Abstract
Diagnosis of many life-threatening illnesses, including acute pulmonary embolism, aortic dissection, and ischaemic bowel disease, requires confirmatory radiological imaging with radiocontrast. It is well established that radiocontrast can induce acute kidney injury, especially in patients with pre-existing renal impairment. The decision to proceed with a radiological study with radiocontrast to confirm or exclude a life-threatening, but potentially reversible, illness in patients with renal impairment is difficult. Theoretically, a radiocontrast study will be justifiable provided its benefits outweigh its harms. Using published prognostic data of contrast-induced nephropathy (CIN), this decision analysis aimed to assess whether a certain threshold of pre-test probability of a life-threatening illness is needed before a radiocontrast study can be justified for patients with different levels of renal impairment. In critically ill patients presenting with a life-threatening illness with hypotension requiring vasopressors or inotropes, the risk of CIN (defined by an increment in plasma creatinine of 40 μmol/l) and the associated attributable mortality after using 50 to 100 ml of radiocontrast was about 30% and 4%, respectively, for patients with baseline plasma creatinine concentrations <400 μmol/l. The risk of CIN and its associated attributable mortality increased substantially and exceeded 80% and 10%, respectively, if patients also had diabetes mellitus and their baseline plasma creatinine concentrations were >400 μmol/l. In the latter high-risk patients, using a radiocontrast study to diagnose or exclude a life-threatening illness could only be justified if the life-threatening illness was readily treatable and the pre-test probability of having such disease was greater than 15%-20%.