Every year, the Cancer Program looks at overall survival for the four major cancer diagnoses, and has compared them with the rest of the Puget Sound regional survival. This year, since regional data was difficult to obtain, we used data from the National Oncology Data Base, which reflects cancer care across the United States.

The four most common cancer diagnoses are Prostate, Breast, Lung, and Melanoma. Melanoma may have a falsely high incidence in comparison to the CSS region because we have a community wide registry rather than a hospital only registry. The standard fourth “major cancer diagnosis” is ordinarily Colo/rectal cancer in CSS and national statistics.

CSS did not utilize TNM staging. It utilized a local, regional, and distant staging system. The National Oncology Data Base (NODB) uses AJCC staging, which is more specific than the system used by CSS. Hence, for the first time, we are able to make comparisons by true staging.

In order to assure that the survival results are comparable, it is necessary to ensure that the incidence by Stage is comparable. Survival can be easily affected by stage migration, so fair comparisons would require fairly comparable incidence by Stage. According to the 2006 cancer statistics table, only prostate cancer has discordance between our stage incidence and national numbers.

Survival curves on CHARTS 1 through 4 demonstrate that overall survival for each of the four major sites is essentially the same, if not slightly better, between St. Joseph’s and NODB.

Top 4 Cancer Sites Survival
Chart 1	Chart 2
Chart 3	Chart 4

CHARTS 5 through 9 compare our survival by stage with NODB survival. Amazingly, the survival curves are nearly superimposable for each cancer and each stage. There is a slight discrepancy in late survival that is not statistically significant and due to small numbers. Interestingly, the best 5-year survival in Prostate Cancer is St. Joseph Stage 3 patients, with somewhat better survival compared to NODB Stage 2 & 3 survival. This is a reflection of “stage migration” and co morbidities that occur when comparing different mixes of clinical and pathologically staged cancers.

Survival by Site of Cancer

Chart 5

Chart 6

Chart 7

Chart 8

In Summary, it appears that our survival results are either slightly better or essentially the same as the survival results of a large national cancer data bank. This is the first year we have been able to report this finding and we can be proud of our excellent results.

Return to previous page