Skip to content
Contents
ZIP Code Destiny

    Type to search. Planned chapters match by title only.

    Chapter 10.3 · The counties that don't count

    Suppressed counties in death data

    CDC WONDER withheld the death count in 47.1% to 57.9% of county-years for overdose, alcohol-induced deaths and suicide from 2018 to 2023, and the counties it published in every year are selected on their death rates.

    Affiliation
    NYU Stern
    Published

    CDC WONDER, the federal database of death certificates, withheld the drug overdose death count of 591591 countiesEstimandcount of counties with every county-year withheldWeightsnone (full count)Populationcounties whose drug overdose count WONDER withholds in every year 2018 to 2023, other than Connecticut'sPeriodcounty-years in 2018 to 2023SourceCDC WONDER underlying cause of death counties in every year from 2018 to 2023. The median one of those counties had 19,76119,761 residentsEstimandmedian county populationn591 countiesWeightsnone (each county counts once)Populationcounties whose overdose count WONDER withholds in every yearPeriodmean WONDER population 2018 to 2023SourceCDC WONDER underlying cause of death residents. In every one of those years, each county's count was above zero and below the limit WONDER sets for publication, and the county table shows Suppressed in its place. Overdose here means drug poisoning of any intent. A suppressedSuppressionThe withholding of a number because the count behind it is too small to publish without risking the identification of a person. A suppressed value is missing, not zero. cell says only that the count is small, so the table records that the county had at least one overdose death that year and not how many.

    Blanks cover many counties and few residents

    WONDER withheld the death count in 47.1%47.1% percent of county-yearsEstimandshare of county-years whose drug overdose count WONDER withholdsn18,852 county-yearsWeightsnone (full count)Populationcounty-years in the CDC WONDER county files, each kind of death counted separatelyPeriodcounty-years in 2018 to 2023SourceCDC WONDER underlying cause of death to 57.9%57.9% percent of county-yearsEstimandshare of county-years whose alcohol-induced count WONDER withholdsn18,852 county-yearsWeightsnone (full count)Populationcounty-years in the CDC WONDER county files, each kind of death counted separatelyPeriodcounty-years in 2018 to 2023SourceCDC WONDER underlying cause of death of county-years for overdose, alcohol-induced deaths and suicide from 2018 to 2023. A county-year is one county in one year, counted separately for each kind of death. The withheld overdose county-years held only 10.1%10.1% percent of residentsEstimandshare of the summed county populations that sit in county-years whose drug overdose count is withheld (WONDER publishes the population)Weightsnone (full count)Populationresidents counted in the county-years of the WONDER county filesPeriodcounty-years in 2018 to 2023SourceCDC WONDER underlying cause of death of residents, so the blanks cover many places and few people.

    Counties whose overdose count is withheld in every year are smaller, more rural and more deprived than counties whose count is published in every year. The median county published in every year has 134,390134,390 residentsEstimandmedian county populationn942 countiesWeightsnone (each county counts once)Populationcounties whose overdose count WONDER publishes in every yearPeriodmean WONDER population 2018 to 2023SourceCDC WONDER underlying cause of death residents.

    Figure 1

    Counties whose overdose count WONDER publishes in every year hold 83.8% of residents, and counties withheld in every year hold 4.1% and are smaller, more rural and more deprived

    Counties by how many years from 2018 to 2023 WONDER withholds the death count: share of counties and residents, median size, nonmetro residents, Area Deprivation Index rank and race and ethnicity of residents, other than Connecticut's counties

    Share of counties (%)

    Share of residents (%)

    Median county residents (thousands)

    Residents in nonmetro counties (%)

    Mean Area Deprivation Index rank (1 least, 100 most deprived)

    Residents who are Hispanic (%)

    Residents who are non-Hispanic Black (%)

    Residents who are non-Hispanic white (%)

    Residents who are American Indian or Alaska Native (%)

