No CT scans for old smokers?


“Traditionally, doctors used to be called in when needed. But this is now changing. Increasingly it is the doctor who calls the person in by issuing an invitation. Healthy people are asked to visit the surgery for a ‘check-up,’ or ‘screening,’ when their computerized records show they are ‘due.’ Non-attendance is known as ‘non-compliance,’ indicating an element of recklessness and irresponsibility.”

– Petr Skrabanek, The Death of Humane Medicine and Rise of Coercive Healthism

If CMS endorses MEDCAC’s recommendations regarding low dose screening CT for lung cancer, we may see a coverage scenario that might be mistaken for an episode of Saturday Night Live.

It will be illegal for private insurance to charge a 64-year-old smoker screened for lung cancer even a $10 co-payment for CT. But a year later on his 65th birthday, when he finally enjoys the security of government from rapacious capitalists, he no longer will be covered for that screening CT.

What science juxtaposes unfettered eligibility and zero eligibility within a year? Does the risk of lung cancer from smoking miraculously cease once covered by Medicare?

This is a result of two trends at odds with one another. Uniformity of coverage and Balkanization of bureaucracy.

Result: inconsistency.

The USPSTF is satisfied that screening CT reduces mortality from lung cancer. The Medicare Evidence Development and Coverage Advisory Committee (MEDCAC, not to be confused with MedPAC) has concerns.

Arbitrariness versus arbitrariness

This is the confusion.

64-year-old smoker: Screening CT saves your life and is an essential medical benefit.

65 year-old smoker:  Overdiagnosis, anxiety from false positive CT and radiation-induced malignancy are far greater existential threats than that death from lung cancer that we’ve been warning you about.

Isn’t variation in health care because we don’t follow evidence-based medicine (EBM)?  But government advisors, who are among the brightest and most accomplished scholars, strictly follow science. Why opposite conclusions of the USPSTF and MEDCAC?

One reason is the lack of a normative frame — i.e., thresholds and ceilings for approval are neither explicitly stated not religiously followed.  Advisors have their pet peeves which they bring to the table. It’s a case of ‘me and my arbitrariness versus you and your arbitrariness.’

Screening is an information problem. If we know who will and who won’t have cancer we won’t have to screen. No screening test is perfect. You won’t get a Nobel Prize for finding faults with screening. But you will get the award for intellectual honesty if you fulfill the following:

  1. Explicitly state a threshold for rejecting a screening test (i.e., “numbers needed to treat (NNT) should not exceed 666”).
  2. Show how you reached the threshold. (Yes, I’m like your math teacher. I want to see the work, not just the answer.)
  3. Apply your reasoning uniformly to all screening tests (regardless if supported by the NFL).

Being concerned about false positives (FPs) in screening CT is like saying this night is terrible because it is dark. If you think there are too many false positives in screening for lung cancer tell us how many is too many. How did you arrive at that number? Is it more than screening for cervical, prostate, breast and colorectal cancer? If not, why not?

Two rights don’t make a wrong any more than two wrongs don’t make a right.

Anxiety from false positives

A thought experiment: You’re a heavy smoker; constantly reminded by media, healthism activists and public health folks that you’ll die an early and miserable death from disseminated lung cancer. You won’t be able to breathe easily because fluid surrounds your lungs; you will turn yellow from jaundice; bones will break easily.

To reduce chances of a miserable end you have a screening CT which shows “6 mm nodule, possible cancer, follow-up in 6 months.” After 6 months CT shows “no change, but follow-up in 6 months to be certain that benign.” After another CT, your doctor says, “It’s a false positive!”

What would you do with the news of false positive?

I suspect you’ll celebrate. Shout out from the roof tops, “I won’t have a miserable death from lung cancer. I have a false positive! Gotcha fate.”

Imagine if a somber professor intervened between you and the first CT scan and said “I can’t let you go on that. You are in danger. Grave danger. Of anxiety from a false positive.”

I suspect a smoker might have a better handle on irony here.

“After driving my anxiety off the scale that I’ll get cancer so that I quit smoking, now you are worried that I might be anxious for having the cancer that you’ve been harping on about all my smoking life!”

“Sir, every day I don’t die is a false positive.”

Thankfully, research confirms common sense. No, smokers aren’t anxious about false positive CTs, just as base jumpers don’t say “I’m really concerned that I’ll land safely when I think I’m going to die and not actually die.”

Numbers needed to treat/ screen

NNT is a good measure for efficacy of screening and is a more tolerable proxy for the unutterable term, “costs.”

How does screening CT for lung cancer do? Not bad, actually. At least, better than mammograms. 320 smokers need to be screened to save one life from lung cancer whereas 1,339 women 50 to 59 need to be screened to save one life from breast cancer.

CMS obviously is in no rush to stop coverage for mammography. Why are smokers different?

Smokers don’t induce the pink sentiment, even though lung cancer kills more women than breast cancer. They lack a lobby in Washington. Smoking the poster child of avoidable harm, smokers gets little sympathy.

Radiation and common sense

All evidence is not equal. Some evidence is more equal than others. Or, has a narrower bandwidth of uncertainty.

The risk of malignancy in an adult from low dose CT, which now barely produces more radiation than a year in Denver, is based on a model with many assumptions — i.e., it remains conjectural (lots of uncertainty that correct). Few would disagree that the risk of lung cancer attributable to smoking isn’t conjectural (far higher certainty that correct).

To be concerned about radiation from CT in a child is prudence.

To be so concerned about radiation-induced malignancy from CT in a 70-year-old, who already has a 20 fold increase in relative risk of lung cancer because of 30-pack year smoking, that one stops the smoker from being scanned to search for the malignancy, is parody.

