Approach to Diagnostic Testing

Introduction

Patients are often preoccupied with the “cause” of their affliction. Although this is understandable, the etiology of a given uveitis syndrome is often elusive, and about half of cases are labeled idiopathic or undifferentiated. In contrast, for the clinician, the first goal of any diagnostic testing is to differentiate between infectious and noninfectious/inflammatory etiologies. A third important but rare category is ocular neoplasm masquerading as uveitis (e.g., retinoblastoma, primary vitreoretinal lymphoma), and a fourth is nonneoplastic masquerade syndromes, such as inherited retinal degenerations or intraocular foreign bodies. The distinction between these four groups is critical because management is so radically different, at times even diametrically opposed.

Once a diagnosis of noninfectious uveitis (NIU) has been made, it is less important to make the diagnosis of a specific uveitis syndrome. The critical determination—that the disease should be treated with antiinflammatory therapy—has already been established. However, it can still be helpful to make a more specific diagnosis for three purposes: first, it informs prognosis; second, it may help guide therapy; and third, it may shed light on other systemic symptoms of heretofore unknown origin. For instance, a patient in her 20s with new-onset, acute unilateral anterior uveitis with a 1-year history of low back pain that she has been ignoring tests positive for HLA-B27. This result raises the diagnosis of ankylosing spondylitis as the origin of her back pain, prompting referral to a rheumatologist and radiographs that confirm the diagnosis. Moreover, HLA-B27 disease is typically recurrent and carries a fairly good prognosis if managed correctly.

The goal of this chapter is to briefly review the appropriate use of laboratory tests based on the uveitic entity determined predominantly by the anatomic site of inflammation. Other common historical or examination features will occasionally be included to help narrow testing. Although uveitis specialists have struggled to decide on a standard set of tests for a given presentation, there is a general consensus that targeted testing, based on key historical or examination features, is strongly preferred over broad “shotgun testing” for every case of intraocular inflammation. Recall that based on Bayes’ theorem, a diagnostic test is most helpful when one’s pretest probability of the diagnosis is closest to 50%. As clinical suspicion of a diagnosis decreases (i.e., pretest probability approaches zero), the positive predictive value (i.e., the probability that a patient truly has the diagnosis given a positive test for it) declines precipitously, such that even highly sensitive and specific tests become nearly useless when positive. Lyme testing is an excellent example of this theorem. As such, each of the following sections outlines essentially a first line of testing with the knowledge that specific features of a case often lead to more targeted and sometimes esoteric testing. The point that specific cases, based on their features, can warrant other tests beyond the ones included here cannot be overemphasized, and the proceeding case chapters serve to illustrate that very concept.