Using chi-square and a PC to assess competency - laboratory management techniques

Medical Laboratory Observer, July, 2001 by Scott Warner

As a check, consult Table 1 and compare the minimum and maximum values -- a rough span of probabilities for eleven values -- and the data in Table 4, above, should be obtained.

Demonstrating competency

Our small study was able to verify that a 100-cell differential allows for a wide range of values and that using the chi-square test helps to produce manual differential results that are statistically equal to an automated standard. Using a standardized table, as suggested in Table 4, does not take into account whole populations and degrees of freedom. The chi-square test is a simple and easy alternative that not only serves to assess the competency of current employees, but that can be used to orient new employees as well.

Note: The program 2Diffs, written in Visual BASIC for the PC, can be downloaded from the author's website at http://members.panax.com/sass. To download the program, click on the Windows logo. A page will appear where programs are listed alphabetically in a table. Find 2diff zip in the left hand column and click on the link. A program to unzip the file, such as Winzip, is necessary to open the download. A file called vbrun300.dll. is also required to run the Visual BASIC program. Instructions for these items are located at the bottom of the page.

Mr. Water is laboratory manager at Mount Desert Island Hospital, Bar Harboor, ME.

References

(1.) Technical consultant responsibilities. Code of Federal Regulations. Title 42, Volume 3, Sec. 493.1413 Standard. October 1, 1999; US Government Printing Office via GPO Access [CITE: 42CFR493]: p.906.

(2.) Bernett R. Clinical Laboratory Statistics, 2nd ed. Boston, MA; Little, Brown end Company: 1979, p. 28.

Table 1
95% confidence limits for a percentage of
cells with a particular characteristic
n=100 cells counted
Observed             Limits
   0                  0-4
   1                  0-6
   2                  0-8
   3                  0-9
   4                  1-10
   5                  1-12
   6                  2-13
   7                  2-14
   8                  3-16
   9                  4-17
   10                 4-18
   15                 8-24
   20                12-30
   25                16-35
   30                21-40
   35                25-46
   40                30-51
   45                35-56
   50                39-61
For X over 50, obtain confidence limits by reading limits for 100
minus X in the table and subtracting them from 100. For example, the
confidence limits for 75% are 65 and 84.
From Rumke CL. Variability of results in differential counts on blood
smears. Triangle, The Sandoz Journal of Medical Science. 1960;4:156.
Table 2
Singificant chi-square for varying df, 95% level
df              1     2     3     4      5      6      7      8
[[chi].sup.2]  3.84  5.99  7.81  9.49  11.07  12.59  14.07  15.51
df                9     10
[[chi].sup.2]  16.92  18.31
Results of comparison of 100-cell manual
differential with Sysmex SF3000
Cell type      Value   1     2     3     4     5     6     7     8
Diff 21S
Neutrophil     79.8    83    80    88    86    87    86    86    85
Lymphocyte      8.2    4     6     5     7     8     5     3     5
Monocyte        8.6    10    12    5     4     4     8     9     9
Eosinophil      2.9    3     2     2     3     1     1     2     1
Basophil        0.5    0     0     0     0     0     0     0     0
[[chi].sup.2]         2.12  1.57  2.79  2.51  3.42  2.15  3.32  2.37
Diff 27R
Neutrophil      56     53    57    51    66    57    49    56    63
Lymphocyte     35.4    41    34    40    32    31    41    37    30
Monocyte        6.1    3     8     7     2     10    9     6     6
Eosinophil      2.2    3     1     2     0     2     1     1     1
Basophil        0.3    0     0     0     0     0     0     0     0
[[chi].sup.2]         1.97  1.04  0.89  5.57  1.55  2.18  0.79  1.61
Cell type       9     10    11
Diff 21S
Neutrophil      82    78    85
Lymphocyte      7     9     2
Monocyte        11    8     10
Eosinophil      0     5     3
Basophil        0     0     0
[[chi].sup.2]  3.82  1.14  4.54
Diff 27R
Neutrophil      50    59    58
Lymphocyte      43    35    37
Monocyte        5     4     4
Eosinophil      2     2     1
Basophil        0     0     0
[[chi].sup.2]  1.49  0.83  1.26
Source of data: Technologist competency assessment comparing 100-cell
manual differentials to automated Sysmex SF3000 differential. Mount
Desert Island Hospital, Bar Harbor, ME; October--December 2000.
Table 4
Observed limits of data compared to theoretical limits from Table 1
Cell type    Value  Observed  Table 1 limits
Diff 21S
Neutrophils  79.8    78-87        70-88
Lymphocytes   8.2     2-9          3-16
Monocytes     8.6     4-12         4-17
Eosinophils   2.9     0-5          0-9
Basophils     0.5      0           0-4
Diff 27R
Neutrophils   56     49-66        44-65
Lymphocytes  35.4    31-43        25-46
Monocytes     6.1     2-10         2-13
Eosinophils   2.2     0-3          0-8
Basophils     0.3      0           0-4