Antibiotic Resistance Deaths Are Rising Even as Mortality Rates Fall, Exposing a Deep Global Health Divide

Antibiotic resistance is creating a troubling contradiction in global public health: the risk of dying from resistant bloodstream infections has declined over three decades, yet the total number of deaths continues to increase. Behind these opposing trends lies a changing distribution of mortality across countries, age groups and bacterial pathogens, raising difficult questions about whether improvements in infection control and treatment are translating into sufficient reductions in the global death toll.

A study titled "Global Mortality Burden of Antimicrobial-Resistant Bloodstream Infections: An Analysis of Global Burden of Disease Study 2021 Estimates," published in the journal Antibiotics, examines this contradiction. Conducted by Hui Yang, Gang Wang and Shaowei Sang of Qilu Hospital of Shandong University, China, the research analyzes global antimicrobial resistance (AMR) mortality between 1990 and 2021, revealing substantial geographical inequalities, changing demographic risks and persistent threats from specific drug-resistant bacteria.

The researchers estimate that antimicrobial resistance was associated with 1.82 million deaths within the bloodstream infection category in 2021, including approximately 460,000 deaths attributable to resistance itself. These figures reveal a significant public health burden that cannot be adequately understood through declining mortality rates alone.

Falling mortality rates conceal a growing absolute death toll

Bloodstream infections are among the most serious bacterial infections because effective treatment frequently must begin before laboratory confirmation identifies the responsible organism. Resistance complicates this process by reducing the effectiveness of available antibiotics and increasing uncertainty over initial treatment decisions. Understanding the resulting mortality burden is therefore essential for determining where healthcare systems should direct prevention, diagnostic and treatment resources.

The researchers used Global Burden of Disease 2021 estimates to examine mortality across seven major geographic regions, accounting for differences in age, sex, bacterial pathogens and antibiotic resistance patterns. Their analysis distinguishes between deaths associated with resistant infections and deaths attributable specifically to resistance, two measures that capture different dimensions of the health burden.

AMR-associated mortality estimates deaths involving resistant infections against a scenario in which those infections do not occur. AMR-attributable mortality instead estimates excess deaths compared with a scenario in which resistant infections are replaced by infections susceptible to treatment. This distinction prevents the broader burden of resistant infections from being mistaken for deaths caused specifically by resistance.

In 1990, the estimated number of AMR-associated bloodstream infection deaths stood at 1.53 million, alongside 370,000 attributable deaths. By 2021, these figures had increased to 1.82 million and 460,000 respectively. However, the age-standardized associated mortality rate declined from 33.03 to 23.30 deaths per 100,000 people, while the attributable rate fell from 7.93 to 5.85.

The divergence illustrates an important limitation in evaluating antimicrobial resistance progress. Age-standardized rates allow comparisons after accounting for differences in population age structure, but they do not reveal the full scale of deaths requiring prevention. Improvements in standardized mortality can therefore coexist with substantial demands on treatment capacity, laboratory services and infection prevention systems.

Sub-Saharan Africa faces the highest risk as regional threats diverge

The global estimates conceal pronounced geographical differences. Sub-Saharan Africa recorded the highest AMR-associated age-standardized mortality rate in 2021, reaching 38.23 deaths per 100,000 people. Meanwhile, Southeast Asia, East Asia and Oceania collectively registered the largest absolute regional burden, with approximately 480,000 associated deaths.

Although standardized mortality declined across all major regions, reductions were slowest in Sub-Saharan Africa and South Asia. The analysis also identified an inverse relationship between socioeconomic development and mortality rates, indicating that regions with lower development levels generally experienced greater mortality burdens. However, the researchers caution that this association does not establish a causal relationship.

The geographical differences reflect a broader challenge in antimicrobial resistance control. The researchers discuss limited healthcare access, inadequate surveillance, inappropriate antibiotic use and weaknesses in water, sanitation and hygiene infrastructure as possible contributors to elevated mortality in resource-constrained settings. These explanations remain informed hypotheses rather than mechanisms demonstrated by the analysis.

