Observation, Bias, and the Burden of Uncertainty in Uveitis

In day-to-day clinical practice, uveitis often tests a clinician far more than we admit. The challenge is not just the inflammation itself, but the many small hurdles that quietly shape how we diagnose and treat this sight-threatening conditions. Interestingly, if we compare our situation with other branches of medicine that encounter inflammatory diseases, we may even realise how privileged we are. Consider a rheumatologist dealing with a case of giant cell arteritis. He must request a surgeon to perform a temporal artery biopsy, and then the sample goes to a pathologist who studies it meticulously to find the features of granulomatous inflammation centred around the internal elastic lamina, dense T-lymphocyte infiltrates, activated macrophages, and other tell-tale features. Only after all this does the rheumatologist arrive at a confirmed diagnosis. In contrast, an ophthalmologist examining an inflamed eye with the slit lamp can directly observe lymphocytes, plasma cells, neutrophils, or macrophages floating in the anterior chamber or deposited over the corneal endothelium as keratic precipitates. We can even classify the inflammation as granulomatous or non-granulomatous based on the clinical phenotype, and often point towards a likely etiology right at the slit lamp. We can not only grade the inflammation but also decide on treatment, monitor its response, and document every nuance in real time. And yet, despite these advantages, it’s not easy to manage a case of uveitis.

I do not have a single definitive answer for this, but several hurdles come to mind. One of the major issues is that a comprehensive eye examination is not always performed. Very often, a red eye is labelled as an allergic problem without checking for cells or subtle anterior chamber activity. Uveitis patients are often photophobic and may not tolerate bright light, making examination more difficult. While a case with 3+ cells is difficult to miss, a case with 1+ or occasional cells can easily escape attention unless we turn the slit-lamp rheostat to its brightest setting. White-eye uveitis, where the eye appears almost normal externally, adds an extra layer of complexity. Children with juvenile idiopathic arthritis are a classic example; the chronic, subtle nature of their inflammation means they may be treated with only topical steroids until they develop some sight-threatening complications. Studies show that even in developed countries, 80–90% of  children with uveitis present for ophthalmic care only after developing complications such as cataract or glaucoma. Comprehensive evaluation in children is nearly impossible without examination under anaesthesia, which becomes another limitation. We must be aware of these challenges while examining patients with uveitis and remain vigilant so our patients do not suffer because of them.

Sometimes in busy clinics, a dilated fundus examination is skipped or becomes difficult. Many posterior uveitic conditions present with some degree of anterior chamber inflammation, and without dilation, posterior lesions can be missed entirely. Viral posterior uveitis may show anterior chamber inflammation, while viral anterior uveitis can occur without any posterior involvement, and we cannot reliably distinguish the two without examining the fundus. Similarly, syphilitic retinitis, ocular tuberculosis, and sarcoidosis can be missed if we focus only on the anterior chamber and never look at the retina and choroid. Standing in 2025, indirect ophthalmoscopy is no longer a skill barrier for most ophthalmologists; it is simply a matter of setting aside time in the middle of a busy OPD. As Osler remarked, “A physician’s day is full of interruptions; the art lies in learning to weave them into the work.” The same applies to us when a complicated case appears during a crowded clinic session. Cognitive load, stress, fatigue, and time pressure can easily impair critical thinking, and even experienced clinicians may overlook important clues.

Bias in clinical reasoning is another obstacle. Even though we frequently discuss various clinical phenotypes, we often become victims of our own familiarity and jump to conclusions based on patterns we recognise. We all enjoy reaching a diagnosis quickly, and this sense of satisfaction can easily feed our “clinician ego.” But this can lead us to ignore contradictory evidence, this is perhaps what we call ‘falling in love’ with our own diagnosis.’ As Bob Dylan said, “You can never be wise and be in love at the same time.” A clinician cannot be wise when he or she is too attached to their own diagnosis. Subtle differences and contradictory evidence against the diagnosis may be overlooked in such situations. I feel such scenarios occur more commonly in subspecialty practice, as when we become too comfortable with what we know, overconfidence walks in and subtle mistakes follow. A case presenting with subretinal fluid may be mistaken for central serous chorioretinopathy, or conversely, for an inflammatory condition. A few cells in the anterior vitreous or a vague complaint of headache may nudge us toward Vogt–Koyanagi–Harada disease when the truth lies elsewhere. Treating such a case with oral steroids without first performing a fluorescein angiogram can have serious consequences.  If you analyse such situations in our clinics, you will realise that such mistake can occur while attempting to treat the clinical sign rather than the underlying cause or disease. A similar mistake happens when clinicians try to treat a laboratory test rather than the clinical condition. The availability heuristic further complicates matters, conditions recently seen at conferences or meetings tend to occupy our minds more, affecting our judgement. That is why some patients with anterior uveitis are started on antituberculosis therapy simply because the Mantoux test reads 15 mm, or given antivirals because herpes serology is positive.  Herpesviruses are remarkably ubiquitous, with some types infecting up to 90–95% of individuals in certain populations. These viruses have coexisted with humans since ancient times and have evolved alongside us for millions of years. One must be careful not to treat intraocular inflammation solely on the basis of a single test. 

From antiquity, medicine has remained largely a science of ‘observation’. Even in remote villages, a barely educated quack may run a flourishing practice because he has a strong sense of observation learnt from years of watching a qualified physician—carefully noting what symptoms and signs led to what treatment. History offers even more striking examples. Greek, Arabic, and Indian physicians could describe diseases with remarkable accuracy despite lacking modern tools. Centuries before germ theory or microscopes, Hippocrates, Rhazes (Al-Razi), and Ayurvedic scholars from India distinguished smallpox from measles purely by observing the pattern and depth of the rash, the sequence of fever, and the overall disease course. Rhazes described how smallpox lesions appeared more superficial, how measles eruptions were deeper and preceded by severe respiratory symptoms, and how smallpox followed a predictable evolution from macules to papules to pustules and finally crusts. Such precision is astonishing, especially when they had no laboratory support, only their eyes, their experience, and their ability to reason.

Finally, perhaps the essence of good clinical practice lies in differential diagnosis. Though in different context, but Osler expressed it beautifully when he said, “Medicine is a science of uncertainty and an art of probability.” Diagnosis begins with considering all possibilities and then slowly refining them into the most likely one. Most importantly, differential diagnosis is not about finding the perfect answer; it is about not missing the dangerous one. Uveitis is full of look-alike conditions and masquerade syndromes, which can be captured perfectly by a single word or aphorism in medicine: the “zebra.” Theodore E. Woodward (1914–2005) was an American physician, a highly respected infectious disease specialist, and a long-serving professor at the University of Maryland School of Medicine. He famously taught his students, “When you hear hoofbeats, think horses, not zebras.” It is true that common diseases are common, and one should consider the most likely diagnosis before jumping to rare ones. But in uveitis, we must remember that zebras also exist, and  one must think also about ‘zebras’ when hear hoofbeats.