Dataset Viewer
Auto-converted to Parquet Duplicate
submission
stringclasses
1 value
variable
stringlengths
3
17
position
int32
1
143
start
int32
1
393
length
int32
1
13
record_length
int32
394
394
sas_label
stringlengths
3
52
naaccr_name
stringlengths
0
38
⌀
naaccr_item
stringlengths
2
4
⌀
applicable_years
stringclasses
26 values
entry
int32
1
141
⌀
title
stringlengths
3
52
⌀
seerstat_name
stringlengths
3
52
⌀
field_description
stringlengths
33
1.99k
⌀
nov2018
PUBCSNUM
1
1
8
394
Patient ID
Patient ID number
20
1
PATIENT ID NUMBER
Patient ID
This field is used in conjunction with SEER registry to uniquely identify a person. One person can have multiple primaries but the same Patient ID. See the sequence number for more information about the primary. This is a dummy number and is not the number used by the registry to identify the patient. The same number i...
nov2018
REG
2
9
10
394
SEER registry
Registry ID
40
2
REGISTRY ID
SEER Registry
A unique code assigned to each participating SEER registry. The number identifies the registry sending the record and what population the data are based on.
nov2018
MAR_STAT
3
19
1
394
Marital status at diagnosis
Marital Status at DX
150
3
MARITAL STATUS AT DX
Marital status at diagnosis
This data item identifies the patient’s marital status at the time of diagnosis for the reportable tumor.
nov2018
RACE1V
4
20
2
394
Race/ethnicity
Race/Ethnicity
160
4
RACE / ETHNICITY
Race/ethnicity
Recode which gives priority to non-white races for persons of mixed races. Note that not all codes were in effect for all years.
nov2018
NHIADE
5
23
1
394
NHIA Derived Hisp Origin
null
null
null
null
null
null
null
nov2018
SEX
6
24
1
394
Sex
Sex
220
5
SEX
Sex
This data item identifies the sex of the patient at diagnosis.
nov2018
AGE_DX
7
25
3
394
Age at diagnosis
Age at diagnosis
230
6
AGE AT DIAGNOSIS
Age at diagnosis
This data item represents the age of the patient at diagnosis for this cancer. The code is three digits and represents the patient’s actual age in years.
nov2018
YR_BRTH
8
28
4
394
Year of birth
Year of Birth
240
7
BIRTHDATE—YEAR
Year of birth
The SEER dates on this file do not have the corresponding date flag included.
nov2018
SEQ_NUM
9
35
2
394
Sequence number
Sequence Number—Central
380
8
SEQUENCE NUMBER--CENTRAL
Sequence number
Sequence Number-Central describes the number and sequence of all reportable malignant, in situ, benign, and borderline primary tumors, which occur over the lifetime of a patient. The sequence number may change over the lifetime of the patient. If an individual previously diagnosed with a single reportable malignant neo...
nov2018
MDXRECMP
10
37
2
394
Month of diagnosis
Month of diagnosis
390
9
MONTH OF DIAGNOSIS
Month of diagnosis
The month of diagnosis is the month the tumor was first diagnosed by a recognized medical practitioner, whether clinically or microscopically confirmed. For analysis purposes, months coded to 99 (unknown) have been replaced with values 01 through 12. Allowable values MM = 01 – 12
nov2018
YEAR_DX
11
39
4
394
Year of diagnosis
Year of diagnosis
390
10
YEAR OF DIAGNOSIS
Year of diagnosis
The year of diagnosis is the year the tumor was first diagnosed by a recognized medical practitioner, whether clinically or microscopically confirmed. Allowable values YYYY = 1975 – forward
nov2018
PRIMSITE
12
43
4
394
Primary site ICD-O-2 (1973+)
Primary Site
400
11
PRIMARY SITE
Primary Site
This data item identifies the site in which the primary tumor originated. See the International Classification of Diseases for Oncology, Third Edition (ICD-O-3) for topography codes. The decimal point is eliminated. Cases diagnosed 1977-1991 were coded using the International Classification of Diseases for Oncology, 19...
nov2018
LATERAL
13
47
1
394
Laterality
Laterality
410
12
LATERALITY
Laterality
Laterality describes the side of a paired organ or side of the body on which the reportable tumor originated. Starting with cases diagnosed January 1, 2004 and later, laterality is coded for select invasive, benign, and borderline primary intracranial and CNS tumors. See http://seer.cancer.gov/manuals/primsite.laterali...
nov2018
HISTO2V
14
48
4
394
Histologic Type ICD-O-2
Histology (92-00) ICD-O-2
420
13
HISTOLOGY (92-00) ICD-O-2
Histology ICD-O-2
NOTE: ALL CASES DIAGNOSED 1975-91 HAVE BEEN CONVERTED TO THE INTERNATIONAL CLASSIFICATION OF DISEASES FOR ONCOLOGY, SECOND EDITION (ICD-O-2, 1992). The SEER program has used several different, but related, coding systems for morphology over time. One should be extremely careful when doing any analysis of trends related...
nov2018
BEHO2V
15
52
1
394
Behavior Code ICD-O-2
Behavior (92-00) ICD-O-2
430
14
BEHAVIOR (92-00) ICD-O-2
Behavior code ICD-O-2
Behavior codes are also defined in ICD-O-2, 1992. Note: For bladder only, all in situs (/2) are converted to invasives (/3) before inclusion on this file. Cervix in situ not required. FOR CASES DIAGNOSED AFTER 2000: All cases diagnosed after 2000 were converted from ICD-O-3 to ICD-O2.
nov2018
HISTO3V
16
53
4
394
Histologic Type ICD-O-3
Histologic Type ICD-O-3
522
15
HISTOLOGIC TYPE ICD-O-3
Histologic Type ICD-O-3
The data item Histologic Type describes the microscopic composition of cells and/or tissue for a specific primary. The tumor type or histology is a basis for staging and determination of treatment options. It affects the prognosis and course of the disease. The International Classification of Diseases for Oncology, Thi...
nov2018
BEHO3V
17
57
1
394
Behavior code ICD-O-3
Behavior Code ICD-O-3
523
16
BEHAVIOR CODE ICD-O-3
Behavior code ICD-O-3
SEER requires registries to collect malignancies with in situ /2 and malignant /3 behavior codes as described in ICD-O-3. SEER requires registries to collect benign /0 and borderline /1 intracranial and CNS tumors for cases diagnosed on or after 1/1/2004. Behavior is the fifth digit of the morphology code after the sla...
nov2018
GRADE
18
58
1
394
Grade
Grade
440
17
GRADE
Grade
Grading and differentiation codes of 1-4, 9 are defined in ICD-O-2; 1992. Grade information may be incomplete for cases diagnosed before 1977. In the early 1980’s, additional codes specifying T-cell, B-cell, or null cell involvement in lymphomas and leukemias (histologies M9590-9940) were introduced by SEER. Because th...
nov2018
DX_CONF
19
59
1
394
Diagnostic confirmation
Diagnostic Confirmation
490
18
DIAGNOSTIC CONFIRMATION
Diagnostic Confirmation
This data item records the best method used to confirm the presence of the cancer being reported. The data item is not limited to the confirmation at the time of diagnosis; it is the best method of confirmation during the entire course of the disease. Microscopically Confirmed
nov2018
REPT_SRC
20
60
1
394
Type of reporting source
Type of Reporting Source
500
19
TYPE OF REPORTING SOURCE
Type of Reporting Source
The Type of Reporting Source identifies the source documents used to abstract the case. This is not necessarily the original document that identified the case; rather, it is the source that provided the best information.
nov2018
EOD10_SZ
21
61
3
394
EOD 10 - size (1988+)
EOD—Tumor Size
780
1988-2003
20
EOD—TUMOR SIZE
EOD 10—size (1988-2003)
This item is part of the 10-digit EOD (Item #779). It records the largest dimension of the primary tumor in millimeters. Unknown size=999. This field is only coded for cases diagnosed 1988-2003. See SEER Extent of Disease, 1988: Codes and Coding Instructions, Third Edition for site- specific codes and coding rules, at ...
nov2018
EOD10_EX
22
64
2
394
EOD 10 - extension
EOD—Extension
790
1988-2003
21
EOD—EXTENSION
EOD 10—extent (1988-2003)
This item codes the farthest documented extension of tumor away from the primary site, either by contiguous extension or distant metastases. Allowable values = 00-99. This field is only coded for cases diagnosed 1988-2003. See SEER Extent of Disease, 1988: Codes and Coding Instructions, Third Edition for site- specific...
nov2018
EOD10_PE
23
66
2
394
EOD 10 - path extension
EOD—Extension Prost Path
800
1985-2003
22
EOD—EXTENSION PROST PATH
EOD 10—Prostate path ext (1995-2003)
This is an additional field for prostate cancer only to reflect information from radical prostatectomy, effective with 1995 diagnoses. The field is left blank for all other primaries. Allowable values = 00-99. This field is only coded for prostate cases diagnosed 1995-2003. See SEER Extent of Disease, 1988: Codes and C...
nov2018
EOD10_ND
24
68
1
394
EOD 10 - lymph node
EOD—Lymph Node Involv
810
1988-2003
23
