Sergio Ginaldi, MD
Director of Ultrasound and CT, Tallahassee Memorial Hospital, FL and Staff Radiologist, Section of Abdominal and Interventional Radiology, Radiology Associates of Tallahassee, Tallahassee, FL

The advanced imaging capabilities of new, fast CT scanners have prompted radiologists to consider adjusting contrast administration protocols, specifically to use contrast media with a higher concentration of iodine and to infuse it at a faster rate. To evaluate the safety of such an approach, my colleagues and I reviewed the incidence, type, severity and predictability of adverse reactions to iodine contrast material in our radiology practice. We also investigated the effectiveness in high-risk patients of premedication for the prevention of allergic reactions and prophylactic hydration for the prevention of nephrotoxicity.

Our experience is that of a community practice in Tallahassee, FL. The 20 radiologists in our group cover 2 major hospitals, where we conduct approximately 26,000 CT scans a year. We also operate a number of satellite locations, including Tallahassee Diagnostic Imaging Center (TDI), where we perform 6,000 to 7,000 CT scans a year. The data and experiences discussed in this article reflect work done at the larger, regional hospital, Tallahassee Memorial Hospital, and at TDI, the largest diagnostic center.

Study Methods

We retrospectively reviewed the medical records of all 30,301 patients who underwent pediatric and adult contrast-enhanced CT examinations between August 1999 and October 2001. Of those, 78% were imaged using Isovue 370 (Bracco Diagnostics, Princeton, NJ). Another 8% were scanned using Ultravist (Berlex Laboratories, Wayne, NJ) with an iodine concentration of 370 mg/mL, and 14% were scanned using Ultravist with an iodine concentration of 300 mg/mL.

The mean age of patients in the study was 58 years, and 63% were female. The vast majority of patients, 21,459 (71%), were scanned in the hospital. In both the hospital and the outpatient imaging center, non-neurological studies were most common, outnumbering neurological examinations by a ratio of almost 5-to-1 in the hospital and >2-to-1 at TDI (Table 1). Non-neurological CT studies include scans of the neck, chest, abdomen, pelvis, and limbs in various combinations. As might be expected in a private practice, the most common combination was a dual scan of the abdomen and pelvis for abdominal pain. The next most common referring diagnoses were appendicitis and diverticulitis. Studies were tailored to the right lower quadrant in cases of suspected appendicitis and consisted of a combined abdominal/pelvic scan in cases of suspected diverticulitis. Chest CT for possible pulmonary emboli was next in order of frequency.

We prepared patients for contrast-enhanced CT by first obtaining a clinical history (Table 2). We focused on whether they had previously experienced an adverse reaction to iodine or been diagnosed with allergies, asthma, cardiac disease, renal disease, multiple myeloma, or sickle cell anemia. We also determined whether the patient had undergone more than one contrast-enhanced procedure within the previous week, and measured baseline serum creatinine levels as needed.

Any patient with a history of allergies to iodinated contrast media or steroid-treated asthma underwent CT scanning at the hospital, rather than the outpatient imaging center. Patients who required premedication were also referred to the hospital for scanning. Premedication consisted of antihistamines, corticosteroids, and occasionally, a sedative administered just before the contrast injection.

Patients whose serum creatinine levels were between 1.5 and 2.0 mg/dL, as well as those with multiple myeloma or recent sickle cell crisis, underwent hydration therapy. This consisted of 100 mL of 0.45% normal saline infused at 100 mL/hr before contrast administration and, if tolerated, 200 mL infused at the same rate after the procedure.

We obtained peripheral venous access by inserting a 20-gauge angiocatheter into the antecubital vein. Most patients were studied in the supine position and, therefore, could be imaged using a breathhold technique. The selection of scanning technique, including slice width and pitch, depended on the size of the region of clinical interest, the size and age of the patient, and the type of CT scanner used. Two of our scanners were equipped with single-detector helical technology and 2 with multidetector technology.

The technique we use to administer contrast is shown in Table 3. It varies with the iodine concentration of the contrast media, as well as the type of scan. For the most common examination, a combination of abdominal and pelvic scans, we use contrast media with an iodine concentration of 370 mg/mL.

