We typically test plasma samples for the animal science department.  I did find a paper specifically about this topic:

TY - JOUR

ID - 27

A1 - Chemnitzer, R.

T1 - ICP mass spectrometry. Direct determination of trace elements in bodily fluids.

Y1 - 2015

VL - 59

IS - 5

SP - 33

EP - 35

KW - ICP mass spectrometry trace element body fluid

AB - The expts. have successfully shown that the ICP-MS Plasma Avatar MS is a simple and effective soln. for direct detn. of trace elements in complex samples, such as plasma and whole blood. [on SciFinder(R)]

N1 - M1: Copyright (C) 2017 American Chemical Society (ACS). All Rights Reserved.; CAPLUS AN 2015:1190976(Journal)

PB - GIT Verlag

JF - GIT Labor-Fachzeitschrift

JA - GIT Labor-Fachz.

M1 - Journal Article

ER -

I believe it's in German but I hope it will be useful anyways.

Best wishes.

You don't state which elements you try to measure. Look at the linked manual from Mayo Labs for some possible alternatives.

http://www.mayomedicallaboratories.com/it-mmfiles/Trace_Metals_Analysis_Specimen_Collection_and_Transport.pdf

Thank you for all of you for valuable inputs.

I would like to further clarify the following:

What fraction of the blood is suitable for the entire panel of elements to be analyzed through ICP MS (which includes heavy & trace metals). I have seen and read several articles which suggest the use of whole blood for certain metals and serum for others. However I am looking for only one matrix which can be used for the entire analysis to avoid variation.

If you are only interested in variation you should use serum/plasma. Then the hematocrit or the distribution on metal ions between erythrocytes and plasma don't influence result variation. If you are concerned about the real meaning of your results you should use the most appropriate sample matrix for each analyte. See link for some examples.

http://www.medsci.uu.se/digitalAssets/158/158840_1schultze_helblod-vs-serum-metaller.pdf