EOD—LYMPH NODE INVOLV
EOD 10—nodes (1988-2003)
This item records the highest specific lymph node chain that is involved by the tumor. Allowable values = 0-9. This field is only coded for cases diagnosed 1988-2003. See SEER Extent of Disease, 1988: Codes and Coding Instructions, Third Edition for site- specific codes and coding rules, at http://seer.cancer.gov/manua...
nov2018
EOD10_PN
25
69
2
394
EOD 10 - positive lymph nodes examined
Regional Nodes Positive
820
1988+
24
REGIONAL NODES POSITIVE
Regional nodes positive (1988+)
Records the exact number of regional lymph nodes examined by the pathologist that were found to contain metastases.
nov2018
EOD10_NE
26
71
2
394
EOD 10 - number of lymph nodes examined
Regional Nodes Examined
830
1988+
25
REGIONAL NODES EXAMINED
Regional nodes examined (1988+)
Records the total number of regional lymph nodes that were removed and examined by the pathologist.
nov2018
EOD13
27
73
13
394
EOD--old 13 digit
EOD—Old 13 Digit
840
1975-1982
26
EOD—OLD 13 DIGIT
Expanded EOD(1) – Expanded EOD (13)
Detailed site-specific codes for EOD used by SEER for selected sites of cancer for tumors diagnosed 1975-1982, except death-certificate-only cases. For details, reference the following website: http://seer.cancer.gov/manuals/historic/EOD_1977.pdf
nov2018
EOD2
28
86
2
394
EOD--old 2 digit
EOD—Old 2 Digit
850
1975-1982
27
EOD—OLD 2 DIGIT
2-Digit NS EOD / 2-Digit SS EOD
Site-specific codes for EOD used by SEER for tumors diagnosed from January 1, 1973, to December 31, 1982, for cancer sites that did not have a 13-digit scheme. For details, reference the following website: http://seer.cancer.gov/manuals/historic/EOD_1977.pdf
nov2018
EOD4
29
88
4
394
EOD--old 4 digit
EOD—Old 4 Digit
860
1983-1987
28
EOD—OLD 4 DIGIT
EOD 4 – size (first two digits)
Indicates the type of SEER EOD code applied to the tumor.
nov2018
EOD_CODE
30
92
1
394
Coding system for EOD
Coding System for EOD
870
1975-2003
null
null
null
null
nov2018
TUMOR_1V
31
93
1
394
Tumor marker 1
Tumor Marker 1
1150
1990-2003
29
TUMOR MARKER 1
Tumor Marker 1 (1990-2003)
This data item records prognostic indicators for breast cases (ERA 1990- 2003), prostate cases (PAP 1998-2003) and testis cases (AFP 1998-2003).
nov2018
TUMOR_2V
32
94
1
394
Tumor marker 2
Tumor Marker 2
1160
1990-2003
30
TUMOR MARKER 2
Tumor Marker 2 (1990-2003)
This data item records prognostic indicators for breast cases (PRA 1990- 2003) and testis cases (hCG 1998-2003). For Breast Cancer Cases (PRA 1990-2003)
nov2018
TUMOR_3V
33
95
1
394
Tumor marker 3
Tumor Marker 3
1170
1998-2003
31
TUMOR MARKER 3
Tumor Marker 3 (1998-2003)
This data item records prognostic indicators for testis cases (LDH 1998- 2003)
nov2018
CSTUMSIZ
34
96
3
394
CS Tumor size
CS Tumor Size
2800
2004-2015
32
CS TUMOR SIZE
CS tumor size (2004-2015)
Information on tumor size. Available for 2004-2015 diagnosis years. Earlier cases may be converted and new codes added which weren't available for use prior to the current version of CS. For more information, see http://seer.cancer.gov/seerstat/variables/seer/ajcc-stage.
nov2018
CSEXTEN
35
99
3
394
CS Extension
CS Extension
2810
2004-2015
33
CS EXTENSION
CS extension (2004-2015)
Information on extension of the tumor. Available for 2004-2015 diagnosis years. Earlier cases may be converted and new codes added which weren't available for use prior to the current version of CS. For more information, see http://seer.cancer.gov/seerstat/variables/seer/ajcc-stage. Note: This item was originally a 2-d...
nov2018
CSLYMPHN
36
102
3
394
CS Lymph Nodes
CS Lymph Nodes
2830
2004-2015
34
CS LYMPH NODES
CS lymph nodes (2004-2015)
Information on involvement of lymph nodes. Available for 2004-2015 diagnosis years. Earlier cases may be converted and new codes added which weren't available for use prior to the current version of CS. For more information, see http://seer.cancer.gov/seerstat/variables/seer/ajcc-stage. Note: This item was originally a...
nov2018
CSMETSDX
37
105
2
394
CS Mets at DX
CS Mets at Dx
2850
2004-2015
35
CS METS AT DX
CS mets at dx (2004-2015)
Information on distant metastasis. Available for 2004-2015 diagnosis years. Earlier cases may be converted and new codes added which weren't available for use prior to the current version of CS. For more information, see http://seer.cancer.gov/seerstat/variables/seer/ajcc-stage.
nov2018
CS1SITE
38
107
3
394
CS Site-Specific Factor 1
CS Site-Specific Factor 1
2880
2004+*
36
CS SITE-SPECIFIC FACTOR 1
CS site-specific factor 1 (2004+ varying by schema)
Each CS site-specific factor (SSF) is schema dependent. They can provide information needed to stage the case, clinically relevant information, or prognostic information. Available for varying years and schemas depending on standard setter requirements. Earlier cases may be converted and new codes added which weren't a...
nov2018
CS2SITE
39
110
3
394
CS Site-Specific Factor 2
CS Site-Specific Factor 2
2890
2004+*
37
CS SITE-SPECIFIC FACTOR 2
CS site-specific factor 2 (2004+ varying by schema)
Each CS site-specific factor (SSF) is schema dependent. They can provide information needed to stage the case, clinically relevant information, or prognostic information. Available for varying years and schemas depending on standard setter requirements. Earlier cases may be converted and new codes added which weren't a...
nov2018
CS3SITE
40
113
3
394
CS Site-Specific Factor 3
CS Site-Specific Factor 3
2900
2004+*
38
CS SITE-SPECIFIC FACTOR 3
CS site-specific factor 3 (2004+ varying by schema)
Each CS site-specific factor (SSF) is schema dependent. They can provide information needed to stage the case, clinically relevant information, or prognostic information. Available for varying years and schemas depending on standard setter requirements. Earlier cases may be converted and new codes added which weren't a...
nov2018
CS4SITE
41
116
3
394
CS Site-Specific Factor 4
CS Site-Specific Factor 4
2910
2004+*
39
CS SITE-SPECIFIC FACTOR 4
CS site-specific factor 4 (2004+ varying by schema)
Each CS site-specific factor (SSF) is schema dependent. They can provide information needed to stage the case, clinically relevant information, or prognostic information. Available for varying years and schemas depending on standard setter requirements. Earlier cases may be converted and new codes added which weren't a...
nov2018
CS5SITE
42
119
3
394
CS Site-Specific Factor 5
CS Site-Specific Factor 5
2920
2004+*
40
CS SITE-SPECIFIC FACTOR 5
CS site-specific factor 5 (2004+ varying by schema)
Each CS site-specific factor (SSF) is schema dependent. They can provide information needed to stage the case, clinically relevant information, or prognostic information. Available for varying years and schemas depending on standard setter requirements. Earlier cases may be converted and new codes added which weren't a...
nov2018
CS6SITE
43
122
3
394
CS Site-Specific Factor 6
CS Site-Specific Factor 6
2930
2004+*
41
CS SITE-SPECIFIC FACTOR 6
CS site-specific factor 6 (2004+ varying by schema)
Each CS site-specific factor (SSF) is schema dependent. They can provide information needed to stage the case, clinically relevant information, or prognostic information. Available for varying years and schemas depending on standard setter requirements. Earlier cases may be converted and new codes added which weren't a...
nov2018
CS25SITE
44
125
3
394
CS Site-Specific Factor 25
CS Site-Specific Factor 25
2879
2004+*
42
CS SITE-SPECIFIC FACTOR 25
CS site-specific factor 25 (2004+ varying by schema)
Each CS site-specific factor (SSF) is schema dependent. They can provide information needed to stage the case, clinically relevant information, or prognostic information. Available for varying years and schemas depending on standard setter requirements. Earlier cases may be converted and new codes added which weren't a...
nov2018
DAJCCT
45
128
2
394
Derived AJCC T
Derived AJCC T
2940
2004-2015
null
null
null
null
nov2018
DAJCCN
46
130
2
394
Derived AJCC N
Derived AJCC N
2960
2004-2015
null
null
null
null
nov2018
DAJCCM
47
132
2
394
Derived AJCC M
Derived AJCC M
2980
2004-2015
null
null
null
null
nov2018
DAJCCSTG
48
134
2
394
Derived AJCC Stage Group
Derived AJCC Stage Group
3000
2004-2015
null
null
null
null
nov2018
DSS1977S
49
136
1
394
Derived SS1977
Derived SS1977
3010
2004-2015
47
DERIVED SS1977
Derived SS1977 (2004-2015)
This item is the derived “SEER Summary Stage 1977” from the CS algorithm, effective with 2004-2015 diagnosis years.
nov2018
SCSSM2KO
50
137
1
394
SEER Combined Summary Stage 2000 (2004+)
