> Skip repeated content

HSS Manual Ch. 2 - Thinking Like a Rheumatologist

From the HSS Manual of Rheumatology and Outpatient Orthopedic Disorders


  • The diagnosis of many rheumatologic disorders is made clinically, and so a detailed medical history and thorough physical examination are unequivocally central in the initial evaluation of the patient. A strong knowledge base of rheumatology streamlines the diagnostic process, enabling quicker development of management plans.
  • Always treat the patient, not the laboratory results. While laboratory, radiologic, and pathologic data can be very useful in the management of rheumatologic disorders, they should always be taken in the context of and never supercede the clinical picture. Appreciating the limitations of diagnostic tests optimizes their clinical utility.
  • Uncertainty is rife in rheumatology and must be accepted. Management decisions must often be made even when the clinical picture is incomplete or atypical or when clinical data is unavailable or inaccessible.
  • Rheumatologic disorders are often variable in course and severity. The aggressiveness of therapy must be appropriate to the aggressiveness of disease, because both treatment modalities and illness carry potential dangers. The chronic nature of many conditions necessitates ongoing vigilance even during periods of disease quiescence.
  • Better education of the patient, especially with respect to the nature of illness and to therapeutic goals and expectations, and trust between the physician and patient optimize compliance and outcome.

Hanging in my examination room are reproductions of two French impressionist paintings. The first is the famous A Sunday on La Grande Jatte by Georges-Pierre Seurat who pioneered the technique of juxtaposing small dots of different colors to create images that become apparent only when seen from a distance. Even then, however, smaller details can remain obscure and subtle. I use this painting to illustrate to patients how I often approach rheumatologic conditions. Firstly, while I am generally called upon to evaluate a specific problem, I do not focus solely on one single “dot” but rather view it in the context of all the “dots” in order to see the whole clinical picture. Secondly, even if the picture is spotted with areas of fuzziness and uncertainty, it can still be fully appreciated and addressed with comfort.



Many rheumatologic conditions are clinical diagnoses and are systemic in nature, and so it cannot be overstated that the skills most important to the rheumatologist are those that are also the most important to an astute internist. These include the ability to obtain an accurate medical history and conduct a thorough physical examination and to be comfortable with handling different organ systems. The review of systems, in particular, often provides crucial pieces of information that may not be spontaneously volunteered by the patient and comprises a large part of my initial evaluations. This process, though seemingly exhausting, can be made very efficient by attaining familiarity with potentially relevant conditions. For example, an elderly man taking diuretics for hypertension who presents with recurrent acute inflammation of the first metatarsophalangeal joint need not necessarily be questioned for a history of sun sensitivity or a malar rash but should be questioned for a history of tophi or renal calculi. A young woman with a history of multiple osteoporotic stress fractures should probably be asked about symptoms suggestive of malabsorptive states. A large fund of knowledge a priori improves the diagnostic process by generating pertinent questions and discarding irrelevant ones.



Laboratory, radiologic, and pathologic studies can be extremely useful to the rheumatologist, but they should only be obtained in the appropriate setting. Inappropriate testing can often increase diagnostic confusion as well as become as source of unnecessary anxiety for the patient. The utility of diagnostic testing is highly dependent on the pre-test probability of a particular condition, and so again, an astute clinical evaluation beforehand remains central. For example, the antinuclear antibody is a highly sensitive test for systemic lupus erythematosus (SLE) but is also notoriously non-specific. Therefore in considering the diagnosis of SLE, a negative test result can be very useful in excluding this diagnosis while a positive test can best be used to support the clinical impression. Conversely, the anti-dsDNA antibody is highly specific but only moderately sensitive for SLE; thus, it is less useful as a screening test and more useful (if positive) as a confirmatory test.

Tests like the erythrocyte sedimentation rate (ESR) and C-reactive protein can provide useful information regarding the degree of activity of a systemic disease. However, one should never be swayed blindly by the results of these tests and should always take the overall clinical picture as the guide for developing the management plan. A patient with polymyalgia rheumatica (PMR) who has a slightly elevated ESR but who is feeling well does not need to have her corticosteroid dosage increased just to normalize the ESR. Conversely, a patient with PMR who complains of a recurrence of significant muscle stiffness in the morning should probably increase her corticosteroid dosage no matter what the ESR is.



Because many rheumatologic diagnoses are made primarily on a clinical basis, one of the greatest challenges in training rheumatology fellows is to teach them to become comfortable with uncertainty. This can only be accomplished by maximizing clinical experience and maintaining a solid knowledge base.

Many criteria and classification schemes for the diagnosis of rheumatologic disorders have been published, but for the most part, these were developed for the purpose of clinical trials and population studies and not as the sole basis for making diagnoses in specific patients. Thus, a young woman with a malar rash, glomerulonephritis, and a positive ANA should be treated as an SLE patient, even if she does not fulfill a fourth criterion for diagnosis as established by the American College of Rheumatology.

Not infrequently, a patient may present with an obvious but undiagnosed systemic inflammatory condition which may be threatening a vital organ or even life, and awaiting a definitive pathologic diagnosis may carry unacceptable risks. An example of this may be an elderly woman presenting with fever, myalgias, and new vision loss, consistent with the diagnosis of giant cell arteritis (GCA). In this situation, therapy should not be delayed pending a temporal artery biopsy because the risk of permanent blindness far outweighs the risks of corticosteroid treatment. Moreover, a strong argument can be made for treating for presumptive GCA, even if the biopsy result is negative.

On other occasions, a patient may have an inflammatory condition such as an interstitial pneumonitis and is steadily declining, but diagnostic testing has been reasonably extensive to exclude infection or malignancy even though no definitive diagnosis is revealed. Here again, systemic anti-inflammatory or immunosuppressive therapy may also be appropriate, so long as continued vigilance is maintained for new or progressive problems.



Two important characteristics of many rheumatologic diseases are chronicity and wide variability in course and severity. SLE may be quite mild for many years with easily managed intermittent arthritic or dermatologic flares or may be aggressive, rapid and fulminant at any moment with endangerment of vital organ or life. It is important to remember to treat the patient at hand, not the diagnosis. There is no one treatment regimen that is universally appropriate for any particular diagnosis, and the management plan needs to be as potent as the severity of a particular case dictates. Unlike an oncologist using very potent (and toxic) medications to treat a malignant neoplasm that if left untreated would kill the patient, rheumatologists tend to walk a finer line between the threats imposed by the disease and the toxicities carried by the therapies. Close and vigilant monitoring for an indefinite period, whether or not the patient requires medical treatment, is the rule.



Rheumatologic conditions are often difficult to understand for the medical professional, let alone the patient. Therefore, every effort must be made to help the patient to the best of his/her ability to understand the nature of the illness and the goals of therapy and to establish a trusting relationship. Without understanding and rapport, compliance and therefore outcome are diminished.

The second painting hanging on my wall is Two Young Girls at the Piano by Auguste Renoir, who himself was afflicted by severe arthritis and towards the end of his life required that his brushes be bound to his hands in order to paint. It is said by some that Renoir sometimes portrayed his subjects as having the arthritic changes of his own hands, and I can make out the synovitis on the hands of the two young girls in the painting. I can only imagine what more he could have produced if he were not so disabled. However, now with very effective treatments for our diseases, there is great promise to prevent their destructive consequences, and perhaps, the full potential of all our patients can be protected and realized.


Headshot of Arthur M. F. Yee, MD, PhD
Arthur M. F. Yee, MD, PhD
Assistant Attending Physician, Hospital for Special Surgery
Assistant Professor of Medicine, Weill Cornell Medical College

Related Professional Content