Outbreak Watch: August 28, 2026
- Heather McSharry, PhD
- 23 hours ago
- 17 min read
Summary
In this Outbreak Watch, we return to the outbreaks and signals we’ve been tracking—and several have changed. New genomic evidence is revealing the hidden early months of the DRC’s Bundibugyo Ebola outbreak, while Uganda has successfully ended transmission across the border. New World screwworm is worsening in Sonora as cattle imports into Arizona resume, and COVID-19 is climbing again across the South amid changes to how CDC categorizes and displays wastewater activity. We also check in on measles, Cyclospora, H5 avian influenza, Vibrio vulnificus, mpox, Oropouche, and other signals on the board—and close the loop on the quarantined riverboat near Kinshasa.
Listen here or scroll down to read full episode.
Full Episode
Welcome back to Outbreak Watch.
It's been about three weeks since our last full check-in, and quite a bit has moved. Tonight we're going back through the active outbreaks and signals we've been following, because some have changed substantially, some haven't crossed the thresholds I've been watching for, and one or two new signals deserve a place on the board. Let's start, as we have for most of this summer, in the Democratic Republic of the Congo.
ACTIVE OUTBREAKS
When we last checked in on August 7, the Democratic Republic of the Congo had reported just under 4,000 confirmed cases of Bundibugyo Ebola and about 1,800 deaths.
As of August 25, that number has climbed to 5,713 confirmed cases and 2,744 deaths, with 1,269 people recovered. The outbreak now affects 58 health zones across six provinces. The crude case fatality rate is about 48 percent.
So this isn't just the largest Ebola outbreak ever recorded in the DRC. As measured by reported cases following recognition of the outbreak, it 's being called the "fastest growing Ebola outbreak ever recorded".
But I want to clarify that.
Over the past several weeks, genomic analysis has given us a much clearer picture of what was happening before this outbreak was officially recognized. And a new analysis released this week strengthens that picture considerably.
The earliest genomic work, based on 139 Bundibugyo virus genomes, suggested that the outbreak's common ancestor probably dated to around March, with uncertainty extending earlier. But researchers now have a much larger dataset. Their latest analysis began with more than 600 genomes, with 525 included in the principal time-calibrated reconstruction. That analysis places the outbreak's most recent common ancestor around late February, with a credible interval extending from mid-January through late March. And importantly, that estimate has remained very stable as researchers have added hundreds of new genomes.
So we can now say with a lot more confidence that this outbreak was already underway well before it was officially recognized in May.
But the genomic reconstruction tells us something else that's important. This doesn't appear to have been an explosive epidemic from the beginning. The evidence instead suggests relatively slow initial growth, probably centered around Mongbwalu, followed by acceleration from March through June as the virus was exported into larger urban areas, including Bunia.
And that gives us some important context for that description we've seen attached to this outbreak: the fastest-growing Ebola outbreak ever recorded.
The extraordinary rise in reported cases after the outbreak was recognized is real. But that doesn't mean the virus suddenly became extraordinarily good at spreading.
In fact, the genomic researchers report no evidence of a fitness-altering mutation and no specific viral lineage taking over the epidemic. Instead, we're seeing the outline of an outbreak that had been building basically unseen for months—at first relatively slowly, then accelerating as it reached larger populations—and was already mostly established by the time the alarm was raised in May.
And there's another potentially encouraging piece of the genomic story. Although cumulative case counts have continued to climb, the reconstruction suggests that the epidemic's underlying growth began to flatten after June. The researchers specifically tested whether that apparent stabilization could be an artifact of declining genomic sampling later in the outbreak, but their analysis suggests the broader pattern is robust. That doesn't mean transmission has stopped, or even that this outbreak is under control. But it may be one of the first indications that its underlying trajectory is beginning to change.
So I think we need to separate three things that can sound contradictory but aren't:
the cumulative number of cases is still rising
newly reported cases appear to have stabilized somewhat
and the genomic reconstruction also suggests that the epidemic's underlying growth has flattened since June.