N/A
2004+
48
SEER COMBINED SUMMARY STAGE 2000 (2004+)
SEER Combined Summary Stage 2000 (2004+)
Combination of NAACCR Items #759 and #3020. For 2004-2015 cases, it is copied from Derived SS2000 and for 2016+, it is the best available between Derived SS2000 and SEER Summary Stage 2000. All benign/borderline cases are set to NA.
nov2018
CSVFIRST
51
141
6
394
CS Version Input Original
CS Version Input Original
2935
2004-2015
49
CS VERSION INPUT ORIGINAL
CS version input (2004-2015)
This item indicates the number of the version used to initially code CS fields. The first two digits represent the major version number; the second two digits represent minor version changes; and, the last two digits represent even less significant changes, such as corrections of typographical errors that do not affect...
nov2018
CSVLATES
52
147
6
394
CS Version Derived
CS Version Derived
2936
2004-2015
50
CS VERSION DERIVED
CS version latest (2004-2015)
This item indicates the number of the version of the CS used most recently to derive the CS output fields. The first two digits represent the major version number; the second two digits represent minor version changes; and, the last two digits represent even less significant changes, such as corrections of typographica...
nov2018
CSVCURRENT
53
153
6
394
CS Version Input Current
CS Version Input Current
2937
2004-2015
51
CS VERSION INPUT CURRENT
CS version input current (2004-2015)
This item indicates the number of the version of the CS after input fields have been updated or recoded. The first two digits represent the major version number; the second two digits represent minor version changes; and, the last two digits represent even less significant changes, such as corrections of typographical ...
nov2018
SURGPRIF
54
159
2
394
RX Summ--surg prim site
RX Summ—Surg Prim Site
1290
1998+
52
RX SUMM—SURG PRIM SITE
RX Summ--Surg Prim Site (1998+)
Surgery of Primary Site describes a surgical procedure that removes and/or destroys tissue of the primary site performed as part of the initial work-up or first course of therapy. The actual coding schemes for individual sites and diagnosis years can be viewed at http://seer.cancer.gov/tools/codingmanuals/historical.ht...
nov2018
SURGSCOF
55
161
1
394
RX Summ--scope reg LN sur 2003+
RX Summ—Scope Reg LN Sur
1292
2003+
53
RX SUMM—SCOPE REG LN SUR
RX Summ--Scope Reg LN Sur (2003+)
Scope of Regional Lymph Node Surgery describes the procedure of removal, biopsy, or aspiration of regional lymph nodes performed during the initial work-up or first course of therapy at all facilities. Changed to Unknown or Not Applicable for breast cancer cases. See http://seer.cancer.gov/seerstat/variables/seer/regio...
nov2018
SURGSITF
56
162
1
394
RX Summ--surg oth reg/dis
RX Summ—Surg Oth Reg/Dis
1294
2003+
54
RX SUMM-SURG OTH REG/DIS
RX Summ--Surg Oth Reg/Dis (2003+)
Surgical procedure of Other Site describes the surgical removal of distant lymph node(s) or other tissue(s) or organ(s) beyond the primary site.
nov2018
NUMNODES
57
163
2
394
Number of lymph nodes
RX Summ—Reg LN Examined
1296
1998-2002
55
RX SUMM-REG LN EXAMINED
RX Summ—Reg LN Examined (1998-2002)
This data item records the number of regional lymph nodes examined in conjunction with surgery performed as part of the first course of treatment at all facilities. This item is only available for cases diagnosed 1998-2002.
nov2018
NO_SURG
58
166
1
394
Reason no cancer-directed surgery
Reason for no surgery
1340
56
REASON FOR NO SURGERY
Reason no cancer-directed surgery
This data item documents the reason that surgery was not performed.
nov2018
SS_SURG
59
170
2
394
Site specific surgery (1983-1997)
RX Summ—Surgery Type
1640
1975-1997
null
null
null
null
nov2018
SURGSCOP
60
174
1
394
Scope of lymph node surgery 98-02
RX Summ—Scope Reg 98-02
1647
1998-2002
58
RX SUMM—SCOPE REG 98-02
Scope of reg lymph nd surg (1998-2002)
This field describes the removal, biopsy or aspiration of regional lymph node(s) at the time of surgery of the primary site or during a separate surgical event at all facilities for cases diagnosed 1998-2002. Changed to Unknown or Not Applicable for breast cancer cases. See http://seer.cancer.gov/seerstat/variables/see...
nov2018
SURGSITE
61
175
1
394
Surgery to other sites
RX Summ—Surg Oth 98-02
1648
1998-2002
59
RX SUMM—SURG OTH 98-02
Surgery of oth reg/dis sites (1998-2002)
This field records the removal of distant lymph nodes or other tissue(s)/organ(s) beyond the primary site given at all facilities as part of the first course of treatment for cases diagnosed 1998-2002. See http://seer.cancer.gov/manuals/historic/AppendC.pdf for a list of valid codes.
nov2018
REC_NO
62
176
2
394
Record number
null
null
null
null
null
null
null
nov2018
TYPE_FU
63
191
1
394
Type of followup expected
SEER Type of Follow-up
2180
61
SEER TYPE OF FOLLOW-UP
Type of follow-up expected
This item codes the type of follow-up expected for a SEER case.
nov2018
AGE_1REC
64
192
2
394
Age recode <1 year olds
Age Recode <1 Year olds
N/A
62
AGE RECODE <1 YEAR OLDS
Age recode with <1 year olds
The age recode variable is based on Age at Diagnosis (single-year ages). The groupings used in the age recode variable are determined by the age groupings in the population data. This recode has 19 age groups in the age recode variable (< 1 year, 1-4 years, 5-9 years, ..., 85+ years).
nov2018
SITERWHO
65
199
5
394
Site recode ICD-O-3/WHO 2008
Site Recode ICD-O-3/WHO 2008
N/A
63
SITE RECODE ICD-O-3/WHO 2008
Site Recode ICD-O-3/WHO 2008
A recode based on Primary Site and ICD-O-3 Histology in order to make analyses of site/histology groups easier. For example, the lymphomas are excluded from stomach and Kaposi and mesothelioma are separate categories based on histology. For more information, see http://seer.cancer.gov/siterecode/icdo3_dwhoheme/index.ht...
nov2018
ICDOTO9V
66
204
4
394
Recode ICD-O-2 to 9
Recode ICD-O-2 to 9
N/A
64
RECODE ICD-O-2 TO 9
Recode ICD-O-2 to 9
The primary site and morphology are recoded to ICD-9 codes using the Conversion of Malignant Neoplasms by Topography and Morphology from the International Classification of Disease for Oncology, Second Edition (ICD-0-2) to International Classification of Diseases, 9th Revision (ICD-9) and the International Classificati...
nov2018
ICDOT10V
67
208
4
394
Recode ICD-O-2 to 10
Recode ICD-O-2 to 10
N/A
65
RECODE ICD-O-2 TO 10
Recode ICD-O-2 to 10
The primary site and morphology are converted to ICD-10 codes using the Conversion of Malignant Neoplasms by Topography and Morphology from the International Classification of Disease for Oncology, Second Edition (ICD-O-2) to International Classification of Diseases and Related Health Problems, 10th Revision, 1998 edit...
nov2018
ICCC3WHO
68
218
3
394
ICCC site recode ICD-O-3/WHO 2008
ICCC site recode ICD-O-3/WHO 2008
N/A
66
ICCC SITE RECODE ICD-O-3/WHO 2008
ICCC site recode ICD-O-3/WHO 2008
A site/histology recode that is mainly used to analyze data on children. The recode was applied to all cases no matter the age in order that age comparisons can be made with these groupings. Based on ICD-O-3. Note that cases diagnosed before 2001 were not coded under ICD-O-3 and were converted to ICD-O-3 and may not ha...
nov2018
ICCC3XWHO
69
221
3
394
ICCC site rec extended ICD-O-3/ WHO 2008
N/A
67
ICCC SITE REC EXTENDED ICD-O-3/WHO 2008
ICCC site rec extended ICD-O-3/WHO 2008
Based on ICD-O-3. For more information on this International Classification of Childhood Cancer (ICCC) site/histology recode, see http://seer.cancer.gov/iccc. While the recode is normally used for childhood cancers, it is on the file for all ages so that childhood cancers could be compared across age groups.
nov2018
BEHTREND
70
224
1
394
Behavior recode for analysis
Behavior Recode for Analysis
N/A
68
BEHAVIOR RECODE FOR ANALYSIS
Behavior recode for analysis
This recode was created so that data analyses could eliminate major groups of histologies/behaviors that weren't collected consistently over time, for example benign brain, myelodysplastic syndromes, and borderline tumors of the ovary. Created from ICD-O-3 behavior and histology. For more information, see http://seer.c...
nov2018
HISTREC
71
226
2
394
Broad Histology recode
Histology Recode—Broad Groupings
N/A
69
HISTOLOGY RECODE—BROAD GROUPINGS
Histology recode - broad groupings
Based on Histologic Type ICD-O-3.
nov2018
HISTRECB
72
228
2
394
Brain recode
Histology Recode—Brain Groupings
N/A
70
HISTOLOGY RECODE—BRAIN GROUPINGS
Histology recode - Brain groupings
