BASICS Child Mortality Survey

Abstract

• Purpose: To investigate deaths in children under five years of age in a community in order to identify the leading biological and socio-cultural factors that contribute to death.
• Type of Design: Descriptive and cross sectional. Could be longitudinal if surveillance system is continued (see below).
• Sample Size: Determined by surveyors and depends on the objective of the survey. Sample size of 350 - 450 deaths will provide good estimates of the true levels of most causes and risk factors (assuming most causes and risk factors having true levels from about 20% to 90% of all deaths).
• Where Used: Bolivia, Cambodia, Guinea and Kazakhstan. The perinatal section has been added in Guatemala and Cambodia.

To survey tool

Basic Information

Name: BASICS Child Mortality Survey

Origin: Henry Kalter and Rene Salgado.

Source: The Partnership for Child Health Care Inc. (BASICS) / Johns Hopkins University, Department of International Health.

Basic Description: A mortality survey or surveillance system which identifies deaths in children under five years (or in perinates and neonates), and then interviews relatives of the dead child (completing a verbal and social autopsy questionnaire) to identify the cause of death and the associated risk factors.

Country Applications: The child mortality survey has been used in Bolivia, Cambodia, Guinea and Kazakhstan. The perinatal mortality version has been used in Guatemala and Cambodia.

Languages Available: English and Spanish.

Technical Scope: Mortality assessment at the community and district levels to raise awareness of factors influencing child and infant mortality, and develop strategies to reduce deaths.

Purpose: The survey determines biological causes of death, investigates the social context and associated risk factors, gathers information from treatment records and assesses the quality of care outside the home from non-Government health providers.

Type of Methods: Quantitative and qualitative.

Design: Descriptive and cross sectional. Could be longitudinal if surveillance system is continued.

Frequency of Administration: Normally it is conducted once, unless the surveillance system is continued, or the survey is used to monitor interventions to reduce child mortality.

Key Users of Information

Primary users: Community members and district level health planners/managers and health providers.

Secondary users: National level health planners/managers and NGO's and other agencies working in the geographic area.

Decisions made

Interventions designed to reduce mortality based on the data obtained, which included:

• Identification of target groups;
• Determination of demographic characteristics of target groups;
• Identification of most frequent causes of deaths;
• Listing of treatments and treatment sources;
• Description of the process of decision making for care seeking;
• Quality of medical services available (both formal and informal);
• Extent of community involvement in health.

Objectives and Scope of Tool

1. Determine the most common causes of child deaths in the project area;
2. Specify demographic characteristics of the decedents that might help select and target a cause of death for intervention;
3. Assess whether the prior health status and care of the children might have increased their risk for a fatal illness;

1. Evaluate in detail the care of sick children inside and outside the home that might have contributed to their deaths;
2. Determine the most important findings to share with the community, in order to develop a health intervention;
3. Determine availability and accessability of health care outside the home.

Key Monitoring Needs and Evaluation Questions Tool Seeks to Address

What are the mortality and morbidity trends in children under five? What are the main causes of mortality? The survey could be used for monitoring or evaluation of interventions to reduce child mortality. Once care seeking behaviors and sources of care are determined, changes in behavior can also be evaluated over time.

Key Indicators

General Child Mortality Indicators

Demographic Indicators

Prior Health Status and Care Indicators

Socio-Cultural Indicators

Research design

Standard protocol:
There are several distinct steps to using this tool. Total time estimated for planning, implementation and analysis is 28 weeks which does not include a full year of surveillance. The perinatal/neonatal section could be carried out more rapidly - perhaps in 12 - 16 weeks.

Formative research is needed to test local terms, etc., and to modify the questionnaires.

Main tasks and suggested time frames are as follows:

• Selection of project coordinator, organization of coordinating committee, expert panel, community partnership group, orient partnership group to study and concepts, select the survey or surveillance area and locate the deaths.
• Conduct formative research to adapt survey forms, training materials, and pre-test, and finalize forms and training materials. Time estimated - eight weeks.
• Choose time for the survey; select sample; select supervisors, surveyors and data clerks; plan for dissemination of results; and continue death reporting. Time estimated - two weeks.
• Make administrative, financial and legal arrangements; train supervisors and surveyors; and continue death reporting. Time estimated - two weeks.
• Conduct survey or surveillance; enter data into computer; monitor the project with the community and problem solve; share interim surveillance data; and continue deaths reporting. Time estimated - eight weeks and 12 month cycle.
• Make analysis plan and tally and analyze data. Time estimated - two weeks.
• Interpret and prepare to share findings with the community. Time estimated - four weeks.
• Share findings with the community; develop and implement interventions; and draft and finalize report. Time estimated - four weeks.

