Identify questions & indicators¶

Note: Content in this section draws on existing FASTR presentation materials and is subject to revision.

Overview¶

This section outlines the process for identifying priority policy and programmatic questions and selecting appropriate indicators for FASTR analysis. It provides a structured approach to ensuring that FASTR analyses are demand-driven, analytically feasible, and aligned with national priorities.

Specifically, this section covers:

1. Introduction to FASTR: gaps and challenges
   An overview of the analytical gaps FASTR is designed to address, its role in reducing fragmentation in routine data analysis, and how FASTR can be positioned as an entry point for engagement with government stakeholders.

2. Development of a data use case
   Guidance on co-developing data use cases through workshops with the Ministry of Health and other stakeholders, including practical examples from country implementations.

3. Defining priority questions and selecting indicators
   A framework for formulating priority analytical questions, selecting suitable indicators, and aligning FASTR analysis with national strategies and decision-making needs.

4. Preparing for data extraction
   A high-level overview of pre-extraction considerations, including understanding the DHIS2 configuration, mapping indicators to data elements, and planning the extraction timeline.

Defining priority questions¶

Effective use of routine data depends on well-defined analytical questions. Priority questions provide direction for FASTR analyses and help ensure that outputs are relevant and actionable for decision makers.

Characteristics of a good priority question include:

• Addresses a priority issue
  Focuses on topics of clear interest to policy makers and program managers.

• Relevant
  Important enough to warrant analysis and to inform decision making.

• Grounded in current realities
  Connected to ongoing challenges, reforms, or shocks affecting service delivery.

• Meaningful to stakeholders
  Addresses issues that matter to specific individuals or groups involved in planning or implementation.

• Answerable
  Can be addressed using available data, methods, and timeframes.

Assessing relevance: key questions to consider¶

When assessing whether a question is a priority, the following considerations are useful:

• Who is the intended audience?
• What do they need or want to know?
• When do they need the information?
• Which period or event is of interest?
• Why is this information needed?
• How will the findings be used?

What do we mean by “answerable”?¶

A question is considered answerable if the following conditions are met:

Data availability
- The required data exist and are of sufficient type, quantity, and quality.

Analytical feasibility
- Appropriate and statistically valid methods are available and feasible to apply.

Timeliness
- The analysis can be completed within the required timeframe (e.g., quarterly reporting cycles).

PICO framework for formulating answerable questions¶

Note: This framework was included in the original presentation material and is retained here as an optional tool.

The PICO framework, commonly used in public health and evidence-based research, provides a structured way to formulate clear and answerable questions.

Selecting indicators: what makes a good FASTR indicator?¶

Indicator selection is critical to the quality and usefulness of FASTR analysis. Indicators should be chosen based on the following criteria:

• Relevance
  The indicator aligns with priority questions and policy objectives.

• Volume
  The indicator is reported at sufficiently high volumes to support robust analysis.

• Completeness
  Reporting completeness is high across facilities and over time.

• Frequency
  The indicator is reported frequently enough (typically monthly) to support rapid-cycle analysis.

• Type
  The indicator represents a count of services delivered.

Why focus on high-volume indicators?¶

One of the core strengths of the FASTR approach is its ability to adjust for data quality issues. High-volume indicators are better suited to this process because:

• Reduced sensitivity to outliers
  In low-volume indicators, individual data points can disproportionately affect trends.

• More stable estimates
  High-volume data reduce random variability and improve the reliability of trend detection.

• Clearer identification of true anomalies
  Larger counts make it easier to distinguish genuine outliers from natural variation.

Count indicators also allow for ongoing validation and adjustment before proportions or coverage measures are derived externally.

Why focus on high-completeness indicators?¶

Indicators with high reporting completeness are preferred because they:

• Improve data reliability
  More complete data reduce bias and provide a more representative picture of service delivery.

• Support consistent analysis
  High completeness enables meaningful comparisons across time and geographic areas.

• Reduce misinterpretation
  Incomplete data can falsely suggest changes in service utilization when changes are driven by reporting gaps rather than real trends.

While statistical methods such as imputation can be used to address incomplete data, these methods require assumptions about missing values. Further detail is provided in Data Quality Adjustment.

Why focus on count indicators?¶

Limitations of proportion indicators

• Proportions limit the ability to adjust numerators and denominators separately for data quality issues.
• Numerators and denominators may each be affected by different sources of error.
• Separating counts from denominator estimation allows for more transparent and flexible adjustment.

Mortality as a rare event

• Mortality indicators are typically low-frequency and not well suited to frequent adjustment.
• These indicators are generally better analyzed using annual rather than monthly or quarterly data.

FASTR core indicators¶

The FASTR approach focuses on a core set of RMNCAH-N indicators that represent key points along the reproductive, maternal, newborn, child, and adolescent health and nutrition continuum in low- and middle-income countries. These indicators typically have higher reporting volumes and completeness and serve as proxies for broader service delivery patterns.

Outpatient consultations are also included as a proxy for overall health service utilization. Country- or program-specific indicators may be added as needed to reflect national priorities.

Preparing for data extraction¶

This step includes a pre-extraction checklist, review of the DHIS2 configuration, mapping of indicators to data elements, and planning of the extraction timeline. These steps ensure that downstream analyses are based on consistent, well-understood inputs.

Last updated: 26-01-2026 Contact: FASTR Project Team