Based on Histologic Type ICD-O-3.
nov2018
CS0204SCHEMA
73
230
3
394
CS Schema v0204
null
null
null
null
null
null
null
nov2018
RAC_RECA
74
233
1
394
Race recode A
Race recode (White, Black, Other)
N/A
72
RACE RECODE (WHITE, BLACK, OTHER)
Race recode (White, Black, Other)
Race recode is based on the race variables and the American Indian/Native American IHS link variable. This recode should be used to link to the populations for white, black and other. It is independent of Hispanic ethnicity. For more information, see http://seer.cancer.gov/seerstat/variables/seer/race_ethnicity.
nov2018
RAC_RECY
75
234
1
394
Race recode Y
Race recode (W, B, AI, API)
N/A
73
RACE RECODE (W, B, AI, API)
Race recode (W, B, AI, API)
Caution should be exercised when using this variable. For more information, see http://seer.cancer.gov/seerstat/variables/seer/race_ethnicity.
nov2018
ORIGRECB
76
235
1
394
Origin Recode NHIA
N/A
null
null
null
null
nov2018
HST_STGA
77
236
1
394
SEER historic stage A
SEER historic stage A
N/A
1975-2015
75
SEER HISTORIC STAGE A
SEER historic stage A (1973-2015)
Derived from Collaborative Stage (CS) for 2004-2015 and Extent of Disease (EOD) from 1973-2003. It is a simplified version of stage: in situ, localized, regional, distant, & unknown. Over time several different EOD schemes have been used. Thus caution should be used when doing trend analysis. For more information inclu...
nov2018
AJCC_STG
78
237
2
394
AJCC stage 3rd edition (1988+)
AJCC stage 3 edition (1988-2003)
N/A
null
null
null
null
nov2018
AJ_3SEER
79
239
2
394
SEER modified AJCC stage 3rd ed (1988+)
N/A
null
null
null
null
nov2018
SSS77VZ
80
241
1
394
SEER Summary Stage 1977 (1995-2000)
N/A
1995-2000
78
SEER SUMMARY STAGE 1977
SEER summary stage 1977 (1995-2000)
Based on SEER Extent of Disease (EOD) following a SEER algorithm. This variable is provided on the NAACCR call for data. For more information, see http://seer.cancer.gov/manuals/historic/ssm_1977.pdf.
nov2018
SSSM2KPZ
81
242
1
394
SEER Summary Stage 2000 2000 (2001-2003)
N/A
2001-2003
null
null
null
null
nov2018
FIRSTPRM
82
245
1
394
First malignant primary indicator
First malignant primary indicator
N/A
80
FIRST MALIGNANT PRIMARY INDICATOR
First malignant primary indicator
Based on all the tumors in SEER. Tumors not reported to SEER are assumed malignant.
nov2018
ST_CNTY
83
246
5
394
State-county recode
State-county recode
N/A
81
STATE-COUNTY RECODE
State-county
This item is a state-county combination where the first two characters represent the state FIPS code. The last three digits represent the FIPS county code.
nov2018
CODPUB
84
255
5
394
Cause of death to SEER site recode
Cause of Death to SEER site recode
N/A
82
CAUSE OF DEATH TO SEER SITE RECODE
COD to site recode
This recode was introduced to account for several newly valid ICD-10 codes and includes both cancer and non-cancer causes of death. See the following website for details: http://seer.cancer.gov/codrecode/1969+_d09172004/index.html
nov2018
CODPUBKM
85
260
5
394
COD to site rec KM
COD to site rec KM
N/A
83
COD TO SITE REC KM
COD to site rec KM
This is a recode based on underlying cause of death to designate cause of death into groups similar to the incidence site recode with KS and mesothelioma. For more information, see http://seer.cancer.gov/codrecode/1969+_d04162012. Study cutoff date has been applied, i.e. coded as alive if death occurred after study cut...
nov2018
STAT_REC
86
265
1
394
Vital status recode (study cutoff used)
Vital Status recode
N/A
84
VITAL STATUS RECODE
Vital status recode (study cutoff used)
Any patient that dies after the follow-up cut-off date is recoded to alive as of the cut-off date.
nov2018
IHSLINK
87
266
1
394
IHS link
IHS Link
192
85
IHS LINK
IHS Link
Incidence files are periodically linked with Indian Health Service (IHS) files to identify Native Americans. The race recode uses information from this field and race to determine if a person is Native American or not. See http://www.seer.cancer.gov/seerstat/variables/seer/race_ethnicity.
nov2018
SUMM2K
88
267
1
394
Historic SSG 2000 Stage
Summary stage 2000 (1998+)
N/A
1998+
86
SUMMARY STAGE 2000 (1998+)
Summary Stage 2000 (1998+)
Summary Stage 2000 is derived from Collaborative Stage (CS) for 2004+ and Extent of Disease (EOD) from 1998-2003. It is a simplified version of stage: in situ, localized, regional, distant, & unknown. Used in the SEER CSR and more recent SEER publications. For more information including sites and years for which it isn...
nov2018
AYASITERWHO
89
268
2
394
AYA site recode/WHO 2008
AYA site recode/WHO 2008
N/A
87
AYA SITE RECODE/WHO 2008
AYA site recode/WHO 2008
A site/histology recode that is mainly used to analyze data on adolescent and young adults. The recode was applied to all cases no matter the age in order that age comparisons can be made with these groupings. For more information, see http://www.seer.cancer.gov/ayarecode/index.html.
nov2018
LYMSUBRWHO
90
270
2
394
Lymphoma subtype recode/WHO 2008
Lymphoma subtype recode/WHO 2008
N/A
88
LYMPHOMA SUBTYPE RECODE/WHO 2008
Lymphoma subtype recode/WHO 2008
A site/histology recode that is mainly used to analyze data on adolescent and young adults. The recode was applied to all cases no matter the age in order that age comparisons can be made with these groupings. For more information, see http://seer.cancer.gov/lymphomarecode/.
nov2018
VSRTSADX
91
272
1
394
SEER cause of death classification
N/A
null
null
null
null
nov2018
ODTHCLASS
92
273
1
394
SEER other cause of death classification
N/A
90
SEER OTHER CAUSE OF DEATH CLASSIFICATION
SEER other cause of death classification
Created for use in left-truncated life table session. This variable designates that the person died of causes other than their cancer. For more information, see http://seer.cancer.gov/causespecific.
nov2018
CSTSEVAL
93
274
1
394
CS EXT/Size Eval
CS Tumor Size/Ext Eval
2820
2004-2015
91
CS TUMOR SIZE EXT/EVAL
CS Tumor Size/Ext Eval (2004-2015)
Available for 2004-2015, but not required for the entire timeframe. Will be blank in cases not collected. For more information, see http://seer.cancer.gov/seerstat/variables/seer/ajcc-stage.
nov2018
CSRGEVAL
94
275
1
394
CS Nodes Eval
CS Lymph Nodes Eval
2840
2004-2015
92
CS LYMPH NODES EVAL
CS Reg Nodes Eval (2004-2015)
Available for 2004-2015, but not required for the entire timeframe. Will be blank in cases not collected. For more information, see http://seer.cancer.gov/seerstat/variables/seer/ajcc-stage.
nov2018
CSMTEVAL
95
276
1
394
CS Mets Eval
CS Mets Eval
2860
2004-2015
93
CS METS EVAL
CS Mets Eval (2004-2015)
Available for 2004-2015,, but not required for the entire timeframe. Will be blank in cases not collected. For more information, see http://seer.cancer.gov/seerstat/variables/seer/ajcc-stage.
nov2018
INTPRIM
96
277
1
394
Primary by International Rules
Primary by international rules
N/A
94
PRIMARY BY INTERNATIONAL RULES
Primary by International Rules
Created using IARC multiple primary rules. Did not include benign tumors or non-bladder in situ tumors in algorithm. No tumor information was modified on any records.
nov2018
ERSTATUS
97
278
1
394
ER Status Recode Breast Cancer (1990+)
N/A
1990+
95
ER STATUS RECODE BREAST CANCER (1990+)
ER Status Recode Breast Cancer (1990+)
Created by combining information from Tumor marker 1 (1990-2003) (NAACCR Item #=1150), with information from CS site-specific factor 1 (2004+) (NAACCR Item #=2880). This field is blank for non-breast cases and cases diagnosed before 1990.
nov2018
PRSTATUS
98
279
1
394
PR Status Recode Breast Cancer (1990+)
N/A
1990+
96
PR STATUS RECODE BREAST CANCER (1990+)
PR Status Recode Breast Cancer (1990+)
Created by combining information from Tumor marker 2 (1990-2003) (NAACCR Item #=1150), with information from CS site-specific factor 2 (2004+) (NAACCR Item #=2880). This field is blank for non-breast cases and cases diagnosed before 1990.
nov2018
CSSCHEMA
99
280
2
394
CS Schema - AJCC 6th Edition
N/A
null
null
null
null
nov2018
CS8SITE
100
282
3
394
Cs Site-specific Factor 8
CS Site-Specific Factor 8
2862
2004+*
98
CS SITE-SPECIFIC FACTOR 8
CS site-specific factor 8 (2004+ varying by schema)
Each CS site-specific factor (SSF) is schema dependent. They can provide information needed to stage the case, clinically relevant information, or prognostic information. Available for varying years and schemas depending on standard setter requirements. Earlier cases may be converted and new codes added which weren't a...
End of preview. Expand in Data Studio

