Scholarlink

Introduction <p>Bacterial antimicrobial resistance (AMR) is a major and growing public health threat in East African Community (EAC) countries, where fragile health systems, inadequate diagnostics, and inappropriate antibiotic use drive high levels of resistant infections. Despite this, robust subregional mortality estimates remain limited.</p> Methods <p>We conducted a secondary pooled analysis of modelled, publicly available, country-level mortality estimates from the Global Research on Antimicrobial Resistance (GRAM) 2019 project. Data were extracted for six EAC countries: Burundi, Kenya, Rwanda, South Sudan, Tanzania, and Uganda. Because GRAM reports age-standardised mortality rates (ASMRs) with 95% uncertainty intervals, we derived standard errors from these intervals, assuming approximate normality, and used them in the pooled analysis. Using random-effects models with restricted maximum likelihood (REML), we pooled ASMRs per 100,000 population for deaths associated with and attributable to AMR. We calculated 95% confidence intervals and prediction intervals, assessed heterogeneity using I<sup>2</sup>, and conducted leave-one-out sensitivity analyses to test robustness.</p> Results <p>Across the six EAC countries, there were an estimated 154,760 deaths associated with AMR and 36,480 deaths attributable to AMR in 2019. The pooled ASMR for AMR-associated deaths was 144.69 per 100,000 (95% CI 129.07–160.30) population, with a 95% prediction interval of 122.57–166.81. Country-specific ASMRs for AMR-associated deaths ranged from 129.5 per 100,000 population in Uganda to 167.0 per 100,000 population in Burundi. For AMR-attributable deaths, the pooled ASMR was 34.62 per 100,000 (95% CI 30.02–39.23) population, with a prediction interval of 28.10–41.14. Country-specific ASMRs for attributable deaths ranged from 30.80 per 100,000 population in Uganda to 41.90 per 100,000 population in Burundi. For both associated and attributable mortality, heterogeneity was negligible (I<sup>2</sup> = 0%), and sensitivity analyses confirmed that no country disproportionately influenced the pooled estimates.</p> Conclusion <p>This pooled secondary analysis indicates a substantial and regionally consistent mortality burden from bacterial AMR in East Africa. The findings reify the need for coordinated investment in AMR surveillance, stewardship, and overall response across the EAC.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Mortality burden of bacterial antimicrobial resistance in East Africa: pooled analysis of modelled estimates

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年12月8日

查看全文
  • Yusuff Adebayo Adebisi,
  • Najim Z. Alshahrani,
  • Theogene Uwizeyimana

摘要

Introduction

Bacterial antimicrobial resistance (AMR) is a major and growing public health threat in East African Community (EAC) countries, where fragile health systems, inadequate diagnostics, and inappropriate antibiotic use drive high levels of resistant infections. Despite this, robust subregional mortality estimates remain limited.

Methods

We conducted a secondary pooled analysis of modelled, publicly available, country-level mortality estimates from the Global Research on Antimicrobial Resistance (GRAM) 2019 project. Data were extracted for six EAC countries: Burundi, Kenya, Rwanda, South Sudan, Tanzania, and Uganda. Because GRAM reports age-standardised mortality rates (ASMRs) with 95% uncertainty intervals, we derived standard errors from these intervals, assuming approximate normality, and used them in the pooled analysis. Using random-effects models with restricted maximum likelihood (REML), we pooled ASMRs per 100,000 population for deaths associated with and attributable to AMR. We calculated 95% confidence intervals and prediction intervals, assessed heterogeneity using I2, and conducted leave-one-out sensitivity analyses to test robustness.

Results

Across the six EAC countries, there were an estimated 154,760 deaths associated with AMR and 36,480 deaths attributable to AMR in 2019. The pooled ASMR for AMR-associated deaths was 144.69 per 100,000 (95% CI 129.07–160.30) population, with a 95% prediction interval of 122.57–166.81. Country-specific ASMRs for AMR-associated deaths ranged from 129.5 per 100,000 population in Uganda to 167.0 per 100,000 population in Burundi. For AMR-attributable deaths, the pooled ASMR was 34.62 per 100,000 (95% CI 30.02–39.23) population, with a prediction interval of 28.10–41.14. Country-specific ASMRs for attributable deaths ranged from 30.80 per 100,000 population in Uganda to 41.90 per 100,000 population in Burundi. For both associated and attributable mortality, heterogeneity was negligible (I2 = 0%), and sensitivity analyses confirmed that no country disproportionately influenced the pooled estimates.

Conclusion

This pooled secondary analysis indicates a substantial and regionally consistent mortality burden from bacterial AMR in East Africa. The findings reify the need for coordinated investment in AMR surveillance, stewardship, and overall response across the EAC.