
The Febrile Patient with Abdominal PainIn this article, we will discuss the work-up of the febrile patient with abdominal pain. We will discuss the current imaging modalities utilized to diagnose the patient's problem. Of course, the work-up is determined by the locale of the patient's pain and laboratory values. In the event that the patient has non-localized abdominal pain, CT is the examination of choice. Administration of oral contrast, and sometimes, rectal contrast, is imperative to differentiate bowel loops from abscesses. Intravenous contrast improves contrast resolution. If there is poor bowel opacification and intravenous contrast cannot be administered, identification of the abscess may be difficult, but not impossible. Ultrasound has limited value in this setting, as it is impossible to image the entire abdomen and pelvis due to obscuration by bowel gas. As such, ultrasound should be reserved for cases with localized abdominal pain.
For acute abscesses, a Technectium-labeled white blood cell scan is the examination of choice. (Indium-labeled white cells have slightly better specificity and sensitivity, but it is difficult to obtain in Hilo.) In either event, 50cc of the patient's blood is withdrawn. The white cells are separated from the red cells, platelets, and plasma. These patients usually demonstrate polymorphonuclear leukocytosis. If the patient is neutropenic, a leukocyte transfusion can be labeled, or 50cc of a whole blood transfusion can be separated and the leukocytes labeled and re-injected into the patient. Imaging is done over the next 4 hours with a 24 hour delayed scan, if necessary. The leukocytes migrate to the areas of inflammation, thereby identifying the abscess. Over 90 percent of patients with positive scans will have subsequently documented culture evidence of infection. After the abscess is identified, percutaneous drainage can be performed surgically or under CT guidance. In the past, radionuclide imaging was performed with gallium citrate, which was suboptimal. Interpretation of gallium scans was difficult due to normal bowel excretion obscuring bowel pathology and normal accumulation of the gallium in the healing wounds of post-operative patients. Gallium is best used for evaluation of inflammatory bowel disease, vertebral infections, chronic osteomyelitis or pulmonary pneumocystis carinii.
Right Upper Quadrant Pain If the patient has localizing symptoms, then the work-up should be directed towards that area. If the patient has has right upper quadrant pain, then the diagnosis considered should include: acute cholecystitis, urinary tract obstruction with pyonephrosis, hepatic abscess, subdiaphragmatic abscess, and bowel perforation. Acute Cholecystitis: Sonographic evaluation of the gallbladder has a gallstone detection accuracy of 99%. Small calculi in the neck of the gallbladder are extremely difficult to detect, and can result in a false negative exam. However, as 15% of the population has asymptomatic cholelithiasis, it is difficult to know if the gallstones are to blame. There is a 66% probability that the patient's cholecystitis is the cause of the pain. We check for several ancillary signs, including the sonographic Murphy's sign, gallbladder wall hyperemia, and gallbladder wall thickening, to further refine our diagnosis. If the patient has gallstones and one of the ancillary signs, then the probability rises to 90%. If the patient has gallstones and two signs, then the probability rises to 95%. The biliary tree is also examined for dilatation. It is difficult to identify 70% of cases of choledocholithiasis, as parts of the common bile duct are obscured by bowel gas. (Magnetic Resonance Cholangiography will be discussed in a future issue of Imaging Update.) If the gallbladder appears sonographically negative but there is a high suspicion of gallbladder disease, then a radionuclide HIDA scan is suggested. This examination evaluates patency of the cystic duct. If the cystic duct is obstructed and the gallbladder is not visualized, then the diagnosis is acute cholecystitis, whether due to cholelithiasis or acalculous cholecystitis. We also evaluate patency of the common bile duct. This examination is 97% accurate in the evaluation of cystic duct patency. Sonography continues to be the initial exam, since it is rapidly performed (30 minutes) and can detect other abnormalities, whereas a HIDA scan may take in excess of 3 hours, if abnormal. Hepatic Abscess: Hepatic abscesses are not usually the primary suspected abnormality in Hawaii. In other parts of the U.S., where cysticerchosis is