction in sore throat, 6 to 12, maybe 24 hours reduction of symptoms, warrant an episode of nausea or diarrhea? That's what the real trade off is here. So antibiotics for strep throat are generally safe. Doesn't mean that there are no cost to these, though. And most studies don't actually measure it, but a few did. So in 1951, Brink measured it and saw that there was a 15% absolute increase in nausea and vomiting and a 10% increase in diarrhea in patients that received antibiotics. And Demeyer in 1992 found similar numbers with a 21% adverse effect rate as compared to only 5 in the placebo group. So similar sorts of numbers, about a 15% harm in that group. It's hard to know exactly how bad all of these adverse events are, because they're not really discussed or documented in the trials. But you get the hint from some of the studies that report on patients who had to stop antibiotics. So antibiotic discontinuation is reported in some trials, such as Danieli. In 1996, 6% of them stopped taking the antibiotics. So that is in that same sort of ballpark. So I think it's fair to say that somewhere in that sort of 10 to 15% of patients have some sort of side effect, and some of them are bad enough to actually stop taking the antibiotics. So that's the sort of level that we're talking about. In terms of harms.

[A]

Yeah, I think in terms of harms, it comes into play when you're counseling patients who really want antibiotics and you believe that they don't. In my experience, when I tell them about the harms from antibiotics, then that seems to convince them that maybe they shouldn't be getting antibiotics. But we'll talk a little bit more about how you counsel patients if you think they don't need antibiotics and they do a little bit later. Okay. Given everything we've talked about so far, I'd like to know from both of you, what's your bottom line takeaway when it comes to antibiotics for pharyngitis, for improving symptoms? And just to be clear here, we haven't even talked about complications of strep throat yet. We're just talking about symptom improvement of pharyngitis. Justin, you go first. Antibiotics for symptom improvement, yay or nay?

[B]

Yeah, I wish it was as easy as the Cochrane review, which makes it sound like it's fantastic. I can't give you any better than, I don't know, there might be a small improvement, which is what the Cochrane review says. Although there's so many problems with this data, I would not be surprised at all if antibiotics literally have zero impact on symptoms. But I'll just assume that they have a small impact, like they're going to make you feel less pain and less fever at 12 hours or something like that. The problem is the last point. It's. It's the harms. When I read this data, I don't think there's any chance that that small benefit out weighs the harms. Now, it's going to be individual for everybody else. But to me, the harms and the trials that I can see, the vomiting, the diarrhea, the abdominal pain, the nausea, seems like it's a lot worse than 12 hours of sore throat. So my read of this is I would not use antibiotics for the sole purpose of Making people feel better. And that's pretty important because that's really, in my experience, the only reason people come to see me in the emergency department. Now, there's other reasons maybe to use antibiotics. We'll talk about that. Going through these supper of complications, rheumatic fever, and so maybe that will push us one way or another. But if your whole purpose is to try to make the person feel better, less fever, less sore throat, I don't think antibiotics are going to be the route for that.

[A]

Yeah, especially since we have dexamethasone and acetaminophen and ibuprofen. We've got some pretty good medications that we know work quite well for improving symptoms. Casey, your take on antibiotics for symptom improvement?

[C]

Yeah, I think I agree with Justin. It's not something that I spend any time arguing with patients about in the emergency department. I tend to be very empathetic. I tend to say, well, you've got a really sore throat. Acknowledge their suffering. I talk about all the other stuff that they can do, you know, all the other sort of over the counter medications that they can try on different techniques. And then I say, great. Do you have any more questions? And in my experience, maybe this is an Australian cultural thing, but most patients just say, yep, that's great, and walk out. And there's maybe a 5 or 10% group that really want to talk about why they're not getting antibiotics. And I don't spend too much time arguing those people. I tend to offer them a prescription if they feel that something that they really, really need. But I definitely counsel them that if they start taking the antibiotics and they get any side effects from them, just stop taking the antibiotics immediately. I say don't continue taking the whole five or seven day course, but because you just, you've already got the side effects and you're just going to get the downsides and we don't really know what the benefit is. So I definitely counsel them to stop taking it as soon as possible. And as you said, Anton, I'm a big dexamethasone user. I tend to use a single dose and I cancel patients, particularly for children and parents. I say, look, use dexamethasone. It's kind of like taking all the Nurofen ibuprofen that you're going to take for the next couple of days all at once. And you don't have to keep torturing your child by getting them to swallow medications every four to six hours. So it's a bit easier for the parents, I find.

[A]