SEER (NCI): U.S. county populations 1969–2024, standard populations, code tables and case-data dictionaries in Parquet

No patient-level data. This repository contains no individual patients, tumors or cases. It holds U.S. county population aggregates (population counts by county, year, sex, race, Hispanic origin and age), reference tables, and documentation of SEER's restricted case files. SEER case data themselves are restricted under a data use agreement and are not included.

This dataset holds the openly published parts of the National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) Program, converted to Parquet:

  • the U.S. county population estimates SEER uses as rate denominators (July 1, 1969–2024, released February 2026). These are county aggregates: one population count per county, year, sex, race, Hispanic origin and age (20 age groups or single ages). One Parquet file is written per year and original SEER file;
  • the standard populations used for age adjustment (2000 U.S. standard, World WHO and Segi, European, Canadian, and others);
  • the site and cause-of-death recode tables (Site recode ICD-O-3/WHO 2008, Site recode ICD-O-3 2023 revision, COD to site recode);
  • the documentation of the SEER case data: the SEER*Stat variable dictionary, Nov 2025, and variable and value dictionaries parsed from the SEER Research Plus data description, Nov 2025. It also includes the ASCII record layout of the last text release, Nov 2018;
  • a pipeline (pipeline/10_case_convert.R) for SEER data-use-agreement holders. It converts their own SEER*Stat case listing exports or legacy ASCII files into the same Parquet layout, on their own computer.

SEER case-level data (SEER Research / Research Plus) are not included and must never be added. They need a signed SEER Research Data Use Agreement, and that agreement forbids passing them on (see Access and redistribution below).

Built for the Bioanalysis R package. Works with any tool that reads Parquet (DuckDB, Arrow, Polars, pandas, R).

File counts, rows and years by category are in Data by category below.

Layout

data/<year>/Population/<original SEER file>.parquet  one row per county x race x origin x sex x age (aggregate counts)
data/<year>/Dictionary/VARIABLES.parquet              one row per file x variable
data/<year>/Dictionary/VALUES.parquet                 one row per file x variable x code (+ computed records, population)
data/reference/StandardPopulations/*.parquet          stdpop.19ages, .18ages, .singleagesthru84/99, web tables
data/reference/CodeTables/*.parquet                   site recodes, COD to site recode
data/reference/SEERStat/seerstat-variable-dictionary-nov2025.parquet
data/reference/CaseDictionary/nov2025/{VARIABLES,VALUES}.parquet   case-file documentation (no cases)
data/reference/CaseDictionary/nov2018/{VARIABLES,VALUES}.parquet   ASCII record layout + codes (no cases)
data/reference/Dictionary/{VARIABLES,VALUES}.parquet  columns of every reference table; standard population codes
metadata/FILES.parquet                                every file: rows, columns, grain, join keys, source URL, MD5
metadata/YEARS.parquet                                every year: U.S. total, counties, estimate source, SEER modifications
pipeline/                                             the R scripts that rebuilt everything from the sources

The four population files and their naming are described in Data by category below.

