Maternal Morbidity Dataset
State and subgroup rates of preterm birth and SMM for Medicaid/CHIP for ML
@kaggle.alamshihab075_maternal_morbidity_dataset
State and subgroup rates of preterm birth and SMM for Medicaid/CHIP for ML
@kaggle.alamshihab075_maternal_morbidity_dataset
This dataset provides annual counts and rates of preterm births and severe maternal morbidity (SMM) among Medicaid- and CHIP-covered live births for 2020–2022. Results are reported overall, by state, and by four subpopulation topics (age group, race/ethnicity, disability eligibility, and SMM category) and include suppression and data-quality flags for careful analysis.
Scope & coverage
Population: Medicaid and CHIP enrollees aged 15–49 with a live birth.
Geography: All 50 states, DC, Puerto Rico, U.S. Virgin Islands (race/ethnicity excludes USVI). Guam, American Samoa, Northern Mariana Islands, and some states with TAF quality issues are excluded.
Years: 2020–2022 (current file).
Data sources: T-MSIS Analytic Files (TAF) Release 1 and Race/Ethnicity Imputation Companion File.
SMM measure: Reported per 10,000 Medicaid/CHIP-covered live births.
Key features
State-level and subgroup breakdowns (age, race/ethnicity, disability eligibility, SMM category).
Counts, denominator counts, and calculated rates (ready for modeling).
Data version field (e.g., “TAF Release 1”) to track provenance.
Suppression code “DS” for cell counts 1–10 and “Unusable data” for states/years with TAF issues.
Methodology highlights
Preterm = live birth before 37 completed weeks of gestation.
SMM defined as a qualifying condition recorded within six weeks before/after delivery.
Race/ethnicity: uses state-reported values when valid; otherwise imputed using enhanced BISG.
Disability eligibility assigned using reported eligibility codes and age in year.
Limitations & data quality
Small counts may be suppressed (“DS”) per CMS policy — these cannot be disaggregated.
Some state-year cells marked “Unusable data.” Do not use these for comparisons.
Imputed race/ethnicity may introduce classification uncertainty; consult methodology notes.
Denominators reflect Medicaid/CHIP enrollees with live births — not the full birth population.
CREATE TABLE preterm_birth_and_smm_2020_2022_01172025 (
"year" BIGINT,
"geography" VARCHAR,
"subpopulation_topic" VARCHAR,
"subpopulation" VARCHAR,
"category" VARCHAR,
"count_of_deliveries" VARCHAR,
"denominator_count_of_deliveries" VARCHAR,
"rate_of_deliveries" VARCHAR,
"data_version" VARCHAR
);Anyone who has the link will be able to view this.