Sounds Like a very reasonable approach. All right. Based on our review so far, I have to agree with you both in terms of the lack of great evidence for giving antibiotics to improve symptoms more than decreasing the duration by a few hours. But my experience, I think, is actually quite different to Casey's. If I spend a few minutes explaining why I don't think antibiotics are especially beneficial for symptoms and even outline the harms and then focus on improving their symptoms with dexamethasone or analgesics or antipyretics, a lot of them still want antibiotics and I feel like I could have spent those few minutes maybe seeing another patient and I just get a bit frustrated, to be honest. So my evidence based answer is I don't think antibiotics work very well for patient symptoms, but practically speaking, I often end up still giving them antibiotics for the patients who I really think clinically have strep throat. But then the question is, who do I really think has strep throat? Well, I don't personally rely on the Centaur score because you're rarely above 50% pretest probability, even with a score of 4 out of 4. But if you take all the elements of the Centaur score and whether or not they have a runny nose, and if sore throat is their chief complaint or just one of their complaints, you know, whether or not their whole family has classic viral uris, are immunocompromised, you know, do they have a hint of trismus? Is their voice muffled at all? Do they have a choking sensation when lying down? These are all questions I ask them, you know, to screen for deep space infections and other things. But, you know, I take, even if they have a hint of any of these things, I'll take my overall gestalt and if their family doctor sent them in to get antibiotics, you know, that kind of thing, I take this entire clinical picture and then I make a decision. So for me there's just, it's so multifactorial as to whether I'm going to give them antibiotics, but certainly based on this journal jam, I'm not convinced that the perhaps few hours of improvement in the duration of symptoms is going to outweigh the potential harms for giving antibiotics. So I guess my bottom line there is, it's not very clear and it totally depends on the patient. Even if we're a bit skeptical about how much antibiotics actually help with symptoms, the Cochrane review gives us another reason to reach for the prescription pad, and that's to prevent the superior complications. So we're talking about peritonslar abscess, retropharyngeal abscess, parapharyngeal abscess, otitis media, sinusitis lemires. And on paper the numbers look pretty compelling. The odds ratio for otitis media is 0.21. For peritonsillar abscess it's 0.16. So when you see numbers like that, your reflex is to think that's a pretty big effect. That's the kind of signal that makes you feel uncomfortable. Not prescribing, but as always on journal jam. Before we get too excited about impressive looking odds ratios, we need to slow down and ask what are the absolute risks and how many events are we actually talking about and how solid is the data underneath those numbers? So Justin, let's talk about preventing complications with antibiotics for our strep throat.

[B]

Yeah, so I think there's probably even more uncertainty about whether antibiotics prevent suppurative complications than there is about their value in symptom control. And that is saying something because we spent a lot of time talking about the uncertainty in the symptom control data. I personally skip over. I ignore the otitis media component of this conversation because that seems silly to me. Lumping it is as a complication seems silly. It's a self resolving condition that doesn't require any medical intervention. So I don't know that it counts as a complication. And if there's any value in his prevention, I think we would have seen it in the symptom sense section because their people would have had less pain. So I'm not as worried about the otitis media, but it'll have the exact same methodology problems as the other complications. So I think we can just focus on the main one, which is peritons abscess. And I'll tell you, I really tried to confirm the Cochrane numbers and I don't have a lot of faith in them. First, there are 25 total cases of peritonsillar abscess in the entire Cochrane review. Sixteen of the 25 come from a single paper in 1951, an unblinded trial from 1951. So that that tells you a little bit about the quality of the data to begin with. All of these patients, as we said, were admitted to hospital. It doesn't really seem to apply to the patients we're talking about today. Amoxicillin and send them home. There's some other problems. Cochrane says that there were 15 abscesses in the placebo group. Read the study, there was only nine. Basic math errors always make me A little bit nervous about the quality of the data that we're doing. And even at 9, that's 5% of patients developing an abscess. 5%. That is unheard of. We're not seeing 5% peritonsillar abscesses. And so this must be a very different population than what we're used to seeing. So they were all admitted, so maybe there's some selection bias there. So again, when the vast majority of your data comes from an unblinded 1951 trial with bizarre numbers, I think you can't just take that Cochrane number as good as the number seems. And it's not just that one trial. I read all the trials and there's almost no information in any of these trials about the actual abscess. What happened in Zwart 2000, there are two abscesses that get mentioned in the abstract, but they don't mention them in the main body of the paper anyway, so I don't know what actually happened with those patients. In Dagnelli 1996, there are two abscesses that get included in the Cochrane review, but in the actual trial they label these as, quote, unquote, imminent abscesses. I don't know what an imminent abscess is. Did an abscess actually develop? Did they get drainage? Like, I literally have no idea what that term means, but they just get thrown into the Cochrane review as if they were an abscess. So there are a lot of reasons, when I look at that Cochrane number, that I have very little faith in the idea that antibiotics are decreasing peritonslar abscesses.

[A]

You are imminently doubtful about this data. Casey, what's your take on the antibiotics for preventing the superior complications?

[C]

Yeah, so some of the best data we have comes from a really big observational trial done by Lidl in 2013, and that looked at 14,600 adolescents and young adults presenting to general practitioners with acute pharyngitis. And the overall rate of suppurative complications, and in Australia we call it quinsi instead of peritonsillar abscess. So if I say quinsi, that's what I mean. Otitis media sinusitis or cellulitis was about 1%. And that number feels a bit more accurate than the trials that Justin was talking about where it was sort of 5% or higher. And in this large observational data set, the rate of complications was identical, as Justin mentioned earlier, whether or not the patients got antibiotics or did not get antibiotics or got a prescription for delayed antibiotics in some of these trials. So that sort of tells us something about these complications. I think now you need to be really careful with observational data because this is not randomized data. The antibiotic group did look a little bit sicker in the little trial, and so they may be more likely to develop complications and the antibiotics seem to eliminate that extra risk. But based on the raw numbers, complications are relatively rare and were not particularly changed by the use of antibiotics. And rare is probably the key word that we need to talk about here, Anton, because if we're using antibiotics for prevention, as we mentioned earlier, most patients get no benefit whatsoever. But all of the patients that take the antibiotics are exposed to potential harms of those antibiotics. So if the complications are preventing, are preventing common issues, then using antibiotics seems like a no brainer. But if a complication is really rare, then you have a much bigger number needed to treat. In other words, more patients need to be exposed to antibiotics just to see one get a benefit.

