A simple program for hematology QC.
In our laboratory, it may take more than 24 hours to collect data from 20 blood specimens, so we can't rely on moving averages to detect sudden loss of calibration. On the other hand, data from our commercial hematology controls may take over a month to be processed and returned to us, so we really can't rely on these controls to detect drifts in calibration.
It occurred to us that a good solution to this problem might be to use Bull's system of moving averages to detect drifts while using commercial controls or duplicate analysis of patient specimens to detect sudden loss of calibration.
Our laboratory has a Commodore 64 microcomputer, which we programmed to do the calculations. We encountered some difficulties in setting up this program due to errors in the published flow charts that we used. We would like now to share a "bare bones" BASIC program that will run on most microcomputers.
It requires the input of red blood cell counts, hematocrits, hemoglobins, and mean corpuscular volumes. We include hemoglobin along with the calculated rBC indices (MCH and MCHC) in the hope that it will help detect drift in diluter calibration.
Users can easily expand the program with error-trapping routines. For example, we display all the entered values on the screen for review before letting the computer proceed to calculations.
Extremely low and high limits for each parameter can also be programmed, in order to flag suspect entries. These work well when an extra digit is mistakenly keyed in or a decimal point is put in the wrong place.
We hope this program proves ueful to those wishing to investigate the use of Bull's method of quality control without investing a lot of time or money.
|Printer friendly Cite/link Email Feedback|
|Publication:||Medical Laboratory Observer|
|Date:||May 1, 1984|
|Previous Article:||The microcomputer as management tool.|
|Next Article:||A time to reexamine values.|