Columns of every population file (layout from the SEER population data dictionary):

Column Type Content
Year int July 1 estimate year
State string postal abbreviation; KR = Katrina/Rita evacuee dummy state (2005)
StateFIPS string 2-digit, zero-filled (99 for KR)
CountyFIPS string 3-digit, zero-filled (999 for KR)
Race int 1969 files: 1 White, 2 Black, 3 Other. 1990 files: 1 White, 2 Black, 3 AI/AN, 4 API
Origin int 0 Non-Hispanic, 1 Hispanic (1990 files); 9 not applicable (1969 files)
Sex int 1 Male, 2 Female
Age int 20ages files: 20 groups, 00 = <1, 01 = 1–4, 02 = 5–9 … 18 = 85–89, 19 = 90+; single-age files: 0–89, 90 = 90+
Population int estimated population

FIPS codes stay as text so leading zeros are kept. Codes in the dictionary keep their published text (code, e.g. "05") and also come as an integer (code_value). Columns 12–13 of the original records are not in the published layout. They were checked to be blank in every record (logs/validation.csv) and are not carried over. SEER's popdic.html still calls the 20-group coding "19 Age group data"; the files hold codes 00–19, i.e. 20 groups.

Data by category

This repository contains no patient-level or tumor-level data. Every file is an aggregate (county population counts), a reference table or documentation. SEER case data are restricted and are not here (see Access and redistribution).

Files are grouped as SEER publishes them: the U.S. county population files (seer.cancer.gov/popdata), the standard populations (seer.cancer.gov/stdpopulations), the code tables (site and cause-of-death recodes) and the case-file documentation (SEER*Stat database documentation).

Category Folder Contents File types Files Rows Years
Population files data/<year>/Population/ U.S. population estimates (July 1) as county aggregates: one row per county × race × Hispanic origin × sex × age group or single age, with the population count. These are SEER's rate denominators; there are no persons or cases in them. 4 182 187,987,202 1969–2024
Standard populations data/reference/StandardPopulations/ Standard populations (age distributions) used as weights for age-adjusted rates. 7 7 1,151 reference (not by year)
Code tables data/reference/CodeTables/ Definitions of SEER's site recodes and cause-of-death recode (ICD-O-3 / ICD codes → recode). 5 5 271 reference (not by year)
Case-file documentation data/reference/SEERStat/, data/reference/CaseDictionary/ Documentation of the restricted SEER case files, containing no cases: the SEER*Stat variable dictionary, the case-data variables and value codes, and the ASCII record layout. 5 5 3,082 reference (not by year)

File naming. Population files keep SEER's names, <area>.<years>.<ages>.adjusted: us = all states combined (SEER's per-state files are subsets and are not repeated); 1969_2024 = White, Black, Other for 1969 onward, 1990_2024 = four races by Hispanic origin for 1990 onward; 20ages = 20 age groups (codes 00 = <1, 01 = 1–4, 02 = 5–9 … 18 = 85–89, 19 = 90+), singleages.through89.90plus = single ages 0–89 and 90 = 90+; adjusted = 2005 Katrina/Rita adjustment. SEER distributes each as one .txt.gz covering all years; here it is split into one Parquet file per year with the same name, data/<year>/Population/<file>.parquet. Reference tables keep the SEER file or page name (stdpop.19ages, siterecode_icdo3_dwhoheme, …).

Population files: 182 files
File type Official title Years Files Rows Grain Join keys
us.1969_2024.20ages.adjusted County-Level Population Files - 20 Age Groups; 1969-2024 White, Black, Other 1969–2024 data/<year>/Population/us.1969_2024.20ages.adjusted.parquet (56 files) 17,485,723 one row per county x race x origin x sex x age (0 zero-population rows; absent cells have no population) Year + StateFIPS + CountyFIPS + Race + Origin + Sex + Age
us.1990_2024.20ages.adjusted County-Level Population Files - 20 Age Groups; 1990-2024 4 Expanded Races by Origin 1990–2024 data/<year>/Population/us.1990_2024.20ages.adjusted.parquet (35 files) 22,178,259 one row per county x race x origin x sex x age (0 zero-population rows; absent cells have no population) Year + StateFIPS + CountyFIPS + Race + Origin + Sex + Age
us.1969_2024.singleages.through89.90plus.adjusted County-Level Population Files - Single-year Age Groups; 1969-2024 White, Black, Other 1969–2024 data/<year>/Population/us.1969_2024.singleages.through89.90plus.adjusted.parquet (56 files) 69,655,933 one row per county x race x origin x sex x age (0 zero-population rows; absent cells have no population) Year + StateFIPS + CountyFIPS + Race + Origin + Sex + Age
us.1990_2024.singleages.through89.90plus.adjusted County-Level Population Files - Single-year Age Groups; 1990-2024 4 Expanded Races by Origin 1990–2024 data/<year>/Population/us.1990_2024.singleages.through89.90plus.adjusted.parquet (35 files) 78,667,287 one row per county x race x origin x sex x age (0 zero-population rows; absent cells have no population) Year + StateFIPS + CountyFIPS + Race + Origin + Sex + Age
Standard populations: 7 files
File Official title Release Path Rows Grain Join keys
singleages.pops.tables Standard Populations - Single Ages (CSV, 2000 U.S., 2011/2021 Canadian, WHO 2000-2025) current release data/reference/StandardPopulations/singleages.pops.tables.parquet 95 reference table Age
stdpop.18ages Standard Populations - 18 Age Groups (0-4, 5-9, ..., 80-84, 85+), file stdpop.18ages.txt current release data/reference/StandardPopulations/stdpop.18ages.parquet 270 reference table Standard + Age
stdpop.19ages Standard Populations - 19 Age Groups (0, 1-4, 5-9, ..., 80-84, 85+), file stdpop.19ages.txt current release data/reference/StandardPopulations/stdpop.19ages.parquet 285 reference table Standard + Age
stdpop.19ages.webtable Standard Populations - 19 Age Groups (0, 1-4, 5-9, ..., 85+), web table with all standards incl. 2011 and 2021 Canadian current release data/reference/StandardPopulations/stdpop.19ages.webtable.parquet 20 reference table Age
stdpop.20ages.webtable Standard Populations - 20 Age Groups (0, 1-4, 5-9, ..., 85-89, 90+), web table; SEER*Stat default since the Nov 2024 submission current release data/reference/StandardPopulations/stdpop.20ages.webtable.parquet 21 reference table Age
stdpop.singleagesthru84 Standard Populations - Single Ages to 84 (85+), file stdpop.singleagesthru84.txt current release data/reference/StandardPopulations/stdpop.singleagesthru84.parquet 258 reference table Standard + Age
stdpop.singleagesthru99 Standard Populations - Single Ages to 99 (100+), file stdpop.singleagesthru99.txt current release data/reference/StandardPopulations/stdpop.singleagesthru99.parquet 202 reference table Standard + Age
Code tables: 5 files
File Official title Release Path Rows Grain Join keys
codrecode_1969_d03012018_neoplasm SEER cause of death (COD) to site recode, neoplasm causes current release data/reference/CodeTables/codrecode_1969_d03012018_neoplasm.parquet 70 reference table Recode
codrecode_1969_d03012018_nonneoplasm SEER cause of death (COD) to site recode, non-neoplasm causes current release data/reference/CodeTables/codrecode_1969_d03012018_nonneoplasm.parquet 27 reference table Recode
codrecode_1969_d03012018_other SEER cause of death (COD) to site recode, alive / unknown codes current release data/reference/CodeTables/codrecode_1969_d03012018_other.parquet 3 reference table Recode
siterecode_icdo3_2023 Site recode ICD-O-3 2023 revision definition current release data/reference/CodeTables/siterecode_icdo3_2023.parquet 83 reference table Recode
siterecode_icdo3_dwhoheme Site recode ICD-O-3/WHO 2008 definition current release data/reference/CodeTables/siterecode_icdo3_dwhoheme.parquet 88 reference table Recode
Case-file documentation: 5 files
File Official title Release Path Rows Grain Join keys
nov2018/VALUES SEER Research Data value codes, Nov 2018 submission (parsed from PDF) Nov 2018 submission data/reference/CaseDictionary/nov2018/VALUES.parquet 1,062 one row per case-data variable x code (documentation only) variable (entry)
nov2018/VARIABLES SEER Research Data ASCII record layout and variables, Nov 2018 submission (1975-2016) Nov 2018 submission data/reference/CaseDictionary/nov2018/VARIABLES.parquet 143 one row per case-data variable (documentation only; no case data) variable (entry)
nov2025/VALUES SEER Research Plus case-data value codes, Nov 2025 submission (parsed from PDF) Nov 2025 submission data/reference/CaseDictionary/nov2025/VALUES.parquet 1,416 one row per case-data variable x code (documentation only) entry (title, seerstat_name)
nov2025/VARIABLES SEER Research Plus case-data variables, Nov 2025 submission (data description + dictionary) Nov 2025 submission data/reference/CaseDictionary/nov2025/VARIABLES.parquet 198 one row per case-data variable (documentation only; no case data) entry (title, seerstat_name)
seerstat-variable-dictionary-nov2025 Dictionary of SEER*Stat variables, Research Plus data, November 2025 submission current release data/reference/SEERStat/seerstat-variable-dictionary-nov2025.parquet 263 reference table Name