[B]

And I spent a lot of time trying to sort this out because originally I thought this was going to be the crux of the issue. I have another thing that we're going to come back to, but I really wanted to know how many people I have to give antibiotics to. And so if you go back to the Cochrane review, there were a little over 15,000 patients and there were 25 total peritonslar abscesses. We can ignore the fact that I think some of those numbers are incorrect. So that would be about a 0.15% rate. Now, maybe antibiotics are driving that rate down. So if we just focus on the placebo group and we ignore the transcription errors, the highest rate I can possibly come up with is about 0.3%, which would mean very rough calculations here. We would need to have an NNT of about 1,000. I would have to give 1,000 prescriptions for antibiotics to prevent one Paraton sclerapsis. Now, I'll admit that's a very rough calculation, but that gives me a sense that even if there is a true scientific benefit from antibiotics, that benefit could pretty easily be outweighed by the harms. But as I mentioned, we haven't even touched on my biggest issue with this as of yet. The idea that we are preventing peritonslar abscess really just doesn't fit with my clinical experience. And I'm wondering about yours. I've managed many, many, many Quincy or peritons of abscesses in my career, and essentially all of them present with one or two days of symptoms at most. Like I've never seen a patient who's had a week worth of pharyngitis that wasn't given antibiotics and then that developed into an abscess. So the timeframe where I see abscesses really doesn't seem, make it seem like a complication of strep throat. It seems like its own primary presentation. And I'll tell you, I spent a lot of time trying to get good literature on, on that and it's harder than you would think. Almost none of the RCTs that look at peritonsal rhapsodists tell you how long patients had symptoms prior to presentations. But one of the RCTs in the Cochrane review does. I already mentioned it, Dagnally, 1996. And both of the peritons loraps in that trial presented on day two of the trial. Now in my mind, if this was really a complication, you'd see a lot more patients presenting later at day five, at day 10, because they have a complication from the pharyngitis. The fact that they presented on day two makes me wonder, did they not already just have the abscess on day one and we just missed it or was too, too small to see at that point? And so based on my clinical experience and this very limited admitted data, I think that peritonsalorapsis might be a different pathology altogether, not a complication. And I'm interested to hear your guys observations, but I think that distinction, whether we're treating peritonsalor abscess or preventing it, might actually be really important in terms of our decisions for these patients.

[C]

Yeah, I agree, Justin. I think there's probably some different virulence factors going on with some strains of streptin. If you get unlucky and you get one that's really good at forming abscesses, then you just arrive with a peritonsal abscess on day one, like you say. And the ones that we're seeing on day two or three of their illness, if they haven't already got the Quincy or peritonslamscis, sorry, then they're probably not going to get it. So I agree. It's probably something to do with the virulence of the various strains of strep that we're seeing in the community.

[B]

So, and then I'll add one little bit of data back talking about that same study that you talked about, Casey. There were two huge predictors of whether you were going to have a quote unquote supertive complication which were. And the two predictors were severe tonsillar inflammation and severe ear pain. And of course, the two common complications were Quincy or otitis media. To me, that doesn't sound like there were complications. That sounds like they were already present on day one. If you have severe tonsillar pain and the next day you show up with an abscess, I think probably you had an abscess.

[A]

You know, my experience is somewhat similar to yours, Justin's. You know, when it comes to timing a peritonsiller abscess presentation, usually they present in the first one to two days. I mean, that brings up the question if they do present on day one or day two with severe pharyngitis, you're not convinced they have a peritonsillar abscess. Let's say you even have a transesophageal pocus probe, and you go in there and you prove to yourself that there's no peritonsillar abscess. You know, do those patients require antibiotics? You know, for me, I'm kind of like, again, if they've got severe pharyngitis and they tick all the boxes that this, you know, very likely isn't viral and they're pretty sick and maybe they're on the edge of something deep, then I'll give them antibiotics. What are your thoughts on that?

[B]

Yeah, I think you've hit the nail on the head. To me, we really need to be clear about whether we are preventing peritonslar abscess or treating it. If you were preventing it, that was your whole goal. That would imply that you need to give antibiotics to everybody in a goal of trying to prevent it, because we don't know which patients are going to go on to have peritons or abscess. But I've already said the number needed to treat is to so high. That strategy doesn't seem very good to me. My read on this data is not that we're preventing it. My read is that we are finding early peritonsillar abscess and we're actually treating it. And there's actually pretty good data that antibiotics alone for peritonsular abscess is as good as draining the peritonsular abscess. So that makes sense to me. And if that is what we're doing, it implies a very different treatment strategy rather than broad empiric antibiotics for everybody. You focus just on the sickest patients. And so my strategy is I still don't give antibiotics to 99% of patients because I don't think that they need it. But in that patient that you're talking about, where they seem to have more unilateral pain, more severe, I wonder whether they're developing an early abscess. This does give me permission to use antibiotics in that case. I think that's an important distinction.

[A]

That's really interesting what you just said, Justin, about antibiotics versus drainage for peritonsillar abscess. Because that goes against everything we know about treating abscesses. I mean, whether it's a skin abscess or an intra abdominal abscess or an abscess anywhere, if it's more than a tiny little abscess, it is standard across all of surgery and medicine to drain that abscess. How do you reconcile that with studies that show that antibiotics versus drainage of peritontsal abscess are equal?

