By Frances Ross, CTR

1
Record Consolidation Test with the 2007 Multiple
Primary/Histology Rules

• By Frances Ross, CTR
• Presented at the
• NAACCR Annual Conference
• June, 2008

2
Background

• First RC test analyzed the variability among
  registries in applying the MP rules in effect in
  2000
• Created a test file of 743 source abstracts for
  320 patients
• Source abstracts represented all types of
  reporting facilities and a broad range of cancer
  types
• Abstracts were selected to test SEER MP rules
  (not a random or representative sample)

3
Background

• An Expert Panel determined the correct number of
  primaries in the test file and the correct values
  for critical data fields the answer file
• 11 central registries subjected the test file to
  their routine RC procedures and created their own
  answer file
• The answer files were compared on count of
  primary tumors by person and by cancer type
• The results were presented in a report to the
  Board and at the 2003 NAACCR Annual Conference

4
R C Test 2 - Purpose

• Repeat the work of the first test, using the same
  test cases and the same general procedures
• BUT
• Apply the new 2007 Multiple Primary and Histology
  coding rules

5
R C Test 2 - Methods

• Re-assembled a work group
• Engaged a new Expert Panel
• Selected all source abstracts for 236 patients
  from the original test file of 320 patients
• Targeted cancers in the new MP/H site groupings
• Not all of the Other sites cases were included
• Experts again determined a correct answer file,
  of tumors for each person, showing diagnosis
  date, primary site, histology, behavior code,
  and laterality

6
R C Test 2 - Results

• In total tumor counts, there were 4 fewer tumors
  under the 2007 MP/H rules
• By cancer site, there were
• 1 tumor for Melanoma
• 2 tumors for Head and Neck
• - 1 tumor for Lung
• -2 tumors for urinary system
• -4 tumors for other sites

7
R C Test 2 - Results

• By person, there were 19 patients with different
  tumor counts

8
R C Test 2 Sample report

• 111110019
• Rules DX Date Site Lat Hist
  Beh Rule
• --------------------------------------------------
  ---------------------------
• Old Rules 11162001 C509 2 8500
  3 Rule 3
• 04272002 C502 2 8500
  3 Rule 3
• New Rules 11152001 C509 2 8500
  3 MP 11 Hist 27
  
• 111115014
• Rules DX Date Site Lat Hist
  Beh Rule
• --------------------------------------------------
  ---------------------------
• Old Rules 11202002 C509 1 8522
  3 Rule 5a, Exc 3
• New Rules 11202002 C504 1 8520
  3 MP 12 Hist 1
• 11202002 C505 1
  8514 3 MP 12 Hist 14
• 111111102
• Rules DX Date Site Lat Hist
  Beh Rule
• --------------------------------------------------
  ---------------------------
• Old Rules 07152003 C189 0 8220
  3 Rule 1
• New Rules 07152003 C180 0
  8140 3 MP 11 Hist 11

9
R C Test 2 - Results

• Results by site Melanoma of the skin 1 tumor
• There was one additional melanoma primary under
  the 2007 rules. This was due to 2 lesions of the
  same site and histology, diagnosed at the same
  time, but one was on the left side of the chest
  and the other was at the midline.
• Under the 2007 rules, midline tumors are counted
  as separate primaries from right and left sided
  tumors.

10
R C Test 2 - Results

• Results by site Head and Neck 2 tumors
• There were 2 patients, each with 1 additional
  tumor according to the 2007 MP/H rules.
• This was due to the removal of the Same Site
  table formerly used under the previous SEER MP
  rules. One patient had lesions on the tonsil
  (C09.0) and oropharynx (C10.8) the other had
  lesions on the base of tongue (C01.9) and other
  parts of the tongue (C02.9). These are separate
  tumors in 2007.

11
R C Test 2 - Results

• Results by site Breast - no change in total
  count
• For the 4 patients with 1 fewer tumor each
• - 3 had new tumors diagnosed within 5 years, so
  the change in the timing rule made these
  recurrences and not second primaries
• - 1 had a different histology on biopsy from
  the mastectomy experts cited the most
  representative specimen instruction and coded
  this as one primary

12
R C Test 2 - Results

• Results by site Breast (cont.)
• For those patients with 1 additional tumor each
• - 3 patients had ductal carcinoma and another
  type (tubular, medullary, desmoplastic type).
  These are not considered types of ductal
  carcinoma under the 2007 rules, so they are
  counted as separate primaries when found in
  separate tumors.
• - 1 patient had tumors in both the right and
  left breast. It was originally interpreted as a
  metastasis, but since it was not explicitly
  stated as such, it was coded under the 2007 rules
  as a new primary.

13
R C Test 2 - Results

• Results by site Lung -1 tumor
• One patient had bilateral synchronous lung tumors
  two on the right side and one on the left.
  Only the right side was biopsied. The 2007 MP/H
  rules for lung include explicit statements
  regarding the number of tumors in each lung.
  This scenario fell to the last rule which states
  that it is to be considered one primary.

