The Dangers of Excessive Medical Tests

Otis: Don Berwick said 30% of medical expenditures in the United States is waste or fraud. That is 30% of what doctors order for patients are things that we don't need to order.

Brent: Welcome to Death Clock. I am your host, Brent Franson. Today, we speak with Doctor Otis Brawley. He's a professor of oncology and the 39th Bloomberg Distinguished Professor at Johns Hopkins. He has served as chief medical and scientific officer at the American Cancer Society, held leadership roles at the National Cancer Institute, and is also a professor at Emory University. Doctor Brawley's work has earned him numerous awards, including the Martin D. Abeloff Award for Excellence in Public Health and Cancer Control.

Brent: Doctor Brawley is an awesome guest, as you'll hear. He's got a deep knowledge of the statistics around screenings, and he's been a vocal advocate about the harms of overscreening, which I think is really interesting because it's so counterintuitive. We tend to believe that the more we screen, the better. So he's a really good resource to walk us through all of the nuances, the harms, and the benefits of screenings so we can make the right decisions for ourselves.

Brent: He's a very impressive guest. I really enjoyed speaking with him, and I hope you enjoy.

Brent: Doctor Otis Brawley, welcome to the show. Hi.

Otis: Thanks for having me.

Brent: Before we jump in, will you just give us a sense of your potted bio?

Otis: I'm a medical oncologist. I also trained as an epidemiologist, and that's actually probably more important because screening is a subspecialty of epidemiology. I went to the University of Chicago for college and medical school and did my internal medicine residency at Case Western Reserve University. I trained in both medical oncology and epidemiology at the National Cancer Institute and ran a research group there looking at health disparities and cancer outcomes, a lot of which is what we're going to talk about today—the outcomes of screening.

Brent: Why did you get into oncology and epidemiology? What motivated you?

Otis: From an early age, I was always interested in two things: science and policy. There is a great deal of policy in the science of oncology and especially in the science of screening. I actually spent 12 years as the chief medical officer of the American Cancer Society because of that interest in policy and screening—what screening tests should be recommended and what should be paid for.

Otis: What are the problems in people accessing that care that they may or may not need? Quite honestly, there's a lot of waste in the screening world right now.

Brent: So let's talk about that. Let's talk first about the standard screening recommendations. Let's use colon cancer screenings as a starting point. For this screening, we used to recommend starting at 50, but we've dropped that recommendation to 45. There's a different set of recommendations for people who have a direct family member who's had colon cancer.

Brent: I have a friend of a friend who passed away recently at 38 from colon cancer, finding out about it just two weeks before it took his life. Why aren't we screening for colon cancer at 30?

Otis: Several reasons. Number one: whenever we do a medical intervention, there is a risk of harm. When we talk about colonoscopy, one out of 600 people getting a colonoscopy suffers a perforated colon. Number two: we have a shortage of capacity to screen. For example, I live in Maryland, and we do not have enough GI doctors to do a colonoscopy every ten years on everyone over the age of 45; they just don't exist.

Otis: If you go to a number of states in the United States, those specialists just don't exist. We could screen everybody every ten years starting at age 20 in New York City, but not across New York State. Another point is the science: we only recommend a screening test when we have good science showing that it saves lives.

Otis: We have several very good studies in colorectal cancer showing that stool blood testing, stool DNA, or colonoscopy all save lives. Stool blood testing every year, stool DNA every three years, or looking in the colon with a colonoscopy every ten years reduces the risk of death by 30 to 35% for people over the age of 50.

Otis: That's an important point; this is a glass-half-full, glass-half-empty situation. We decided to drop the recommended age from 50 to 45, realizing that's a stretch. If you read the recommendation document—and I'm one of the coauthors from the American Cancer Society—it states that this is a projection that screening would be useful for people in their late 40s.

Otis: There are arguments as to why we think it's a reasonable projection, but again, it's a projection. We have solid data showing that if people aged 50 and above get any of those three screening tests, they reduce their risk of death by 30%. The reason this is important is that if I have a thousand people destined to die from colorectal cancer and I screen them all regularly, 700 of the thousand will still die from colorectal cancer, while our screening test, with its 30% relative risk reduction, will prevent 300 of those 1,000 deaths.

Otis: A lot of people think screening is far better than it actually is and that every person who dies represents a failure of screening. But for colorectal cancer—which is one of the best screening tests we have—a 30% relative risk reduction is what clinical studies have documented.