[B]

Yeah, it's interesting. The mucous membrane abscesses seem to respond differently than the skin abscesses. And maybe I'll make it a quick. I'll promise your audience, because there is a systematic review and meta analysis from 2020 that seems to show that medical management is identical to surgical management for peritonsillar abscess. Before I knew that, I had some heated conversations with ENT over the phone when I asked them to come in and see my patients and they said, bah, just put them on some antibiotics and I'll see them in my office. So it is an important bit of information to know so that you understand why ENT might be suggesting the things they are suggesting. I still tend to drain it if I'm pretty sure there's pus. There's. I do a needle aspiration just to get most of the pus out. But it is worth knowing. I mean, we're not talking About,000 person RCTs and, you know, giant, giant numbers of data, but There are multiple RCTs that show no difference between just putting somebody on dexamethasone and antibiotics as compared to actually draining the pus out.

[A]

Fascinating.

[C]

I'll give you my ultrasound take on that one, Anton. We do a lot of ultrasounds of, of tonsils in my practice just to look and see if there is an abscess there. And I'd say the most common thing that we see is someone who's been clinically diagnosed with a peritonsillar abscess. We look at the ultrasound and there isn't actually anything drainable there. Or there's a tiny little sort of 2 millimeter pocket of pus, which is really not Worth traumatizing your patient to try and clear. So, having used ultrasound for a while, I can sort of see why the data points that way, Because I think our old clinical diagnosis of peritonsillar abscess, or quincy, was probably not that accurate. And so I think we're actually treating a lot of these minor abscesses quite successfully with antibiotics in the past, maybe without even knowing it. And I think we've probably saved people a bit of trauma because it's not pleasant having your peritonsal abscess drained. And if you can avoid it, I think that's a reasonable thing to do. Although I think if. If you're to the point where you can't swallow and you've got major anatomical disruption from a really big abscess, those are the ones that I think probably are still worth draining in that a lot of the symptoms do seem to get better if you can reduce the mechanical effects of the abscess.

[A]

Yeah, that sounds perfectly reasonable. I mean, we know that for sure. We are not good at just looking at someone's throat and knowing whether they have a peritonsillar abscess or not. I mean, it's kind of similar to skin abscesses, unless it's totally obvious. We often think that patients just have a skin cellulitis and they actually have an abscess there. So, like many things in medicine, it's not exactly cut and dry, black and white. There's a lot of gray area in there. Just to summarize, that's a lot of information. Can you guys help me here Wrap up just in a sentence or two? Should we be relying on antibiotics to prevent superior complications of strep throat like peritonsillar abscess? You know, again, for me, giving antibiotics to everyone with strep throat to prevent peritonsillar abscess is not evidence. Is not evidence based. For certain, but in a certain subset of patients who present early with severe pharyngitis. And I'm worried about them based on all the clinical features, not just the centro score, et cetera, I might give antibiotics in that case. But certainly for all comers, the evidence is very weak, with an astronomically high number needed to treat. So we certainly shouldn't be giving antibiotics to prevent peritonsillar abscess for everyone. Anything to add there, guys, before we move on?

[B]

No, I think that summary is perfect. I think we just need to drop the word prevention from our vocabulary. That's not the goal. Just treat and then remember that periodontsylar abscess or these complications are rare, less than 1%. So this isn't permission to give antibiotics to everybody, but the worst 1 to 2% of pharyngitis that you see definitely warrants antibiotics. So when I say I don't give antibiotics for strep throat, this is not black and white. There is judgment. But this isn't like, I still give it to 75% because I'm trying to eliminate peritons. It's like the worst 1, 2, 3% that we're going to use antibiotics for, for this indication, at least.

[A]

All right, let's move on to another potential complication of strep throat, and that's glomerulonephritis. We all learned in medical school that glomerul nephritis is a bad kidney disease that is caused by group A, strep. So it seems logical that treating people with strep throat would prevent glomerulonephritis. Now, glomerulonephritis is even less common than peritonsillar abscess. So, again, I have a feeling we're going to run into some problems saying anything definitive about antibiotics preventing gnarly, because a big enough study could never be done. Justin, what's your take on the evidence for antibiotics preventing glomerulonephritis?

[B]

Yeah, this is a really interesting one. In medical school, I had this burned into my brain that the infectious disease does not believe antibiotics affect glomerulonephritis, whereas they do believe it affects rheumatic fever. I think it was something to do with strains and immunology that I didn't quite understand. So I had to memorize that for my tests. Multiple major guidelines that agree with what I was taught, at least. The Infectious Disease Society of America and the American Academy of Family Physicians both have very definitive statements that antibiotics have zero effect, no effect on glomerulonephritis. And that's what I expected. And then I went into the literature, and I don't know that the evidence is strong enough to make that kind of definitive claim. In the entire Cochrane Review, there was one, a single case of glomerulonephritis. And for what it's worth, it was in the control group, not the antibiotica group. So I don't know that you can make a scientifically literate statement of any sort at all. But I think we can say if there's only a single case in all of the history of all strep throat Literature. This is so rare that I don't think you're going to be able to make any clinical decisions based on this. So when you're focused on the patient in front of you, the answer is very clear. We're just not going to make a decision. We're not going to worry about preventing glomerulonephritis. It's too rare. It just doesn't happen. There's a different context that we're not used to in emergency medicine, the public health approach. And there are some arguments that if you have an outbreak of glomerulonephritis in your community, maybe that's a time for using some antibiotics. I've never seen an outbreak in Toronto, but that's sort of why we invited Casey onto our show.

[A]

Casey, you have seen patients with glomerulonephritis in your career and I understand there are outbreaks there. Tell us about that. And in the context of treating strep throat.

