Analyzing the reporting of lung cancer CT screening

It’s been another challenging week for journalists covering various screening stories.

First the new analysis of Swedish mammography studies: Some familiar flaws surfaced in some stories.

Then came the followup analysis of the National Lung Screening Trial – first reported last fall.

This week various headlines announced:

  • More evidence CT scans better at detecting lung cancer
  • Study bolsters evidence that screening reduces lung cancer deaths
  • Nat’l Study Shows Long-Term Smokers Should Get Lung Cancer CT Scan

Most (but not that last one) commented on the tremendous health care policy and cost implications of more aggressive lung CT screening. Most also reported something about the very high rate of false positives in the study – although some downplayed the false positives – allowing one researcher, for example, to say that “there are no significant harms due to screening.” That is a comment that must be challenged, and we’ll provide more perspectives on this later.

But one story stood out in our searches for how imbalanced and incomplete it was – and because of the broad reach it could have because it was broadcast on a national television network. Among national news stories we saw, the following ABC News piece offered no discussion of false positives, no discussion of cost, but used hyperbolic, promotional language.

  • It called CT scans “simple.” There’s nothing simple about this decision.
  • It cited a cost of a “mere $99″ – not to be matched in many locations across the US and failing to take into account the followup costs of the considerable number of false positives.
  • It used as its primary source a researcher who is facing significant conflict of interest charges.
  • It promoted screening advice that simply hasn’t been established and didn’t cite the source of that advice.
  • It offered to help viewers find hospitals who could scan them – journalism? Or advertising?

Someone at ABC might say, you should have checked our website, which had more information. But the readers/viewers who asked me to review this story did not see the ABC website; they saw the ABC network TV piece.

One of our HealthNewsReview.org medical editors, Harry Demonaco, who directs the Innovation Support Center at the Massachusetts General Hospital, critiqued the ABC piece:

“The NLST paper is really a tour de force that was masterfully crafted and operationalized. The authors presented the results in a well balanced fashion. Unfortunately, the ABC News report did not.There are 94 Million smokers at risk for lung cancer in the United States today. According to the NSLT authors only about 7 million of them would meet the eligibility criteria to have been included in the study. This is important because the results may not be generalizable to the remainder 87 million smokers.

Yes, there were more lung cancers diagnosed with spiral CT as compared to an x-ray. The survival of patients depends a good deal on how early or late the cancer is detected, so early detection is better than late. But people need to dig a bit deeper into the data to get a better understanding of the study results. There were 427 deaths from lung cancer in the CT group as compared to 503 in the x-ray group. That’s a difference of 76 people in almost 54,000 people screened. Or, put another way 22.9% of the deaths in the group who had CT scans were caused by lung cancer as compared to 25.3% in the chest X-ray group.

There was a cost to those lives saved. 26,722 people were screened with CT scans and 6413 had a suspicious lesion that required follow up. Of those, 6182 were found to be false positives. This compares to 1844 positive findings with X-ray and 1743 false positives. More suspicious lesions found is a good thing but not without a cost. 75 of the folks who had a false positive finding on CT had a major complication associated with the follow up studies and compared to 24 in the chest X-ray group. Finding more cancers early is a good thing. Unfortunately there is a negative aspect to those findings.

The ABC report provides the listener with an amazingly unbalanced report on what is a landmark study. Focusing solely on the benefits without acknowledging the limitations of the study, its generalizabilty to the average smoker or ex-smoker and the complications associated with CT scans certainly does not inform.”

How will this story play out at the local level across the country? The local ABC station in Minneapolis-St. Paul posted a virtual commercial for a local hospital’s scanning.

But a far more cautious approach was demonstrated in the local St. Paul Pioneer Press. Its reporter looked at the same offer of discount scans from the same hospital and reported that “screening is not a free lunch”:

“Abbott Northwestern said it was offering the screening at a special low price that covers both scan and interpretation of the imaging study. Screening experts, however, said the hospital will more than make up for any discount on the front end.
“All those abnormalities that will be found will lead to more scanning,” said Peter Bach, a physician at Memorial Sloan-Kettering Cancer Center. “It will be a financial boon to the hospital …

… “While the study showed a substantial and statistically significant reduction in mortality from lung cancer, it has to be balanced by the potential harm from screening that has not yet been fully evaluated,” said Timothy Church, a researcher at the (University of Minnesota). “

It’s worth noting that when we reviewed news coverage of the initial release of NLST findings last fall, ABC also had the least balanced report of the ones we saw.

There may be institutional biases at play – reflections of individual biases on the part of key people at some news organizations.

And lest anyone accuse me of bias on this issue, let me be clear: I would never tell anyone not to be screened just as I would never tell anyone to be screened. First, I’m not a physician. Second, I don’t think journalists should give health care advice. I react when news stories don’t adequately inform consumers who may make unwise decisions based on what they’ve read or heard. This ABC piece wouldn’t be a good starting – or ending point – for such information.

Instead, see the comments from Dr. Otis Brawley of the American Cancer Society

“Guidelines groups have yet to carefully evaluate these and other data to determine who should and should not consider undergoing screening for early lung cancer detection and how often. Best practices have yet to be defined and in place in all clinical settings that will offer screening and diagnostic procedures, nor are we prepared to articulate the details that high risk individuals should hear before making a decision to undergo screening.
As this study demonstrates, the rate of findings suspicious for lung cancer was high in each screening round (over 27% in the first two rounds), but low-dose CT exams also can identify other non-lung related abnormalities, and this positivity rate also was high. So managing abnormal findings and avoiding doing harm in individuals with false positive findings are among the major challenges we will confront.

A very sobering observation in this report is that there were 26 deaths associated with invasive diagnostic procedures. We are reminded of the period right after the results of the first mammography screening trial were published in the late 1970s. Our country did not rush into screening, but rather the ACS and the NCI rushed to implement the Breast Cancer Detection Demonstration Project as a strategy to gain experience in the community based implementation of breast cancer screening. We must prepare to implement lung cancer screening in an even more deliberate and organized manner to ensure steady progress towards maximizing benefits, minimizing harms, and insuring that appropriate screening takes place based on risk.

While the study showed that screening can reduce the risk of lung cancer death in current and former smokers, we must always emphasize that screening is a secondary risk reduction strategy to smoking cessation. Proven smoking cessation strategies and the policies that promote them, including strong smoke-free laws, higher tobacco taxes, and fully funded smoking cessation programs, remain critical in the fight against cancer. The findings of this study do not diminish their importance, nor do they suggest that resources could be shifted from prevention to early detection. We estimate that quitting smoking will in ten years time reduce a smoker’s risk of death from lung cancer as much as CT screening did in this study.

Finally, if and when major groups do make a recommendation for screening, it will be important that those considering screening be made aware of the significant number of false positive findings and potential other harms associated with downstream testing that can occur with spiral CT scanning.”

We know that journalists struggle with screening stories. A simple reminder may help them do a better job: All screening tests do harm; some may also do good. If you don’t reflect that in your story, you’re probably doing harm as well.

Gary Schwitzer has specialized in health care journalism in his more than 30-year career in radio, television, interactive multimedia and the Internet.  He is publisher of HealthNewsReview.org.

Submit a guest post and be heard on social media’s leading physician voice.

Comments are moderated before they are published. Please read the comment policy.

Most Popular