Many ghosts stalk the hearts and minds of emergency physicians, but few are encountered quite as frequently as the specter of contrast-induced nephropathy (CIN). It is far from uncommon to have the momentum of a resuscitation come to a definitive halt as when the critical ill patient is denied the CT scan he urgently requires because the serum creatinine is above a facility-mandated threshold, a limit that seems impervious to logic or reason, unbreakable by claims that the risks of the illness outweigh those of the mysterious injurious properties of iodinated contrast.
We have been given amulets and wards meant to protect us from its tubular mischief, and we ritualistically wave these charms above our patients' beds in the hope that they will prevent any ill will that may befall them.
Nijssen, et al., examined the efficacy of one of our most steadfast and dependable protective wards. (Lancet 2017; Feb 20. doi: 10.1016/S0140-673630057-0.) These authors randomized 660 patients who were scheduled to undergo elective procedures requiring iodinated contrast administration to receive prophylactic fluid administration or nothing.
The patients were eligible if they met criteria to be at high risk for CIN (an estimated glomerular filtration rate [eGFR] between 45 and 59 mL per min/1.73 m. combined with diabetes or at least two predefined risk factors [age over 75; anemia; cardiovascular disease; nonsteroidal anti-inflammatory drug or diuretic nephrotoxic medication]; or eGFR between 30 and 45 mL per min/1.73 m.; or multiple myeloma or lymphoplasmacytic lymphoma with small chain proteinuria).
Patients who were randomized to receive intravenous hydration underwent a long or short protocol. These consisted of the administration of 0.9% NaCl 3-4 mL/kg/hr either four or 12 hours prior to procedure and four or 12 hours after contrast administration. Screening serum creatinine was performed prior to procedure two to five days following contrast administration and 26-35 days after receiving a contrast exposure.
The authors found no difference in their primary outcome, the rate of CIN (defined as proportion of patients with an increase in serum creatinine by more than 25% or 44 μmol/L within two to six days of contrast exposure) between the groups. A total of 2.7 percent of the patients experienced a CIN in the intravenous-hydration group vs 2.6 percent of the patients in the no-hydration group.
Because the authors utilized a noninferiority methodology, they were barely capable of demonstrating significance because the upper level of their confidence interval flirted with their pre-set noninferiority margin of 2.1 percent. This, of course, says nothing about the clinical inferiority of a no-hydration strategy, but once again speaks to how noninferiority trials are traditionally powered.
Now there are a number of reasons the data may not be directly translatable to our patients in the emergency department. For one, these were stable patients who were scheduled for an elective procedure that required iodinated contrast dye. This is represented in the extremely low rate of CIN, with only eight patients in each group meeting the authors' primary endpoint. Contrast this to the rate of CIN in a recently published article by Valette, et al., where the rate of CIN in critically ill patients admitted to the ICU who received normal saline prophylaxis was 33.3 percent. (Crit Care Med 2017; 2017;45:637.) Was the study by Nijssen, et al., too underpowered to detect a difference?
The study could not rule out a two percent benefit of IV hydration, but the point estimates were almost identical. It is unlikely that the addition of more patients would have produced a different result. These results and many other recent failed attempts to reduce the rate of CIN may be because we are attempting to treat an entity that does not exist.
It's a shadow left over from the days of high-osmolar contrast agents, enduring despite numerous trials questioning its continued clinical relevance. Two large retrospective cohorts, in fact, have failed to demonstrate any association between iodinated contrast administration and an increased risk of kidney injury. (Ann Emerg Med 2017; Jan 19. doi: 10.1016/j.annemergmed.2016.11.021; Radiology 2013;267:106.)
All ghost stories are based on some underlying truth. This may well be the case for CIN. The potential for kidney injury in patients given iodinated contrast exists, but this rate and clinical significance have been overstated at the cost of patient care. Most of the cases mistakenly identified in the current literature as being caused by CIN are likely due to the underlying illness that caused these patients to present to the ED in the first place.
The surrogate endpoints used to define CIN rarely lead to clinically important outcomes such as the need for renal replacement therapy or death. There are many reasons patients presenting to the ED should not undergo imaging, but in the subset of patients that requires contrast-enhanced radiography, the specter of CIN should not dissuade or delay us from obtaining the information we require.