[C]

Yeah, I've had the unfortunate pleasure of living through a couple of epidemics of acute post trip glomerulonephritis in the northern parts of Australia. And certainly it seems to happen in clusters. And so we see that go through a school where you'll just have a heap of kids that suddenly get the clinical symptoms of glomerulonephritis. And I've seen it enough that I can actually pick it from the end of the bed without actually having to do too much because they come in hypertensive and puffy usually and you can sort of see that on their faces. So it's, it's a, it's an interesting clinical syndrome. Presumably there's some specific serovariants of the streptococcal bugs that have this particular immunological cross reactivity that attacks kids kidneys and we just see it happen from time to time. It happens every few years in the northern part of Australia. Now there's a bit of observational data from Johnson in 1989 that recorded nine previous outbreaks of post rep GN. And they concluded that targeting children in the community with antibiotics or just focusing on the children with the skin and soft tissue infections may help prevent subsequent cases. And the data is really hard to interpret and obviously there's no control groups because these are sort of outbreak situations, so you're not going to be able to run a trial. But glomerulonephritis is a tricky one because the actual kidney part of the disease happens about two weeks after the initial infection. So it can be really hard to put that back into the can. And you need a time machine to go backwards to treat the kids that were exposed two weeks ago, essentially. So that can be really difficult. And in Australia, when we've had these outbreaks happen in more recent years, basically they've had like a SWAT team of public health nurses that go out to the school and grab all the kids and give them antibiotics in that small community. So it works okay if you've got a small contained community or a family group, which is not too bad. So we do that. And I think we're not talking about prevention here, we're talking about really targeted prevention of just a small outbreak. The second point, which I've just mentioned, but is often overlooked and I think runs through the rest of what we're going to talk about today is that streptococcal disease involving the skin and soft tissue is really the problem. Like, we don't even really think about this in terms of a strep pharyngitis because that's just not the disease process that we see in Australia, where we have high incidence of this in the indigenous population. So this will come up again when we talk about rheumatic fever. But a sore throat that you may see in Toronto or in suburban Australia is not what we should be worried about. We need to be thinking more about pyoderma and skin disease, because that's where the money is. There's a little bit of data to back that up. In the, in that outbreak that I mentioned before the data, only two of the 25 cases, index cases, actually had a sore throat, whereas the rest had documented pyoderma as their precursing streptococcal infections. So that's really what we think about in terms of preventing these horrible diseases.

[A]

So it sounds like to me, I shouldn't really worry at all about glomerulonephritis in the usual low risk patients that I'm seeing on every shift. If I was in Australia and there was an outbreak, it would be a whole different story. Post strep throat, GN is very, very, very rare and antibiotics probably aren't going to change the risk for my patient at all. And that's interesting. So it's really about skin infections that you should worry about. So post strep skin infections are the ones that really cause glomerulonephritis?

[C]

Yeah, that's right. And the children that we see tend to have chronic skin infections which are usually related to underlying scabies infections as well as. And so these are the kids that we tend to see presenting again and again with chronic pusy sores on their skin. And these are the ones that tend to get sick. I honestly think that if you really want to prevent post glomerulonephritis, you're probably better off being either a public health physician or maybe a plumber, because I think access to clean running water and skin hygiene is probably more important than anything that we dull ed doctors ever dream about.

[A]

When we talk about strep. One issue that always gets me a bit nervous is invasive group A strep infections, because these are horrendous infections, life threatening. You know, where I work we've had cases of kids coming in, you know, being sick for 24 hours and then end up in the ICU very quickly. This is a very, very nasty, very, very quick acting entity, invasive group A strep. And every time there's an outbreak in invasive group A strep, I see everyone in our department getting really aggressive about swabbing and treating sore throats. Casey, what's the relationship between strep throat and invasive group A strep infections?

[C]

That's an interesting question. And the short answer is that there probably is no relationship between the incidence of pharyngitis and the subsequent invasive group A strep infections. And so you do need to have obviously strep on your body in order to get an invasive infection. And pharyngitis could theoretically be one of the sources of the invasive infection, but it just doesn't seem to work out that way. And in the recent European outbreak, they looked through all the 63 children that suffered with eyegas and they found only two of them had a recent pharyngitis. And that was in the erlakka paper in 2024. So, so it's, it's a pretty weak predictor, I think, of who's going to get invasive streptococcal disease.

[B]

Yeah, I look in a lot of places and unfortunately most studies don't even list this information because I think infectious disease doctors just know that eye gas isn't related to pharyngitis. So they don't even put the stat in their papers. But I did some again, back of the napkin calculations on a recent Canadian data set. This was Tyrell in 2024 and there was the biggest outbreak of invasive group A strep in Alberta's history occurred around the time of COVID and there were about 3,000 total cases of invasive group A strep out of 3,000, only 111 had a throat swab that was positive in the seven days prior. And that is in the context of what Anton says. As soon as you have an outbreak, everybody is swabbing and over swabbing as you're going to find a lot of strep carriers. And even in that context where everybody's over swabbing, only 3% of all invasive group A strep infections had a preceding strep throat. Doesn't sound like that's the source.

[C]

Yeah. And I guess the other way of looking at this is if you look at, in terms of pure epidemiology, invasive group A strep pretty much happens in outbreaks, but at any time of the year, whereas strep throat is very much a seasonal disease. So we tend to see it, you know, more through the wintertime. Of course, if you live in a tropical country like Australia, we don't have a winter, so that stat may not work. So it is what it is. So if you really thought that invasive streptococcal disease was a result of strep throats, you would expect to see a nice overlay where it peaked in the winter. But that isn't what we're seeing in terms of our strep throat pathology.

