Data Collection Methodology

Data Collection Methodology

All data presented in the survival curves, or clinical outcomes charts, on these web pages are derived from the National Cancer Data Base (NCDB)--a joint program of the Commission on Cancer (CoC) of the American College of Surgeons and the American Cancer Society.  It is a resource maintained by the American College of Surgeons. The information below, based on the NCDB Survival Reports User Help Sheet1, outlines the methods and limitations associated with the NCDB data. 

Five-Year Survival Curves from NCDB Survival Reports

The survival curves presented here provide overall survival estimates for selected cancer sites at one, two, three, four, and five years from the date of diagnosis. The curves show the percent of the group alive at these different time points. There is a separate chart for each stage of disease. Stage is classified using the American Joint Committee on Cancer (AJCC) combined stage group (5th edition). Reports were run separately for Seattle Cancer Care Alliance (SCCA), other Academic/Research Centers, Community Cancer Centers, and Comprehensive Community Cancer Centers. The NCDB Survival Reports provided the methods for generating the curves and the definition of center type.

Case Selection

Cases were diagnosed between 2003 and 2005. At the time the survival curves displayed here were generated, 2005 was the most recent year that five-year follow-up information was available (vital status and date of last contact or date of death). As described in the NCDB Survival Report manual, cases are limited to adult patients (over the age of 18) with only one diagnosis of cancer who have received all or part of their first course of therapy at SCCA or the relevant hospital from the NCDB. As many patients treated at SCCA receive second-line therapy after primary treatment elsewhere, the total number of patients treated at SCCA for each type of cancer is generally substantially larger than the numbers presented here. Other exclusions include unknown AJCC stage of disease and incomplete vital status information. Cases from hospitals without current accreditation from the Commission on Cancer (CoC)2 are excluded.

Methods and Limitations

The five-year observed survival rates are estimated using the actuarial method with one-month intervals. The endpoint is death from any cause (not cancer specific death); patients may have died from causes unrelated to their cancer. Calculations were performed using the NCDB Survival Reports software tool. Survival rates are not displayed when fewer than 30 cases are available, as survival rates calculated from small numbers of cases can yield misleading results and may have very wide confidence intervals.

The figures present unadjusted outcomes. That is, the NCDB did not account for demographic differences (e.g., age at diagnosis, gender, socioeconomic status, and insurance status), prognostic factors, and comorbidities for SCCA and other hospitals. Also, the NCDB did not account for subjective differences in staging practices among hospitals. For example, it is possible that a cancer considered Stage I at one hospital might be considered Stage II at another hospital due to practice pattern variations. The outcomes comparisons presented here might have differed if the NCDB had accounted for such demographic and staging differences in our analyses.

1 The full NCDB Survival Reports User Help Sheet is available here.  

2 The Commission on Cancer is a program of the American College of Surgeons.