Sunday, November 7, 2010

Can CAT Scans Prevent Lung Cancer? Smoke and Mirrors

I still marvel at the accomplishments of CAT scans and imaging studies.  These technologies have revolutionized the medical profession.  Imagine foretelling 50 years ago that a new technology would emerge that could perform ‘exploratory surgery’ without an incision.   This prediction would be greeted with disbelief, if not scoffing.  Of course, there would be similar reactions if decades ago other technological feats, such as fax machines, personal computers, emails and internet, were predicted.
CAT scans save patients’ lives, prevent surgeries, make accurate diagnoses and refine treatment plans.  This is the good news.  However, as expressed repeatedly on this blog, there is a darker side of the story.
  • Without doubt, too many CAT scans are performed.
  • Accumulated medical radiation has health consequences.
  • Resources expended on unnecessary CAT scans could be devoted to worthy health care endeavors
  • Radiologists suffer from litigophobia and identify every tiny abnormality, which generates patient anxiety and more medical tests chasing these trivial results.   Off the record, radiologists will confide that these ‘abnormalities’ are inconsequential, but their formal dictations always suggest that these lesions may be ominous.  Their dictated recommendation?  Mores scans!
  • CAT scans can provide patients with false security, which is the premise behind total body scans that are scamming the public.
  • CAT scanning and other technologies have eroded physicians’ bedside physical examination skills.   Does anyone think we are as skilled using a stethoscope as our predecessors were?  Of course, one can argue that we don’t really need these antiquated skills anymore. 

Recently, the press went ga ga over a new study that concluded that screening CAT scans on cigarette smokers could save lives. They hyped headline of The New York Times was CT Scans Cut Lung Cancer Deaths.   The study concluded that annual chest CAT scans on smokers could reduce the risk of death by lung cancer by 20%.   Why is my reaction so cynical to what was heralded a groundbreaking development in oncology? 
First, I am always wary when study results are presented in relative terms, such as a 20% decrease in the number of deaths.  Relative percentage terms always exaggerate the clinical benefits and are misunderstood by the public, as well as the press.  For example, assume that a new cholesterol-lowering medication is prescribed to 500 people who are compared with a control group of untreated individuals to determine if the drug can prevent heart attacks.    If 4 folks in the medication group develop a heart attack, and 5 control patients do also, then the drug company can correctly claim it lowers heart attack rates by 20%.  This sounds impressive, although the true benefit that an individual realizes is trivial, since 99% of all patients suffered no cardiac event.   Can’t you just see the headline Drug Cuts Heart Attack Rate by 20%?
In the chest CAT scan study, 300 people must be screened to save one life.  Is this worth it?  Of course, if the lucky individual is in your family or mine, then any cost would justify the outcome.  However, we cannot make public policy based on anecdotes or rare favorable outcomes.   For those who argue that saving lives is worth enormous expense at the expense of others, do they also support the following positions which would save lives?
  • Lowering the highway speed limit to 40 mph
  • Outlawing air travel
  • Prohibiting swimming
  • Eliminating skiing
  • Forbidding contact sports
Assuming the CAT scan study’s conclusions are correct, 299 smokers have to be screened to save a fellow inhaler.  Twenty-five percent of the scans had ‘abnormalities’ discovered that led patients into the medical labyrinth described above.  So, in the group of 300, 1 life was saved and 75 were targeted for subsequent medical assault.  When you consider that there are nearly 50 million smokers in America, imagine the billions of dollars that will be spent and the hundreds of thousands of patients who will be subjected to medical care evaluating harmless lesions.   
In addition, as doctors who have been around a while know, one study shouldn’t change the course of medical practice.  Over the past decade or two, many other groundbreaking studies were subsequently refuted.   I’m sure that physician readers can cite many examples of these medical retreats.  In six months’ time, we could learn of new study that concludes that smokers who are screened with CAT scans have a higher death rate.
A better strategy that is considerably cheaper and more effective is to try to reduce the number of cigarette smokers.  The CAT screening study cost $250 million.  What if those funds were used instead to treat nicotine addiction?  Would this have saved more lives than annual CAT scan screening?  The low tech approach, at least, is more focused on low hanging fruit, rather than aiming buckshot style at every smoker.  Moreover, I would imagine that if a smoker is told that his annual CAT scan is negative, that he will have less incentive to consider quitting.  For him, a negative scan may reinforce his belief that he will avoid a malignant fate.
Other bloggers, including KevinMD and Gary Schwitzer didn’t drink the CAT scan Kool Aide.  My advice?  When the press is serving up Kool Aide, hire a food taster.
Do I think there may a reasonable role for CAT scans in smokers?  I’m not holding my breath on this one. 

