Read Original Story

Fox News regurgitates a hot mess from the UK Sun


2 Star


Artificial intelligence can predict Alzheimer’s 6 years earlier than medics, study finds

Our Review Summary

This brief, 375-word story published by Fox News was sourced from The Sun. It describes a study using an artificial intelligence algorithm to recognize patterns from brain scans.

The story at least warns readers that because of the small study size (just 40 subjects), larger studies would be needed to know if the reported “100% accuracy” for predicting Alzheimer’s six years before diagnosis, can actually be replicated.

But beyond that, the story is full of holes and canned quotes. The news release was far more informative.


Why This Matters

Early diagnosis of Alzheimer’s is a popular research area at this time. Stories about it often highlight the presumption that “catching” Alzheimer’s early will help doctors slow or stop the disease “before it starts.”

Although this may have intuitive appeal there is currently no evidence to suggest this is currently feasible, or even possible in the near future. It’s an important point that should be made in all stories touching on both early diagnosis and “prevention” of this degenerative neurologic disease.

This sort of framing runs a huge risk of misleading patients and their caregivers and — perhaps more disturbing — providing false hope based on what are usually preliminary findings that are by no means conclusive.


Does the story adequately discuss the costs of the intervention?

Not Satisfactory

Cost is not mentioned. Nor that the type of scan was specifically a PET scan.

According to this survey of six Veteran Affairs hospitals, the cost of a PET scan using FDG is approximately $1,900.

Without insurance (and without the FDG tracer) CostHelper lists the cost of a brain PET scan as $6,700.

Using the AI would presumably add additional costs.

Does the story adequately quantify the benefits of the treatment/test/product/procedure?

Not Satisfactory

The story highlights a single benefit:

“The [artificial intelligence algorithm] was able to identify dementia in 40 patients an average of six years before they were formally diagnosed … with 100% accuracy”

But who these patients were, and how they were chosen is not mentioned. Nor is it mentioned that the specificity was 82%. In other words, nearly 1 in 5 subjects (18%) were predicted to develop dementia who were never diagnosed with it (aka a “false positive test”).

Does the story adequately explain/quantify the harms of the intervention?

Not Satisfactory

Not mentioned. As with all screening tests, there is a risk of a false-positive or a false-negative result. Either can create anxiety, confusion, and incorrect medical treatment.

Regarding the PET scan itself, the American College of Radiology lists the following:

  • Although the amount of radiation in the tracer is low, potential harms should be considered in pregnant or nursing women.
  • Major allergic reactions to the tracer are rare but do occur.

The small circular opening of the scanner can elicit anxiety in some patients that may require sedation or stopping the scan.

Does the story seem to grasp the quality of the evidence?

Not Satisfactory

Other than selectively choosing to highlight the dramatic 100% sensitivity of the test (but neglecting to mention the 82% specificity mentioned above in benefits) the story doesn’t mention some key context:

Not only was the study group very small (n=40) it was also very select. All the subjects had already been referred to a memory clinic and their attending neurologist had been concerned enough to order a brain scan. That means it’s completely unknown how well this AI model would predict Alzheimer’s disease in the general public.

The story at least included quotes from people who made it clear that the results of this small, pilot study need to be refuted or confirmed by larger studies, but it should have been made clear to readers the primary reason for this caution is the highly select/non-representative nature of the study population.

Also, the story should have explained that to definitely diagnose Alzheimer’s, an autopsy is needed (or biopsy–rarely done). In this study, only one patient’s diagnosis was confirmed by autopsy.

Does the story commit disease-mongering?


The story does not disease monger.

However, we found this sentence problematic and think it’s worth noting:

Early detection of Alzheimer’s could open the door to new ways of slowing down or even halting the progression of the disease

Although the tone is speculative, it still erroneously suggests that early detection could result in slowing or stopping Alzheimer’s. At this time there is no treatment — regardless of the timing of the diagnosis — that can significantly slow or definitively stop this progressive neurologic disease.

Does the story use independent sources and identify conflicts of interest?

Not Satisfactory

The story included quotes that were collected by the Science Media Centre, yet this wasn’t disclosed. The other quote was from the news release, which also wasn’t attributed.

Does the story compare the new approach with existing alternatives?

Not Applicable

There is a lack of proven alternatives, though the story could have provided some context on this by discussing other efforts underway to detect Alzheimer’s early on.

Does the story establish the availability of the treatment/test/product/procedure?


It’s clear from the story that this machine-learning technology is not yet available and needs further study.

Does the story establish the true novelty of the approach?

Not Satisfactory

It’s unclear from the story if this is the first time such an algorithm has been used to “predict” the diagnosis of Alzheimer’s diseases or other diseases. Some context was greatly needed.

Does the story appear to rely solely or largely on a news release?

Not Satisfactory

One of the quotes came directly from the news release but was not attributed.

It’s worth noting that the news release was superior to the story in terms of useful details.

Total Score: 2 of 9 Satisfactory


Please note, comments are no longer published through this website. All previously made comments are still archived and available for viewing through select posts.