NR 555 Week 3 Disparity Data Analysis Example

Disparity Data Analysis: Postpartum Visits Among 15,000 Medicaid-Covered Deliveries Split by Race and Ethnicity, With the Overall Row Kept in View and White Enrollees as the Stated Reference

Student Name

College of Nursing, Chamberlain University

NR 555: Health Policy: Advancing Health Equity & Outcomes

Instructor Name

Month Day, Year

Disparity Data Analysis: Postpartum Visits Among 15,000 Medicaid-Covered Deliveries Split by Race and Ethnicity, With the Overall Row Kept in View and White Enrollees as the Stated Reference

One Outcome, One Split

This analysis uses a single outcome, a postpartum visit in the window from day 7 to day 84 after delivery, and a single grouping variable, race and ethnicity as recorded on the Medicaid application. It covers the 15,000 deliveries counted in our state's Medicaid postpartum care measure last year. A second split, for example by plan or by county, would produce small cells and a table too complex to read; those analyses can follow once the main gap is established. The limits of the measure's denominator, particularly its continuous enrollment rule, were set out in the measure appraisal and apply throughout.

The Table

Table 1 shows counts, rates and differences, with the overall row kept in place.

Postpartum visit within 7 to 84 days, Medicaid-covered deliveries, one year

Table 1

The Overall Row Stays

The overall rate of 67.0% is the figure the state usually reports. Printed beside the split, it shows how an average conceals a gap. A reader told only that two-thirds of enrollees had a postpartum visit would not know that the rate among Black enrollees is 16 points lower than among white enrollees. Removing the overall row would make the table tidier but would hide the very comparison that makes the case for looking below the average.

A Comparison Group With a Reason

White non-Hispanic enrollees are used as the reference group because they are the largest group and have the highest rate, so differences from them show how far each group is from the best rate currently achieved in the program. This choice follows common practice in disparities measurement, but it is a choice, not a law. An alternative would be to measure every group against the overall rate, which would make the gaps look smaller, because the overall rate already includes the lower-rate groups. Another would be to measure against a target, such as a state goal of 80%, which would show every group falling short. Braveman (2006) noted that choices like these shape how large a disparity appears and should be stated. Here, the reference was chosen to answer a specific question: how far is each group from what the program already achieves for its best-served group?

Absolute and Relative

The table reports both both the percentage-point difference and the ratio of rates. The absolute difference of 16 points for Black enrollees means that, if Black enrollees had the reference rate, about 672 more of the 4,200 deliveries would have been followed by a visit. The rate ratio of 0.78 means Black enrollees had a visit at 78% of the rate of white enrollees. Absolute differences translate most directly into missed visits and workload; ratios are more comparable across outcomes with different baseline rates. Showing the two side by side stops either being picked for impact.

Precision and Missing Data

With several thousand deliveries per group, the rates are precise enough that a 16-point gap is far larger than chance variation would produce. The "other or unknown" group is more problematic. It combines people who selected a smaller category, such as Asian or American Indian, with people whose race was not recorded, and the two may have very different experiences. At 900 deliveries it is too mixed to interpret, and the state should report the known smaller groups separately where counts allow and work to reduce missing data.

What a Second Split Would Add

Once the main gap is established, a second analysis could split each racial and ethnic group by managed care plan, by county or by the hospital where the birth occurred. That would show whether the gap is uniform or concentrated. If one plan has a small gap and another a large one, the difference points to plan practices, such as care coordination or network adequacy, that could be copied. If the gap is concentrated in births at particular hospitals, it points to discharge and scheduling practices. The cost of a second split is smaller cells: with four racial and ethnic groups, five plans and three hospitals, some cells would have fewer than 100 deliveries, and rates based on so few births are unstable. The second split should therefore be run for Black and white enrollees only, where counts are large enough, and reported with confidence intervals.

Change Over Time

A single year shows the size of the gap but not its direction. The state has reported the postpartum care rate by race for only two years. Last year's gap of 16 points compares with 17 points the year before, a change too small to distinguish from year-to-year variation. The more important comparison will come after the 12-month extension has been in place for a full measurement year. Because the extension changes who is counted as well as who is seen, the trend analysis should report the size of the denominator by group each year, so that a change in the gap can be attributed to care rather than to counting.

What the Analysis Shows and Does Not Show

The analysis shows a clear gap: Black enrollees are substantially less likely than white enrollees to have a postpartum visit within 12 weeks, and Hispanic enrollees fall in between. It does not show why. The gap could reflect coverage loss, scheduling practices, transport, clinic location, experiences of discrimination or other factors, and the next analysis must examine those mechanisms. Bailey and colleagues argued that structural racism operates through interconnected systems, including insurance, health care and housing, and that analyses must move from describing gaps to identifying the institutional arrangements that produce them (Bailey et al., 2017).

The analysis also does not include people excluded by the continuous enrollment rule, who are disproportionately Black. The true gap among all Medicaid-covered births is probably larger than 16 points. Given the higher risk of pregnancy-related death among Black women in the United States (Petersen et al., 2019), the gap in postpartum visits is a gap in the care most likely to detect serious complications.

References

Bailey, Z. D., Krieger, N., Agénor, M., Graves, J., Linos, N., & Bassett, M. T. (2017). Structural racism and health inequities in the USA: Evidence and interventions. The Lancet, 389(10077), 1453-1463. https://doi.org/10.1016/S0140-6736(17)30569-X

Braveman, P. (2006). Health disparities and health equity: Concepts and measurement. Annual Review of Public Health, 27, 167-194. https://doi.org/10.1146/annurev.publhealth.27.021405.102103

Petersen, E. E., Davis, N. L., Goodman, D., Cox, S., Syverson, C., Seed, K., Shapiro-Mendoza, C., Callaghan, W. M., & Barfield, W. (2019). Racial/ethnic disparities in pregnancy-related deaths: United States, 2007-2016. MMWR. Morbidity and Mortality Weekly Report, 68(35), 762-765. https://doi.org/10.15585/mmwr.mm6835a3