More trials? Really?

The National Lung Screening Trial (NLST) randomized smokers to low dose CT and chest radiographs. It is a rigorous, multicenter study which found that low dose CT reduced mortality from lung cancer by 20 percent.

Critics say the NLST is impressive but it is only an n of 1: i.e. we need more evidence. Like an observational study? Or more RCTs?

If evidence from an RCT, the Brahmin of EBM, which enrolled over 50,000 patients (and cost $250 million) can’t be trusted who can we trust these days? And how many RCTs are enough? Precise answer to these questions is more difficult than academic hand waving.

Incidentally, I doubt many Institutional Review Boards will approve another RCT for CT lung cancer screening. Equipoise no longer exists.

I mean you can no longer say to a trial participant with a straight face, “We are not sure screening CT really saves lives.” Particularly when the NLST was stopped when researchers found the evidence of benefit of screening CT so compelling that randomization (50 percent chance of not receiving treatment with benefits) would be unethical.

Reflect on this for a moment. It is unethical for a researcher to give smokers 50 percent chance of not receiving screening CT which might save their lives. But it’s not unethical for CMS to give smokers no chance of screening CT which might save their lives.

Why do we need both USPSTF and MEDCAC?

Ostensibly, because Medicare recipients have different health care needs. What works for a thirty year old doesn’t necessarily apply to a nonagenarian. But there’s no reason that what is beneficial for a 62-year-old isn’t for a 68-year-old.

Here is an important point. In screening age and comorbidities matter. Benefits of screening diminish with age. MEDCAC could certainly have recommended an upper limit of age when coverage for lung (and breast and colorectal) cancer ceases. That is, for example, CMS will reimburse screening CT for a 68-year-old but not an 85-year-old smoker.

This requires political courage as accusations of death panels are never far off. Academics who advise the government can either shield politicians from addressing inconvenient truths for their electoral prospects or expose them to the rhetoric of democracy so that they are forced to address difficult issues. My preference is latter.

The upper age for screening CTs can be set by the common sense of radiologists. When we see a script for a 90-year-old for lung cancer screening, we should pick up the phone and say “no, this is wasteful” without fear of that hackneyed term “rationing.”

Failing that, the American College of Radiology and other specialist societies which have tirelessly lamented MEDCAC’s decision, ought to take the lead enforcing an age limit and demonstrating that they understand in which patients screening is futile.

Saving lives cost $$$

The most compelling argument against screening CT for lung cancer dare not be used: cost. This is estimated at a $9 billion over the first five years.

(That’s 9 billion. “B” as in “be very afraid.”)

But basing coverage on a threshold of costs per quality associated life years, as in Britain’s NHS, is anathema to American culture. Politically dangerous. Reinforced by the ethos of comparative effectiveness, as opposed to cost effectiveness, which has the noble but unsustainable ideal “Let them have cake … if cake is evidence-based.”

ACA and Screening

The ACA endorses prevention and wellness. The oft-quoted “an ounce of prevention is better than a pound of cure” runs through many of its statutes. There’s widespread belief that screening reduces net future medical costs for populations. There’s not an ounce of evidence to support this for many screening tests presently used.

Furthermore, patient adherence to screening, including highly contentious mammograms, is a quality metric for physicians. The rejection of screening CT for lung cancer would send a mixed message.

OK, short break for irony spotting. Think about this. We’re vigorously debating whether screening mammograms prolong life. But recommending mammograms is a quality metric. That’s like saying, “An ice axe won’t help in the hike. But you’re a terrible leader for not reminding us to bring it along.”

Upside down morality and downside up economics

Does this make moral sense?

A middle class family on a bronze plan making just over four times the federal poverty limit must pay for child’s antibiotics. But a healthy CEO of a large non-profit hospital will receive a screening colonoscopy without even incurring a co-payment. His physician will tick quality boxes; one for reminding him to get screened and one for his kind compliance.

Got that? Sick (and middle class): cost share. Well (and 1 percent): no cost sharing.

That’s an own goal for equality.

This upside down economics seem nonsensical until you blink. There is price elasticity in screening, meaning the well are likely to forego the test at even modest copayments. Thus, the ACA forbids cost sharing for screening tests approved by USPSTF.

Screening is a gift that keeps on giving. Yes, ACA has dug its own cost grave here.

If the cake is too expensive say so, and we’ll stick to bread

Blink again and screening is an intrusion in to the lives of the healthy, an inversion of the doctor-patient relationship that Skrabanek observed.

Like Skrabanek, I believe the Hippocratic Oath is a contract to heal the ill not hound the well. Were it up to me, screening tests (barring very few) would be the first to drop out of essential medical benefits of private insurance, to be released to the vicissitudes of the market beside caviar and flight insurance. Assuming money is no object, screening CT for lung cancer would be one of the last to go though. At least not until the much coveted annual physical examination of highly dubious value is still covered and men are still rendered incontinent from prostatectomy and live no longer as a result.

Foolish consistency is the hobgoblin of little minds. But inconsistency undermines public trust in government institutions. For logical consistency that’s syntonic with the rhetoric of prevention, an exaggerated optimism no doubt, CMS should approve low dose screening CTs for lung cancer.

But if costs are the issue, CMS should be bold enough to say so, rather than cloak the rejection in the pretense of science. I, for one, will respect them for that.

Saurabh Jha is a radiologist and can be reached on Twitter @RogueRad.  

Image credit:

View 3 Comments >

Most Popular

Get KevinMD's 5 most popular stories.