A Cause for Thoroughness

Contributed by readers. Edited by Sheldon Jacobson, MD

A 56-year-old man presented to the emergency department complaining of low-grade fever, chills, and periumbilical pain of two days' duration. The pain was constant and penetrating and radiated into the mid-back. The patient did not report having nausea, vomiting, diarrhea, or hematochezia. His medical history included hypothyroidism but was otherwise unremarkable. The only medications he was taking were aspirin and levothyroxine.

On physical examination, the patient was in mild distress. His oral temperature was 101.1°F; pulse, 88 and regular; and blood pressure, 122/88 mm Hg. The only significant physical finding was mild mid-abdominal and left upper quadrant tenderness. The rectal examination and stool examination for occult blood yielded normal results.

Screening laboratory tests in the emergency department showed a hematocrit of 38% and a white blood cell count of 12,500/mm3. Liver function, SMA 7, amylase, and lipase tests and urinalysis were all normal. Making a presumptive diagnosis of diverticulitis, the emergency physician prescribed clindamycin and cephalexin and referred the patient to his primary care provider.

The next day, the patient was already feeling considerably better by the time he visited his physician, who instructed him to continue the antibiotic therapy and undergo an abdominal computed tomography (CT) scan, which revealed several diverticula in the sigmoid and the descending colon but no abscess formation. The working diagnosis of diverticulitis was maintained, and the patient's condition continued to improve over the next five days, allowing him to resume his normal activities without difficulty.

About a year later, the patient, who had a family history of hereditary nonpolyposis colorectal cancer, underwent his regularly scheduled colonoscopy. The endoscopist found a large area of mucosal irregularity in the area of the splenic flexure. The results of biopsies from this area were consistent with adenocarcinoma. After the biopsy, the patient again began to have abdominal pain and fever--symptoms identical to those that had led him to the ED the previous year. A colon resection was then performed, which revealed cancer at Duke's stage B2.

COMMENT

This is a very instructive case. It reminds us that the link between symptoms or findings and a particular diagnosis should not be assumed and that we must construct diagnostic criteria that are specific as well as sensitive. However, the most important teaching point in this case is the clinicians' failure to consider the patient within the context of all the available information. The pain was located in the mid-abdomen and left upper quadrant, an unusual site for diverticulitis. The hematocrit was at the lower limits of normal. The patient had a strong family history of colon cancer. The CT scan did show diverticula, but this finding by itself is not diagnostic of diverticulitis. To make a definitive diagnosis of diverticulitis, the clinician must find--in addition to the diverticula--evidence of extramural air or fluid collection, air in the bowel wall, obstruction, a mucosal discontinuity, or mass lesion.

In this case, because the colon cancer was flat rather than polypoid, it was not identified on the CT scan. In retrospect, and in good medical practice, the patient should have undergone colonoscopy as well as an abdominal CT scan as part of the diagnostic work-up when he first presented with abdominal pain and fever.

Those symptoms were probably caused by a localized perforation or microabscess within the tumor site. It is well known that patients with colon cancer can present with bacteremia and sepsis before the tumor is detected. Streptococcus bovis is an unusual pathogen that has produced bacteremia in such patients.