PU685 Unit 4 Disparity Data Analysis Example

A Ratio of 1.84, a Difference of 52 per 10,000 and 74 Excess Events: A Unit 4 Disparity Data Analysis of Severe Maternal Morbidity in a Composite State, Stated Both Ways Before Any Explanation Is Weighed

[Student Name]

Purdue University Global

PU685: Issues and Concepts in Maternal and Child Health

Unit 4 Disparity Data Analysis

[Instructor Name]

[Date]

The state, its hospital discharge counts and its strata are composite; the outcome definition, the methods and the research cited are real.

The Outcome and the Data

Severe maternal morbidity is identified in hospital discharge data with the federal code-based definition: 21 indicators of unexpected outcomes of labor and delivery, such as eclampsia, acute renal failure, sepsis, hysterectomy and blood transfusion, recorded during the delivery hospitalization. The state's two most recent years of discharge data include 14,200 deliveries to non-Hispanic Black women and 46,800 to non-Hispanic White women. Race is as recorded in the discharge record.

Counts, Rates and the Gap Stated Three Ways

1. Black women: 162 events in 14,200 deliveries, 114.1 per 10,000.

2. White women: 290 events in 46,800 deliveries, 62.0 per 10,000.

3. Rate ratio: 1.84, 95 percent confidence interval 1.52 to 2.23.

4. Rate difference: 52.1 per 10,000 deliveries.

5. Excess events: if Black women had experienced the White rate, about 88 events would have been expected; 162 occurred, so 74 are excess.

The ratio and the difference answer different questions and can move independently. If both rates fell by half, the ratio would stay at 1.84 while the difference would shrink to about 26 per 10,000 and the excess to about 37 events, a real improvement that a ratio alone would hide. If both rose, the opposite. Reporting both, plus a count of events, keeps the disparity tied to people. Seventy-four is the number that a hospital quality committee can picture, and it is the one most disparity reports leave out.

Stratified by Payer

1. Medicaid deliveries: Black 107 in 9,100, 117.6 per 10,000; White 116 in 15,600, 74.4. Ratio 1.58, interval 1.22 to 2.05.

2. Privately insured deliveries: Black 55 in 5,100, 107.8 per 10,000; White 174 in 31,200, 55.8. Ratio 1.93, interval 1.43 to 2.61.

The most common objection to a racial disparity in maternal outcomes is that it reflects insurance or income. Stratifying by payer is a partial test. The ratio narrows among Medicaid deliveries and widens among privately insured ones, and it persists in both. Privately insured Black women had a higher rate than White women covered by Medicaid. Payer is a rough proxy for income and does not capture wealth, neighborhood or education, so this test cannot rule out socioeconomic explanations, but it shows that insurance status alone does not account for the gap.

Without Transfusion-Only Cases

Transfusion is the most common indicator and the most sensitive to practice differences between hospitals. Excluding cases identified by transfusion alone leaves 129 events among Black women and 210 among White women: 90.8 against 44.9 per 10,000, a ratio of 2.02, interval 1.63 to 2.52. The disparity grows when the most discretionary indicator is removed, which argues against the gap being an artifact of transfusion practice.

Four Explanations, Weighed

Chronic conditions before pregnancy. Hypertension, diabetes and obesity are more prevalent among Black women entering pregnancy and are strong risk factors for severe morbidity. In California data, the prevalence of severe maternal morbidity was 1.63 percent among Black women and 0.84 percent among White women, and after accounting for comorbidities, anemia, cesarean birth and other characteristics the adjusted risk ratio fell to 1.27 but remained elevated (Leonard et al., 2019). Comorbidities explain part of the gap, not all of it. This state's discharge data can test this directly by adjusting for coded comorbidities, which is the next analysis. Where women deliver. In New York City, 74 percent of Black deliveries occurred at hospitals serving high or medium shares of Black patients, and those hospitals had higher severe morbidity rates even after adjustment for patient characteristics (Howell et al., 2016). If the same concentration holds in this state, part of the disparity is a difference in hospital quality rather than in the women. The state can test it by computing the share of Black deliveries by hospital and estimating how much of the gap remains within hospitals. Quality of care within the same hospital. Even within hospitals, differences in recognition of warning signs, timeliness of treatment and patient experience have been described, and the residual adjusted gaps in both studies above leave room for them. Discharge data cannot measure this; chart review or maternal mortality review committee findings can. Structural racism and cumulative stress. The persistence of the gap among privately insured women and after adjustment is consistent with explanations that operate over women's lives, including residential segregation and chronic stress. These are harder to measure and should be stated as plausible contributors rather than demonstrated causes in an analysis of discharge data.

Why the Explanation Matters for Action

The four explanations point to different remedies, which is why weighing them is not an academic exercise. If chronic conditions drive most of the gap, the remedy lies before pregnancy, in blood pressure and diabetes care for women of reproductive age. If hospital of delivery drives it, the remedy is raising quality at the hospitals where Black women most often give birth. If care within hospitals differs, the remedy is standardized protocols that leave less to individual judgment. There is evidence that the third route works. In a California quality collaborative on obstetric hemorrhage, severe morbidity among women with hemorrhage was 28.6 percent for Black women and 19.8 percent for White women at baseline, both groups improved after hospitals adopted standardized hemorrhage protocols, and the absolute reduction was larger for Black women, 9.0 against 2.1 percentage points (Main et al., 2020). Standardizing care for one complication narrowed the gap for that complication. That result suggests the state's hospitals could begin with hemorrhage and hypertension protocols while the comorbidity and hospital analyses are completed, since those protocols are likely to help whichever explanation turns out to carry the most weight.

What the Data Support, and What Comes Next

The data support three statements: the disparity is large in relative and absolute terms, it persists within payer groups, and it is not a transfusion artifact. They do not yet show how much is explained by comorbidities or by hospital of delivery. The next two analyses are adjustment for coded comorbidities and decomposition by hospital, both possible with the same discharge file.

Limits

Discharge coding varies between hospitals. Race is recorded administratively and may be misclassified. Two years of data give stable overall rates but wide intervals within strata.

References

Howell, E. A., Egorova, N., Balbierz, A., Zeitlin, J., & Hebert, P. L. (2016). Black-white differences in severe maternal morbidity and site of care. American Journal of Obstetrics and Gynecology, 214(1), 122.e1-122.e7. https://doi.org/10.1016/j.ajog.2015.08.019

Leonard, S. A., Main, E. K., Scott, K. A., Profit, J., & Carmichael, S. L. (2019). Racial and ethnic disparities in severe maternal morbidity prevalence and trends. Annals of Epidemiology, 33, 30-36. https://doi.org/10.1016/j.annepidem.2019.02.007

Main, E. K., Chang, S. C., Dhurjati, R., Cape, V., Profit, J., & Gould, J. B. (2020). Reduction in racial disparities in severe maternal morbidity from hemorrhage in a large-scale quality improvement collaborative. American Journal of Obstetrics and Gynecology, 223(1), 123.e1-123.e14. https://doi.org/10.1016/j.ajog.2020.01.026