05 / Evidence & Sources

Every number has a denominator.

Colonoscopy First is an advocacy campaign, but the evidence comes first. This page shows what each study actually measured, what it did not prove, and which campaign claims rely on it.

How we use evidence Core evidence Younger adults & age 45 Sources by post Full references

How we use evidence

Read the endpoint before the headline.

A screening study can be completely accurate and still answer a different question from the one a patient thinks it answers. These are the interpretation rules used throughout this campaign.

Cancer sensitivity is not the same as advanced-precancer sensitivity.

Single-round test performance is not the same as long-term program effectiveness.

Post-positive colonoscopy yield is not upstream screening-test sensitivity.

Evidence from organized FIT programs is not automatically evidence for multitarget stool DNA.

Invitation-level outcomes differ from outcomes among people who actually complete a test.

Industry funding is disclosed and scrutinized; it does not automatically invalidate a result.

Core evidence

What the studies actually show.

Use the source link on each card to read the primary publication, regulatory document, or guideline. Campaign posts link back to these evidence anchors rather than carrying a full bibliography on every image.

E01 · GUIDELINE · SCREENING OPTIONS

ACS 2026 colorectal cancer screening guideline

Average-risk adults should begin screening at age 45. Stool-based tests and visual exams are preferred options; positive non-colonoscopy tests require timely colonoscopy.

Start at 45 · Timely colonoscopy after a positive non-colonoscopy test

What it establishes: Multiple screening options are legitimate. Informed choice and completed screening matter.
Important limits: Guideline support does not mean identical advanced-precancer detection, burdens, or same-session treatment capability.
Funding / conflicts: Guideline evidence synthesis; conflicts disclosed in the publication.

Primary source ↗

E02 · PRIMARY STUDY · COLOGUARD PLUS

BLUE-C published diagnostic-accuracy study

Prospective diagnostic-accuracy study of the next-generation multitarget stool DNA test with screening colonoscopy as the reference standard.

93.9% colorectal-cancer sensitivity · 43.4% advanced-precancer sensitivity · one screening round

What it establishes: Strong direct evidence for single-round next-generation stool DNA performance.
Important limits: Not a randomized trial of colorectal-cancer incidence, mortality, or lifetime prevention. Repeated screening and adherence matter.
Funding / conflicts: Funded by Exact Sciences; multiple authors were Exact Sciences employees.

Primary source ↗

E03 · FDA · REGULATORY REVIEW

FDA Cologuard Plus primary-effectiveness analysis

The FDA primary-effectiveness population is a different analysis set from the published BLUE-C paper and should be cited separately.

95.3% colorectal-cancer sensitivity (81/85) · 43.3% advanced-precancer sensitivity (849/1,962)

What it establishes: Regulatory performance estimates used in FDA labeling and campaign assets such as “The Other Number.”
Important limits: Do not splice these numbers into the published BLUE-C 93.9% / 43.4% pair.
Funding / conflicts: Regulatory source; the underlying pivotal study was sponsored by Exact Sciences.

Primary source ↗

E04 · FDA TABLE 18 · DENOMINATORS MATTER

What a negative Cologuard Plus result can still contain

FDA Table 18 describes residual findings among people whose Cologuard Plus result was negative.

1,113 of 16,414 negative results had advanced precancer (6.8%) · 1,113 of 1,962 people with advanced precancer tested negative (56.7%)

What it establishes: Supports the “about 1 in 15” and “about 57%” messages when the denominator is stated correctly.
Important limits: 6.8% is not sensitivity. 56.7% applies to people who already had advanced precancer. Both describe one screening round.
Funding / conflicts: Regulatory source; the underlying pivotal study was sponsored by Exact Sciences.

Primary source ↗

E05 · CAPABILITY · DIRECT EXAMINATION

What colonoscopy can do

Colonoscopy directly examines the colon and rectum and can remove detected polyps during the procedure.

Direct examination · biopsy/removal can occur during the same exam

What it establishes: Supports the campaign’s core functional distinction between stool-based screening and colonoscopy.
Important limits: Colonoscopy can miss lesions. Preparation, sedation, bleeding, perforation, and other burdens/risks matter. “Can remove” is not a guarantee.
Funding / conflicts: U.S. government source.

Primary source ↗

E06 · RANDOMIZED TRIAL · COLONOSCOPY

NordICC randomized colonoscopy-invitation trial

A pragmatic randomized trial compared invitation to one-time screening colonoscopy with usual care.

10-year CRC risk 0.98% vs 1.20% · risk ratio 0.82 · 42% of invited participants underwent colonoscopy

What it establishes: Randomized evidence that invitation to colonoscopy reduced colorectal-cancer incidence at 10 years.
Important limits: Invitation-level effect. The primary mortality estimate was imprecise and did not establish a statistically significant mortality reduction.
Funding / conflicts: Funded by the Research Council of Norway and others.