Regional variation extends beyond mortality levels to the bacteria responsible for deaths. While Staphylococcus aureus accounted for the largest global burden, Acinetobacter baumannii led attributable mortality in South Asia, whereas Klebsiella pneumoniae was the leading pathogen in Sub-Saharan Africa. Such differences complicate strategies that prioritize bacterial threats solely according to worldwide rankings.

The findings support approaches that combine internationally coordinated surveillance with locally informed interventions. Countries require information about their dominant pathogens, resistance patterns and vulnerable populations to determine appropriate diagnostic capacity, infection prevention measures and antimicrobial treatment priorities. Identical strategies across regions with substantially different mortality profiles may fail to address their most pressing threats.

From vulnerable newborns to ageing populations, the mortality burden is shifting

Neonatal mortality declined markedly over the study period, while the absolute number of deaths among older adults increased. These opposing developments reveal how the populations requiring protection from resistant infections have changed.

Among newborns, AMR-associated deaths fell from approximately 586,000 in 1990 to 350,000 in 2021. Deaths attributable to resistance declined from 143,000 to 89,000. Despite this progress, neonates continued to experience the highest age-specific mortality rates, demonstrating that reductions in deaths have not eliminated their exceptional vulnerability.

The burden among older adults moved in the opposite direction. Among people aged 70–74, AMR-associated deaths increased from 112,000 to 193,000, while attributable deaths rose from 26,000 to 47,000. Importantly, age-specific mortality rates declined across all age groups, meaning increasing death counts among older populations cannot be interpreted as evidence of rising individual mortality risk.

The researchers identify population ageing, underlying medical conditions and age-related physiological changes as possible explanations for the increasing concentration of deaths among older adults. Improvements in newborn care may also have contributed to declining neonatal mortality, although the study does not directly evaluate the effects of specific healthcare interventions.

Sex differences introduce another dimension of inequality. Men consistently experienced higher age-standardized AMR-associated and attributable mortality rates than women across all seven regions. Mortality reductions were also generally faster among females for most pathogens, indicating persistent demographic differences that warrant further investigation rather than assumptions about their underlying causes.

For healthcare planning, these patterns create simultaneous priorities. Infection prevention and effective antimicrobial treatment remain essential for newborns, while older populations require increasing attention to resistant infections. Age-specific surveillance can help distinguish groups facing the highest mortality rates from those accounting for growing numbers of deaths.

Drug-resistant pathogens expose the limits of a uniform treatment response

Beyond regional and demographic differences, the research identifies a concentrated burden among six bacterial pathogens. Staphylococcus aureus accounted for approximately 410,000 AMR-associated deaths in 2021, followed by Escherichia coli with 280,000, Acinetobacter baumannii with 230,000 and Klebsiella pneumoniae with 220,000. Streptococcus pneumoniae and Pseudomonas aeruginosa were also major contributors.

The analysis reveals that resistance to different antibiotics produces distinct mortality rankings. Fluoroquinolone resistance accounted for the largest AMR-associated burden among antibiotic classes, at approximately 1.095 million deaths. Carbapenem resistance, however, represented the largest attributable burden, with approximately 110,000 deaths, emphasizing the significance of resistance affecting these important antibiotics.

At the individual pathogen–drug level, methicillin-resistant Staphylococcus aureus accounted for approximately 68,000 attributable deaths, while carbapenem-resistant Acinetobacter baumannii accounted for approximately 41,400. The researchers emphasize that these are modeled estimates, and their rankings do not establish statistically significant differences between resistance categories.

Another concerning pattern involved Escherichia coli, whose resistance-related mortality rates increased among adults aged 25 and older, particularly males aged 25–49. This departure from broader declining mortality trends demonstrates why improvements in aggregate indicators should not obscure emerging pathogen-specific risks.

The evidence also carries important methodological cautions. Estimates rely on statistical models rather than directly observed deaths, while microbiological surveillance remains uneven across regions and periods. Bloodstream infection classifications do not necessarily represent laboratory-confirmed cases, and overlapping resistance to several antibiotics means associated mortality estimates cannot simply be added together.

These limitations reinforce the importance of stronger surveillance and more reliable clinical data. The researchers advocate targeted infection prevention, antimicrobial stewardship, rapid diagnostics, equitable treatment access and development of new therapies. Their findings identify where attention is warranted without establishing which interventions would deliver the greatest mortality reductions.