6.2 The National Cancer Database (NCDB), RCRS Submission & CoC Quality Measures
Key Takeaways
The NCDB is a joint program of the Commission on Cancer and the American Cancer Society that began in 1989, receives data from more than 1,500 CoC-accredited programs and captures about 72% of newly diagnosed U.S. cancers.
RCRS, released in September 2020, replaced the Rapid Quality Reporting System (RQRS) and the separate annual NCDB Call for Data submission; the CP3R report tool was retired and its measures moved into RCRS.
CoC Standard 6.4 requires submitting all new and updated cases to RCRS at least once each calendar month and all complete analytic cases once a year as specified by the annual Call for Data.
The cancer committee must review RCRS data and quality measure performance rates at least twice each calendar year, and these reports cannot double as the CLP's Standard 2.2 reports.
NCDB quality measures fall into accountability, quality improvement and surveillance categories, and programs compare their performance with the CoC's expected Estimated Performance Rate (EPR).
What the NCDB is
The National Cancer Database (NCDB) is a joint program of the Commission on Cancer and the American Cancer Society. It began in 1989 and is a nationwide oncology outcomes database built from hospital registry data. It receives data from more than 1,500 CoC-accredited programs in the United States and Puerto Rico, captures about 72% of newly diagnosed U.S. cancers at the institutional level, and holds roughly 40 million records. Accredited programs use NCDB tools to compare their care with state, regional and national patterns. Researchers use de-identified Participant User Files (PUFs).
NCDB data are hospital-based, not population-based. They come only from accredited programs and include only analytic cases. NCDB therefore does not measure incidence. That job belongs to SEER and NPCR central registry data.
STORE notes that NCDB reporting takes place under the facility's American College of Surgeons Business Associate/Data Use Agreement. Services that share the hospital's Federal Employer Identification Number (FEIN) are generally covered by the hospital's accreditation and are reportable to NCDB.
From annual batches to RCRS
For three decades, programs submitted a year's worth of completed cases in the annual NCDB Call for Data, and some breast and colorectal measures were tracked concurrently in the Rapid Quality Reporting System (RQRS). In September 2020 the NCDB released the Rapid Cancer Reporting System (RCRS). RCRS replaced RQRS and the separate Call for Data submission, and the Cancer Program Practice Profile Reports (CP3R) tool was retired, with several of its measures moved into RCRS. CoC then retired Standards 6.2 (Data Submission) and 6.3 (Data Accuracy) in 2021 and made Standard 6.4 the data submission standard.
Standard 6.4: Rapid Cancer Reporting System – Data Submission
| Requirement | Detail |
|---|---|
| Monthly submission | All new and updated cancer cases are submitted at least once each calendar month. A calendar month runs from the first day to the last day of the month. |
| Annual complete submission | Once each calendar year, the program submits all complete analytic cases for all disease sites through RCRS, as specified by the annual Call for Data |
| Active participation | The program participates in RCRS through the whole accreditation cycle and follows the RCRS terms and conditions |
| Committee review | RCRS data and the required quality measure performance rates are reported to the cancer committee at least twice each calendar year and documented in the minutes |
| No double counting | RCRS reports cannot also satisfy the CLP's NCDB presentations under Standard 2.2 |
The standard describes RCRS as a way for programs "to report data on patients concurrently and receive notifications of treatment expectations." Because cases are submitted while treatment is still under way, RCRS can flag patients who have not yet received expected care, such as radiation after breast-conserving surgery or adjuvant chemotherapy for stage III colon cancer, while there is still time to act. It also shows performance rates for each CoC quality measure compared with state, peer and national rates.
Which cases go to RCRS and NCDB?
- Analytic cases (Class of Case 00–22) for tumors the CoC requires. STORE defines these.
- Class of Case 00 cases are analytic, but CoC does not require them to be staged or followed.