Primary source ↗

E07 · COUNTEREVIDENCE · PROGRAM EFFECTIVENESS

COLONPREV: colonoscopy invitation vs biennial FIT

COLONPREV tested screening strategies under real-world participation: invitation to one-time colonoscopy versus repeated biennial FIT.

Participation 31.8% vs 39.9% · 10-year CRC mortality 0.22% vs 0.24% · FIT strategy met noninferiority criterion

What it establishes: A well-run repeated FIT program can achieve excellent population-level outcomes when participation is higher.
Important limits: FIT is not multitarget stool DNA. This is invitation/program-level evidence, not intrinsic performance of completed tests in identical adherers.
Funding / conflicts: Funded by Fundación Científica de la Asociación Española contra el Cáncer and Instituto de Salud Carlos III.

Primary source ↗

E08 · REAL WORLD · ORGANIZED FIT + COLONOSCOPY

Kaiser organized screening: the system matters

Kaiser Permanente Northern California used annual FIT outreach, registries, reminders, tracking, and colonoscopy follow-up.

Up-to-date screening 38.9%→82.7% · CRC incidence −25.5% · CRC mortality −52.4%

What it establishes: Organized outreach and dependable follow-up can make noninvasive screening highly effective at a population level.
Important limits: Observational, not randomized. FIT, not Cologuard. Integrated-system infrastructure is part of the intervention.
Funding / conflicts: Research support reported in the publication.

Primary source ↗

E09 · EMERGING EVIDENCE · SELECTED DENOMINATOR

What the reported 66.3% post-positive ADR means

A 2026 conference abstract reported adenoma detection at colonoscopy after a positive multitarget stool DNA test.

66.3% adenoma detection rate after positive mt-sDNA and completed colonoscopy

What it establishes: Positive stool tests enrich the subgroup that reaches colonoscopy and may help triage limited colonoscopy capacity.
Important limits: This is post-positive colonoscopy yield, not stool-test sensitivity. It excludes test-negative patients and uses an any-adenoma endpoint. Publication-only abstract.
Funding / conflicts: Research funded by Exact Sciences.

Primary source ↗

Younger adults & age 45

Start on time. Keep the numbers in context.

Rising younger-age cancer incidence supports timely screening. Colonoscopy’s ability to examine the colon and remove detected precancer supports this campaign’s prevention-focused preference. These are two distinct lines of evidence.

Registry surveillance · United States

What the current age shift means

Siegel and colleagues’ 2026 report describes U.S. colorectal-cancer incidence through 2022.

About 3% annual incidence growth at ages 20–49, 2013–2022.
Half of diagnoses before 50 occur at ages 45–49.
Under-55 share of all diagnoses: 11% in 1995 → 22% in 2022.

Different denominators: The annual trend is a population incidence rate. The other figures are shares of diagnoses, not individual probabilities or a doubling of personal risk. The paper states the half-of-under-50 finding in its screening discussion without assigning that sentence a separate year window; it is not a projection of lifetime risk.
Detection and disease: Screening uptake at ages 45–49 contributed to recent acceleration through detection of existing, asymptomatic cancers. Regional and distant-stage incidence also rose. Trend rates exclude appendiceal cancer and account for age and reporting delays.
Funding & conflicts: Authors were ACS employees, paid from ACS funds, and declared no conflicts. ACS receives foundation grants, including health-sector-linked grants, for other work; these authors were not funded by or key personnel on those grants.

Read Siegel et al., 2026 ↗

Modeled birth cohorts · United States

Generations, with uncertainty

Siegel et al. analyzed 490,305 invasive colorectal-cancer diagnoses at ages 20 and older in nine SEER registry areas, 1974–2013.

Born around 1990 versus around 1950:
Colon cancer: 2.40× (95% CI 1.11–5.19).
Rectal cancer: 4.32× (95% CI 2.19–8.51).

What these estimates mean: Age-period-cohort models compared age-specific incidence, accounting for age and calendar-period effects. The approximate birth years label cohorts centered on 1989 and 1949; the 1950-era cohort was the low-risk reference.
What they do not mean: These are modeled relative comparisons, not observed lifetime risks, a comparison with your parents, or an estimate of your absolute risk. The youngest cohort has limited follow-up and wide confidence intervals. Registry coverage and modeling assumptions limit generalization. The study does not compare screening tests.
Funding & conflicts: Supported by the ACS Intramural Research Department and NIH/NCI Intramural Research program. The paper reports no funder role in study conduct or publication. No separate author conflict statement was identified in the article; this is not a declaration that no conflicts exist.