[B]

Yeah. So the data does not seem to be good. But the inclusion. What every infectious disease doctor has ever said during every outbreak that I've lived through, they've always said the exact same thing during an invasive group A strep outbreak. Do not change your management of sore throats at all. They're not worried. Invasive group A strep is not coming from the strep from a strep throat. It's not coming from pharyngitis. That shouldn't be our focus.

[A]

All right, let's move on to the dreaded rheumatic fever. Now, this one scares me too, because it's really easy to miss in the edge and can have some devastating consequences for the patient. And again, this one's in North America, at least, pretty darn rare. So, Justin, do antibiotics for strep throat prevent rheumatic fever?

[B]

You might be surprised given how wishy washy my answers have been so far, but the short answer here is yes.

[A]

Did you just say yes, Justin? Wow.

[B]

Absolutely.

[A]

That's the first time in a long time you said yes to me.

[B]

We're gonna have to add some caveats to how we actually use it in modern standards. So the data here is very old and it's still imperfect, but I Think there appears to be a very clear decrease in rheumatic fever when you're using antibiotics. In the Cochrane review, they have 12 studies that include 12,000 patients and the risk of rheumatic fever goes down by an odds ratio of 0.32. But we always want to absolute numbers. And so in these trials back in the 1950s, the rate of rheumatic fever without antibiotics was about 3% and antibiotics reduced that to about 1%, a 2% absolute reduction. Although there are some issues, I do tend to believe that is a real decrease. But there's one massive problem that we have to consider. We do not see anything like a 3% rate of rheumatic fever in most areas of high income countries in 2026. In fact, they even say that in the Cochrane Review, they say that they can't really make an estimate of modern benefit because there has not been a single case of rheumatic fever in any study done since 1961. So the real question in the modern era is not whether antibiotics work, because they do. It's whether the population you are looking after has any risk at all of rheumatic fever or whether that risk is high enough to warrant antibiotic therapy. And I have some data on this, but perhaps before we get into the data again, I think I've never seen acute rheumatic fever in my practice in Toronto. You can tell me if you have, Anton, but maybe we should let Casey talk us through a little bit because again, he works in a slightly different environment than we do.

[A]

Like ditto for me. Justin, never seen a case. I understand there are some cases in the far north of Canada, but never seen a case in Toronto. Casey, what's your experience?

[C]

Yeah, sadly we say this quite a lot. I work in a population that includes a lot of, you know, well, affluent families, but also many Aboriginal kids from remote communities who are very high risk for rheumatic fever. The prevalence of rheumatic fever in my part of the world is about one to two per thousand school age kids, so, you know, orders of magnitude greater than what you have in Toronto. I imagine it's still not super common, but it's enough that we see it regularly in the emergency department. I saw a girl with Sydenham's Courier just last week in the emergency department, which is remarkable to say in living in a rich developed country like Australia, but it's sadly common enough that we actually have a protocol for how to work that up and treat it. It's a terrible disease. And the cardiac consequences of rheumatic fever are a huge burden on the young aboriginal people in the north part of Australia and parts of New Zealand and Southeast Asia as well. And prevention is definitely worth it because this is a major morbidity with lifelong complications for relatively young, otherwise healthy people. And we have a pretty strong public health program that tries to provide secondary prevention for people that have already had an episode of acute rheumatic fever. And we also do screening for rheumatic heart disease in young people who may have been missed when they had their index case. So. So this is a very different practice, I imagine to what your, your listeners are used to seeing. When I think about antibiotics for prevention of acute rheumatic fever, assess the risk, ask about risk factors like overcrowding, chronic skin disease, look at nutrition and, and other things. But we have a very low threshold for offering treatment to all young aboriginal people who are at high risk. In my practice though, if you're otherwise low risk and non indigenous, you probably get exactly the same treatment that you guys would be offering over there in Canada.

[B]