Brent: So you're saying if we waved a magic wand and had a thousand people in a lab following our instructions exactly, regularly screening them for colon cancer—and assuming we knew all thousand would eventually develop colon cancer...

Brent: Even if we catch it early, 700 of the thousand would still die of colon cancer?

Otis: That's correct. For a thousand people destined to die from colon cancer, we have the power to prevent 30% of those deaths.

Brent: The power to prevent 30% from dying. Maybe that is a common misconception with colon cancer. My understanding was that catching it early yields a high survival rate. Are we extending the lives of those 700 in this scenario?

Otis: Yes. Clinical studies are specifically designed to answer whether we prevent patients from dying from the disease.

Brent: How should we think about this at a cohort and policy level versus an individual level? Take me—I'm 42. A 1 in 600 risk of a perforated colon feels acceptable to me individually. The shortage of doctors is a societal issue, but individually, if I can access a doctor, I want to take advantage of it. Why shouldn't I screen early? What is the argument at the individual level?

Brent: That's kind of a society problem. But for me individually, I don't care if I can get access to a doc, I'm going to be I'm going to be glad to have it. Why in in that case, am I still not screening early? What is the argument at the individual level as opposed to the cohort based math?

Otis: A lot of people get screened early, wasting time and money fixating on something they shouldn't be focused on. I very frequently see people—this happens with prostate cancer and breast cancer especially—who get fixated on one specific disease.

Otis: They don't realize that the odds of being negatively affected by that particular disease are actually pretty small. People will see that a movie star died of colon cancer at age 41 after a diagnosis at 37, so they run out, spend a lot of money, and find a doctor willing to perform colorectal screening. In reality, their risks of dying from heart disease, lung cancer, diabetes, or stroke are much higher.

Otis: I try to encourage people not to get fixated on one specific condition. In my field, we frequently see patients who insist on lung cancer screening whose lungs turn out fine, but who actually have stage three breast cancer that was overlooked.

Otis: They never got mammography, colon cancer, or cervical cancer screenings because they were fixated on lung cancer. We see that often. Furthermore, when you perform extensive radiologic imaging on an average 30-year-old with rectal bleeding, for instance...

Otis: The most likely cause is hemorrhoids. If they insist on getting a CT scan, data shows that doing enough CT scans on 30-year-olds causes cancers in their 50s or 60s. That 30-year-old is entering a lottery.

Otis: They will probably get the CT scan and never get cancer from it, but as population scientists, we worry about causing cancer through the overuse of radiation. One-half of one percent of all cancers in the United States today are caused by medical radiation.

Brent: This perspective is new and compelling to me. Individually, discovering colon cancer late is the worst outcome, so I don't mind rolling the dice regarding colon perforation from a colonoscopy. But the most interesting point you raised is that people take their eyes off the ball regarding things that really matter—sleep, diet, exercise, and heart disease—which are far more likely to shorten life.

Brent: In the case of colon cancer, with the perforation of the colon from the from the colonoscopy. But what I do think is a really interesting point is the you taking your eye off the ball in terms of the things that really matter sleep, diet, exercise, you know, heart disease, like these things that are going to be much more likely to shorten the length of your life.

Brent: So people over-obsess over something that might matter for a small group, but statistically matters less for them than foundational health factors they ignore.

Otis: You're absolutely correct. I worry about people taking their eyes off risk reduction like good diet and physical activity. We should all eat 5 to 9 servings of fruits and vegetables daily, get at least 30 minutes of moderate exercise four times a week, and maintain a healthy body weight. Unfortunately, those messages get lost among people fixated on obtaining specific tests.

Otis: Screening means testing an asymptomatic person who is at risk for a disease. If someone has symptoms—abdominal complaints, blood in the stool, or changes in stool caliber—they ought to be assessed, and we may perform some of these tests.

Otis: At that point, those are diagnostic tests, not screening tests. Insurance reimburses differently for a diagnostic colonoscopy versus a screening colonoscopy. People in their 20s, 30s, or early 40s with symptoms should receive diagnostic testing. We also evaluate family history: if someone has a strong family history of colon cancer, we recommend they start screening—usually with a colonoscopy—ten years before the age at which their relative was diagnosed.

Brent: On the policy side, when writing these recommendations at the American Cancer Society, what is the North Star? There seems to be tension between what might be best individually regardless of economic or system limits versus guidelines that reflect what is realistic across socioeconomic groups and healthcare infrastructure.

Brent: What is the primary guiding principle when formulating these recommendations?

