2011-12-31-世界卫生组织-Review_of_the_WHO_Verbal_Autopsy_VA_Instruments_53页_575kb

Meeting Summary: Review of the WHO Verbal Autopsy (VA) Instruments

Executive Summary

The 2011 meeting reviewed the WHO Verbal Autopsy (VA) instrument to develop a simplified version for routine use in civil registration systems, especially where medical certification of death is unavailable. The goal was to create a concise yet effective tool with a list of 62 causes of death, validated through expert review, item reduction, and comparison with automated methods like InterVA and PHMRC. The simplified VA instrument includes standardized questions to be adapted via IT for data analysis, with plans for future validation and implementation.

Key Outcomes

• CAUSE OF DEATH (CoD) List: A finalized, simplified list of 62 causes of death, reduced from the original 106, was agreed upon based on feasibility, public health relevance, and evidence from InterVA and PHMRC.

  • Removed Causes: Typhoid/paratyphoid, leishmaniasis, non-communicable respiratory diseases (unless specified), mental disorders, certain other infectious and unspecified diseases.
  • Retained/Promoted Causes: HIV/AIDS, malaria, neonatal tetanus, perinatal asphyxia, maternal-related causes, tuberculosis, etc., prioritized for their public health impact and VA feasibility.

• VA Instrument Simplification:

  • A total of 221 items were consolidated into four sections with skip patterns based on age and context.
  • The maximum number of questions (140 for reproductive-age women) was determined after balancing simplicity and accuracy.
  • Tariff Method: Applied for item reduction, prioritizing symptoms relevant to specific causes, ensuring better performance than physician review for adults/ neonates; physician review is strongly recommended for children.
  • Question Design: Focuses on closed-ended questions suitable for automated analysis and avoids subjective or unreliable questions on culturally sensitive topics.

• Implementation Challenges:

  • Resource scarcity and lack of qualified personnel may hinder deployment.
  • Integration with CRVS systems and IT tools (like InterVA) requires careful planning.
  • Cultural factors (e.g., sensitive topics, mourning periods) affect VA questionnaire administration.

• Recommendations:

  • Validation of the simplified VA tool in selected countries (Thailand, Bangladesh, Ghana, Philippines).
  • Further work to adapt software for automated cause-of-death assignment.
  • Training programs focusing on interviewers and local health professionals.
  • Future refinements may include expanded IT systems for decision support.

Significance

This review provides the foundation for improving global mortality surveillance in low/middle-income countries through a standardized, simplified VA tool compatible with automated analysis and integrating lessons from diverse field experiences.