Flow cytometric cell counts: not perfect? (How getting a calculator out can suggest an anomalous result)
Background/introduction:  
HIV patients have their CD4 counts measured regularly using flow cytometry, ‘single platform’ measurement being the current standard.  Recently, two HIV patients attending for routine follow up blood tests showed unexpectedly low CD4 counts compared to previous results.  Using the patients’ total lymphocyte counts (obtained from contemporaneous testing on a haematology analyser) and CD4 % (from the flow cytometry report), we calculated a more expected result.   This prompted a review of our CD4 counts comparing single and dual platform results.
Aim(s)/objectives:
To identify any anomalous results when comparing flow to calculated CD4s.

Methods:
Fifty-nine CD4 counts from 38 HIV patients (27 males and 11 females) attending clinic for routine bloods from 18/01/2015 to 09/02/2016 were reviewed.

Results:
 

The table shows the comparison between the dual platform CD4 and the single platform CD4. The two patients that triggered the query are in green (male) and red (female).  The sequential before/after CD4 counts for the male patient (pale green) and the female patient (pale pink) are also highlighted on the table.
Discussion/conclusion:
Reassuringly, statistical analysis showed very close correlation between the two methods, apart from the two odd results. Previous and subsequent counts in these two patients were normal, as expected.  Twenty years ago only the % was available so absolute number was calculated using the simple method: lymphocyte count × CD4%.  As usual in medicine, no methodology is perfect. Unexpected results should be questioned and, if necessary, repeated, especially if important therapeutic decisions depend on them.