PCV vs. HCT, TS vs. TP

Measuring manual packed cell volume (PCV) and refractometric plasma total solids (TS) has long been part of the emergency database; moreover these parameters are often reassessed during hospitalisation.

With that in mind, I was interested to read the following recently published article and wanted to share some of their findings and other information/discussion points from the paper:

Tamborini A, Papakonstantinou S, Brown A, et al. Comparison of manual and laboratory PCV and total protein using EDTA and lithium heparin canine samples. J Sm Anim Prac 2014. 55(5):258-264.
DOI: 10.1111/jsap.12198


Manual PCV is measured using a capillary tube and microhaematocrit reader, i.e. is a direct measurement, whereas haematocrit as reported by laboratory analysers is calculated from red blood cell count and mean cell volume (MCV); errors can occur for example if platelets and red cells are misidentified or when osmolality affects MCV.

But how do manual PCV measurements compare with laboratory HCT values?

In this study:

  • 176 EDTA samples analysed
  • Median PCV 41% (interquartile range 36-46%)
  • Median HCT 42% (36-47%)
  • Excellent correlation (r = 0.99)
  • No statistically significant difference in values

TS vs. TP

Refractometers measure the refractive index of a liquid that is the result of its total solids/solutes (TS) constituents. As far as plasma is concerned, proteins make up a significant proportion of the solutes but there is a variable contribution from non-protein solutes. Some of the more simple handheld refractometers purely measure solutes and the values obtained should be referred to as concentration of total solids/solutes. However some refractomers are designed to use a conversion factor for converting the refractive index to protein concentration to more closely provide a total protein reading.

But how do the values obtained from both types of refractometer compare to laboratory total protein?

In this study:

Plasma from 238 EDTA samples analysed by refractomer for total solids (TS)
133 blood samples in serum tubes analysed by laboratory analyser for total protein (TP)
105 blood samples in lithium heparin (LiH) tubes analysed by laboratory analyser for total protein (TP)

Serum samples showed good correlation (r = 0.86) between TS and TP:

  • Mean EDTA plasma TS 69.4 g/L (SD ± 11.1 g/L)[using one refractometer with protein conversion factor]
  • Mean serum TP 59.8 g/L (SD ± 9.6 g/L)
  • BUT statistically significant difference in values; EDTA plasma TS overestimated serum TP with a mean bias of 9.6 ± 5.7 g/L

Lithium heparin samples showed fair correlation (r = 0.74) between TS and TP:

  • Median EDTA plasma TS 79 g/L (71-90 g/L)[using the other refractometer without protein conversion factor]
  • Median plasma TP 63.8 g/L (59.7-68.2 g/L)
  • BUT statistically significant difference in values; EDTA plasma TS overestimated LiH TP with a mean bias of 16.3 ± 10 g/L

Two hand-held screw-type refractometers were used. These were temperature-compensated as is recommended and calibration with distilled water was performed daily. However the devices were different in so far as one included a conversion factor as described above but the other did not; as the authors acknowledge, each sample was not run on both devices as shown above and their magnitude of overestimation of plasma TS versus TP differed (the one without the protein conversion factor reading higher).

Further analysis of substances such as urea, glucose, bilirubin and triglycerides failed to identify reasons for why EDTA plasma TS by refractomer overestimates TP by laboratory chemistry analyser. Only cholesterol was shown to have some statistically significant effect but its true significance in clinical terms remains equivocal.

One other aim of this study that I will not detail much was to evaluate whether lithium heparin and EDTA samples can be used interchangeably to obtain manual PCV and plasma TS readings in canine blood.

Their results:

  • PCV and TS was measured in 43 corresponding LiH and EDTA samples
  • There was excellent correlation (r = 0.97) but a statistically significant difference between the PCV values obtained…
  • ….compared to EDTA, the PCV determined from LiH samples was overestimated with a mean bias of 1.29 ± 2.06%....
  • …but as the authors recognise in practical clinical terms this is insignificant despite the statistical significance!
  • There was excellent correlation (r = 0.97) and no statistically significant difference between the TS values obtained.

Some noteworthy comments about this study, some of which are acknowledged by the authors in the paper:

The fact that the study was looking at something with clinical significance is clearly a strength.
Sample sizes are reasonable for a veterinary study
Authors acknowledge that very few of the samples had values that were significantly outside of reference intervals – and this is potentially a problem with applying these findings to ECC patients who often have PCV/TS values outside of reference intervals.
Samples were placed into EDTA containers before in-house analysis; my personal clinical experience is that mostly these samples are run on non-anticoagulated blood samples as point-of-care tests – but not always.
It was not clear to me who performed the PCV and TS measurements and if/how this was quality controlled. Personal experience suggests that there can be some variability when the sample measurement is read by different individuals and if a single blood sample from the same patient is analysed multiple times in a short space of time. There was perhaps an argument for at least having multiple operators assess each sample but this is not the reality on most clinic floors so this methodology would not mirror practical reality.
The plasma total solids (EDTA plasma TS) values reported in this study anecdotally seem higher than the sorts of values I am accustomed to seeing in ECC patients.
The in-house methodology with respect to centrifugation etc. seemed akin to what many/most clinics seem to do. It is of note that samples were centrifuged for 5 minutes. Capillary tubes were filled adequately but not excessively (approximately 2/3-3/4) as recommended to minimise erroneous readings. This methodology would ideally need to be adhered to when extrapolating their results to our own patients.
HCT was determined using flow cytometry methodology which according to the authors is a modern and more accurate technique. I am unaware how many external laboratories presently employ this methodology.
Laboratory analysers underwent twice daily quality control procedures and an external procedure every 14 days.

Clinical bottom line:

Overall I felt the methodology was suitably robust to be able to take their findings on board with some level of confidence, although why the EDTA plasma TS readings seemed higher than my personal anecdotal experience is unclear.This study suggests that:

  • Should the scenario arise, manual PCV and laboratory HCT can be used interchangeably for clinical decision-making in dogs [at least when a flow cytometry-based quality assured laboratory analyser is being used (added 23.5.14 following feedback)].
  • Plasma TS and laboratory TP should not be used interchangeably in dogs
  • Ideally clinicians should know whether their refractometer is one that includes a protein conversion factor
  • However we have to remember that many of the values in the study were within reference intervals and a similar study done with more abnormal values would be useful for ECC patients.

Lithium heparin and EDTA samples can be used interchangeably to obtain manual PCV and plasma TS readings in canine blood

Final comments:

I think in reality in clinical practice the scenario in which we could consider using these parameters interchangeably is one that we encounter only seldomly. PCV and TS are measured regularly (e.g. twice daily in some cases) whereas laboratory tests are, or at least should be (!), performed much less frequently. Nevertheless I found this paper interesting and it expanded my knowledge and understanding about a common clinical practice.

It was disappointing – though not through the lack of trying! – that an explanation for why refractometric plasma TS overestimates laboratory TP was not identified but it does allow me to say “I don’t know” with more confidence! And I would like to see some results for patients with abnormal PCV/TS values as a significant proportion of ECC patients fall into this category. And of course a similar study that includes cats would be great.

Please do leave any comments or questions below.