Assessment of acute iron poisoning by laboratory and clinical observations

A survey of medical records at a tertiary children's medical center was performed to determine whether selected laboratory and clinical observations and the serum iron concentration exceeding the total iron-binding capacity serve as indicators of acute iron poisoning. Patients were hospitalized for treatment of iron poisoning during January 1976 through June 1992. A total of 128 patients were identified; 92 met selection criteria and 65 had a serum iron concentration and total iron-binding capacity obtained simultaneously. The present study was unable to confirm that vomiting, diarrhea, leukocytosis, hyperglycemia, and radiopacities were associated with a serum iron concentration in excess of 300 micrograms/dL (54 mumol/L); only one observation, coma, was associated (P < .02) with a serum iron greater than 500 micrograms/dL (90 mumol/L). Coma, radiopacities, leukocytosis, and an elevated anion gap will be concurrently present (predictive value positive = 100%) when the serum iron concentration is greater than 500 micrograms/dL; moreover, they will be absent collectively (predictive value negative = 95%) when the serum iron concentration is below 500 micrograms/dL. Individually, these signs and symptoms have a low positive predictive value, but the absence of any one of these variables is likely to be associated with serum iron concentrations less than 500 micrograms/dL (predictive value negative > or = 93%). The ratio of serum iron concentration to the total iron-binding capacity was not associated with symptoms of systemic iron toxicity or presence of vin rosé urine after parenteral deferoxamine administration. A serum iron concentration in excess of the total iron-binding capacity may not identify patients with serious iron poisoning.