Yeah, so this is going to be the key is to figure out what patients actual risk is because to me you go back to the data. In the 1950s, if my patients had a 3% risk of rheumatic fever, I'm giving antibiotics to everybody. That's high enough that it just does, it doesn't matter. It's not like everything else we've talked about so far. Right. I don't think there's probably any effect or that the harms probably outweigh the benefits. When we're talking about symptoms. It doesn't seem like we're helping in peritonsalorapsis or glomerular nephritis. But here we might make a difference if you're looking at the right population. And so let's try to figure out what the risk actually is. And unfortunately rheumatic fever is actually not a reportable disease in almost any country. And so we don't actually have very accurate numbers to work with. In most high income countries rheumatic fever has basically 100% disappeared in the last hundred years. I have some graphs in my write up if people want to look. But the graphs are remarkable. In 1900 the rate of rheumatic fever was more than 200 cases per hundred thousand. By 1962 it had dropped to 10. So in just 60 years went from 200 to 10. And now in modern day days the numbers are so low they're not even readable on that scale. Now I think we need to hit something early because A lot of people might think that maybe this drop is because of antibiotics, maybe because this is a strong argument for using antibiotics because we're getting rid of rheumatic fever. But actually, if you look at the graphs, rheumatic fever started disappearing in the mid-1800s, well before we had discovered any antibiotics. We didn't even know that streptococcus existed when rheumatic fever started to disappear. There's a question why that is, and the best answer that we have right now is that this is entirely driven by the stuff Casey was talking about as risk factors. It's living conditions, crowding, hygiene, other socioeconomic factors. It's not the antibiotics that's driving, driving this down. So I don't think it's an argument for using antibiotics. I still think you want to use antibiotics if you're in a high risk community. So let's focus on Canada, because that's where you and me are. And Anton, the most recent numbers I could find were in 1990. At that point, Canada wide, there were three cases of rheumatic fever per million. And if you remember, initially I said 200 per hundred thousand. So we've changed the scale here. Instead of 200 per 100,000, we're now talking 3 per million. And I think the thing to recognize is that's an average across all of Canada. That includes the very high risk populations. And unfortunately, like the aboriginal populations in Australia, our first nation populations still have a very high risk. We're talking about 20 per 100,000 or 200 per million. 200 as compared to three for the rest of Canada. And that three includes those, those first nation populations. So if you're working in suburban Toronto or Montreal or Halifax, your risk is essentially zero. You will work an entire career without seeing a case of acute rheumatic fever in those populations. I'm pretty convinced antibiotics cannot help. The benefits cannot outweigh the harms in those populations. However, if you're working with our first nation populations in Canada or any other high risk population, I think your approach needs to be very different. Now, Casey's already mentioned this. I don't think that antibiotics for sore throat is necessarily the solution. My read on this data is that a lot of this is a societal issue. It's a public health issue. It's better housing, better access to showers, baths, less crowded living environments. But if I was working in that high risk population, and I have, when I worked in New Zealand, our Maori population had a very high risk. I was broadly using antibiotics, empiric, amoxicillin, for pharyngitis in that population. But in Toronto, I just don't. And I think that's the distinction. You need to figure out what your local risk is to decide whether this is a worthwhile thing.

[C]

In terms of northern Australia, we're very liberal with treating streptococcal disease. As I say, we focus mainly on things like skin sores and scabies. We typically treat these with oral antibiotics or we tend to use a lot of long acting intramuscular penicillin because it's a treat and street sort of drug where you can give it and you get a week's worth of of penicillin in one injection. It's not very popular with the kids, but it is what it is. And most of the kids that we see are chronically colonized, so we usually don't bother sending off swabs from their skin or their throat. We do a lot of testing of the ASOT titers to confirm exposure because that's one of the Jones criteria that you have to satisfy to diagnose acute rheumatic fever. But to be honest, in my career working in the northern part of Australia, I've never seen a kid with a negative ASOT titer in that this, these kids are chronically in a state of immune response to strep. It's just part of their life, essentially. So it's probably that chronic exposure that creates the syndrome rather than any sort of acute single infection that we as ED doctors may treat.

[A]

So let's get to the bottom line on who needs treatment and who you might swab. And I'll go first. I'll just say my bottom line is that antibiotics certainly should not be given routinely for strep throat. We really need a much more nuanced approach to our decision making if we're going to give antibiotics. And I have a very short list of patients. You know, one is those that are at risk for complications like immunocompromised patients and for rheumatic fever. If I'm working in an area that's very high risk, I'm going to be going by my local protocols and giving antibiotics for those patients who present early, like within the first 48 hours with severe pharyngitis. And, you know, they might have some maybe red flags on history or physical that make me maybe worried about a deep space infection, or they might already have a non drainable abscess or simply my gestalt says, I think they really need antibiotics. Those patients I'm going to give antibiotics to and then there's those patients that I really don't think need antibiotics, but they may have been sent in for antibiotics. They have really high expectations for getting antibiotics and maybe they just don't understand my rationale for not giving antibiotics or for whatever reason, they just won't accept my explanation for why they don't really need the antibiotics. There are some of those patients that I end up giving antibiotics to, even though I know I shouldn't really be doing that. So that's my honest opinion of who I'm going to be giving antibiotics to. Justin, what about you? Who do you give antibiotics to?

[B]

Yeah, I don't disagree much. I think the summary was in everything we talked about. If you're looking to try to improve symptoms, I'm pretty convinced that you're causing more harm than you are.

[A]

Good.

[B]

So I'm not using antibiotics for patients symptoms. In terms of all the complications, we don't think that we're doing anything for invasive group A strep. We don't think we're doing anything for glomerulonephritis. I'm not preventing peritonsillar abscess, just like you. If somebody's presenting early and seems to be more unilateral, seems to have the most severe, maybe, maybe the most severe 1 or 2% of patients that I see, I'll be a little bit more liberal with antibiotics, thinking that I might actually be treating an early peritonsillar abscess. And then for me, the crux of the issue is, is the patient at any risk for rheumatic fever? In Toronto, I have gone most of my career basically using zero antibiotics for strep pharyngitis because we have no risk here. I lived for a year in New Zealand and I basically gave everybody antibiotics because there was risk. And I think at that point it tipped over into benefits outweighing the harms. And so I think most likely, if you live in a community at risk of rheumatic fever, you already know it, but if you don't, it's worth looking into a little bit and it's worth thinking about the sub populations around you that might be at higher, higher risk, because that to me, is the crux of the issue.

[A]

Wow, Justin, we actually seemed like we agreed there. It's amazing.

[B]

I will say that patient expectations almost never push me over the edge. I'm not going to take a patient complaint to the college for this, but when I tell patients that harms outweigh benefits, they tend to believe me. So I'm not sure about that as an indication, but we'll leave it there.

[A]

Okay. Well, maybe I need some tips from you, Justin, on how I should be counseling patients.