17 comments:

A. Bailey said...

Yes, I'm quibbling, but no one asserts that CAT's can prevent lung cancer. Some assert that they can prevent, or reduce, lung cancer DEATHS.

The best prevention is, of course, smoking cessation, but that is a different linguistic matter.

Marilyn Mann said...

You make some very good points. Another thing that occurs to me is that this trial enrolled a select group of people who were at very high risk for developing lung cancer. If I recall correctly, they had to have smoked for 30 pack years, which is one pack a day for 30 years or 3 packs a day for 10 years, and so forth. What tends to happen in medical practice is that many doctors see results like this and will recommend that their patients start undergoing CT scans, even if they are not as high risk. Say, someone who has smoked for 20 pack years. In a person at lower risk, the benefit will be smaller, the cost will be the same, and the harm may be just as large, or nearly as large.

Anonymous said...

Each cat scan will require prior auth from most insurance plans. Whenever a patient with hematuria gets a ct of the abd/pelvis and the lower lung fields yield a 2-3 mm nodule, the recommendation becomes do a CT q 3 months for a year, then q 4-6 months the next year, and then annually to demonstrate stability.

Glenn Laffel, MD, PhD said...

Excellent post Michael, and I agree with your position. It's appalling but not altogether surprising that a study like this--which can trigger billions of dollars in unnecessary testing--can receive so much attention in the press when the study has yet to be published in a peer-reviewed journal.

What's a poor doctor to do when her next patient, a smoker, comes into her office with a copy of this article?

Anonymous said...

First of all, its been 'CT scan' for a few decades. It would help if the author of the article new what he was talking about.

Michael Kirsch, M.D. said...

@A.Bailey, point acknowledged
@Marilyn, agree with your fear of CAT scan mission creep.
@anonymous #1, welcome to my world
@Glenn, suggest you shred the article and discuss her smoking habit
@anonymous #2, is that all you could come up with?

B. RAD, M.D. said...

Dr. Kirsch

It's surprising to me that this CT screening is big news now. People have short memories. About a decade ago, CT screening, including whole body, was making headlines and imaging centers were popping up all over the place. Not so much anymore. You make excellent points many of which I agree. I can speak from the other side, being a Rad who's practiced for a while. Radiologists, if they see it, they most likely will report it. As I have seen more than one little incidental "ditsel" on a chest CT turn into a cancer over time, I don't know what size nodule should be ignored or reported. Management and follow up, of course should depend on the patient's risk for developing cancer.
Adverse effects of radiation exposure, I believe will become a big issue in years to come, which may bring on litigation such as we have seen in the tobacco industry. This, I realize, is another issue, but who should be the gatekeeper for this? Anyone working around xrays has to monitor his or her exposure. But the patient? Not that I,m aware of. Is 16 CT scans of the abdomen and pelvis in the last 23 month (real patient with pancreatitis) too much?

Michael Kirsch, M.D. said...

@B.Rad, I know you're a radiologist, but you're surname isn't really Rad, is it?

B. RAD M.D. said...

No it's not.

Keep up the nice writing.

Juliet Kottak Mavromatis, MD said...

Dr. Whistleblower Kirsh,
Interesting post--thanks. I've been thinking a lot about the politics and industry of screening. Just wrote a post on KevinMD on the prospect of using the new biomarker test to screen for preclinical Alzheimer's:
http://www.kevinmd.com/blog/2010/11/questions-early-diagnosis-alzheimers-disease.html
One of the respondents to my blog appears to be highly invested in an industry that has a lot to gain from detecting Alzheimer's Disease early--think of all of those MRIs, PET's and neuropsych tests that might be done as a consequence, all with lack of a clear therapeutic that impacts clinical outcome. Now, think about lung CTs in smokers...actually, I was fairly impressed with the data that I read in the NY Times journal of medicine.

Michael Kirsch, M.D. said...

