UNIVERSITY OF WASHINGTON Maternal Mortality in Mexico: lessons learned from improving the data quality Prof. Rafael Lozano, IHME and former General Director of Health Information at the Ministry of Health in Mexico Measuring the Progress on Maternal and Child Mortality: Data, alternative methods, and findings Washington, D.C. May 24 th , 2010
14
Embed
Maternal mortality in Mexico: lessons learned from improving data quality
IHME's Dr. Lozano discusses how the quality of maternal mortality data was improved in Mexico, leading to better estimates to more reliably measure progress.
For more information please visit www.healthmetricsandevaluation.org
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
UNIVERSITY OF WASHINGTON
Maternal Mortality in Mexico:lessons learned from improvingthe data quality
Prof. Rafael Lozano, IHME and former General Director of Health Information at the Ministry of Health in Mexico
Measuring the Progress on Maternal and Child Mortality:Data, alternative methods, and findings
Washington, D.C. May 24th, 2010
La democratización de la salud en México
Hacia un sistema universal de salud
Programa Nacional de Salud 2001-2006
National Health Program 2001-2006 Annual Report of Accountability
How do we measure and report progress?
•Benchmark by state
•Selected health indicators
1)Health Outcomes2)Health System Performance
o Qualityo Fair Financeo Health Resourceso Health expenditure
•The best performance for specific health interventions
Annual Report of AccountabilityMexico-Salud (2001)
Under-five mortality rate
MMR was not included in first Annual Report of Accountability !!!• There was evidence that maternal deaths were under registered
in Mexico
o WHO (2000) used a correction factor of 1.4
o National Survey of Demographic Dynamic (ENADID 1997) showed 38% under registry
• Evidence was national only; not sufficient at subnational level
• In 2001, evidence gathered showed misclassification of causes of death as a major component of the problem (study in 9 states)
• After deliberation, Minister supported decision not to publish MMR
…We need an instrument that at the same time can improve the quality of the measurement and include all the actors in the exercise .....
Implications of the decision
• Skepticism of the program managers of Reproductive Health
• Disbelief of many … we were giving more importance to the use of healthy information, even for a very sensitive indicator…
o Presidential Goal; NHP Goal; MDG 5
• Opportunity for NGOs to give input
The challenge of counting maternal deaths
Objective: To produce timely and reliable indicator (MMR) by state and involve all relevant participants
•Designed a passive identification of maternal deaths in women of reproductive age using death certificate records and a special list of potential causes of maternal- or pregnancy-related deaths
•Participants
o State Health Systems (MOH); other health providers (social security, private, etc.); other statistics agencies (INEGI, RENAPO, etc.)
o Information areas, epidemiological surveillance, reproductive health, at federal and state level
The challenge of counting maternal deaths
• Process
o Collect all information available of each “potential maternal death”: Medical records, Verbal Autopsy, Death Certificate, Maternal Mortality Committee Report, Critical Links of maternal deaths, etc.
o Verify one by one with INEGI database
o Incorporate those which qualify as new registries into the “official” mortality database (INEGI)
o Include the reliable numbers by state in the Annual Report of Accountability (August). These numbers should match with the consolidated statistics of causes of death in the official INEGI release (November)
Passive identification of maternal deaths from VR
Confirmed maternal deaths in
2002
38 597 5 849 1 330
Possible maternal deaths
Deaths of women 10-
54 years
Short-term effects of the publication 2002• Following a long debate about the limitations of the method
selected, we received the support of the authorities of the States Health Systems and other health providers
• Special attention to NGOs, without total success
o Small correction of underreporting, because deaths from indigenous areas still not included in the statistics
• Newspapers treated the publication harshly
o Rates seemed to be getting worse instead of improving
Mid-term effects of the publication (2002-2008)
• Empowerment of neglected areas in the state health systems, such as: information and evaluation (more training)
• More attention to keeping the data healthy and the process transparent
• More credibility from NGOs
• Better culture of information1,221 maternal deaths addeddeaths %