So, the bigger explanation for how we got here still seems to be operational rather than evolutionary. Transmission got ahead of surveillance. Cases were occurring outside known contact networks. Patients were reaching care late or not at all. Treatment capacity became overwhelmed in some areas. Health workers faced shortages and delayed pay. And conflict, displacement and insecurity made some communities too difficult to reach.
None of that makes this outbreak any less serious. But it does mean we're watching what Ebola can do when it gets a months-long head start and the systems required to find cases, trace contacts, isolate infections and interrupt transmission struggle to catch up. And the genomic data are allowing us to reconstruct that hidden head start.
But there have been a couple of genuinely encouraging developments.
The first is across the border in Uganda.
Today, WHO and Africa CDC formally announced that Uganda has completed the 42-day period—twice Ebola's maximum incubation period—without another confirmed case linked to its last patient. Uganda recorded 20 cases during this outbreak, including 15 imported from the DRC and five locally transmitted from those cases, with two deaths. So Uganda's outbreak is officially over.
What a contrast, though. The same virus crossed the border repeatedly. Uganda detected the introductions, followed contacts, isolated cases and successfully interrupted onward transmission.
That's evidence that this virus can be contained when the response gets ahead of it.
But the second development I want to mention may be even more interesting scientifically. In an effort to protect frontline workers and other high-risk people during this Bundibugyo outbreak, the DRC has begun using Ervebo, the licensed vaccine developed against Zaire ebolavirus. More than 70,000 doses have reportedly been made available, with some also being used in clinical research to determine whether the vaccine provides meaningful protection against Bundibugyo.
And separately, today, CEPI (the Coalition for Epidemic Preparedness Innovations) announced additional support for another experimental candidate moving toward clinical testing.
And one quick follow-up from our last Outbreak Watch: remember the riverboat that was quarantined just outside Kinshasa after a former passenger died with Ebola-like symptoms? Well, no Ebola was detected among the passengers who were tested, and I haven't seen evidence of transmission associated with that incident in Kinshasa. So that particular alarm, thankfully, appears to have been a false one.
OK, so my Ebola watch points have changed slightly. I'm still watching geographic spread and whether contact tracing begins catching up with transmission. But I'm also watching the epidemic curve more carefully now. If reported new cases begin falling, we need to distinguish a real reduction in transmission from changes in case finding. I'm watching whether those 58 affected health zones keep increasing. And I'll be watching very closely for the first meaningfuler data from the vaccine effort.
Because right now, this outbreak is still not contained. But for the first time in several weeks, we're beginning to see some tools moving into place that could change the trajectory.
New World Screwworm
Next, New World screwworm—and this one has become considerably more complicated since our last episode. Let's start with the United States.
The good news is that I have not found a major new geographic expansion here. Texas still reports 16 counties with confirmed cases, although the number of affected Texas premises has now increased to 37. USDA's public confirmed-detection dashboard has not reported a third affected state, a confirmed wildlife infestation or a positive wild-screwworm fly trap that would fundamentally change the U.S. picture. So, at least for the moment, I would still describe the U.S. situation as geographically contained.
But the story immediately south of the border has changed quite a bit. On August 19, Sonora confirmed its first New World screwworm case. At that point, an isolated introduction was plausible but that's no longer the picture. As of August 26, Sonora is reporting seven active cases: six in the municipality of Álamos and one in Yécora.
Those places are still hundreds of miles south of Douglas, Arizona, so I want to be really clear: this is not evidence that screwworm is suddenly approaching the Arizona border.
But six cases in Álamos plus another in neighboring Yécora is materially different from one infected animal. It raises the possibility that we're now looking at a small local transmission focus in southeastern Sonora, or repeated nearby introductions into the same corridor, rather than one aberrant case.
And the timing is uncomfortable because on August 24, the United States reopened the Douglas, Arizona, port of entry to Mexican cattle. Of all the entry points, they opened the one bordering Sonora. So, more than 700 cattle crossed the first day, and traffic has been ramping up under enhanced inspection protocols. Every animal is inspected and treated under the new import system, and Arizona has added its own tracking and screening requirements. USDA has not paused the reopening. This doesn't mean imported cattle are suddenly unsafe. The whole point of the inspection system is to create multiple barriers between infected animals and the United States. But epidemiologically, source pressure matters.