- Nonanalytic cases (Class of Case 30–99) are not required by CoC. A program may still collect them for the cancer committee or the state.
CoC quality measures
The CoC's Quality Assessment and Data Committee reviews the measure portfolio every year, adds new measures and retires those that have been "maximized." Measure specifications, evaluation years and expected performance thresholds are posted on the NCDB site and in the RCRS user library. Standard 7.1 requires the cancer committee to monitor the program's rates against the expected Estimated Performance Rate (EPR) and to act when a rate falls below it.
| Measure category | Meaning |
|---|---|
| Accountability | Strong evidence that the care process improves outcomes; suitable for comparing and holding providers accountable |
| Quality improvement | Evidence supports the process, but it is used mainly to guide internal improvement |
| Surveillance | Used to monitor patterns and trends in care, without judging performance |
Examples of long-standing measure concepts, with wording that shows the "considered or administered" convention:
- Breast, radiation after breast-conserving surgery: radiation therapy is considered or administered within a set time after diagnosis for women under 70 treated with breast-conserving surgery.
- Breast, radiation after mastectomy (MASTRT): radiation is considered or administered within 1 year (365 days) of diagnosis for women with 4 or more positive regional lymph nodes.
- Colon, adjuvant chemotherapy (ACT): chemotherapy is considered or administered within 4 months (120 days) of diagnosis for patients under age 80 with AJCC stage III (lymph-node-positive) colon cancer.
- Colon, 12 regional lymph nodes (12RLN): at least 12 regional lymph nodes are removed and pathologically examined for resected colon cancer.
"Considered" matters. A patient who refused recommended therapy, or for whom the physician documented a contraindication, can still meet a measure when the registry codes the reason correctly (for example, chemotherapy code 87 for refusal or 82 for contraindication). This is one reason accurate treatment coding and follow-up are quality work, not just clerical work.
The registrar's role in RCRS
- Submit monthly, even for cases that are still incomplete. RCRS expects new cases to be updated as treatment is given.
- Watch for treatment-expectation notifications, and find out whether the care was given elsewhere, refused, contraindicated or still pending.
- Correct errors that EDITS or RCRS flag, and resubmit.
- Prepare the twice-yearly report of RCRS data and measure performance for the cancer committee.
Worked example
RCRS shows 70% performance on a breast radiation measure. The ODS reviews the 18 patients who "failed" and finds that 12 received radiation at an unaffiliated center, 2 refused it, and 4 have no documentation. After obtaining the outside treatment summaries, coding the refusals with the reason-for-no-treatment codes and resubmitting, the performance rate is recalculated. Only then does the committee decide whether a real care gap remains and needs a Standard 7.3 initiative.
Under CoC Standard 6.4, how often must an accredited program submit new and updated cases to the Rapid Cancer Reporting System?
At least once each calendar month
Quarterly, before each cancer committee meeting
Once a year, during the annual Call for Data
Only when a case is complete, including all first-course treatment
A Cancer Liaison Physician wants to use the twice-yearly RCRS quality measure report to also satisfy the CLP's own NCDB presentation requirement. What does the CoC allow?
Nothing; the RCRS reports required by Standard 6.4 cannot also satisfy the CLP reports required by Standard 2.2.
One combined report per year is enough for both standards.
The combination is allowed if the CLP presents the RCRS report personally.
The combination is allowed for programs with fewer than 500 analytic cases.
A 72-year-old patient with resected stage III colon adenocarcinoma is offered adjuvant chemotherapy within 4 months of diagnosis and refuses it, and the refusal is documented. How can the registry make sure the colon adjuvant chemotherapy measure reflects this care accurately?
Code chemotherapy as given (03), because it was offered
Leave chemotherapy blank until the patient changes their mind
Code chemotherapy 87 (recommended but refused by the patient), because the measure counts chemotherapy that was considered or administered
Exclude the patient from the registry, because refusals are not analytic cases
Sections you finish are checked off in the contents.