Laboratory responsibility includes all pre-analytical aspects of testing. Recently there has been a focus on test utilization – ordering the right test, on the right patient, at the right time. But the next step is also critical – collecting the specimen appropriately.  If the patient has a line available, it is common to draw a blood specimen from the line. This is even more common in children where, in order to avoid the trauma of repeated needle sticks, a line may be placed specifically for specimen collection. An informal survey through Children’s Hospital Association indicated that heparin levels were commonly drawn from the same line used for heparin administration. In addition, specimens from children tend to be a smaller volume then from adults. A given volume of IV fluid will have a larger effect of a smaller specimen. Tested immediately, a 1 ml blood specimen (Hct 40) contaminated with 1 drop (0.075 ml) of D5 (glucose 5000 mg/dl) would have a glucose of 644 mg/dL.

At Children’s Mercy Hospital, the critical value for glucose is 400 mg/dL. The laboratory recently established guidelines for suspecting IV fluid contamination in specimens with a glucose value > 300 mg/dL with a CO2 > 12 mol/L. The inclusion of a CO2 value in the algorithm avoided delaying results for patients in ketoacidosis. When contamination was suspected, the laboratory technologists called the nurse to determine how the specimen was drawn and to request a redraw, if appropriate. After implementing this procedure, approximately 1% of our specimens were determined to be contaminated. This is a minimum estimate based on one analyte, and many specimens with glucose values < 300 mg/dL may also be contaminated, but remain unidentified.

This is an important safety issue. Hypoglycemia in a newborn, which is a common problem, may be missed based on contamination of a specimen with glucose. Dilution with IV fluid of a specimen drawn for a CBC could lead to an unnecessary transfusion. Drug dosing may be inappropriately altered based on incorrect results secondary to contamination.

At my institution, the laboratory engaged with nursing in a joint Quality Improvement project and investigated this problem. Factors that contributed to contaminated specimens included: (1) Using the pause button to stop the IV fluid flow instead of stopping the infusion by clamping or using a stopcock, (2) Not clamping all of the lumens of a catheter, (3) Ineffective slide clamps, (4) Not waiting the appropriate time between stopping IV fluid and obtaining specimen and (5) High negative pressure draws which may pull fluid even from a non-flowing line.

Starting in the high risk intensive care units, the following steps were taken (1) Revise the procedures for drawing specimens from lines , as appropriate, (2) Bedside observation and re-education, (3) Development of reminder posters kept with required equipment for drawing specimens, (4) Development of reminder cards that attach to the nurse’s ID badge.

At this time, there has been a > 50% reduction in contamination of specimens from the intensive care units identified using the glucose algorithm as discussed above. Hospital-wide test cancellation due to possible specimen contamination has decreased by 42% for all analytes. The next step in the process is to expand this program to all care units in the hospital and to implement a system of personal responsibility of the nursing staff for each specimen.


  1. Allen TL, Wolfson AB. 2003. Spurious Values of Serum Electrolytes due to Admixture of Intravenous Infusion Fluids, J Emergency Med. 24: 309-313.