Otis: The first North Star is: do we have evidence that this screening test saves lives? Answering that usually requires a prospective randomized clinical trial that can take 10 to 15 years. We've lost that North Star a few times. With prostate cancer, we did extensive screening from the 1990s through 2010 without studies proving that it actually saved lives.

Otis: We misled ourselves into thinking we had data showing it saved lives when we didn't. That's an example where we lost our North Star, advocated widespread screening, and ended up treating many men who had prostate cancer but did not require treatment.

Otis: Unnecessary treatment can cause impotence, incontinence, and even death. So our first North Star must be whether a screening test saves lives. Our next North Star should be how to save as many lives as possible through effective policy.

Otis: In discussions about rising colorectal cancer rates in 40-year-olds, people forget that 80% of colorectal cancer deaths occur in individuals diagnosed after age 50. We also overlook that only 60% of people over age 50 receive appropriate screening today.

Otis: Flipping that around, 40% of people over 50 are not getting properly screened. Mathematically, increasing screening rates from 60% to 90% among those 50 and older saves significantly more lives than expanding screening to all 40-year-olds, which would be far more expensive.

Otis: Screening everyone aged 40 to 49 costs more, whereas a 30% increase in screening among people over 50 yields greater life savings.

Brent: Focusing on increasing screening percentages in higher-risk groups saves more lives than expanding recommendations to lower-risk populations.

Otis: That's right.

Brent: There's mistrust of the healthcare system. How do considerations like insurance coverage and doctor availability influence this process? It would be concerning if the top priority was saving the most lives, yet recommendations were limited by system capacity.

Brent: I imagine committee members asking whether broad recommendations are actually realistic.

Otis: I can be critical of insurance companies, but I will defend them here. Don Berwick, who ran CMS for Barack Obama, stated that 30% of US medical expenditures go toward waste or fraud.

Otis: That means 30% of what doctors order is unnecessary. When insurers require justification before approving tests, it is reasonable given that unnecessary procedures are frequently ordered in American medicine.

Otis: A major problem in American medicine is that doctors and patients often fail to let science guide practice. I wrote a book on how unnecessary medical tests cause harm and kill people.

Otis: We frequently practice based on desire rather than evidence. I see patients requesting a $10,000-a-month drug over a $300 option proven more effective in trials simply because they assume higher cost equals better quality.

Brent: Is this 30% waste driven by clinician financial incentives, lack of competence, or a combination of factors?

Otis: It's a combination. In oncology, practices buy drugs wholesale and sell retail. Sometimes it's financial motivation, sometimes clinical laziness or incompetence, and sometimes it's placating patients who request advertised drugs or tests to retain their business.

Otis: We often fail to emphasize prevention. Comparing countries is illustrative: Costa Rica's life expectancy is two years higher than the United States'.

Otis: The US spends $10,000 to $12,000 per person annually on healthcare, whereas Costa Rica spends around $1,000. Costa Rica focuses heavily on risk reduction and early prevention rather than expensive, high-tech imaging and treatments that are often no better than cheaper alternatives.

Otis: In America, we demand new technologies simply because they are newer or more expensive.

Brent: That segues into your work regarding the downsides of overscreening. Most people assume more screening is strictly beneficial.

Brent: What is your perspective on full-body scans like Prenuvo?

Otis: Full-body scans are unproven and represent a significant waste of money. They trigger follow-up scans involving unnecessary radiation. For context, 15% of chest CT scans reveal incidental abnormalities unrelated to the original reason for the scan.

Otis: Those incidental findings represent normal human variation, yet they lead to additional invasive studies. I am similarly concerned about 3D mammography, which gained rapid adoption over the last decade. It was FDA-approved based on equivalence to 2D mammography, not superiority.

Otis: Women in some states pay extra out-of-pocket for 3D mammograms, while in other states, manufacturers lobbied legislatures to mandate insurance coverage for 3D over 2D mammography.

Otis: Meanwhile, the National Cancer Institute is currently running a massive trial to determine if 3D mammography is actually superior to 2D. State legislatures mandated coverage before scientists answered the question. 3D mammography may identify harmless lesions that lead to unnecessary follow-up, making it potentially inferior to 2D mammography.

Otis: That's an example of some waste that we have right now.

Brent: How do you view the research lag? If full-body scans eventually prove beneficial, there will be a period where we suspect their utility before clinical trials confirm it.

Brent: Early adopters might pay out-of-pocket hoping for a benefit that gets validated later. How do we balance that adoption lag?