Read Siegel et al., 2017 ↗

What the campaign takes from this

Start at 45 if you are at average risk. ACS supports both visual examinations and stool-based tests; it does not endorse this campaign’s colonoscopy-first preference. Screening guideline ↗

Colonoscopy can find and remove detected precancerous polyps in one exam. Stool tests cannot remove polyps; positive stool screening requires colonoscopy. Not every polyp becomes cancer, and colonoscopy cannot prevent every cancer. Preparation, risks, examination quality, access, and patient preferences matter. NCI evidence ↗

These incidence studies do not establish colonoscopy’s superiority or show that advanced polyps are increasing. Stool screening can be valuable for people who would otherwise remain unscreened; do not wait indefinitely for colonoscopy. Symptoms or elevated risk can require earlier evaluation, separate from routine screening below 45. Compare the options ↗

Make your screening plan ↗

Sources by post

What supports each campaign asset.

This public cross-reference lets a reader move from a social post to the evidence behind it. Internal publication status and legal-review notes remain in the private publication matrix.

Campaign asset	Evidence	Interpretation note
94% of What?	E02 · E05	Published BLUE-C pair; one-round patient-level sensitivity; AI illustration disclosure.
The Other Number	E03 · E05	FDA 95.3% / 43.3% pair only; do not relabel as the BLUE-C publication pair.
Both Recommended. Not Equivalent. — 1 in 15	E01 · E04 · E05	6.8% is residual advanced precancer among negative results, not sensitivity.
Don’t Stay Unscreened	E01 · E08	Expansion message: a completed alternative screen is better than remaining unscreened.
These Can Both Be True	E03 · E04 · E05	A negative result can coexist with advanced precancer; illustrative report/image labeling retained.
Before Cancer	E05 · E06	Functional prevention capability; no numeric claim on the creative.
What One Number Doesn’t Tell You carousel	E02 · E05	Published 93.9/43.4 pair; funding disclosed; one screening round.
100 Patients reel	E02 · E05	100-person sensitivity illustration; never reinterpret as prevalence among negative tests.
Dad Ads — Camping / Grill / Rental	E01 · E03 · E05	FDA 43.3% implies 56.7% test-negative among APL-positive participants; standardized caption.
Professional series — Barber / Home Inspector / Wedding	E05	Analogy is creative; medical bridge is a capability statement.
Test After the Test	E01 · E05

Counterevidence

Evidence that keeps the campaign honest.

A colonoscopy-forward message is stronger when it acknowledges what organized stool-based screening can accomplish.

COLONPREV matters. Repeated biennial FIT invitation achieved higher participation and was noninferior to colonoscopy invitation for 10-year colorectal-cancer mortality in that program. It was FIT—not multitarget stool DNA—and it tested screening strategies under observed participation, not intrinsic performance of completed tests in identical adherers.

Kaiser matters. Organized annual FIT plus colonoscopy follow-up dramatically increased screening and was associated with lower colorectal-cancer incidence and mortality. The intervention included registries, outreach, repeat testing, tracking, and follow-up—not simply mailing a kit.

Independence & disclosures

Advocacy, with the conflicts visible.

Colonoscopy First is independent physician-led education. The campaign is not presented on behalf of a medical practice, hospital, health system, professional society, screening-test manufacturer, or endoscopy-equipment manufacturer unless specifically stated.

When a study used in campaign content was funded by a manufacturer—or included manufacturer employees as authors—that relationship is disclosed with the post or on this page. Funding is treated as context to evaluate methods and replication, not as a reason to dismiss a study automatically.

Full references

Primary sources and guidance.

American Cancer Society colorectal cancer screening guideline, 2026. Open source ↗
Imperiale TF, et al. N Engl J Med. 2024;390:984–993. Open source ↗
U.S. FDA. Cologuard Plus PMA P230043. Open source ↗
U.S. FDA. Cologuard Plus PMA P230043, Table 18. Open source ↗
National Cancer Institute. Screening Tests to Detect Colorectal Cancer and Polyps. Open source ↗
Bretthauer M, et al. N Engl J Med. 2022;387:1547–1556. Open source ↗
Castells A, et al. Lancet. 2025;405:1231–1239. Open source ↗
Levin TR, et al. Gastroenterology. 2018;155:1383–1391.e5. Open source ↗
Estes C, et al. J Clin Oncol. 2026;44(suppl 16):e15647. Open source ↗
Siegel RL, Wagle NS, Star J, et al. Colorectal cancer statistics, 2026. CA Cancer J Clin. 2026;76:e70067. Open source ↗
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. Open source ↗

Last evidence review: September 27, 2026. Evidence pages should be re-reviewed when FDA labeling, major screening guidelines, or pivotal/longitudinal studies materially change.