Comparing Health Indicators Across Sources

4.5.1 What is an indicator?

An indicator is a defined measure used to describe a condition, activity, output or result. Examples include reporting completeness, reporting timeliness, service coverage, the number of services delivered, and the percentage of facilities meeting a reporting target. Indicators need consistent definitions if we want to compare them meaningfully.

4.5.2 Worked example: reporting completeness

A fictional district expects 40 facility reports and receives 36. Reporting completeness is 36 ÷ 40 × 100 = 90%. A second district expects 50 reports and receives 40. Its reporting completeness is 40 ÷ 50 × 100 = 80%. The first district has a higher reporting completeness percentage. We should still examine whether the expected reporting units and reporting rules are defined consistently in both districts.

4.5.3 Worked example: survey estimates and routine records

At first glance, the facility and outreach figures add up to 1,150 doses. The survey estimates 800 unique children. Can we conclude that the programme vaccinated 1,150 different children? No. A child may receive multiple doses, and the facility and outreach reports may overlap. The survey also uses a different measurement method and may have sampling uncertainty.

Before calculating coverage, we need the defined vaccine dose or service, the eligible population, the reporting period, the geographical coverage, whether outreach activity is included in facility totals, and whether the survey estimates the same target population and period.

4.5.4 Understanding denominator problems

Suppose a fictional programme records 800 eligible children receiving a specified service. The population estimate is 1,000 eligible children. The calculated coverage is 800 ÷ 1,000 × 100 = 80%. If the population estimate is later found to be 1,200, the coverage becomes 800 ÷ 1,200 × 100, about 66.7%. The numerator has not changed. The calculated coverage changed because the denominator changed. This does not automatically establish which estimate is correct. It shows why denominator quality matters.

4.5.5 How to compare indicators responsibly

1. Confirm the definition.
2. Confirm the numerator.
3. Confirm the denominator.
4. Confirm the reporting period.
5. Confirm the geographical coverage.
6. Identify the data collection method.
7. Consider missing information and uncertainty.
8. Explain any limitations.
9. Avoid conclusions that exceed the evidence.

Comparing indicators requires more than comparing percentages. The definitions, populations, methods and reporting periods must be understood.