The findings were presented at the congress of the Spanish Society of Hospital Pharmacy (SEFH) and at the most recent European congress in the field. They are based on a 2019 report by the government of Spain's Valencia region that was never published, to which EL PAÍS has now obtained access.
Metamizole, also known as dipyrone, is the active ingredient in Nolotil, Spain’s best-selling branded drug, prescribed for pain and high fever. It is banned in the United States and has been withdrawn in the United Kingdom, France, Sweden and some 30 other countries because of its link to agranulocytosis, a drastic reduction in granulocytes, a type of white blood cell, that can be fatal. In Spain, the Spanish Medication and Health Products Agency (AEMPS) says the drug’s benefit-risk balance remains positive.
The study analysed severe adverse reactions reported in 2016-2017 in five coastal areas of Alicante province with a high presence of foreign residents: Denia, Villajoyosa, Torrevieja, San Juan and Orihuela. These areas together had 916,000 registered residents at the time, of whom 62,000 (6.7%) were originally from the UK and 28,000 from other northern European countries.
The report identified 41 patients hospitalised with agranulocytosis caused by metamizole: 26 British, 11 Spanish, one Danish, one Dutch, one Swedish and one Moroccan. Nine patients died, seven of them British and two Spanish.
Researchers cross-checked hospital discharge databases and outpatient consumption of the drug in people over 45 during 2016 and 2017, and reviewed medical records individually. Patients whose immune suppression could have had other causes were excluded.
Using two different methods, the study compared the relative risk for British versus Spanish patients. One method calculated 0.33 cases per 100,000 defined daily doses among Spaniards and 39.8 cases among British people, a 120.7-fold higher risk. The other method found 0.084 cases per 1,000 treated patients among Spanish nationals (about one in 12,000) and 6.62 among Britons (one in 151), a 78.8-fold higher risk.
The AEMPS downplayed the findings in a statement, saying the Valencia report "has important limitations since it is not analytical but descriptive. It does not allow calculation of incidences or relative risks, which is crucial to evaluate the safety profile." The agency added that "neither the study design, nor the method for calculating frequency measures, nor the estimation of the relative risk, nor the absence of control for other biases allow the offered conclusion to be drawn."
The AEMPS pointed to its own study, published in 2024 in the European Journal of Clinical Pharmacology, which concluded that "agranulocytosis was very rare, but more common with metamizole than with other analgesics (more than four times higher)." The paper noted the small number of cases analysed prevented ruling out that the differences were due to chance, and it did not distinguish the population by country of birth.
In 2018, the AEMPS recommended against using the drug for longer than one week and advised against it in transient populations. The study’s author, José Luis Marco Garbayo, criticised this, saying at an SEFH event: “The recommendation referred to tourists, when [the risk] is actually with people from these countries who are residing here.” He has declined to be interviewed.
Between 2023 and 2024, the European Medicines Agency (EMA) reevaluated metamizole after Finland decided to withdraw it. The EMA concluded the benefits still outweighed the risks but pushed for changes to package leaflets to help detect complications.
Vicente Palop, of the working group on medication use at the Spanish Society of Family and Community Medicine (semFYC), criticised the handling of the drug's safety: "There is a serious problem of transparency and reporting. We do not know the real impact of adverse effects. And not only in the northern European population, but also among Spaniards. While other countries act firmly, like Finland, the AEMPS seems to prefer to ignore or discredit reports that point to the problem instead of promoting the measures necessary to determine its true magnitude."
