“Overdiagnosis is a myth that has been created by a handful of individuals who provide no care for women with breast cancer.”

This quote by Dr. Daniel Kopans encapsulates a long-standing and intense debate in the medical community regarding cancer screening.

In the Dec. 9, 2025, issue of the New York Times, Gina Kolata quoted H. Gilbert Welch, MD, on his perspective that “treatments need to be balanced with the consequences of finding a cancer that did not need to be found.” Based largely upon a recent article in JAMA Internal Medicine, entitled “The Rise in Early-Onset Cancer in the US Population: More Apparent Than Real,” the Times reported on a rise of eight cancers occurring in people under age 50. The article noted an increase in incidence with no corresponding increase in mortality.

The core concept: overdiagnosis

The underlying concept being discussed is “overdiagnosis” (OD), which refers to the diagnosis of a disease that will never cause symptoms or lead to death during a patient’s lifetime. Accordingly, when cancer is detected in an asymptomatic person, it is sometimes uncertain whether the growth is relentlessly aggressive or biologically harmless.

The Times quoted Welch as stating that many cancers “go away on their own. Others stop growing or pose no risk; they cause no symptoms and do not spread.”

However, identifying which specific cancers will regress or remain dormant is currently not possible in clinical practice. The authors of the JAMA article acknowledge this limitation, noting that “it is impossible to know if someone’s cancer will be deadly or not.”

As a response, the authors propose a shift in perspective: “Our findings highlight the need for a more nuanced approach to early detection.”

The debate over population screening

This perspective is part of Welch’s long-standing research into the potential harms of population-wide cancer screening. He has argued for decades that when small lesions are found inadvertently, the subsequent treatments may do more harm than good.

Welch has been a leading researcher on the topic of overdiagnosis for more than 25 years, publishing more than 30 peer-reviewed articles. He has questioned the efficacy of widespread screening for esophageal adenocarcinoma, melanoma, breast, prostate, thyroid, and lung cancers. In 2011, he authored the book Overdiagnosed: Making People Sick in the Pursuit of Health. His work has helped initiate an international movement examining the risks of screening, including a series of “Preventing Overdiagnosis” conferences.

His core argument relies on “epidemiologic signatures.” He posits that if screening shows an increased incidence of cancer but a stable stage distribution and mortality rate, it is strong evidence of overdiagnosis. He has previously estimated overdiagnosis rates for cancer screening at 25 percent for breast, 50 percent for lung, and 60 percent for prostate. In 2007, Welch and other physicians theorized that as many as 80 percent of lung cancers diagnosed by CT screening might be overdiagnosed. For context, the subsequent National Lung Screen Trial found only a 1 percent excess of lung cancers in the CT screening arm.

Evaluating incidence and mortality data

Critics of the overdiagnosis theory argue that long-term data does not support these high estimates. Following a peak incidence in 1995, both all-cancer incidence and mortality have declined substantially in the United States.

This decline is evident for the five cancers where population screening is most commonly recommended:

• Cervical
• Breast
• Prostate
• Colon
• Lung

A steady and progressive decline in mortality for each of these cancers has occurred between 1990 and the present. Furthermore, cancer screening has resulted in a well-documented shift toward lower-stage disease at the time of diagnosis for these cancers.

Proponents of screening offer an alternative explanation for the data. Since screening is specifically designed to diagnose cancers today that would otherwise remain undetected until years later, incidence rates will naturally increase with the initiation of population screening. With rising screening participation and the successful treatment of early-stage disease, a pattern of rising incidence and falling mortality is the expected outcome of an effective program.

Additionally, modern algorithms now allow for the identification of slowly growing cancers, for example, in the thyroid, kidney, lung, and prostate, and incorporate active surveillance options to minimize potential overtreatment.

Media representation and public health

The debate extends beyond medical journals into the public sphere. Clinicians have expressed concern that an outsized media focus on overdiagnosis may foster unreasonable fear of cancer screening and contribute to reduced participation rates. They argue that minimizing the benefits of screening can lead to delayed diagnoses and preventable mortality.

Some medical professionals argue that journalists have a responsibility to update their reporting when earlier, high estimates of overdiagnosis are challenged by newer trial data. The New York Times and Kolata have faced specific criticism from the medical community for frequently relying on Welch as a primary source, which some argue presents a skewed view of the consensus on screening.

Media observers have noted this dynamic in the past:

“… because the media, blindly following the lead of Gina Kolata of The New York Times, have largely taken a one-sided viewpoint on the issue of screening: against. They are finely attuned and devote column space to the findings of any study, however weak, that appears to downgrade the value of screening, while ignoring the preponderance of studies that demonstrate positive results.”

As cancer diagnostics continue to evolve, the challenge for both medicine and the media remains how to accurately convey the complexities of screening, balancing the real risk of overdiagnosis against the proven life-saving benefits of early detection.

Frederic W. Grannis, Jr. is a thoracic surgeon.