14
R C Test 2 - Results

• Results by site Kidney and Urinary System -2
  tumors
• Pre-2007 MP rules defined 2 site groupings
• Bladder
• Urinary system (kidney, renal pelvis, ureter,
  and other parts of the urinary tract)
• 2007 MP/H rules define 2 site groupings
• Kidney
• Urinary system (bladder, renal pelvis, ureter,
  and other parts of the urinary tract)

15
R C Test 2 - Results

• Results by site Kidney and Urinary System -2
  tumors
• Two patients had 1 fewer tumor each
• One had multiple tumors in the kidney and
  bladder, all diagnosed as PTCC. Since renal
  pelvis and bladder were in separate site
  groupings before 2007, this was 2 primaries.
  Under the 2007 rules, this is one site grouping
  and one primary.
• One patient had PTCC of the ureter and 2 years
  later, PTCC of the bladder. The 2007 site
  groupings and timing rules make this one primary
  now.

16
R C Test 2 - Results

• Results by site Other Sites grouping -4
  tumors
• One patient with a gastroesophageal tumor had a
  recurrence 14 months later. Under the 2007 MP
  rules, this is a new primary.

17
R C Test 2 - Results

• Results by site Other Sites grouping -4
  tumors
• Two patients with VIN3 had recurrences within a
  year but not within 2 months. Under the 2007 MP
  timing rules, these are not new primaries.

18
R C Test 2 - Results

• Results by site Other Sites grouping -4
  tumors
• The patients with the small bowel and unknown
  primary cancers had very limited text. These
  cancer reports were interpreted differently by
  the second EP (not as new primaries), but this is
  not related to changes in the 2007 MP/H rules.

19
R C Test 2 Issues and Discussion

• Histology coding changes
• - code the final diagnosis from the most
  representative specimen
• Breast
• - new instructions regarding use of combination
  codes Table 2 defines ductal types
• - instruction to invasive portion only
• - code inflammatory carcinoma only from final
  path
• - give precedence to comedocarcinoma if /2

20
R C Test 2 Issues and Discussion

• Histology coding changes
• Colon
• - code 8210 if a polyp is mentioned
• - use of 8220 (FAP) and 8221
• Urinary
• - confusion arose over which rule to use when
  biopsy says renal cell, clear cell and final
  path says renal cell carcinoma

21
R C Test 2 Issues and Discussion

• Breast cancers The changes in MP rules resulted
  in more synchronous tumors and fewer subsequent
  tumors, so counts by diagnosis year could differ.
• Urinary system Since bladder is now grouped
  with renal pelvis, ureter, and other urinary
  sites, there may be fewer primaries abstracted.
  A new question arose regarding which topography
  code to use for multiple tumors of the bladder
  and other urinary sites.

22
R C Test 2 Issues and Discussion

• For example, which topography code should be used
  for multiple tumors of the bladder and another
  urinary site, i.e., ureter?
• C67.__ Bladder
• C66.9 Ureter
• C68.8 Overlapping lesion of urinary organs
• C68.9 Urinary system, NOS

23
R C Test 2 Issues and Discussion

• Example (cont.)
• If you code this scenario to C67._ (bladder) and
  the patient has a new bladder tumor after 3 yrs.,
  it is NOT a new primary, but
• If you code this to C68._ (urinary system) and
  the patient has a new bladder tumor after 3 yrs.,
  it IS a new primary per rule M7. By using code
  C68 the information about which urinary sites
  were involved gets lost.

24
R C Test 2 Conclusion

• - Study was a descriptive analysis of test cases
  (not a random or representative sample from a
  population)
• - Identified specific scenarios where tumor
  counts will be different after 2007, especially
  by cancer site, but could not quantify impact
• - Changes occur at patient level therefore
  counts by race and gender may be affected. POC
  and other studies will be affected as well, by
  the mix of tumor records available

25
R C Test 2 Recommendations

• Specify additional site coding guidelines,
  particularly for urinary system
• Review SEER site recode groupings, since they do
  not align with the 2007 MP/H site groupings
• Develop RC guidelines for central cancer
  registries to
• - identify the appropriate site specific rule
  set (i.e, solid tumors, benign brain,
  hematopoietic diseases)
• - evaluate accuracy of data source to use during
  RC
• - define how to handle situations with
  ambiguous, conflicting or insufficient
  information

26
Acknowledgements

• Many thanks to the Record Consolidation Test
• Experts and Work Group Members
• Peggy Adamo, RHIT, CTR Jerri Linn Phillips, MA,
  CTR
• NCI/SEER American College of Surgeons
• Deborah Bringman, MPH Frances E. Ross, BA, CTR
• University of California, Irvine Kentucky Cancer
  Registry
• Susan J. Capron Wendy Scharber, RHIT, CTR
• SEER Contractor Registry Widgets
• Cynthia Dryer, BA, CTR John Young, Jr., DrPH,
  CTR
• State Health Registry of Iowa Georgia SEER
  Registry
• Fran Michaud, BS, CTR
• CDC/NPCR