    Value in the unit of each panel title

    Show data
    Data for Counties whose overdose count WONDER publishes in every year hold 83.8% of residents, and counties withheld in every year hold 4.1% and are smaller, more rural and more deprived, Drug overdose
    causeestimategroupmeasuren_units
    Drug overdose30.06Published every yearShare of counties (%)942
    Drug overdose83.81Published every yearShare of residents (%)942
    Drug overdose134.4Published every yearMedian county residents (thousands)942
    Drug overdose4.16Published every yearResidents in nonmetro counties (%)942
    Drug overdose43.04Published every yearMean Area Deprivation Index rank (1 least, 100 most deprived)942
    Drug overdose20.97Published every yearResidents who are Hispanic (%)942
    Drug overdose13.37Published every yearResidents who are non-Hispanic Black (%)942
    Drug overdose55.34Published every yearResidents who are non-Hispanic white (%)942
    Drug overdose1.18Published every yearResidents who are American Indian or Alaska Native (%)942
    Drug overdose51.08Withheld in some yearsShare of counties (%)1,601
    Drug overdose12.09Withheld in some yearsShare of residents (%)1,601
    Drug overdose16.08Withheld in some yearsMedian county residents (thousands)1,601
    Drug overdose61.06Withheld in some yearsResidents in nonmetro counties (%)1,601
    Drug overdose67.71Withheld in some yearsMean Area Deprivation Index rank (1 least, 100 most deprived)1,601
    Drug overdose9.66Withheld in some yearsResidents who are Hispanic (%)1,601
    Drug overdose9.38Withheld in some yearsResidents who are non-Hispanic Black (%)1,601
    Drug overdose75.91Withheld in some yearsResidents who are non-Hispanic white (%)1,601
    Drug overdose2.12Withheld in some yearsResidents who are American Indian or Alaska Native (%)1,601
    Drug overdose18.86Withheld in every yearShare of counties (%)591
    Drug overdose4.10Withheld in every yearShare of residents (%)591
    Drug overdose19.76Withheld in every yearMedian county residents (thousands)591
    Drug overdose75.52Withheld in every yearResidents in nonmetro counties (%)591
    Drug overdose70.17Withheld in every yearMean Area Deprivation Index rank (1 least, 100 most deprived)591
    Drug overdose10.00Withheld in every yearResidents who are Hispanic (%)591
    Drug overdose8.11Withheld in every yearResidents who are non-Hispanic Black (%)591
    Drug overdose77.24Withheld in every yearResidents who are non-Hispanic white (%)591
    Drug overdose2.13Withheld in every yearResidents who are American Indian or Alaska Native (%)591
    Note: Shares of counties and residents and the median are full counts of the 3,134 counties other than Connecticut's 8 and carry no interval. Residents are the mean WONDER population over 2018 to 2023. The other panels are population-weighted means across counties, so an interval describes the spread of county values.Note: Deprivation is the mean of the Area Deprivation Index national ranks of a county's tracts, weighted by 2019-2023 tract population, and the index understates deprivation in high-cost counties. Nonmetro counties have a Rural-Urban Continuum Code of 4 or higher.Note: Race and ethnicity shares come from County Health Rankings 2024 (Census Population Estimates) and use that source's categories: Hispanic of any race, and non-Hispanic Black and white. They describe a county's residents, not the people who died.Source: CDC WONDER underlying cause of death, county files, 2018-2023, Area Deprivation Index 2023, Rural-Urban Continuum Codes 2023 and County Health Rankings 2024; full counts and population-weighted means
    1. Step 1
      Across the panels, counties withheld in every year are smaller and more deprived than counties published in every year. Their residents more often live in nonmetro counties and are more often non-Hispanic white or American Indian or Alaska Native. Counties other than Connecticut's are grouped by the years from 2018 to 2023 in which WONDER withholds the overdose count, and means carry 95% intervals from resampling counties.
    2. Step 2
      Grouped by the years in which WONDER withholds the alcohol-induced count, counties withheld in every year differ from counties published every year in the same direction as for overdose. That holds for size, rurality, deprivation and the four race and ethnicity shares.
    3. Step 3
      Grouped by the years in which WONDER withholds the suicide count, counties withheld in every year again differ from counties published every year in the same direction on every measure. Values in Table 4.

    These are properties of counties and of their residents, and not of the people who died (ecological inferenceEcological inferenceDrawing conclusions about individuals from data about places or groups. A pattern across counties need not hold for the people inside them.). Published counts therefore describe the larger, more urban and less deprived counties and leave out the rural counties where deaths are few.

    How we know

    The counts come from the county files of WONDER's Underlying Cause of Death database for 2018 to 2024, which carry a row for each county, year and kind of death. WONDER publishes a county's population whether or not it withholds the deaths, so the residents behind each blank can be counted. Connecticut's counties are left out of the county comparisons because WONDER has no population for them in 2022 and 2023. The groups of counties are formed on whether WONDER publishes the count, which depends on the deaths themselves, so every comparison between the groups compares sets of counties selected on their deaths. The report gives every estimate with its interval and sample size.