endemic, hepatic abscesses are not infrequently the presenting problem. These are readily detected at sonography and CT, both of which can be used for guidance for the performance of percutaneous drainage. At sonography, they present as complex, posterior acoustic enhancing, well-defined masses, which can be as large as 12cm in size, which may be multiple. At CT, they are seen as low density masses. They are readily differentiated from hepatic cysts and solid masses, such as focal nodular hyperplasia, hepatoma, and metastasis. Urinary Tract Disease: Patients with urinary tract obstruction usually present with flank pain and abnormal urinalysis, though if the obstruction is severe and complete, there may be no hematuria. Most patients are not febrile, unless there is an infection associated with the obstruction. Thus, the work-up of these patients will be directed to ultrasound of the kidneys or spiral KUB, both of which are highly accurate. Sonographic accuracy of detecting urinary tract obstruction is 95%. Calyceal dilatation is seen sonographically on the affected side. If there is minimal questionable dilatation, then we perform doppler analysis and measure the resistive indices (RI). If there is a significant difference in the RI, with the suspect side having a RI of 0.1 or greater than the assymptomatic side, then we are led to suspect urinary tract obstruction. We also examine the bladder for ureteral jets, which is the sonographic depiction of urine exiting the ureter, entering the bladder. If bilateral ureteral jets are seen, it is highly unlikely that there is proximal obstruction. Sonography can demonstrate renal calculi if they are of sufficient size, on the order of 3mm or larger. It is difficult but not impossible to visualize smaller calculi. Ureteral calculi, however, cannot usually be demonstrated. As the ureters are retroperitoneal structures, they are obscured by bowel gas. Therefore, if a patient has documented urinay tract obstruction, we recommend the performance of an AP supine film of the abdomen (KUB). We are able to visualize 80% of radiopaque calculi, which comprises 80% of all calculi. This documents the etiology of the obstruction. The urologist also then uses the KUB as a baseline as he follows the calculus to passage into the bladder. Since ultrasound does not utilize radiation, it is the exam of choice for pregnant patients. For the past several years, we have been performing noncontrast spiral CT scanning of the urinary system (Spiral KUB) in lieu of Excretory Urography (IVP)and Sonography for the evaluation of the urinary tract for obstruction. The examination takes 10 minutes and does not utilize oral or intravenous contrast. Thin section CT is performed of the entire urinary tract. Accuracy is up to 99%. Findings at CT include intra-renal dilatation, ureteral dilatation, renal swelling and perinephric fluid due to the distal obstruction. We are able to detect the size and location of the urinary tract calculi. We can detect radio-opaque and radiolucent ureteral calculi. Following the exam, we perform a KUB, so that the urologist can follow the calculus to resolution. However, there will be calculi that cannot be seen on the AP abdomen film due either to their small size or the fact that they are radiolucent. (Click to see image) As there are patients who are initially suspected to have urinary tract abnormality, but rather have another etiology of their illness, we are sometimes able to visualize other disease processes, such as appendicitis, ovarian masses, and pancreatic masses. In the past several months of performing spiral KUB's, we have discovered three cases of incidental renal carcinomas in patients who also had ureteral calculi. At least one of these might have been too small to be detected on an IVP, though it would have been visualized at sonography. (Click to see image) Pyelonephritis, renal carbuncles and carcinomas, however, are better evaluated with intravenous contrast. Note that spiral KUB is not the examination of choice for evaluation of painless hematuria, which is best evaluated via an IVP since urothelial tumors produce filling defects in the contrast column. (Remember that we do not typically administer IV contrast for a spiral KUB.) Perforated Viscus: A pneumoperitoneum is readily detected on an upright chest or abdomen film. If the initial imaging exam is a CT of the abdomen, a pneumoperitoneum is also readily detectable, with the added benefit of possible localization of the site of perforation, whether it be the right colon, duodenum or stomach. Right Lower Quadrant Pain If