Type of design: Descriptive and cross sectional.

Units of Observation and Analysis: Individual interviews with relatives of dead child and examination of medical records.

Sample Size: Depends on objectives of survey, but 350 - 450 deaths is an adequate size for most common etiologies. Researchers may choose to investigate all deaths in a population, if feasible.

Target Population: Communities with levels of infant and child mortality which would benefit from more information on causes and risk factors related to those deaths.

Training

A minimum of six days is needed for training the surveyors and the supervisors. Training should include thorough discussions, role plays and problem-solving exercises surrounding the questionnaires. Self assessment checklists are required for the surveyors. A supervisor's checklist assists in standardizing and maintaining quality.

Lessons from Experience:
Experience in other countries suggest that at least two weeks of training are needed to assure quality. In some cases, up to three weeks have been needed for training.

Implementation

• Formative research should be performed prior to finalizing the questionnaires and starting data collection.
• Standardized job descriptions are needed for hiring surveyors, supervisors and computer data entry clerks.
• Informed consent statements should be drafted and signed by respondents before interviews.
• Some form of legal agreement should be drafted for project personnel to pledge that they will not disclose study data/information.
• Training materials including a surveyors' training schedule and curriculum, role play scenarios for surveyors and trainers, and surveyors' procedures manuals for the verbal autopsy, social autopsy and treatment and records questionnaires must be prepared and translated.
• During the training, surveyors should be sensitized to the problems of interviewing relatives about sad and stressful events and the psychological stress on the surveyors themselves. Support for mothers and the surveyors should be anticipated and provided.
• Tables, graphs, figures and a sample illness narrative to help share the project findings with the community should be drafted.
• A profile of a hypothetical health program, e.g. an intervention against pneumonia, to illustrate the process by which a community group can develop an integrated intervention, should be prepared.
• Key staff required: Survey manager, Project coordinator/trainer, formative research team (usually health program staff, possibly social workers), census takers who locate houses where children have died, surveyors who conduct the interviews, supervisors who supervise the surveyors and various support staff, such as data entry clerks, etc.
• To monitor survey implementation, the Survey Manager should ask:

- do the number of deaths reported in survey or surveillance system correspond with estimated rates?

- how many interviews are being completed?

- are there any problems associated with locating households, carrying out interviews etc.?

• Supervisors should check the quality and consistency of the data obtained. It may be helpful if supervisors repeat a certain percentage of interviews to check interviewer consistency.

Analysis

• Data entry and analysis can be performed using EPI INFO.
• Close supervision is required. It is desirable for surveyors to use self-assessment checklists and for supervisors to use checklists. Randomly selected repeat interviews in a certain percentage of cases (5% - 10%) should be performed as a quality check.
• The instrument uses the BASICS 'Pathway to Survival' as a framework for the data collection and analysis. Suggestions to guide the analysis include:
• A list of suggested indicators;
• A table to assist data analysis, specifying the numerator, the denominator and data source for calculating each indicator;
• Guidelines for determining appropriate child care actions are needed to calculate indicators for the "Pathway to Survival";
• Suggested verbal autopsy algorithms to determine the biological diagnoses of children who died;
• Information on the computer data entry and analysis programs supplied with the manual;
• Examples of tables, charts and graphs providing ideas on how to present results of the data analysis.
• Data input and programming of the databases and analysis programs is key.

Reporting
See above.

Dissemination of Results

Dissemination of results is clearly a vital part of using this instrument. Plans for use and dissemination of data should be made in the initial planning phase. Plans will reflect the intended use of the tool, i.e., community motivation, awareness raising for politicians or Ministry of Health officials, etc.

Manual and Guidelines

A comprehensive manual with sample job description, data collection forms, indicator lists, examples of results will be produced under BASIC II.

Contact Person
Rene Salgado
BASICS Project
1600 Wilson Blvd.
Arlington, VA 22209
703-312-6800