    A zero, a blank and an unnamed county differ

    WONDER publishes a county-year when its count is zero or at least 1010 deaths in a county-yearEstimanddefinition: WONDER withholds a county-year count above zero and below this numberWeightsnone (definition)Populationcounty-years in the CDC WONDER county files, each kind of death counted separatelyPeriodnot applicable (definition)SourceCDC WONDER underlying cause of death deaths. A default WONDER query leaves out county-years with no deaths, and the pulls behind this book asked for zeros, so a zero here is a count and not an absent row.

    A blank is not a zero. It is missing, and the book never fills it in. A withheld cell is counted and left out of both the deaths and the residents of a rate. Filled to the largest count below that limit, the withheld cells could hold at most 13.4%13.4% percent of deaths in the county filesEstimandlargest share of the county files' drug overdose deaths that the withheld cells could hold at the largest withheld count each; the public limit and the published cells onlyWeightsnone (bound from the public limit)Populationdeaths in county-years in the CDC WONDER county filesPeriodcounty-years in 2018 to 2023SourceCDC WONDER underlying cause of death of the deaths in the county files for overdose, 29.1%29.1% percent of deaths in the county filesEstimandlargest share of the county files' alcohol-induced deaths that the withheld cells could hold at the largest withheld count each; the public limit and the published cells onlyWeightsnone (bound from the public limit)Populationdeaths in county-years in the CDC WONDER county filesPeriodcounty-years in 2018 to 2023SourceCDC WONDER underlying cause of death for alcohol-induced deaths and 28.4%28.4% percent of deaths in the county filesEstimandlargest share of the county files' suicide deaths that the withheld cells could hold at the largest withheld count each; the public limit and the published cells onlyWeightsnone (bound from the public limit)Populationdeaths in county-years in the CDC WONDER county filesPeriodcounty-years in 2018 to 2023SourceCDC WONDER underlying cause of death for suicide.

    A third kind of gap belongs to WONDER's birth products. Natality names a county only if it has 100,000100,000 residentsEstimanddefinition: the population at which WONDER natality names a county (and a boundary between county size classes)Weightsnone (definition)Populationcounties that WONDER natality namesPeriod2018 to 2024SourceCDC WONDER natality or more residents, and linked birth and infant death only if it has 250,000250,000 residentsEstimanddefinition: the population at which WONDER linked birth and infant death names a county (and a boundary between county size classes)Weightsnone (definition)Populationcounties that WONDER linked birth and infant death namesPeriod2018 to 2022SourceCDC WONDER linked birth and infant death or more. WONDER pools every other county in a state into one row labeled Unidentified Counties, so a smaller county has no cell at all, and there is nothing to suppress or to recover. In natality, Cape May County NJ and Portsmouth city VA are listed with no births because WONDER has moved them into the pooled row, so their zero is not a count either. The mortality county files name every county, so a missing county there is a blank inside a named county. Chapter 2.1, Birthweight and preterm birth and Chapter 2.2, Infant mortality work with the birth files.

    A pooled rate is biased in a direction set by size

    A rate pooled over a county's published years keeps its years with no deaths and drops its years with a few. Bias here means a difference from the county's true rate in a systematic direction. The chapter works it out under a model in which a county has a constant true rate and its yearly deaths are a Poisson count, the standard model for deaths in a small population.[1] Under that model the pooled rate is below the true rate when a county expects only a few deaths a year, because nearly every published county-year is then a zero. The ratio of the pooled rate to the true rate starts near zero and crosses one near 4.24.2 expected deaths a yearEstimandleast expected number of deaths, before the peak, at which the ratio reaches oneWeightsnone (model)Populationcounties that share one expected number of deaths a year under the model of the publication rulePeriodnot applicable (definition)SourceCDC WONDER underlying cause of death expected deaths. It peaks at 1.81.8 ratio of the pooled rate to the true rateEstimandlargest value of the ratio of the expected pooled rate over published county-years to the true rate, Poisson counts and a constant rateWeightsnone (model)Populationcounties that share one expected number of deaths a year under the model of the publication rulePeriodnot applicable (definition)SourceCDC WONDER underlying cause of death when a county expects 5.45.4 expected deaths a yearEstimandexpected number of deaths at which the ratio is largestWeightsnone (model)Populationcounties that share one expected number of deaths a year under the model of the publication rulePeriodnot applicable (definition)SourceCDC WONDER underlying cause of death deaths a year, where nearly all published county-years have a high count.

