Congenital Syphilis

Limitations

Congenital syphilis cases per 100,000 live births, not per resident, by county of residence, 2000-2023. Georgia's cases rose from 23 in 2015 to 127 in 2023. OASIS hides county counts of 1-4 and nonzero counts on under 100 births (NULL), and withholds rates on 1-4 cases. Race groups' cases or births are NULL where over 20% of the place-year's lack race (our mask; it can hide a 0). Race groups are presumably the mother's; male and female are the baby's sex, with no births count. OASIS publishes no age-adjusted rate for this cause. Georgia's rate went from 17.5 in 2015 to 101.6 in 2023 (5.8 times; cases 5.5 times); the fewest cases were 2 in 2005, and 2023 is the most. In 2023 77 of the 127 cases (61%) were black, 33 white and 12 hispanic. Counts are new cases reported to DPH (OASIS's definitions: incidence, by place of residence). OASIS does not say whose residence, race or year it uses for this cause; CDC's national congenital syphilis counts use the mother's residence, the mother's race and Hispanic ethnicity, and the year of birth, and CDC's case definition also counts syphilitic stillbirths, while the denominator is live births. num_births is OASIS's births module's count for the place, year and group (all, and the seven race groups by the mother's race and ethnicity): cases x 100,000 / births reproduces every published rate. Male and female rates are OASIS's, per 100,000 live births of that sex (each implies a whole number of births, and the two add up to the place's births), but OASIS publishes no births by sex, so num_births is NULL there and they cannot be pooled. OASIS hides a county's count of 1-4 cases (code -7) and, at every level, the count of a group with fewer than 100 live births in the place that has a case (-6), with its rate; it shows Georgia's own small counts and withholds every rate on 1-4 cases. The platform adds no disclosure suppression of its own and computes nothing that would reveal a hidden count; its one quality mask, on race groups (below), can NULL a published count, 0 included. In 2023, 8 counties show cases, 40 are hidden and 111 had none: county rates rest on few cases and swing widely, so combine years before comparing counties. A place and group with no live births (num_births 0) has no rate (NULL): OASIS shows 0 there, but a rate on no births is undefined (our mask; the count stays 0). Race groups follow the platform convention from OASIS's race x ethnicity cross: 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 wherever every count is published; it has no births and so no rate, and it is NULL wherever OASIS hides a count it subtracts (all, hispanic or a non-Hispanic race). Where OASIS hides only its own unknown race or ethnicity cells, race_unknown is still given (its own table gives the same figure), so in a county it can be 1-4. Georgia's race_unknown is published every year. Our race mask (OASIS's own 20% threshold for missing items) acts on the seven race groups of a place and year, one count at a time: where race or ethnicity is unknown for more than 20% of the cases (race_unknown / all), their case counts are NULL, since each would miss its share of the unknowns; where the mother's race is unknown for more than 20% of the live births (all's births minus the race groups'), their num_births are NULL, since each denominator would run low; and their rate is NULL wherever either count is. A place-year with no case has no case share, so its zeros stay. The all, sex and race_unknown rows are kept. 113 place-years are masked (never Georgia): 16 by cases, 97 by births. Where a count is hidden, the case share is bounded on OASIS's whole race x ethnicity table of the place and, where that leaves it open, on every county's table together (OASIS's county total is the counties' sum), and masked only where surely over 20%, including 14 county-years whose total OASIS hides (race_unknown at least 1 of at most 4 cases). The 1 county-year whose own table leaves the share straddling 20% is bounded on the county tables together: 0 surely over 20% (masked), 1 at most 20% (kept); 0 still straddle it (unmasked). pacific_islander has no rows in 2000-2007 and multiracial none in 2000-2007 and 2009: they were not recorded on STD case reports (no STI case of any kind anywhere those years; our reading). Race-group rates divide by births by the mother's race, whose recording changed with the revised birth certificate in mid-2007: births to multiracial mothers were 237-648 a year in 2000-2006, then 3,538 in 2007 and 6,961 in 2008, while births to white mothers fell from 66,940 (2007) to 57,568 (2008), so don't read race-group rates across 2007-2008 as trends. Cases of unknown sex have no platform group: they count in all only (Georgia 2000). From 2005 OASIS places cases by geocoded address; 2000-2004 use the county the case report states (OASIS's data issues page), so county series can shift between 2004 and 2005. OASIS has revised congenital syphilis counts after first release (2009-2012, 2014 and 2019; its data issues page), so older reports can differ. Not served: OASIS's age-adjusted rate (never published for this cause), unknown sex and age (cut platform-wide). 2023 is the newest year OASIS's STD module offers. tract_geoid is always NULL.