In the early 2000s, public health investigators in Taiwan traced a cluster of Stevens-Johnson syndrome and toxic epidermal necrolysis, two of the most dangerous reactions a medicine can cause, to a common treatment pairing: the worm medicine mebendazole taken together with the antiprotozoal and antibiotic metronidazole. This article explains what those conditions are and how the investigation was carried out. It sets out what the case-control study actually showed and where its evidence is limited. It also explains why prescribing information for mebendazole now advises against combining the two drugs. The aim is to give patients and families an accurate picture, neither alarming nor dismissive, so they can talk it through sensibly with their own physician.

What Stevens-Johnson syndrome and toxic epidermal necrolysis are

Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) are now understood as a single disease that varies in severity. In both, the immune system turns against the skin and the moist linings of the body. Cells in the outer skin layer die in large numbers, and the skin blisters and peels away. Doctors classify the condition by how much of the body surface detaches:

The illness usually begins one to four weeks after a new medicine is started. Early symptoms include fever, sore throat, burning eyes and general malaise. Within days a painful, dusky rash appears, along with sores in the mouth, eyes and genital area. Patients often need care in a burn unit or intensive care setting. Mortality rises with the extent of skin loss. It is generally under 10% for SJS and substantially higher for TEN, where published figures commonly fall in the range of 25–35% or more depending on age and other illnesses. Survivors can be left with lasting eye damage, scarring and other complications.

These reactions are rare, on the order of a few cases per million people per year. Most are triggered by medicines. Drugs most often implicated include certain anticonvulsants, allopurinol, sulfonamide antibiotics and some anti-inflammatory drugs. Neither mebendazole nor metronidazole is on the list of usual suspects. That is part of why the Taiwan cluster drew attention.

The setting: deworming foreign workers in Taiwan

Taiwan has long relied on contract workers from Southeast Asia, including many from the Philippines. As part of health screening, these workers were tested for intestinal parasites, and those found to be infected were treated. In practice, one person might carry both a worm infection and a protozoal infection such as amoebae or Giardia. Treating both together therefore made clinical sense. Mebendazole is a benzimidazole that kills many intestinal worms by disrupting their internal cell scaffolding. Metronidazole is a standard treatment for amoebic and giardial infections.

In 2001, Taiwanese health authorities were alerted to an unusual number of SJS and TEN cases among Filipino laborers. Investigators from Taiwan's national Center for Disease Control conducted the inquiry. It was published in 2003 in the American Journal of Public Health by Chen and colleagues, under a title describing an outbreak of SJS/TEN associated with mebendazole and metronidazole use among Filipino laborers in Taiwan.

It is worth pausing on who these patients were. They were working people far from home and family, earning a living under contracts and health rules they did not write. They took the medicines in good faith as part of a screening programme. Their suffering is a reminder that every person on the receiving end of a public health measure is an individual whose life has dignity. They are not an entry in a ledger. The investigators' careful work to find the cause was, in its own way, an act of respect for those lives.

What the case-control investigation found

A case-control study works backwards. Investigators identify people who developed the illness (cases) and comparable people who did not (controls). Both groups are then asked, or their records checked, about what they were exposed to beforehand. If an exposure is much more common among cases than controls, it becomes a suspect.

In the Taiwan study, the cases were Filipino workers who developed SJS or TEN. Controls were drawn from comparable workers who had not. Investigators examined medication histories and other potential exposures.

The central finding was striking. Cases were far more likely than controls to have taken both mebendazole and metronidazole. Use of the combination was strongly associated with the reaction, and that association stood out from the other exposures examined. The number of cases was small, as is typical of outbreaks of rare conditions. That means the precise size of the risk estimate is imprecise even though the direction of the association was clear. The authors concluded that concurrent use of the two drugs was the likely explanation for the cluster. They recommended that the combination be avoided.

Readers sometimes ask whether the study proved that either drug alone is dangerous. It did not. The signal centred on the combination. Both medicines have been taken safely, separately, by enormous numbers of people over decades. Both carry only rare individual reports of severe skin reactions.

How strong is the evidence?

Honesty requires holding two things together. The finding is important, and it is also limited.

Reasons to take it seriously:

Reasons for caution in interpreting it:

In short, this is a credible epidemiological signal that has never been replicated or explained. It is neither a proven, quantified hazard nor a statistical fluke to be waved away.

Why the combination is avoided today

Regulators and clinicians treat the finding with practical caution. United States prescribing information for mebendazole lists metronidazole under drug interactions. It notes that a case-control study of a single SJS/TEN outbreak suggested a possible association with concomitant use. It acknowledges that no additional data exist on the interaction, and it advises that concomitant use be avoided. Other drug-interaction references carry similar warnings.

The reasoning is straightforward. When the possible harm is catastrophic and alternatives exist, a precautionary approach costs little. A person with both a worm infection and a protozoal infection can usually have the two treated in sequence rather than at the same moment. Alternatively, a different agent can be chosen for one of them. Some antiprotozoal options and other anthelmintics are available, and the right choice depends on the specific parasite, the patient's other conditions and local availability. Those decisions belong with the treating physician.

It is also worth saying what the warning does not mean. It does not mean mebendazole is unsafe as ordinarily used. For most people it is a well-tolerated medicine. It is minimally absorbed from the gut and is used worldwide, including in mass deworming programmes that have relieved a great deal of childhood illness. Nor does the warning mean that metronidazole should be feared on its own terms. The concern is specific to taking the two together.

Practical points for patients and families

The Taiwan cluster carries a lesson beyond these two drugs. Many dangerous drug reactions are caught early, or prevented outright, because a patient or family member knew what was being taken and spoke up. Caring for one's own household includes keeping track of medicines, especially when several are started at once.

Informed consent is more than a signature. It means understanding why a medicine is offered, what the known risks are, and what the alternatives would be. The skin is a remarkable organ, a self-renewing barrier designed with real wisdom, and it usually repairs itself quietly throughout life. When a reaction like SJS overwhelms that design, the consequences are grave. That is reason enough for patients and physicians to weigh combinations thoughtfully together.

Key takeaway: A single, well-conducted but unreplicated case-control investigation of a 2001 outbreak among Filipino workers in Taiwan linked taking mebendazole together with metronidazole to Stevens-Johnson syndrome and toxic epidermal necrolysis, and because the possible harm is severe and alternatives exist, prescribing guidance advises avoiding the combination.