    Figure 2

    A rate pooled over published county-years is 1.8 times the true rate at 5.4 expected deaths a year, below it under 4.2 and within 5% of it above 14.2

    Expected pooled rate over the county-years WONDER publishes (counts of zero and of at least 10) divided by the true rate, against the expected number of deaths in a county-year, Poisson counts and a constant rate

    Pooled rate over published county-years relative to the true rate (1 = unbiased)

    Show data
    Data for A rate pooled over published county-years is 1.8 times the true rate at 5.4 expected deaths a year, below it under 4.2 and within 5% of it above 14.2, Bias of the pooled rate
    lampublishedratio
    0.10090.480
    0.20081.870
    0.30074.080
    0.40067.030
    0.50060.650
    0.60054.880
    0.70049.660
    0.80044.930
    0.90040.660
    136.790
    1.1033.290
    1.2030.120
    1.3027.250
    1.4024.660.000
    1.5022.310.000
    1.6020.190.000
    1.7018.270.000
    1.8016.530.001
    1.9014.960.001
    213.540.002
    2.1012.250.003
    2.2011.090.004
    2.3010.040.006
    2.409.090.009
    2.508.240.014
    2.607.460.020
    2.706.770.028
    2.806.150.040
    2.905.590.055
    35.090.075
    3.104.640.101
    3.204.250.134
    3.303.910.177
    3.403.610.230
    3.503.350.295
    3.603.130.372
    3.702.960.463
    3.802.820.567
    3.902.710.683
    42.650.808
    4.102.610.938
    4.202.611.07
    4.302.651.20
    4.402.721.32
    4.502.821.43
    4.602.961.52
    4.703.131.61
    4.803.341.67
    4.903.581.73
    53.861.77
    5.104.171.79
    5.204.521.81
    5.304.911.82
    5.405.331.83
    5.505.791.83
    5.606.281.82
    5.706.821.81
    5.807.391.80
    5.908.001.78
    68.641.77
    6.109.321.75
    6.2010.041.73
    6.3010.801.72
    6.4011.591.70
    6.5012.411.68
    6.6013.271.66
    6.7014.171.64
    6.8015.091.62
    6.9016.051.61
    717.041.59
    7.1018.061.57
    7.2019.111.56
    7.3020.191.54
    7.4021.291.52
    7.5022.411.51
    7.6023.571.49
    7.7024.741.48
    7.8025.931.46
    7.9027.141.45
    828.371.44
    8.1029.621.42
    8.2030.881.41
    8.3032.151.40
    8.4033.431.39
    8.5034.721.37
    8.6036.021.36
    8.7037.331.35
    8.8038.641.34
    8.9039.961.33
    941.271.32
    9.1042.591.31
    9.2043.901.30
    9.3045.221.29
    9.4046.521.28
    9.5047.831.27
    9.6049.121.26
    9.7050.411.25
    9.8051.691.25
    9.9052.951.24
    1054.211.23
    10.1055.461.22
    10.2056.691.22
    10.3057.901.21
    10.4059.101.20
    10.5060.291.20
    10.6061.461.19
    10.7062.611.18
    10.8063.741.18
    10.9064.861.17
    1165.951.16
    11.1067.031.16
    11.2068.081.15
    11.3069.111.15
    11.4070.131.14
    11.5071.121.14
    11.6072.091.13
    11.7073.041.13
    11.8073.971.12
    11.9074.881.12
    1275.761.12
    12.1076.621.11
    12.2077.471.11
    12.3078.281.10
    12.4079.081.10
    12.5079.861.10
    12.6080.611.09
    12.7081.341.09
    12.8082.061.09
    12.9082.751.08
    1383.421.08
    13.1084.071.08
    13.2084.701.07
    13.3085.311.07
    13.4085.901.07
    13.5086.471.07
    13.6087.031.06
    13.7087.561.06
    13.8088.081.06
    13.9088.581.06
    1489.061.05
    14.1089.531.05
    14.2089.971.05
    14.3090.411.05
    14.4090.821.04
    14.5091.221.04
    14.6091.611.04
    14.7091.981.04
    14.8092.341.04
    14.9092.681.04
    1593.021.03
    15.1093.331.03
    15.2093.641.03
    15.3093.931.03
    15.4094.211.03
    15.5094.481.03
    15.6094.741.03
    15.7094.991.03
    15.8095.221.02
    15.9095.451.02
    1695.671.02
    16.1095.881.02
    16.2096.081.02
    16.3096.271.02
    16.4096.451.02
    16.5096.631.02
    16.6096.791.02
    16.7096.951.02
    16.8097.101.02
    16.9097.251.01
    1797.391.01
