If you've been reading this book, you know that I believe MCED testing has a trajectory that points toward a future where MCED testing may be integrated into preventive care, like mammograms and colonoscopies are. But I also believe the evidence for implementing MCED testing at a large scale is incomplete (as of 2026), which many critics have pointed out.
Criticism #1: No Mortality Evidence
Traditionally, for a screening test to be implemented at a national level, a randomized controlled trial showing that screened individuals live longer than unscreened individuals has been the gold standard. The entire rationale for population-based cancer screening rests on the idea that finding cancer earlier leads to better outcomes. But the demand for randomized controlled trials before any clinical use creates a paradox. A mortality endpoint trial for MCED testing requires enormous sample sizes (likely on the order of hundreds of thousands of participants), long-term follow-up (on the order of several years to even decades), and massive resources. While we wait for that trial, patients are dying of cancers that have no screening test at all. However, in my opinion, insisting on this type of evidence before any adoption may not be the right policy, when the alternative is doing nothing for people who ultimately get types of cancer that have no recommended screening.
Criticism #2: Sensitivity in Early-Stage Cancer
MCED tests are marketed as "early detection" technologies. The implicit promise is that they find cancer before it becomes advanced, so that it can be treated with curative intent. However, across the studies mentioned in this book, the sensitivity for detecting early-stage cancers is low (on the order of 25-40%), while the sensitivity for detecting advanced-stage cancers is nearly two to three times higher.
Criticism #3: Lead-Time Bias
Lead-time bias, as described in the post “Statistical Terms”, describes a phenomenon where a screening test finds cancer before it would have been found without a screening test, but the patient dies at the same time they would have died had they never undergone the screening test. In this scenario, the test did not really add survival time. The cancer type in which this concept has the most credence is probably prostate cancer. Lead-time bias can only be resolved by mortality analyses in randomized controlled trials, which, as stated previously, MCED tests currently lack.
Criticism #4: False Positives at Scale
The specificities of the MCED tests described in this book are over 95%. Assuming even a fraction of individuals in the US undergoes MCED testing, a false positive at this level generates tens (or perhaps even hundreds) of thousands of people each year who receive a positive test result and then undergo weeks (or months) of diagnostic workup, and ultimately learn that they do not have cancer. Each of those false positives represents someone who spent that time in a state of anxiety, underwent procedures with some degree of risk and cost, and, perhaps more importantly, consumed healthcare resources that could have been allocated to patients who have cancer.
Criticism #5: Commercialization Before Evidence
One of the most unsettling criticisms of MCED testing is about the overall trajectory of the technology's incorporation into society. MCED tests are being sold commercially, legislated into Medicare coverage, and marketed aggressively, all before the most basic question about their ability to reduce mortality has been answered. A reimbursement framework, regulatory pathway, and commercial market are being built for a technology whose clinical utility has not been established by the kind of evidence normally required. This is not unprecedented, as the FDA often approves drugs based on surrogate endpoints, particularly in oncology. The history of cancer screening tests includes periods of time where tests were adopted before definitive mortality data existed, such as low-dose CT screening for lung cancer, which was broadly implemented in some settings before a mortality benefit was shown. But one of the critics' points is that past precedents of premature adoption don't justify current premature adoption.
What I Think
I agree with many of the criticisms above. The lack of mortality data is the biggest issue in my opinion, and the poor sensitivity for early-stage cancers, concerns about downstream costs, harms of false positives at the population scale, and commercialization of MCED tests ahead of the establishment of a definitive evidence base are reasonable things to feel uncomfortable about. However, I think the critics' overall framing has a blind spot, which is that the implicit comparator in most of these criticisms is perfection. The reality is that more than half of cancer deaths come from cancers with no recommended screening, more than half of all cancers are diagnosed at an advanced stage, and over 500,000 Americans die of cancer each year.
The critics are doing essential work by holding the MCED field to the evidentiary standards that cancer screening demands by forcing companies, regulators, and clinicians to confront the limitations of the technology honestly.