The global estimates are recent and large: the GRAM study put deaths directly attributable to bacterial resistance at about 1.3 million in 2019, with nearly five million deaths associated, and its 2024 update projected the attributable toll rising to around two million a year by 2050, with the increase concentrated among people over 70. The estimates depend on a counterfactual — what would have happened had the infection been susceptible — and on sparse data from the countries with the highest burden, so the uncertainty is wide. In high-income countries the pattern is clearer and different: MRSA has fallen, resistant Gram-negative infections have risen, and the deaths are increasingly in older, frailer patients in hospital. Surveillance networks such as EARS-Net and GLASS record resistance well and deaths poorly; almost no country links microbiology to its death registration routinely, so national mortality from resistance is inferred rather than counted.