USDA made the reopening decision partly because Sonora and Chihuahua were considered comparatively low-risk states with strong animal-health programs and significant geographic distance from the main Mexican outbreak.
Sonora going from essentially no established problem to seven active cases so quickly, is important, even if the cluster remains pretty far south of the border.
But Mexico and Sonora are responding aggressively. More than seven million sterile screwworm flies have reportedly been dispersed in the state, including roughly 2.5 million around Yécora. Which means Sonora has now become a really interesting real-world test of the sterile insect technique and whether the border protocols are effective.
So, can an intensive, geographically targeted sterile-fly campaign extinguish this cluster before it moves? I guess we're gonna find out.
There is better news next door in Chihuahua, though. The active burden there has continued to decline overall, although there has been occasional recurrence in previously quiet municipalities.
Nationally, Mexico reported 2,015 active cases and 39,759 cumulative cases as of August 26. That's a modest increase from the approximately 1,970 active cases being reported immediately before the border reopening, so the tentative plateau we were hoping to see hasn't really held yet.
My assessment tonight would be: the United States remains relatively stable; Chihuahua is improving but still transmitting; and southeastern Sonora has become the new watch point.
The threshold I care about most at this point is geographic. Another case in Álamos matters, but it doesn't fundamentally change the risk picture.
A new affected municipality farther north—like movement toward Hermosillo, Agua Prieta or the Arizona border—would change the risk picture and be concerning.
The U.S. Cyclospora Outbreak
Now to Cyclospora, where the numbers are still climbing even though the contamination event itself appears to be mostly behind us. But I mean, who the fuck knows, amiright?
The iceberg-lettuce outbreak linked to Taylor Farms de Mexico has continued growing as state and federal investigators work through the reporting backlog.
FDA's August 27 update counted 11,458 confirmed illnesses in 20 states, with 495 hospitalizations and two deaths.
But those numbers are way behind state reporting. Michigan alone has reported 14,510 cases and 356 hospitalizations, so check your state's public health reporting for information relevant to you.
FDA says the recalled lettuce associated with this outbreak should now be off the market. Most illnesses began before the July 17 recall, and because Cyclospora has an incubation period and outbreak investigations have a substantial reporting delay, new cases can continue being added after the contaminated food itself is gone.
So a rising case count doesn't necessarily mean people are continuing to encounter contaminated lettuce today. But can we trust these guys? I don't know. I guess we'll see.
At this stage, we're still wondering if investigators can determine exactly where and how contamination occurred. The FDA has been conducting inspections and sampling at Mexican lettuce-growing operations but we've gotten no information on the specific contamination pathway that caused this.
So I'll keep this one on the active list for now, but unless a new exposure source appears or the investigation identifies an important underlying cause, I'm hoping we're getting closer to moving this one off the watch board.
GLOBAL SIGNALS
Now let's move to the slower-moving biological signals I've had my eye on, starting with
one that's actually changed.
H5 Avian Influenza: I've said throughout this series that individual spillovers into mammals aren't the threshold I'm waiting for. H5 already infects mammals and we've been seeing that. What I'm looking for is evidence that the virus has become better at transmitting between mammals, specifically, sustained chains of mammal-to-mammal transmission or biological adaptations that would meaningfully alter that risk.
We have not crossed that threshold. But Australia just gave us another data point worth logging. Australia has confirmed its first H5 infection in a mammal on the mainland: a long-nosed fur seal found dead in South Australia.
H5 arrived in Australian wildlife this year through migratory birds and has since spread through multiple wild bird species. They've already seen devastating impacts among seals on sub-Antarctic Heard Island, but this mainland detection shows the virus entering another mammalian host as it expands through Australian ecosystems.
Again: one infected seal is not evidence of seal-to-seal transmission.
The animal could have encountered infected birds or a contaminated environment. But every mammalian infection gives influenza another opportunity to replicate in mammalian cells. Most of those opportunities go nowhere interesting. Occasionally they produce adaptations worth paying attention to. So the H5 signal stays exactly where it was: no evidence yet that the fundamental risk has changed, but the number of biological opportunities continues to grow.
