American author Mark Twain credited former British prime minister Benjamin Disraeli with coining the phrase, “lies, damned lies, and statistics.”1 For this inaugural National Cancer Database (NCDB) cancer bytes column, I would like to describe the NCDB, its relationship to the American College of Surgeons (ACS) Commission on Cancer (CoC) programs, and summarize how the database is used. For many of us in surgical oncology, the NCDB helps us get past the “lies and damned lies” about cancer care with verifiable statistics from hospital registry data that are collected from CoC-accredited facilities.
What are the CoC and NCDB?
The CoC is an ACS quality improvement program. According to the American Hospital Association, there are 5,564 acute health care facilities in the U.S.2 More than 1,500 programs (28 percent) are accredited by the CoC and an estimated 70 percent of all cancer patients receive care from CoC-accredited programs.3 These programs submit deidentified case-level data for every cancer patient they diagnose or treat to the NCDB. Data abstraction (submission) is performed by the cancer program registry within six months of the patient’s first encounter with the reporting facility.
The NCDB was established in 1989 and is jointly funded by the College and the American Cancer Society. It contains records on more than 34 million patients and captures 250 data points for each patient. Data points include site of care (75 in all), demographic data, American Joint Committee on Cancer (AJCC) stage, histologic type, time to first treatment, distance traveled for treatment, type of first treatment, long-term all-cause mortality, and other data point information. The data are password protected and are accessible only to CoC-accredited programs. Just imagine the power of this database: it is accessible via the web, can crunch 8.5 billion cancer bytes (34 million patients multiplied by 250 data points apiece), and produces a result within a few seconds.
Other cancer registries
In collaboration with the National Cancer Institute (NCI), the Centers for Disease Control and Prevention (CDC) receives cancer-related data from all health care facilities in the U.S. This information is initially sent to each state’s Department of Public Health and then forwarded to the CDC in Atlanta, GA. The CDC reports demographic data, screening and risk factors, cancer incidence, prevalence, and mortality, and tracks related trends. These results can be searched at the county, state, and national level. This information is accessible to the public on the National Cancer Institute website.
Overall, the CDC database provides a quick glimpse into what is happening in cancer care, including tracking trends related to prevalence and mortality, whereas the NCDB allows a deeper dive, since it collects case-level treatment and staging data. Using both of these databases in conjunction with one another paints a bigger picture about cancer throughout the U.S. by state, county, and all CoC programs.
Complementing NCDB data with CDC data
Since 1997, the American Cancer Society’s National Colorectal Cancer Roundtable has been educating the American public on how colon cancer mortality rates may decrease by following screening recommendations. In 2014, the roundtable announced the “80% by 2018” initiative to have 80 percent of Americans older than 49 or younger persons in high-risk groups screened for colon cancer by 2018. By achieving this screening goal, the American Cancer Society estimates that 21,000 lives will be saved and 43,000 colorectal cancer cases prevented annually by 2030.4
Using CDC and NCDB data to get a better picture of geographic and temporal trends could help inform roundtable efforts and other policies. Figure 1 depicts the results from an opinion poll conducted through the CDC in 2005 concerning the public’s knowledge about getting colon cancer. People in north central and western states are apparently less aware of the steps they can take to lower the risk of getting colon cancer.
Figure 1. Health Information National Trends Survey (HINTS) mapping knowledge of lowering chances of colon cancer
Source: U.S. Department of Health and Human Services, State Cancer Profiles
Taking a deeper dive into colon cancer using the NCDB, we can look at AJCC combined pathologic and clinical staging groups during a 10-year period. In Figure 2 we have split these data between 2004–2009 and 2010–2014 because the AJCC definitions for colon staging changed between the release of 6th and 7th editions of the AJCC Staging Manual, which could affect trends over time. At first glance, it would appear that more changes in stage distribution occurred during the 6th edition, but that finding is likely due to data quality; that is, changes in the number of “unknown” stage cases fluctuated over time. Overall, the graph shows a progressive decrease in colon cancer incidence and no substantial change in the stage of colon cancer at diagnosis 2004–2014.
Figure 2. NCDB colon cancer diagnoses by combined AJCC pathologic and clinical stage, 2004–2014
Figure 3 shows age at time of diagnosis by year. There is a decrease in older age (70+) at diagnosis and, more importantly, an unexpected increase in the number of younger people developing colon cancer. This surprising shift was also identified this year using Surveillance Epidemiology and End Results (SEER) Program data.5 Five years earlier, Nancy You, MD, MHSc, FACS, reported this disturbing trend using NCDB data.6 Dr. You obtained her dataset by obtaining a participant user file (PUF) from the NCDB. This increase in younger people developing colon cancer was a hot topic in the national news during Colon Cancer Awareness Month in March 2016. Its etiology is still being studied, but obesity seems to play a role.
Figure 3. NCDB colon cancer diagnoses by age group, 2004–2014
The CoC cannot be certain these findings represent the general U.S. population since their data reflect the makeup of CoC-accredited hospitals and their patients, but SEER data is collected by the National Cancer Institute and corroborates these NCDB findings.
As these findings suggest, when NCDB data are viewed alongside CDC results, it is possible to get a comprehensive picture of trends in cancer. This perspective of using the NCDB to investigate trends among CoC-accredited facilities will be the purpose of this quarterly column in the future.
For more information about the NCDB, go to the ACS website. For more information about CDC State Cancer Profiles online.
Acknowledgement
The author thanks NCDB specialists Ryan McCabe, PhD; Ashley Loomis, MPH; and Nina Miller, MSSW, OSW-C, ACS Cancer Programs staff, Division of Research and Optimal Patient Care, for their assistance with this column.
References
- Twain M. Autobiography of Mark Twain. New York City, NY: Harper & Brothers; 1924.
- American Hospital Association. Fast facts on U.S. hospitals. Available at: www.aha.org/research/rc/stat-studies/fast-facts.shtml. Accessed May 31, 2017.
- American College of Surgeons. 27 accredited cancer programs receive the Outstanding Achievement Award from American College of Surgeons Commission on Cancer. Available at: facs.org/media/press%20releases/2016/oaa0216. Accessed May 31, 2017.
- Meester R, Doubeni CA, Zauber AG, et al. Public health impact of achieving 80% colorectal cancer screening rates in the United States by 2018. Cancer. 2015;121(13):2281-2285.
- Siegel RL, Fedewa SA, Anderson WF, et al. Colorectal cancer incidence patterns in the United States, 1974–2013. J Natl Cancer Inst. 2017;109(8):djw322.
- You YN, Xing Y, Feig BW, et al. Young-onset colorectal cancer: Is it time to pay attention? Arch Intern Med. 2012;172(3):287-289.