@Juliet, welcome to the blog! I was less impressed with the data in the New York Times Annals of Medicne. Seems like the benefits are slim, and the pitfalls and expense are substantial. I look forward to reviewing your Alzheimer's post.

Anonymous said...

Is the difference between this report and the report regarding women and mammograms from a year ago that those who contract lung cancer "deserve" it and those who contract breast cancer do not? Seems like I recall that even more women had to be screened and actually overtreated to save one life. And so many concerned parties came out of the woodwork on that one - the outcry was particularly loud from those who stood to lose substantial revenue if the new guidelines were embraced.

Big difference in the response from the medical community on this issue than the mammogram issue, and yet the scenarios seem really similar. It seems that patients, in order to deserve compassion, must provide evidence of having lived a spotless life.

Marya Zilberberg said...

Right on! Interesting point in comment above about the difference between mammo and this study -- value-laden, is it not? Especially since the number needed to invite to save a cancer death for CT (~300) is nearly an order of magnitude lower than for mammo in 40-49 (~2,000).

My take is here:
http://evimedgroup.blogspot.com/2010/11/more-thoughts-on-lung-cancer-screening.html
http://evimedgroup.blogspot.com/2010/11/ct-screening-for-lung-ca-epidemiology.html

Joyce Neifert said...

Dr. Kirsch,
Is it not true that even if all smoking ceased today, we would not see a significant decrease in lung cancer for decades? Do you propose that we simply sit on our hand (the other being in use for waggling the "shame on you" finger in the face of former smokers) and write off these casualties as collateral damage in the war against smoking? Pretty much has been the policy over the last 30 years and consequently, lung cancer mortality has remained stagnant while that of the other major cancers has skyrocketed.

Perhaps you have overlooked the fact that the majority of diagnoses are in former smokers or those who never smoked. Ah yes, that's right, collateral casualties, since every person diagnosed with lung cancer is painted with the same tobacco stain resulting from decades of telling people to just stop smoking.

And, by the way, just how IS that war on smoking going? Seems recent reports were that smoking appeared to be climbing once again after years of dropping. Guess those lung cancer deaths will just have to continue to mount. Heaven forbid we save a bunch, even those who wised up and quit. People should know better, right?

From NCI's "Report of the Lung Cancer Progress Review Group" August 2001 - "Although the link to tobacco is the clearest etiologic relationship for a human cancer, the development of lung cancer in persons who have never smoked and in former smokers and the failure of the majority of heavy smokers to develop the disease are poorly understood. The complex inter-relationships among genetic, molecular, and other biologic processes in modulating the carcinogenic response to tobacco smoke need to be further explored." Forgive my poor, obviously non-MD, interpretation, but it would appear that there are a whole lot more factors than smoking we at which we ought to be looking.

Catching lung cancer early is key to not only reducing the human cost of this disease, but also in changing the tide of public opinion regarding research funding. CT scans are the best chance we have for doing this NOW - TODAY. Some of the top lung cancer researchers in the country have long supported CT scans for lung cancer screening. It is interesting that strong criticism against it is usually accompanied by a position that we just need to get people to stop smoking.

I second the comparison with mammograms. Did you post a similar response when the use of mammograms was challenged?

One last comment - you see wasted $$$. Funny thing - I see only the face of my dead, non-smoking husband and eight other friends (most non-smokers), I lost to the disease in three years. That and the face of 10-year-old Gabby Wilson who has lung cancer. True this trial would not have been for them, but the potential of what this trial has produced has everything in the world to do with every lung cancer patient.

Joyce Neifert said...

Correction in above posting:
"mortality" in first paragraph should read "survivial rates" (obviously)

- long day = tired advocate

Michael Kirsch, M.D. said...

@Joyce, I am truly sorry for your loss and am grateful for your comments. The reason I am antagonistic to the CAT scan screeing strategy is because I believe that this test performs very poorly, as I discussed in the post. I fear that if this strategy is widely implemented, that it will cause much more harm than healing. With regard to my views on the mammogram debacle, I have a blog post you may wish to read. http://bit.ly/656CwP

lung cancer alternative treatments said...

I've heard how CAT scans help but I am not sure how effective this scan could be. Anyhow, I am so excited to see more studies about it soon. Keep blogging!

Add this