Considered Decisions
Can a glaucoma diagnosis do more harm than good? Let's see what the experts have to say...
Keith Barton, Chelvin Sng, Dan Lindfield, Malik Kahook, David Garway-Heath | | Opinion
We asked five glaucoma experts for their opinions on the issues brought to the fore by Raymond Radford's feature: “Physician, Heal Thyself.” Here’s what they told us.
Keith Barton, Consultant Ophthalmologist, Moorfields Eye Hospital, London, UK
The idea that being labeled a “glaucoma suspect” affects the patient’s life forever does not ring true for me. Sometimes a diagnosis of glaucoma is not clear cut, although accuracy has improved considerably in recent years, with advanced imaging technologies. However, these improvements mean that more abnormalities to the optic nerve are detected, but they might not result in the patient developing glaucoma.
I do see a lot of over-treated mild glaucoma, and a lot of under-treated severe glaucoma, with many patients losing their vision unnecessarily. Two decades ago, general ophthalmologists were adept in trabeculectomy, but with therapeutic advances, the rates of trabeculectomy dropped, and it is now seen as a more specialized procedure. Many glaucoma specialists do not perform a lot of glaucoma surgery, resorting heavily to the use of drops, but that has real implications for patients who would benefit from surgical approaches, even if they are a very small minority.
Malik Y. Kahook, The Slater Family Endowed Chair in Ophthalmology. Professor of Ophthalmology, Sue Anschutz-Rodgers Eye Center, University of Colorado School of Medicine, Aurora, USA
It is without doubt that our inherent biases can shape the way we interact with patients. Physicians are human, after all – and we are prone to all of the factors that influence decisions both personally and professionally. This is, in large part, why we call what we do an “art” rather than a concrete science dependent on a “check-box” approach to patient care. Daniel Kahneman and Amos Tversky, who partnered in research at the crossroads of psychology and economics, dissected our decision-making process and championed a path that involved undoing our assumptions, which they believed caused the human mind to err systematically when forced to judge situations in the presence of uncertainty (1). The term attached to their work was “heuristic,” which Wikipedia defines as “any approach to problem solving or self-discovery that employs a practical method, not guaranteed to be optimal, perfect, logical, or rational, but instead sufficient for reaching an immediate goal.” This definition sounds very much like the decision-making technique employed by many, if not most, physicians around the world every day. To what degree should our clinical decisions leverage concrete data? To what degree should we lean on subconscious decision making based on past experiences? As with most things in life, a balance between the two is likely the best path. Readers of “Physician, Heal Thyself” would be well-served to take some time for self-reflection on what factors guide us in clinical decisions and to continue to contemplate these factors the next time they see patients in clinic. I am left wondering how much of my own inherent biases shaped the writing of this commentary!
David Garway-Heath, IGA Professor of Ophthalmology for Glaucoma and Allied Studies, Moorfields Eye Hospital and UCL Institute of Ophthalmology, London, UK
Labeling a patient using certain terminology can cause anxiety, but it is the explanation that goes with the “label” that is important. I always tell glaucoma patients that if they are diagnosed early, they rarely lose vision to an extent that would be noticeable to them. Identifying that glaucoma may be present is an important step to diagnosis, but a patient should not remain a “glaucoma suspect” forever – if there is no deterioration identified on monitoring, then the patient should be reassured and discharged.
Unconscious bias is a complex issue to consider. We are moving towards objective imaging methods, including objective methods to measure IOP, so any bias should diminish as new technology is adopted. Personally, I don’t think preconceptions have much effect on assessments of glaucoma patients.
When talking about surgery risks, context matters. Surgery done well has fewer risks and shouldn’t be relegated as an option because of perceived risks. Appropriate discussions with patients are paramount, so that they understand the potential risks and benefits.
Dan Lindfield, Consultant Ophthalmologist and Glaucoma Lead, Royal Surrey County Hospital, UK
I believe that, as physicians, we have to highlight our use of language. Despite my repeated pleas, my hospital sends all patients a letter confirming their “glaucoma clinic” appointment prior to even meeting a diagnostic professional.
Radford’s article makes an evocative and provocative point about the “cost” of carrying the label of glaucoma. However, there is also the unmentioned flip side whereby patients with glaucoma often present late with significant visual impairment, and threat to their lifestyle; for example, keeping their driving license.
Immediate previous experience certainly subconsciously (and often consciously) impacts our decision making. We’re high-functioning humans after all, not binary machines outputting a “yes/no” answer. For example, last week I saw a patient referred over five years ago with suspected glaucoma, who had been reassured and discharged back to optometric care. However, the patient didn’t attend routine annual checks as instructed, believing that the optometrist “got it wrong” the first time. Five years later, the patient presented with central visual field defects in both eyes.