    17.1097.521.01
    17.2097.651.01
    17.3097.771.01
    17.4097.881.01
    17.5097.991.01
    17.6098.091.01
    17.7098.191.01
    17.8098.291.01
    17.9098.381.01
    1898.461.01
    18.1098.541.01
    18.2098.621.01
    18.3098.691.01
    18.4098.761.01
    18.5098.831.01
    18.6098.891.01
    18.7098.951.01
    18.8099.011.01
    18.9099.061.01
    1999.111.00
    19.1099.161.00
    19.2099.211.00
    19.3099.251.00
    19.4099.291.00
    19.5099.331.00
    19.6099.371.00
    19.7099.411.00
    19.8099.441.00
    19.9099.471.00
    2099.501.00
    20.1099.531.00
    20.2099.561.00
    20.3099.581.00
    20.4099.611.00
    20.5099.631.00
    20.6099.651.00
    20.7099.671.00
    20.8099.691.00
    20.9099.711.00
    2199.721.00
    21.1099.741.00
    21.2099.751.00
    21.3099.771.00
    21.4099.781.00
    21.5099.801.00
    21.6099.811.00
    21.7099.821.00
    21.8099.831.00
    21.9099.841.00
    2299.851.00
    22.1099.861.00
    22.2099.871.00
    22.3099.881.00
    22.4099.881.00
    22.5099.891.00
    22.6099.901.00
    22.7099.901.00
    22.8099.911.00
    22.9099.911.00
    2399.921.00
    23.1099.921.00
    23.2099.931.00
    23.3099.931.00
    23.4099.941.00
    23.5099.941.00
    23.6099.941.00
    23.7099.951.00
    23.8099.951.00
    23.9099.951.00
    2499.961.00
    24.1099.961.00
    24.2099.961.00
    24.3099.971.00
    24.4099.971.00
    24.5099.971.00
    24.6099.971.00
    24.7099.971.00
    24.8099.971.00
    24.9099.981.00
    2599.981.00
    25.1099.981.00
    25.2099.981.00
    25.3099.981.00
    25.4099.981.00
    25.5099.981.00
    25.6099.981.00
    25.7099.991.00
    25.8099.991.00
    25.9099.991.00
    2699.991.00
    26.1099.991.00
    26.2099.991.00
    26.3099.991.00
    26.4099.991.00
    26.5099.991.00
    26.6099.991.00
    26.7099.991
    26.8099.991
    26.9099.991
    2799.991
    27.10100.001
    27.20100.001
    27.30100.001
    27.40100.001
    27.50100.001
    27.60100.001
    27.70100.001
    27.80100.001
    27.90100.001
    28100.001
    28.10100.001
    28.20100.001
    28.30100.001
    28.40100.001
    28.50100.001
    28.60100.001
    28.70100.001
    28.80100.001
    28.90100.001
    29100.001
    29.10100.001
    29.20100.001
    29.30100.001
    29.40100.001
    29.50100.001
    29.60100.001
    29.70100.001
    29.80100.001
    29.90100.001
    30100.001
    30.10100.001
    30.20100.001
    30.30100.001
    30.40100.001
    30.501001
    30.601001
    30.701001
    30.801001
    30.901001
    311001
    31.101001
    31.201001
    31.301001
    31.401001
    31.501001
    31.601001
    31.701001
    31.801001
    31.901001
    321001
    32.101001
    32.201001
    32.301001
    32.401001
    32.501001
    32.601001
    32.701001
    32.801001
    32.901001
    331001
    33.101001
    33.201001
    33.301001
    33.401001
    33.501001
    33.601001
    33.701001
    33.801001
    33.901001
    341001
    34.101001
    34.201001
    34.301001
    34.401001
    34.501001
    34.601001
    34.701001
    34.801001
    34.901001
    351001
    35.101001
    35.201001
    35.301001
    35.401001
    35.501001
    35.601001
    35.701001
    35.801001
    35.901001
    361001
    36.101001
    36.201001
    36.301001
    36.401001
    36.501001
    36.601001
    36.701001
    36.801001
    36.901001
    371001
    37.101001
    37.201001
    37.301001
    37.401001
    37.501001
    37.601001
    37.701001
    37.801001
    37.901001
    381001
    38.101001
    38.201001
    38.301001
    38.401001
    38.501001
    38.601001
    38.701001
    38.801001
    38.901001
    391001
    39.101001
    39.201001
    39.301001
    39.401001
    39.501001
    39.601001
    39.701001
    39.801001
    39.901001
    401001

