5 Measuring and Addressing Health Inequities

Why measure health inequities

Health inequities are avoidable and unfair differences in health and in the conditions that shape health. Measuring them helps identify who is not benefiting, where barriers arise, and whether interventions or policies are reducing those barriers. A population average can improve while serious gaps remain, so assessment should consider both overall health and how health is distributed across groups.

Choose measures and groups

Begin with a clearly defined question and choose an Indicator that reflects the issue. Indicators can describe social conditions such as housing stability, income, food security, education, neighborhood conditions, or exposure to discrimination. They can also measure access to care, including insurance coverage, usual source of care, travel time, appointment availability, affordability, or unmet need; care and health outcomes may include screening, treatment, preventable hospitalization, maternal health, disease burden, or mortality.

Disaggregate results by characteristics relevant to the question, such as race and ethnicity, income, education, disability, sex, geography, or housing status. Combining dimensions—for example, examining outcomes by both race and rural residence—can reveal patterns hidden by single-category comparisons. Define groups carefully, explain the definitions, and protect privacy, especially for small populations. The data must link the health indicator to the group characteristics being examined.

Assess data quality

Suitable data sources include population surveys, vital records, health-care records, and community data. Each has limitations: surveys may omit people who are difficult to reach; clinical records describe people who accessed services, not everyone who needed care; and small samples can produce unstable estimates.

Check data for completeness, consistency, and missingness, and assess whether measures are comparable across groups and over time. Where possible, report uncertainty as well as estimates.

Compare disparities

Two common measures compare a group's rate with a reference group's rate. The Absolute difference is calculated by subtracting one group's rate from the other; it retains the indicator's units, such as percentage points or deaths per 100,000 people. If the rates are represented by $R_1$ and $R_2$, their difference is $R_1-R_2$; the order of subtraction determines the direction of the result.

The Rate ratio is calculated by dividing one group's rate by the other's, or $\frac{R_1}{R_2}$. A ratio of $1$ indicates equal rates, while values above or below $1$ indicate the direction and relative size of the difference.

For example, if timely prenatal care reaches $60\%$ of Group A and $80\%$ of Group B, the difference is $20$ percentage points and the ratio for Group A relative to Group B is $0.75$. Absolute gaps can show the scale of unmet need, while relative measures show how large the difference is in proportion to the comparison rate. No single measure fully describes a disparity, so choose measures that fit the question and interpret them together.

Interpret gaps over time

Compare trends as well as snapshots. In a later period, if rates rise to $75\%$ for Group A and $90\%$ for Group B, the gap narrows to $15$ percentage points, but the groups have not reached parity.

Report group-specific rates alongside the disparity measure. A gap can shrink because the less-served group improved, because the better-served group worsened, or because both changed. Benchmarks and population-wide summary measures can add context, but should not obscure the experiences of smaller or disadvantaged groups.

Frame an evaluation

Evaluation asks whether an approach was implemented as intended, whom it reached, what changed, and whether it contributed to the change. Begin by describing the problem, the people affected, the proposed strategy, and the pathway by which it is expected to improve health or reduce inequity. For example, a transportation-support program might reduce missed appointments, improve continuity of care, and eventually improve disease control. Make assumptions in this pathway explicit so the evaluation can test them.

Useful questions include whether the strategy reached groups facing the greatest barriers; whether services were accessible, acceptable, and delivered consistently across communities; and whether access, care, social conditions, or health outcomes improved, and for whom. Assess whether absolute and relative gaps narrowed, remained unchanged, or widened; whether there were unintended harms, added burdens, or unequal effects; and whether benefits were proportionate to the costs and resources required.

Select an evaluation method

Match the evaluation method to the question. Process evaluation examines implementation, reach, and barriers. Outcome evaluation measures whether intended changes occurred, but by itself cannot establish that the strategy caused them. Impact evaluation compares outcomes with a credible estimate of what would have happened without the strategy. When random assignment is not feasible, comparison groups and before-and-after trends may help, provided important differences and other influences are considered. Economic evaluation can compare effects with costs.

Combine evidence and community input

Use quantitative and qualitative evidence together. Disaggregated rates can reveal which groups benefited; interviews, focus groups, observation, and community feedback can help explain why.

For community strategies, evaluate the quality of participation as well as health results. Consider whose input shaped decisions, whether it influenced action, and whether participants experienced the process as respectful and useful. Engage community members and relevant partners in setting questions, interpreting findings, and deciding what to change.

Use findings to improve action

Evaluation is a continuing improvement cycle: assess the context, describe the strategy, focus the questions, gather credible evidence, draw supported conclusions, and act on findings. Share results in clear, accessible formats with the people and organizations who can use them. Collaboration, equity, learning, transparency, rigor, and ethical practice are important throughout the cycle.

Use findings to adapt delivery and monitor again. If a program raises overall access but leaves a gap unchanged, consider whether its location, hours, cost, language access, eligibility rules, or outreach fit the needs of the groups most affected. If outcomes improve but participation is unequal, adapt the approach and continue monitoring.