Mpox: For mpox, there hasn't been a major change in the signal we're tracking. WHO issued an updated global risk assessment this week, but I haven't seen evidence of a new geographic or transmission shift in clade I that changes our assessment.
There is new activity in Guinea-Bissau, where ten cases have been confirmed and community transmission may be occurring, but sequencing identifies those viruses as clade IIb, not the clade I virus we've specifically been watching. So: worth noting, but no threshold crossed.
Oropouche: And Oropouche is even simpler tonight. I haven't seen a significant new development that changes the picture we've been following. The signal remains whether the virus establishes sustained transmission in additional regions outside its historical range. For now: still watching, no major change.
SURVEILLANCE MILESTONES
I have one important follow-up from the last Outbreak Watch and two new ones I'm bringing in.
Measles: When we last talked, the United States had already surpassed its entire 2025 measles total. That trajectory has continued. By August 20, CDC was reporting roughly 2,777 confirmed cases nationally, making 2026 the worst U.S. measles year in more than three decades. And this week Pennsylvania's Department of Health announced two measles-associated deaths in Lancaster County, both in unvaccinated people.
We now know that one was a newborn who acquired measles before birth. The Lancaster County coroner says the baby's immediate cause of death was a lacerated spleen, although postmortem testing confirmed congenital measles infection. Pennsylvania health officials are therefore describing this carefully as a measles-associated death, and at this point we don't have enough evidence to say exactly what role the infection played in the splenic injury.
But this has also prompted a lot of discussion about whether people are dying from measles or merely with measles, and that's a distinction we need to handle carefully.
Measles doesn't have to be the final event on a death certificate to contribute to someone's death. The virus can cause pneumonia and encephalitis directly, but it also profoundly disrupts the immune system. Measles can erase portions of the immune memory built through previous infections and vaccinations, leaving people more vulnerable to other infections for months to years afterward.
In fact, many measles deaths historically have resulted from complications and secondary infections rather than the acute viral infection alone.
So we should absolutely be precise about what we know in these two Pennsylvania cases—and in the newborn's case, there are still important unanswered questions. But we shouldn't mistake that uncertainty for evidence that measles is harmless.
It's not.
And with 2,777 confirmed U.S. cases as of August 20, this remains the country's worst measles year in more than three decades, just as millions of children are heading back to school.
Now the first new surveillance signal is Vibrio vulnificus
I missed this by essentially hours during our last Outbreak Watch. Right around the time that episode came out, Louisiana announced that it had recorded nine Vibrio vulnificus infections this year—and five of those patients had died. All nine had been hospitalized, all had underlying health conditions, and all of the infections were associated with wounds exposed to seawater. For comparison, Louisiana has averaged about seven cases and one death during the same period over the previous decade.
I haven't seen additional Louisiana deaths reported since then, but Florida's numbers have continued to rise. As of August 20, Florida had recorded 17 cases and three deaths this year.
So this remains a relatively small number of infections, but a very serious Gulf Coast summer risk, especially for people with underlying health conditions who expose open wounds to warm coastal or brackish water.
COVID-19
And the last domestic signal I want to put on the surveillance board tonight is COVID.
COVID activity is rising again across the United States, particularly in the South and West, with wastewater surveillance showing substantial summer transmission in parts of the country.
And here in Texas, we're beginning to see some real-world effects of that increase. San Perlita Independent School District in South Texas temporarily closed all of its campuses this week because of a growing number of COVID cases, saying the closure was intended to protect the school community and prevent further spread. A smart move. And last week, Park Ridge Achievement Academy in Louisiana temporarily closed for a day after a COVID outbreak affected students and staff.
Two school closures don't make a national trend. But they reflect increasing infections in actual communities, and in some places those infections are already causing enough illness to disrupt schools.
There's also an interesting surveillance wrinkle here that I want to mention. Earlier this year, CDC changed both, the thresholds it uses to classify wastewater activity and the colors used to display those categories.
Under the previous system, for example, a wastewater activity level above 4.5 was classified as high. Under the current system, that same value can be classified as moderate, because “high” now begins above 5.3. The threshold for “very high” actually moved slightly in the opposite direction—from above 8 to above 7.8—so this isn't just a case of CDC redefining everything downward.