    The line starts near zero, climbs to a peak at a few expected deaths, then declines toward one and stays near it for larger expected counts. It plots the expected pooled rate over the county-years WONDER publishes, divided by the true rate, against the expected number of deaths in a county-year. It is a calculation from the rule, not an estimate from data, and the second view plots the share of county-years published. Values in Table 6.

    Note: A county-year with a given expected number of deaths is published when its Poisson count is zero or at least 10. The ratio is the average death count in published county-years divided by the average count in all county-years, which is the ratio of the pooled rate to the true rate when counties share a population.Note: The model assumes a constant rate in each county, independent years and Poisson counts. Real counts vary more than Poisson counts, so the curve gives direction and rough size, and it is not used to correct any estimate in this book.Source: Calculation from the publication rule of the CDC WONDER county files, 2018-2024, applied to Poisson counts; full counts of the model, no sampling error, equal weight for every county-year

    The rate of a complete countyComplete countyA county whose death count is published in every year of a window, possibly as zero, so its rate over the window is exact. Counties that are complete are selected by their counts, so as a group they are not typical of counties their size., one whose count is published in every year, is exact for that county, so complete counties escape the bias in their own rates. They do not escape it as a group. Under the same model, the rate pooled over the counties published in every year is the same multiple of the true rate as the rate pooled over published county-years.

    The data follow the model's sign, not its sizes

    Testing the model needs a rate that does not depend on WONDER's rule. CDC's Mapping Injury, Overdose, and Violence panel supplies one for overdose and suicide, with modeled estimatesModeled estimateA figure produced by a statistical model instead of a direct count or measurement, usually for small areas where direct data are thin, borrowing strength from surveys, neighboring areas or other data. Neighboring areas can share errors, so single extremes call for caution. that borrow from neighboring counties where deaths are few. The chapter divides the WONDER rate pooled over published county-years in 2019 to 2023 by the panel's rate for the same years, in counties with a stable panel rate in every year. For suicide the pooled rate is 0.18 (95% CI 0.00 to 0.42)0.18 ratioEstimandthe suicide rate pooled over county-years WONDER publishes (a county never published adds nothing), divided by the panel's population-weighted rate in all counties of the class; bootstrap over countiesInterval95% CI 0.00 to 0.42n406 countiesWeightscounty-years with a published count (numerator); mean WONDER population 2019 to 2023 (panel)Populationcounties with a stable panel suicide rate in every year 2019 to 2023 and under 10,000Periodcounty-years in 2019 to 2023 against annual modeled rates for 2019 to 2023SourceCDC injury and violence mapping times the panel's rate in counties under 10,00010,000 residentsEstimanddefinition: a boundary between county size classesWeightsnone (definition)Populationcounties in the size classes of this chapterPeriodmean WONDER population 2018 to 2023SourceCDC WONDER underlying cause of death residents and 1.51 (95% CI 1.47 to 1.55)1.51 ratioEstimandthe suicide rate pooled over county-years WONDER publishes (a county never published adds nothing), divided by the panel's population-weighted rate in all counties of the class; bootstrap over countiesInterval95% CI 1.47 to 1.55n484 countiesWeightscounty-years with a published count (numerator); mean WONDER population 2019 to 2023 (panel)Populationcounties with a stable panel suicide rate in every year 2019 to 2023 and 25,000 to 49,999Periodcounty-years in 2019 to 2023 against annual modeled rates for 2019 to 2023SourceCDC injury and violence mapping times in counties of 25,00025,000 residentsEstimanddefinition: a boundary between county size classesWeightsnone (definition)Populationcounties in the size classes of this chapterPeriodmean WONDER population 2018 to 2023SourceCDC WONDER underlying cause of death to 50,00050,000 residentsEstimanddefinition: a boundary between county size classesWeightsnone (definition)Populationcounties in the size classes of this chapterPeriodmean WONDER population 2018 to 2023SourceCDC WONDER underlying cause of death residents. In counties of the same size, the overdose rate pooled over counties published every year is 2.09 (95% CI 1.92 to 2.28)2.09 ratioEstimandthe overdose rate pooled over counties published in each year 2019 to 2023, divided by the panel's population-weighted rate in all counties of the class; bootstrap over countiesInterval95% CI 1.92 to 2.28n602 countiesWeightsperson-years (numerator); mean WONDER population 2019 to 2023 (panel)Populationcounties with a stable panel overdose rate in every year 2019 to 2023 and 25,000 to 49,999Periodcounty-years in 2019 to 2023 against annual modeled rates for 2019 to 2023SourceCDC injury and violence mapping times the panel's. The panel counts only unintentional and undetermined-intent poisonings. In counties of the largest size, where every count is published, the overdose ratio of 1.05 (95% CI 1.05 to 1.06)1.05 ratioEstimandthe overdose rate pooled over county-years WONDER publishes (a county never published adds nothing), divided by the panel's population-weighted rate in all counties of the class; bootstrap over countiesInterval95% CI 1.05 to 1.06n273 countiesWeightscounty-years with a published count (numerator); mean WONDER population 2019 to 2023 (panel)Populationcounties with a stable panel overdose rate in every year 2019 to 2023 and 250,000 or morePeriodcounty-years in 2019 to 2023 against annual modeled rates for 2019 to 2023SourceCDC injury and violence mapping measures that difference in definition.

