One of the most overused cliches in medicine is the oft referenced: "When you hear hoofbeats, think horses, not zebras."
It's a valid reasoning in which to guide one's thought processes. After all, common things are common, uncommon things... aren't. But part of the responsibility of a physician also is to provide comfort and reassurance. It's our job to think "worst case scenario," to work up patients for those conditions, and provide reassurance when evidence is sufficient to quell our suspicions. Another common phrase in medicine is "until proven otherwise." Vaginal bleeding in a postmenopausal woman is cancer until proven otherwise. Acute onset of dyspnea or hypoxemia is a pulmonary embolism until proven otherwise. Severe epistaxis in an adolescent is a nasopharyngeal angiofibroma until proven otherwise. I recently had two patient who elucidated just how true this axiom can be.
A 62 year old woman presented with lateral chest pain of two weeks duration. On physical exam, her pain seemed very musculoskeletal in nature. Pain to palpation, pain on deep inspiration and with sneezing/cough, etc. The horse in this situation is a simple intercostal muscle strain. Regardless, we ordered a chest xray which showed ambiguous opacification of her right lower lung. It just didn't quite add up with the lack of any pulmonary symptoms. So, congresspersons and escalating health costs be damned, we decided we couldn't quite be comfortable with just writing things off, and sent the patient off with a referral for a CT scan and instructions for prn ibuprofen and heat. We saw her back today. The CT scan showed findings pathognomonic for lung cancer. Turns out, her pain was musculoskeletal in nature, as the cancer had begun to invade into her 8th rib. It had also spread to her spine. Zebra. Ironically, the patient returned to say that the heat and ibuprofen had really helped with the pain. If it wasn't for the CT, she would have been sent on her way with the belief that it was all just an intercostal muscle strain, while the cancer continued to grow in her chest.
A 22 year old woman presents with a painful unilateral cervical lymphadenopathy which had been present for 1 month. The horse in this situation is some form of infectious etiology: mononucleosis, cat-scratch fever, occasionally HIV (though this didn't jive with her history). She had been to several urgent care centers, and, going with horses instead of zebras, prescribed two antibiotic regimens, with no improvement of her symptoms. There was still a high likelihood her neck mass was viral in etiology, but we ordered a chest xray "just in case." It ended up showing an extensive mediastinal mass. One biopsy later, the diagnosis returned nodular sclerosing Hodgkin's. Zebra. Luckily, her prognosis is excellent and the delay in diagnosis likely will have no significant effect on her therapy. But it is never easy telling a previously healthy 22 year old that they have cancer, and there is a certain level of embarrassment that it took 5 visits to a physician to reach a diagnosis.
I think the most telling thing I've taken away from these experiences is how important the differential diagnosis remains in clinical practice. Most common symptoms can be attributed to the relatively benign conditions that afflict the gross majority of the general population. But it is important to always consider what else can be consistent with a clinical picture that is truly dangerous, as just because a condition is rare does not mean it cannot be affecting the patient sitting in front of you. Bacterial pharyngitis is common and fairly benign. A retropharyngeal abscess is not, and can often present identically. It is the responsibility of the clinician to use their clinical judgment and work up a patient to the point that they can confidently feel the patient is safe in the context of their illness.
After all, just one day, you may come across a zebra in downtown New York.