But when surveillance systems change their categories or visualizations, the words and colors on a map can change even when the underlying biological signal doesn't.
So for COVID, whether a map says “moderate” or “high” isn't as important as the trend itself. And right now, that trend is upward.
It's also important not to equate relatively low rates of severe acute disease with low overall risk. Long COVID and other post-acute complications can follow infections that were initially mild, and evidence shows that risks and burdens can accumulate with reinfections, affecting multiple organ systems.
So hospitalization remains an important measure of COVID severity but it isn't the only one. More transmission means more infections, more reinfections, and more opportunities for both acute and long-term harm.
This isn't some fundamentally new COVID threat. It's another summer wave with the same risks as last years. But it's one worth paying attention to and one where this year the underlying trend tells us more than the color of the map.
That's all for tonight's Outbreak Watch.
I'll continue keeping an eye on these stories and if the evidence changes, I'll bring you the update. And before I go, stay tuned—I’ll be revealing the newly redesigned Infectious Dose website soon, including a new Outbreak Watch tracker map where you’ll be able to follow some of the outbreaks and signals we talk about here as they evolve.
Until next week, stay healthy, stay informed, and spread knowledge not diseases.
ANNOTATED CITATIONS
Accessing the literature: Whenever possible, I've linked directly to free full-text articles and books that are legally available online. For subscription-only journal articles, many researchers are happy to share a personal copy of their work if you contact the corresponding author. For books, I've included links to free online lending copies or library catalogs when available. If a title isn't freely available online, I've linked WorldCat to locate it at a nearby library or you can ask your local library about interlibrary loans.
Each citation is annotated to explain how the source connects to the episode and to help listeners who want to explore the science in greater depth.
Democratic Republic of the Congo Ministry of Health. 2026. Ebola Situation Report No. 103.
🌐 Free online resource https://insp.cd/sitrep-n103-mvebdb-25-08-2026/
➡ Official outbreak situation report providing the most current case counts, deaths, affected health zones, and operational updates from the DRC Ministry of Health.
USDA Animal and Plant Health Inspection Service (APHIS). 2026. Current Status of Confirmed U.S. Cases of New World Screwworm.
🌐 Free online resource https://www.aphis.usda.gov/animals/animal-health/livestock-and-poultry-disease/current-status/us-confirmed-cases-new-world
➡ Official USDA dashboard tracking confirmed New World screwworm detections in the United States and summarizing the federal response.
Michigan Department of Health and Human Services. 2026. Infectious Disease Outbreaks.
🌐 Free online resource https://www.michigan.gov/mdhhs/keep-mi-healthy/infectious-diseases/infectious-disease-outbreaks
➡ Official Michigan outbreak dashboard providing current cyclosporiasis case counts, hospitalizations, and ongoing investigation updates.
Centers for Disease Control and Prevention. 2026. Measles Cases and Outbreaks. CDC.
🌐 Free online resource
https://www.cdc.gov/measles/data-research/index.html
➡ CDC’s current national surveillance page for U.S. measles cases and outbreaks. Provides regularly updated case counts, outbreak information, geographic distribution, and historical comparisons used to place the current resurgence in the context of U.S. measles activity over previous decades.
Kamale, J.-Y. 2026. All passengers on quarantined river boat test negative for Ebola, authorities say. Associated Press, 9 August 2026.
🌐 Free online resource
➡ Reports the resolution of the Ebola scare involving a riverboat quarantined near Kinshasa after a former passenger died with Ebola-like symptoms. Passengers tested for Ebola were negative, providing follow-up to concerns that the outbreak might have reached the DRC capital.
Amuri-Aziza, A., et al. 2026. Phylodynamics and evolution of the 2026 Bundibugyo virus circulating in the Democratic Republic of the Congo: Insights from a 100-day window of genomic sequencing. Virological, August 2026.