We use only nonionic contrast media and attempt to record all adverse reactions that occur within 24 hours of contrast administration. We observe patients for at least 48 hours for evidence of worsening renal function, as gauged by an increase in the serum creatinine level of 0.5 mg/dL or more above baseline. In some cases we also check creatinine clearance and electrolyte levels.

Study Results

Nonfatal Reactions

Nonfatal allergic-type reactions were equally common among patients who received contrast media with an iodine concentration of 370 mg/mL as among those who received contrast media with an iodine concentration of 300 mg/mL (Table 4). In the outpatient clinic, the rates of allergic-type reactions were 1.4% and 1.3%, respectively, for a total of 120 patients. In the hospital, where more than 21,000 patients were scanned using only higher-concentration contrast media, the rate of allergic-type reactions was even lower, <1%.

Only 0.3% of the patients scanned at the outpatient center required medication for an allergic-type reaction, and of the 29 who did, only 3 were injected with epinephrine. There was no appreciable difference in the rates of allergic-type reaction on the basis of iodine concentration.

Of the more than 21,000 patients who were scanned in the hospital, 95% had no clinical risk factors in their history to prompt premedication. In this group, the incidence of allergic-type reactions was very low, approximately 0.7%. In the 1,175 patients who were premedicated, the incidence of allergic-type reactions was substantially higher, approximately 4%. In other words, the chance of an allergic-type reaction was more than 5.5 times as high in our premedicated patients (those with risk factors) than in those without risk factors for an allergic-type reaction (Table 5). The most common allergic-type reactions were urticaria, itching, sneezing, transient warmth, slight headache, flushing, and nausea.

Of the 191 patients who had allergic-type reactions in the hospital, 48, or 25%, were premedicated. Only 0.6% of patients who were scanned at the hospital required medication for an allergic-type reaction. Of the 137 who did, only 9 were injected with epinephrine.

Fatal Allergic Reactions

Of the 4,249 patients who were scanned using contrast media with an iodine concentration of 300 mg/mL, 1 patient died. An additional 2 patients died of the more than 26,052 who were scanned with 1 of 2 contrast agents with an iodine concentration of 370 mg/mL. There was no statistically significant difference between the two groups in the incidence of death. None of the 3 patients was considered to be at clinical risk or was premedicated.

Nephrotoxicity

We were able to collect data on renal function in a total of 8,543 patients. Of these, 8,122 were scanned in the hospital, and 421 were admitted from our outpatient facility. A total of 258 patients were identified as being at high risk for contrast nephrotoxicity before scanning, and underwent hydration. Nonetheless, we observed a high incidence of contrast nephrotoxicity, in 87 (34%) of 258 high-risk patients (Figure 1). Five patients required dialysis, 1 of whom continues on chronic dialysis. This level of contrast-associated nephrotoxicity is lower than is usually reported in the literature. ^1,2 It may have been reduced by following strict pre-scan protocols, including determination of baseline serum creatinine and the prophylactic use of hydration.

We identified an additional 12 patients with nephrotoxicity in the remaining group, none of whom were considered to be at high risk and, therefore, did not receive hydration before contrast administration. None required dialysis. Overall, about 1% of patients monitored for renal function developed nephrotoxicity, and only 0.06% of the monitored patient population required dialysis.

Conclusion

There was little difference in the incidence of allergic-type reactions among patients who received high-concentration contrast agent (370 mgI/mL) and those who received a lower concentration (300 mgI/mL); the rates for both were approximately 1%.

The majority of adverse reactions were mild. Reaction severity also does not appear to depend on the concentration of iodine in the contrast media.

Among patients with established risk factors, allergic-type reactions were 4 times more frequent than among patients without risk factors, despite premedication. Posttreatment medication for allergic-type reactions, in our experience, was needed in <1% of patients.

Contrast nephrotoxicity occurred in about 1% of patients without predisposing risk factors. In patients known to be at high risk, the incidence was >30%, despite hydration.

It is important to remember that, even if the incidence of adverse reactions to nonionic contrast agents is very low, reactions can happen at any time with any product. Physicians who administer or authorize the use of intravenous contrast media need to know how to monitor and treat adverse reactions. Patients at risk for a reaction or nephrotoxicity should be well hydrated and premedicated when possible. Guidelines for the diagnosis and treatment of adverse reactions to radiocontrast media are available from the American College of Radiology. ³ *