    Figure 3

    A suicide rate pooled over published county-years is 0.18 times the panel's rate in counties under 10,000 residents and 1.51 times in counties of 25,000 to 50,000 residents

    Death rate pooled over the county-years CDC WONDER publishes, 2019 to 2023, divided by the CDC panel's population-weighted modeled rate for 2019 to 2023 in all counties of each size with a stable panel rate, suicide and drug overdose

    • Pooled over published county-years
    • Pooled over counties published every year
    Show data
    Data for A suicide rate pooled over published county-years is 0.18 times the panel's rate in counties under 10,000 residents and 1.51 times in counties of 25,000 to 50,000 residents, Suicide
    causeconf_highconf_lowestimaten_unitsseriessize
    Suicide0.41900.182406Pooled over published county-yearsUnder 10,000
    Suiciden/an/a041Pooled over counties published every yearUnder 10,000
    Suicide1.070.4960.781406Pooled over published county-years10,000 to 24,999
    Suicide1.551.471.51484Pooled over published county-years25,000 to 49,999
    Suicide1.761.391.5612Pooled over counties published every year25,000 to 49,999
    Suicide1.101.071.08379Pooled over published county-years50,000 to 99,999
    Suicide1.181.111.14145Pooled over counties published every year50,000 to 99,999
    Suicide1.011.001.00325Pooled over published county-years100,000 to 249,999
    Suicide1.021.011.01304Pooled over counties published every year100,000 to 249,999
    Suicide1.001.001.00273Pooled over published county-years250,000 or more
    Suicide1.001.001.00273Pooled over counties published every year250,000 or more

    The suicide dots for the two smallest sizes of county lie left of the line at one, the second with an interval that reaches the line. The dot for the next size lies right of it, and the dots for larger counties sit on or near it. Each dot divides a WONDER rate for 2019 to 2023 by the panel's rate for the same years. The rate is pooled over published county-years or over counties published every year, in counties with a stable panel rate in every year, other than Connecticut's. Values in Table 7.

    Note: The panel is a model, so these ratios compare WONDER's rule with a modeled rate and not with the true rate. The panel's rates for small counties borrow from neighboring counties. WONDER counts and panel rates both cover 2019 to 2023. In counties of 50,000 to 100,000 residents the ratios are 1.09 for suicide and 1.18 for overdose, and in counties of 100,000 to 250,000 residents 1.00 and 1.07.Note: Counties are all counties with a stable panel rate in every year 2019 to 2023, other than Connecticut's. A county that WONDER never publishes adds nothing to the pooled WONDER rate and stays in the panel rate. Restricting both sides to counties with a published county-year, the ratios for counties of 25,000 to 50,000 residents are 1.34 for suicide and 1.38 for overdose. A series is plotted for a size of county only when it has at least 10 counties; where every complete county has a rate of zero, the ratio is zero and has no interval.Note: Overdose in the panel is unintentional or undetermined-intent drug overdose and in WONDER drug poisoning of any intent, so the levels differ by definition. In counties of 250,000 or more residents, where every count is published, the overdose ratio is 1.05, which measures that difference.Source: CDC WONDER underlying cause of death, county files, 2019-2023, and CDC Mapping Injury, Overdose, and Violence county file, 2019-2023; crude rates, modeled estimates, population-weighted

    The data match the sign pattern and not the sizes of the model, which assumes one rate for every county of a given expected count. The panel is itself a model, so these ratios compare WONDER's rule with a modeled rate and not with the true rate. The panel's rates for counties that WONDER never publishes are modeled in every year.

