Date of Diagnosis: If a clinician states his current diagnosis of malignancy is based on a CT scan done at an early date that contained a diagnosis of only "neoplasm" or "worrisome for carcinoma" should the date of diagnosis be the date of the scan?
Yes. Code the Date of Diagnosis field to the date of the scan. The physician's clinical impression upon reviewing the earlier scan, is that the malignancy was confirmed by the scan. If there is a medical review of a previous scan that indicates the patient had a malignancy at an earlier date, then the earlier date is the date of diagnosis, i.e., the date is back-dated.
EOD-Size of Primary Tumor--Melanoma: How do you code tumor size for a melanoma diagnosed by a positive lymph node biopsy when the primary site is coded C44.9 because no primary site was identified? See discussion.
Should the size be 000 because no primary was found or 999 for unknown?
For cases diagnosed 1998-2003:
Code the EOD-Size of Primary Tumor field to 000 [No mass; no tumor found] when primary site is coded to C449.
Date of Diagnosis: When doing follow-back at nursing homes on DCO cases, we find it difficult to code diagnosis date because the nursing home records are often vague or incomplete. Should the diagnosis date be coded as unknown (excluded from SEER database), the date of death, or the approximate date of diagnosis as reported on the death certificate?
If the nursing home record indicates that the patient had cancer, use the best approximation for date of diagnosis.
If the record says the patient had cancer when admitted, but it does not provide a date of diagnosis, use the date of admission as the date of diagnosis.
If there is no mention of cancer in the nursing home record and/or all work-up in the record is negative, assume the cancer was discovered at autopsy. Use the date of death as the date of diagnosis, and leave as a Death Certificate Only case.