Brent: And so I'm going to be early on the adoption curve for that. And then, you know, in this case, that ends up being justified later. How do we think about that lag. And when you get the findings, you're right.

Otis: If an individual wants to spend their own money on an unproven scan, that is fine, but insurance shouldn't be forced to cover it before evidence exists. History provides important lessons here.

Otis: In the 1950s and 60s, medicine assumed chest X-rays saved lives by detecting lung cancer early. Widespread X-ray screening was implemented after Luther Terry's 1964 Surgeon General report linked smoking to lung cancer.

Otis: When the Mayo Clinic eventually conducted a prospective randomized trial comparing X-ray screening to no screening, the annual mortality rate from lung cancer and associated diagnostic procedures was 3.2 per thousand in the screened group versus 2.8 per thousand in the control group.

Otis: Unscreened individuals were actually less likely to die than those who received screening.

Brent: Why did the screened group experience higher mortality?

Otis: Complications from diagnostic procedures caused the deaths. Biopsying a suspicious finding can collapse a smoker's lung or trigger a fatal heart attack, and bronchoscopies can cause oxygen deprivation.

Otis: Low-dose spiral CT screening for lung cancer has proven beneficial, but consider the trial data: 54,000 smokers were randomized to spiral CT or control. Over ten years, there were 437 lung cancer deaths in the control arm versus 350 in the CT arm.

Otis: That represents an 87-life difference, or a 20% mortality reduction. However, 16 people in the screened arm died from diagnostic procedures triggered by abnormal CT scans.

Otis: Autopsies revealed that six of those 16 individuals did not actually have lung cancer; they died from collapsed lungs or bronchoscopy complications. The net ratio was 5.4 lives saved for every life lost to diagnostics.

Otis: There were 5.4 lives saved for every two patients admitted to an intensive care unit and every one life lost. Because of these trade-offs, Medicare requires a counseling session where clinicians educate patients on the risks and benefits before covering spiral CT screening.

Brent: That trial illustrates tragedies on both sides: people in the control group who would have benefited from screening, and screened individuals killed by diagnostic complications despite not having cancer.

Brent: They have some risk. They don't have lung cancer, but the screening kills them.

Otis: That's right. The diagnostics caused by the screening. Yes.

Brent: Regarding mammograms and incidental findings, an intuitive reaction is that finding incidental abnormalities is helpful. However, investigating random findings often opens Pandora's box, leading to invasive procedures that cause more harm than good.

Otis: Oh, good, we.

Brent: Wouldn't have found it otherwise. But you presented in a way where it's where that's where it's not good. I don't think you view it the same way. And it seems that this is the reason, which is, wait, if we're opening up Pandora's box, then we're sticking needles into you and we're going to do more harm than good, for lack of a better term, sticking needles into random abnormalities that we see on screens that are meant to find something else.

Otis: The issue is pronounced in women in their 40s. The median age of breast cancer diagnosis is 63, and the median age of death is in the early 70s. Mammograms are much harder to interpret in 40-year-olds than in 60- or 70-year-olds.

Otis: Women who begin screening in their 40s often experience multiple false alarms and call-backs, which causes immense anxiety. Consequently, many stop getting mammograms altogether.

Otis: When they reach their 50s, 60s, and 70s—when mammography is most effective at saving lives—they are absent from screening due to negative past experiences in their 40s.

Brent: To summarize: screening at age 30 against general recommendations carries the risk of a perforated colon or severe injury that outweighs the low probability of finding early-stage cancer.

Brent: For an average-risk individual, early screening presents a higher likelihood of harm than benefit.

Otis: For normal-risk individuals in their 20s, the potential harm outweighs the benefit. Higher-risk individuals—such as those with ulcerative colitis or strong family histories—warrant early screening, and insurance should cover them.

Otis: The trade-offs become a gray area in one's 30s and 40s. If an average-risk individual without a family history wants a colonoscopy and pays $2,000 out-of-pocket, that is their choice, but insurance shouldn't be mandated to cover it.

Otis: Insurers generally do not cover unindicated early screenings, and patients who demand them are often the first to complain when overall premium costs rise.

Brent: Does this framework apply to emerging tools like multi-cancer liquid biopsies (e.g., Galleri)? Should patients stick strictly to established guidelines from bodies like the American Cancer Society or the US Preventive Services Task Force?

Brent: Would you apply the same framework there? Hey, look, we might find something that we don't know what to do with. And so we're going to have to do more screening. And there there can be negative consequences from screening. And so like kind of stick to the recommendations of the American Cancer Society or the US Preventive Services Task Force.