    What a published county rate describes

    A county rate built from WONDER counts describes the counties and years that cleared the publication rule. A county of middling size has to have a high count in every year to be published in every year, and a very small county has to have no deaths at all. The rule selects on the outcome in both directions. Every chapter of this book that uses WONDER therefore leaves withheld cells out and reports how many there were and the largest share of deaths they could hold. It draws county rates from complete counties or from a modeled source. National series are summed from the state files, which have no withheld cell. The maps of the three kinds of death are in Chapter 8.2, Overdose, alcohol and suicide. A rate among counties describes places and is not any resident's risk.

    What this cannot tell you

    Withheld counts cannot be recovered, and the bound printed above is a ceiling that says nothing about any one cell. The bias calculation assumes Poisson counts, a constant rate and independent years, and real counts vary more than that, so the curve gives direction and rough size and corrects no estimate. The comparisons with the panel show the direction of the selection and cannot give the true rate of a withheld county. The sizes of county and their boundaries are choices, and the figures move with them.

    Go deeper

    • The report: every estimate with its interval and sample size.
    • Tables: every figure of the report as a table.

    Data sources

    • CDC WONDER underlying cause of deathCDC WONDER underlying cause of death, county files for drug poisoning (overdose), alcohol-induced causes and suicide, 2018 to 2024, and state files for the same causesYears: 2018-2024 · Weights: none (full counts of deaths; crude rates); WONDER county population · License: Public domain (NCHS, CDC); WONDER data-use restrictions apply
    • CDC WONDER natalityCDC WONDER natality, county totals 2018 to 2024: births in the counties that WONDER names and in the pooled Unidentified Counties rowsYears: 2018-2024 · Weights: none (full counts of births) · License: Public domain (NCHS, CDC); WONDER data-use restrictions apply
    • CDC WONDER linked birth and infant deathCDC WONDER linked birth and infant death, county totals 2018 to 2022: births in the counties that WONDER names and in the pooled rowsYears: 2018-2022 · Weights: none (full counts of births) · License: Public domain (NCHS, CDC); WONDER data-use restrictions apply
    • CDC injury and violence mappingCDC Mapping Injury, Overdose, and Violence county file: modeled overdose and suicide death rates, annual 2019 to 2023Years: 2019-2024 for counties, 2022-2024 for tracts (data as of 2026-05-14) · Weights: mean WONDER population of each county · License: Public domain (CDC, U.S. Department of Health and Human Services)
    • County Health RankingsCounty Health Rankings 2024: counties with no rate for overdose, suicide, homicide, firearm deaths and premature death, and race and ethnicity shares of residents (Census Population Estimates)Years: 2024 release · Weights: county population in County Health Rankings · License: CC BY 4.0 (University of Wisconsin Population Health Institute)
    • Area Deprivation IndexNeighborhood Atlas Area Deprivation Index 2023, tract national rank averaged to countiesYears: 2023 release (version 4.0.1; built on ACS 2019-2023) · Weights: 2019-2023 tract population · License: Free for non-commercial use; cite the Neighborhood Atlas and Kind and Buckingham
    • Rural-Urban Continuum CodesUSDA Rural-Urban Continuum Codes 2023, countyYears: 2023 · Weights: none (classification) · License: Public domain (U.S. federal)
    • American Community SurveyTotal population of 2020 tracts, 2019-2023, summed to counties as population for deprivation and for the shares of residents in the tableYears: 2019-2023 (5-year) · Weights: none (full count) · License: Public domain (U.S. Census Bureau)

    All data sources

    References

    1. Brillinger, David R. (1986). The natural variability of vital rates and associated statistics. Biometrics 42(4): 693-734. doi:10.2307/2530689

    How to cite

    @misc{singh2026suppressedcountiesindeathdata,
  author = {Singh, Vishal},
  title = {Suppressed counties in death data},
  year = {2026},
  url = {https://vishalsingh.org/zipcode/suppressed-counties-in-death-data},
  note = {In ZIP Code Destiny. Published 9 October 2026.}
}