[C]

It would be an interesting study to look at how you go throughout your shift. If you're more likely to put up resistance at the start and at the end of the shift, if you're just giving in and handing out scripts sent on, because maybe it's that that's what's going on, how busy the rest of the department is. I think that often weighs on our minds more than the actual science. Unfortunately, I would agree with you. I think that I'm a very simple doctor. I have a spectrum. If you're way over on the right end of the spectrum in terms of the severity at the time of your presentation, then you know you're probably going to get some antibiotics. If you're, if you're already looking like you've got a peritonsiller abscess, or you, your throat's so swollen that you can't drink and you come in dehydrated. In my world, you're going to get admitted to the short stay ward. And if you're in that very end of the spectrum, then you can have some antibiotics. But I don't routinely prescribe them for anyone for symptom control. And in terms of preventing the non separative complications such as glomerulone, arthritis or rheumatic fever, even my high risk population, I'm really not that interested in this sore throat. I'm more interested in what their skin looks like because I honestly think that's more likely the source of the badness. So if you've got active impetigo, then you get treated very liberally.

[A]

Okay, so that's our opinions based on the world literature on strep throat when it comes to giving antibiotics, then the second closely related question is who to swab? Casey, I think we already know your answer. You said at the top that you don't swab anyone. Do you have any modifications to that? After our long discussion?

[C]

I think that the swabbing question is a really interesting conversation to have around the evidence based medicine side of this discussion in terms of what we think of as test thresholds. So if you're a well person who isn't from a high risk population, and you know, you don't come from a community where it's a high background rate of all these horrible complications, you're already well below the test threshold. Because as we've just discussed, we don't think that giving antibiotics to treat your potential strep throat is going to make a difference. So it's a wasted swab, in my opinion. And if you're a person from a high risk background that is at high risk of things like glomerulonephritis or acute rheumatic fever, then you're already well over that test threshold. And the swab, even if it's negative, is not going to stop me from giving you antibiotics to try and prevent that complication. So I think the swab doesn't really help in either of those situations.

[A]

That was an excellent explanation. That totally makes sense. Justin, any indication for swabbing pharyngitis in your practice?

[B]

Yeah, I might end up just saying what Casey said in different words. I want to start with two things. Number one, I am a little bit biased because there's nothing I hate more, more than having a long list of results that I need to follow up, especially if they were sent by another doctor and they're just not going to change management. And now I'm phoning patients all shift long rather than seeing new patients. So I sort of hate these swabs because they come back two days later when you're not not at work to begin with. So that's my bias going into it. I think. The other thing that's interesting as a general observation is you said earlier that you get the feeling that everybody in North America is swabbing liberally. But I've been working on this topic for a while and I know that at least a third of doctors in both departments I work in never swab for strep throat. So we work in our little bubbles. And I don't think we always know the degree of testing that everybody is doing. And I think it's worth knowing that because sometimes you feel some external pressure to do things when that external pressure isn't actually there. And it's worth talking about these kind of things a little bit more. To me, the answer to the swab is basically what Casey says. But you can put it just about everything we talked about. If you were trying to treat symptoms, I don't even think you can treat symptoms. But a swab 48 hours later isn't going to help you with in the management of symptoms. It's not going to help you with the management of a peritonsillar abscess that occurs in the first one or two days. And we don't think that we can prevent eye gas or glomerular nephritis. So really the only area where the swab could possibly come in is rheumatic fever. Now, rheumatic fever is interesting because the majority of cases do occur, occur after that 48 hour period. So the swab will catch a lot of them, but not all of them. About 10 to 15% of rheumatic fever will occur in those first 48 hours. And I just, I tend to agree with Casey. In the two populations that I have worked in, in Toronto, I will argue very strongly there is zero role for a swab because these patients have zero risk of rheumatic fever. So what the heck are you doing? If you really want to treat that most severe 1%, treat them on day number one and send the antibiotics. But then the results of the swab two days later are irrelevant. Relevant. We could get into an argument in the north of Australia or in rural New Zealand, where I've, where I've worked, about whether you should swab to try to limit some of the antibiotics, even when you're being liberal about antibiotics to prevent rheumatic fever. But to me, as much as I talk about the harm of antibiotics, we're talking about penicillin or amoxicillin, I think you're probably just better off being liberal with antibiotics in the communities where this is a risk and pretty strict with antibiotics in communities where rheumatic fever is not a risk. And a swab 48 hours later is just not going to change that practice pattern for you. So I have been in a zero swab group, but I think just as long as people are thinking through their practice, if you're swabbing specifically for rheumatic fever, then it could make some sense and that's fine with me.

[A]

Great. I have to say, I just really appreciate both of your intellects. You guys are super smart and totally awesome. I really enjoyed that. So thank you.

[B]

We're very good at lying on podcasts or painting ourselves under the good brush.

[C]

I've been riding on Justin's coattails for 10 years now, so it's doing me.

[A]

Oh, no, you're just being humble, Kasey. All right, thanks so much, guys. That was super fun and totally interesting. And what I'm hoping is that everyone out there can reflect on their swabbing and antibiotic prescribing of sore throats. And if you have any thoughts or comments, please email me or on the EM Cases website, show notes, just throw in your comments there and we'll do our best to respond to any questions you might have. Thanks, guys.

[B]

Thank you so much for having us. Anton and thank you to everybody who stuck out to the end of like an hour or more on just strep pharyngitis. Really appreciate it it.

[C]

Thanks for having me. It's been a pleasure.

[A]

Anton, fantastic.