Hugging Face subsets and splits

On Hugging Face a subset (also called a config) is one named table: it is the drop-down list above the Dataset Viewer and the second argument of load_dataset("jamesjiadazhan/SEER", "<subset>"). A split is a part of a subset, chosen with split=. Here most subsets have a single split, all, which stacks the yearly files: use the Year column (population subsets) or the year column (dictionary subsets) to pick years. standard_populations has one split per standard-population file (ages19, ages18, single_ages_to_84, single_ages_to_99); seerstat_variable_dictionary and case_dictionary_values have one split named after the SEER submission (nov2025).

Subset What it holds Splits
files Catalogue of every file: year, category, title, rows, columns, grain, join keys, path, source URL, MD5. all (1 file)
years One row per year: U.S. total, counties, estimate source, SEER modifications, denominator guidance. all (1 file)
dictionary_variables Population dictionary, every year stacked: one row per year × file × variable (year column). all (56 files)
dictionary_values Population value codes, every year stacked: one row per year × file × variable × code (year column). all (56 files)
population_1969_20ages County aggregates, file us.1969_2024.20ages.adjusted, every year stacked (Year column). all (56 files)
population_1969_singleages County aggregates, file us.1969_2024.singleages.through89.90plus.adjusted, every year stacked (Year column). all (56 files)
population_1990_20ages County aggregates, file us.1990_2024.20ages.adjusted, every year stacked (Year column). all (35 files)
population_1990_singleages County aggregates, file us.1990_2024.singleages.through89.90plus.adjusted, every year stacked (Year column). all (35 files)
standard_populations Standard populations from the .txt files: Standard, Age, Population. ages19, ages18, single_ages_to_84, single_ages_to_99
standard_populations_20ages_webtable 20-age-group standard populations (web table; one column per standard). all (1 file)
siterecode_icdo3_dwhoheme Site recode ICD-O-3/WHO 2008 definition. all (1 file)
siterecode_icdo3_2023 Site recode ICD-O-3 2023 revision definition. all (1 file)
codrecode_1969_d03012018_neoplasm COD to site recode, neoplasm causes of death. all (1 file)
seerstat_variable_dictionary Documentation of the restricted case files, contains no cases: SEER*Stat variable dictionary (Nov 2025). nov2025 (1 file)
case_dictionary_variables_nov2025 Documentation of the restricted case files, contains no cases: case-data variables (Nov 2025). all (1 file)
case_dictionary_values Documentation of the restricted case files, contains no cases: case-data value codes (Nov 2025). nov2025 (1 file)
case_dictionary_layout_nov2018 Documentation of the restricted case files, contains no cases: ASCII record layout (Nov 2018). all (1 file)
reference_dictionary_variables Columns of every reference table, with labels and sources. all (1 file)

Files that are not subsets (6): data/reference/CaseDictionary/nov2018/VALUES.parquet, data/reference/CodeTables/codrecode_1969_d03012018_nonneoplasm.parquet, data/reference/CodeTables/codrecode_1969_d03012018_other.parquet, data/reference/StandardPopulations/singleages.pops.tables.parquet, data/reference/StandardPopulations/stdpop.19ages.webtable.parquet, data/reference/Dictionary/VALUES.parquet. Each is a small one-off table whose columns differ from every other file (or, for the Nov 2018 case-data values, from the Nov 2025 values subset), so it cannot share a subset; read it by path, for example hf://datasets/jamesjiadazhan/SEER/data/reference/CaseDictionary/nov2018/VALUES.parquet with DuckDB or read_parquet(). The Files and versions tab shows every file.

The population subsets are large (187,987,202 rows in all); for one year or a few counties, read the yearly file by path (see Quick start) instead of load_dataset.

Data dictionary

File Subset One row per Main columns
data/<year>/Dictionary/VARIABLES.parquet dictionary_variables year × population file × variable file, file_label (official title), variable, label, start_column, length (position in SEER's fixed-width record), type, notes, universe, n_nonmissing, doc_url
data/<year>/Dictionary/VALUES.parquet dictionary_values year × file × variable × code code (published text, e.g. 05), code_value (integer as stored), description, context (which files the code applies to), published (FALSE = filled from the pattern the page elides), records, population (computed from that year's file)
data/reference/Dictionary/VARIABLES.parquet reference_dictionary_variables reference table × column component, file, file_label, variable, type, label, n_nonmissing, source_url
data/reference/Dictionary/VALUES.parquet (by path) standard-population file × code variable (Standard, Age), code, description, records
data/reference/CaseDictionary/nov2025/VARIABLES.parquet, VALUES.parquet case_dictionary_variables_nov2025, case_dictionary_values case-data variable; variable × code title, seerstat_name, sas_variable_name, naaccr_item, field_description; code, description, table (documentation only, no cases)
data/reference/CaseDictionary/nov2018/VARIABLES.parquet, VALUES.parquet case_dictionary_layout_nov2018; values by path ASCII field; field × code variable, start, length, record_length, naaccr_item, applicable_years, field_description (documentation only, no cases)

Example: in 2020, Race of us.1990_2024.20ages.adjusted has codes 1 = White, 2 = Black, 3 = American Indian/Alaska Native, 4 = Asian or Pacific Islander (file = 'us.1990_2024.20ages.adjusted' AND variable = 'Race' in data/2020/Dictionary/VALUES.parquet). The official population codebook is popdic.html (doc_url); the case-data sources are the SEER*Stat database documentation; every file's source URL is in the files subset.

Joining

  • Population files: one row per StateFIPS + CountyFIPS + Race + Origin + Sex + Age within a year (checked for every file). Join across years on the same columns plus Year. Cells with no population are absent, not zero.
  • To SEER case data (your own export, see below): StateFIPS || CountyFIPS = the case State-county code. Then match year of diagnosis to Year, sex to Sex, and the age recode to Age (20-group or single-age file). For race, use Race recode (White, Black, Other) with the 1969 files, or Race recode (W, B, AI, API) + Origin recode NHIA with the 1990 files.
  • Standard populations: Standard + Age (codes in data/reference/Dictionary/VALUES.parquet; e.g. 203 = 2000 U.S. Std Population, 19 age groups, Census P25-1130). The 20-age-group standards (85–89 and 90+, the SEER*Stat default since the Nov 2024 submission) and the 2021 Canadian standard exist only as web tables: stdpop.20ages.webtable, stdpop.19ages.webtable.
  • Code tables join to case data on Recode (e.g. Site recode ICD-O-3/WHO 2008 = 20010 Lip).

