Results of a prepilot study of potential test material for the external quality assessment of reticulocyte haemoglobin content

Sir, The increasing sophistication of automated red cell analysis has led to the routine availability of new measures shown to have clinical utility in the assessment of iron status. Of these, reticulocyte haemoglobin content (RHC) has shown considerable promise [1, 2], as it is at the reticulocyte stage of development that clinically important fluctuations in cellular measures of functional iron status are first seen. RHC has recently been recommended as a measure of choice in the diagnosis and management of iron deficiency in patients with chronic renal failure [3] creating a need for external assessment of laboratory performance. We report here the results of a prepilot study to assess the suitability of material for potential use in such a scheme with instruments available in the UK capable of providing this measure. Four samples were produced by the UK National External Quality Assessment Scheme for Haematology (UK NEQAS Haematology). Two (RH2 and RH4) were derived from a single donation from a healthy donor taken into citrate-phosphate-dextrose anticoagulant and supplied by National Health Service Blood and Transplant. The other two (RH1 and RH3) were obtained with informed consent from patients with iron-deficient primary polycythaemia treated by venesection and taken into acid-citrate-dextrose anticoagulant. All donations were processed according to the protocol used for the production of samples in the UK NEQAS Haematology full blood count scheme. Five aliquots of each sample, labelled ‘Day 1’–‘Day 5’ and each containing sufficient material for 10 analyses, were issued to six volunteer laboratories together with an instruction sheet giving guidance on sample analysis. In order for analysis to take place over five consecutive working days, donations were collected 6 days beforehand and processed and despatched the following day in order to arrive at the volunteer laboratories before the weekend prior to the commencement of the trial. The volunteer laboratories comprised two users each of Abbott (Maidenhead, UK), Siemens (Frimley, UK) and Sysmex instruments (Milton Keynes, UK). Mean (SD) RHC and reticulocyte per cent values were obtained for each sample on each instrument over five consecutive days. One Sysmex user requested additional samples and analysed them on two instruments. Bartlett’s test was used to investigate differences in betweenday precision and one-way analysis of variance to investigate differences in between-day mean values for samples analysed on each instrument. To allow for multiple comparisons, only P values <0.001 were considered statistically significant. Findings for RHC (pg; reported as MCHr with Abbott, CHr with Siemens and Ret-He with Sysmex instruments) are shown in Table 1. With Abbott instruments, statistically significant differences in between-day precision were found for five of the eight sets of data. Between-day mean values differed significantly in all eight data sets, with daily mean values differing by ≥2 pg in six. There were fewer statistically significant differences with Siemens and Sysmex instruments. Reticulocyte per cent values showed acceptable between-day precision for all samples and for all instruments, with no statistically significant findings. There were a few statistically significant differences for between-day mean values with both Abbott and Sysmex instruments, although differences were small: with Abbott instruments, the greatest difference between highest and lowest mean values over the 5 days was 0.28% (sample RH4 and Abbott instrument two), whilst with Sysmex instruments the comparable difference was 0.21% (sample RH3 and Sysmex instrument one). Daily mean values across all four samples and both makes of instrument ranged from 0.57 to 1.96%. Similar results were obtained with Siemens instruments on samples from a single donation (RH2 and RH4), whilst with the iron-deficient samples (RH1 and RH3), mean values were higher and increased significantly over the trial period for both instruments, from 3.92 to 5.67% and 4.33 to 6.37% with RH1 and 2.58 to 3.40% and 3.06 to 3.94% with RH2.