the patient's pain is in the right lower quadrant, then the diagnoses considered should include acute appendicitis, urinary tract disease, pelvic inflammatory disease, and diverticulitis with or without an abscess. Acute Appendicitis: Historically, appendicitis had been a clinical diagnosis. However, when the clinical impression alone is relied upon for diagnosis, approximately one out of every five patients undergoes appendectomy for a normal appendix. In addition, diagnosis is initially overlooked in every fifth patient who actually has appendicitis, according to national statistics. The national average false positive rates of 20% on clinical examination rises even further in the subgroup of young women in their reproductive years, where the clinical impression of appendicitis may be wrong 40-45% of the time. Sonography has been performed in an attempt to visualize the appendix, however, this is a difficult operator-dependent exam, especially if the patient is obese. The size and compressibility of the appendix is evaluated. The normal appendix can be demonstrated occasionally in normal patients. In cases of appendicitis, the abnormal appendix can be demonstrated as an abnormally enlarged, non-compressible tubular structure, perhaps with an appendicolith, and perhaps with surrounding inflammation. However, the abnormal appendix cannot always be demonstrated due to technical difficulties. Ultrasound, however, still may be the appropriate examination in pregnant females or where gynecological pathology is of primary concern. A recent study from Massachusetts General Hospital documents that spiral CT had a 100% sensitivity for the diagnosis of appendicitis with an accuracy rate of 98%. A normal appendix was seen in all patients without appendicitis who were scanned using their technique. Although we frequently see the appendix on noncontrast CT's of the kidneys, ureters, and bladder, the use of a dedicated appendix protocol with both oral contrast and a contrast enema will be helpful in optimizing our technique. If necessary, the examination can also be performed without intravenous contrast. Spiral CT scans show extensive periappendiceal and cecal inflammatory changes, abscesses, and extensive free fluid, which some surgeons also feel are a relative contraindication to laparoscopic appendectomy. If these findings can be confirmed by imaging preoperatively, then the patient can proceed directly to open appendectomy. If these findings are discovered at the time of the initial laparoscopic appendectomy, then the surgeon may be obligated to convert to an open appendectomy, which results in a significantly longer procedure time. (Click to see image) Imaging is no replacement for a clinical judgement, especially in today's era of cost containment. However, in the diagnosis of acute appendicitis, currently available technology has the potential to spare many patients an unnecessary invasive procedure and may often establish other alternative diagnoses which may be non-surgical. The use of preoperative scanning may also result in a significant decrease in the cost of treating patients with suspected appendicitis, by deferring the cost of unnecessary surgery in 20% of patients, as well as the cost of delayed diagnosis in another 20%. Pelvic Inflammatory Disease: PID will usually be diagnosed via vaginal exam prior to imaging. PID with TOA's is readily diagnosable via sonography. Diverticulitis: For the evaluation of diverticulitis, CT is the modality of choice. CT of the abdomen and pelvis is performed with oral contrast. Diverticulitis is identified as a focal area of bowel wall thickening with luminal narrowing. Inflammation into the peri-colonic fat is sometimes the only radiographic sign. Abscesses are not always present. Microscopic perforations result in diverticulitis, but often, a drainable abscess is not present. If an abscess is present, it can be drained surgically or percutaneously via the abdominal wall, per-rectal, or per-vaginal. In the past, we performed gastrograffin enemas. However, this exam is not as accurate, since we can only see luminal narrowing and very infrequently intra-peritoneal spillage of the gastrograffin. Left Abdominal Pain If the patient's pain is in the left hemi-abdomen, then the diagnoses should include urinary tract disease, perforated viscus, diverticulitis, and pelvic inflammatory disease. In conclusion, we have attempted to direct the imaging evaluation of a febrile patient with abdominal pain to the most appropriate imaging modalities. | ||
© Hawaii Radiologic Associates, Ltd. All rights reserved. |