I will welcome AI into this process, but judgement is the hardest thing to teach (both to the doctor and the machine). Our patient’s own views and beliefs should be at the center of our care. No two patients are alike. We must not fall into the trap of just seeing mmHg, RNFL thickness, and mean deviation.
Chelvin Sng, Consultant Ophthalmologist, National University Hospital, Singapore
Being labeled a “glaucoma suspect” can induce significant anxiety, but this is certainly not an inevitable outcome, and indeed is often the consequence of inadequate communication and patient education. Just as well-intentioned doctors dole out diagnoses and labels in order to neatly categorize each patient within a well-established management framework, these labels must always be accompanied by an adequate explanation of the relevant implications and prognoses. Indeed, the vast majority of “glaucoma suspects” do not have the disease, and will never develop it. Hence, when informing patients that they are “glaucoma suspects”, the doctors should emphasize that this label is not a cause of undue concern, but is most likely a mere inconvenience, requiring regular monitoring. With adequate patient education and counseling, the “glaucoma suspect” label does not necessarily “reduce the patients’ quality of life forever” or “make them worry about blindness for the rest of their lives.”
Unconscious bias can indeed skew management and result in imperfect clinical decisions. And that has led to significant excitement about the role of artificial intelligence and its applications, not only in ophthalmology, but also in other clinical specialties. Currently, AI in glaucoma is still in its infancy, and future developments may be hindered by the lack of a clearly defined gold standard for determining the presence and severity of glaucoma, which undermines the training of artificial intelligence algorithms. In addition, what we negatively brand as “unconscious bias” may indeed be beneficial for patient management. A doctor’s clinical experience and training may hone an innate intuition that cannot be captured by AI algorithms, and may influence clinical decision-making positively (2). Improvements in optic disc imaging techniques are also likely to reduce clinician subjectivity in glaucoma diagnosis and monitoring.
The decision to escalate glaucoma therapy is often based on inadequate intraocular pressure control rather than clinical evidence of glaucoma progression. This is unsurprising as the philosophy behind glaucoma treatment favors prevention rather than reaction. When faced with the prospect of irreversible glaucoma progression, most would err on the side of over-treatment rather than under-treatment (especially in the context of advanced glaucoma), even if there is no current evidence of progression. Nevertheless, a consultative rather than prescriptive approach to management decisions is advocated, with each patient’s preferences and life expectancy taken into account. This is especially important when conventional glaucoma surgery (such as trabeculectomy) is considered, as potential sight-threatening complications must be weighed against the risk of losing vision from glaucoma. With the recent renaissance in glaucoma surgery, safer surgical options (for example MIGS) can now be offered earlier in the glaucoma treatment algorithm, with less fear of blinding complications.
- D Kahneman, A Tversky, Choices, Values and Frames. New York: Cambridge University Press: 2000.
- K Yamada, S Moris, “The day when computers read between lines.” Jpn J Radiol, 37, 351 (2019). PMID: 30911986.
Barton is a Consultant Ophthalmologist and Director of The Glaucoma Service, Moorfields Eye Hospital, London, UK. His principal interests are the surgical management of glaucoma (especially with aqueous shunt devices and MIGS approaches) and secondary glaucomas – Keith runs an NHS clinic at Moorfields that’s uniquely dedicated to the management of glaucoma in uveitis. However, Keith’s skillset goes beyond surgery. He’s well known for organizing and running glaucoma educational symposia that have attracted significant acclaim, and he famously co-founded and organized (with Kuldev Singh) the Ophthalmology Futures Forum investor networking events that take place in both Europe and Asia. He was voted to #10 on the 2018 Power List.
Sng is Consultant Ophthalmologist at the National University Hospital, Singapore; Assistant Professor at the National University of Singapore; Honorary Consultant at Moorfields Eye Hospital, London, UK. She was voted #9 in The Ophthalmologist Rising Star Power List in 2017.
Dan Lindfield is Consultant Ophthalmologist and Glaucoma Lead, Royal Surrey County Hospital, Guildford, UK, Training Programme Director, Health Education England, and Director, Farnham Eye Care.
With over 30 patents filed, Kahook’s research is focused on novel devices and surgical instruments for glaucoma and cataract surgery, and advanced imaging techniques. He is Director of Clinical and Translational Research at the University of Colorado, and is also the editor of Essentials of Glaucoma Surgery, MIGS: Advances in Glaucoma Surgery, and author of over 250 papers, chapters and abstracts. He was named New Inventor of the Year 2009 and Inventor of the Year 2010 at the University of Colorado.
IGA Professor of Ophthalmology for Glaucoma and Allied Studies, Moorfields Eye Hospital and UCL Institute of Ophthalmology, London, UK
