DELTA CHECK

This is also a method for monitoring quality of results. The difference between a patient's present laboratory result and the previous result which exceeds a predefined limit is referred to as a delta check. Delta checks are investigated by the lab internally to rule out: 1) mislabeling, 2) clerical error or 3) possible analytical error. When patient’s clinical condition is stable and differences between repeated test results are small, the difference between successive results may be used as a form of quality assurance. 

The delta check concept is applied to two successive values regardless of the time interval between them. Delta check values are generated in one of two ways: the first is derived from the differences between the collected consecutive values for an analyte in many individuals which are then plotted in a histogram with the central 95% or 99% of all values used to identify a clinically significant change in values. Delta checks may involve the absolute difference or the percent change between the consecutive numbers. 

The second approach relies on a laboratorian’s or clinical best estimate of an appropriate delta to yield a manageable number of flagged results for follow up. A more refined means of using patient data for assessing statistically significant changes is through rate checks that involve dividing a delta check value by the time interval between successive measurements. Delta checks have been based on,,

a.       Delta difference: current result – previous result;

b.      Delta percent: (current-previous result)x100

c.       Delta difference: delta difference/delta time;

d.      Rate percent change: delta percent change/delta time (where delta time is the interval between the current and previous specimen collection times)

For a given patient, delta check method compares the differences (deltas) between today’s test values and corresponding previous test values with given thresholds. If a delta exceeds its threshold, the value for today fails the delta check and is suspected of being erroneous. Any source of laboratory error may cause one or more of a set of test values to fail a delta check. The method is particularly interesting as a method for detecting two important types of error, specimen mislabeling (i.e., assigning today’s values to a patient different from the one from whom the sample was actually taken) and error in reporting test results, because these types of errors are not detectable by any other posteriori means, such as checking control specimen against quality control limits.  

In healthy individuals and in stable patients the delta value between any two results should be small.