Brent: And you're going to be good at this juncture.

Otis: I am hopeful about many of these tests. Several companies are conducting proper clinical trials to establish their true value. Within five years, data will clarify whether these tests offer genuine clinical benefit.

Otis: So I'm hopeful about many of these tests.

Brent: How do healthcare disparities impact screening access and outcomes?

Otis: When low-risk individuals consume screening capacity, high-risk or underserved populations lose access. Overuse directly exacerbates disparities. I worked with a hospital that had four CT scanners and a two-week waitlist for symptomatic patients needing diagnostic scans.

Otis: The cancer center director chose not to offer elective lung cancer screening to avoid delaying diagnostic scans for patients with known disease, explicitly managing resources to mitigate health disparities.

Otis: Numerous studies confirm that racial, ethnic, socioeconomic, and geographic disparities lead to lower screening rates, delayed follow-up diagnostics, and reduced treatment access.

Otis: In a study conducted by one of my fellows in Atlanta, 7% of Black women and 3% of white women diagnosed with localized, curable breast cancer received no treatment within the first year following diagnosis.

Otis: 7% of the black women and 3% of the white women in metropolitan Atlanta who got the screening message and got diagnosed, got biopsy proven diagnosis of a localized curable breast cancer. Got no treatment in the first year after the diagnosis.

Brent: Where does the breakdown occur? Is it a lack of available doctors, educational barriers, or patient non-utilization?

Otis: It's a combination of factors. In our Atlanta study, patients had sufficient healthcare access to receive initial screening and biopsies, but dropped out before receiving treatment. Studies point to transportation barriers, childcare obligations, and fear.

Otis: Patient navigation and education help address these issues, but our healthcare system lacks systematic infrastructure to ensure patients remain in care once diagnosed.

Otis: Social workers assist when patients raise concerns, but if a patient simply stops showing up, few systems systematically follow up to bring them back for treatment.

Brent: Are there specific factors impacting minority populations independent of socioeconomic status, or is socioeconomic status the primary driver?

Otis: Socioeconomic status and educational attainment outweigh race. Women with a college degree have a US cancer mortality rate of 60 per 100,000 annually. For women who did not complete college or dropped out of high school, the cancer mortality rate doubles to 120 per 100,000.

Brent: Regardless of race?

Otis: Yes, education and socioeconomics trump race. However, Black and Hispanic populations face disproportionate rates of lower educational attainment, lack of insurance, and reduced access to early preventive care during childhood.

Otis: Energy imbalance—high-calorie diets paired with physical inactivity—is the second leading cause of cancer overall and the leading cause for Black and Hispanic women. Tobacco remains the top cause for other groups, but energy imbalance will be the primary cause of cancer for all Americans by 2030.

Otis: The average age at which people begin smoking in the United States is 15.

Brent: Has the decline in smoking rates and associated lung cancer deaths resulted primarily from older smokers passing away or from active smoking cessation?

Brent: Or is it just the generation of smokers? Has died and they're we aren't creating new smokers.

Otis: Tobacco causes 18 types of cancer. When smokers quit and remain cancer-free for ten years, their risk drops dramatically. Since Surgeon General Luther Terry's 1964 report, older generations of smokers passed away and were replaced by generations with lower smoking rates.

Otis: The proportion of teenagers starting to smoke has steadily declined. Because tobacco is highly addictive, quitting is difficult once started.

Otis: In 1955, 55% of American men smoked; in 1965, 35% of American women smoked. Today, overall smoking rates are under 15%. Cessation programs helped, but the primary factor in reduced smoking rates is that older smokers passed away without being replaced by new ones.

Brent: I've really enjoyed this conversation. Where can listeners find more information about your work?

Otis: For general scientific information, the American Cancer Society offers resources at cancer.org. For screening science and interpretation, the US Preventive Services Task Force is the best source. Be cautious of organizations with corporate influences or financial ties to screening manufacturers that overly promote specific screening tests.

Otis: Certain organizations advocating aggressively for prostate or breast cancer screenings receive funding from companies that profit from those interventions.

Brent: Doctor Otis Brawley, thank you so much for joining us and for your important work.

Otis: Thank you.

Brent: Death Clock is recorded in Boulder, Colorado and San Francisco, California. Produced by Patrick Gudino, with music by Patrick Lee, and hosted by yours truly, Brent Franson, founder and CEO of Death Clock.