Fetal Deaths and Fetal Mortality Rates

Limitations

Fetal deaths at 20+ weeks of gestation, by the mother's county of residence; rates are per 1,000 live births plus fetal deaths. From 2008 OASIS dates gestation by the obstetric estimate, lowering the rate: compare across 2008 with care. Race is the mother's; in 2000-2010 9-17% of fetal deaths a year lack it, so Georgia's race-group rates then run low; race-group rates are NULL where over 20% of fetal deaths or births lack it (our mask). OASIS withholds rates on 1-4 fetal deaths (NULL here). A fetal death is OASIS's: a death before complete delivery, with no sign of life after it, reported on a fetal death certificate; OASIS counts those at 20 or more weeks of gestation, of Georgia residents wherever the fetal death occurred. num_live_births_and_fetal_deaths is OASIS's live births (its births module, the same place, year and group, the mother's race and age) plus the fetal deaths, our sum; num_fetal_deaths x 1,000 / num_live_births_and_fetal_deaths reproduces every published rate, which OASIS defines the same way (live births = num_live_births_and_fetal_deaths - num_fetal_deaths). Gestational age: on 2015-12-15 OASIS moved from a gestational age based on the last menstrual period to the obstetric estimate (the national standard) and applied it back to 2008; its known data issues say the new method lowers gestational ages slightly ('~0.5 per 1,000'), so the fetal mortality rate is lower than in 2007 (https://oasis.state.ga.us/dtt/content/KDIFetalDeath.htm). Georgia's fetal deaths fell from 1,292 (8.49) in 2007 to 1,088 (7.37) in 2008, its largest one-year fall; its rate was 8.29-9.63 a year in 2000-2007 and 7.12-8.36 since. Race groups are the mother's race and ethnicity in the platform convention: hispanic is Hispanic of any race; white, black, asian, native_american (American Indian or Alaska Native), pacific_islander (Native Hawaiian or Other Pacific Islander) and multiracial are non-Hispanic. race_unknown is our calculation, the total minus those seven (unknown ethnicity, or non-Hispanic of unknown race), so the race groups add up to all where every group's count is published (our race mask NULLs them in the place-years where race is mostly unknown); it has no rate. demographic and mother_age are never crossed: a race group's row has mother_age = 'all', and a mother's age row has demographic = 'all'. Both the fetal deaths and the live births carry the mother's race and ethnicity. Fetal death certificates leave the mother's ethnicity or race unrecorded far more often than birth certificates: race_unknown is 9-17% of Georgia's fetal deaths a year in 2000-2010 (at most 7% since) but at most 5% of its live births plus fetal deaths, so those fetal deaths fall in no race group while their group's births stay in the denominator. The births' race recording break of 2007-2008 reaches the denominator: multiracial live births plus fetal deaths were 654 in 2006, 3,542 in 2007 and 6,970 in 2008 (see the births topic). Counties break harder and for longer: in 84 county-years of 2000-2012 (17 counties, most often Catoosa, Dade and Walker on the Tennessee border) more than 25% of the live births plus fetal deaths have no recorded race or ethnicity of the mother, up to 78%. There the live births lack it more often than the fetal deaths, so, unlike Georgia's, those counties' race-group rates ran high: every nonzero race-group rate OASIS published in those county-years (5) is above the county's all-group rate (Ben Hill's black rate in 2009 was 86.2: 5 fetal deaths, 58 live births plus fetal deaths, against 16.6 for all); they are NULL here. The rate counts two recorded events with the mother's race on each, so our mask (OASIS's own 20% threshold for missing items) measures the unknown share on each, on the mother_age = 'all' rows: where race or ethnicity is unknown for more than 20% of a place-year's fetal deaths, every race group's fetal deaths for that place and year are NULL; its denominator (which contains them) is NULL there and also where the mother's race is unknown for more than 20% of the live births (the county-years the births topic masks), and the rate goes with the denominator. Each group would miss its share of the unknowns, so its rate runs low or high with the recording. A place-year with no fetal deaths has no fetal deaths' share: its 0s stay. The all, mother's age and race_unknown rows are kept. That covers 399 place-years (2000-2025, 122 counties, 44 of them from 2013; never Georgia): 328 on the fetal deaths' share (a county's few fetal deaths) and 97 on the live births' (71 of them keep their fetal deaths). Compare a county's race-group rates over time from 2013 only (at most 17% since). mother_age is the mother's age at the delivery in OASIS's groups (10_14, 15_17, 18_19, then five-year groups to 40_44, and 45_55); the groups add up to all. Rates on few deliveries swing widely: the highest county fetal mortality rate (all groups) is 63.6 (county 13193, 2025: 7 fetal deaths, 110 live births plus fetal deaths), and the highest county rate of any row is 157.9 (county 13001, 2002, mother_age 30_34: 6 fetal deaths, 38 live births plus fetal deaths). 78% of rows with at least one fetal death have no rate (1-4 fetal deaths in small counties and groups, or a race group's denominator our mask NULLs). OASIS publishes every count, however small, and withholds rates on 1-4 fetal deaths (NULL here); the platform adds no small-cell suppression of its own (its race mask is a quality mask) and never recomputes a withheld rate. A place and group with neither live births nor fetal deaths (num_live_births_and_fetal_deaths 0) has no rate (NULL): OASIS shows 0 there, but the rate is undefined and would rank as the lowest (our mask). Counties add up to Georgia, and the race groups (with race_unknown) and mother's ages to all, except a race group's counts in the place-years our race mask NULLs them: there a race group's counties fall short of Georgia's row. Not served: OASIS's SES-vulnerability rows (its SES rates here divide by each group's own live births plus fetal deaths, but the platform's demographics have no SES category yet); the fetal death page has no education, payer or sex breakdown. 2025 is the newest year. tract_geoid is always NULL.