How Accurate Is Shield? Cancer Detection, Missed Polyps and False Results

Use the accuracy information to choose and complete a screening plan. If you are considering Shield, ask a provider whether it fits your history and what follow-up either result would require before arranging an order. If you already have a result, send the actual report for follow-up rather than using a percentage to interpret it on your own.

Care starts by chat with a licensed provider. Message us and we will confirm coverage in your state.

The headline percentage needs a denominator

Shield detected 83.1% of colorectal cancers in the ECLIPSE study. That number describes how often the blood test was positive among participants who actually had cancer. It does not describe the chance that a positive result means cancer, and it does not measure how well the test finds precancerous polyps. [1]

Those are different questions. If you are deciding whether to use Shield, the useful discussion includes what it detects well, what it often misses, and what happens after either result. A single "accuracy" percentage cannot answer all three.

This guide separates the study's cancer-detection results from its precancer findings and the meaning of an individual report. It also explains why a high negative predictive value should never be used to dismiss new symptoms.

What the study measured

ECLIPSE compared the blood test with screening colonoscopy in an average-risk population. Of the clinical validation cohort, 7,861 participants were eligible and evaluable. The study asked how often the blood test found cancer, how often it was negative without advanced neoplasia, and how often it detected advanced precancerous lesions. [1]

In plain language, sensitivity starts with people who have the target condition and asks how many the test finds. Specificity starts with people who do not have the defined target and asks how many test negative. You need to know what the target is before comparing the percentages.

• Cancer sensitivity: 83.1%. About 17 in 100 people with colorectal cancer had a negative blood result in that study. [1]
• Advanced precancer sensitivity: 13.2%. The test detected only a small share of these lesions. [1]
• Specificity without advanced neoplasia: 89.6%. About 10 in 100 participants without colorectal cancer or advanced precancerous lesions nevertheless had a positive blood test. [1]

These are study estimates. They are not promises about the next person tested. ECLIPSE was funded by Guardant Health, the test's manufacturer, and used a defined screening population. That context belongs beside the performance claims. [1]

Early cancer and precancer are the weak points

The FDA labeling warns of limited detection of stage I colorectal cancer, approximately 55%-65%. It also warns that Shield misses about 87% of precancerous lesions. Those limits are more informative for a screening decision than an overall cancer-detection percentage on its own. [2]

A cancer test can perform differently at different stages. Combining stages into one figure can hide poorer performance at the earliest stage. Do not read a statistic for several stages grouped together as a stage I result.

Precancer detection is a separate goal because screening can identify a growth before it becomes cancer. Shield's low detection of advanced precancerous lesions limits that opportunity. The ACS's 2026 guidance therefore places blood tests after preferred stool or visual options that a person declines or does not complete. [2,3]

This does not turn every missed polyp into an inevitable future cancer. It means that the test has a substantial blind spot in the category of findings that screening programs also try to detect. Avoid both extremes: treating a negative blood result as a clear colon, or assuming that every negative result has missed a dangerous lesion.

A positive result has a different denominator

The question "How likely is cancer after a positive result?" starts with everyone who tested positive. That is different from starting with everyone who had cancer. A positive report is not a cancer diagnosis. The labeling reports false positives and requires follow-up colonoscopy to investigate a positive result. [2]

Predictive values depend on how common the condition is in the population being tested. A number from average-risk screening cannot simply be transferred to someone with symptoms, a previous cancer, or a different risk profile. The FDA labeling excludes high-risk use and states that Shield does not replace diagnostic or surveillance colonoscopy. [2]

The same caution applies to the approximately 10% false-positive rate. It does not mean that a person with a positive Shield result has a 90% chance of cancer. That would reverse the calculation. A clinician interprets the report in context and arranges the examination that can establish what is present.

Why 99.9% negative predictive value is not an all-clear

The labeling reports a negative predictive value for colorectal cancer of 99.9% in the studied population. It separately reports a substantially lower negative predictive value for advanced adenoma, about 90%. The impressive-looking cancer number is not a statement that a negative test rules out precancerous polyps. [2]

A high negative predictive value and missed cancers can coexist. When relatively few screened people have cancer, most negative results will come from people without cancer even if the test misses some cancers. That is why the label tells people with negative results to continue screening. [2]

New blood in the stool, a bowel change that persists, unexplained weight loss or anemia still deserves assessment. Tell your clinician about the symptoms and the test result together. Do not let the result become a reason to omit the symptoms from the conversation. [4]

Finally, one round of detection is not proof of a reduction in deaths over years. The approval labeling states that repeated programmatic screening with Shield had not been studied; the ACS's 2026 assessment describes modeled outcomes and the need for ongoing evaluation. A modeled benefit is a prediction, not an observed mortality result from ECLIPSE. [2,3]

To put these numbers to work, read how Shield compares with other screening options and what to do after a result. You can message Well Revolution with the result and your screening history to begin a discussion of the next step. If you decide to go ahead with testing, our guide to getting Shield online in the US walks through the order, the draw and the result.

What it costs with Well Revolution

$50 for a prescription. If approved, you will receive a prescription for the Shield lab order and 30 days of support. Guardant runs the test and reports the result. We are there for those 30 days to interpret it, arrange referrals, and treat if needed. Perfect for a single screening test.

Or $90 for the Unlimited Membership. If approved, you will receive 3 months of care with 7 days a week support, plus unlimited prescriptions, lab tests, imaging orders, letters and referrals with same day care, cancel anytime, no fees. Perfect for established patients looking for full care.

Getting started with Well Revolution

Care is chat-based with a licensed provider, same day, no appointment needed. If the numbers here leave you unsure which test to choose, message us, tell us where you are, and we will confirm coverage in your state and talk through the options with you.

References

• [1] Chung DC et al. A Cell-free DNA Blood-Based Test for Colorectal Cancer Screening. 2024.
• [2] Guardant Health / FDA. Shield instructions for use and provider labeling, P230009C. 2024.
• [3] Wolf AMD et al. Colorectal cancer screening: An update to the American Cancer Society guideline, 2026. 2026.
• [4] CDC. Symptoms of Colorectal Cancer. 2026.

Reviewed by PA Michael Rubio, US Medical Director