In 2021, 4.71 Million Deaths Were Associated With Bacterial Antimicrobial Resistance Worldwide. By 2050, That Number Is Projected to Reach 8.22 Million Per Year.

Every time a clinician prescribes an antibiotic, every time a farmer uses one prophylactically in livestock, every time a patient does not complete a course of treatment, the selective pressure on bacteria increases. The organisms that survive are more resistant. The ones that reproduce pass that resistance on. This is not a hypothetical future problem. It is a present one, and it has been growing for three decades without the sustained global response it requires. This systematic analysis from the Global Burden of Disease 2021 Antimicrobial Resistance Collaborators, coordinated through the Institute for Health Metrics and Evaluation at the University of Washington, represents the first comprehensive assessment of antimicrobial resistance (AMR) mortality covering historical trends from 1990 to 2021 and projections through 2050. What it finds should change how every clinician thinks about the antibiotics they prescribe, and how every policymaker thinks about the surveillance and research infrastructure they fund.

The GBD 2021 AMR Collaborators estimated all-age and age-specific deaths and disability-adjusted life-years (DALYs) attributable to and associated with bacterial AMR for 22 pathogens, 84 pathogen-drug combinations, and 11 infectious syndromes in 204 countries and territories from 1990 to 2021. The analysis drew on more than 520 million individual records or isolates and 19,513 study-location-years, including multiple cause of death data, hospital discharge data, microbiology data, pharmaceutical sales, antibiotic use surveys, and outpatient and inpatient insurance claims data across every world region. Two counterfactual scenarios were used to estimate AMR burden: one replacing all drug-resistant infections with drug-susceptible infections, yielding deaths attributable to AMR; and one replacing all drug-resistant infections with no infection, yielding deaths associated with AMR. Forecasts were produced for three scenarios through 2050: a reference scenario reflecting the most likely future, a Gram-negative drug scenario assuming new drug development targeting Gram-negative pathogens, and a better care scenario assuming improvements in healthcare quality and access to appropriate antimicrobials.

In 2021, an estimated 4.71 million deaths (95% uncertainty interval 4.23 to 5.19 million) were associated with bacterial AMR, and 1.14 million deaths (1.00 to 1.28 million) were directly attributable to bacterial AMR. Trends over the 31-year study period varied substantially by age. From 1990 to 2021, deaths from AMR decreased by more than 50% among children younger than 5 years, driven by improvements in vaccination coverage, water and sanitation infrastructure, and infection control. At the same time, AMR mortality increased by over 80% for adults aged 70 years and older. AMR mortality decreased for children younger than 5 in all super-regions, while AMR mortality in people 5 years and older increased in all super-regions.

Among specific pathogens, meticillin-resistant Staphylococcus aureus (MRSA) increased the most globally, rising from 261,000 associated deaths in 1990 to 550,000 associated deaths in 2021. Among Gram-negative bacteria, resistance to carbapenems showed the steepest rise of any antibiotic class, increasing from 619,000 associated deaths in 1990 to 1.03 million associated deaths in 2021. The reference scenario forecasts project 1.91 million deaths attributable to AMR and 8.22 million deaths associated with AMR globally in 2050. Super-regions projected to have the highest all-age AMR mortality rate in 2050 are South Asia and Latin America and the Caribbean. Under the better care scenario, 92.0 million deaths could be cumulatively averted between 2025 and 2050 through improved care of severe infections and appropriate antibiotic access. Under the Gram-negative drug scenario, 11.1 million AMR deaths could be averted through new drug development.

• —The analysis relied heavily on statistical modelling to produce estimates for locations with limited or no primary data, which introduces uncertainty that is wider in low and middle income settings where the data is sparsest and the burden may be greatest.
• —There was a notable decrease in non-COVID-related infectious disease in 2020 and 2021, which may reflect real reductions from pandemic-era infection control measures or may reflect underdiagnosis and underreporting during a period of extreme healthcare disruption. The authors acknowledge this complicates trend interpretation around the pandemic years.
• —The forecasts are projections based on modelled trends and carry substantial uncertainty intervals, particularly for the 2050 horizon. Alternative scenarios represent policy-relevant counterfactuals rather than predictions.
• —The study covers bacterial AMR and does not encompass antifungal, antiviral, or antiparasitic resistance, which represent additional and growing dimensions of the broader AMR problem.

AMR is not a future threat. It is a present one that has been building for thirty years and is projected to accelerate. The data shows what works, infection prevention, vaccination, appropriate antibiotic access, and stewardship, and quantifies what is at stake if those interventions are not scaled. Ninety-two million deaths between now and 2050 sit on one side of a policy decision. The question is whether the global health community will treat AMR with the urgency that number demands before it becomes the next pandemic no one can say they did not see coming.