🌐 Free online resource
➡ Provides the most extensive genomic reconstruction of the 2026 DRC Bundibugyo Ebola outbreak to date. Analysis of hundreds of viral genomes places the outbreak’s most recent common ancestor around late February, with uncertainty extending from mid-January through late March, supporting substantial transmission before recognition in May. The reconstruction suggests relatively slow early growth followed by acceleration and later stabilization, with no evidence that a fitness-enhancing viral lineage drove the outbreak.
Amuri-Aziza, A., et al. 2026. Genomic epidemiology of the ongoing 2026 Bundibugyo Virus Disease outbreak in the Democratic Republic of the Congo. Virological, 2026.🌐 Free online resource
➡ Earlier genomic analysis of the DRC outbreak establishing that Bundibugyo virus had likely been circulating before the outbreak was formally recognized. Provides the initial phylogenetic evidence for hidden transmission and geographic spread subsequently strengthened by the much larger 100-day genomic analysis.
World Health Organization Regional Office for Africa & Africa Centres for Disease Control and Prevention. 2026. Uganda ends Ebola outbreak following completion of 42-day countdown. WHO Regional Office for Africa, 27 August 2026.
🌐 Free online resource
➡ Announces the official end of Uganda’s Bundibugyo Ebola outbreak after 42 consecutive days without a new confirmed case following discharge of the last imported case. Provides an important contrast with the continuing DRC outbreak and demonstrates successful interruption of transmission through surveillance, contact follow-up, case management and border preparedness.
Africa News Agency. 2026. DRC: WHO makes 70,000 doses of Ervebo vaccine available against Ebola. Africa News Agency, 2026.
🌐 Free online resource
➡ Reports WHO’s provision of 70,000 doses of the Ervebo Ebola vaccine for use during the DRC Bundibugyo outbreak. Relevant to efforts to evaluate whether a vaccine licensed against Zaire ebolavirus can provide useful cross-protection against Bundibugyo virus during a large ongoing outbreak.
Coalition for Epidemic Preparedness Innovations. 2026. African Bundibugyo ebolavirus vaccine candidate to be advanced to clinical trials. CEPI, 27 August 2026.🌐 Free online resource
➡ Announces new support to advance an African-developed vaccine specifically targeting Bundibugyo ebolavirus into clinical trials. Provides context for longer-term efforts to develop a dedicated Bundibugyo vaccine rather than relying on possible cross-protection from vaccines designed for Zaire ebolavirus.
Kravinsky, N. 2026. Sonora registers more screwworm cases as cattle imports to Arizona resume. KJZZ Fronteras Desk, 27 August 2026.
🌐 Free online resource
➡ Reports the increase in New World screwworm detections in Sonora as cattle imports through the Douglas–Agua Prieta border crossing resume. Provides current context for the emerging Sonora cluster and why developments there matter for assessing pressure on the northern containment front and risk to Arizona.
Bowe, B., Xie, Y., & Al-Aly, Z. 2022. Acute and postacute sequelae associated with SARS-CoV-2 reinfection. Nature Medicine.
🌐 Open access
➡ Examines acute and post-acute health risks associated with SARS-CoV-2 reinfection in a large U.S. Veterans Affairs cohort. Finds that reinfection adds risk of death, hospitalization, and sequelae across multiple organ systems, with cumulative risks and burdens increasing with the number of infections. Supports the importance of considering post-acute and cumulative health effects when assessing the burden of repeated SARS-CoV-2 infections, even when acute disease does not require hospitalization.
Louisiana Department of Health. 2026. LDH reports additional Vibrio vulnificus cases and deaths. Louisiana Department of Health, 26 August 2026.
🌐 Free online resource
Louisiana Department of Health update
➡ Provides Louisiana’s updated 2026 surveillance data for Vibrio vulnificus, including cases, deaths, hospitalization and exposure information. Documents an unusually severe year for V. vulnificus in the state and provides comparison with Louisiana’s historical case and mortality averages.
Florida Department of Health. 2026. Vibrio infections. Florida Department of Health.🌐 Free online resource
Florida Department of Health Vibrio surveillance
➡ Florida’s official surveillance resource for Vibrio vulnificus, providing current and historical case and death counts by year and county. Useful for tracking Florida’s 2026 V. vulnificus activity and comparing the current season with previous years.
Comments