Wednesday November 12 2014
Who do you think you are kidding, Mr Health Journalist?
"Fancy an episode of Dad's Army? How watching TV and films can save your eyesight," is the curious headline in the Daily Express.
Its headline is a rather abstract interpretation of research testing the potential for new computer eye-tracking software to help diagnose chronic glaucoma.
In glaucoma, pressure in the eyeball rises, damaging the optic nerve and threatening sight. Chronic glaucoma develops gradually, and loss of peripheral vision is usually the first sign.
The software being studied was designed to detect differences in eye movements between people with healthy eyes and those with glaucoma.
This study included just 44 older people with chronic glaucoma and 32 people of a similar age with healthy vision.
The computer software produced "scan paths", mapping eye movements while people watched three different film and TV clips, which indicated areas of visual loss.
As the news reports noted, one of the clips was from the ever-popular BBC sitcom "Dad's Army", although what was in the TV clips was irrelevant to the study or the patients' eyesight.
The computer software had fairly good accuracy for detecting glaucoma – about three-quarters of the people with glaucoma were correctly identified as having the condition using this test.
But we can only draw very limited further conclusions currently. We don't know whether the software will be affordable and become widely available, or whether it would offer any improvements on current methods used to detect chronic glaucoma.
Where did the story come from?
The study was carried out by researchers from City University London and was funded by a project grant awarded by Fight for Sight.
It was published in the peer-reviewed medical journal, Frontiers in Aging Neuroscience.
The media headlines give a misleading interpretation of this study. It is not possible for you to tell whether you have chronic glaucoma simply by watching an episode of "Dad's Army".
The historical sitcom just happened to be one of the TV clips that researchers showed people while tracking their eye movements using specialised computer software.
Even then, the software was not completely accurate at distinguishing which people did and did not have glaucoma. And we don't know that this test is an improvement on standard diagnostic tests.
What kind of research was this?
This was a diagnostic study where a control sample of elderly people with healthy vision, and another sample of people with glaucoma, received standard visual examinations. They also watched film and TV clips while a computer tracked their eye movements.
Researchers wanted to see whether they could differentiate between people with and without glaucoma by examining eye movements while someone watches a film.
Glaucoma is a condition where there is raised pressure in the eyeball. This can damage the optic nerve that carries visual information from the retina to the brain. The eye pressure increases because there is a blockage to the channels that drain aqueous fluid from the eye.
The patients in this study had chronic glaucoma, where the pressure in the eye gradually rises, causing a gradual loss of peripheral vision. Chronic glaucoma is more common with increasing age and can often run in families.
Current checks for chronic glaucoma include testing someone's peripheral visual fields, using a machine to measure the pressure in the eyeball, and looking at the back of the eye (retina) to check that the area where the optic nerve attaches to the eye looks healthy. Treatments can involve eye drops and laser surgery.
Chronic glaucoma is different from acute glaucoma, where the pressure in the eye suddenly rises very rapidly. Acute glaucoma is a medical emergency and needs immediate treatment to save the sight in the eye.
The researchers wanted to provide evidence that people with a diagnosis of chronic glaucoma can be distinguished from a group of age-matched healthy people by only using their visual scan paths while they watch a film or TV programme.
What did the research involve?
The researchers recruited 44 adults aged 63 to 77 with chronic glaucoma from Moorfields Eye Hospital in London. They deliberately recruited a sample of people who had variable degrees of visual field loss.
A comparison group of 32 adults (aged 64 to 75 years) with healthy vision were recruited from an eye clinic where they had received standard eye examinations. Both people with glaucoma and controls had no other significant health problems.
All participants had their visual fields tested using the optimal test designed to identify the early visual field loss associated with early glaucoma, the Glaucoma Hemifield Test (GHT), using a Humphrey Field Analyser (HFA).
The GHT was "outside normal limits" for all people with glaucoma and "within normal limits" for the controls.
The HFA mean deviation is the overall measure of the severity of the clinical field defect, and people with glaucoma were classed as having early disease if their mean deviation was better than -6dB in both eyes, and advanced disease if worse than -12dB.
The researchers outlined how people in the latter category would normally have symptoms and would most likely fail the visual field component for fitness to drive.
Best corrected visual acuity was also tested for all participants. There was little difference between people with glaucoma and healthy controls.
The main experiment involved participants viewing three separate TV and film clips taken from the 1970s TV comedy "Dad's Army", the 2006 film "The History Boys", and the 2010 Vancouver Winter Olympics men's ski cross event.
While they watched, the movements of the eye were tracked using special optical software. The software builds a scan path, illustrating the person's quick eye movements (called saccades) and fixations while they are watching. This scan path can indicate areas of vision loss.
What were the basic results?
Scan paths were built for each of the three film clips taken for both people with glaucoma and the controls – a total of 205 film clips.
Using a statistical measure known as the ROC curve, the researchers found the use of scan paths to detect chronic glaucoma was 0.85 (95% confidence interval [CI] 0.82 to 0.87) – with 1 indicating a perfectly accurate test, and 0.5 a useless diagnostic test with results no better than chance.
The result of 0.85 suggests scan paths obtained from this computer programme were a good – but not completely accurate – method of distinguishing between people with and without glaucoma.
The technique had a sensitivity of 76% (95% CI 58 to 86%), indicating roughly three-quarters of people with glaucoma would be accurately detected by using this test.
At this detection rate, the specificity was 90%, meaning 9 out of 10 people without glaucoma would accurately test as being free from the condition.
How did the researchers interpret the results?
The researchers concluded that, "Huge data from scan paths of eye movements recorded whilst people freely watch TV-type films can be processed into maps that contain a signature of vision loss.
"In this proof of principle study we have demonstrated that a group of patients with age-related neurodegenerative eye disease can be reasonably well separated from a group of healthy peers by considering these eye movement signatures alone."
This research demonstrates that a particular software application has fairly good accuracy for distinguishing between people with and without chronic glaucoma.
The scan paths that the software built, mapping eye movements while watching TV or film clips, were able to accurately pick up about three-quarters of those with glaucoma. Meanwhile, 9 out of 10 people without the condition accurately tested as being free from glaucoma.
The researchers appropriately call this a proof of concept study, in that they have demonstrated that the technique can reasonably separate people with and without chronic glaucoma.
But we can only draw limited further conclusions at this time. This study only tested a fairly small sample of people, and we don't know whether the same accuracy results would be obtained if a separate, bigger sample were tested.
We also don't know whether this test could offer any improvements on current methods for detecting chronic glaucoma. For example, it is not known whether the test could detect peripheral field defects any earlier than current standard visual field tests (combined with pressure testing), and so ultimately lead to the earlier detection and treatment of chronic glaucoma.
Of course, the ultimate aim of earlier detection is to improve outcomes for people in terms of preserving their vision. However, the current stage of research can offer no indication of whether this treatment could help "save your eyesight", as the Express headline suggests. As yet, no study has examined the longer-term outcomes of people with chronic glaucoma detected solely using this test.
Overall, these results suggest this software could have potential as a diagnostic technique to detect visual field loss in chronic glaucoma. However, it remains to be seen whether this test will ever be widely used in diagnostic practice, or how it would supplement or replace current standard tests.
Analysis by Bazian. Edited by NHS Choices. Follow Behind the Headlines on Twitter. Join the Healthy Evidence forum.