Quick start

import duckdb
con = duckdb.connect()
base = "hf://datasets/jamesjiadazhan/SEER/data"
# 2020 U.S. population by sex, 20 age groups and Hispanic origin
con.sql(f"""
    SELECT Sex, Age, Origin, sum(Population) AS pop
    FROM read_parquet('{base}/2020/Population/us.1990_2024.20ages.adjusted.parquet')
    GROUP BY ALL ORDER BY ALL
""").df()
# Age-adjustment weights: 2000 U.S. standard population, 19 age groups
con.sql(f"""
    SELECT Age, Population / sum(Population) OVER () AS weight
    FROM read_parquet('{base}/reference/StandardPopulations/stdpop.19ages.parquet')
    WHERE Standard = 203 ORDER BY Age
""").df()

R (arrow):

library(arrow)
base <- "https://huggingface.co/datasets/jamesjiadazhan/SEER/resolve/main/data"
pop <- read_parquet(file.path(base, "2020/Population/us.1990_2024.20ages.adjusted.parquet"))

Survey design: none. SEER is a registry, so use population denominators

SEER is a population-based cancer registry. It aims to record every cancer diagnosed among residents of its areas. It is not a sample, so there are no survey weights, strata or PSUs. Do not apply survey-design methods to SEER case data. Rates use the populations in this dataset as denominators:

  • Crude rate = cases / population × 100,000, for the same area, years, sex, race/origin and ages.
  • Registry denominators: sum Population over the counties of the SEER registries in your case file (the case State-county field). Note two exceptions:
    • The Alaska Native Tumor Registry uses the AI/AN population only.
    • In 2005, Louisiana and the Katrina/Rita dummy state KR need care. SEER's own SEER*Stat populations are built from these same county files.
  • Age-adjusted rate = Σ over age groups of (age-specific rate × standard weight). The standard weight is the standard population share; SEER default: 2000 U.S. standard (Standard = 203 with 19 groups; 20 groups in stdpop.20ages.webtable).
  • Pooling years: add cases and add populations over the years (person-years). Do not average rates.
  • metadata/YEARS.parquet gives, per year, the U.S. total, the estimate source and the SEER modifications:
    • 1969–1989: Census intercensal estimates.
    • 1990–2009: Census/NCHS bridged-race intercensal estimates.
    • 2010–2019: Woods & Poole intercensal estimates for NCI.
    • 2020–2024: Census Vintage 2024 bridged-race estimates.
    • Hawaii adjustment (all years).
    • 2005 Katrina/Rita time-weighted adjustment for 62 counties, plus the KR dummy state.
  • Estimates for different SEER releases differ (each release revises past years). Use the population release that matches your case data release, or note the difference.

SEER case data: pipeline for data-use-agreement holders (not included here)

pipeline/10_case_convert.R turns your own SEER data into this layout, on your own computer, in case_local/. It refuses to write into the upload folder.

  • SEER*Stat case listing export (current releases): pass the data file and the .dic that SEER*Stat writes (File > Export > Results as Text File; the .sas program is optional): Rscript 10_case_convert.R --seerstat mylisting.txt [--dic mylisting.dic] [--sas mylisting.sas]
  • Legacy ASCII files (through the Nov 2018 submission; e.g. yr1975_2016.seer9/BREAST.TXT) with SEER's SAS read program: Rscript 10_case_convert.R --ascii yr1975_2016.seer9 --layout read.seer.research.nov2018.sas

Output: data/<submission>/<registry group>/<year of diagnosis>/<FILE>.parquet, plus data/<submission>/Dictionary/{VARIABLES,VALUES}.parquet, metadata/FILES.parquet and logs/validation.csv.

  • Dictionary: built from the export's .dic (variable names, [Format=…] value labels) and the published data description (NAACCR item, field description).
  • Values: unchanged. Fields are read as text. A column becomes an integer only if every value is a canonical integer, so codes such as 06037 or survival months 0005 stay text. The export's declared missing character becomes null.
  • Key: Patient ID + Record number recode (ASCII: REG + PUBCSNUM + REC_NO). The pipeline checks that the key is unique.
  • Patient ID across releases: Patient IDs are dummy numbers and differ between releases. SEER: "The same number is not used across all submissions for each patient." The SEER DUA also forbids linking releases.
  • Record counts: --expect-rows checks the total against the published record counts (SEER*Stat database documentation; e.g. SEER 8, 1975–2023: 5,475,910 tumors, all cases).
  • Tests: the converter is tested on synthetic files that follow the documented layouts, with a cell-by-cell round trip (11_make_synthetic.R, 12_test_case_convert.R).

Access and redistribution

  • SEER Research / Research Plus case data: restricted, not redistributable. Access requires a SEER*Stat account and a signed SEER Research Data Use Agreement: "No Release to Others. Authorized User will not release, nor permit others to release, the Data to any other person." The agreement also bars linking data at the individual level, and bars sending case data to AI tools that transmit data outside your secure environment. Suppress counts of 1–4 in publications. Request access at seer.cancer.gov/data/access.html.
  • The files here are open:
    • The county population files download from seer.cancer.gov/popdata without login, registration or agreement. This differs from the census-tract populations, which need a data use agreement and are not included.
    • The standard populations, recode tables and documentation are public NCI web pages and files.
    • NCI's reuse policy (cancer.gov/policies/copyright-reuse): "Unless otherwise indicated, all text within National Cancer Institute (NCI) products is free of copyright and may be reused without our permission. Credit the National Cancer Institute as the source."
    • The 2010–2019 county estimates were produced for NCI under contract by Woods & Poole Economics, Inc., and NCI distributes them openly as part of the SEER population data.

Sources and processing

Part Source Processing
County populations SEER U.S. population data (.txt.gz, Feb 2026 release) 26-byte fixed-width records; split by Year; integers as above
Population dictionary popdic.html Codes the page elides ("…") were filled from the pattern (published = FALSE); records / population per code computed from the data
Standard populations stdpopulations: stdpop.19ages.txt, .18ages.txt, .singleagesthru84.txt, .singleagesthru99.txt, web tables, singleages.pops.tables.csv 14-byte records per the archived dictionary (removed from the SEER site); web tables as published, thousands separators removed
Code tables Site recode ICD-O-3/WHO 2008, 2023 revision, COD to site recode HTML tables with row/column spans expanded; heading rows become group; footnote markers kept as published
SEER*Stat dictionary seerstat-variable-dictionary-nov2025.xlsx As published
Case-data dictionaries Data description Nov 2025 (PDF); Nov 2018 record description (PDF) and read.seer.research.nov2018.sas (archived) Parsed from PDF text (pdftotext -layout); 198 / 141 field entries, about 1,400 / 1,060 codes. Code tables with unusual layouts can be incomplete; the full field description is kept

Validation

logs/validation.csv (40 checks, all passed):

  • Row counts: rows in each population file, summed over years, equal the records in the source file, for all four files. Every expected year is present.
  • Consistency: in every year, the four files give the same U.S. total.
  • Published totals: U.S. July 1 resident population totals match the Census Bureau figures exactly:
    • 1990: 249,622,814
    • 2000: 282,162,411
    • 2023: 336,806,231
    • 2024: 340,110,988
  • Standard populations: the 2000 U.S. standard population totals 274,633,642 in every file. Every standard million sums to 1,000,000. The published 2000 U.S. Std Million is reproduced from the Census P25-1130 population to within 2 per million (SEER rounded it so it sums to 1,000,000).
  • Case pipeline: 22 synthetic round-trip tests pass:
    • tab and comma exports, gzipped exports, exports with and without a header row;
    • missing characters (space, period), quoted labels, and leading zeros;
    • detection of a duplicate key;
    • legacy fixed-width files;
    • refusal to write inside the upload folder.

Licence and citation

  • Population data: Surveillance, Epidemiology, and End Results (SEER) Program Populations (1969–2024) (www.seer.cancer.gov/popdata), National Cancer Institute, DCCPS, Surveillance Research Program, released February 2026. 2010–2019 estimates: Woods & Poole Economics, Inc., 2010 to 2019 Intercensal Estimates of County Population by Age, Sex, and Bridged Race for Vintage 2024 Postcensal Estimates, Washington D.C., January 2026.
  • Standard populations, recode tables and documentation: National Cancer Institute, SEER Program (seer.cancer.gov). NCI text is free of copyright; credit NCI as the source.
  • This compilation (pipeline, dictionaries, layout) is released under CC0-1.0. NCI does not endorse it.
Downloads last month
123