2013-03-13-世界卫生组织-Report_on_the_second_meeting_of_the_WHO_Evidence_Review_Group_on_malaria_burden_estimation_18页_220kb

Malaria Burden Estimation Report Summary

Introduction

The WHO Evidence Review Group (ERG) meeting in 2013 focused on reviewing current methods for estimating malaria morbidity and mortality, achieving consensus on WHO methodologies, and identifying research gaps. Key experts presented approaches, and recommendations were made for methodology and future studies.

Morbidity Estimation

Cartographic Approach (MAP Projects)

• Method: Uses parasite prevalence surveys and longitudinal studies combined with age-adjustment and population data to estimate incidence at the local level.
• Strengths: Not dependent on surveillance systems, global consistency, quantifies uncertainty; useful where surveillance is weak.
• Limitations: Data sparse in low-transmission areas, temporal and spatial gaps, lack of seasonality adjustment.
• Key Findings: Most reliable in Africa; least reliable in South Asia. Relies on high-resolution population maps (e.g., AfriPop).
• Future Work: Improve models for low-transmission settings, incorporate covariates (e.g., ITN coverage), validate models through local cohort studies.

Surveillance-Based Approach

• Method: Used by WHO for high-transmission African countries and low-transmission regions. Adjusts data for reporting completeness and test positivity.
• Key Data Inputs: HMIS reports, household surveys (DHS, MIS) for treatment-seeking behavior.
• Strengths: Countries can apply the method independently.
• Limitations: Complete data dependence; wide uncertainty intervals; recall bias; may overestimate cases by excluding non-tested groups.

Mortality Estimation

WHO’s Approach

• Low-Transmission Countries: Fixed Case Fatality Ratio (CFR) of 0.3% on total cases.
• High-Transmission (Africa): Age-stratified protocol using CHERG-derived child mortality rate and immunity models.

IHME Approach (Global Burden of Disease - GBD)

• Method: Multi-model ensemble (CODEm) using verbal autopsy data; aims to predict trends through multiple model testing.
• Key Findings: High uncertainty in adult mortality estimates; driven by empirical data rather than direct measurement.
• Limitations: Over-reliance on VA; time-period misalignments; uncertain cause fraction redistribution from unspecified deaths.
• Future Work: Incorporate updated data sources and improve predictive validity testing.

CHERG’s Approach (Child Mortality)

• Method: Uses verbal autopsy studies to estimate cause-specific mortality fractions (CSMF); incorporates covariates (skilled birth attendance, ITN coverage).
• Limitations: Covariate selection process not fully validated; imputation issues may misallocate deaths in low-transmission settings.

Recommendations

Morbidity

• Pilot surveillance-transition criteria (e.g., comparing risk-based vs surveillance-based estimates).
• Enhance data coordination: Collect MIS, use RDTs in sentinel sites, explore integration of molecular diagnostics.
• Develop simplified cartographic tools and apps for local application.

Mortality

• Immediate Actions: Clearly present uncertainties in WHO reports, include P. vivax mortality estimates.
• Long-Term Actions: Strengthen high-quality data collection, explore new methods like clinical trials and enrollment in prospective cohort studies.
• Validation: Engage INDEPTH sites to audit verbal autopsy coding and improve dataset completeness.

List of Recommendations

1. Use hybrid methods for integrated burden estimation.
2. Prioritize strengthening surveillance systems in African countries.
3. Validate current tools with local clinical data (RDTs, imaging).
4. Communicate uncertainty transparently in estimates.
5. Establish ongoing advisory support for methodology updates.